Katholieke Universiteit Leuven

Transcription

1 Katholieke Universiteit Leuven FACULTY OF BIOSCIENCE ENGINEERING INTERUNIVERSITY PROGRAMME (IUPFOOD) MASTER OF SCIENCE IN FOOD TECHNOLOGY Major Food Science and Technology Academic year Influence of composition and processing on the heat stability of recombined evaporated milk by Pham Thi Tuyet Mai Promotor : Prof. Dr. Ir. Paul Van der Meeren Tutor : Marios Kasinos Department of Applied Analytical and Physical Chemistry Master's dissertation submitted in partial fulfilment of the requirements for the degree of Master of Science in Food Technology

2 The author and the promoter give the permission to use this thesis for consultation and to copy parts of it for personal use. Every other use is subject to the copyright laws; more specifically the source must be extensively specified when using results from this thesis. August 2012 Promoter Author Prof. Dr. Ir. Paul Van der Meeren Pham Thi Tuyet Mai

3 Acknowledgements First and foremost, I would like to express my gratefulness to my promoter, Prof. Dr. Ir. Paul Van der Meeren for his instructions and help concerning my thesis works. I especially thank him for carefully correcting and giving helpful comments several times that helped encourage me and improve my dissertation manuscript a lot. My heartfelt thanks go to my supervisor, Marios Kasinos, for his helpful discussions and laboratory guidance. From my first days doing thesis, he helped me to have a good start in my research. I acknowledge his discussion and ideas, guidance and patience with me. I thank him for going along with me in every step of my research. I would like to thank Prof. Dr. Ir. Bruno De Meulenaer, head of the Research group Food Chemistry and Human Nutrition of the Department of Food Safety and Food Quality, for allowing me to use chemicals, the Kjeldahl equipment for protein load determination and Ellman s reagent. I would like to acknowledge Geert Meesen, technician in the Department of Molecular Biotechnology, for his willing help in operating the Beckman centrifuge. The deep thanks go to all technicians, assistants and staff members of the Laboratory of Applied Analytical and Physical Chemistry of the Faculty of Bioscience Engineering in Gent University who assisted me a lot. Especially, I would like to show gratitude to Quenten Denon and Saskia Van der Looven for their friendly encouragements, warm helps and good advices during my living and studying here. My special thanks to Prof. Dr. Ir. Koen Dewettinck and Prof. Dr. Ir. Marc Hendrickx for organizing a very helpful InterUniversity programme (IUPFOOD) Master of Science in Food Technology for International students. Besides important courses with the good teaching and evaluating methods, the studying and living environments, that I followed in both Universities (UGent and K.U. Leuven), I had the chance to know the cultures from all over the World, which not only broaden my knowledge and friendship but also is a really good experience in my life. I especially thank them for their forgiving my mistake and giving me a chance to reunite with my family for the whole summer holidays. I also wish to thank Ir. Katleen Anthierens, Ruth Van den Driessche (UGent) and Katrien Verbist and Dr. Chantal Smout (K.U. Leuven) for their valuable support in administrative issues for the whole time I was in Belgium. I appreciate their willing help and encouragement in not only issues related to studying but also in living and personal problems. I could not have had the opportunity to study in Belgium if there had been no financial support. My appreciation goes to the Belgian government for being generous to students from developing countries, including those from Vietnam. I thank God for giving me the Master scholarship granted by Belgian Technical Cooperation (BTC). The fellowship gave me very wonderful years in Belgium. They supported us a lot including arranging us to stay in OBSG, where gather cultures for international students from all over the World. Especially, it is

4 really the house for many helpful and happy activities of the Vietnamese student family in Gent. I am also happy to have many good friends among the Vietnamese student communities. Thanks for their friendship and help during my studying and living in Belgium. The times of barbecue, party and cooking together are very useful for me to learn about social knowledge and behavior and overcome stress in the life living far from home. I also want to thank to Molly Gabaza, my closest classmate from Zimbabwe, for her friendship and sharing during my staying and studying in Belgium. The times that we studied together during exam preparation are unforgettable for me. My family is all the time my motivation. From the bottom of my heart, I would like to thank my parents-in-law, my mother and sisters who are support, encourage me, pray for me and wait for my success. I am greatly thankful to my parents-in-law who take care of my son, when both my husband and I go abroad for studying. I am lucky to have my husband studying PhD in Moscow Russia. He is my real helper, consultant, guiding me in both my social life and scientific knowledge. And especially, I am really proud of my little son, Tran Hoang Thai, for supporting my spirit. I would not have gone so far without their support and encouragement. I thank Belgium with kind and nice people, society and environment for all the things that I obtained, and that I met through two years. I wish them all a health-happiness and success. I will not forget the time studying in Belgium and wish to have the opportunity to come back for PhD studying in the future. Pham Thi Tuyet Mai Gent, August 3 rd

5 Table of contents Table of contents... V Abstract... VII Introduction... 1 Chapter 1 Literature review Composition Milk proteins Milk fat Lactose Homogenization Heat treatment Heat stability Effect of ph and temperature on heat-induced interaction Influence of homogenization on heat stability Effect of sulfhydryl- group reaction on the heat stability of milk Chapter 2 Materials and methods Materials Emulsion preparation Standard procedure Different casein:whey protein (C:W) ratios ph adjustment Different sugars Addition of Ellman s reagent Homogenization Heat treatment Viscosity measurement Particle size distribution V

6 2.7. Protein recovery Chapter 3 Results and discussion Effect of heating time and temperature on heat stability Effect of protein concentration on heat stability Change in particle size distribution Change in viscosity Effect of homogenization pressure on heat stability Effect of Casein: Whey (C:W) protein ratio on heat stability Without ph adjustment and without sugar addition With ph adjustment but without sugar addition With ph adjustment and sugar addition Effect of sugar on heat stability Heat treatment before or after homogenization Effect of Ellman s reagent addition on heat stability Chapter 4 Conclusions and future perspectives Conclusions Future perspectives References...62 VI

7 Abstract Milk has been one of the nutritional foods for human consumption, containing a high protein, fat and carbohydrate (lactose) content. Heat treatment has been applied from a long time ago in order to prevent microbial deterioration. However, heating milk has a range of effects on its functional and sensorial characteristics, which must be considered as well. Heat stability of milk has been the objective of many studies. In this thesis, the influence of several processing and composition factors was evaluated on the heat stability of recombined evaporated milk. The latter was mainly evaluated by measuring the viscosity and particle size distribution of the emulsions after heat treatment. More precisely, denaturation, coagulation and/or aggregation of whey proteins was investigated as a function of protein concentration, heat treatment, homogenization pressure, casein to whey protein ratio, ph, sugar content, the order between heat treatment and homogenization and the addition of Ellman s reagent. The results showed that the protein concentration had a direct proportion to the degree and rate of heat-induced denaturation of whey proteins. Both increasing the homogenization pressure (at a microfluidisation driving air pressure within the 1 to 4 bar range) and the addition of Ellman s reagent ( %) reduced the heat stability of milk. It was also observed that the onset of aggregation of emulsions containing 23 % SMP occurred at 100 o C, while in the presence of 12 % SMP it was only observed at around sterilization temperature (i.e. 121 o C). Heating duration also influenced the rate of coagulation of emulsions, depending on the thermal processing (heating temperature). Besides, the balance in the content of caseins and whey proteins could be optimized to improve the milk heat stability. Reduced ph values from neutral ph (7.1) to 6.6 resulted in the faster heat-induced denaturation and aggregation of emulsions containing predominantly caseins. This behavior was more enhanced in the presence of carbohydrates such as β-d-lactose. Heat treatment followed by homogenization improved the heat stability of recombined evaporated milk. Sucrose, galactose and lactose did not have impact on the heat-induced denaturation of proteins in the emulsions containing a mixture of casein micelles and whey protein isolate. VII

8 Introduction Milk is an important source of nutrients in human diets, providing energy, as well as highquality proteins, containing all of the essential amino acids that play important functions in the human body. Milk is also an important source of essential minerals, mainly phosphorous, potassium, magnesium and especially calcium. Fat- soluble vitamins such as A, D, E, and K are also present in milk. Milk is consumed in its natural form but is also widely used for the production of numerous products, such as cheese, yoghurt, butter and ice-cream. Due to its high nutrient content, heat processing is needed in order to prevent microbial deterioration. In dairy industry, heat treatment is by far the most commonly used technique in order to ensure the production of stable products with prolonged shelf life. Various thermal processes exist nowadays in dairy industry. However, heat treatment causes several changes in milk and its constituents, which must be considered. For that reason, several studies have been carried out in order to investigate the behavior of milk under different experimental conditions, in order to understand insightfully the effect of composition and processing on the heat stability of milk. In this Master dissertation, the heat stability of a recombined evaporated milk analogue was studied, consisting of 6.5 % oil as well as 16.5 % skim milk powder. Hereby, the effect of variations in heating time and temperature were considered. In addition, the influence of the overall protein content (by changing to skim milk powder content to either 12 % or 23 %), as well as of the casein to whey protein ratio was evaluated. Finally, the relevance of free sulfhydryl groups was studied by addition of Ellman s reagent. 1

9 Chapter 1: Literature review Chapter 1 Literature review 2

10 Chapter 1: Literature review 1.1. Composition Milk constitutes an important part in human s diet, mainly because of its high nutritional value (Creamer and MacGibbon, 1996). It is possible to be consumed as such or as raw material for the production of other dairy products (Euston, 2008; Raikos, 2010). From a molecular point of view, milk is a complex colloidal system. Milk is an emulsion of fat globules in an aqueous phase (O/W emulsion) consisting of dissolved and suspended components such as casein micelles, serum proteins, lactose, minerals and vitamins (Euston, 2008; Dewettinck, 2011; Raikos, 2010). The average composition of cow s milk is shown in Table 1. Table 1.1. The composition of cow s milk (Dewettinck, 2011) Component Content (%) Water 87.3 Fat 3.9 Proteins 3.3 Lactose 4.7 Minerals 0.8 According to the PhD summaries of Blaskó (2010) based on FAO data, the consumption of milk and milk products per capita was kg in 2009, in which the average level in developing countries, where financial, geographical or climatological problems are not able to meet consumers demands, was only 66.2 kg/capita/year, whereas in developed countries it was 245 kg/capita/year. Currently, only 40 percent of the world population consumes milk daily, in which a strong demand for milk and milk products is seen among the Asian countries. However, the prices have doubled as compared to the prices within the period of , and the consumption of these products significantly depends on the income. Thus, recombining dairy products is important and emerges to meet the World dairy products 3

11 Chapter 1: Literature review challenges. Recombined Evaporated Milk (REM) is a product which results from the recombining of skim milk constituents with milk fat and water and which contains not less than 20% of milk solids. Protein, fat and carbohydrates which constitute the most important substances in human diet, are also the main components present in milk. However, its composition can be influenced by a number of factors such as season, stage of lactation and breed Milk proteins Milk protein constitutes about 3.5 % (w/w) of milk and is one of the most important protein sources in human nutrition (Sawyer et al., 2002; Creamer and MacGibbon, 1996). The nitrogen content of milk is distributed among caseins, whey proteins, milk fat globule membrane proteins and non-protein nitrogen (NPN). Caseins, comprising % of the total milk protein, are defined as the phosphoproteins that precipitate from raw skim milk upon acidification to ph 4.6 at 20 o C. Caseins are divided into four main categories: α s1 - caseins, α s2 -caseins, β-caseins and κ-caseins. The lack of tertiary structure requires caseins to expose their hydrophobic residues to water contributing to a strong association with insoluble components. As caseins contain negatively charged groups such as phosphates, sulfhydryls and terminal carboxyls, they are able to bind different cations. Especially, binding with calcium is of great interest, because it involves the formation of micelles which affect milk stability (Dewettinck, 2011). In normal milk, approximately 95 % of casein exists as micelles, colloidal particles with a diameter of about 100 nm containing about casein molecules. In dry weight, the micelle comprises about 94 % of casein and 6 % of ions, mainly calcium and phosphate. The detailed structure of casein micelles is unclear but the hypothesis about the porous structure with spherical submicelles is of most interest. K-caseins having very hydrophilic C-terminal parts as flexible hairs are located on the surface, acting as the micelle-stabilizing factor (Dewettinck, 2011). Apart from caseins, the whey protein fraction, comprises approximately 20 % of the protein in bovine milk and is soluble in the aqueous phase. Whey proteins are characterized by a well-defined tertiary structure and are highly susceptible to elevated temperatures. Β lactoglobulin (β lg), α-lactabumin (α-la), bovine serum albumin (BSA) and immunoglobulin 4

12 Chapter 1: Literature review are the major representatives of whey proteins; β lg contributes more than 50 % of the total whey protein fraction (Rade-Kukic et al., 2010; Dewettinck, 2011). The monomer of β lg contains five cysteine residues, four of which are involved in intramolecular bridges. The remaining free thiol group is of great importance for changes occurring during heating, since it has long been recognized that the heat-induced aggregation of β lg involves both disulfide bond interchange and modification to the hydrophobic interactions that are intramolecular in the native whey protein and become intermolecular in the aggregates and gels (Havea, 2004) Milk fat Fat is an important compound of milk, whose content and fatty acid composition are unstable, depending on factors such as feed, individual and season. Milk contains about 3-5 % fat. The major group of milk fat, the triglycerides, with a variety of fatty acids, contribute to the nutritional, sensorial, physical and manufacturing properties of dairy products (Smet, 2010). Milk fat is present in milk in the form of fat droplets, each of which is surrounded by a thin membrane (8-10 nm) which is known as milk fat globule membrane (MFGM). The milk fat globule membrane contains approximately 50 % protein and accounts for about 1 % of the total protein of the milk (Dewettinck, 2011). In raw milk, the size of fat globules is in the range from 0.1 to 15 µm and hence they can be observed under a light microscope. The size of the fat globules can be modified by technological processes, such as homogenization Lactose The most prominent carbohydrate in milk is lactose. Lactose comprises about 52 % of milksolid-non-fat, about 70 % of whey solids and more than 90 % of the solids in milk ultrafiltrate. Chemically, lactose is a disaccharide consisting of one residue of D-glucose and D-galactose joined in a β-1,4-glycosidic linkage. Lactose is a reducing sugar causing nonenzymatic browning reactions with amino acids at high temperature heat treatment known as Maillard reaction. This property is used to evaluate the nutritional quality of milk. Lactose is not as sweet as other sugars such as sucrose, glucose and fructose (Dewettinck, 2011) and is a slowly digestible carbohydrate contributing to the sugar content balance in blood (Van Camp, 2012). 5

13 Chapter 1: Literature review 1.2. Homogenization Homogenization is a very important step during milk processing. Changes in fat globules and milk proteins during homogenization play a crucial role on the heat stability and the renneting properties of homogenized milk (van Boekel and Walstra, 1989). During homogenization, the fat surface area increases and this newly exposed fat surface is stabilized by the skim milk proteins. If homogenized milk is heated, serum proteins bind to the adsorbed micelles and cause rearrangement of the micelles themselves. When milk is heated before homogenization, serum proteins denature and associate with casein micelles. The following homogenization induces these complexes to adsorb on the newly created fat surfaces (Cano-Ruiz and Richter, 1997). Homogenization in dairy industry is used principally to prevent or delay the formation of a cream layer in full cream milk, by reducing the diameter of the butterfat globules. During homogenization of milk, the size of the fat globules is reduced and at least a 10-fold increase of the fat globule area is obtained. As a result the stability of the emulsion is improved (Dewettinck, 2011). Homogenization is usually accomplished by forcing the hot mix through a small orifice under suitable conditions of pressure and temperature. The actual mechanism of fat disruption within the homogenizer is thought to result from turbulence, cavitation, and velocity gradients (energy density) within the valve body (Smet, 2010). With single stage homogenization, fat globules tend to cluster as bare fat surfaces come together or adsorbed molecules are shared. Therefore, a second homogenizing valve is frequently placed immediately after the first, allowing more time for surface adsorption to occur (Pandolfe, 1982). Natural fat globules are coated with phospholipids to which other lipids and proteins are adsorbed. The reduction of fat globule size during homogenization makes the adsorbed amount of phospholipids limited. After homogenization, the newly formed fat globule is practically out of any membranous material due to its tremendous increase in surface area and readily adsorbs amphiphilic molecules, i.e. proteins and emulsifiers, from the immediate microenvironment to reduce the oil-water interfacial tension (Smet, 2010). 6

14 Chapter 1: Literature review Homogenization is usually performed by high pressure homogenizers (HPH) or microfluidizers. The latter operates by a different mechanism and at different pressures than conventional valve homogenizers. The operating efficiency of a laboratory scale microfluidizer has been compared to that of a HPH using either pasteurized whole milk or recombined milk. Microfluidization is thought to be a very effective method for reducing fat globule size and is slightly affected by changes in operating pressure. The back pressure module on the microfluidizer has a marginal effect on particle size increase. As a result, the homogenization effect is always greater than that of conventional homogenization. Only very small amounts of serum protein but high levels of casein are found on fat surfaces after microfluidization. The protein load is higher than predicted on the basis of decreases in globule size. Microfluidization has a little effect on the formation of fat clusters in milk. The higher protein load probably inhibits fat clustering (McCrae, 1994) Heat treatment Thermal processing of milk is an essential step in dairy industry. It aims to an increase of shelf life of the product and to ensure safety for human consumption, by diminishing microorganisms, inactivating enzymes and changing the chemical composition (Lehmann and Buckin, 2005). Therefore, it is important that milk and milk powders used in various food applications to be heat stable. Heat stability is defined as the ability of milk to withstand high temperatures without flocculation, gelation or protein separation (Singh, 2004). When milk is subjected to thermal processing and depending on the heating conditions, whey proteins may undergo structural changes, known as heat denaturation, accompanied by protein unfolding and exposure of hydrophobic groups. Thus the functional properties of whey proteins are changed, and this is important in food applications (de Wit, 1984). Generally, whey proteins are more suceptible to denaturation in skimmed milk than in whole milk, probably because fat hinders heat transfer and increases the viscosity of milk. The denaturation temperature among whey proteins constituents is different. More precisely, the most labile is α-la (62 o C), BSA follows at 64 o C, immunoglobulins at 72 o C, and β-lg has the highest denaturation temperature at 78 o C (Brown, 1988; Tran Le, 2010). However, the denaturation of α-la is reversible, while in β-lg it is irreversible. Β-lg/κ-casein interactions seem to be the main complex formed during heat treatment of milk due to the intermolecular disulphide bond (Raikos, 2010). 7

15 Chapter 1: Literature review Due to its high content in whey protein fraction, β-lg is considered a thermal marker in processed milk (Verheul and Roefs, 1998). Upon heating, β-lg undergoes conformational changes and partially unfolds and nonpolar together with the thiol group are exposed. Aggregation phenomena are irreversible and therefore the whole denaturation process becomes irreversible (Sawyer, 1986). Whey protein aggregation and formation of complexes between whey proteins, caseins and fat globules is a mechanism which depends on many factors, such as electrostatic and hydrophobic interactions, hydrogen bonding, disulfide crosslinking (Corredig and Dalgleish, 1999). Moreover, processing parameters such as ph, heating temperature and salt concentration also affect the kinetics of heat-induced aggregation of β-lg (Tran Le, 2010; Verheul and Roefs, 1998). The association between whey proteins and casein varies with heat treatment. For example, heating at 146 o C for 16 s causes less denaturation than heating at 90 o C for min (Harwalkar, 1989). Heat treatment affects the rheology and structure of milk emulsions and contributes to the consumer s acceptability. It has been proved that heating influences the particle size of emulsions stabilized with milk proteins. This increase in particle size distribution was attributed to fat globules aggregation, which resulted from interactions between non adsorbed protein molecules in the serum phase and proteins adsorbed at the interface of fat globules (Raikos, 2010). Caseins do not possess tertiary structure and they are stable towards heating. Mild heat treatment does not affect the micelles; however it does undergo changes, mostly hydrolytic, when subjected to severe heating temperature. Casein is completely dephosphorylated in 5h at 120 o C and about 50 % dephosphorylation occurs within the first 1h (Fox, 1989). According to Pesic et al.(2012), β-lg interacts with surface-bound κ-casein upon heating at 90 C (Figure 1.1). Hence, heat treatment induces significant changes in the distribution of denatured bovine whey proteins between soluble and micelle-bound complexes. For the sake of completeness, it should be mentioned that about 30 % of the total bovine whey proteins were in soluble complexes, whereas about 65 % of β-lg and about 40 % of α-la were found in micelle-bound complexes. 8

16 Chapter 1: Literature review Figure 1.1. A schematic representation of the interactions between casein micelles and denatured whey proteins in bovine milk after heat treatment at 90 o C for 10 min and at natural ph (6.71) (reproduced from Pesic et al., 2012). In order to clearly determine the quantitative distribution of the denatured whey protein, the observed degree of denaturation of α-la (left) and β-lg (right) is shown in Figure 1.2., based on the results of Vasbinder (2002). Both figures represent the decrease in native whey protein with the increase in heating temperature. At 90 o C heating, more than 95 % of β-lg is denatured, while this value in α-lac is nearly 80 %. This means that β-lg is more susceptible to heat denaturation than α-lac, which is in agreement with several studies (Dannenberg and Kessler, 1988; de Wit, 1984; Singh and Creamer, 1992). Figure 1.2. Percentage of denatured α-la (left) and β-lg (right) present in milk as soluble whey protein aggregates (white bars) and associated with the casein micelle (black bars) after 10 min heat treatment at temperatures ranging from o C. The grey bars represent native whey proteins (reproduced from Vasbinder, 2002). 9

17 Chapter 1: Literature review 1.4. Heat stability Heat stability has been a subject of research for about a century. During the First and Second World War, condensed milk was of the major dairy products due to its easy transport and a long shelf life; and hence solving heat coagulation of this kind of milk was focused (Singh, 2004). Between 1960s and 1980s, apart from condensed milk, factors affecting on heat stability of normal (unconcentrated) milk were investigated, such as ph dependence and milk composition, mainly β-lg, κ-casein and milk salt. However, the explanation of the mechanism of heat coagulation of milk was not completed in the molecular viewpoint (Singh, 2004). Numerous investigators gradually have enlarged the literature on heat stability of milk over several decades (Singh and Fox, 1985; Fox, 1989; Singh and Creamer, 1992; Singh, 2004). Heat stability is most widely defined as the ability of milk to withstand high temperature with no observation in coagulation and gelation and is calculated by the time required for a sample of milk in an oil bath at a definite temperature to initiate the coagulation indicated by flocculation, gelation or changes in the sedimentability of protein (Singh, 2004; Fox, 1989). Less commonly the temperature required to form coagulation in a constant time is measured. In order to evaluate the heat stability, the most frequent method is heating a small sample of milk in a narrow sealed glass tube in a thermostatically controlled oil bath at 140 o C for unconcentrated milk or at 120 o C for concentrated milks. The time is measured until the onset of coagulation of protein particles (Fox, 1989; Singh, 2004) Effect of ph and temperature on heat-induced interaction Heat-induced interaction between whey proteins and casein micelles and heat stability of milk are highly influenced by ph (Singh and Creamer, 1992). At ph values above 6.8, the whey protein complexes remain associated in the serum, whereas at lower ph values, these complexes remain associated with casein micelle surface. When heated at ph values above 6.8, both whey protein aggregates and micellar κ-casein dissociates in the serum (Smits and Van Brouwershaven, 1980; Singh and Fox, 1985). Heat coagulation time (HCT) depends on a number of experimental factors such as degree of tube filling, head space gas, rocking rate during heating and especially temperature and ph (Fox, 1989). Singh stated that the ph plays the most important role in the HCT of milk (Singh, 2004). Figure 1.3 shows that the maximum HCT occurred at ph values around 6.7, 10

18 Chapter 1: Literature review with a minimum at ph of 6.9 and further HCT increases with the increase of ph leading to the better stability of milk (Singh, 2004; Tran Le, 2010; Fox, 1989). The HCT-pH curves differ among milks. Some milks have noticeable maximum and minimum peaks, whereas others do not show these pronounced trends. For example, in Figure 1.3, milks are classified into type A and B milks with completely different curves. Figure 1.3. Heat coagulation time (HCT)-pH profiles for normal skim milk heated at 140 o C. Type A milk (o), type B milk (Δ), serum protein-free casein micelle dispersions ( ) and concentrated milk (20 % total solids) ( ) (Singh, 2004). Many researches indicated that protein usually has the least solubility at the isoelectric point (pi), because the lowest electrostatic forces of molecules increase the protein-protein interactions and reduce protein-water interaction (Van der Meeren, 2011; Tran Le, 2010; Pelegrine and Gasparetto, 2005). This condition easily leads to aggregation or possible precipitation at high temperature. As the ph is different from the pi, the protein charges can interact with water, which increases the protein solubility. The pi value of whey protein is at about ph 4.5. However, with a variety of proteins, the pi values are in a wide range of (Pelegrine and Gasparetto, 2005). Thus, the initial ph of milk also influences the HCT: the temperature for the coagulation induction is reduced with a lower initial ph. The rate of coagulation is inversely proportional to ph at a fixed temperature (Tran Le, 2010). 11

19 Chapter 1: Literature review Many studies have proposed that the ph has a strong influence on determining the dissociation from caseins and the association of whey proteins with the micelles (Singh and Fox, 1985; Visser and Jeurnink, 1997 ; Smits and Van Brouwershaven, 1980). Smits and Van Brouwershaven (1980) found that β-lg is more dependent on ph and temperature than α-la. The free cysteine residue in β-lg is considered as the fundament of the denaturation process, which involves several steps: dissociation of quaternary structure, changes in the conformation of the polypeptide chain and aggregation via disulphide bridges. Four disulphide bridges without free cysteine residues in α-la maintain the reversible denaturation of α-la, in the presence of calcium (Swaisgood, 1992). The heat stability-ph profile of milk is strongly influenced by this complex (Singh and Creamer 1992). Other intermolecular complexes, such as β-lg/α-la and β-lg/α s2 -casein are also formed upon heating. Dalgleish (1997) found that the rate of interaction between β-lg and α-la is directly proportional to their concentration. Β-lg seems to less effectively interact with casein micelles than α-la during the presence of whey protein in the casein micelles. The ph at which milk is heated is important in determining both the extent of casein dissociation from the micelle, and also whey protein association with the micelle (Singh and Fox, 1985; Smits and Van Brouwershaven, 1980). Above ph 6.9, κ-casein/β-lg complexes dissociate from the micelle upon heating, and in the ph range the κ-casein/β-lg complex remains within the micelle, stabilizing the micelle and reducing casein dissociation (Singh and Fox, 1985). Data obtained from studies of Singh and Fox (1985), when milk and serum protein-free casein micelle dispersions were heated at 140 o C for 1 min at different ph values, showed that at ph values lower than 6.7, the amounts of soluble κ-casein and total soluble protein from heated milk were lower than in a corresponding unheated milk, whereas at ph values above 6.7, these amounts were greater than in the unheated milk and increase with increasing ph. Studying in a broad ph range from 2 to 8 and heating from o C to check the sensitivity of β-lg (9 g/l), Verheul and Roefs (1998) observed two main features, i.e. a minimum in reaction rate of the denaturation /aggregation of β-lg in water at ph between 3-4 at o C and a strong increase between ph 6-8 at o C.The latter resulted from the decreasing stability in conformation of β-lg around ph 7 and an increased reactivity (dissociation) of the thiol groups. They commented, however, that above ph 8 accurate experiments were impossible, since the β-lg concentration decreased significantly in the heating up time (even 12

20 Chapter 1: Literature review at 65 o C), and actually denaturation reacted already at room temperature (data not shown) due to the formation of intermolecular disulphide bonds Influence of homogenization on heat stability During homogenization smaller fat droplets are formed leading to an increased surface area. More adsorption of caseins on the fat globules occurs, and hence more enhanced interactions will take place between adsorbed caseins and whey proteins during heating (McCrae, 1994). High pressure homogenization is also considered as a method of preservation and improvement of the quality of milk and dairy products alternative to the thermal treatment. High pressure homogenization is able to increase the stability of milk from the viewpoint of both colloidal, enzymatic and microbial stability (Thiebaud et al., 2003). Sweetsur and Muir (1983) observed that there was no effect on heat stability of homogenized skim milk below pressure of 31 MPa. In the beginning of the 20 th century, a number of publications demonstrated that homogenization can decrease the heat stability of milk (McCrae and Muir, 1991). However, this reduction is dependent on the type of homogenizer, homogenization temperature, pressure and the relative concentrations of protein and fat (Sweetsur and Muir, 1983). The reason for that phenomenon is due to more adsorption of protein, of casein in particular at the fat/ serum interface, but its mechanism has not been completely understood Effect of sulfhydryl- group reaction on the heat stability of milk It has been documented that the presence of disulfide (S-S) bonds and sulfhydryl (SH) groups in cystine and cysteine is important in the heat-induced gelation of proteins. Covalent cross-linking of protein molecules can be brought by SH oxidation into S-S bonds and/or by SH-induced S-S interchange reactions (Kazuko and Cheftel, 1989). Moreover, sulfhydryl groups in milk contribute to the cooked flavor, possibly due to the formation of H 2 S, affecting on the qualities of evaporated and dried milk and on other properties (Zweig and Block, 1953). Β-lg, the major protein component in whey (50-60 % of total bovine whey proteins) contains an important feature in its primary structure. The presence of two disulphide bridges (Cys 66 - Cys 160 and Cys 106 -Cys 119 ) and a free thiol group (Cys 121 ) lead to β-lg inaccessibility to solvent 13

21 Chapter 1: Literature review (Kinsella et al., 1989). The later group is equally distributed between positions 119 and 121, and hence it is located in a structural region which limits its accessibility to reagents, especially for the case of the C variant (Harold, 1989). Subsequently, aggregates are formed via intermolecular thiol-disulphide exchange, thiol-thiol oxidation and noncovalent hydrophobic interactions (Hoffmann and van Mil, 1997; McSwiney et al., 1994; Mulvihill and Kinsella, 1987). This may lead to heat-induced gelation of whey protein. Several researchers stated that the polymerization reaction initiates from dissociation of the dimers into monomers, then a change in the conformation of β-lg at o C at neutral ph conditions, which exposes the buried SH group of Cys 121 and starts sulfhydryl/disulphide (SH/S-S) interchange reactions, resulting into irreversible aggregation/ polymerization (Verheul and Roefs, 1998; Sawyer, 1968). Thiols and inorganic sulfides can be quantitated by using 5,5 -dithiobis-(2-nitrobenzoic acid) (DTNB or Ellman s reagent) in a spectrophotometric method, by analysis of thiol-disulfide exchange reactions and oxidative thiol modifications. The reaction is described by Equation (1) (Russell et al., 1997). (1) 14

22 Chapter 2: Materials and methods Chapter 2 Materials and methods 15

23 Chapter 2: Materials and methods 2.1. Materials High-heat skim milk powder (SMP) was obtained from Friesland Campina (Deventer, The Netherlands). SMP, according to the manufacturer, contained 37.3 % (w/w) protein, 0.5 % (w/w) fat and 0.15 % (w/w) phospholipids. The high oleic sunflower oil (Hozol, Contined, The Netherlands) contained maximum 0.05 % free fatty acid as oleic. Its melting point is at 0 o C and hence the oil remains clear even after 10 hours at 4 o C. Hozol oil was added in order to make the emulsions. Whey protein isolate (WPI) containing 92.6 % of protein (85 % β-lg) was obtained from Davisco Foods International, Inc. (BiPro, Le Sueur, MN, USA). Micelllar casein MPI 85 MC containing not less than 85.0 % of protein, not more than 5.0 % of water, not more than 1.0 % of lactose, 1.5 % of fat and 7.5 % of ash, was obtained from the Hungarian Dairy Research Institute (Budapest). The ph of a 10 % solution is situated in the range of , and the casein to whey protein ratio has a value of 95 : 5. Ellman s reagent or 5,5 -dithiobis (2-nitro-benzoic acid) (DTNB) was obtained from Sigma- Aldrich NV/SA (Bornem, Belgium). D (+)-Glucose (reagent ACS, anhydrous) and β-d-lactose (ca.80 % beta and 20 % alpha) were obtained from Acros Organics. Sucrose was obtained from VWR International bvba/sprl (Leuven, Belgium) Emulsion preparation Standard procedure Evaporated milk contains a total dry matter content of 23 %, consisting of 6.5 % oil and 16.5 % SMP. In order to check the effect of protein content on the heat stability of recombined evaporated milk emulsions, samples were prepared by mixing 23 % and 12 % (w/w) SMP, corresponding to 8.6 and 4.5 g of protein, respectively, with 6.5 % (w/w) Hozol oil as the dispersed phase, and the rest was sodium azide NaN 3 (0.02 %) aqueous solution. The latter was used as a preservative, to prevent any potential microbial growth. 16

24 Chapter 2: Materials and methods Different casein : whey protein (C:W) ratios In this case, model solutions (58.74 g protein/l corresponding to 16.5 % SMP) were prepared by mixing micellar casein and whey protein isolate to provide C:W ratios of 100:0, 90:10, 80:20, 70: 30, 50:50, 30:70 and 0: % (w/w) Hozol oil and β-d-lactose were added to have the same lactose content like in SMP (50 % lactose). Finally, NaN 3 solution (0.02 %) was added to prevent microbial growth ph adjustment In this series of experiments, two lots were prepared, characterized by the absence and presence of β-d-lactose, respectively. In the first one, emulsions were prepared by mixing micellar casein and whey protein isolate with C:W ratios of 100:0, 90:10, 80:20 and 70: % (w/w) Hozol oil was added as the dispersed phase, and NaN 3 (0.02 %) aqueous solution, without β-d-lactose, was the continuous phase. In the second case, the emulsion preparation was similar to the section with four ratios of C:W in the range of 100:0 to 70:30, and the addition of Hozol oil and β-d-lactose. The samples were homogenized as described in section 2.3 at a driving air pressure of 4 bar. Afterwards, the ph of the samples was adjusted to 6.6 and 6.8 with HCl 0.1 M Different sugars The emulsions were prepared by mixing micellar casein and whey protein isolate powder with C:W ratio of 80: 20. In order to evaluate the effect of sugars on the heat stability of recombined evaporated milk, three different types of sugar were selected in these experiments: β-d-lactose, β-d-galactose and sucrose. The same amount of each sugar was added to provide 50 % of carbohydrates as it is in SMP (16.5 %). The rest was 6.5 % (w/w) Hozol oil as the dispersed phase and NaN 3 (0.02 %) aqueous solution. The heat stability of these samples were compared with the blank containing 16.5 % SMP, Hozol oil and NaN 3 (0.02 %) aqueous solution. 17

25 Chapter 2: Materials and methods Addition of Ellman s reagent Each sample contained 23 % SMP or 16.5 % SMP with 6.5 % Hozol oil and the rest was 0.02 % NaN 3 aqueous solution. These mixtures were premixed in an Ultra-Turrax before adding Ellman s reagent in different concentrations: 0.05, 0.1 and 0.2 % (w/w). The blanks in this case were samples containing 23 % and 16.5 % SMP without the addition of Ellman s reagent, respectively Homogenization Emulsions were pre-homogenized by an IKA Ultra- Turrax TV45 (Janke & Kunkel,Staufen, Germany) for about 1 minute and then homogenized by a Microfluidizer 110S (Microfluidics Corporation, Newton, Massachusetts, USA) having its coil immersed in a water bath at 55 o C. In order to evaluate the effect of homogenization pressure on heat stability, the samples were microfluidized at a compressed air pressure of 1, 2 and 4 bar, corresponding to a liquid pressure of 140, 280 and 560 bar, for about 2 minutes. In other cases (check the effect of ph, of heat treatment, of protein concentration, with addition of Ellman s reagent, with sugars), the emulsions were homogenized using a compressed air pressure of 4 bar Heat treatment Prior to heat treatment, a volume of 10 ml from each sample was transferred to 20 ml headspace glass vials (75.5 x 22.5 mm, 1 st hydrolytic class, DIN-crimp neck, long neck, HCbottom; Caps: 20 mm combination seal: aluminum cap, plain, with centre hole, silicone transparent blue/ PTFE white, 35 o shore, 3.0 mm) which were clipped on a platform and positioned in a temperature-controlled oil bath (Fritel turbo SF, 5 L capacity). In order to check the effect of heating temperature on heat stability of recombined evaporated milk, the samples were heated at lower temperatures ( o C) in a water bath for minutes, whereas, a temperature controlled oil bath was used for samples which were treated at sterilization temperatures (i.e. 121 o C). For other cases (check effect of ph, of sugars and of Ellman s reagent), samples were heated at sterilization temperature (121 o C) in different time intervals. 18

26 Chapter 2: Materials and methods After heating, the samples were cooled down to room temperature in water for 5 minutes Viscosity measurement A Brookfield Programmable LV DV-II + viscometer (USA), at speeds ranging from 0 to 200 RPM (rotational speed per minute), was used to measure the viscosity of samples. Spindle 21(with an 8 ml small volume adapter) was used to measure the viscosity of samples which maintained their liquid structure after heating. On the other hand, spindle 34 was used to measure the viscosity of samples with a semi-liquid, gel-like or solid structure. The measurements were performed at 25 o C, controlled by a temperature-controlled circulating water bath. The conversion of RPM into shear rate (s -1 ) is carried out by multiplying the value of rotational speed (in RPM) with specific coefficients: 0.93 and 0.38 s -1 for spindle 21 and 34, respectively Particle size distribution The particle size distribution of the emulsions was determined using the Mastersizer S long bed (version 2.15, Malvern Instruments Ltd) with MSX 17 wet sample dispersion unit sample. The 300RF lens was used in the wet laser diffraction analysis and for the particle size distribution determination, the polydisperse model was chosen in the Malvern Mastersizer software. Sample was added dropwise until an obscuration of about 10 % was obtained. For the result calculation, the presentation code incorporated a continuous phase refractive index of 1.33 and a real and imaginary dispersed phase refractive index of and , respectively Protein recovery The protein recovery was calculated based on the content of protein in the serum phase after centrifugation. Samples containing 23 % of SMP were heated at 100 o C for 10, 20, 30, 40 and 60 minutes and at 121 o C for 10 and 20 minutes. After cooling down, sucrose was used as a density increasing agent in order to make centrifugation more effective. For this purpose, 5 ml sample was mixed with 2.5 g sucrose, dissolved completely and then 5 ml of each final mixture was transferred into a plastic centrifugation tube (13 x 51 mm, Beckman Instruments, Inc., California). Centrifugation was executed by Beckman L7-55, at RPM, for 90 minutes at 20 o C. After centrifugation, the cream layer was carefully removed, whereas the 19

27 Chapter 2: Materials and methods sediment and the serum phase were mixed thoroughly. Finally, the protein content was determined by the Kjeldahl method, by using 0.5 g of each mixture. The protein content was calculated according to the formula: % P = V * T * 14 * C * 100 / 1000 * M V: number of ml of the HCl solution T: normality of HCl C: conversion factor (6.38 for milk and milk products) M: mass (g) of the sample 20

28 Chapter 3: Results and discussion Chapter 3 Results and discussion 21

29 Chapter 3: Results and discussion 3.1. Effect of heating time and temperature on heat stability In a first series of experiments, recombined milk containing 23 % SMP and 6.5 % Hozol oil was prepared by homogenization at a driving air pressure of 4 bar to evaluate the effect of heating temperature and duration on its heat stability. To that end, the recombined milk samples were heated at temperatures from 60 to 121 o C for intervals of 10, 20, 30, 40, 60 and 120 minutes and the effect was evaluated by visual observation and by means of viscosity, particle size distribution measurement. No significant changes occurred, when heating from 60 to 80 o C for 10 till 120 minutes: the samples remained completely white and liquid. However, after heating at 100 o C, there was a significant effect of the immersion time and temperature on the particle size distribution and the volume-weighted average particle diameter. As there is already a slight increase in diameter at 90 o C, most probably aggregation starts from that temperature, as shown in Figure 3.1. Similarly, table 3.1 demonstrates that the average particle diameter d(4,3) slightly increased from 0.34 μm for unheated samples, to 0.61 μm for samples heated at 90 o C for 60 minutes, while upon heating at 100 o C a clear increase in diameter was observed with values of more than 2 μm and even up to 7 μm after 2 hours heating. On the other hand, several publications mentioned that from 70 o C onwards, some moderate thermal changes occur, especially protein denaturation (Chen et al., 2005; de Wit, 1984). The mechanism of protein denaturation is related to the unfolding of the secondary and tertiary structure of globular whey proteins, leading to the release of hydrophobic as well as sulfhydryl groups. The thiol/disulphide interaction may lead to intermolecular disulphide bridges, resulting in the formation of complexes with casein micelles or in whey proteinwhey protein interactions. This in turn leads to aggregation, then coagulation and precipitation. Hence, denaturation decreases the solubility of whey proteins (Dannenberg and Kessler, 1988; Vasbinder, 2002; Pelegrine and Gasparetto, 2005; Smits and Van Brouwershaven, 1980). Similarly, Anema and Li (2003a) reported that a rapid increase in the casein micelle size, measured by photon correlation spectroscopy, occurred during the early stages of heating of skim milk at ph 6.55 from 75 to 100 o C for up to 60 minutes. Overall, the casein micelle size increased by about nm upon heating. Moreover, they stated that the mass of the particles did not significantly increase as a consequence of the heating which was derived from the fact that the scattering intensity of the particles did not change significantly. 22

30 Chapter 3: Results and discussion It is interesting that there was a small decrease in particle size d(4,3) and viscosity (results not shown) during heating at temperatures below the whey protein denaturation temperature in this study, for example, samples containing 12 % SMP and heated at 90 and 100 o C (Figure 3.8.) and samples containing 23 % SMP at 60 o C. A similar decrease was also mentioned by Jeurnink and De Kruif (1993), when measuring the viscosity and by Anema and Li (2003a), when measuring the casein micelles size of heated milk samples. The explanation for this decrease has not been clarified yet. Jeurnink and De Kruif (1993) ascribed this effect to the precipitation of calcium phosphate on the casein micelles, which results in the slight shrinkage of the micelles. However, this precipitation would be expected to be reversible on subsequent cooling and storage unless the native colloidal calcium phosphate is solubilized in preference to the heat precipitated calcium phosphate. 23

31 Volume (%) Chapter 3: Results and discussion o C 0 min 10 min 20 min 30 min 40 min 60 min 120 min 80 o C 90 o C 100 o C 121 o C Particle size (µm) Figure 3.1. Cumulative (upper row) and differential (lower row) volume-weighted particle size distributions (μm) of recombined evaporated milk containing 23 % SMP homogenized in a microfluidizer at a driving air pressure of 4 bar, after heating from 60 o C to 121 o C for 0, 10, 20, 30, 40, 60, and 120 minutes. 24

32 Chapter 3: Results and discussion Table 3.1. Effect of heating time and temperature on the volume-weighted average particle diameter d(4,3) of recombined milk containing 23 % SMP homogenized in a microfluidizer at a driving air pressure of 4 bar, after heating from 60 o C to 121 o C for 0, 10, 20, 30, 40, 60, and 120 minutes. 60 o C 80 o C 90 o C 100 o C 121 o C 10 min min min min min min Similar particle size increase with increasing heat treatment results were found by Raikos (2010), as well as Tran Le (2010). This effect was attributed to fat globule aggregation, which resulted from the increased intermolecular interactions between the protein molecules adsorbed onto different droplets. Furthermore, studies hypothesized that droplet aggregation was induced by interactions between non-adsorbed protein molecules in the serum phase and proteins adsorbed at the interface of fat globules (de Wit, 1984; Raikos, 2010; Tran Le, 2010). On the other hand, according to Demetriades (1997), the increase in particle size followed a different pattern. When emulsions were heated from 65 to 80 o C for 30 min, there was an increase in particle size due to flocculation of emulsion droplets. However, further heat treatment (80-90 o C) led to a decrease in particle size. This behaviour was attributed to the extent of protein denaturation at the oil-water interface. Adsorbed whey proteins were considered partially unfolded when heated at temperatures as high as 80 o C, promoting surface hydrophobicity and droplet flocculation. At higher heating temperatures, proteins were completely unfolded, leading to rearrange effectively all non-polar amino acids towards the oil phase, and hence reducing the tendency for aggregation. In this case, it was concluded 25

33 Chapter 3: Results and discussion that upon initial aggregation of oil droplets, the aggregates become deaggregated with further heat treatment, causing the formation of smaller, more compact emulsion droplets (Raikos, 2010). According to Dickinson and Parkinson (2004), the decrease in protein solubility upon heating involves an increase in viscosity, which is in agreement with our results. Figure 3.2 depicts that the viscosity was also a function of the heat treatment extent, resulting in the formation of a very viscous gel (almost solid) for samples heated for at least 20 minutes at 121 C, whereas semi-liquid samples were induced after 10 minutes heating at 121 o C. A similar behavior in viscosity was observed compared to the volume-weighted particle size distributions of recombined evaporated milk containing 23 % SMP. An increased viscosity was assessed after heating recombined evaporated milk emulsions at 90 o C and certainly the increase evolved upon increasing both temperature and time of heating. However, a slight increase on viscosity was determined even at 80 o C. More precisely, the viscosity values ranged from 10.1 to 12.7 mpa.s (within the shear rate range of s -1 ) for samples heated for minutes. On the contrary, all samples heated at 60 o C, had relatively constant viscosity throughout the heating process, with values around mpa.s. According to Corredig and Dalgleish (1996, 1999), both the rate and degree of protein denaturation and complex formation depend on the processing conditions, especially heating temperature. At temperatures from 70 to 90 o C, both β-lg and α-la participate in the formation of coaggregates with casein similarly. The degree of this complex formation increases with heating time and temperature in the region of o C due to more denatured whey protein. However, at higher temperatures, this interaction is replaced by faster protein-protein interactions. 26

34 Viscosity (mpa.s) Chapter 3: Results and discussion o C 0 min 10 min 20 min 30 min 40 min 60 min 120 min 80 o C 90 o C 100 o C 121 o C Shear rate (s -1) Figure 3.2. Apparent viscosity (mpa.s) of recombined milk homogenized at a driving air pressure of 4 bar and heated from 60 o C to 121 o C for 0 ( ), 10 ( ), 20( ), 30 ( ), 40 ( ), 60 ( ), and 120 ( ) minutes. 27

35 Chapter 3: Results and discussion According to the studies of Galani and Owusu Apenten (1999), the importance of noncovalent interactions varied with temperature. They found that at low temperatures (< 75 o C), non-covalent interactions hardly contribute to aggregation. Thus, past studies could not find clear evidence for the formation of non-covalently associated aggregates when they emphasized on disulphide bond mediated aggregation at low temperatures within the range o C in the works of Hoffmann and van Mil (1997). In addition, they stated that noncovalent interactions become important at temperatures of o C, which agreed with our results as well. In order to evaluate the effect of severe heat treatment (i.e. at 100 o C for minutes and 121 o C for minutes) on the protein load, percentage of protein recovery in the serum phase (aqueous phase) was also determined after centrifugation at RPM for 90 minutes at 20 o C. According to Anema, et al. (2004), higher levels of whey proteins were deposited with the pellet, although there was little change in the serum casein content as a high centrifugation speed was used, and therefore a high g force. This suggests that, under high centrifugation force, some of the soluble whey protein aggregates were deposited even though these aggregates were not associated with the micelles. To reduce the possibility of centrifuging down soluble whey protein aggregates, the centrifugation speed chosen was the minimum required to effectively deposit the casein micelles/associated whey proteins as a firm pellet. From the obtained results, depicted in Figure 3.3, it is obvious that after heating at 100 o C for long duration, the protein recovery declined remarkably from more than 12 % in unheated sample to about 8-9 % after minutes of heating. This was in accordance with previous studies and explanations about formation of complexes of denatured proteins due to their heat-induced interactions. However, after the first 10 minutes of heating at sterilization temperature (121 o C), the protein recovery reduced to nearly one third compared to the samples before heating. This implies that heating temperature is a decisive factor in the maintenance of protein in the serum phase. Upon further heating for an additional 20 minutes, the slightly decreased trend continued from 5.4 % (for 10 minutes) to 3.9 % of protein recovery. This was in line with our expectation that upon increasing both temperature and heating time, the protein load of the serum phase was diminished. This implies an increased heat coagulation resulting in the formation of bigger entities, due to denaturation of whey proteins and hence protein-protein interactions. Figure 3.4 indicates that the formation of 28

36 Chapter 3: Results and discussion cream occurs due to the lower density of the fat globules compared to water and therefore they accumulate at the top. The larger is the degree of heat coagulation, the more the whey proteins are denatured and form complexes with caseins and fat droplets and hence more cream arises to the surface. Consequently, a smaller amount of intact free protein will remain in the serum phase after centrifugation. 100 o C 121 o C Figure 3.3. Protein recovery (%) as a function of heating time and temperature for recombined evaporated milk containing 23 % SMP prepared by microfluidization at a driving air pressure of 4 bar compressed air, upon heating at 100 o C for 10, 20, 40 and 60 minutes and 121 o C from 10 to 20 minutes. T o C Figure 3.4. Recombined evaporated milk samples containing 23 % SMP that were either unheated, or heated at 90 o C, 100 o C or 121 o C (from left to right, respectively) for 1 hour, and then centrifuged at high speed and the cream layer, respectively. 29

37 Chapter 3: Results and discussion Moreover, the color change is also a function of heat treatment. The color became more visible yellowish and brownish with increasing heating time. These color changes occurred due to the Maillard reaction between lactose and amino acids in milk, and agreed with the results of Tran Le (2010). The higher the temperature of the heat treatment, the larger the sterilizing effect and the more marked the change in the colour and taste of the milk. When milk is heated, off-flavours occur. At first, a cooked flavor is obtained, caused by the production of volatile sulfur compounds. Then, when the intensity of the heat treatment is increased, a sterilization taste originates, which is mainly caused by the reaction between the sugar and the protein constituents (Maillard s reaction). As described above, the latter process impairs not only the taste of the milk but also its color, which becomes brownish. According to studies about sterilization of milk by Galesloot (1962), the temperature-time combination has exerted an important influence upon the choice of techniques. Table 3.2 illustrates the results of the effect of heat treatment with a number of temperature-time combinations on the relative degree of browning with the same sterilizing effect. Moreover, he also stated that with increasing temperature the rate of spore-destruction increases far much more than the influence on the taste and color of the milk. This is the reason why the method of high temperature-short time (HTST) is often applied in milk sterilization in dairy industry. Another important finding is the whitening effect in milk upon heat treatment. By sufficiently intensive heating the serum proteins are more or less denatured, leading to the milk reflecting more light and hence appearing whiter. Table 3.2. Influence on colour of milk of exposure to different temperatures for different durations of time (reproduced from Galesloot 1962). Temperature of heating* ( o C) Time * Relative degrees of browning minutes 60 minutes 6 minutes 36 seconds 3.6 seconds 0.36 seconds

38 Chapter 3: Results and discussion *: All these time-temperature combinations produce the same sterilizing effect. In fact, information concerning off flavors in heated milk is limited; the majority of milk flavor research has investigated the oxidized off flavors occurring during storage of milk. According to Dewettinck (2011), two popular flavors arise in heat-treated milk due to the heat treatment such as cooked, caramelized and scorched flavors and oxidation flavours. The flavor of heat-treated milk is affected by several factors namely the severity and type of heat treatment, the milk quality, packaging material, storage temperature and time, as well as oxygen content during heat treatment and storage. Immediately after UHT-heat treatment, no undesirable odours are found in raw milk, but the cooked flavors, especially caused by sulfhydryls and other sulfur compounds, are very intense. They decrease during storage with a sufficient oxygen content in milk. In general, it can be concluded that thermal processing (i.e. heating time and temperature) is one of the most important unit operations in the dairy industry, and hence has been an important object to investigate over several decades Effect of protein concentration on heat stability Whereas regular recombined evaporated milk contains 16.5 % SMP, in this series of experiments, either 12 % or 23 % SMP was used in order to check the effect of protein concentration on the heat stability of recombined evaporated milk Change in particle size distribution The particle size distribution of the recombined evaporated milk samples containing 12 % and 23 % SMP was determined by laser diffraction. From Figure 3.5 (right), it is obvious that in 12 % SMP the onset of aggregation only occurred at sterilization temperature (about 121 o C), while that was observed in 23 % SMP already at 100 o C (left). In addition, at sterilization temperature, the increase in volumeweighted particle size distribution was seen from the first minutes of heating (10 min) in 23 % SMP, while this behavior was only obvious after 1 hour of heating in the case of 12 %. Thus, it can be concluded that the rate and the degree of heat coagulation of recombined milk is strongly protein content-dependent. This result suggests that larger heat induced aggregates 31

39 Chapter 3: Results and discussion were created with higher protein content. Similar observations were published by Graciá- Juliá et al. (2008), when they compared dispersions of 6 % and 10 %, w/w, protein. Moreover, they also observed a clear and significant decrease in protein solubility above 69 o C for dispersions at 10 % (w/w) protein and above 74 o C at 6 % protein. A similar result is shown in Figure 3.6, which depicts the average particle diameter upon heating. It becomes very clear that the recombined evaporated milk emulsions containing 23 % SMP had an increased diameter at temperatures in the range between 80 and 121 o C, whereas in the presence of 12 % SMP, the diameter did not show any particular change at temperatures below 121 o C, even after heating for 2 hours. Particle size measurements of the recombined evaporated milk samples after heating showed that increasing the proportion of SMP, meaning increasing the casein micellar as well as the whey protein content, resulted in an increase in the diameter of the particles. Hoffmann (1997) stated that both the size and the amount of aggregates increase with increasing protein concentration. Indeed, heating of dispersions at high protein concentration increases the rate of aggregation propagation (but not the rate of the unfolding step). This probably explains the wider particle distributions with a predominance of large aggregates observed at 23 % SMP as compared to 12 % SMP at o C (Figure 3.5). 32

40 Volume (%) Volume (%) 1 bar 2 bar 4 bar Chapter 3: Results and discussion o C 0min 10min 20min 30min 40min 60min 120min o C 121 o C Particle size ( m) o C unheated 10 min 20 min 30 min 40 min 60 min 120 min o C Particle size ( m) 121 o C Figure 3.5. Cumulative volume-weighted particle size distribution of recombined milk containing either 23 % SMP (left graph) or 12 % SMP (right), prepared by microfluidisation at a driving air pressure ranging from 1 to 4 bar compressed air, after heating at 90, 100 and 121 o Cfor 10 min up to 120 minutes. 33

41 Diameter, d 4,3 (um) 23% SMP 12% SMP Chapter 3: Results and discussion 90 o C 100 o C 121 o C 75 Unheated min min 30 min 40 min 60 min 120 min Unheated 10 min 20 min 30 min 40 min 60 min 120 min bar 2 bar 4 bar 1bar 2 bar 4 bar 1bar 2 bar 4 bar Pressure Figure 3.6. Effect of SMP content of a recombined milk emulsion with 6.5 % Hozol, prepared at a driving pressure ranging from 1 to 4 bar compressed air, on the volume-weighted average diameter (μm) after heating at 90, 100 and 121 C for 10 up to 120 minutes. 34

42 Chapter 3: Results and discussion Change in viscosity In order to check the effect of protein concentration on the viscosity of recombined evaporated milk containing either 12 or 23 % SMP, samples were premixed by an Ultra- Turrax and eventually homogenized by a microfluidizer 110S operating at a driving air pressure of 1bar, 2 bar and 4 bar of compressed air for 1 minute, followed by heating for different times at different temperatures. Due to its higher protein content, the viscosity of the unheated emulsion containing 23 % of SMP was roughly twice higher as compared to that containing 12 % SMP with viscosity values ranging from mPa.s (corresponding to a shear rate range of s -1 ) for samples homogenized at a driving air pressure of 4 bar. The same observation of whey protein denaturation and aggregation can be conducted from the viscosity values for both emulsions. As shown in Figure 3.7 and 3.8, the apparent viscosity of samples containing 23 % SMP showed a gradual increase with heating time at 100 o C, while this trend in 12 % was only observed at sterilization temperature (i.e. 121 o C, even after 1 hour of heating in the case of 1 bar homogenization). This means that the lower is the protein concentration, the slower and the smaller amount of whey protein is denatured and hence the slower the degree and rate of protein aggregation is. Therefore, we can conclude that although the volume-weighted average particle diameter d(4,3) of unheated samples is hardly affected by the protein concentration, still the latter significantly influences the rate and degree of heat-induced denaturation and aggregation of whey proteins, as illustrated by the increase in viscosity, and in particle size distribution values as well as by the visible observation of the structure of the heated samples. 35

43 Viscosity (mpa.s) 1 bar 2 bar 4 bar Chapter 3: Results and discussion 10 7 unheated 10 min 20 min 30 min 40 min 60 min 120 min o C 80 o C 90 o C 100 o C 121 o C Shear rate (s -1 ) Figure 3.7. Apparent viscosity (mpa.s) as a function of heating time and temperature for recombined evaporated milk containing 23 % SMP prepared by microfluidization at a driving air pressure ranging from 1 to 4 bar compressed air. 36

44 Viscosity (mpa.s) 1 bar 2 bar 4 bar Chapter 3: Results and discussion o C unheated 10 min 20 min 30 min 40 min 60 min 120 min 100 o C 121 o C Shear rate (s -1 ) Figure 3.8. Apparent viscosity (mpa.s) as a function of heating time and temperature for recombined evaporated milk containing 12 % SMP prepared by microfluidization at a driving air pressure ranging from 1 to 4 bar compressed air. 37

45 Chapter 3: Results and discussion 3.3. Effect of homogenization pressure on heat stability In the next series of experiments, the combined effect of composition and processing of recombined milk on its heat stability was evaluated. To that end, the heat stability of recombined milk containing either 23 % or 12 % of high-heat SMP and 6.5 % hozol oil, homogenized in a microfluidizer operating at a driving air pressure within the range from 1 to 4 bar, was evaluated. The main purpose of this experiment was to evaluate the impact of the residual free (i.e. not adsorbed) protein in the recombined milk emulsion, which was thought to be increased by selecting a higher protein content as well as a lower homogenization pressure (and hence a larger particle size and lower interfacial area). High pressure homogenization is a unit operation to decrease the particle size, which is important in dairy industry in order to decrease the creaming rate (Van der Meeren, 2011; Dewettinck, 2011), and to increase the resistance to cold agglutination as well as the heat stability of concentrated milk (Whiteley and Muir, 1996). In order to evaluate the effect of homogenization on the heat stability of recombined milk, the samples were pre-homogenized by Ultra-Turrax and then homogenized by Microfluidizer 110S at a constant temperature of 55 o C at a compressed air pressure in the range of 1, 2 and 4 bar, corresponding to a liquid pressure of 140, 280 and 560 bar, for about 1 minute. The results, as shown in Figure 3.9, indicate that the volume-weighted mean diameter d(4,3) of the recombined milk decreased when the microfluidization driving air pressure was increased from 1 bar to 4 bar for unheated milk. A similar trend was also seen in the region from o C in emulsions containing 23 % SMP and at higher heating temperature ( o C) in the case of 12 % SMP. This result was in the agreement with the results obtained by Olson (2004) when skim milk, semi-skimmed milk (2 % fat), and whole milk were homogenized in a microfluidiser with a homogenization pressure in the range from 50 to 100 MPa. He observed, however, that there was an increase in volume-weighted mean diameter d(4,3) in whole milk from 304 to 383 nm with increase in pressure from 100 to 200 MPa, and no significant change in skim milk and semi-skimmed milk compared to 100 MPa. Desobry-Banon (1994) also found a drop in the mean size of casein micelles when reconstituted milk was pressurized from 230 to 430 MPa. This was attributed to the change in conformation from spherical particles into chains or clusters of submicelles. 38

46 Diameter ( m) 23 % SMP 12 % SMP Chapter 3: Results and discussion o C 1 bar 2 bar 4 bar 80 o C o C 100 o C Heating time (minutes) Figure 3.9. Volume-weighted mean diameter d(4,3) (μm) of recombined evaporated milk with 6.5 % Hozol oil and either 23 % SMP at 60 o C and 80 o C(above) or 12 % SMP at 90 o C and 100 o C (below) from 0 to 120 minutes, homogenized at a driving air pressure of 1, 2 and 4 bar. On the other hand, the cumulative volume-weighted particle size distributions as a function of homogenization pressure (Figure 3.5) demonstrate the reversed effect at higher heating temperature from 90 o C onwards in the case of 23 % SMP. It can be concluded that increasing the homogenization pressure leads to a decrease in heat stability. This finding is consistent with work by other investigators (McCrae et al., 1994; Sweetsur and Muir, 1983; McCrae and Muir, 1991). Similarly, McCrae (1991) found a decrease in heat stability of recombined milk at ph 6.7, corresponding with a shorter heat coagulation time with increasing homogenization pressure: the heat coagulation time decreased from 17.5 over 15.5 to 14.0 min when increasing the homogenization pressure from 5.2 over 10.3 to 20.7 MPa. For the sake of completeness, it should be mentioned that no significant effect of homogenization pressure was observed at both lower and higher ph values. In addition, Lopez-Fandino (1996) reported that homogenization of whole milk at pressures from 100 to 400 MPa improved its coagulation properties leading to promote protein and moisture retention in the 39

47 Chapter 3: Results and discussion curd of fresh cheese, and consequently increasing the cheese yield. Muir (1985) mentioned that the reduction in size of milk fat globules due to an increase in homogenization pressure leads to the increase in fat droplet surface area, which is stabilized by adsorption of an interfacial layer of milk proteins, with predominance of caseins. Consequently, the heatinduced interactions of denatured whey proteins with the adsorbed casein layer decrease the heat stability (Cano-Ruiz and Richter, 1997). Another reason involves the decrease of casein particles, which results in an increase in the specific surface area, and an increase in the number of smaller particles leading to increased interparticle collision, reduced steric repulsion, and consequently, the enhancement of aggregation, including acid and rennet aggregation (Desobry-Banon et al., 1994; Lopez-Fandino et al., 1996) Effect of Casein: Whey (C:W) protein ratio on heat stability In fact, very few studies that are related to the effect of the C:W ratio on the heat-induced denaturation/aggregation of whey proteins have been published. As whey proteins are thought to play an important role in the heat coagulation process, recombined milk samples were prepared of varying whey protein to casein ratio. To that end, SMP was replaced as protein source by a combination of a casein micellar powder and whey protein isolate. Hereby, the amount of both powders added was adjusted to obtain the same protein content (i.e %) as in emulsions that contained 16.5 % SMP. In order to investigate the effect of the C:W ratio on the heat-induced denaturation/aggregation, model emulsions were prepared from whey protein isolate powder and casein micellar powder with C:W ratios of 100:0, 90:10, 80:20, 70:30, 50:50, 30:70, and 0:100. Protein and fat contents were kept constant at all ratios like in recombined evaporated milk with 16.5 % SMP. Heat treatment was performed at 121 o C for 20, 30, 35 and 40 minutes Without ph adjustment and without sugar addition Figure 3.10 demonstrates that in the first minutes of heating, the emulsion containing 100 % whey protein isolate seemed to be the most susceptible to heat denaturation with slightly increased viscosity. This was expected as it is known that whey proteins are more susceptible to heat denaturation than caseins due to their conformational structure. Surprisingly, upon further heating for additional 35 minutes, another trend was observed. 40

48 Viscosity (mpa.s) Chapter 3: Results and discussion Whereas the dispersion with 100 % whey protein isolate gradually increased in viscosity, much more denaturation occurred in the case of C:W ratios of 100:0 and 90:10. The mechanism by which heat destabilization of these emulsions proceeds is not yet understood but one could expect that the presence of casein micelles is involved. From the result after heating for 40 minutes, it can be concluded that an unbalanced ratio of C:W (i.e. when either casein or whey is dominant) leads to emulsions that become more sensitive to heat denaturation, whereas the equilibrium ratio, as found in milk, seems to lead to more heat stable emulsions. Figure 3.11 more clearly depicts the trend of viscosity upon heating. From these findings, it can be suggested that the mixture of casein and whey should be in a balanced ratio in order to obtain the highest heat stability. unheat c/0w 90c/10w 80c/20w 70c/30w 50c/50w 30c/70w 0c/100w 20 min 30 min 35 min 40 min Shear rate (s -1 ) Figure Apparent viscosity (mpa.s) of recombined evaporated milk with 6.5 % Hozol oil and 5.87 % milk proteins (containing a variety of C:W ratios) heated at 121 o C for 20 up to 40 minutes. 41

49 Diameter, d 4,3 (um) Consistency coefficient(mpa.s n-1 ) Chapter 3: Results and discussion c/0w 90c/10w 80c/20w 70c/30w 50c/50w 30c/70w 0c/100w unheat 20 min 30 min 35min 40 min Figure Consistency coefficient (mpa.s n-1 ) of emulsions with 6.5 % Hozol oil and 5.87 % protein with different C:W ratios heated at 121 o C for 20 up to 40 minutes c/0w 90c/10w 80c/20w 70c/30w 50c/50w 30c/70w 0c/100w unheat 20min 30 min 35 min Figure Effect of C:W ratio on the volume-weighted mean diameter (μm) of recombined evaporated milk containing 6.5 % Hozol oil and 5.87 % dairy proteins upon heating at 121 o Cfor 20, 30 and 35 minutes. 42

50 Chapter 3: Results and discussion Figure 3.12 shows the volume-weighted mean diameter d(4,3) of emulsions with different ratios of casein micelles to whey proteins, was consistent with the viscosity findings. An average diameter of about 0.45 μm was found for all unheated emulsions; increasing the proportion of whey proteins in the model emulsions had little effect (from 0.37 μm in C:W of 0:100 to 0.55 μm in 100:0 ratio) on the average particle diameters. After heating at 121 o C up to 30 minutes, there was only a slight increase in diameter in emulsions with 100 % whey proteins (up to 0.88 μm), while at all other ratios not any significant change in diameter was observed. A similar behaviour was noticed after 35 minutes heating, with a dramatically increased diameter in emulsions containing almost only casein micelles: the volume-weighted diameter increased up to 370 and 338 μm upon 35 min heating at a C:W ratio of 100:0 and 90:10, respectively. The results published by Beaulieu et al. (1999) confirmed that the proportion of whey proteins in the heated solutions had an important effect on the formation of aggregates. Higher whey protein concentrations led to the formation of bigger particles mainly composed of whey proteins. The increase in particle size upon heating is typically related to the aggregation of whey proteins at the surface of casein micelles via the β-lg/κ-casein complexes and disulfide-linked aggregates between κ-casein and whey proteins (Singh and Creamer, 1992; Smits and Van Brouwershaven, 1980). Moreover, these researches noted that particles formed by aggregation between whey proteins and casein micelles did not form infinite aggregates, and an excess proportion of whey proteins could aggregate with themselves. By using electron microscopy, Beaulieu et al. (1999) noted that heat treatment of milk not only increased the size of casein micelles but also increased the number of protein particles smaller than casein micelles. The smaller particles were composed of heat-denatured whey proteins which did not present at the surface of micelles. The difference between our results and the results of Beaulieu et al. (1999) may be explained by the difference in experimental procedure. While our model emulsions were set up in the original ph (within a range from 7.0 to 7.2), and were heated at 121 o C in a temperaturecontrolled oil bath for various time intervals from 20 to 40 minutes. While in the latter experiment, raw milk was skimmed, and model solutions were prepared by mixing skim milk, its ultrafiltration permeate and industrial whey protein isolates to provide different C:W ratios. Then, the ph was adjusted to 6.7 and the samples were heated at 95 o C for 5 minutes in a water bath. 43

51 Chapter 3: Results and discussion From the results of this series of experiments, it can be concluded that the ratio of caseins to whey proteins is important. A properly selected ratio significantly increased the heat stability of recombined evaporated milk With ph adjustment but without sugar addition In recombined evaporated milk containing SMP as protein source, the SMP has a profound buffering effect and hence all samples have a quite similar ph. Replacing the SMP by a combination of casein micelles and whey protein isolate, the buffering effect may be lost and hence the effect of C:W ratio might also be (partly) caused by ph variations. In order to overcome the latter complication, the ph was adjusted in the below experiments. C:W ratios of 100:0, 90:10, 80:20 and 70:30 were selected in similarity with the C:W ratio of about 80:20 in bovine milk. The sensitivity of heat coagulation towards ph was studied by adjusting the ph with 0.1 M HCl from its original ph (around 7.1) to 6.8 and 6.6. In the absence of lactose, the viscosity is a function of the C:W protein ratio and the ph of the emulsions before heating, as shown in Figure However, these effects are quite small. There was a small deviation in viscosity from 4.0 to 4.9 mpa.s (within a shear rate range of s -1 ) among samples containing different C:W ratios in neutral ph, while in ph 6.8 this was seen lowest within the mpa.s range. In addition, the effect of ph on the heat stability of reconstituted milk with varying casein/whey protein ratios was assessed by measuring the viscosity upon heating as a function of ph (Figure 3.15). After 2 minutes heating at 121 o C, only the emulsion with 100 % casein and without whey protein isolate and adjusted to ph 6.6 was heat denatured, while in the emulsion at that ph in the presence of whey protein isolate no change in viscosity was observed. A similar result was obtained in the case of ph 6.8 after 5 minutes of heating with denaturation in the 100 % casein containing emulsion and maintenance in (low) viscosity of the other samples containing whey protein isolate. Surprisingly, after heating 5 minutes at 121 o C, all samples adjusted to ph 6.6 became solid, whereas no change in viscosity was found in those with the original ph. It can be concluded that the heat stability of protein-stabilized emulsions or more precisely, the level of denatured whey proteins associating with the micelles is strongly dependent on the ph of the milk during heating, especially in the predominance of caseins. This may be explained as a consequence of the ph-dependent interactions (attraction or repulsion) between the particles or in the serum induced upon heating, and these relationships were directly reflected in the viscosity of the milk samples (Anema et al., 2004). 44

52 Chapter 3: Results and discussion Our findings were in accordance to those obtained by Anema, et al. (2004). Figure clearly shows that the change in relative viscosity (η/η 0 ) of milk samples heated for up to 30 minutes at 90 o C was a function of ph with values ranging from 6.5 to 6.7. While an increase in viscosity at ph 6.5 for 30 minutes was seen by about 15 %, this behavior at ph 6.7 was only by 5 %. Moreover, the relative particle size also followed a comparable trend to the observed viscosity, with increasing particle diameter as the ph was reduced from 6.7 to 6.5. Figure Relative viscosity (η/η 0 ) of milk (A) and relative particle size (d/d 0 ) of casein micelles in milk (B). the milk samples were heated at 90 o C for various times at different ph values: ph 6.5 ( ), ph 6.55 (o), ph 6.6 ( ), ph 6.65 (Δ) and ph 6.7 ( ). Error bars are the standard deviation for any mean value where two or more samples were analysed (redrafted by Anema, et al.(2004)). 45

53 Viscosity (mpa.s) Chapter 3: Results and discussion C/0W 90C/10W 80C/20W 70C/30W Original ph ph = 6.8 ph = Shear rate (s -1 ) Figure Effect of casein to whey protein ratio and ph on viscosity (mpa.s) of unheated emulsions without lactose. 46

54 Viscosity (mpa.s) ph = 7.2 ph = 6.8 ph = 6.6 Chapter 3: Results and discussion c/0w 90c/10w c/20w 70c/30w 2 min 5 min Shear rate (s -1 ) Figure Effect of ph on the viscosity (mpa.s) of emulsions with variable casein to whey protein ratio without added lactose; the emulsions were heated at 121 o C for 2 minutes and 5 minutes. 47

55 Chapter 3: Results and discussion In another study, Anema and Li (2003b) also observed a strong dependence of the change in casein micelle size on ph, with the largest increases at the lowest ph and vice versa, when reconstituted skim milk samples were heated at o C for times up to 60 minutes after adjusting to ph 6.5, 6.6 and 6.7. a similar trend was seen in the level of denatured whey proteins associating with the micelles. In contrast, the rate of denaturation of the major whey proteins, α-la and β-lg was weakly influenced by the ph of the milk upon heating within the narrow ph range in their research. The mechanism for the large change in the level of association of denatured whey proteins with the casein micelles in the small ph range has not been clarified yet. Some hypotheses were given related to the dissociation of κ-casein from the micelles at higher ph, or the change in reactivity or availability of sulfhydryl and disulphide groups with changes in ph With ph adjustment and sugar addition Upon replacing SMP by a mixture of casein micelles and whey proteins, not only the protein composition and the ph differed from the SMP-containing reconstituted milk, but also the lactose content. Therefore, in this series of experiments, β-d-lactose powder was added as well to obtain the same sugar content as in emulsions that contained 16.5 % SMP. The first row in figure 3.16 shows that lactose did not influence the heat coagulation of recombined evaporated milk emulsions after heating for 5 minutes at 121 o C without adjustment of ph. However, at ph 6.8 the emulsion containing 100 % casein showed coagulation after 5 minutes heating. Interestingly, after heating 2 minutes, not only the emulsion containing 100 % casein was denatured like in the absence of lactose, but the emulsions of lower casein and higher whey protein content were also coagulated. This means that lactose promoted the heat coagulation of casein with a stronger ph-dependence. This result was confirmed by the findings of Sweetsur and White (1975), who noted that carbohydrates addition reduces the heat stability of milk. Smits and van Brouwershaven (1980) proposed that at ph values below 6.8, the majority of whey protein complexes remain associated with the casein micelle surface whereas at higher ph values, these complexes remain associated in the serum. Upon heating at ph values above 6.8, not only do the whey protein aggregates remain in the serum but also micellar κ-casein dissociates in the serum. 48

56 Viscosity (mpa.s) ph = 7.1 ph = 6.8 ph = 6.6 Chapter 3: Results and discussion c/0w 90c/10w 80c/20w 70c/30w unheated 1 min 2 min 5 min Shear rate (s -1 ) Figure Effect of ph on the apparent viscosity (mpa.s)of emulsions with added lactose before and after heating at 121 o C from 1 up to 5 minutes. 49

57 Chapter 3: Results and discussion 3.5. Effect of sugar on heat stability Although sugars are extremely important biochemicals, it has been little interest for their effect on the heat stability of milk. It is known that SMP is composed of about 35 % protein with a casein:whey protein ratio of about 80:20, and contains about 50 % lactose. In order to evaluate the effect of sugar on the heat stability of recombined evaporated milk, the content of proteins was kept constant by mixing the same ratio of casein micellar powder and whey protein isolate powder, and 50 % of carbohydrates such as lactose, galactose and sucrose were added to replace SMP. Figure 3.17 shows that after 10 minutes of heating at 121 o C, not any significant change occurred in samples containing 16.5 % SMP as well as in the blends of whey protein isolate and micellar casein powder with different kinds of sugars. Interestingly, the recombined evaporated milk composed of 16.5 % SMP was the most susceptible to heat denaturation, while others still maintained their liquid structure. It is known that sucrose is a non-reducing disaccharide since neither of the rings is capable of opening, and hence after heating for 20 minutes the sample of sucrose did not change in color and structure. Galactose and lactose, on the other hand, are known as reducing sugars. Hence, the Maillard and caramelization reactions led to a brown color in these emulsions. On the other hand, the thermal stability of recombined evaporated milk was not affected by the replacement of sugars in SMP. Actually, we found out that the heat coagulation in this series of experiments was independent of sugars, whereas ph was involved and played the main role in the difference between emulsions prepared using SMP or using mixtures of sugar, whey protein isolate and micellar casein (table 3.3). Table 3.3. ph of recombined evaporated milk emulsions prepared using SMP or using mixtures of sugar, whey protein isolate and micellar casein. Milk with ph 16.5% SMP 6.6 Lactose 7.3 Galactose 7.2 Sucrose

58 Chapter 3: Results and discussion Other studies stated that sucrose inhibits sulfhydryl-disulfide exchange of α-la and β-lg during heat treatment (Yamauchi, 1961). Garrett (1988) also established that sugars inhibited the thermal coagulation of whey proteins when a whey protein solution was heated to 80 o C for 5 minutes. They also found that sucrose, lactose and glucose were the most effective at inhibiting whey protein coagulation in these conditions. In fact, the sugars compete with the proteins for the available water and as such can increase the protein denaturation temperature by some degrees in concentrated solutions. Moreover, increasing the sucrose concentration from 20 to 200 mg/ml could increase the amount of protein remaining in solution and decrease the amount of precipitate as measured by the increase in light scattering (Garrett et al., 1988). Although the denaturation of whey proteins is affected by the lactose concentrations at 80 o C, no change was found during denaturation at 133 o C (Hillier and Lyster, 1979). Thus, it can be concluded that heat treatment at 100 o C is sufficiently severe to eliminate the effect of sugars in promoting the conformational change. 51

59 Viscosity (mpa.s) Chapter 3: Results and discussion 1000 SMP Lactose Galactose Sucrose unheat 10 min 20 min Shear rate (s -1 ) Figure Effect of added sugars on the apparent viscosity (mpa.s) of emulsions containing 6.5 % Hozol oil and 5.87 % proteins (with a casein to whey protein ratio of 80:20) when heated at 121 o C for 0, 10 and 20 minutes. 52

60 Chapter 3: Results and discussion 3.6. Heat treatment before or after homogenization Heat treatment and homogenization are essential steps in processing of recombined evaporated milk. Their sequence maybe one of the factors affecting the heat stability because both unit operations influence the protein interactions and redistribution. Hence, the order of the 2 steps may alter the properties of the proteins and hence may further change the functional characteristics of recombined milk. Several studies have already published results about the influence of the sequence of processing steps on the changes of the properties of milk (Garcia-Risco et al., 2002; van Boekel and Walstra, 1989). In all previous experiments, the homogenization step was performed prior to the heat treatment of the emulsions. However, according to van Boekel and Walstra (1989), heat coagulation of milk is more detrimental if the milk is homogenized first and then heated, rather than being heated before homogenization. Therefore, in order to check the effect of the order of processing conditions on the heat stability, heat treatment was performed before homogenization in this experiment with the same composition and conditions as applied before. Hereby, a coarse emulsion was first obtained by premixing all ingredients in an Ultra- Turrax. The micellar casein, whey protein and β-d-lactose powders were mixed at different ratios (Table 3.4). Table 3.4. Composition (in g/100 ml) of recombined evaporated milk containing different Casein:Whey protein ratios, but the same total protein concentration as in 16.5 % SMP. All samples contained 6.5 g Hozol oil as well as 8.5 g β-d-lactose per 100 ml. Component (g) Sample Casein powder Whey protein isolate % aqueous NaN

61 Viscosity (mpa.s) Chapter 3: Results and discussion A B Figure Samples heated before homogenization (A) and samples heated after homogenization using a Microfluidizer operated at 4 bar driving air pressure (B). Upon homogenization of the samples of series A, homogenous milky samples were obtained as well C/0W 90C/10W 80C/20W 70C/30W 50C/50W 30C/70W 0C/100W A B Shear rate (s -1 ) Figure Apparent viscosity (mpa.s) of samples heated at 121 C for 20 min before homogenization (A) and after homogenization at a driving air pressure of 4 bar in a Microfluidizer (B). 54