Functionality has been defined as: "any property of a food or food ingredient, except its nutritional ones, that affects its utilization." For proteins then, there must be a large number of functions and functional properties listed in Table 1 are some of the most important ones to consider when discussing proteins.

Functions of proteins in foods

Protein Structure

A protein manifests functionality by interacting with other compnents within the food system. These interactions may involve solvent molecules, solute molecules, other protein molecules or substances that are dispersed in the solvent such as oil or air.

In order to describe the forces involved in these interactions, it is essential that the forces and energies involved in the achievement and maintenance of native protein structure be described. While a complete discussion of the forces involved is beyond the scope of this chapter, some observations on the nature of protein structure will be useful.

Proteins exist in the lowest kinetically attainable state of free energy. The free energy of the protein may not be the global minimum, but it will be the lowest that the protein can achieve in a reasonable period of time.

Protein structure is highly dependent upon the environment and the protein will assume different conformation as the environmental conditions change. Factors of importance include pH, temperature, dielectric constant, ionic strength and the presence of other molecules including air, fat, denaturants, etc.

One of the main ways that proteins lower their free energy involves the removal the removal of hydrophobic groups from the aqueous environment. This may provide the greatest single decrease in free energy of all the types of binding that occur within proteins. The strength of hydrophobic binding is, however, very sensitive to changes in temperature and dielectric constant and thus changes in these parameters can have large influences on protein structure.

Protein molecules may contain crosslinks of a non-covalent, eg. salt bridges or a covalent, eg. disulfide bonds, nature. These crosslinks lower the conformational entropy of the molecule which must be compensated for by a decrease in binding energy. The presence of crosslinks adds greatly to the stability of the native protein structure and makes the molecules resistant to unfolding or denaturation. For example, simple unfolding is inhibited sterically by the presence of crosslinking because portions of the molecules are held in place by the crosslinks. Denaturation is also less likely because one of the driving forces of denaturation, an increase in conformational entropy is greatly reduced. In a non-crosslinked protein, if unfolding to a random coil structure can be induced, there is a very large gain in the number of conformations the molecule can assume. This gain in conformational entropy is a large driving force for the maintenance of the denatured state when the denaturing agents are removed. In contrast, a highly crosslinked protein cannot assume the same degree of random conformations and thus the increase in entropy is much less. This helps explain why molecules that contain large numbers of disulfide bonds are often resistant to denaturation.

The structures attained by proteins are not rigid, but are very dynamic. There is rotational freedom about many of the bonds within the protein molecule and the entropy gain of this freedom lowers the total free energy of the native structure. There are also portions of the protein structure that are stabilized by rather weak secondary forces and these are often free to assume different conformations. These alternate conformations lead to structures of higher free energy and thus are not stable or long lived. A protein may be envisioned as a dynamic entity that is constantly sampling a variety of structures. These new structures are usually only slightly different from the native conformation and almost always lead to a situation where the free energy of this system increases. The increase in free energy causes the protein to spontaneously refold into the state of lowest free energy. Thus, the native structure of a protein is not the only structure it can assume, but rather the one of lowest free energy and hence of greatest probability. Slight changes in the environment can cause alternate structures to be of lowest free energy and thus lead to protein denaturation.

In order for a protein to exhibit functionality, it must interact with other components of the food system. These interactions may often require that the protein be free to either move throughout the system or to alter its structure in such a way to allow interactions with other components.

In some cases the simple presence of other molecules in the protein solution will allow interaction to occur, but more commonly, the interactions require an input of energy into the system to insure adequate mixing. This energy may alter the physical nature of the molecules being mixed, eg. decrease in average fat globule size and also alter the conformation of the protein molecule.

Flavor

In order for a compound to have flavor it must either interact with one of the four basic taste receptors on the tongue ( sweet, sour, bitter or salty ) or it must be volatile enough to interact with odor receptors in the olfactory organ. Most proteins can do neither of and so they have little direct impact on flavor. While free proteins as such have little affect, they may contribute to food flavor through one of the following:

bound molecules
absorption
break down products
chemical reactions.

The hydrophobic portions of proteins can interact with lipids. These interactions can be strong and it may be difficult to remove residual lipid from protein. If the lipid molecule has a flavor or if it reacts to produce flavor compounds, the protein may be considered to have contributed to to the flavor of the product. Usually flavors associ-ated with residual lipids are undesirable and result from lipid oxidation. This may be a relatively slow process and the off flavor may develop only upon storage of the product. Freshly processed soy flour have a bland flavor, but in a few months it often develops a beany taste. This has been attributed to the oxidation of protein bound lipid components. Most protein products may be susceptible to this type of off flavor development. Care is required during the processing of protein products for utilization as food ingredients to do all that is possible to prevent the development of off flavors due to the oxidation of residual lipid components.

Most flavor compounds are volatile and lipid soluble. As such, they may be absorbed to hydrophobic portions of protein molecules. Whether such absorption is beneficial or detrimental will depend on the types of compounds that are absorbed. In some formulated food products, proteins make important, positive contributions to flavor through interactions with small molecules.

The chemical or biological products of protein degradation may also contribute to the flavor of a food product. With certain exceptions like the controlled fermentation of soy beans to form soy sauce, such degradation usually has a negative affect on flavor quality. Proteolysis, for example, often results in a bitter flavor. This has been attributed to the formation of hydrophobic peptides. Some small hydrophobic peptides have been shown to be extremely bitter.

Further degradation can lead to decarboxylation of free amino acids with the formation of volatile amines. The products of lysine and or-nithine fermentation are called cadaverine and putrescine respectively. As their name suggests, these compounds have unpleasant odors. The production of these type of compounds is whey the fer-mentation of proteins is called putrefication. It should be obvious that these types of reactions must be prevented in food products.

The reactions of proteins and amino acids with reducing sugars can, on the other hand, lead to the formation of pleasant flavor com-pounds. The aroma of freshly baked bread being one example. The generation of these type of compounds requires considerable energy input and generally result from the thermal processing of foods. Normal protein isolation or production procedures would not be expected to result in the formation of these compounds.

Water Binding
The interactions of water with proteins are very important both to the structure of the proteins and to their behavior in food systems. Water molecules tend to associate with themselves through a network of hydrogen bonds. When solute molecules are placed in water, these molecules will be soluble if water solute interactions have a lower free energy than do the separate solute-solute and water-water interactions. The nature of these interactions are very complex and have been the subject of many reviews. While there is still much to be learned about the interactions of water both with itself and with solute molecules, the types of interactions that are important to protein structure and functionality can be described.

Proteins contain a number of amino acids that have side groups that contain electrical charges at certain pH values. The ion-dipole interactions between water and these charged groups are fairly strong with an energy of about 5 Kcal/mole. With model peptides, it has been shown that from four to seven water molecules can be associated with each residue of charged amino acid.

Proteins also contain amino acids that have polar side chains, but that do not have a charge. These polar molecules are dipoles and thus water can interact with them through dipole-dipole interactions. Because the molecules involved all contain hydrogen as part of the dipoles, the special class of dipole-dipole interactions known as hydrogen bonds can occur. These are typically stronger than other dipole-dipole interactions and can have energies of from 2 to 6 Kcal/mole. The nonionized polar amino acids in proteins typically have two water molecules strongly associated with them.

Nonpolar amino acids are not soluble in water and thus, the interaction of water with these molecules is minimal. Sometimes, how-ever, nonpolar groups are forced into water as a part of a specific protein structure. The intrusion of hydrophobic groups into the aqueous environment causes an ordering of the water molecules in their vicinity and thus a decrease in entropy. It has been estimated that the removal of a hydrophobic group from contact with water yields a reduction in the free energy of the protein of about 4 Kcal/mole. This is a strong driving force for the removal of these groups from the aqueous environment. Any water associated with these groups is highly ordered and has been termed hydrophobic hy-dration water.

Kinsella has listed the factors in Table 2 as being those that affect water binding by food proteins.

Table 2. Factors affecting water binding by food proteins

Amino acid composition
Protein conformation
Surface polarity/ hydrophobicity
Ionic concentration
Ion species
pH
Temperature

As noted earlier, some amino acids bind more water than do others and hence, proteins that contain large amounts of the charged amino acids will tend to bind large amounts of water. A number of schemes have been devised to predict the water binding of proteins based on amino acid composition. These generally work fairly well and demonstrate the importance of amino acid composition has on water binding.

The next three factors in Table 2 relate to how the amino acids are arranged in the final protein structure. If groups that are capable of forming hydrogen bonds are removed from contact with the water and allowed to form hydrogen bonds with themselves, the amount of water bound to the protein will be decreased On the other hand, if the protein molecule is unfolded, these groups may make better contact with the aqueous phase and water binding may actually increase with denaturation

Some proteins require the presence of trace amounts of electrolytes to be soluble. These ions interact with the charged groups on the protein surface and also with water present. This tends to increase the amount of water bound to the protein. At very high salt concentrations, the salt and protein may compete for the water present and protein water binding and solubility may be decreased. The extent of these effects depend to a large extent on the nature of the ions present. The size of the hydrated radius of the ion seems to be of prime importance for determining the dehydration effect of concentrated salts.

The effect of pH on the water binding of proteins can be manifested in many ways. As the pH of a protein solution changes, so does the number of charged groups on the molecule. As previously noted, the number of charged groups is strongly related to the amount of water associated with a protein. Changes in pH can also alter the conformation of the protein which may either expose or bury potential water binding sites.

Any discussion of water binding must include a definition of what is meant by bound. The types of bound water associated within pro-teins have been described by Chou and Morr and are listed in Table 3.

Table 3. Types of water associated with proteins.

Structural
Monolayer
Unfreezable
Hydrophobic Hydration
Capillary

In some proteins, a molecule of water may be intimately associated with the final structure of the molecule. In these cases, the water serves as a bridge to hydrogen bond charged groups within the molecule. These molecules of structural water do not behave like bulk phase water and are unavailable for chemical reactions. The water molecules that are bound through ion-dipole or dipole-dipole interactions are described as monolayer water, in which the molecules form a layer of tightly bound water around the protein molecule.

Hydrophobic groups that are exposed to the aqueous phase tend to cause an increase in the order of the water molecules near them. The nature of the water bound near these hydrophobic groups is not well defined, but it does not behave as normal bulk water.

All of the above types of water interact strongly with the protein molecule and exhibit properties that are different from those of free water. One of the most notable characteristics of bound water is that it does not freeze at normal temperatures and is thus termed unfreezable water. Many of the methods utilized to measure the amount of water bound to proteins are based on this phenomenon. This is also the type of water that can be predicted from a knowledge of the amino acid composition of proteins as previously mentioned.

Capillary water refers to water that is associated with proteins, but that freezes at normal temperatures and is free to act as a solvent for small molecules. This water behaves very much like bulk phase water, but is very difficult to remove from the protein mass. The amount of this type of water associated with a protein is very dependent upon the nature of the measurement being utilized. A good physical description of the forces that bind this water to proteins is not available, but at least some of it seems to be entrained in a three-dimensional network of protein molecules. The incorporation of large amounts of water into a protein structure can lead to the formation of a gel which is the next property to be discussed.

Viscosity

Viscosity in food is often considered with flow properties. Ideal solutions maintain a constant viscosity coefficient independent of shear stress or shear rate. Viscosity can be related to the volume occupied by solute compared to the volume occupied by solvent. As a protein occupies more volume in solution it has a greater chance of interacting with other molecules and for there to be a resistance to flow.

When most protein solutions are exposed to increasing shear the viscosity coefficient decreases. This behavior is called shear thin-ning and solutions that exhibit this behavior are called pseudoplastic. As shear is applied interactions between protein molecules can be weakened. If this results in the breaking of a network, shear thinning will occur and the apparent viscosity will decrease with increased shear. This leads to an alinement of molecules in the direction of shear which decreases resistance to flow. When the shear is removed it is possible for interactions to occur again. If these interactions lead to an increase in apparent viscosity, the solution is said to be thixotropic. A number of protein solutions exhibit pseudoplastic and thixotropic behaviors.

Factors that can cause proteins to unfold can increase their viscosity. Factors having an important effect on viscosity include:

pH
temperature
concentration
ionic strength.
Gelation

As was discussed in the chapter on protein denaturation, extremes of pH can cause proteins to unfold and to increase their water bind-ing. As the compact protein unravels its effective diameter increases and it occupies a larger percentage of the available volume. Interactions are more probable and viscosity increases. Near the isoelectric point proteins can form aggregates more easily than when they are highly charged. If these aggregates are poorly hydrated or precipitate, the volume of the protein in solution will de-crease and viscosity will be low.If the aggregates are highly hydrated and remain is suspension the viscosity may increase. Heating proteins can cause them to unfold. If this unfolding leads to increased interactions the viscosity will increase.

Gelation
Definition by Ziegler and Foegeding:

A gel is a continuous netwrok of macroscopic dimensions immersed in a liquid medium exhibiting no steady-state flow.

Protein gels can be formed by the addition of salts, the action of enzymes, changes in pH or by the application of heat. Whey protein gels are obtained by heating so this discussion will be limited to the mechanisms of heat induced protein gelation.

Ferry first suggested that the process of gelation was a two-stage one involving: an initial denaturation or unfolding of a protein molecule followed by subsequent aggregation. The steps involved in thermal gelation are summarized in Table 4.

Table 4. Stages in heat induced gelation

• Protein unfolding
• Water binding
• Protein-protein interactions
• Water immobilization

When a protein is heated, the bonds that maintain its secondary and tertiary structures are weakened and, at some temperature, broken. This breaking of non-covalent bonds with its resulting alteration of protein structure is termed denaturation . In the early stages of thermal denaturation, most protein molecules begin to unfold. This unfolding often leads to a slight increase in the amount of water tightly bound to the protein. If protein-protein interactions lead to the formation of a three-dimensional network capable of entraining water molecules, a gel can form.

If the network is too weak, the viscosity will increase, but fluid flow will be possible and a true gel will not form. If, on the other hand, the protein-protein interactions are too strong, the network will collapse and water will be expelled from the structure. A balance between the attractive forces necessary to form a network and the repulsive forces necessary to prevent its collapse is required for gel formation.

Consider a fibrous protein being held together by a series of ionic bonds. This could just as easily be a protein that is stabilized by hydrogen bonds such as the collagen triple helix or a globular protein whose structure is stabilized by a variety of non-covalent and possibly some covalent bonds. In any case, water does not interact with the interior of the protein because there are too many bonds present. If water were to attempt to interact at some point, not only would that bond have to be broken, but a number of neighboring bonds would have to be weakened. This is not possible at room temperature and the protein remains in the conformation of lowest free energy.

The application of heat will weaken these bonds and allow water to interact with the charged groups. At some temperature, the attractive forces will have been weakened enough to allow for extensive water-ion interactions. This causes an unfolding of the molecule and an increase in water binding. If a network can be imagined to extend into three-dimensions, we could envision the existence of "pockets" of bulk phase water that would be associated with the network. This water would freeze at normal temperatures and would be freely available to participate in chemical reactions.

Thus, protein-protein interactions are necessary for a network to form. The interactions could involve calcium or other ionic bridges, hydrophobic interactions, disulfide bonds or others. As long as there are enough of them to form a network and there are enough repulsive forces to prevent the network's collapse, a gel can be formed.

Many factors have been reported to have an effect on the formation and properties of protein gels. Some of the more important of these are presented in Table 5.

Table 5. Factors affecting gel formation.

• Temperature
• Protein concentration
• pH
• Salt concentration
• Calcium concentration
• Free sulfhydryl concentration

Temperature can affect the formation of protein gels in a variety of way. Heat can affect both the rate of denaturation and the rate of protein-protein interaction. If the rate of protein aggregation is rapid compared to the rate of protein unfolding, the gel structure will be adversely affected. If the rate or extent of aggregation is too great, a precipitate will form without appreciable water immobilization. A temperature must be selected for gel formation that balances the rate of protein unfolding with that of aggregation. Often times at the temperature selected, there is only minimal aggre-gation and the majority of the protein-protein interactions occur as the system is cooled.

The protein concentration determines both the likelihood of gel formation and also the characteristics of the gel that forms. Below a certain level that varies according to the protein utilized, gelation will not occur. When the level of protein is too low, a protein net-work is difficult to establish. Protein-protein interactions then tend to occur within molecules rather than between molecules and a gel framework cannot be established. As the protein content in-creases, the likelihood of intermolecular crosslinks increases and at some concentration, gelation can occur. Further increases in protein content change the strength and texture of the gel resulting in firmer gels as more water is tightly bond to protein molecules.

As discussed previously, pH can have a marked affect on the structure of proteins and the amount of water that can be bound to proteins. pH can also affect the extent of protein-protein interaction if they are of an ionic nature. In other cases, maintenance of proper pH can prevent the collapse of a gel network due to charge repulsion.

The denaturation of proteins is highly pH dependent and so the rate of protein unfolding can be influenced by the pH of heating. Proper pH adjustment may be necessary to achieve the proper balance be-tween the rate of denaturation and aggregation as well as the forces of attraction and repulsion between adjacent protein chains that are necessary to achieve a protein gel.

Salts in general can affect the structure of protein molecules as well as the nature of protein-water interactions. These affects markedly influence both the solubility of protein and their rate of thermal denaturation. As with heating and pH there is generally an optimal level of salt that favors gel formation.

The calcium ion concentration can have affects on gelation beyond what would be predicted due to changes in ionic strength or to its hydrated diameter. The calcium ion concentration can affect both the rate of whey protein denaturation and the solubility of the denatured protein molecules. Calcium has also been shown to be an effective protein crosslinking agent in casein systems as well as mediating the interaction between whey proteins and caseins. Calcium concentration has been demonstrated to be important in the gelation of a number of protein systems. Once the protein has been unfolded, a network capable of immobilizing water must be formed. Calcium is capable of forming crosslinks between adjacent protein chains and aids in the formation of such a network. The effects of calcium concentration on the strength and texture of gels produced by WPC have been studied by a number of workers. It has generally been noted that in samples with low calcium levels, the addition of calcium increases gel strength. WPC with low calcium contents were usually prepared from commercially available products by dialysis. Above 11 mM, calcium causes a decrease in gel strength. This has been attributed to protein aggregation occurring at high a rate which limits protein unfolding and network formation.

It has been reported that the calcium content of commercially available acid WPCs was in the range that inhibited gel formation. Similarly, it has been reported that the level of calcium in sweet WPC was also in a range that decreased gel strength. A number of workers have noted that the texture of gels involves more than just gel strength.

The major application of WPC in gel systems is to produce meat-like gels and in these applications the strength of the gel is the major factor producers consider. For the products available commercially, chelation of calcium will increase gel strength. The single most important factor manufacturers of WPC for utilization in gelled products can control is calcium content. Changes in processing conditions that reduce the calcium content of the finished WPC or the addition of materials that chelate calcium will make these products more marketable.

An effect of sulfhydryl content on the strength and texture of WPC gels has been demonstrated. As was the case with calcium, there is an optimal sulfhydryl concentration for gel strength. In WPC, b- lactoglobulin is the main source of free sulfhydryl groups and it would seem logical that the concentration of native b- lactoglobulin should be related to gel strength. It has been reported that the compressive strength, elastic modulus and impact strength of whey protein solutions increased with increasing b- lactoglobulin content. A correlation between the sulfhydryl content of WPC and the content of b-lactoglobulin has also been reported. The appearance and strength of the gels formed from whey protein concentrates is dependent on the pH of gelation. At low pH values, the gels are rather soft and opaque and can be described as coagula. At higher pH values the gels become more elastic, transparent and have greater gel strengths. The effects of sulfhydryl groups on the strength of WPC gels is strongly pH dependent. Below about 7, there is little effect of sulfhydryl content on gel strength. At higher pH values the effects of sulfhydryl groups becomes significant.

A relationship between protein hydrophobicity and the gelation characteristics of whey protein concentrates has been reported by a number of workers. As with calcium and sulfhydryl content, there appears to be an optimal level of hydrophobicity beyond which gel strength is weakened. In a study with commercially available whey protein concentrates the optimal level, as measured by alkane binding, was not reached reached and there was a direct positive correlation between gel strength and protein hydrophobicity. As is the case for calcium and sulfhydryl groups, a certain number of hydrophobic associations should provide sites for protein chain crosslinking and would be expected to increase gel strength. Too many such sites could lead to protein aggregation. Studies with commercially available WPC have demonstrated that increases in protein hydrophobicity will increase the strength of the gels formed. Only in a model system of b-lactoglobulin and added milk fat globule membrane has the level of hydrophobic groups been high enough to lower gel strength. It is difficult to reproducibly alter the hydrophobicity of protein solutions. It may be that a positive effect may be obtained by decreasing the lipid content of the WPC. This would free hydrophobic sites on the proteins and make them available to interact during gel formation.

Emulsification

Emulsions are thermodynamically unstable mixtures of immiscible liquids . If energy is applied the systems may be dispersed, but increased surface energy causes the phases to coalesce unless an energy barrier to coalescence is established. Emulsified droplets can be stabilized by the addition of molecules that are partially soluble in both phases. In foods a number of small emulsifier molecules can serve this function. Proteins capable of unfolding at the interface may also serve this function. Protein coats the lipid droplet and provides an energy barrier to particle association and phase separation.

The study of emulsions is complicated by the interactions that can occur when multiple components are present and by the fact that conditions that are important in dilute solutions where the systems are easier to study may not apply to conditions likely to be found in foods. Generally it is possible to explain much of how proteins function in emulsions from a knowledge of the forces that are operational during emulsion formation and a knowledge of protein structure.

Generation of an emulsion involves the mixing of two immiscible liquids. The reason that the liquids are immiscible can be related to their relative polarities. When the non-polar portions of a protein or any other molecule are exposed to the aqueous phase, they tend to spontaneously associate in a manner that minimizes contact with water. Measurements of the enthalpy of hydration of a number of non-polar molecules yield values that are similar and negative . This suggests that interaction between non-polar molecules and water should be favorable. When solubility data are examined, how-ever, it is found that non-polar molecules are only slightly soluble in water. Measurements of the free energy of transfer of nonpolar molecules from organic solvents to water give values that are posi-tive. The negative values for the enthalpies of hydration and the positive free energy of transfer to the aqueous phase suggest that an entropically driven aggregation of nonpolar molecules occurs in an attempt to minimize their contact with water. The reason for this is that the intrusion of a nonpolar molecule interferes with the nor-mal structure of water in such a way as to increase its order. When a liquid of low polarity such as fat is mixed with water there is a strong driving force to limit the contact between the two liquids. This happens when phase separation occurs. To increase the interfacial area and thus the energy of the system requires the input of work. If the liquids are dispersed through the application of work, the system attempts to achieve the conformation of lowest free energy. The total energy can be minimized if the area of contact between the two liquids is kept to a minimum. This can initially be achieved by the formation of spherical particles. Thus, when two immiscible liquids are forced into contact by the application of work, the result will be the formation of a number of spherical droplets within the dispersed phase. Larger spheres have a smaller ratio of surface to volume than do smaller spheres and hence a lower surface energy. If there is no energy barrier to prevent coalescence, the system will continue to lower its total energy content by the formation of larger droplets from smaller ones. Given enough time this leads to the situation of minimum contact and phase separation

The dispersed system can be stabilized against coalescence and phase separation if another component that is is partially soluble in both phases is added. Molecules that are composed of portions that are soluble in water and portions that are soluble in lipids can serve as emulsifiers. Phospholipids are a class of naturally occurring compounds that can serve this function. When mixed with lipid in an aqueous environment, the fatty acid portion of the phospholipid molecule is inserted into the oil phase, while the phosphate ester head group remains in contact with the aqueous phase. Thus, a portion of the molecule is in contact with the lipid phase while another portion of it is in contact with the aqueous phase. More impor-tantly the two immiscible phases are not in contact with each other and the total energy of the system is lower. The head portions of the phospholipid molecules contain like charges and tend to repel each other causing an energy barrier to coalescence and phase sepa-ration.

An emulsion formed from a mixture of oil, water and emulsifier is at a higher energy level than the deemulsified system. Thus, an emulsion is thermodynamically unstable and given enough time, breaks or separates. The goal of the food scientist is to make the energy of activation high enough to give the emulsion a reasonable lifetime.

In order to form an emulsion, energy must be provided in excess of the of that due to the creation of the new interfacial area of the emulsion. The size of the droplets and thus their interfacial energy, depends on the amount of work done on the system. As more work is applied, the droplet size becomes smaller. As soon as new interfa-cial area is created, the system attempts to reach a lower energy state by coalescence of fat globules. The rate of coalescence depends on the energy barrier and the rate of droplet collision. For uncoated fat globules the energy barrier to coalescence is so small that it can be ignored. In most lipid systems the density of the lipid phase is much less than the aqueous phase and the fat droplets tend to rise to the surface. This increases the rate of collisions and the rate of droplet coalescence and phase separation.

The presence of an emulsifier makes the situation more complex. The emulsifier has a portion of the molecule oriented away from the phase that it is dispersed in. In the case of phospholipid in water, a micelle forms with the fatty acid tail groups oriented away from the aqueous phase. When such a micelle approaches the lipid/water interface it tends to reorient. The charged groups on the phospholipid resists removal from the aqueous phase. If the micelle approaches a lipid droplet the structure reorients in an attempt to prevent the dehydration of the charged phospholipid head groups. This causes an exposure of the fatty acid tail portions of the molecules. As these come into contact with the lipid phase, the fatty acid portions of the molecules orient into the lipid, while the charged head groups remain in contact with the water. This results in the formation of a monolayer of phospholipid molecules at the surface of the droplet oriented such that their hydrophobic groups are inserted into the lipid phase and their charged head groups are in contact with the aqueous phase.. For lipid in water with no emul-sifier present, a rapid coalescence of fat globules occurs and in time phase separation follows. If emulsifier molecules are present, they diffuse to the fat lipid interface as coalescence is occurring.

If the newly created surface could be instantaneously coated with emulsifier molecules, the emulsion would consist of particles hav-ing the same size distribution as they did at the moment of homoge-nization. In real emulsions, the emulsifier molecules require a finite time to diffuse to the interface and to be absorbed in order to provide a barrier to coalescence. The rate of droplet coating by phospholipid is a complex function of the rate of fat droplet coalescence, the rate at which the phospholipid molecules reach the lipid surface and the rate at which the micelles are able to reorient. Small emulsifier molecules are able to diffuse rapidly and the amount of reorientation required to interact in the surface is small. In general, the use of small emulsifier molecules results in a relatively narrow particle size distribution and in the formation of emulsions with relatively small fat globules.

Proteins are often included in emulsions to aid in their formation and to increase their stability. Proteins are much larger and more complex that are simple emulsifier molecules and the formation of a protein stabilized emulsion requires that the protein molecule must first reach the water/ lipid interface and then unfold so that its hydrophobic groups can contact the lipid phase. To illustrate the forces involved, the situation of a protein molecule approaching a static water/lipid interface will first be considered. In native proteins most of the nonpolar amino acid side chains are located in the interior of the molecules. Proteins have charged groups at the surface of the molecule and in contact with water molecules. The favorable interaction of water with surface charge lowers the total energy of the protein molecule. In some respects the protein may be envisioned as resembling the micelles of phospholipid in the previ-ous example. The hydrophobic groups are removed from contact with the aqueous phase while charged groups maximize solvent contacts.

As a protein molecule approachs the interface, there is less oppor-tunity for the charged groups to interact with solvent. In the extreme case, charged groups are removed from the aqueous phase and enter the lipid phase. This is energetically unfavorable and these groups are repelled from the interfacial area. If the groups nearing the interface are in a region of the protein molecule that contains some flexibility, the molecule may begin to unfold. This unfolding causes the exposure of hydrophobic groups to the surface. If these groups are exposed to the the aqueous environment, there is an increase in total energy and random fluctuations in protein structure cause these groups to return to the interior of the molecule. If the exposure occurs at an interface, the state of lowest free energy depends on the nature of the interface. In the case of a protein un-folding near lipid, the hydrophobic groups are inserted into the lipid phase. This insertion has a very low energy of activation and pro-ceeds spontaneously For most proteins studied, the size of the hydrophobic region inserted is about 6 to 8 amino acid residues. The enthalpy for this step is positive so that the driving force must be an increase in the entropy of the system. This increase in entropy has two components, one due to the conformational entropy of the protein and one due to the structure of water near hydrophobic groups. There is an increase in the conformational entropy of the protein as the the hydrophobic groups are removed from the interior of the molecule and placed into another nonpolar environment. The original protein had a limited number of ways of arranging its components to attain a low energy state. The partially unfolded molecule has many ways of inserting a hydrophobic group into a non-polar environment and once there the group can assume more conformations than before. The solvent molecules at the interface are arranged in highly ordered structures as was previously noted. The approach of the protein with the insertion of hydrophobic groups into the oil phase will in essence coat the nonpolar material and will al-low for the release of solvent from the surface. The release of this water is responsible for a significant increase in the entropy of the system.

While the original insertion of a hydrophobic group proceeds sponta-neously with a small energy of activation, the reaction is not readily reversible . In time other sections of the protein molecule approach the surface and if these occur in flexible portions of the protein they too may be inserted into the lipid phase. As this continues the protein will unfold at the interface.

Proteins that become attached by more than one hydrophobic group desorb very slowly from the surface, if at all. Langmuir and Schaeffer calculated that if absorption were completely reversible and the Gibb's absorption equation applied that changes in surface pressure of the magnitude they observed in ovalbumin stabilized emulsions there should result in essentially complete desorption of protein form the interface. This does not occur for protein stabilized emulsions suggesting that a significant energy barrier to protein desorption exists. Removal of hydrophobic groups from the lipid ex-poses lipid to the aqueous phase as well as the hydrophobic groups that are being removed. Even if the removed hydrophobic groups could be buried in the protein interior, the protein would remain at-tached to the fat globule at other points and reattachment would be likely. If other hydrophobic molecules are available to cover the exposed lipid area, desorption is easier to achieve. It has been shown, for instance, that gelatin molecules can be replaced by more hydrophobic casein molecules from the water/ lipid interface.

Once a layer of protein has been adsorbed additional protein cannot be added in the same manner because an energy barrier to absorption exists. In order for more protein to be absorbed, the protein already at the surface must be compressed to make room. The amount of compression that is possible depends on the rigidity of the protein and also on the amount of residual charge near the surface. At some level of compression, the absorption of more protein will require more energy than can be gained by the insertion of hydrophobic groups into the lipid layer. Further interaction involves the interaction of protein molecules in the bulk phase with those already ab-sorbed to the lipid and the formation of multilayers.

When proteins are used for generation of emulsions the system be-comes highly complex. Some form of shear is generally be responsible for the creation of new surface area. The high energy state is relieved by rapid coalescence of fat globules. For prevention of coalescence protein molecules need to diffuse to the fat/ water in-terface and then unfold and coat the surface. When enough of the new surface is covered, coalescence ceases. With proteins the rate of diffusion to the interface is a significant variable in the amount of protein that absorbs to the interface during emulsion formation. Anything that tends to decrease the rate of diffusion of the protein molecules decreases the protein load.

Once a protein molecule reaches the surface it must be able to unfold enough to expose hydrophobic groups if it is to function as an emulsifier. Molecules like the various caseins are extremely flexible and contain little secondary structure. Caseins are excellent emulsifiers because of their ability to easily unfold at interfaces. Molecule that contain crosslinks such as disulfide bonds are more rigid and less able to unfold. Such molecules are less effective in emulsion formation. Reduction of disulfide bonds enhances the emulsifying ability of some proteins as long as the molecules do not unfold to the point that there is a large increase in viscosity. The content of disulfide bonds has been related to the emulsion capacity of complex mixtures of proteins such as whey protein concentrates. Small highly crosslinked protein molecules tend to perform poorly as emulsifiers.

Once proteins begins to unfold, there must be hydrophobic groups present to insert into the nonpolar phase. In theory, a measure of the relative hydrophobicity of a protein should be related to its ability to function as an emulsifying agent. In practice, relative hydrophobicity measurements have been difficult to obtain. The early methods generally assigned some relative value to each amino acid and then the value for the protein was calculated from its composition. These procedures have rarely correlated well with func-tionality because they measure the total potential hydrophobicity of the protein rather than those hydrophobic groups which can actually reach the surface upon unfolding. Recently a number of procedures have been developed which measure what is termed the effective hydrophobicity of proteins. Generally this means obtaining a quantitative measure of those hydrophobic groups that are capable of binding to a selected probe molecule. The groups that are deeply buried in a portion of the protein that does not unfold are not mea-sured. The groups that interact with the probes are generally acces-sible to the surface and are the groups capable of interacting in emulsions or foams. Both surface hydrophobicity as determined by the binding of cis paranaric acid and effective hydrophobicity as determined by alkane binding have been related to the ability of various proteins to form stable emulsions.

The distribution of hydrophobic groups is also important. In proteins like b-lactoglobulin, the hydrophobic groups are rather evenly distributed throughout the molecule. There are no large portions of the molecule where hydrophobic amino acids are grouped, nor are there large sections of the molecule that do not contain charged amino acids. This makes it difficult to find portions of the molecule that are sufficiently hydrophobic or find residues that do not also contain amino acids with charged groups that would resist their removal from the aqueous phase. In molecules like b casein there are large sections of the protein that contain hydrophobic amino acids without the presence of charged groups. The molecule has such an uneven distribution of charge and hydrophobic groups that is amphipathic. It is easy to find portions of this molecule that contain at least six nonpolar amino acids and no charged groups.

Once formed an emulsion can undergo a number of changes. The most striking change would be phase inversion. In order for phase inversion to occur, the surface of a number of fat globules have to be exposed and allowed to coalesce. With protein stabilized emulsions phase inversion is generally not a problem because when fat globules near each other, the proteins usually provide an effective barrier to coalescence. The removal of protein from the surface of a fat globule is energetically unfavorable and does not occur at any appreciable rate. In order to destabilize the emulsion large changes in the structure of water in the system, large inputs of energy or both are required. In food products, fluctuations in temperature are a common cause of emulsion destabilization. As the temperature is lowered, water attains more and more structure. As the water becomes more ordered, there is less of an energy difference between hydrophobic groups exposed to the aqueous phase and those buried in the oil phase. Low temperature alone does not usually cause an emulsion to break, but it can be the deciding factor in the stability of an otherwise poorly emulsified system.

The largest temperature induced changes to emulsions occur upon freezing and subsequent thawing. Not only is the energy difference between the associated and free state minimized by the low temper-ature, but the formation of ice crystals can cause physical damage to the emulsion. When the system is thawed, coalescence occurs if the physical damage has been extensive. One of the best ways to minimize this type of damage is to add substances that will modify the size and extent of water crystal formation .

A more common defect in food emulsions results from the phenomenon known as creaming. If density differences between the dispersed and continuous phase exist, particles of the dispersed phase either sediment or rise depending on the relative densities. In most emulsions the dispersed phase is less dense than the continuous phase and creaming occurs. Given enough time a depletion of lipid from the bulk aqueous phase occurs with the formation of a compact cream layer containing the majority of the lipid. The rate of creaming is given by Stoke's law:

Stoke's law:

Where v equals the velocity of the fat globule, r is the radius of the fat globule, g is the force of gravity, Dp is the density difference between the two phases and m is the viscosity of the continuous phase. In theory for an emulsion to have an extended shelf life either the density of the fat globules must be made identical to that of the continuous phase or the viscosity must be high enough so that the yield value is greater than the acceleration due to buoyant differences. Few emulsions can be made totally stable to creaming without the formation of some sort of matrix. Homogenized milk, for example, can be shown to have approximately 1% of the fat globules entering the cream layer for each day of storage. Thus a product with a fat globule size distribution similar to that found in homogenized milk would have 90% of its fat globules in the cream layer after only 3 months of storage. For products with shelf lives approaching 24 months, such as infant formulas, even small differences in density between the dispersed and continuous phases results in the formation of a cream layer during the useful shelf life of the product unless additional measures are taken to increase stability.

It has been shown that the addition of polysaccharide stabilizers to emulsions has little effect on the stability of the systems unless they increased the viscosity to the point of imparting a yield value. It has been hypothesized that carrageenan stabilization of heated milk products is the result of the formation of a network that physically prevents the fat globules from coming into contact. This has been likened to the formation of a loose gel matrix within the fluid phase. In fluid emulsions the network must be firm enough to retard or prevent fat globule coalescence, but must be weak enough to allow fluid flow when the product is poured. It has been calculated that a yield stress of greater than 0.1 Pa would be enough to prevent creaming. If the yield stress were less than about 10 Pa the gel would be readily reversible and and would be fluid when poured. Even if there is not a yield value of more than 0.1 PA, most emulsions are sufficiently pseudoplastic to exhibit higher than expected viscosities at very low shear rates and thus creaming is often slower that predicted. Thus, while Stoke's law is important in predicting the rate of emulsion creaming, for most products with any appreciable shelf life, other factors, especially viscosity, pseudoplasticity and yield stress, must also be considered.

The amount of damage done to a product by the formation of a cream layer depends on the product type and the tenacity of the formed layer. If the layer can be readily dispersed by shaking, little damage ensues. If on the other hand, the proximity of the particles in the cream layer leads to coalescence or the layer is some how crosslinked to an extent that prevents its from being readily redispersed, considerable economic loss can occur.

The determination of meaningful emulsion data with complex food products is difficult. Much of the experimental work with model systems has been done in very dilute solutions. The surface pressure or interfacial tension is often the quantity measured. With a food product the relevant information is concerned with the question: How much lipid can be emulsified and how long will it be stable to coalescence and / or creaming? The situation in food products is also complicated by the presence of other surface active molecules in addition to the proteins present.

Three main types of test have been devised to give an indication of the efficiency of proteins to serve as emulsifiers in food products. A number of tests measure emulsion capacity. These generally involve adding lipid to an aqueous solution of the protein to be tested. Addition of lipid is continued until phase inversion occurs. The test measures the capacity of the protein to emulsify fat at very high lipid to protein ratios. Values for emulsion capacity are commonly in the range of a few hundred milliliters of oil emulsi-fied per gram of protein. The test does not measure the stability of the emulsion formed. While these values can predict which protein of a group will emulsify the most fat or what conditions of pH, salt content, etc. lead to maximal emulsion capacity, the conditions do not usually resemble those found in food systems. Extrapolation of these values to determine the stability of a food emulsion with storage are difficult at best.

Another means of estimating emulsion stability is to form an emulsion under conditions that resemble those in the product. The emulsion is then allowed to separate either under the influence of gravity or after exposure to a centrifugal field. The change in lipid distribution throughout the sample with time can be measured and the phase separation with time noted. Within a centrifugal field the fat globules are compacted into a cream layer and an aqueous layer devoid of fat is formed. The ratio of either the cream layer or the aqueous layer formed to the volume of the initial emulsion is often utilized as an indicator of emulsion stability. The emulsion volume index is an example of such a test. Generally forces of many hundreds or thousands times that of gravity are employed to obtain separation in fairly short periods of time. These high forces tend to distort fat globules and to overcome forces of repulsion that would probably still be operative in a product stored at 1 g for an extended period of time. Studies that have measured differences in distribution of lipid with time in emulsions have been generally limited to times in the order of hours.

The size distribution of the particles in an emulsion can also be utilized as an indicator of the effectiveness of the emulsifier. In general the the more efficient the homogenization, the smaller the particles. Determination of the size distribution can be tedious and may involve the introduction of artifacts due to sample preparation. This is especially true when electron microscopy is used to measure particle diameters. A method based on light scattering has become very popular recently. This method is easy and requires only small amounts of samples. The procedure as described makes a number of assumptions that may not be valid for a number of samples.

When comparing results in the literature it is important that the method of evaluation be considered since the different methods of emulsion evaluation may be more important than the variables be-ing examined to explain differences in results obtained by workers. Recently it has been reported that none of the above methods gave results that were able to predict the shelf life of either a coffee whitener or a salad dressing. A method has been reported which may be applicable for both of these food systems. The original formulations were modified so that a test product could be produced that was not as stable as the commercial product, but for which the stability was strongly correlated with the commercial product. Such an approach is difficult and a new test formulation must be applied for each commercial product investigated. If properly done, however, a system can be developed that is capable of predicting in a few weeks the shelf life of a product that is measured in months or years.

Foaming

The formation of a foam is analogous to the formation of an emulsion. The forces involved when air is incorporated into the aqueous phase are similar to those involved in the mixing of oil and water. In the case of a foam, water molecules surround air droplets. Air is essentially non-polar and an ordering of the water molecules adjacent to the air cells occurs. This results in a high surface tension and a high surface energy. Emulsifiers work in an air-water mixture much as they do in an oil-water system to lower the interfacial energy and to provide a kinetic barrier to bubble coalescence. A protein that is utilized to form a stable foam will require many of the same properties as are required to form an emulsion. The protein must be able to rapidly diffuse to the interface and unfold in such a manner as to lower the interfacial tension between the air and water phase. Once a foam is formed, there are three main forces that can lead to its decay. The forces are listed in Table 6.

Table 6. Factors that affect the stability of foams.

A foam will initially have a number of small air cells separated by an interfacial layer and by areas of water. As the foam ages, gravitational forces will cause water to drain and the air cells will come closer together. A phenomenon called capillary drainage works to further thin the walls separating the air cells.

Even though extensive drainage occurs, there is still an interfacial layer between the air cells that should be capable of preventing air cell coalescence. Mechanical disturbances, such as vibration, can cause the cells to collide and rupture. When this occurs, the smaller cells will give their contents to the larger cells because the pres-sure is lower in the larger ones.

If a foam were examined within one minute of whipping, a number of large and small air cells would be apparent. In some cases, the lamella would have ruptured and small air cells would be seen merging with larger ones. Also evident would be areas rich in bulk phase water as well as areas where the air cells were coming closer to each other and capillaries were forming.

A few minutes later, it would be apparent that the average size of the air cells had increased and that much of the free water had been lost through gravitational drainage. After an additional three min-utes, more gravitational drainage would have occurred and many distinct capillaries would be seen. As these capillaries form, they tend to become thinner due to capillary drainage. The amount of free water present would become so small that gravitational drainage would not occur to as great an extent. A calculation of the pressure in the areas between the air cells would show that the pressure is lowest in the region where three capillaries meet to form the so-called Plateau borders. The water tends to migrate from the high pressure areas to those of lower pressure causing a further thinning of the capillaries. After an additional time, almost all the non-capillary water would be removed and the air cells that remained would be fairly stable. Rupture would occur primarily due to disturbances in the air cells due to vibration. Some evaporative loss of water must also occur which would also weaken the foam. After an extended period of time, some very stable small cells would be seen as well as a number of very large.ones. Further losses of structure would be almost all due to mechanical disturbances. While the reasons for foam collapse have been discussed, other forces must be present which stabilize the film from collapse for fairly long periods of time. The main stabilizing forces are listed in Table 6.

If the protein at the air-water interface binds water tightly and is rigid, there will be an increased surface viscosity that will tend to reduce the rate of water drainage. If the viscosity is high enough, all flow may stop and the film will be stabilized. Thus, proteins that are rigid and have a high surface viscosity, like lysozyme, do not form high overrun foams; but the foams once formed, are stable. Modifications of a protein that tend to increase its viscosity at a surface, will increase the stability of foams formed by the protein.

The Gibbs-Marangoni effect relates to the increase in surface ten-sion that occurs in areas where the drainage of water has also re-moved some of the interfacial material. The increase in surface tension, as well as the increase in concentration of the emulsifier molecules in areas adjacent to the location of water drainage, will cause a diffusion of emulsifier back to this location. The emulsifier molecules will carry water with them and the thickness of the film will be restored. This self "healing" of films is very important to their stability.

Foams may also gain stability due to the electrical properties of the protein or emulsifier molecules. As the lamella thin and the emulsifiers move closer to each other, they may either be attracted due to the Van der Waal's forces or be repelled due to charge repulsion. In films that drain to this stage, residual charge on the emulsifier molecules can be very important to film stability. Too much charge, however, should be avoided as this can lead to charge repulsion dur-ing the formation of the film and to incomplete coverage by the protein molecules. Also there is an energy barrier to the removal of charged groups from the aqueous phase and to their insertion into air cells. Most studies have shown that foaming is maximal at pH values near to the isoelectric point of soluble proteins involved. At these pH values, the number of charges to be removed from the aqueous phase is minimal and the surface viscosity of the proteins then is maximal. Thus, to function well in a foam formation, protein molecules must be able to unfold at an air-water interface and to spread rapidly to cover the entire interfacial area. They must then possess either enough surface viscosity or charge or both to prevent drainage.

Fiber Spinning

Most proteins assume a three dimensional structure that is not too far from that of a sphere. Most proteins are also soluble to a greater or lesser degree. For most functions, protein solubility is required. There are certain functions, however, where solubility would be a liability. It would not be good to have skin or muscle that was soluble. Often proteins assume a specialized structure to insure insolubility.

The major protein in muscle is myosin. This protein has a soluble globular portion attached to a long thin strand of tightly wound a helix. The helix is so tightly wound that water can not interact with the amino acids and the protein is insoluble. In muscle, a number of these proteins are wound around each other to form a fibrous structure. The texture of meat is directly related to the content of these fibers.

There are food application where it would be desirable to utilize a less expensive protein source to simulate the texture of meat. In these cases, the structure of the protein must be made to resemble that of muscle in order to attain the proper texture. The most commonly utilized method to simulate muscle fibers is called fiber spinning.

If a sample of globular protein is exposed to high pH values, the proteins will obtain a high net negative charge. If enough charge is present, adjacent portions of the protein will start to repel each other due to the high charge density. When this occurs, the protein begins to unfold. If the pH is high than the pK of Cysteine, disulfide bonds will be broken and the protein will be able to further unfold.

In practice 20 to 50% suspensions are made with soy protein concentrates, although other protein sources would work too. The pH is adjusted to about 10. This is high enough to unfold the protein and the viscosity of the suspension will increase rapidly as the proteins begin to unfold. Prolonged exposure to high pH values should be avoided to minimize the loss of sulfur containing amino acids and to avoid the formation of potentially toxic degradation products.

The proteins will repel each other in solution. They are then passed through a die with openings of from 0.005 to 0.015 mm in diameter. Pressure causes the strands to align as they pass through the narrow opening. The proteins are spun into a solution that contains calcium ions and is at a low pH. The calcium forms crosslinks between the negatively charged groups. The low pH neutralizes the high negative charge. If properly done, the final pH will be near the isoelectric pH of the protein. At this value, the number of positive and negative groups is equal and attraction between adjacent portions of the molecule is maximal. When the protein unfolded, hydrophobic groups were exposed to the aqueous environment. The high charge repulsion prevented protein aggregation. At the isoelectric point, these groups will be removed from contact with water. In the unfolded protein, this may be best achieved by the formation of intramolecu-lar crosslinks. The net result is the formation of asymmetric fibers that are insol-uble at neutral pH values. These fibers are some what similar to those found in meat and the texture of products containing spun fibers resembles meat.

Extrusion

Method of protein modification - simpler than spinning.
Does not give well defined fibers, but rather, fibrous particles
Goal is to achieve mouthfeel similar to meet.
Can use either defatted or full fat meals.

Extrusion

Protein is dispersed at high temperature and pressure.
Extruded from high pressure to atmospheric.
Water flashes off and product swells creating large voids.
Often other proteins utilized to give better texture.

Dough formation

Dough can be defined as an extensible, viscoelastic protein network formed upon the mixing of an appropriate amount of water to cereal proteins, eg. wheat, rye and barley. The proteins involved in dough formation are gliaden and glutenin.

The gliadins are a mixture of proteins that are soluble in 70% alco-hol. They range in molecular weights of from about 25,000 to 100,00. They are usually associated with small amounts of glycol-ipid. The molecules are held together with disulfide linkages and are fairly extensible, but are not very elastic.

The glutenins average about 100,00 molecular weight. They are elastic molecules with low extensibilities. Glutenins are soluble in solutions of dilute acid or alkali and are associated with non-polar lipids. These molecules contain both inter and intra molecular disulfide bonds.

Glutenin and gliadin have very few charged amino acids and thus are not very water soluble. They do contain large amounts of glutamine ( around 35% ), however, and thus are excellent at immobilizing water. They also contain a large number of hydrophobic groups that are buried in their interiors.

When glutenin and gliadin are suspended together in an aqueous medium, little occurs. With the input of mechanical energy, both molecules start to unfold. This unfolding exposes hydrophobic groups which cause the molecules to aggregate. Also during the mixing process the glutenin and gliadin are covalently crosslinked through the formation of inter molecular disulfide bonds. The crosslinked pro-tein is known as gluten. If enough crosslinking occurs, a viscoelastic structure results. This structure is able to expand when gas is produced either through the metabolism of yeast or through the expansion of water vapor during heating. As the temperature increases, the proteins begin to denature. The denaturation of gluten along with gelationization of starch gives structure to products like bread.

If too much mixing occurs, there will be too many disulfide bonds formed and the gluten will be unable to expand properly. This results in bread of low loaf volume that is generally tough. The redox potential is important to optimal gluten formation. A ratio of about 15 disulfides for each free sulfhydryl is desirable. If there are too many free sulfhydryls, it will be difficult for enough intermolecular disulfide bonds to form. Too many disulfides if present before mixing will make it difficult for disulfide interchange to occur. Too much oxidation after mixing will result in excessive crosslinking and low loaf volume.

Functional requirements of food protein ingredients.

Food Protein requirements for application in different food products.

Functional characteristics of some common food proteins.