Table

Influence of Droplet Size on the Time It Takes for the Flavor to Diffuse Out of the Droplets

is shown in Table 9.2. Flavor release is extremely rapid in emulsions that contain droplets less than about 10 |im, but becomes appreciably slower at large sizes. It therefore seems that the movement of flavor molecules from the droplets to the aqueous phase will not be the rate-limiting step in flavor release provided the emulsion is well-agitated and the droplets are relatively small. Instead, it is more likely that the breakdown of the structure and the diffusion of the molecules through the aqueous phase to the receptors are rate limiting. If the droplets are coated by interfacial membranes, the release rate may be slowed considerably and would be dependent on the type of emulsifier present, although little work has been carried out in this area (Harvey et al. 1995).

9.2.2.2. Release of Volatile Compounds (Aroma)

The release of volatile compounds from a food involves the mass transfer of the compounds through the emulsion and into the vapor phase (Overbosch et al. 1991). We are therefore interested in the change in the concentration of the flavor in the vapor phase with time.

Flavor Release from Homogeneous Liquids. The release rate of volatile flavors from a homogeneous liquid (e.g., oil or water) has been modeled by assuming that static diffusion conditions apply (Overbosch et al. 1991). A finite thickness of the liquid containing the flavor is assumed to be in contact with an infinite gas phase. The flavor moves from the liquid into the gas phase in order to try and establish the equilibrium partition coefficient (which can never be achieved because of the infinite extent of the gas phase). The flux of the flavor across the boundary separating the liquid and the gas and the change in the flavor concentration in the liquid with time are given by:

where

J = Fc0 V Dl / nt M(t) = 2Fcl^jDLt / n f = KGL4DG / DL

Here, DL and DG are the translational diffusion coefficients of the flavor in the liquid and the gas phase, respectively; tis the time; KGL (= cG/cL) is the equilibrium partition coefficient between the gas and liquid; and c0 is the initial concentration of flavor in the liquid. The factor F is the driving force for the diffusion process: its value varies from 0 (low flavor release rate) to 1 (high flavor release rate). The value of F is determined principally by the equilibrium partition coefficient of the flavor between the gas and the liquid (KGL), because there are much larger variations in this parameter than in diffusion coefficients for different flavors.

These equations indicate that the rate of flavor release from a good solvent (KGL low) is much slower than the rate from a poor solvent (KGL high). Thus a nonpolar flavor will be released more slowly from a nonpolar solvent than a polar solvent and vice versa. This accounts for the fact that the ranking of the release rates of flavors from water is opposite to that from oil (Overbosch et al. 1991). It also indicates that anything which decreases the diffusion coefficient of the flavor molecules in the liquid will decrease the release rate (e.g., increasing viscosity or molecular size).

Flavor release does not occur under static conditions, and therefore the above model has limited applicability in practice. In reality, there will be a flow of gas over the food in the mouth due to respiration and swallowing (Thomson 1986). When the flavor is constantly swept away from the surface of the food, the concentration gradient is increased, and so the rate of flavor loss is more rapid than under static conditions. Overbosch et al. (1991) developed a model to take this convective diffusion process into account. A numerical solution of this model indicated that the release rate due to convective diffusion is much greater than that due to static diffusion when KGL is small, because the rate-limiting step is the movement of the flavor compound from the liquid surface into the gas. At large KGL values, the difference between the two theories is much smaller because the rate-limiting step is the diffusion of the molecules through the liquid, rather than from the surface into the gas. Under convective conditions, the rate of flavor release therefore depends on the equilibrium partition coefficient of the flavor, as well as the flow rate of the gas through the mouth and into the nose.

Influence of Ingredient Interactions. A number of ingredients commonly found in food emulsions are capable of decreasing the rate of flavor release because of their ability to either bind flavors or retard their mass transfer (e.g., proteins, carbohydrates, and surfactant micelles).

The effect of flavor binding on the rate of flavor release can be accounted for using the same approach as for static diffusion in homogeneous liquids (i.e., Equations 9.20 to 9.22) (Overbosch et al. 1991). For reversible binding, DL is replaced by an effective diffusion coefficient DL = DL/(K* +1), and KGL is replaced by an effective partition coefficient KeGL = KglI(K* +1), where K* is the binding coefficient (K* = cLBlcLF). The release rate is reduced because DL and KL are smaller than DL and KGL, although the total amount of flavor released when the process is allowed to go to completion is unchanged. For irreversible binding, c0 is replaced by cLF in Equations 9.20 to 9.22 because the rest of the flavor is "lost." In addition, the total amount of flavor released when the process is allowed to go to completion is reduced by a factor cLf F/c0.

The release rate may also be reduced because of the ability of certain food ingredients to retard the movement of flavor molecules to the surface of the liquid, which may be due to an enhanced viscosity or due to structural hindrance (Kokini 1987). The diffusion coefficient of a molecule is inversely proportional to the viscosity of the surrounding liquid, and so increasing the viscosity of the liquid will decrease the rate of flavor release because the movement of the flavor molecules is reduced. The presence of a network of aggregated biopolymer molecules may provide a physical barrier through which the molecules cannot directly pass. Instead, they may have to take a tortuous path through the network, which increases the time taken for them to reach the surface. If the flavor molecules are associated with surfactant micelles, their release rate will depend on the diffusion coefficient of the micelles, as well as the kinetics of micelle breakdown (Section 4.5).

Highly volatile flavors (high KGL) are most affected by viscosity or structural hindrance effects because the rate-limiting step in their release from a food is the movement through the liquid rather than the movement away from the liquid surface. On other hand, low-volatility flavors (low KGL) are affected less, because the rate-limiting step in their release is the movement away from the liquid surface rather than through the liquid (Roberts et al. 1996).

A great deal of research has been carried out to establish the relative importance of binding and retarded mass transfer mechanisms. Many experimental studies have shown that increasing the biopolymer concentration decreases the rate of flavor release, but have been unable to establish the relative importance of the two mechanisms (Hau et al. 1996, Guichard 1996, Roberts et al. 1996). The importance of rheology has been demonstrated by studies which have shown that the intensity of flavors decreases as the viscosity of a biopolymer solution or the strength of a biopolymer gel increases (Baines and Morris 1989, Carr et al. 1996). On the other hand, solutions with the same viscosity often have different flavor release rates, which may be because they have different microstructures or because the flavors bind to them differently (Guichard 1996). It is clear that more systematic research is needed to establish the role of biopolymers and other ingredients in retarding flavor release. The influence of ingredient interactions on release rates has important consequences for the formulation of many food products. For example, it may be necessary to incorporate more flavor into a food to achieve the same flavor intensity when the biopolymer concentration is increased.

Flavor Release from Emulsions. Overbosch et al. (1991) used the same mathematical approach as for homogeneous liquids to describe the rate of flavor release from emulsions. For static diffusion conditions, the flux and time dependence of the mass of flavor in an emulsion are given by:

where

These equations provide some useful insights into the major factors which determine flavor release in emulsions. They indicate that the release rate depends on the diffusion and partition coefficients of the flavor in the oil, water, and gas phases. The larger the value of KGE or DE, the more rapid the flavor release. Similar conclusions were recently drawn by Harrison et al. (1997) using another mathematical model based on penetration theory.

The above equations predict that the release rate from an oil-in-water emulsion is the same as that from a water-in-oil emulsion with the same composition, because they are symmetrical with respect to the physical properties of the two phases (Overbosch et al. 1991). This will only be true if the flavor in the droplets and continuous phase is in equilibrium. Some studies have indicated that the taste perception of oil-in-water and water-in-oil emulsions of the same composition is approximately the same (Barylko-Pikielna et al. 1994, Brossard et al. 1996). Nevertheless, other studies have shown that the release rate is faster from oil-in-water emulsions than from water-in-oil emulsions, which suggests that the emulsion cannot be simply treated as a homogeneous liquid with "averaged" properties (Overbosch et al. 1991, Bakker and Mela 1996).

The above equations are most likely to be suitable for describing systems in which the rate-limiting step is the movement of the flavor molecules from the emulsion surface to the gas phase (i.e., when KGE is small), rather than those where the movement of the flavor molecules through the emulsion is rate limiting (i.e., when KGE is large), especially for emulsions with relatively large droplet sizes.

It is possible to develop more sophisticated theories to describe the kinetics of flavor release from emulsions which take into account the partitioning and diffusion of the flavor molecules in the oil, water, gas, and interfacial membrane. Nevertheless, these theories are much more complex and can usually only be solved numerically.

9.2.3. Measurements of Partition Coefficients and Flavor Release

A variety of experimental techniques can be used to measure equilibrium partition coefficients and the kinetics of flavor release in emulsions.

9.2.3.1. Head Space Analysis

The concentration of flavor in the vapor phase above a liquid can be determined by head space analysis (Franzen and Kinsella 1975, Overbosch et al. 1991, O'Neill 1996, Landy et al. 1996, Guyot et al. 1996). The liquid is placed in a sealed container which is stored under conditions of constant temperature and pressure. Samples of the gas phase are removed from the head space using a syringe which is inserted through the lid of the sealed container, and the concentration of flavor is measured, usually by gas chromatography or high-performance liquid chromatography. The concentration of volatile flavors in many foods is too low to be detected directly by conventional chromatography techniques, and therefore it is necessary to concentrate the samples prior to analysis. Head space analysis can be carried out over time to monitor the kinetics of flavor release or after the sample has been left long enough for equilibrium to be attained to determine equilibrium partition coefficients.

9.2.3.2. Concentration Analysis in Static Binary Liquids

The equilibrium partition coefficient of a flavor between a bulk oil and bulk water phase can be determined by measuring the concentration of flavor in the two phases after the system has been left long enough to attain equilibrium (McNulty 1987, Guyot et al. 1996, Huang et al. 1997). The flavor is usually added to one of the liquids first, and then the water phase is poured into the container and the oil phase is poured on top. The container is sealed and stored in a temperature-controlled environment until equilibrium is achieved, which can be a considerable period (a few days or weeks), although this time can be shortened by mild agitation of the sample. The method used to determine the concentration depends on the nature of the flavor molecule. The most commonly used techniques are spectrophotometry, chromatography, and radiolabeling. In some systems, it is possible to analyze the solutions directly, whereas in others it is necessary to extract the flavors first using appropriate solvents.

The partition coefficient of flavors in emulsions can be determined using a similar procedure. The emulsion to be analyzed is placed into a container and a known amount of the flavor is added to the continuous phase. The container is filled to the top and sealed to prevent any of the flavor from partitioning into the vapor phase. It is then stored in a temperature-controlled environment until it reaches equilibrium. Equilibrium is attained much more rapidly than in a nonemulsified system because the molecules only have to diffuse a short distance through the droplets. The emulsion is centrifuged to separate the droplets from the continuous phase, and then a sample of the continuous phase is removed for analysis of the flavor concentration. The partition coefficient can then be determined from a knowledge of the overall flavor concentration: KDC = cD/cC = (ctotal - cC)/cC. One important limitation of this technique is that it cannot distinguish between the flavor which is contained within the droplets and that which is associated with the interfacial membrane.

9.2.3.3. Concentration Analysis in Stirred Diffusion Cells

Flavor release occurs under highly dynamic conditions within the mouth (Land 1996), and so a number of workers have developed experimental techniques which attempt to mimic these conditions. McNulty and Karel (1973) developed a stirred diffusion cell for monitoring the mass transport of flavor compounds between an oil and aqueous phase under shear conditions. The flavor compound is initially dissolved in either the oil or aqueous phase. A known volume of the aqueous phase is then poured into the vessel, and a known volume of oil is poured on top. The oil and aqueous phases are sheared separately using a pair of stirrers, and samples are extracted periodically using syringes which protrude into each of the liquids (Figure 9.8). These samples are then analyzed to determine the concentration of the flavor within them. The liquids are stirred at a rate which ensures a uniform flavor distribution, without significantly disturbing the air-water or oil-water interfaces. By carrying out the measurements over a function of time, it is possible to determine the kinetics of flavor transport between the oil and water.

A similar system can be used to study the movement of flavor from a solution or emulsion to the vapor phase above it. The liquid to be analyzed is placed in a sealed vessel and stirred at a constant rate. A syringe is used to withdraw samples from the head space above the liquid as a function of time. Alternatively, a continuous flow of gas can be passed across the stirred liquid and the concentration of flavor within it determined by chromatography (Roberts and Acree 1996). Thus it is possible to simulate the agitation of the food within the mouth, as well as the flow of the gas across the food during manufacture.

FIGURE 9.8 Diagram of stirred diffusion cells used to measure the kinetics of flavor release in liquids and emulsions.

9.2.3.4. Sensory Analysis

The ultimate test of the flavor profile of a food is its acceptance by consumers. Analytical tests carried out in a laboratory help to identify the most important factors which determine flavor release, but they cannot model the extreme complexity of the human sensory system. For this reason, many researchers use sensory analysis by human subjects to assess the overall flavor profile of a food sample (Buttery et al. 1973, Williams 1986, Barylko-Pikielna et al. 1994, Guyot et al. 1996).

Was this article helpful?

0 0
Drop Fat The Low Carb Way

Drop Fat The Low Carb Way

Sick Of Going Round In Circles With Your Dieting? You're About To Discover The Easiest Way To Drop The Fat Once And For All, And Start Living The Life You've Always Dreamed Of This book is one of the most valuable resources when looking at starting a low carb die.

Get My Free Ebook


Post a comment