One of the most important decisions a food manufacturer must make when developing an emulsion-based food product is the selection of the most appropriate emulsifier (Fisher and Parker 1985, Charalambous and Doxastakis 1989, Dickinson 1992, Hasenhuettl 1997). A huge number of emulsifiers are available as food ingredients, and each has its own unique characteristics and optimum range of applications (Hasenhuettl and Hartel 1997). The efficiency of an emulsifier is governed by a number of characteristics, including the minimum amount required to produce a stable emulsion, its ability to prevent droplets from aggregating over time, the speed at which it adsorbs to the droplet surface during homogenization, the interfacial tension, and the thickness and viscoelasticity of the interfacial membrane. These characteristics depend on the food in which the emulsifier is present and the prevailing environmental conditions (e.g., pH, ionic strength, ion type, oil type, ingredient interactions, temperature, and mechanical agitation) (Sherman 1995). For this reason, it is difficult to accurately predict the behavior of an emulsifier from a knowledge of its chemical structure (although some general prediction about its functional properties is usually possible). Instead, it is often better to test the efficiency of an emulsifier under conditions which are similar to those found in the actual food product in which it is going to be used (Sherman 1995). A number of procedures commonly used to test emulsifier efficiency are discussed in this section.
It is often important for a food manufacturer to know the minimum amount of an emulsifier that can be used to create a stable emulsion. The emulsifying capacity of a water-soluble emulsifier is defined as the maximum amount of oil that can be dispersed in an aqueous solution that contains a specific amount of the emulsifier without the emulsion breaking down or inverting into a water-in-oil emulsion (Sherman 1995). Experimentally, it is determined by placing an aqueous emulsifier solution into a vessel and continuously agitating using a highspeed blender as small volumes of oil are titrated into the vessel (Swift et al. 1961, Das and Kinsella 1990).* The end point of the titration occurs when the emulsion breaks down or inverts, which can be determined by optical, rheological, or electrical conductivity measurements. The greater the volume of oil which can be incorporated into the emulsion before it
* The emulsifying capacity of an oil-soluble emulsifier can be determined in the same way, except that the water is titrated into the oil phase.
breaks down, the higher the emulsifying capacity of the emulsifier. Although this test is widely used to characterize emulsifiers, it has a number of drawbacks which limit its application as a standard procedure (Sherman 1995, Dalgleish 1996a). The main problem with the technique is that the amount of emulsifier required to stabilize the emulsion is governed by the oil-water interfacial area rather than by the oil concentration, and so the emulsifying capacity depends on the size of the droplets produced during agitation. As a consequence, the results are particularly sensitive to the type of blender and blending conditions used in the test. In addition, the results of the test have also been found to depend on the rate at which the oil is titrated into the vessel, the method used to determine the end point, the initial emulsifier concentration, and the measurement temperature (Sherman 1995). The emulsifying capacity should therefore be regarded as a qualitative index which depends on the specific conditions used to carry out the test. Nevertheless, it is useful for comparing the efficiency of different emulsifiers under the same experimental conditions.
A more reliable means of characterizing the minimum amount of emulsifier required to form an emulsion is to measure the surface load (D, which corresponds to the mass of emulsifier required to cover a unit area of droplet surface (Dickinson 1992). A stable emulsion is prepared by homogenizing known amounts of oil, water, and emulsifier. The mass of emulsifier adsorbed to the surface of the droplets per unit volume of emulsion (Ca/kg m-3) is equal to the initial emulsifier concentration minus that remaining in the aqueous phase after homogenization (which is determined by centrifuging the emulsion to remove the droplets and then analyzing the emulsifier concentration in the serum). The total droplet surface area covered by the adsorbed emulsifier is given by S = 60 Ve/d32, where Ve is the emulsion volume and d32 is the volume-surface mean droplet diameter. Thus the surface load can be calculated: rS = CaVJS = Cad32/60, which is typically a few milligrams per meter squared. A knowledge of the surface load enables one to calculate the minimum amount of emulsifier required to prepare an emulsion that contain droplets of a given size and concentration. In practice, an excess of emulsifier is usually needed because it does not all adsorb to the surface of the droplets during homogenization due to the finite time it takes for an emulsifier to reach the oil-water interface and because there is an equilibrium between the emulsifier at the droplet surface and that in the continuous phase (Hunt and Dalgleish 1994, Dalgleish 1996a). In addition, the surface load is often dependent on environmental conditions, such as pH, ionic strength, temperature, and protein concentration (Dickinson 1992, Hunt and Dalgleish 1994, Dalgleish 1996a).
An efficient emulsifier produces an emulsion in which there is no visible separation of the oil and water phases over time. Phase separation may not become visible to the human eye for a long time, even though some emulsion breakdown has occurred. Consequently, it is important to have analytical tests which can be used to detect the initial stages of emulsion breakdown, so that their long-term stability can be predicted.
One widely used test is to centrifuge an emulsion at a given speed and time and observe the amount of creaming and/or oil separation which occurs (Smith and Mitchell 1976, Tornberg and Hermannson 1977, Aoki et al. 1984, Das and Kinsella 1990). This test can be used to predict the stability of an emulsion to creaming using relatively low centrifuge speeds or to coalescence by using speeds which are high enough to rupture the interfacial membranes. The greater the degree of creaming or oil separation that occurs, the greater the instability of an emulsion and the less efficient the emulsifier.
An alternative approach which can be used to accelerate emulsion instability is to measure the degree of droplet coalescence when an emulsion is subjected to mechanical agitation (Britten and Giroux 1991, Dickinson et al. 1993b, Dickinson and Williams 1994). The droplet size distribution of the emulsions can be measured either as a function of time as the emulsions are agitated at a constant stirring speed or as a function of stirring speed after the emulsions have been agitated for a fixed time. The faster the increase in droplet size with time, the greater the instability of the emulsion and the lower the efficiency of the emulsifier.
Although these tests are widely used and can be carried out fairly rapidly, they do have a number of important limitations: the rate of creaming or coalescence in a centrifugal field or during agitation may not be a good indication of emulsion instability under normal storage conditions, and it does not take into account chemical or biochemical reactions that might alter emulsion stability over extended periods.
A more quantitative method of determining emulsifier efficiency is to measure the change in the particle size distribution of an emulsion with time using one of the analytical techniques discussed in Section 10.3. An efficient emulsifier produces emulsions in which the particle size distribution does not change over time, whereas a poor emulsifier produces emulsions in which the particle size increases due to coalescence and/or flocculation. The kinetics of emulsion stability can be established by measuring the rate at which the particle size increases with time. These tests should be carried out under conditions similar to those found in the final product (e.g., pH, ionic strength, composition, temperature, etc.).
The analytical instruments used for measuring particle size distributions are often fairly expensive and are not available in many small institutions. In this case, a measure of the flocculation or coalescence in an emulsion can be obtained using a simple UV-visible spectrophotometer (Walstra 1968, Pearce and Kinsella 1978, Reddy and Fogler 1981, Pandolfe and Masucci 1984). The turbidity of light is measured at a single wavelength or over a range of wavelengths, and the mean particle size is estimated using light-scattering theory. This technique should be used with caution, because the relationship between particle size and turbidity is fairly complex in the region where the droplet radius is the same order of magnitude as the wavelength of light used (see Figure 9.12).
One of the most valuable means of obtaining information about the characteristics of an emulsifier is to measure the reduction in the surface (or interfacial) tension when it adsorbs to a surface (or an interface). Surface tension measurements can be used to determine the kinetics of emulsifier adsorption, the packing of emulsifier molecules at an interface, critical micelle concentrations, surface pressures, and competitive adsorption (Chapter 5). A variety of analytical instruments are available for measuring the surface or interfacial tension of liquids, as discussed in Section 5.10.
The stability of food emulsions to creaming and droplet coalescence depends on the rheologi-cal characteristics of the interfacial membranes which surround the droplets (Chapter 7), and so it is often important for food scientists to be able to quantify the rheological characteristics of interfaces. In addition, interfacial rheology measurements can also be used to provide valuable information about other characteristics of emulsifiers, such as adsorption kinetics, competitive adsorption, and interfacial interactions. Interfacial rheology is the two-dimensional equivalent of bulk rheology (Chapter 8), and consequently many of the principles and concepts are directly analogous.* An interface can be viscous, elastic, or viscoelastic depending on the type, concentration, and interactions of the molecules present (Murray and Dickinson
* Although it is convenient to consider interfacial rheology in two dimensions, it must be remembered that the interfacial membrane has a finite thickness (usually a few nanometers) in reality, and this may also influence its rheological characteristics.
1996). Two types of deformation are particularly important at an interface: shear and dilatation. The shear behavior of an interface is characterized by its resistance to the "sliding" of neighboring regions past one another, without any change in the overall interfacial area. The dilatational behavior of an interface is characterized by its resistance to the expansion or contraction of its surface area. Instruments for measuring the shear and dilatational rheology of interfaces were reviewed in Section 5.11.
10.3. MICROSTRUCTURE AND DROPLET SIZE DISTRIBUTION 10.3.1. Microscopy
The unaided human eye can resolve objects which are greater than about 0.1 mm (100 |im) apart (Aguilera and Stanley 1990). Many of the structural components in food emulsions are smaller than this lower limit and therefore cannot be observed directly by the eye (e.g., emulsion droplets, surfactant micelles, fat crystals, gas bubbles, and protein aggregates) (Dickinson 1992). Our normal senses must therefore be augmented by microscopic techniques which enable us to observe tiny objects (Aguilera and Stanley 1990, Kalab et al. 1995, Smart et al. 1995). A number of these techniques are available to provide information about the structure, dimensions, and organization of the components in food emulsions (e.g., optical microscopy, scanning and transmission electron microscopy, and atomic force microscopy) (Kirby et al. 1995, Kalab et al. 1995, Smart et al. 1995). These techniques have the ability to provide information about structurally complex systems in the form of "images" which are relatively easy to comprehend by human beings (Kirby et al. 1995). Each microscopic technique works on different physicochemical principles and can be used to examine different levels and types of structural organization. Nevertheless, any type of microscope must have three qualities if it is going to be used to examine the structure of small objects: resolution, magnification, and contrast (Aguilera and Stanley 1990). Resolution is the ability to distinguish between two objects which are close together. Magnification is the number of times that the image is greater than the object being examined. Contrast determines how well an object can be distinguished from its background.
Although the optical microscope was developed over a century ago, it is still one of the most valuable tools for observing the microstructure of emulsions (Mikula 1992; Hunter 1986, 1993). An optical microscope contains a series of lenses which direct the light through the specimen and magnify the resulting image (Figure 10.1). The resolution of an optical microscope is determined by the wavelength of light used and the mechanical design of the instrument (Franklin 1977, Hunter 1986). The theoretical limit of resolution of an optical microscope is about 0.2 | m, but in practice it is difficult to obtain reliable measurements below about 1 |im (Hunter 1993). This is because of technical difficulties associated with the design of the instrument and because the Brownian motion of small particles causes images to appear blurred. The optical microscope therefore has limited application to many food emulsions because they contain structures with sizes below the lower limit of resolution. Nevertheless, it can provide valuable information about the size distribution of droplets in emulsions which contain larger droplets and can be used to distinguish between flocculation and coalescence (Mikula 1992), which is often difficult using instrumental techniques based on light scattering, electrical pulse counting, or ultrasonics.
The natural contrast between the major components in food emulsions is often fairly poor (because they have similar refractive indices or color), which makes it difficult to reliably distinguish them from each other using conventional bright-field optical microscopy. For this
Scattered Specimen ^ Light
FIGURE 10.1 Comparison of conventional bright-field and dark-field microscopes.
reason, the technique has been modified in a number of ways to enhance the contrast, improve the image quality, and provide more detailed information about the composition and microstructure of food emulsions. Various types of chemical stains are available which bind to particular components within an emulsion (e.g., the proteins, polysaccharides, or lipids) and therefore enable specific structural features to be highlighted (Gurr 1961, Smart et al. 1995). These stains must be used with caution because they can alter the structure being examined. In addition, they are often difficult to incorporate into concentrated or semisolid emulsions.
The contrast between different components can be improved without using chemical stains by modifying the design of the optical microscope (e.g., by using phase contrast or differential interference contrast microscopy) (Aguilera and Stanley 1990). These techniques improve contrast by using special lenses which convert small differences in refractive index into differences in light intensity.
The structure of optically anisotropic food components, such as fat crystals, starch granules, and muscle fibers, can be studied using birefringent microscopy (Aguilera and Stanley 1990). Anisotropic materials cause plane polarized light to be rotated, whereas isotropic materials do not, and so it is possible to distinguish anisotropic and isotropic materials using "crossed polarizers." This technique is particularly useful for monitoring phase transitions of fat and for determining the location and morphology of fat crystals in emulsion droplets (Walstra 1967, Boode 1992).
The characteristics of particles with sizes less than a micrometer can be observed by dark-field illumination using an instrument known as an ultramicroscope (Shaw 1980, Farinato and Rowell 1983, Hunter 1986). A beam of light is passed through the specimen at a right angle to the eyepiece (Figure 10.1). In the absence of any particles, the specimen appears completely black, but when there are particles present, they scatter light and the image appears as a series of bright spots against a black background. This technique can be used to detect particles as small as 10 nm; however, the particles appear as blurred spots rather than well-defined images whose size can be measured directly. The particle size is inferred from the brightness of the spots or from measurements of their Brownian motion. The ultramicroscope can also be used to determine the number of particles in a given volume or to monitor the motion of particles in an electric field (Section 10.6).
Certain food components either fluoresce naturally or can be made to fluoresce by adding fluorescent dyes which bind to them (Aguilera and Stanley 1990, Kalab et al. 1995).
Fluorescent materials adsorb electromagnetic radiation at one wavelength and emit it at a higher wavelength (Skoog et al. 1994). A conventional bright-field optical microscope can be modified to act as a fluorescence microscope by adding two filters (or other suitable wavelength selectors). One filter is placed before the light enters the sample and produces a monochromatic excitation beam. The other filter is placed after the light beam has passed through the sample and produces a monochromatic emission beam. A variety of fluorescent dyes (fluorophores) are available which bind to specific components within a food (e.g., proteins, fats, or carbohydrates) (Larison 1992). Fluorescence microscopes usually use an ultraviolet light source to illuminate the specimen (which is therefore invisible to the human eye), whereas the light emitted by the fluorescent components within a specimen is in the visible part of the electromagnetic spectrum, and so they appear as bright objects against a black background. Fluorescence microscopy is a very sensitive technique that is particularly useful for studying structures that are present at such small concentrations that they cannot be observed using conventional optical microscopy. In addition, it can be used to highlight specific structures within an emulsion by selecting fluorescent dyes which bind to them.
One of the major drawbacks of optical microscopy is the possibility that sample preparation alters the structure of the specimen being analyzed (Aguilera and Stanley 1990, Kalab et al. 1995). Sample preparation may be a simple procedure, such as spreading an emulsion across a slide, or a more complex procedure, such as fixing, embedding, slicing, and staining a sample (Smart et al. 1995). Even a procedure as simple as spreading a specimen across a slide may alter its structural properties and should therefore be carried out carefully and reproducibly. Other disadvantages of optical microscopy are that measurements are often time consuming and subjective, it is often necessary to analyze a large number of different regions within a sample to obtain statistically reliable data, and it is limited to studying structures greater than about 1 |im (Mikula 1992). Many modern optical microscopes now have the capability of being linked to personal computers which can rapidly store and analyze images and thus enhance their ease of operation (Klemaszeski et al. 1989, Mikula 1992).
This is a fairly recent development in optical microscopy which can provide extremely valuable information about the microstructure of food emulsions (Blonk and van Aalst 1993, Brooker 1995, Smart et al. 1995, Vodovotz et al. 1996). Laser scanning confocal microscopy (LSCM) provides higher clarity images than conventional optical microscopy and allows the generation of three-dimensional images of structures without the need to physically section the specimen. The LSCM focuses an extremely narrow laser beam at a particular point in the specimen being analyzed, and a detector measures the intensity of the resulting signal (Figure 10.2). A two-dimensional image is obtained by carrying out measurements at different points in the x-y plane, either by moving the specimen (and keeping the laser beam stationary) or by moving the laser beam (and keeping the specimen stationary). An image is generated by combining the measurements from each individual point. Three-dimensional images are obtained by focusing the laser beam at different depths into the sample and then scanning in the horizontal plane. Observation of the microstructure of multicomponent systems is often facilitated by using the natural fluorescence of certain components or by using fluorescent dyes that bind selectively to specific components (e.g., proteins, lipids, or polysaccharides) (Larison 1992). The LSCM technique suffers from many of the same problems as conventional optical microscopy. Nevertheless, it has a slightly better resolution and sensitivity, and the sample preparation is often less severe.
LSCM has been used to study the size, concentration, and organization of droplets in emulsions (Jokela et al. 1990); to examine the microstructure of butter, margarines, and low-
Was this article helpful?