1. Top of page
  2. Abstract
  3. Introduction
  4. Biopolymer Particle Properties
  5. Structural Design Principles
  6. Biopolymer Particle Formation Methods
  7. Cross-Linking Biopolymer Particles
  8. Conclusions
  9. References

Abstract:  Biopolymer nano- and micro-particles, fabricated from either proteins and/or polysaccharides, can be utilized as delivery systems or to modulate the physicochemical and sensory characteristics of food products. This article reviews the principles underlying the design, fabrication, and application of biopolymer particles fabricated from globular proteins, used either alone or in combination with polysaccharides, within the food industry. The properties of biopolymer particles and their impact on the physicochemical and functional properties of foods are described. The molecular characteristics and interactions of the building blocks (proteins and polysaccharides) used to assemble these particles are briefly reviewed. The major structural design principles that can be used to fabricate biopolymer particles from food-grade proteins and polysaccharides are outlined. Finally, some of the potential applications of functional biopolymer particles within foods are highlighted.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Biopolymer Particle Properties
  5. Structural Design Principles
  6. Biopolymer Particle Formation Methods
  7. Cross-Linking Biopolymer Particles
  8. Conclusions
  9. References

Designed colloidal particulate systems are finding increasing utilization within the food industry for application as encapsulation and delivery systems, or to modulate the physicochemical and sensory properties of foods (Aguilera 2000; Augustin and others 2001; Mezzenga and others 2005; Sanguansri and Augustin 2006; McClements and others 2007). To be commercially viable these systems must be prepared entirely from food-grade ingredients using economic and reliable processing operations. One of the most promising routes to produce food-grade colloidal particulates is to create biopolymer particles from proteins and/or polysaccharides (Norton and Frith 2001; Benichou and others 2002; Dickinson 2003; Malone and Appelqvist 2003; Chen and others 2005; Burey and others 2008). Nano- and micro-scale biopolymer particles can be created from proteins and polysaccharides using a number of different physicochemical principles, including controlled complexation, segregation, gelation, or disruption. These biopolymer particles must be carefully designed and manufactured so that they exhibit the required functional attributes within the final product, for example, optical properties, rheological properties, release characteristics, encapsulation properties, and physicochemical stability. A great deal of research has focused on understanding the behavior of protein and protein-polysaccharide systems, and their potential for forming biopolymer particles (Aguilera and Stanley 1999; Benichou and others 2002; Tolstoguvoz 2007; van der Goot and Manski 2007). It is the purpose of this article to provide an overview of the properties of globular protein and polysaccharide building blocks, methods to manufacture biopolymer particulates with controlled properties, strategies for stabilizing these particulates, and potential applications of these particulates within food systems.

There is a growing trend within the food industry toward the development of innovative products through the rational design of functional structures based on the application of fundamental physicochemical principles, rather than the use of the more traditional “trial and error” approach (Ubbink and Krueger 2006; McClements and others 2009). A major reason for this tendency is that there are currently only a limited number of substances that are legally allowed in foods, and it is extremely time-consuming and costly to get new ingredients approved. Consequently, food manufacturers must introduce innovation through controlling the structural organization of existing food-grade ingredients, rather than through introducing new functional ingredients (Champagne and Fustier 2007). It is for this reason that many technologies developed in other industries cannot simply be adopted by the food industry because they use either ingredients or processes that are not legally acceptable or economically viable (Gouin 2004).

Biopolymer particles may be utilized for the encapsulation, protection, and delivery of various functional food ingredients, such as bioactive lipids, minerals, enzymes, peptides, and dietary fibers (Gouin 2004; Chen and others 2006; Madene and others 2006; Champagne and Fustier 2007). In these applications it is important that the biopolymer particles do not adversely affect the physicochemical characteristics of the food in which they are incorporated, but that they are capable of delivering the encapsulated component at the point where it needs to be released (mouth, stomach, small intestine, colon). Biopolymer particles may also be used to modulate the physicochemical and sensory properties of foods, such as their appearance, texture, stability, or flavor (Chen 2007; Foegeding and Drake 2007; Cernikova and others 2008; Macku and others 2008; Sala and others 2008; van den Berg and others 2008). This application may be desirable to produce foods with novel properties, or to replace ingredients associated with health problems. For example, biopolymer particles could be used to simulate the properties of fat droplets in emulsified food products, thereby lowering dietary lipid intake (Tamime and others 1995; Sandoval-Castilla and others 2004; Liu and others 2007).

Biopolymer particulates can be fabricated using many different physicochemical approaches, for example, extrusion/gelation, segregation, aggregation, size-reduction, and solvent removal. All of these approaches demand a firm knowledge of the properties of food biopolymers, their possible interactions in solution, and structural assembly principles. A creative food scientist must carefully select the appropriate biopolymer(s), particle-creation method(s), and finishing step(s) in accordance with the particulate's ultimate functional purpose.

Biopolymer Particle Properties

  1. Top of page
  2. Abstract
  3. Introduction
  4. Biopolymer Particle Properties
  5. Structural Design Principles
  6. Biopolymer Particle Formation Methods
  7. Cross-Linking Biopolymer Particles
  8. Conclusions
  9. References

The functional performance of a biopolymer particle ultimately depends on its composition, physicochemical properties, and structural characteristics. It is therefore useful to be aware of the most important characteristics of biopolymer particles and their relationship to the bulk physicochemical and sensory properties of foods. The composition and structure of biopolymer particles can then be rationally designed to obtain the desired functional attributes.

Particle composition

Biopolymer particulates can be fabricated from a variety of different food-grade proteins and polysaccharides, for example, whey protein, casein, gelatin, soy protein, zein, starch, cellulose, and various other hydrocolloids. In addition, they may contain other components, such as water, lipids, minerals, and sugars. As well as the overall composition of the biopolymer particles, it is often important to control the spatial organization of the different components (such as homogeneous, dispersion, or core-shell structure), as well as the nature of the molecular interactions acting between the components (physical or covalent bonds). The type, concentration, interactions, and location of the different components present within a biopolymer particle can alter many of its physicochemical properties, including density, refractive index, rheology, environmental sensitivity, and enzyme digestibility, and therefore its functional characteristics, such as stability to gravitational separation, optical properties, bulk rheology, and in vivo digestibility.

It is therefore important to design and fabricate biopolymer particles with specific compositions and structures. For example, if one were designing a biopolymer particle to deliver an anticancer component to the colon, then it would be necessary to construct it from components that were resistant to disruption/digestion within the mouth, stomach, and small intestine, but that do break down in the colon to release the encapsulated component at the required time or site of action.

The selection of particular proteins and polysaccharides to form biopolymer particles will depend on a number of factors: (1) the ability of the biopolymers to assemble into particles; (2) the functional attributes required in the particles formed, such as their size, structure, charge, permeability, and stability to environmental and solution conditions; (3) the legal status, cost, ease of use, and consistency of the biopolymer ingredients and processing operations.

Particle dimensions, shape, and internal structure

Biopolymer particulates can be created with a wide range of different sizes, shapes, and internal structures depending on the nature of the ingredients and assembly conditions used to fabricate them. The mean particle diameter of biopolymer particles typically lies somewhere in the range 0.01 to 1000 μm. Particle size is important because it impacts both the physicochemical properties and sensory attributes of foods (see below). The dimensions of the particles in a colloidal dispersion are usually expressed as either the particle size distribution (PSD) or as a mean particle diameter (d) and associated polydispersity index (σ) (McClements 2005). Sometimes it is important to know the fraction of particles that fall above or below some critical size, namely, the cumulative distribution (McClements 2005).

Many types of biopolymer particles are approximately spherical in shape, although other shapes are also possible, such as spheroids, fibers, or clusters (Figure 1). Nonspherical-shaped particles can be produced by extrusion or molding methods, or by applying shear forces to a solution during particle formation (Norton and Frith 2001; Norton and others 2006). The internal structure of a biopolymer particle may also be important for particular applications. In principle, a biopolymer particle could have various different types of internal structure depending on the type of ingredients present, their interactions with one another, and their relative spatial organization. The particle interior may be (Figure 2):


Figure 1–. Structured biopolymer systems. Spheres represent particles aggregates or separated phases, while solid lines represent linear polysaccharides or proteins.

Download figure to PowerPoint


Figure 2–. Particle internal structuring with pure or mixed components.

Download figure to PowerPoint

  • Homogeneous: A biopolymer particle interior may be comprised of a single kind of biopolymer or a combination of biopolymers that are intimately mixed with each other so that they can be considered to be homogeneous on the length scale considered, for example, particle diameter.

  • Heterogeneous—Dispersion: A biopolymer particle may consist of 2 or more discrete phases, with 1 or more of the phases being dispersed within another phase. The nature, size, shape, interactions, and organization of the dispersed and continuous phases within the particle may vary considerably. For example, the dispersed phase may be a network of aggregated biopolymer molecules dispersed within water. In this case, the size, shape, and connectivity of the pores between the biopolymer molecules may be important. Alternatively, the dispersed phase may be lipid droplets, solid particles, or air bubbles dispersed within a hydrogel phase. In this case, the size, concentration, and location of the dispersed particles may be important.

  • Heterogeneous—Core-Shell: A biopolymer particle may consist of 2 or more discrete phases, with at least 1 of the phases forming a shell around the other phase. This shell may vary in its composition, thickness, and structure, for example, it may be single- or multiple-layered.

The internal structure of the biopolymer particle may play an important role in determining its functional characteristics, such as encapsulation efficiency, loading capacity, permeability, integrity, and digestibility. For example, an active ingredient encapsulated within a highly porous particle may be released much faster than one encapsulated within a dense solid particle, since it can diffuse through the particle matrix more easily through the former. Similarly, a lipid encapsulated within a highly porous biopolymer particle may be more digestible than one encapsulated within a dense solid particle because the digestive enzymes (lipases) should be able to penetrate it more easily.

Particle electrical characteristics

The electrical characteristics of biopolymer particulates are important for a number of reasons. First, they influence the stability of the biopolymer particles to aggregation with each other. If the biopolymer particle charge is sufficiently large, then there will be a high electrostatic repulsion between the particles that may help prevent aggregation. Second, the electrical characteristics of biopolymer particles determine their interactions with other charged species in the surrounding medium. If a biopolymer particle has an opposite charge to another ionic ingredient within a food, then it may form an electrostatic complex that leads to product instability (for example, precipitation and sediment formation). Third, the electrical charge on a biopolymer particle determines how it interacts with biological surfaces within the mouth, stomach, small intestine, and colon. A cationic biopolymer particle may bind to the anionic surface of the tongue thereby causing a perceived astringency. On the other hand, a cationic biopolymer particle may be designed as a delivery system that can bind to a specific location within the gastrointestinal tract (“mucoadhesion”), so as to delay its transit, and thereby deliver its payload over a longer period. Finally, the electrical characteristics of the molecules within a biopolymer particle may determine its internal structure, and might lead to either swelling or shrinking in response to environmental changes such as pH or ionic strength. Similarly charged biopolymers will tend to repel each other, whereas oppositely charged ones will attract each other.

The electrical characteristics of biopolymer particles are determined by the properties of the various components used to fabricate them, as well as the pH and ionic composition of the surrounding medium. The electrical characteristics of proteins and polysaccharides are discussed in more detail in a later section.

Particle physicochemical properties

The type of components a biopolymer particle contains, their interactions with each other, and their relative location determine many of the particle's physicochemical properties, such as density, refractive index, rheology, polarity, and porosity (mesh size). In turn, these physicochemical characteristics determine the way that other molecular species interact with the particle, such as equilibrium partition coefficients (KOW), diffusion coefficients (D), and permeability characteristics (P). In addition, the physicochemical properties of the particles will determine the bulk physicochemistry of the overall system, such as appearance (optical properties depend on particle refractive index), texture (rheological properties depend on particle porosity), and stability (sedimentation rate depends on particle density). It is therefore important to be able to define, measure, and control the physicochemical properties of biopolymer particles so that they exhibit the particular functional attributes required. Some of the impacts of biopolymer particles on the macroscopic physicochemical properties of materials are reviewed in the section below.

Particle integrity and environmental sensitivity

The integrity of a biopolymer particle determines its ability to maintain its composition and structure under a given set of environmental conditions. In many applications it is important for a biopolymer particle to preserve its integrity under 1 set of environmental conditions, but then break down under another set of conditions. For example, if a biopolymer particle is intended to be used as a delivery system for a bioactive component that must be released in the small intestine, then it will have to be designed to maintain its integrity within the product, mouth, and stomach, but break down within the small intestine. A biopolymer particle may lose its integrity by a number of different physicochemical mechanisms: it may completely dissociate; it may erode throughout; it may erode from the edges; or it may swell or shrink (Figure 3). The mechanism involved depends on the type of biopolymer components present, the nature of the interactions holding these components together (such as electrostatic, hydrogen, hydrophobic, van der Waals, or covalent bonds), and the prevailing environmental conditions (pH, ionic strength, temperature, agitation, or enzyme activity).


Figure 3–. Common methods for particle integrity loss (gray spheres) and release of entrapped components (black spheres).

Download figure to PowerPoint

Impact of particle properties on physicochemical and sensory properties

In practical applications it is important that a biopolymer particle either improves or at least does not adversely impact the physicochemical and sensory properties of the food product in which it is incorporated. It is therefore important to understand how biopolymer particle characteristics impact the bulk physicochemical and sensory properties of foods.

Optical properties

The optical properties of food materials may be altered when biopolymer particles are incorporated into them. The 2 most important optical properties of foods are their opacity (which is primarily determined by light scattering) and their color (which is primarily determined by selective absorption of light waves). Biopolymer particles typically have a different refractive index than the surrounding medium and so they scatter light, thereby altering the optical properties and appearance of foods. The overall impact of added biopolymer particles on the optical properties of a particular food depends on their concentration, size, and refractive index (McClements 2002). The impact of particle size and refractive index on the optical properties (turbidity) of a biopolymer particle suspension is shown in Figure 4. The turbidity is relatively low for small particles (d < 50 nm), increases to a maximum value around d = 1000 nm, and then decreases as the particle size is increased further. The turbidity increases as the refractive index contrast increases, which would occur if the total biopolymer concentration within a biopolymer particle increased (if the porosity decreased). The impact of biopolymer particles on optical properties has important practical consequences for their application within different types of foods, since some products should be transparent (such as clear beverages), whereas others should be opaque (yogurt, dressings, sauces). A biopolymer particle may therefore have to be designed so that its properties do not adversely impact the final optical properties and appearance of a product, for example, by controlling its size or refractive index contrast. For transparent products it would be necessary to use small biopolymer particles (d < 50 nm) that do not scatter light strongly, whereas for opaque products it would be necessary to use biopolymer particles that scatter light strongly (d ≈ 500 to 2000 nm).


Figure 4–. Theoretical predictions (Mie theory) of the turbidity dependence for a 0.1 wt% spherical protein particle suspension (n2= 1.50) in water (n1= 1.33) on particle diameter (λ= 500 nm).

Download figure to PowerPoint

Rheological properties

The rheology or texture of food materials may also be affected by the incorporation of biopolymer particles. The impact of biopolymer particles on rheology depends mainly on their concentration, composition, interactions, shape, and size. To a first approximation, the impact of biopolymer particles on the viscosity of fluid foods can be described by the following equation:

  • image(1)

Here, η0 is the shear viscosity of the liquid surrounding the particles, ϕeff (=Rϕ) is the effective volume fraction of the biopolymer particles, ϕ is the actual volume fraction of the biopolymer molecules that make up the particles, ϕc is the critical packing parameter (about 0.6) where spherical particles become closely packed, and R is the effective volume ratio (the total effective volume occupied by the biopolymer particle divided by the volume occupied of the actual biopolymer chains). The effective volume of a biopolymer particle may be considerably greater than the actual volume of the biopolymer molecules for a number of reasons: (1) solvation—biopolymer particles may entrap solvent molecules; (2) flocculation—aggregated particles trap solvent between them; (3) nonsphericity—nonspherical particles have a greater effective volume than the equivalent mass of spherical particles.

The impact of biopolymer concentration and the effective volume ratio on the viscosity of colloidal dispersions of biopolymer particles predicted by the above equation is shown in Figure 5. Overall, the viscosity of a biopolymer suspension increases with increasing biopolymer concentration, gradually at first and then steeply as the particles become more closely packed and the particle concentration approaches ϕc (McClements 2005). Above ϕc the system gains solid-like characteristics, such as a yield stress and an elastic modulus. The effectiveness of biopolymer particles at increasing the viscosity of the system increases as they entrap more solvent (higher R) within their structure (Figure 5).


Figure 5–. Predictions (Eq. 1) of the relative viscosity (η/η0) dependence on the biopolymer concentration for spherical biopolymer particle suspensions of different particle density (ϕi).

Download figure to PowerPoint

Biopolymer particles may be designed to provide desirable rheological attributes to a product (such as thickness or creaminess), or they may be designed so that they do not adversely impact the anticipated textural attributes of a product (by not greatly increasing or decreasing the expected viscosity). Some foods are required to have a low viscosity (such as beverages) while other foods are meant to be highly viscous or gel-like (such as dressings, dips, sauces, or desserts). Based on the above discussion the main characteristics of biopolymer particles that can be designed to control their impact on food texture are: composition, shape, and interactions. Their ability to increase solution viscosity will increase as they become more highly solvated, more asymmetrical, and more strongly aggregated.


The biopolymer particles incorporated into a food system should remain stable through the expected lifetime of the product, which includes specific processing, transport, storage, and utilization conditions. In addition, the biopolymer particles should not adversely impact the normal shelf-life of the product itself. Biopolymer particles may become unstable in a food product through a variety of mechanisms, including gravitational separation (creaming or sedimentation), aggregation (flocculation or coalescence), volumetric changes (swelling or shrinking), and dissociation (erosion or disintegration). It is important to identify the major physicochemical mechanism that promotes biopolymer particle instability in a particular food product so as to successfully combat instability issues. The dominant instability mechanism depends on the characteristics of the molecules from which the particles are assembled (for example, type, concentration, interactions, organization), the characteristics of the particles themselves (such as physicochemical properties, size, and charge), and the environmental conditions (including pH, ionic composition, temperature, and enzyme activity).

In dilute Newtonian fluid solutions, the creaming rate of noninteracting rigid spherical particles is given by Eq. 2.

  • image(2)

Here, U is the creaming velocity (positive U for creaming; negative U for sedimentation), g is the acceleration due to gravity, r is the radius of the particle, ρ is the density, η is the shear viscosity, and the subscripts 1 and 2 refer to the continuous phase and particles, respectively. More sophisticated mathematical models are available that take into account polydispersity, nonspherical particles, particle fluidity, particle–particle interactions, and non-Newtonian fluids (McClements 2005). The overall density of a biopolymer particle will depend on the densities (ρ) and concentrations (ϕ) of the various components within the particle. Typically, biopolymer particles contain biopolymer and water in different ratios, and so the overall particle density is given by:

  • image(3)

Here the subscripts B and W refer to biopolymer and water, respectively. In many applications, biopolymer particles may contain other components and so the above equation must be extended. For example, the density of hydrogel particles filled with lipid droplets of density ρL can be described by:

  • image(4)

In the absence of lipid droplets the sedimentation rate (−U) increases as the biopolymer particle size increases and the biopolymer content of the particles increases (since the density is then further away from that of the surrounding water) (Figure 6a). The impact of filled particle composition (biopolymer, lipid, and water contents) on the sedimentation stability of an aqueous biopolymer particle suspension is shown in Figure 6a and 6b, with the assumption that particles do not form lasting interactions upon contact and maintain a compact, spherical shape. The rate and direction of gravitational separation depends on the composition of the biopolymer particles. At low lipid droplet concentrations and high biopolymer concentrations the particles tend to sediment (−U). On the other hand, at high lipid droplet concentrations and low biopolymer concentrations the particles tend to cream (+U). Interesting from a practical point of few is the fact there are particular combinations of lipid, biopolymer, and water that provide density matching between the filled biopolymer particles and the surrounding aqueous phase, for example, 50% lipid, 10% biopolymer, and 40% water (Figure 6b). If the biopolymer particles are to be used in a low viscosity product (such as a beverage), then it may be necessary to use small or density-matched particles that do not sediment during the shelf-life of the product, but if the biopolymer particles are going to be used in a highly viscous or gelled product (such as a desert, dressing or sauce) then this issue will be less important.


Figure 6–. Theoretical dispersed-phase separation velocity (U; Eq. 2 to 4) as a function of (a) particle radius with increasing lipid concentration and (b) lipid concentration with increasing biopolymer incorporation.

Download figure to PowerPoint

The impact of other major forms of instability of the shelf-life of biopolymer particles can also sometimes be predicted using mathematical models. For example, the tendency for particle aggregation to occur can be predicted by calculating the relative strength of the various attractive and repulsive colloidal interactions operating between them, for example, van der Waals, steric, electrostatic, and hydrophobic forces (McClements 2005). When the attractive forces dominate the particles have a tendency to aggregate, but when the repulsive forces dominate they are stable to aggregation. The tendency for swelling, shrinking, erosion, or dissociation to occur is highly system-specific and will depend on the type of bonds holding the biopolymer molecules together in the particles, such as covalent, ionic, hydrophobic, or hydrogen bonding. More in-depth treatments of particle and colloidal stability can be found in the writings of McClements (2005) and Walstra (2003).

Release characteristics

A biopolymer particle may be designed to encapsulate, protect, and release a specific functional food component, such as a flavor, or antimicrobial, antioxidant, or bioactive nutrient. Consequently, it may have to be designed to release the active component at a particular site, which may be the mouth, stomach, small intestine, or colon. Developing a mechanistic model for such processes requires understanding of the physicochemical mechanisms leading to release. Four main mechanisms, which mainly differ in the role the carrier particle plays in controlling release, have been described:

  • Diffusion: The active component simply diffuses into the surrounding medium through the biopolymer particle matrix, which remains intact. In this case, the mass transport rate will depend on the solubility of the substance in the particle matrix and its diffusion coefficient through the matrix. For biopolymer networks, the diffusion rate may depend on the mesh size of the biopolymer network compared to the size of the diffusing active component, as well as any specific interactions between the biopolymer network and the active component (electrostatic or hydrophobic attraction).

  • Erosion: The active component is released into the media due to erosion processes taking place at either the outer layer or throughout the entire volume of the biopolymer matrix. Matrix erosion may be due to physical, chemical, or enzymatic degradation processes, such as dissociation of physical bonds (electrostatic, hydrophobic, or hydrogen bonds) or chemical or enzymatic hydrolysis of covalent bonds.

  • Fragmentation: The active component is released into the media due to the physical disruption of the carrier, which is either fragmented or fractured, such as by applying shear or compression forces. The bioactive will still diffuse out of the particles, but the rate of release will be quick due to the increased surface area and decreased diffusion path.

  • Swelling/Shrinking: Core release may be induced by the uptake of solvent by the biopolymer particles, which causes the particles to swell. For example, an active component could be encapsulated within a solid biopolymer particle or within a hydrogel biopolymer particle with a pore size small enough to prevent it from leaching out. Once the particle absorbs solvent molecules, it swells and the active component can then diffuse out. The active component could be loaded into the biopolymer particles by initially swelling them in its presence, and then changing the solution conditions to induce shrinkage.

Mathematical theories have been developed that can be used to model different types of release mechanisms involving particulate systems (Pothakamury and BarbosaCanovas 1995; Siepmann and Siepmann 2008). Selecting the most appropriate mathematical model requires understanding the physicochemical origin of the release of the active component, diffusion, erosion, swelling, or fragmentation. Information about the structure and physicochemical properties of the biopolymer particles and surrounding medium are required to utilize these mathematical models, such as the initial PSD, the concentrations of the active component within the particle and surrounding medium (equilibrium partition coefficients), and the transport rate of the active component in the system (translational diffusion coefficients). While such empirical data are often available for drug delivery systems, the data for food delivery systems are only now starting to accumulate (Sereno and others 2009). The utilization of these models will help food scientists to rationalize the design, fabrication, and utilization of biopolymer particle systems with specific release characteristics.

An example of the usefulness of mathematical theories for modeling and predicting the release rates of encapsulated active components is shown in Figure 7. These calculations were carried out assuming that a lipophilic active component was encapsulated within spherical filled biopolymer particles suspended in water, and that the release mechanism was a diffusion process. The mathematical equation used to make these predictions is a modified version of the Crank diffusion equation that takes into account the presence of lipid droplets within the biopolymer particles (Lian and others 2004). The calculations indicate that the release rate increases with decreasing biopolymer particle size, as would be expected when the diffusion path length decreases. Nevertheless, the calculations do give an indication of the size of particle that is needed to significantly slow down the release of active components from biopolymer particles.


Figure 7–. Mathematical prediction (Crank diffusion equation) for the release of a lipophilic encapsulated component from spherical particles in water as a function of time and particle size.

Download figure to PowerPoint

The above calculations suggest that nano-scale particle systems (1 to 50 nm) are unlikely to possess sufficient size to limit release, but instead may rely on direct physical interaction with the entrapped component. In these cases, molecular binding theories (such as for enzymes) may be more appropriate.

Structural Design Principles

  1. Top of page
  2. Abstract
  3. Introduction
  4. Biopolymer Particle Properties
  5. Structural Design Principles
  6. Biopolymer Particle Formation Methods
  7. Cross-Linking Biopolymer Particles
  8. Conclusions
  9. References

The rational design of biopolymer particles with specific functional attributes requires knowledge of the building blocks used to assemble them (proteins and polysaccharides), the forces holding these building blocks together (physical and covalent), and the physicochemical principles used to assemble the building blocks (top-down and bottom-up methods).

The building blocks: biopolymer characteristics

The functional properties of biopolymer particles are ultimately determined by the type, concentration, arrangement, and interactions of the biopolymer molecules used to assemble them. In addition, the type of preparation method that can be used to fabricate a biopolymer particle depends on the specific characteristics of the biopolymer molecules involved (Table 1). For example, electrostatic-driven assembly depends on the biopolymers having particular electrical characteristics under the environmental conditions used. The 2 major classes of food-grade biopolymers that are used to fabricate functional biopolymer particles are proteins and polysaccharides. In this section, we provide a brief overview of the molecular and physicochemical attributes of some of the most widely used biopolymers, restricting ourselves to a discussion of proteins (Table 2) and ionic polysaccharides (Table 3) for the purpose of this review.

Table 1–. Summary of important protein and polysaccharide molecular properties in the assembly of biopolymer particles.
Molecular propertyProteinPolysaccharide
Molecular conformationGlobular: nonpolar residues within interior; linear: restricted secondary structureLinear or random coil
Electrical characteristicsPositive: -NH3+; Negative: -CO2; pI: positive = negativePositive: -NH3+; Negative: -CO2, -SO4
Hydrophobic characteristicsFunctionality: Surface (effective) compared with Internal (ineffective)Interaction with nonpolar residues; protein conjugation
Physical interactionsvan der Waals, electrostatic, hydrophobic, and hydrogen bondingvan der Waals, electrostatic, hydrogen bonding
Chemical reactivityDisulfide interchanges, dehydration, phenolic oxidation, Maillard, and transglutaminase reactionsConjugation, esterification, etherification, radicalization; depolymerization
Table 2–. Summary of important molecular characteristics among common food-grade proteins for assembling biopolymer particles. Derived from Damodaran (1997) and Fox and McSweeney (2003).
NameSourceMain structural typepITm (°C)
  1. A: Type A gelatin; B: Type B gelatin; S: S-type ovalbumin; 7: 7S soy glycinin fraction; 11: 11S soy glycinin fraction.

β-LactoglobulinWhey proteinGlobular4.8–5.175
Bovine serum albuminBovine blood/milkGlobular4.770–90
OvalbuminEgg whiteGlobular4.5–4.774; 82S
Soy glycininSoybeanGlobular∼5677; 8711
GelatinAnimal collagenLinear7–9.4A; 4.8–5.5B40
Table 3–. Summary of important molecular characteristicsA among common food-grade polysaccharides for assembling biopolymer particles. Derived from Stephen and others (2006).
NameSourceMain structure typeMajor monomerGelation
  1. APolysaccharide ingredients available commercially generally possess appreciably different molecular and functional properties; the listed information describes general characteristics for industrial usage.

CarrageenanAlgalLinear/helicalSulfated galactanCooled set
Xanthan gumXanthomonas campestris exudateLinear/helical (high MW)β-D-Glucose (backbone)None; thickens with concentration
Methyl celluloseWood pulpLinearMethylated glucoseHeat-set (rev.)
PectinPlant cell wallsHighly branched coilGlucuronate (backbone)Sugar/heat (HM); calcium (LM)
Beet pectinSugar beet pulpBranched coil with proteinGlucuronate (backbone)Sugar/heat (HM); calcium (LM)
Gum arabicAcacia sapBranched coil domains on protein scaffoldGalactoseConc.-dependent
InulinPlants or bacteriaLinear with occasional branchesβ-D-FructoseConc.-dependent
ChitosanCrustaceans, invertebratesLinear2-Amino-2-deoxy-β-D-glucoseNo common application
AlginateAlgalLinearβ-D-Mannuronic acidCalcium cross-linking

Proteins are polypeptide chains consisting of 20 common amino acids linked together through peptide bonds. Amino acids are distinguished by their side groups, notably amines, carboxylates, hydroxyls, phenolics, and sulfhydryls (Damodaran 1996). The type, number, and sequence of amino acids along the polypeptide chain determine the molecular weight, conformation, electrical charge, hydrophobicity, physical interactions, and chemical reactivity of proteins. Detailed information on protein structure and conformation has been given by Damodaran (1997). Some basic characteristics of proteins (Table 1) are discussed, as well as specific food protein properties (Table 2).

In globular proteins, the polypeptide chain is usually folded into a compact spheroid conformation with most of the nonpolar amino acids located in the interior and the polar amino acids located at the exterior (Seno and Trovato 2007). The major driving force for this configuration is the hydrophobic effect (see below), yet other interactions (such as van der Waals, hydrogen bonding, disulfide bonds, and electrostatic forces) also play a role. Disulfide bonds, specifically, play a key role in maintaining the durable internal structures of certain globular proteins, such as β-lactoglobulin, bovine serum albumin, and soy glycinin (Wolf 1993; Hoffman and van Mill 1997; Fox 2003). Loss in this globular conformation is termed denaturation, which is further discussed, below.

Electrical charge on proteins may be both positive and negative, contributing to electrostatic interactions (attractive or repulsive) among differing environments. Protonated amino side groups (-NH3+) contribute positive charges below pH 10, whereas deprotonated carboxylate side groups (-CO2) contribute negative charges above pH 2. The pH value where positive and negative charge contributions are equal is usually referred to as the isoelectric point (pI); when the pH = pI, a protein has zero net-charge. Some common food protein pI values are listed in Table 2.

Protein hydrophobicity is divided into 2 functional categories: surface hydrophobicity and internal hydrophobicity. Surface hydrophobicity plays a key role in determining the functional attributes of a bulk protein, such as interactions with nonpolar materials (for example, fatty acids and flavors), nonpolar surfaces (oil or air), and other proteins (in the formation of aggregates) (Damodaran 1996). Internal hydrophobicity results from globular conformations and is relatively ineffective in bulk protein interactions.

Proteins interact with other components through a variety of physical or chemical interactions. Physical interactions (for example, electrostatic or hydrophobic forces) play a role in the interchange with solvents, cosolvents, surfactants, phospholipids, polysaccharides, sugars, and minerals. Chemical reactions among proteins include disulfide interchanges, dehydration, phenolic oxidation, Maillard, and transglutaminase reactions. These reactions have been summarized in a review by Gerrard (2002). Disulfide interactions play a stronger role in high temperature protein aggregation (Monahan and others 1995; Hoffmann and van Mill 1999), specifically at high pH (where the cysteine residue is partially deprotonated) (Sava and others 2005). The remaining reactions are discussed in more detail in the section titled “Cross-Linking of Biopolymer Particles.”

Protein denaturation and aggregation. In nature, globular proteins have certain biological functions (including enzyme activity, signaling, transport, and molecular recognition), which require a specific 3-dimensional conformation usually referred to as the native state. Globular proteins in food may lose their native states during the extraction, isolation, and purification of functional food ingredients (such as milk, egg, or soy protein concentrates or isolates) or after incorporation into a food product due to changes in their environment, such as pH, ionic strength, solvent type, temperature, adsorption to interfaces, high pressure, dehydration, or chemical treatments (Damodaran 1996). Understanding the specific factors and mechanisms of protein aggregation is helpful in the rational formation of protein structures, such as spheroids, filaments (Akkermans and others 2007; Jung and others 2008), and nano-tubes (Graveland-Bikker and others 2009).

In principle, denaturation may be either reversible or irreversible, but it is typically irreversible for the globular proteins commonly used in foods. When a globular protein becomes denatured its physical and chemical interactions change appreciably through exposure of nonpolar and sulfur-containing groups that were originally present within the compact interior of the globular protein. Consequently, denatured proteins have a greater tendency to aggregate, irreversibly, with each other through hydrophobic bonding and disulfide bond formation.

Globular proteins experience a number of physicochemical phenomena during thermal denaturation and aggregation, which is often characterized by the model protein β-lactoglobulin (Sawyer 2003). At ambient temperatures globular proteins may self-associate into various types of quaternary structures (for example, dimers or octamers), depending on solution conditions (such as pH, ionic strength, and solvent type). Upon heating, these quaternary structures dissociate (Galani and Apenten 2000), followed by partial unfolding of the secondary and tertiary structures of the protein (Qi and others 1997; Bryant and McClements 1998). Hydrophobic groups are then exposed to the surrounding aqueous phase (Lametti and others 1996; Relkin 1998), leading to protein self-association or binding of nonpolar components (Carrotta and others 2001). Hydrophobically driven aggregation is especially prevalent near the isoelectric point or at high ionic strengths, where the electrostatic repulsion between protein molecules is low (Hoffmann and van Mill 1999). At sufficiently high pH values, disulfide interchanges may also occur (Relkin and others 1998; Sava and others 2005) leading to the formation of irreversibly denatured structures (Creamer and others 2004).

The temperature at which a given protein denatures is termed as the thermal denaturation temperature (Tm). This temperature depends on its unique molecular characteristics, as well as solution conditions (such as pH, ionic strength, solvent composition, and the presence of surface-active substances) (Ragone 2004). Thermal denaturation tends to be less favorable (higher Tm) at pH values close to the isoelectric point, which has been attributed to increased hydrogen bonding within secondary structures, increased electrostatic attraction between oppositely charged regions, and reduced electrostatic repulsion between similarly charged regions along the polypeptide chain (Kella and Kinsella 1988; Boye and others 1996; Relkin and others 1998; Baeza and Pilosof 2002).

The nature of the protein aggregates formed after thermal denaturation is highly dependent on the relative strength of the various molecular interactions in the system. Large spheroid particulate aggregates (dia about 100 to 1000 nm) tend to be formed under conditions where there is only a weak electrostatic repulsion between the protein molecules: pH ≈ pI and/or high ionic strength. On the other hand, thin filaments (dia about 1 to 10 nm) (Arnaudov and others 2003; Gosal and others 2004; Akkermans and others 2008) tend to be formed under conditions where there is a relatively strong electrostatic repulsion between the protein molecules: pH ≠ pI and low ionic strength (Belloque and Smith 1998; Verheul and Roefs 1998; Hoffmann and van Mill 1999; Verheul and others 1999; Foegeding 2006; Donald 2008; Osaka and others 2008). Recently, it has been shown that nano-tubes can be formed by inducing thermal denaturation and aggregation of enzymatically modified globular proteins (Graveland-Bikker and others 2009).


Polysaccharides are polymers of monosaccharides that possess varying molecular weights, conformations, branching, electrical characteristics, flexibility, and hydrophobocity (Table 1) (Rinaudo 2008). Compared to protein, the monomer composition of polysaccharides is more uniform (Table 3), granting particular physicochemical and functional properties (for example, solubility, binding properties, viscosity enhancement, gelation, and surface activity). However, polysaccharides within 1 batch often possess variation in monosaccharide arrangement, molecular weight, and electrical charge, making initial characterization a necessary step in all molecular-level research (Autio 2006). The following section is far from exhaustive; a thorough discussion on polysaccharide structure and relevant analytical methods can be found in books (Eliasson 2006; Stephen and others 2006) and reviews (Muralikrishna and Rao 2007; Rinaudo 2008).

Molecular conformations of polysaccharides are limited to random coil or helical structures. Linear structures are favored as glycosidic bond rotation and chain flexibility are restricted. Chain flexibility in ionic polysaccharides is especially restricted due to intra-chain charge repulsions. Helical polysaccharides (such as carrageenan and agar) are constructured from hydrogen bonds and undergo conformational changes with heating. Helix-to-coil transition occurs at a specific temperature (Tm), whereupon the polysaccharide may reassociate into intermolecular junctions. These junctions are responsible for many polysaccharide gels.

Polysaccharide charge may be anionic, cationic, or nonionic depending on the nature of their functional groups and varies with pH depending on the pKa value of the ionizable side-groups. Common ionized groups on polysaccharides include carboxylates (-CO2, pKa about 2.5 to 4.5), sulfates (-SO4, pKa < 0), and amines (-NH3+, pKa about 9.4). Polysaccharides will generally only have 1 ionized group; for example, pectins contain carboxylates (Ridley and others 2001), carrageenans contain sulfates (Piculell 1995; Whistler and BeMiller 1997), and chitosan contains amines (Kurita 2006; Rinaudo 2006).

The ability of polysaccharides to self-associate and form gels has important implications for biopolymer particle formation. In an excellent review by Williams (2007), gelling biopolymers are categorized by their gel formation mechanism: cooling (for example, carrageenan); heating (methyl cellulose); ion addition (alginate); and retrogradation (amylopectin). Physical interactions in polysaccharides are largely based on hydrogen bonding and, among ionic species, electrostatic interactions (often via divalent ions). These play a role in the formation of intermolecular junctions and solubilization. Nonpolar-like interactions are also possible among helical, uncharged polysaccharides (for example, maltodextrin) and nonpolar residues (such as fatty acids) (Siswoyo and Morita 2003). True nonpolar properties among polysaccharides generally result from conjugation with amphiphilic protein fractions (for example, gum arabic or pectins) (Gaspar and others 2001; Funami and others 2007; Yapo and others 2007; Kirby and others 2008; Mahendran and others 2008; Nakauma and others 2008). Other physicochemical properties may be attained in polysaccharides through chemical modification, such as changes in molecular weight (Coffey and others 2006; Gidley and Reid 2006; Wurzburg 2006) or the addition of specific functional groups (for example, phosphorylation) (Whistler and BeMiller 1997).

The cement: physicochemical interactions

Biopolymer molecules interact with each other and with other molecules through a variety of physical and chemical bonds. The sign, magnitude, strength, and direction of these forces can often be modulated by changes in environmental conditions or solution composition, such as pH, ionic strength, temperature, and solvent type. These forces are responsible for holding biopolymer particles together, and determine the way a biopolymer particle responds to different environments and solutions, whether it will stay intact, swell, shrink, erode, or disintegrate. Understanding the nature of these forces and the factors that impact them is therefore essential for the rational design of functional biopolymer particles. A detailed treatment on physicochemical interactive forces, along with predictive theoretical or semi-empirical models, is found in a variety of books (Hunter 1986; Israelachvili 1992; McClements 2005). In this section, we briefly summarize some relevant physicochemical interactive forces (Table 4).

Table 4–. General features of some physicochemical interactions among food biopolymers and relevant factors.
Interaction typeSignMagnitude, factorsRange, factorsEffect of environment
  1. aBiopolymer conformation refers to any environmental factor that significantly alters the conformation of the biopolymer, changing possible physicochemical interactions.

Electrostatic interactionsAttractive: opposing charge; Repulsive: similar chargeStrong; Factors: charge density, surface charge density, solvent dielectric constantLong range; Factors: solvent dielectric properties, ionic screening (ions or dipoles)Magnitude: pH, dielectric constant, biopolymer conformationa; Range: pH, ionic strength, dielectric constant
Hydrogen bondingAttractiveRelatively weak; Factors: density of interactive speciesShort rangeMagnitude: temperature, solvent dielectric constant
Hydrophobic interactionsAttractiveStrong; Factors: interfacial tension with solvent, exposed surface areaMedium rangeMagnitude and Range: biopolymer conformationa, dielectric constant, temperature
Excluded volume effectsAttractiveVariable; Factors: concentration and biopolymer gyration radiusProportional to gyration radiusMagnitude and Range: biopolymer conformationa

Physicochemical interactive forces include electrostatic interactions, hydrogen bonding, hydrophobic interactions, and excluded volume effects. Some important roles of these forces in protein/polysaccharide particle formation are:

Electrostatic interactions. Electrostatic interactions occur between charged and/or partially charged species among biopolymers, particularly on their surface (hence, surface charge density). The electrical charge on a cationic (chitosan) or anionic (pectin, alginate, or carrageenan) polysaccharide depends on the solution pH relative to the pKa values of their ionizable groups. The net electrical charge on proteins varies from positive, to neutral, to negative as the pH is increased from below to above their isoelectric points (pI). Nevertheless, it should be stressed that the electrical charge distribution on protein surfaces is heterogeneous, with varying amounts of negative and positive regions. Thus, a globular protein that has a net negative charge may have substantial patches of positive charge on its surface, which has important implications for its ability to self-assemble and to interact with other molecules (Cooper and others 2005). Electrostatic interactions are affected by ionic species and are modeled by the Debye screening length (κ−1 about 0.304/√I nm; I: ionic strength).

Hydrogen bonding. Hydrogen bonds are a type of partial-charge (weakly electrostatic) interaction among strongly polarized bonds of hydrogen and oxygen, nitrogen, fluorine, and sulfur. Hydrogen bonds play a major role in stabilizing intra-molecular structures in many proteins and polysaccharides (such as helices, β-sheets), as well as in the formation of junction zones between different biopolymer molecules (as in gelatin, cellulose, starch) (McClements 2005). These forces are weakened at increased temperature, which contributes to the loss in native biopolymer conformation (denaturation).

Hydrophobic interactions. Hydrophobic interactions result from a thermodynamically favorable exclusion of nonpolar species from water. Practical effects of this are: (1) proteins with a significant number of nonpolar residues fold into globular conformations to reduce solvent exposure, (2) upon denaturation (for example, at high temperature), (3) intermolecular nonpolar biopolymer residues interact together to form aggregates, and (4) small nonpolar molecules have a tendency to interact with hydrophobic pockets on biopolymers. The magnitude of this force increases with temperature.

Excluded volume interactions. The “excluded volume” or “steric exclusion” effect is the result of a competition for available volume within a system. Different food components may not be able to share the same space because of their shape, conformation, or charge effects, which reduces the configurational entropy of the system. Above a critical concentration, there is an osmotic driving force that favors phase separation of the different kinds of molecular species involved. Further treatment of this type of interaction is given when discussing incompatible phase separation (see below).

Biopolymer Particle Formation Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Biopolymer Particle Properties
  5. Structural Design Principles
  6. Biopolymer Particle Formation Methods
  7. Cross-Linking Biopolymer Particles
  8. Conclusions
  9. References

In general, one can categorize biopolymer particle formation methods as either physicochemical methods or processing operation methods. The former rely principally on the utilization of physical forces (such as hydrogen bonds, electrostatic interactions, or steric exclusion effects) to direct the formation of biopolymer particles, whereas the latter rely principally on the utilization of specific processing operations (such as injection, extrusion, homogenization, and evaporation). In practice, most biopolymer formation methods rely on a combination of both different approaches.

Physicochemical methods

Under certain conditions many biopolymer solutions will separate into 2 or more phases, with each phase having a different biopolymer composition (Dickinson 1998). These biopolymer solutions may contain either a single type of biopolymer or they may contain a mixture of different types of biopolymers. The driving force for phase separation may be either attractive (associative) or repulsion (segregative) interactions between the biopolymers involved (de Kruif and Tuinier 2001). When the driving force for phase separation is repulsive, then the resulting systems are referred to as “thermodynamically incompatible” systems. On the other hand, when the driving force for phase separation is attractive, then the resulting systems are referred to as “thermodynamically compatible” systems. Once a system has phase-separated into 2 or more phases, then it can be sheared or extruded to form a system consisting of 1 aqueous phase dispersed in the other aqueous phase in the form of spheroid particles. This kind of system is often referred to as a water-in-water (W1/W2) emulsion in contrast to an oil-in-water (O/W) emulsion, which consists of oil droplets dispersed in an aqueous phase. Biopolymer particles can then be formed by gelling the internal aqueous phase using an appropriate method for the biopolymers involved, for example, adding calcium ions for an alginate-rich or pectin-rich phase (see below). In the following sections, we briefly describe the physicochemical basis of phase separation for segregative and aggregative systems.

Segregative systems

Consider a binary biopolymer mixture where there is no attractive interaction between the 2 biopolymers. At sufficiently low biopolymer concentrations the system forms a single phase consisting of an intimate mixture of the 2 types of biopolymer. However, at higher biopolymer concentrations the system exists as a 2-phase system, with each phase having a different biopolymer composition, due to the steric exclusion effect mentioned above. If the system is left long enough or is centrifuged, the biopolymer phase with the higher density forms a separate layer at the bottom of the container, whereas the biopolymer phase with the lower density forms a layer at the top of the container. The upper phase is rich in one kind of biopolymer and depleted in the other kind of biopolymer, whereas the opposite is true of the lower phase.

The behavior of thermodynamically incompatible biopolymer mixtures can be conveniently represented using phase diagrams (Figure 8). These phase diagrams can be used to describe the number of phases existing in a biopolymer solution of a particular composition, the relative volumes of the different phases present, and the biopolymer composition of each of the phases. For a mixed biopolymer system, a phase diagram consists of an x-y plot with the concentration of one of the biopolymers as the x-axis and the concentration of the other biopolymer as the y-axis. The boundary between 1-phase and 2-phase regions in the phase diagram is then described using a binodal line: a thermodynamic boundary condition between fully miscible and separating phases. A 1-phase system forms at biopolymer concentrations below the binodal line, while a 2-phase system forms at biopolymer concentrations above the binodal line. The volume fraction and compositions of the upper and lower phases can be determined from the phase diagram by using tie-lines, which are lines drawn from one point on a binodal line to another point. It has been shown that mathematical models can be used to qualitatively relate the phase diagrams of biopolymer mixtures to their molecular characteristics (Wu and others 1998; Clark 2000). These models are useful for identifying those molecular features of proteins and polysaccharides that will optimize the formation of 2-phase systems that can be used to create biopolymer particles.


Figure 8–. Theoretical phase diagram of mixed solutions between biopolymers 1 and 2 that experience incompatible phase separation. Hypothetical biphasic solution A separates along the tie line to points B and C, with respective concentrations B1 and C1.

Download figure to PowerPoint

Incompatible phase separation represents a thermodynamically favorable event, but the spontaneity of this reaction will vary. The binodal line, separating incompatible and miscible phases on a phase diagram, represents a condition of thermodynamic equilibrium (ΔG = 0). Incompatible solutions near the binodal line will phase separate (G < 0), yet the separation is not spontaneous (G″ > 0). Nonspontaneous phase separation requires a nucleation site for growth of a 2nd phase; such separation is termed “nucleation and growth” (Norton and Frith 2001). Crossover to a spontaneous phase separation occurs at the spinodal line (Turgeon and others 2003), where free energy of mixing passes through an inflection point (G[RIGHTWARDS ARROW]G″ < 0).

The most useful regions for forming biopolymer particles would appear to be those in which phase separation occurs by a nucleation and growth mechanism. It has been shown that the type of W/W emulsion formed in the 2-phase region depends on the relative volume fractions of the 2 biopolymer phases: W1 and W2. A W1/W2 emulsion tends to be formed when the W2 phase is in excess, whereas a W2/W1 emulsion is formed when the W1 phase is in excess. When the 2 phases have similar volume fractions the mixed biopolymer system tends to form a bicontinuous structure consisting of interconnecting regions of each biopolymer. These emulsion types are analogous to the O/W and W/O emulsions formed with oil and water when either water or oil is in excess. Interestingly, there currently appears to be no analogy to an emulsifier in W/W emulsions, which would be a surface-active substance that absorbs to the boundary between the 2 phases and prevents it from coalescing.

The morphology of the mixed biopolymer solution during phase separation is determined by the magnitude of the driving force for phase separation and the rheological properties of the 2 phases (Doublier and others 2000). The morphology of the phase-separated system may change appreciably over time or after application of mechanical forces (such as shearing), which has important consequences for the formation of biopolymer particles with specific properties. Initially, a W/W emulsion containing relatively small water droplets may be formed, but these may coalesce with other water droplets, leading to the formation of larger droplets, and eventually complete phase separation. At biopolymer compositions where phase separation occurs through a spinodal decomposition mechanism, the system tends to rapidly separate into small inter-dispersed networks of 1 phase distributed throughout the other phase. On the other hand, at biopolymer compositions where phase separation occurs through a nucleation and growth mechanism, the system tends to form a W/W emulsion consisting of larger spheroids that are highly prone to coalescence. To utilize these network differences, further separation must be arrested by a solidification or gelation procedure (see below). In these cases, the shapes of the separated phases are defined by the biopolymer that gels first, regardless of the concentration or cause of gelation (Doublier and others 2000). The rate of gelation may be accelerated in incompatible separations because the biopolymers in each phase are more concentrated (Capron and others 1999). Gelation may be induced using a variety of methods depending on the nature of the biopolymer phase that makes up the particles, such as temperature alterations, solvent changes, or addition of cross-linking agents (see below). As with other particulate-creation methods, application of shear to biphasic systems prior to gelation induces elongation of spheroid shapes and produces irregular shapes (Wolf and others 2000). Utilization of controlled shear conditions for creation of particular morphologies will be discussed further below.

Knowledge of the phase diagram of a biopolymer mixture is useful to select the most appropriate biopolymer types, solution composition, and preparation conditions required to rationally design and fabricate biopolymer particles with specific functional characteristics. The phase diagram of a mixed biopolymer solution depends on the molecular characteristics of the biopolymer molecules involved (for example, their molecular weight, electrical charge, hydrophobicity and conformation), as well as the prevailing environmental conditions (such as temperature, pH, ionic strength and solvent quality). One of the major driving forces for phase separation in thermodynamically incompatible mixed biopolymer systems is the excluded volume effect. The physicochemical origin of this driving force is analogous to the depletion effect observed in oil-in-water emulsions containing nonabsorbed biopolymers, as in the steric exclusion of biopolymers (Chatterjee and Schweizer 1998; Doublier and others 2000; McClements 2000). The large majority of protein-polysaccharide pairs exhibit this type of phase separation (Bourriot and others 1999; Hemar and others 2001; Simeone and others 2004; Tolstoguzov 2006). Phase separation is a function of the effective biopolymer volume via conformational entropy. Thus, more compact biopolymers (such as globular proteins) require larger concentrations for noticeable phase separation to occur (Tolstoguzov 2006). Also, processes that increase the effective volume (like protein denaturation) will then decrease the necessary concentrations for phase separation.

A variety of analytical techniques are needed to characterize the phase behavior, morphology, and properties of thermodynamically incompatible mixed biopolymer systems. A phase-separated biopolymer mixture tends to separate into an upper phase and a lower phase due to the density difference between the 2 different biopolymer phases. This process may occur naturally due to gravity, but complete phase separation may take a considerable time if the external biopolymer phase has a sufficiently high viscosity or even gel-like characteristics. In these situations, applying centrifugal forces to the biopolymer mixture can accelerate phase separation. Once the biopolymer mixture has been completely separated into an upper and lower layer, each layer can be isolated and its composition measured using standard analytical techniques such as chemical, gravimetric, spectroscopic, or chromatrographic methods (Durrani and others 1993; Butler and Heppenstall-Butler 2003; Pudney and others 2003; Yakimets and others 2005). The bulk physicochemical properties of each biopolymer phase can also be measured, such as its refractive index, density, and rheology (Ding and others 2002; Simeone and others 2004). This information may be important for designing biopolymer particles with particular functional attributes (see above). The morphology of the biopolymer mixture can be measured during phase separation or after application of shear forces using scattering and microscopy techniques, such as small angle light scattering (van Puyvelde and others 2002), rheo-optical microscopy (Antonov and others 2004), or confocal fluorescence microscopy (Firoozmand and others 2008). The driving force for phase separation can be characterized in terms of the interfacial energy at the boundary between the 2 biopolymer phases, as by spinning drop tensiometry (Scholten and others 2004; Nan and others 2006).

Ultimately, one would like to be able to control the composition, physicochemical properties, size, and shape of the biopolymer particles formed. Previous studies indicate this can be achieved by selecting an appropriate initial pair of biopolymers and solution conditions, as well as understanding the type of structures formed in different regions of the phase diagram.

Aggregative systems

Aggregative systems are those in which there tends to be an attractive interaction between the biopolymer molecules, which causes them to aggregate with each other. Aggregation may occur between similar biopolymer molecules (self-association) or between different types of biopolymer molecules. The nature of the aggregates formed depends on the type and concentration of biopolymer molecules present in solution, as well as prevailing environmental conditions (such as temperature, solvent type, pH, and ionic strength). Knowledge of the physicochemical mechanisms that drive aggregation can often be used to rationally control the characteristics of the biopolymer particles formed using this method.

Single biopolymer systems. Biopolymer particles can sometimes be formed from aqueous solutions containing a single biopolymer type by promoting self-association of the biopolymer molecules. This approach involves altering the solution conditions so that biopolymer–biopolymer interactions are favored over biopolymer–solvent interactions. The specific method used to promote this kind of self-association of the biopolymer molecules depends on their precise molecular characteristics. For example, heating globular protein solutions above their thermal denaturation temperature under conditions where there is a relatively weak attraction between the proteins may lead to the formation of protein nano- or micro- biopolymer particles (Jones and McClements 2008; Jones and others 2009a, 2009b). The size of these biopolymer particles can be controlled by altering the initial biopolymer concentration, holding temperature, holding time, pH, and/or ionic strength. Biopolymer particles can also be formed by changing the quality of the solvent surrounding the biopolymer molecules in solution, as by adding alcohol to an aqueous biopolymer solution. The biopolymer molecules will tend to spontaneously self-associate once a critical alcohol content has been exceeded, leading to the formation of biopolymer aggregates. Biopolymer particles can also be formed by adding a cross-linking agent to an aqueous solution with a biopolymer concentration below the threshold level required to form a macroscopic gel, with c < c*. These cross-linking agents may be chemicals (such as glutaraldehyde or formaldehyde), enzymes (such as transglutaminase or laccase), or mineral ions (such as potassium, calcium, or tripolyphosphate). Finally, the temperature of the biopolymer solution may be either increased or decreased to promote biopolymer–biopolymer interactions. As mentioned above, globular proteins tend to self-associate when they are heated above their thermal denaturation temperature due to the increase in hydrophobicity when they unfold. Some types of modified cellulose will also self-associate upon heating, which has been attributed to an increase in the strength of the hydrophobic attraction between them. Other biopolymers, such as gelatin, alginate, and carrageenan, may self-associate when they are cooled below their thermal transition temperature due to helix formation and association through hydrogen bonding. At present, there is still a relatively poor understanding of how to control solution conditions so that biopolymer particles with well-defined characteristics can be formed using many of these methods.

Mixed biopolymer systems: electrostatic complexes. Different kinds of biopolymers may associate with each other through a variety of interactions, including electrostatic, hydrophobic, and hydrogen bonding. Nevertheless, the most common form of complexation used to form biopolymer particles is the attractive association between oppositely charged biopolymer molecules. The size and charge of the electrostatic complexes formed depends on the precise experimental variables used in their preparation, such as biopolymer types, biopolymer concentrations, pH, and ionic strength (Sanchez and others 2006). Knowledge of the physicochemical basis for complex formation is, therefore, paramount for the rational fabrication of biopolymer particles with specific functional properties.

Electrostatic complexes can be formed directly be mixing a cationic biopolymer solution with an anionic biopolymer solution. However, it is often difficult to control the nature of the complexes formed using this direct approach. Instead, it is common for 2 biopolymers (usually a protein and a polysaccharide) to be mixed together at a pH where they have similar charge signs (either positively or negatively charged), and then to adjust the solution to a pH where they have opposite charges so that complexation is promoted. For anionic polysaccharides (such as pectin), complex formation is achieved by mixing the 2 biopolymers at a pH above the protein's isoelectric point (pI) where the protein and polysaccharide are both negative, and then adjusting the pH to around or below the pI (Girard and others 2002). Conversely, for cationic polysaccharides (such as chitosan), complex formation is achieved by mixing the biopolymers at a pH below the pI of the proteins where the protein and polysaccharide are both positive, and then adjusting the pH to around or above the pI (Guzey and McClements 2006).

A number of different types of biopolymer particles can be formed based on electrostatic complexation of a single biopolymer pair depending on solution conditions. Consider the events that occur when a mixed biopolymer solution containing an anionic polysaccharide and a globular protein is adjusted from a pH above to below the protein's pI. The charge on the protein molecule will go from highly negative to highly positive as the pH is decreased from above to below the pI, whereas the negative charge on the anionic polysaccharide molecules will remain relatively constant until the pKa value of its charged groups is reached when it will start to decrease. A number of different pH regions can be distinguished:

  • 1
    No complexation (pH ≫ pI). Initially, both the protein and polysaccharide molecules have a sufficiently strong negative charge that the electrostatic repulsion between them prevents them from coming close enough to associate.
  • 2
    Soluble complexes (pHc < pH < pHs). When the pH is reduced below a critical value, which we refer to as pHs, then the protein and polysaccharide weakly associate with each other to form soluble complexes. These complexes are relatively small and so they do not scatter light strongly, leading to a transparent or slightly turbid solution. The complexes are prevented from aggregating with one another because of their relatively high net charge. Soluble complexes often form above the pI of the protein, even though both the protein and polysaccharide have a net negative charge, because of the presence of positive patches on the protein's surface to which the anionic polysaccharide can bind (Xia and others 1993; Turgeon and others 2003). Soluble complexes are highly dynamic and reversible structures because of the weak physical interactions holding them together. Consequently, they tend to dissociate when the pH is altered or sufficient salt is added to weaken the electrostatic interactions. Once formed it is possible to create more permanent structures by inducing covalent interactions between the 2 biopolymers, as by using physical, chemical, or enzymatic methods.
  • 3
    Coacervates (pHp < pH < pHc). Upon further reduction of the pH another critical value is reached, which we refer to as pHc, where the protein and polysaccharide associate to form complexes usually referred to as coacervates. These complexes are relatively large (in the 100 to 10000 nm range) and so they scatter light relatively strongly, leading to turbid solutions. In addition, they are highly prone to coalescence because they have a relatively low net charge and so the electrostatic repulsion between them is insufficient to prevent them from merging together. As a result, biopolymer mixtures that form coacervates tend to separate into a 2-phase system, with a dense lower layer that is rich in both biopolymers (the coacervate phase) and a less dense upper layer that is depleted in both biopolymers (the serum phase). Typically, the coacervate phase is a highly viscous or gel-like material that contains a high amount of solvent (>70%). Coacervates usually form in a narrow pH range that is below the globular protein's pI, since the protein and polysaccharide molecules then have opposite charges so that charge neutralization and bridging may occur. Coacervates are also highly dynamic and reversible structures that tend to fall apart when the pH or ionic strength of a solution is altered to weaken the electrostatic interactions. To convert coacervate particles into more permanent structures with well-defined sizes and shapes it is usually necessary to cross-link the biopolymer molecules within them.
  • 4
    Precipitates (pH < pHp). When the electrostatic attraction between the protein and polysaccharide molecules is sufficiently strong, then precipitates may be formed rather than coacervates. The biopolymer molecules within precipitates are packed much more densely than in coacervates (that is, there is much less solvent present). Consequently, they tend to scatter light more effectively and sediment more rapidly.
  • 5
    No complexation (pH ≪ pKa). Eventually, the pH may decrease below the pKa value of the anionic groups on the polysaccharide molecule so it loses its charge and can no longer associate with the protein molecule through electrostatic interactions. This effect is more prevalent with anionic polysaccharides with relatively high pKa values (such as pectin, pH 3.5), rather than those with lower pKa values (such as carrageenan, pH < 2).

The positions of the various critical pH values listed above, as well as the nature of the complexes formed in each region, depend on many factors, including: the molecular characteristics of the protein and polysaccharide molecules (such as charge density, molecular weight, conformation, and flexibility); the total biopolymer concentration; the protein-to-polysaccharide ratio; and the solution conditions (particularly pH and ionic strength). Mathematical models have been developed to describe the binding interactions of charged polysaccharides with charged proteins (Chodanowski and Stoll 2001; de Vries and others 2003; Girard and others 2003). Various analytical methods have been used to obtain information about the composition, structure, and properties of the complexes formed (Weinbreck and others 2003). The value of pHs is higher for highly charged carrageenans (sulfate groups) than for less charged pectins (carboxylate groups), and it deceases with increasing ionic strength due to screening of charged interactions. The pH where soluble complex formation begins is largely independent of protein-to-polysaccharide ratio because it depends on the ability of individual globular proteins to bind to particular sites on the polysaccharide backbone (Sperber and others 2009). On the other hand, the pH where coacervates form is dependent on the protein-to-polysaccharide ratio, since coacervates tend to form when charge neutralization occurs, so a certain number of protein molecules need to be bound per polysaccharide molecule for this to occur. For example, decreases in the protein-to-polysaccharide ratio have been shown to decrease pHc (Weinbreck and others 2004b).

The fact that a variety of different complexes can be formed, based on electrostatic attraction between protein and polysaccharide molecules, means that a variety of different biopolymer particles can be created. Biopolymer particles can be fabricated from soluble complexes of globular proteins and ionic polysaccharides by heating them above the thermal denaturation temperature of the proteins (Yu and others 2006; Hong and McClements 2007; Jones and others 2009a). Biopolymer particles can be fabricated once a coacervate phase has been formed using a number of approaches. The coacervate phase can be stirred with the serum phase to form a W/W emulsion where the coacervate forms the dispersed phase droplets. As mentioned earlier, the coacervate droplets are highly prone to coalescence and the system may quickly separate into an upper layer and a lower layer. The coacervate droplets can be stabilized against aggregation using a suitable method to cross-link the biopolymers within them. In food systems, there are a number of methods that can be used to cross-link the biopolymers, such as physical, chemical, or enzymatic methods (see below). The size and shape of the biopolymer particles produced using this method can be manipulated by controlling the coacervate formation and gelation conditions. For example, the biopolymer particle size and shape can be controlled by shearing the coacervate dispersion at different rates (Weinbreck and Wientjes 2004) or by altering the time when a cross-linking agent is added (Sanchez and others 2006). Alternatively a coacervate phase could be collected and then extruded or injected into a cross-linking solution. The size of the biopolymer particles formed will then depend on the injection conditions (such as nozzle diameter, injection rate, injection volume, and stirring conditions).

The utilization of protein-polysaccharide precipitates as a basis of forming biopolymer particles is rarely investigated. This is probably because it is difficult to control the size of the precipitates formed, and they are highly dense so have a tendency to sediment. Nevertheless, this may be an interesting area for future research since they may exhibit some useful encapsulation properties.

Process operation methods

In this section, we focus on methods of biopolymer particle formation that primarily rely on the utilization of specific processing operations to form and stabilize the particles.

Molding techniques

Particulates of well-defined size and morphology may be formed using molds or lithographs. In this case, biopolymer particles are formed by pouring a biopolymer solution into a cavity with a specific size and shape, and then adjusting the solution or environmental conditions to promote gelation. This method is a small-scale analog of the formation of gelatin gels (“jellies”) that is routinely carried out in home kitchens. In the past, the production of micro- or nano-sized cavities was limited by the available technology. Nevertheless, developments in soft lithography techniques within the past decade have enabled the formation of polymer molds from finely tooled metal plates (Xia and Whitesides 1998). Molten substrates are trapped between these polymer molds and flat polymer plates until the particulate is hardened. A number of different kinds of molding technology have been developed, including replica molding, micro-contact printing, micro-transfer molding, capillary micro-molding, and solvent-assisted micro-molding (Xia and Whitesides 1998).

Classic soft lithographic techniques used polydimethylsiloxane (PDMS) molds. However, PDMS has functional limitations, such as substrate adhesion to the mold. This has been overcome by applying polymeric spacers, such as monolayers of hexa-(ethylene glycol) or bovine serum albumin (Tang and others 2003). Also, simple PDMS molds only produce embossed gelled films, rather than discrete particles. In the PRINT method, nonwetting mold materials, such as photocurable perfluoropolyether (PFPE), reduce wall-adhesion of hydrophilic substrates (Rolland and others 2005). Sufficient pressing causes a complete entrapment of molten substrates within the formed recesses. Using PRINT, particles between 200 and 500 nm have been created (Rolland and others 2005).

To our knowledge, food-grade molding methods have not been attempted commercially, possibly due to cost and scale-up issues. However, some closely related systems have been investigated. For instance, alginate particles have been made by placing a calcium-releasing plate on 1 side of the mold (TaleiFranzesi and others 2006). Diffusion of the calcium ions from the plate caused gelation of alginate within the mold. Photosensitive biopolymers, such as hyaluronic acid, have been similarly created using UV-radiation for cross-linking and particulate formation (Yeh and others 2006). In theory, a variety of oxidation and gelation methods could be combined with PRINT methods to create food-grade particulates. Malone and co-workers (Malone and Appelqvist 2003) used a variety of molding methods to create biopolymer particles with different morphologies and compositions.


Spray-drying is a technique wherein atomized suspensions or solutions are quickly dried using a heated gas (Fellows 2000). Atomization can be accomplished using different types of nozzles within the spray drier, including those that use centrifugal forces, high/differential pressures, or high-intensity ultrasound. The operating temperatures within a spray drier are typically between 150 and 300 °C, but the temperature of the material within the atomized particles is considerably less because of the latent heat of evaporation. In addition, due to the high surface-to-volume ratio of the atomized particles, drying is very quick and minimizes thermal damage. Spray-drying is capable of continuous operation and produces a dry/stable product. The diameters of the particles formed within the dried powder are usually in the micron range. The characteristics of the particles formed in the powder depend on the composition of the initial fluid fed into the spray drier, as well as the operating conditions of the spray drier.

Spray-drying has been widely used within the food industry for the production of dry powders from a variety of materials, including proteins, flavor oils, and lipid droplets (Desai and Hyun Jin 2005; Ameri and Yuh-Fun 2006; Abdul-Fattah and others 2007). Excellent reviews exist on spray-drying of oils for encapsulation (Vega and Roos 2006; Jafari and others 2008). Reconstitution of dispersed droplets is an important issue for application (Vega and Roos 2006). For many spray-dried capsules, the ultimate purpose is to have the encapsulated material released at a controlled rate when the powder is dispersed in a product or comes into contact with saliva during ingestion. During the process of wetting, submersion, dispersion, and dissolution, the particles are subjected to increasing levels of water. The speed and conditions of these steps are important for proper reconstitution (Vega and Roos 2006). Also, changes could occur during storage (for example, caking), which adversely affect the redissolution process.

Spray-drying, despite minimal thermal damages, may induce aggregation or chemical interactions in the material. Protein aggregation is a recognized limitation of spray-drying (Ameri and Maa 2006), particularly among proteins in dried milk powders (Singh 2007; Anandharamakrishnan and others 2008). Studies with IgG indicated that up to 17% of protein aggregation occurs from the spray-drying, while much greater aggregation occurs during the subsequent dry storage (Maury and others 2005). Since dry storage is an unavoidable factor of spay-drying, protein aggregation is a major consideration. Stabilizers, such as carbohydrates and surfactants, are often utilized to reduce protein aggregation in pharmaceutical applications (Ameri and Maa 2006). Carbohydrates (for example, maltodextrin and lactose) can also be used in spray-drying to accelerate the drying of hygroscopic materials (Bhandari and others 1997; Adhikari and others 2004), while also influencing the size and porosity of dispersed particles (Elversson and Millqvist-Fureby 2005). Materials containing both protein and carbohydrate possess the possibility of enhanced Maillard reaction (see below) during spray-drying due to the increased temperature and low moisture content. Maillard reactions are actually limited by reduced feed moisture contents, as intermolecular reactions are slowed when approaching the glass transition (Miao and Roos 2006).

When spray-drying a biopolymer solution containing lipid droplets, one can vary the amount of dispersed oil, the quantity of biopolymer, the composition of the wall material, or the heating conditions to form different kinds of spray-dried particles. Studies have found that higher oil contents will increase the amount of oils in the produced capsules, but leads to a decrease in encapsulating efficiency (Adamiec and Kalemba 2006). Another study found oil type did not have much effect on the entrapped oil content, but it did have an effect on the surface oil content (Baranauskiene and others 2006). Also, the type of biopolymer could change the particle size and surface oil content (Baranauskiene and others 2006).

Solvent desorption

Solvent desorption refers to a class of encapsulation techniques that rely on change in the solvation properties of biopolymers. In this process, the biopolymer and the active component to be encapsulated are solubilized in a particular solvent. Biopolymer particle formation is then induced by changing the solvent conditions to promote solvent desorption, such as by adding incompatible co-solvents, by adding aggregating reagents, or by solvent evaporation (Mosqueira and others 2000; Galindo-Rodriguez and others 2004). A fraction of the active components dispersed within the initial solution become entrapped within the biopolymer particles during solvent desorption (Galindo-Rodriguez and others 2004).

Solvent desorption has a variety of pseudonyms, including nanoprecipitation (Barichello and others 1999; Duclairoir and others 1999; Lee and others 1999; Jiminez and others 2004; Leo and others 2004), interfacial deposition/solvent displacement (Murakami and others 1999; Mosqueira and others 2000, 2001), salting-out (Galindo-Rodriguez and others 2004), simple coacervation (Mauguet and others 2002; Mohanty and Bohidar 2003), and emulsification-diffusion (Galindo-Rodriguez and others 2004; Cirpanli and others 2005; Lai and Tsiang 2005). These different pseudonyms represent differences in the approaches used to induce solvent desorption and biopolymer particle formation, for example, polymer polarity, use of salts, speed of solvent addition, and removal of solvent (as by evaporation). Nevertheless, each of these approaches involves promoting desorption of the solvent from the biopolymer to induce its aggregation and entrapment of the compound of interest. The entrapment efficiency depends on the biopolymer used and the method of solvent desorption.

Solvation of biopolymers can be followed using free energy equations. The free energy of solvation is related to the Flory-Huggins parameter, X12 (Schmitt and others 1998). X12 is a representation of the free energy between lattice segments and is affected by solvent properties. A “good” solvent reduces X12, while a “poor” solvent increases X12 above the critical level required for aggregation (Scott 1948). Directly above this critical level, solvent desorption occurs. Solvent requirements for solvent desorption are large (Mohanty and Bohidar 2003) and may be difficult to incorporate in practical food applications. However, it may be possible to initiate solvent desorption with less solvent by lowering the temperature, which theoretically increases X12 (Schmitt and others 1998).

Charged biopolymers can be desolvated into particulates through the use of alcohols and salts. Nanoparticles of BSA (Rahimnejad and others 2006) or gelatin (Mohanty and Bohidar 2003) were made with the addition of ethanol. Gelatin can also be desolvated by certain salt solutions (Izmailova and others 2001). Chitosan particles have been formed by subjecting aqueous solutions to sodium hydroxide and methanol (Peniche and others 2003). Even casein might be used for desolvation, as it is aggregated in the presence of ethanol and cationic solutions (Gupta and Bohidar 2005; Tsioulpas and others 2007).

Predominantly nonpolar biopolymers might be desolvated through the use of aqueous solutions. Gliadin, a wheat protein, is soluble in ethanol, yet precipitates in water. Desolvation is generally performed by introducing large quantities of aqueous solution and other polar cosolvents (Duclairoir and others 1998) or salts (Mauguet and others 2002; Lazko and others 2004a) to an ethanol-dispersion. Particle size of these particles can be optimized using salts and surfactants (Duclairoir and others 1999). Such gliadin particles (d ∼ 500 nm) were capable of encapsulating retinoic acid (Duclairoir and others 1999). Similar alcohol-soluble biopolymer materials, such as celluloses, could theoretically be created using this methodology. In a parallel study, cellulose has been formed into particulates by promoting its precipitation using acetone and water (Hornig and Heinze 2008).

Along phase interfaces, such as emulsion droplets, solvent desorption can be utilized to precipitate polymers onto the interfacial layer. A study investigating precipitation onto an interface found that the size of emulsion droplets was dependent upon diffusion of the desolvated polymer (Yow and Routh 2006). Reduced affinity of the polymer for the droplet induces larger adsorbed layers by forming more loosely packed structures (Cosgrove and Heath 1987), despite the reduction in driving force.

Injection methods

Biopolymer particles can be formed by injecting a biopolymer solution into another solution that promotes gelation. This 2nd solution may contain a chemical substance that promotes biopolymer cross-linking (such as an acid, base, mineral, or enzyme) or it may be at a different temperature (hot or cold gelation). A common example of this kind of system is the alginate microbead system. Alginate forms a physical gel in the presence of calcium (see below), so microbeads can be formed by controlled injection of an alginate solution into a calcium solution (Liu and others 2006; Shin and others 2007; Amici and others 2008). The microbeads formed have shown great promise as microencapsulation devices for pharmaceutical drugs (Jiao and others 2002; Tonnesen and Karlsen 2002; Shanmugasundaram and others 2005) and probiotics (Sheu and Marshall 1993). It is a relatively easily controlled and inexpensive encapsulation methodology that is suitable for industrial production of microgel particles (Shin and others 2007). Other biopolymers can also be utilized to form microbeads: injecting a pectin solution into a calcium solution (ionic gelation); injecting a chitosan solution into a tripolyphosphate solution (ionic gelation); injecting a whey protein solution into a hot liquid (hot gelation); and injecting a gelatin solution into a cold liquid (cold gelation).

Gelation methods

Biopolymer particles can be formed by disrupting a macroscopic gel, or by promoting biopolymer association at a concentration below the level required for macroscopic gel formation. The interactions that form gel networks can be grouped into 2 main categories: physical and chemical (Totosaus and others 2002; Williams 2007). Physical gels are typically held together by hydrogen bonding, hydrophobic, and divalent ion cross-linking (see below), while chemical gels are covalently bonded via inherent (for example, disulfides) or added substances (see below). Gel properties depend strongly on biopolymer conformation and interaction sites. For instance, linear polymers tend to have more open, porous structures, while particulate aggregates form highly compact agglomerate structures (Walstra 1996). Points of contact determine the gel's fractal dimension, which determine properties such as fracture stress and water-holding capacity.

Gel systems may be valuable as delayed-release agents and texture modifiers. Production of hydrogels or gel micro-capsules as nutraceutical carriers has been well reviewed (Chen and others 2006). There are now reliable methods for tracking the diffusion of various components in a gel, such as water (Walderhaug and Nystrom 1997; Traore and others 2000), aroma/lipophilic volatiles (Malone and Appelqvist 2003; Malone and others 2003; Lian and others 2004), and nutraceuticals (Draye and others 1998). By comparing diffusion rates, a food chemist can tailor his/her particles toward specific release characteristics. Texture, which is a factor of various tactile attributes (Jianshe 2007), can also be controlled through the use of gelled systems (Foegeding 2006). Hydrogel structures have been used as fat mimetics (Schmitt and others 1998; Liu and others 2007) or meat analogs (Schmitt and others 1998).

While gels may be composed of either protein or polysaccharide, combinations of these biopolymers may add unique properties. These diverse combinations have been reviewed previously (de Jong and van de Velde 2007). Often, the incorporation of polysaccharides into protein-based gels adds a synergistic benefit. For many neutral and like-charged polysaccharides, an incompatible interaction with the protein induces added protein agglomeration (Sittikijyothin and others 2007; Spahn and others 2008). The polysaccharide then viscosifies the incompatible interpenetrating phase. On the other hand, associative interactions can also contribute to gel formation through bridging junctions (Oakenfull and others 1999).

Emulsion-templating methods

Biopolymer particles with well-defined sizes can be produced using emulsion-templating methods (Figure 9). In this method, an aqueous biopolymer solution is homogenized with an oil phase containing an oil-soluble emulsifier to form a water-in-oil emulsion (W/O). The size of the water droplets produced can be controlled by varying either the homogenization conditions (homogenization time and intensity, temperature) or system composition (oil-to-water ratio, oil type, and emulsifier concentration). The biopolymers within the water droplets can then be gelled using a number of methods, including heating, cooling, mineral addition, pH changes, and enzyme addition. The gelation method used depends on the nature of the biopolymers (see above). This might be done by mixing a W/O emulsion containing the biopolymer (such as alginate) with a W/O emulsion containing the mineral ion (such as calcium). Once the particles have been gelled they can be separated from the oil phase by filtration or centrifugation, and then washed using an organic solvent to remove any residual oil. The resulting biopolymer particles can then be dispersed in an aqueous solution or dried.


Figure 9–. Emulsion-templating method for particle formation via internal gelation.

Download figure to PowerPoint

Shaping particles through shear

The application of shear forces before or during biopolymer gelation can be used to modulate the size and shape of biopolymer particles. Shear forces are capable of breaking up dispersed droplets into smaller sizes, or of causing spheroid particles to become increasingly elongated. On the other hand, biopolymer association processes can sometimes be halted or limited by applying sufficiently high shear forces. Shearing also influences the rate that gelling agents can be incorporated into biopolymer solutions or W/O emulsions.

For biopolymer particles created by the aggregative phase separation approach it has been found that smaller coacervate particles can be formed by using higher stirring speeds (Jegat and Taverdet 2000). Biopolymer particles with controllable sizes were produced using a combination of controlled shear forces and cross-linking techniques. However, when the shearing rate used was too high, the biopolymer particles began to flocculate. The particle size produced in a coacervate system has been related to the localized “microeddies” produced during stirring (Dobetti and Pantaleo 2002). These microeddies are zones of lower shear forces induced by turbulent flow, which depend on the Reynold's number, the vessel dimensions, and the stirrer characteristics.

For biopolymer particles created by the segregative phase separation approach it has been found that the particle size and shape can also be controlled using shear forces (Stokes and others 2001). At low shear, spherical droplets were obtained, while elongated fibrils formed at higher shear (Wolf and others 2000). Gelation of the internal aqueous phase locks in this morphology to create stable biopolymer particles (see below). The extent of biopolymer particle deformation is related to the viscosity of the solution, the shear rate, the particle diameter, and the interfacial tension (van Puyvelde and others 2002).

Microfluidics is a relatively new technique that utilizes precisely shaped microchannels to create uniform droplets (Seo 2005; Zhang and others 2007; Oh and others 2008). Microchannels are often created using lithographic techniques (see above) to obtain nanoscale diameters. During operation, dispersed phase flow creates pressure at the tapered ends of the channels and forces the solution into specific droplet shapes. Droplets can be expelled into a bulk carrier solution or alongside carrier solution from a parallel microchannel. Utilization of co-eluting parallel streams allows for a purely mechanical atomization effect, giving droplet sizes in the tens of microns (Gañán-Calvo and Barrero 1999). By selecting the appropriate stabilization method (see below) the morphology and shape of the droplet can be retained (Oh and others 2006; Zhang and others 2007; Oh and others 2008). Often, microspheres are on the micron-scale (Nisisako and others 2004). The great advantage of microfluidics is the production of highly reproducible droplet sizes (Utada and others 2005) and the capability of producing multiple-phase systems (Okushima and others 2004; Utada and others 2005).

Cross-Linking Biopolymer Particles

  1. Top of page
  2. Abstract
  3. Introduction
  4. Biopolymer Particle Properties
  5. Structural Design Principles
  6. Biopolymer Particle Formation Methods
  7. Cross-Linking Biopolymer Particles
  8. Conclusions
  9. References

In many situations the initial biopolymer particles formed are not physically stable and are highly prone to coalescence and phase separation, for example, when using aggregative or segregative methods. Consequently, it is necessary to cross-link the biopolymers within the particles in order to increase their stability. Cross-linking can be carried out using various physical, chemical, and enzymatic approaches depending on the specific characteristics of the biopolymers involved.

Physical cross-linking

Physical cross-linking refers to any process of polymer–polymer interaction that adds stability without covalent bonding. This includes ionic cross-linking, hydrogen bonding, hydrophobic bonding, and others. These methods may be used for the induction of gelation processes among already formed discrete biopolymer particles.


Temperature changes may either strengthen or weaken the interactions holding biopolymer molecules together, depending on the nature of the dominant forces involved. Typically, hydrophobic forces are strengthened with increasing temperature, hydrogen bonding is weakened, and entropy effects are increased (“The Cement: Physicochemical Interactions”). Therefore, hydrophobic-driven association tends to increase with heating (heat-set gelation), whereas hydrogen-bonding driven association tends to increase with cooling (cold-set gelation) (Burey and others 2008). The changes in biopolymer association with temperature may be either reversible or irreversible, which may be important for their functional application. In the following sections, we will discuss how temperature changes can be used to cross-link various kinds of biopolymer particles.

Heat-set gelation is widely used to cross-link globular proteins, such as many milk, soy, and egg proteins. The majority of nonpolar amino acids are buried within the interior of native globular proteins so as to minimize the hydrophobic effect (Brownlow and others 1997). The application of thermal energy promotes globular protein unfolding so that hydrophobic residues are exposed to the aqueous environment (Belloque and Smith 1998). Consequently, there is an increase in the hydrophobic attraction between the protein molecules, which may be sufficiently strong to promote protein aggregation (Carrotta and others 2001; Donald 2008). Once globular protein molecules have been brought into close proximity through hydrophobic association, they may be further held together by formation of covalent linkages, such as disulfide linkages (Lametti and others 1996; Verheul and others 1998; Hoffmann and van Mill 1999). Heat-set gelation of globular proteins tends to be irreversible. That is, once the aggregates are formed at higher temperatures they remain intact when the system is cooled below the thermal denaturation temperature.

Heat-set gelation may also be used to cross-link polysaccharides with some hydrophobic character, such as methyl cellulose (Joshi and Lam 2005). Methyl cellulose is soluble in aqueous solutions at relatively low temperatures, but it tends to aggregate when the temperature is increased to above approximately 40 °C, which has been attributed to an increase in the hydrophobic attraction (Sarker and Walker 1995; Joshi and Lam 2005). This type of heat-set gelation tends to be reversible, once the system is cooled below the thermal aggregation temperature the molecules tend to dissociate.

Cold-set gelation systems are typically stable at relatively low temperatures, but they tend to dissociate when heated above a critical temperature. This type of gelation mechanism usually involves those biopolymers that are capable of forming helical regions that associate with each other through hydrogen bonding upon cooling such as gelatin (Mackie and others 1998; Izmailova and others 2001) and carrageenan (Piculell 1995; Ikeda and others 2001). Globular proteins can also be induced to form cold-set gels. In this case, a globular protein solution is heated under conditions where the protein molecules unfold and form linear filaments, but macroscopic gelation does not occur. This protein solution can then be made to gel at ambient temperatures by altering the solution conditions to reduce the electrostatic repulsion between the filaments, for example, by adjusting the pH close to the pI or by increasing the ionic strength (Ju and Kilara 1998; Alting and others 2003). Globular proteins such as β-lactoglobulin, BSA, and soy proteins can form cold-set gels (Bryant and McClements 1998).

Among mixed biopolymer systems that segregate, decreased temperatures can be utilized to favor separation, which in turn increases biopolymer interaction within the dispersed phase (Loren and others 2001). Increased biopolymer interactions within these excluded volumes result into gel-like matrices. Subsequent heating may be utilized to further solidify these compact phases into particulates. Uniquely shaped particulates have been created in this fashion using gelatin, gellan, alginate, and carrageenan (Wolf and others 2000).

Salt incorporation often has an important influence on thermal aggregation by reducing biopolymer charge and electrostatic interactions (“The Cement: Physicochemical Interactions”). As the ionic strength of a solution increases the electrostatic forces are progressively screened, which means that other interactions become more important, for example, hydrophobic forces or hydrogen bonding. For instance, methyl cellulose aggregation occurs at lower temperatures through hydrophobic attraction with increasing ionic strength (Joshi and Lam 2005). Also, low levels of salts are required for gelation of carrageenan systems, as some sulfate repulsions must be overcome for full association (Piculell 1995). In the aggregation of β-lactoglobulin, the extent of thermal aggregation varies depending on the concentration of added salts (Bryant and McClements 1998). This ionic strength effect is more apparent farther away from the protein's isoelectric point (Verheul and others 1998). Generally, salt increases aggregation when electrostatic repulsion plays a more dominant stabilizing role (that is, pH ≠ pI or pKa).

Effects of pH are important during thermal aggregation, especially among proteins. As mentioned, disulfide interactions dominate at higher pH, while hydrophobic interactions become more important closer to the pI, leading to the formation of more linear/fibrillar aggregates or particulate aggregates, respectively (see above). Between neutral pH and the pI, a mix of disulfide linkages and hydrophobic aggregation during heating may create useful particulate structures. Small suspended particulates have been formed in this manner using whey proteins (Ju and Kilara 1998) and β-lactoglobulin (Schmitt and others 2007). These particulates could be covered with a complexing polysaccharide layer, which showed good stability (Santipanichwong and others 2008).

Addition of interacting components may also limit the extent of thermal aggregation. Certain proteins, such as κ-casein (Donato and Dalgleish 2006) and soy (Roesch and Corredig 2005), interact with whey proteins upon heating and reduce aggregation. They could accomplish this by forming disulfide cross-links, thereby terminating further interaction. Polysaccharides are also known to associate with proteins (see above) and to alter their denaturation profiles. Protein aggregation does seem to be limited during heating when complexed with a polysaccharide. For instance, complexes of ovalbumin and chitosan heated to 80 °C at pH 5.45 formed defined particulates with high pH stability (Yu and others 2006). Similar results were found for complexes of β-lactoglobulin/chitosan (Hong and McClements 2007) and β-lactoglobulin/sugar beet pectin heated at pH 5.0 (Jones and others 2009a). Stability of these particles to pH changes has been improved by immersion in medium ionic strengths (Jones and McClements 2008).

Mineral ions and pH

Changes in pH and addition of mineral ions may be used to promote biopolymer association through alterations in electrostatic interactions. As discussed above, electrostatic interactions between oppositely charged biopolymers are used to assemble biopolymer particles in associative systems. The charge on proteins can be made to change from positive to negative by adjusting the pH from above to below their isoelectric point. Hence, the tendency for aggregation to occur can be controlled by controlling solution pH.

The propensity for biopolymers to associate with each other in solution depends on the balance of attractive and repulsive forces acting between them. Association tends to occur when the attractive forces dominate over the repulsive forces. In many systems containing charged biopolymers, the biopolymer molecules are prevented from aggregating because they have similar charges and there is an electrostatic repulsion acting between them. The addition of oppositely charged ions to the system can promote biopolymer association by screening the electrostatic interactions, or by acting as a salt bridge. For example, calcium ions (Ca2+) are frequently used to cross-link anionic polysaccharides such as pectins, alginates, or carrageenans (Williams 2007). Potassium ions (K+) are used to cross-link anionic carrageenans through the formation of an “egg-box” structure (Williams 2007). Mineral ion addition may be used to cross-link biopolymer particles in various ways (Burey and others 2008). The most common approach is to extrude the biopolymer particles into a salt bath using micro-channels. An alternative method is to utilize slow-releasing salt devices to initiate cross-linking only after biopolymer particles have been formed by phase separation. An example of such an approach is the addition of calcium carbonate salts, which slowly release calcium ions upon acidification. Other slow-release devices include diffusion setting systems, such as osmotic filters or W/O/W emulsions (Burey and others 2008).


Hardening methods can be employed to reduce solvency of the particulates to a varying degree. Two methods that generally follow this pattern are use of co-solvents and drying. Co-solvents’ main effect is to alter the solvent characteristics, generally through the dielectric constant. Reduced dielectric constant can diminish repulsive electrostatic interactions between proteins, leading to a net attraction between biopolymer chains. Drying can be performed through spray-drying (see above) or other methods (such as freeze-drying). These methods remove the water quickly, leaving an intact dry particulate. The speed and flow are important factors in reducing particle aggregation or destruction.

Some desolvation methods have been compared on a coacervate encapsulation system (Lamprecht and others 2001). Techniques involved were spray-drying, ethanol addition, and cross-linking by dehydroascorbic acid (see above). Spray-drying was shown to create the smallest particles, yet gave less oxidation protection and encapsulation efficiency.

Chemical cross-linking

Chemical cross-linking refers to the introduction of a covalent linkage between polymer functional groups. For instance, chemical agents may act as a bridge between similar or dissimilar amino acids (Andrea Sinz 2003). They may also be used to directly bond 2 amino acids together. Functional groups capable of chemical bonds include amines, thiols, hydroxyl groups, and phenyl rings.


Under certain conditions, esters and amides are capable of undergoing interchange reactions in aqueous solutions. Hence, certain carboxylate-containing compounds may interact with amines or hydroxyls to create cross-linking bridges.

Genipin, a compound of recent interest, is able to form physical links between biopolymers containing amines. Genipin is a naturally occurring heterocyclic compound derived from Genipa americana. Primary amine groups located on biopolymers attack at either the α- or β-carbon of the genipin ester (Mi and others 2000). While attachment at the α-carbon forms a simple amide linkage, attack at the β-carbon causes ring cleavage and further cross-linking capability. Examples of biopolymers that have been cross-linked using genipin include chitosan, BSA, soy protein, and gelatin (Butler and others 2003). Diffusion (Mi and others 2002) and concentration (Bigi and others 2002) of genipin are both factors in the rate and extent of cross-linking. Reaction kinetics increase with increasing amine content and accessible conformational structures of biopolymer substrates (Butler and others 2003). Genipin's major drawback is the production of blue pigments via free-radical polymerization of amines in the presence of oxygen. However, genipin is considered to be less cytotoxic than other cross-linking agents, such as glutaraldehyde (Mi and others 2001).

Maillard reaction

Maillard reactions chemically link aldehydes and amines through a well-established oxidation-reduction pathway (Oliver and others 2006). Linkages are caused by a series of imido- and redox reactions, which are favored at higher pH values. For common protein and sugar products, the ultimate products are a wide array of aryl composites and Strecker degradation products, providing both desirable flavor components and undesirable side-products. Polysaccharides, lacking adequate reducing ends, are generally incapable of significant cross-linking. Dextran and whey protein have been conjugated, however (Akhtar and Dickinson 2003; Zhu and others 2008), through the free reducing ends by dry-heating for prolonged periods. These Maillard conjugates have found use as emulsifiers and gelling agents. Another useful application of Maillard reactions is to cross-link protein components with mono- and disaccharides to create novel structures (Rich and Foegeding 2000). The Maillard reaction could also be used to cross-link proteins and polysaccharides within the biopolymer particles formed by spray- or freeze-drying. After preparation of the biopolymer particles by drying, the system can then be subjected to dry-heating to promote the Maillard reaction.

Aldehyde reactions

Aldehydes, such as glutaraldehyde and formaldehyde, can be used to chemically cross-link protein particulate systems. These highly reactive aldehydes may form a variety of compounds, such as acetals, cyanohydrins, and oximes, with common functional groups (McGregor and others 2006). This reactivity extends to most biological materials, inducing either cell death (when cytotoxic) or mutagenesis (when carcinogenic). Unfortunately, even incorporation of 3 ppm was found to induce cytotoxic effects to human fibroblasts (Speer and others 1980). Formaldehyde and glutaraldehyde, although not food-grade, are used as processing aids in the cross-linking of coacervate particles (Junyaprasert and others 2001; Thimma and Tammishetti 2003; Lazko and others 2004b; Weinbreck and others 2004a). A number of food-grade alternatives are being sought to replace these highly effective, yet potentially toxic, compounds.

Enzymatic methods

Enzymes can be utilized to catalyze specific cross-linking reactions involving biopolymer particles in food-grade systems. They are particularly useful in applications where alternative methods might cause damage to some encapsulated component. For instance, sulfhydryl or phenolic residues can be oxidatively cross-linked using high levels of gaseous oxygen (Strauss and Gibson 2004), but this would be deleterious to high-value lipids or phenolics. The use of specific enzymes avoids such problems. Also, amide cross-links, difficult to control by general chemical means, may be formed with high specificity by enzymes.

Transglutaminase has been widely used to cross-link various types of proteins in foods, and an excellent review on its applications can be found in the literature (de Jong and Koppelman 2002). This enzyme functions by a nonoxidative transamidation between glutamine and lysine, whether intra- or intermolecularly. Transglutaminase is commonly produced from bacterial sources (Ando and others 1989), although they are found in nearly all living organisms. Transglutaminase has been used to produce cross-linked protein films from gelatin (Lim and others 1999), egg-white (Lim and others 1998), gluten (Larre and others 2000), soybean, and whey proteins (Su and others 2007). Cross-linking has also been performed at the monomeric scale to produce thermally resistant components (Yildirim and Hettiarachchy 1997). The extent of cross-linking can be controlled by changes in pH, enzyme inhibitors, or heating (de Jong and Koppelman 2002).

Laccase is a copper-containing oxidase obtained from fungal sources, and its characteristics and capabilities have recently been reviewed (Couto and Herrera 2006; Widsten and Kandelbauer 2008). Laccase functions by cross-linking phenoxyl-containing polymers through oxidation of phenol to a benzylperoxy intermediate in the presence of dispersed oxygen. It has been found to oxidize the amino acids tryptophan, cysteine, and tyrosine within food proteins (Mattinen and others 2005, 2006). Laccase is a large enzyme (MW about 70000), and has limited access to tightly grouped polymeric structures, such as core residues within globular proteins. Laccase, along with other peroxidases, can be used to cross-link ferulic acids in food biopolymers. Ferulic acid is a hydroxycinnamic acid found in various plant materials with possible health benefits (Mathew and Abraham 2004). Dimers, and possibly trimers, can be formed through the enzymatic actions of peroxidases or laccases on ferulic acid side-chains (Ward and others 2001; Minussi and others 2002; Couto and Herrera 2006). Ferulic acid cross-linking occurs naturally within sugar beet pectin pulp and maize brans (Saulnier and Thibault 1999), as plant matter has oxidative enzymes and exposure to an oxidative environment. Numerous studies exist on laccase-induced ferulic acid cross-linking in the formation of strong gels (Norsker and others 2000; Vansteenkiste and others 2004; Carvajal-Millan and others 2005). Recent work has also demonstrated the successful use of this cross-linking combination with a multi-layer β-lactoglobulin-sugar-beet-pectin emulsion (Littoz and McClements 2008). Despite these promising results, industrial application might prove challenging. Ferulic acid esterification is believed to be irregular (Mathew and Abraham 2004), and the action of the oxidase enzyme is rate-limiting.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Biopolymer Particle Properties
  5. Structural Design Principles
  6. Biopolymer Particle Formation Methods
  7. Cross-Linking Biopolymer Particles
  8. Conclusions
  9. References

Particulate systems can be prepared from common food biopolymers, such as proteins and polysaccharides, using a variety of techniques. At present, our understanding of how to create biopolymer particles with specific functional attributes is still rather limited, and a better understanding of structure–function relationships is required. Identification of the most appropriate ingredients and conditions required to create a desirable particulate requires knowledge of the molecular and functional characteristics of the biopolymers used, as well as of the physicochemical mechanisms underlying particle formation and cross-linking. Limitations exist in the food industry concerning the biopolymer types, physicochemical mechanisms, and processing operations that can be economically utilized.

Future studies on particulate systems in food products should be focused on both novel techniques and refining of current methods. Many of the techniques that have been developed for particulate creation in other industries (such as pharmaceuticals) are currently unsuitable for use in food systems, either due to non-food-grade materials or high operational costs. Most likely, new techniques can be developed based on existing knowledge of biopolymer interactions and structures. Even currently accepted techniques can be retuned for the creation of smaller, more reproducible particulate structures. Future successes in food particulate structures will require knowledge of physical forces and a creative assembly of available techniques.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Biopolymer Particle Properties
  5. Structural Design Principles
  6. Biopolymer Particle Formation Methods
  7. Cross-Linking Biopolymer Particles
  8. Conclusions
  9. References