Journal of Colloid and Interface Science xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science www.elsevier.com/locate/jcis

Structuring of colloidal particles at interfaces and the relationship to food emulsion and foam stability Eric Dickinson School of Food Science and Nutrition, University of Leeds, Leeds LS2 9JT, UK

a r t i c l e

i n f o

Article history: Received 12 August 2014 Accepted 24 September 2014 Available online xxxx Keywords: Emulsion stability Colloidal interactions Depletion flocculation Thin films Polydispersity Aerated emulsions

a b s t r a c t We consider the influence of spherical colloidal particles on the structure and stabilization of dispersions, emulsions and foams. Emphasis is placed on developments in the use of the methods of liquid state theory and computer simulation to understand short-range structuring of concentrated colloidal dispersions and ordering of particle layers near surfaces and within liquid films. Experimental information on the structuring of surfactant micelles and caseinate particles in thin liquid films is described, including an assessment of the effect of particle polydispersity on depletion interactions and kinetic structural stabilization. We specifically discuss the relevance of some of these structural concepts to the stability of food colloids. Ó 2014 Elsevier Inc. All rights reserved.

Contents 1. 2. 3. 4. 5. 6. 7.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structuring of particles in dense systems and near hard surfaces Depletion flocculation in food colloids . . . . . . . . . . . . . . . . . . . . . . Colloidal structuring in thin films . . . . . . . . . . . . . . . . . . . . . . . . . . The polydispersity factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aerated emulsions and emulsion-stabilized foams . . . . . . . . . . . . Future outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . .

1. Introduction For budding investigators aspiring to establish a body of work capable of recognition by their academic peers as both rigorous and significant, the fields of food science and food technology might be considered unpromising areas of research. Pragmatic justification for this view would have its origin in the well-founded perception of the ‘messiness’ of food systems, which presumably conspires to inhibit the systematic planning and execution of traditional hypothesis-driven experimental investigation and associated theoretical analysis. One way to try to get around the problem of the compositional and structural complexity of foods is to adopt a strongly reductionist E-mail address: [email protected]

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

00 00 00 00 00 00 00 00

approach. This involves the specification and systematic investigation of much simpler model systems that are perceived to contain the key structural attributes or essential functional ingredients of real food products. One then argues intuitively [1] that the physicochemical behaviour of the real food product is essentially similar in certain crucial respects to the well-chosen model system. A successful exposition of this approach is to be found in the research work of Professor Darsh Wasan and his colleagues on the structuring of colloidal particles in thin films, foams and emulsions [2–5]. The outstanding stabilizing properties of familiar natural ingredients such as casein, gelatin, albumin and gum arabic were recorded by the pioneering colloid scientists. According to one of the earliest quantitative measures of the ability to act as a ‘protective agent’, published back in 1901 [6], the food proteins casein and gelatin were assigned low values of the so-called ‘gold number’

http://dx.doi.org/10.1016/j.jcis.2014.09.080 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.

Please cite this article in press as: E. Dickinson, J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.09.080

2

E. Dickinson / Journal of Colloid and Interface Science xxx (2014) xxx–xxx

(Goldzahl) in recognition of their impressive effectiveness as colloid stabilizers. In contrast, other natural polymers like starch and dextrin (‘non-stabilizers’) were placed at the opposite end of the scale. Based on such information, the essential principles of colloid science were gradually established within the physical chemistry community in the early years of the last century. Nevertheless, the relevance of colloid science to food technology was not properly recognized until after the Second World War. Of particular importance in this connection was a pioneering paper published by Hauser [7] in the month of this author’s birth (April 1948) entitled ‘‘The Importance of Colloid Chemistry in Food Technology’’. Hauser’s wide-ranging article (including a record of lively symposium-based discussion) contains straightforward explanations of the basic colloid science principles underlying the main areas of food production. In particular, attention is directed towards the putative role of functional ingredients involved in the formulation and stabilization of a familiar culinary emulsion—mayonnaise. Hauser’s article also contains a quotation from a certain Mr. J.J. Corran, the then Chief Chemist of J. & J. Colman Ltd (UK), which seems relevant today: ‘‘A complete understanding of the digestion of our food, particularly of fats, is dependent, amongst other things, on a knowledge of emulsions.’’ Indeed, the investigation of the stability properties of emulsions, both before and after eating, does still remain a vibrant area of research activity within the food colloids community. The year 1948 could also be regarded as the start of the modern colloid science era with the publication of the classic treatise of Vervey and Overbeek on the stability of lyophobic colloids [8]. This monograph describes how two independently derived concepts— interacting electrical double-layers and van der Waals dispersion forces—can be combined into a general theory of colloid stability (the DLVO theory) which is both rigorous and predictive [8]. Unfortunately, however, in the field of food colloids we have come to recognize that the range of usefulness of the DLVO theory is rather limited [1]. One notable situation of successful applicability, though, is the predicted ionic strength dependence of the flocculation stability of model emulsions prepared with globular proteins [9]. But the failure of the DLVO theory more broadly in the food colloid arena is a consequence of the molecular and structural complexity of food ingredients. That is, multicomponent mixtures of adsorbed biopolymers and particles are typically involved in food emulsion stabilization, and their combined presence at interfaces leads to a synergy of polyelectrolyte and steric contributions to the overall droplet– droplet interactions. In these complex multicomponent systems, the structure and dynamics of the stabilizing interfacial layer is affected by a range of molecular phenomena such as complexation, self-assembly and coacervation, as well as by the competitive adsorption of biopolymers and small-molecule surfactants [10–15]. A further complication is that the properties of food emulsions (and foams) are affected also by the presence of non-adsorbed macromolecules and particles located in the continuous medium between pairs of dispersed oil droplets (or gas bubbles). The attractive interdroplet forces induced by the presence of these nonadsorbed species (protein particles, hydrocolloids, surfactant micelles) are responsible for the commonly encountered phenomenon of depletion flocculation [9,16,17]. Furthermore, the structuring of these non-adsorbed species in the narrow gap between fat droplet surfaces may have a positive stabilizing influence on food emulsions [2,4,18]. The role of the interparticle interactions on the structuring of concentrated colloidal dispersions requires a knowledge of the statistical mechanical concepts developed originally for simple liquids [19]. An influential contribution to this approach was made by Professor Wasan and his colleagues in modelling the structuring of colloidal particles at liquid interfaces, and in characterizing the essentials of particle-based structure formation in model food-related systems.

This article reviews how the principles of liquid state theory and computer simulation have proved useful for describing structuring of colloidal dispersions at surfaces and in thin films. We show how such concepts have provided mechanistic insight into the stability properties of emulsions and foams containing nanoparticles, surfactant micelles, and protein aggregates. While limitations of space require that the discussion here be restricted to the role of non-adsorbed particles, one should be aware that there exist many other situations for which the application of statistical modelling and/or computer simulation is a valuable tool for describing the properties of food colloid systems. Examples include the study of structure and rheology of aggregated particle gels [20–24], the reversible heteroflocculation of mixed emulsions [25], adsorbed layer structure and interactions of biopolymercoated surfaces [26–28], surface rheology of globular protein layers [29,30], and the competitive displacement of proteins from interfaces by surfactants [12,31].

2. Structuring of particles in dense systems and near hard surfaces The theoretical representation of concentrated colloidal dispersions relies on exploiting the analogy with simple liquids [19]. According to the classical theory of van der Waals, the structure of a dense liquid is determined primarily by the short-range repulsive interactions between the molecules, with the intermolecular attractive forces merely providing the uniform background potential in which the molecules move. Consequently, an infinitely repulsive particle-based system, the hard-sphere model, provides a reliable reference system for describing the properties of real

Fig. 1. Two-dimensional representation of the structure of a concentrated dispersion of spherical colloidal particles. The graph shows a plot of the radial distribution function, g(r), where r is the centre-to-centre distance between a test particle and the neighbouring particles. Peaks labelled (a) and (b) indicate first and second coordination shells.

Please cite this article in press as: E. Dickinson, J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.09.080

E. Dickinson / Journal of Colloid and Interface Science xxx (2014) xxx–xxx

liquid systems. Hence the free energy of a system of N molecules in volume V at temperature T is given by [32]

A ¼ A0 þ ðN2 =2VÞ

Z r

1

uðrÞg 0 ðrÞ4pr2 dr;

ð1Þ

where A0 is the free energy of the hard-sphere system of diameter r, u(r) is the intermolecular pair potential, and g0(r) is the radial distribution function of the hard-sphere liquid. The function g(r) is defined as the average local molecular density at distance r from a test molecule, divided by the overall density N/V. It is normalized to approach unity at large separations and to approach zero at very close separations; the hard-sphere distribution function g0(r) is exactly zero for r < r. For a dense system of colloidal particles interacting with pair potential u(r), the form of the radial distribution function g(r) provides a convenient statistical description of its time-averaged structure. Fig. 1 illustrates how the decaying oscillatory features of g(r) can be interpreted in terms of the successive coordination shells of particles around the test particle in a dense liquid-like dispersed system [33]. The effective interaction between a pair of colloidal particles A and B in a concentrated system is known as the potential of mean force W(r). It is related to the radial distribution function via the Boltzmann relation:

WðrÞ ¼ kT ln gðrÞ:

ð2Þ

Therefore particles A and B are most likely to be located at separation r where the value of W is large and negative. The quantity W(r) is a free energy function: it is dependent not only on the interaction energies of A and B with all the other colloidal particles in the system (including the A–B interaction), but also on their time-averaged (ensemble-averaged) statistical arrangement (the entropic contribution). As well as other colloidal particles, the potential of mean force may include the structural effects of additional entities such as polymer molecules or surfactant micelles. The potential of mean force is a useful theoretical concept because it allows us to reduce

3

the number of effective entities in a complex colloidal system by averaging over one or more of the other components present [33]. Because the energetic and entropic contributions to W(r) have different temperature sensitivities, the potential of mean force is temperature sensitive. Therefore it is important to be aware that the free energy W(r) is essentially a different kind of quantity from the interparticle pair energy u(r). Nevertheless, the formal distinction between the two can be conveniently overlooked in certain situations, so long as it is recognized that W(r) is not an intrinsic property of the particles alone, but is dependent on the state variables (T, p, etc.) and on the concentrations of the various other species in the system (nanoparticles, polymers, micelles, etc.). Suppose we consider a pair of large particles immersed in a dispersion of small hard-sphere particles. When the separation between the large particle surfaces is less than the diameter of the small spheres, a net attractive force is induced between the large particles as a consequence of the osmotic pressure difference between the bulk phase and the interparticle gap region which is depleted of small hard spheres [34]. This attractive depletion interaction is indicated by the short-range part of the potential of mean force between the large particles as illustrated in Fig. 2. For increasing surface-to-surface separations, the form of the potential of mean force is oscillatory in character, reflecting the structuring of the dense dispersion of hard spheres in the gap between the large particles [35]. The hard walls of the large particles induce an ordering of the small spheres into discrete layers, and for high particle concentrations this layering can propagate over several particle diameters. At a surface-to-surface separation just exceeding the hard-sphere diameter there is a structural stabilization barrier, and the height of this structural barrier increases with the concentration of the small particles. It was recognized by Wasan and colleagues [2,4,36] that the presence of such a kinetic barrier can have significant implications for the stability of concentrated systems of dispersed colloidal particles (or emulsion droplets) with respect to their coagulation (or coalescence).

Fig. 2. Schematic presentation of the structuring of a dense dispersion of small hard-sphere particles between a pair of large (hard) particles (diameter D). For surface-toseparation below the hard-sphere particle size, the potential of mean force exhibits an attractive depletion well as described by Asakura and Oosawa [33]. At larger separations there is a structural stabilization barrier and an oscillatory decay. Reproduced with permission from Ref. [4].

Please cite this article in press as: E. Dickinson, J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.09.080

4

E. Dickinson / Journal of Colloid and Interface Science xxx (2014) xxx–xxx

3. Depletion flocculation in food colloids One of the most common instability mechanisms exhibited by a food oil-in-water emulsion is depletion flocculation caused by the presence of a small amount of water-soluble polymer (hydrocolloid) [33,37]. Depletion flocculation may also be induced by nonadsorbing entities such as surfactant micelles [38] or casein(ate) nanoparticles [17,39]. The depletion contribution to the droplet pair interaction, udep(d), as a function of the surface-to-surface separation d, is related to the osmotic pressure difference between the interdroplet gap region depleted of polymer molecules (or nanoparticles) and the bulk phase beyond the interdroplet gap region. For the ideal case of two large spheres of radius R in a fluid of small spheres of diameter ds, the Asakura–Oosawa depletion potential takes the form [34]

udep ðdÞ ¼ ð3kTR/s =ds Þðds  dÞ;

ðd < ds Þ

ð3Þ

where /s is the volume fraction of the small spheres. While this simple relationship neglects completely any interdroplet structuring beyond d = ds, it provides a useful expression for estimating the approximate strength of the depletion pair interaction. We note that depletion forces are relatively weak compared with some other kinds of interdroplet forces (electrostatic, polymer bridging, etc.). For instance, for droplets of diameter 1 lm in a 2 vol% solution of nanoparticles of diameter 10 nm, Eq. (3) predicts a maximum attraction at contact (d = 0) of udep  3kT. A widely used ingredient for making and stabilizing dairy-type emulsions under neutral pH conditions is sodium caseinate. In situations where most of the caseinate emulsifier present during homogenization adsorbs to give a saturated protein monolayer, the resulting emulsions are Newtonian fluids, they contain small uniform droplets, and they possess excellent stability due to a combination of electrostatic and steric mechanisms [40,41]. But when homogenization is carried out at high caseinate/oil ratios, so that a considerable proportion of the protein remains unadsorbed, the systems exhibit shear-thinning rheology and instability with respect to creaming and macroscopic phase separation [42,43]. This instability associated with the presence of excess sodium caseinate in the continuous phase is attributed [17,42] to reversible depletion flocculation induced by non-adsorbed caseinate nanoparticles (‘sub-micelles’). The strength of the depletion interaction in any particular formulation is dependent on the number concentration of the nanoparticles (related to osmotic pressure) and their average size (related to depletion layer thickness). A detailed theoretical analysis has demonstrated [39] that the mean size of the nanoparticles present in sodium caseinate systems under low ionic strength conditions is rather close to the optimum size of particles necessary to induce the maximum depletion attraction between emulsion droplets. It is noteworthy that depletion flocculation is not found in emulsions prepared with extensively aggregated dairy protein ingredients such as calcium caseinate, whey protein concentrate, or skim milk powder. Furthermore, the depletion flocculation observed in sodium caseinate emulsions at low ionic strength may be diminished or completely inhibited by changing the concentration and mean size of the nonadsorbed protein aggregates via an increase in ionic strength [44], the addition of calcium ions [45–47], or a reduction in pH [48]. 4. Colloidal structuring in thin films The thin liquid films between gas bubbles in a polyhedral foam are the key structural elements controlling stability with respect to bubble coalescence. Hence the acquisition of information on drainage, thinning and rupture of thin aqueous films is useful for understanding the mechanistic factors determining the lifetimes of

aqueous foams [49]. The properties of thin films containing dispersed particles are especially interesting to food colloid scientists, because it is well established [33,50] that aerated food systems may be either stabilized or destabilized by colloidal particles, depending on their concentration and their hydrophilic/hydrophobic character. Against this background, pioneering studies on plane parallel films by Nikolov and Wasan have shown [51,52] that there is a strong tendency for colloidal particles to form ordered layers in films when present at sufficiently high particle concentrations. Moreover it has been suggested [53] that this particle-induced ordering produces an additional stabilizing force in the film called ‘the structural force’. In following the evolving structure of the particle-laden film, it was discovered [51,52,54] that, as the liquid film drains to thicknesses of the order of a few particle diameters, further thinning occurs in a stepwise manner, i.e., by a process of stratification. Although the early experiments were carried out on surfactant micellar systems [51,55–57], the same kind of ordered layering and stratification was also observed with dispersions of solid silica particles or latices [51,57,58], as well as for sodium caseinate systems [18]. Moreover the observed behaviour has been clearly established as a general colloidal phenomenon [59] that is fully consistent with theory and computer simulations [60–62]. Fig. 3 illustrates experimental data [58] for the stepwise thinning of a curved film observed between two gas bubbles in the presence of 8 nm hydrophilic silica particles at effective particle volume fractions of / = 0.25 and 0.35. The nearly monodisperse particles exhibit a tendency to form distinct layers, and the evolving stratification is rationalized [58] on the basis of the transport of the colloidal particles from the film to the meniscus by a diffusive osmotic mechanism. The layering generates a steric barrier against bubble coalescence in bulk foams containing the same nanosized silica particles [3]. Fig. 3 shows that, as the particles become removed layer-by-layer from the film, there are abrupt changes in the measured film thickness. The width of each step transition corresponds to the effective particle size. Finally, at late observation times, particles are expelled from the last particle monolayer, and there develops a ‘black spot’ region: the film thickness then reduces to the equilibrium thickness of a Newton black film before ultimate rupture. It is observed [3,58] that the total number of stepwise film transitions increases with the particle concentration and decreases with increasing particle size. These generic filmthinning trends appear consistent with observations of the ‘foaminess’ of surfactant-free silica particle suspensions [3]. The microlayering of sodium caseinate nanoparticles (‘submicelles’) in thin foam films (and emulsion films) was reported by Koczo et al. [18]. The authors found that the stepwise film thinning behaviour was qualitatively similar to that found with surfactant micelles or silica particles. The width of the step transitions was determined to be approximately equal to the mean size of the caseinate nanoparticles (20 nm), and the number of steps observed was found to increase with the protein concentration. Another potential influence on film thinning kinetics is the complexation of the caseinate particles with surface-active lipids such as lecithin [63]. It was inferred [18] that the layering of the caseinate nanoparticles would provide a kinetic barrier to instability in emulsions and foams containing high concentrations of nonadsorbed sodium caseinate. As illustrated in Fig. 2, this depletion stabilization by solid spherical protein particles is predicted [53,64–66] to be oscillatory in nature with a spatial periodicity of the order of the size of the caseinate sub-micelles. The ordered assembly of protein layers in the thin film imparts a substantial disjoining pressure, and individual layers must be successively shed in order for the overall separation distance between adjacent bubbles or droplets to be fully diminished. Since this process involves cooperative diffusional transport over a structural

Please cite this article in press as: E. Dickinson, J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.09.080

E. Dickinson / Journal of Colloid and Interface Science xxx (2014) xxx–xxx

5

Fig. 3. Stratification and stepwise thinning of films containing silica nanoparticles (8 nm diameter) at two effective volume fractions: / = 0.25 (3 particle layers: 25a, 25b, 25c) and / = 0.35 (4 particle layers: 35a, 35b, 35c, 35d). The films observed at short times are labelled 25a and 35a. The films with black spots observed at long times are labelled 25c and 35d. Reproduced with permission from Ref. [58].

stabilization barrier (see Fig. 2), the inner primary nanoparticle monolayer is expected to be the slowest one to disappear. 5. The polydispersity factor A characteristic feature of any real colloidal system is the distribution of particle sizes. To describe the influence of polydispersity on the properties of dispersions of spherical particles, a useful generic theoretical form is the Schultz distribution [67]:

f ðrÞ ¼

 1þz   1 1þz ðz þ 1Þr : r2 exp Cðz þ 1Þ < r >

ð4Þ

(/D)(W/kT)

In Eq. (4), the average diameter is hri, and the degree of polydispersity is expressed in terms of the standard deviation (z + 1)–1/2, where z = 1 corresponds to a monodisperse system. In comparing the properties of polydisperse and monodisperse systems, one

d /

Fig. 4. Effect of polydispersity on the reduced interaction energy (/D)W (in units of kT) as a function of the surface-to-surface separation d (in units of < r>) between two large spheres of diameter D suspended in a hard sphere fluid of volume fraction / = 0.20: —, monodisperse; - - -, 30% polydisperse; – –, 50% polydisperse. Numbers in parentheses indicate particle number concentrations (in reduced units) for each degree of polydispersity. Reproduced with permission from Ref. [67].

should properly specify whether the comparison is made at constant particle number density or constant particle volume fraction [65]. From the experimental perspective, it is clearly more convenient to consider the case of constant particle volume fraction. One of the main consequences of polydispersity in a concentrated dispersion of hard spherical particles is the disruption of ordered packing arrangements. Whereas highly monodisperse assemblies may readily form ordered crystalline phases under favourable thermodynamic conditions, such behaviour is inhibited for most practical polydisperse colloids. Computer simulation of assemblies of spherical particles interacting with various types of colloidal forces has consistently demonstrated [68–72] that the presence of just a few percent polydispersity in the system’s particle-size distribution is sufficient to prevent the formation of an ordered crystalline state. This modest extent of polydispersity is, of course, routinely exceeded in colloids containing food-grade ingredients. Polydispersity of the depletant has a significant influence on the magnitude of the short-range depletion energy and on the height of any structural energy barrier [4,64,65]. Fig. 4 shows the theoretical results of Henderson and coworkers [67] for the interaction energy W between a pair of large hard spheres immersed in a fluid of small spheres at volume fraction / = 0.2. The change in W as a function of the surface-to-surface separation distance d of the large particles is calculated for a Schultz distribution with degrees of polydispersity of 0%, 30% or 50%. We can see that the effect of increasing the polydispersity is to reduce the depth of the attractive well at d = 0 and also the height of the repulsive structural peak at d hri. The practical implications of these calculations are twofold: firstly, the extent of depletion flocculation in the corresponding emulsion is expected to be greatest in the presence of a depletant of low polydispersity; and, secondly, the phenomenon of structural stabilization by non-adsorbing particles will be most effective when the structured particles are close to being perfectly monodisperse. In the case of surfactant micelles, this predicted influence of polydispersity on thin film stratification and stability has been confirmed experimentally [73].

Please cite this article in press as: E. Dickinson, J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.09.080

6

E. Dickinson / Journal of Colloid and Interface Science xxx (2014) xxx–xxx

Fig. 5. Representation of the effect of particle bidispersity on thin film stability: (a) large particle is pushed to wall; (b) water is sucked from small-particle-depleted region due to osmotic pressure difference (Posm); (c) film is locally reduced in thickness; (d) unstable thinner section of film ruptures rapidly; (e) spot formation appears in micrograph recorded during foam lamella thinning. Reproduced with permission from Ref. [3].

Another kind of polydispersity (strictly, pauci dispersity) occurs in foaming systems containing binary mixtures of particles of very different sizes. The presence of just a small fraction of larger particles can have a very pronounced destabilizing effect. For instance, it has been reported [3] that the addition of 2 vol% of 100-nm silica particles to an aqueous system of 8 vol% of 8-nm silica particles (a ratio of 1:7800 on a particle number basis) leads to a tenfold reduction in foaming ability. The reason for this dramatic effect is schematically illustrated in Fig. 5. As the film thickness approaches the size of the large particles during film thinning, they become pinned to the walls. This constrains their diffusivity while small particles continue to diffuse freely. Subsequent trapping of the large particles in the film produces regions rich in the large particles and devoid of small particles. The resulting osmotic pressure difference draws water from the region rich in the large particles, and the film around each large particle bends locally to reduce the film thickness. Finally the foam lamella in the deformed film region becomes highly unstable to rupture [3]. 6. Aerated emulsions and emulsion-stabilized foams Some food systems contain both emulsion droplets and gas bubbles. The stability implications of the combined presence of these two kinds of dispersed entities may be positive or negative depending on the conditions. On the one hand, a small quantity of fatty particles in a polyhedral foam is usually found to be detrimental to stability [74,75]. The destabilization of egg-white foam by a trace of egg yolk is an everyday example, as is the diminished foaming of whole milk (3 wt% fat) as compared to skim milk. On the other hand, a high concentration of emulsified oil droplets can substantially enhance foam stability by inhibiting film drainage through the formation of structured layers between gas bubbles and by means of droplet densification in Plateau borders [76]. These two mechanisms can retard the rates of bubble coalescence and disproportionation. Nevertheless, in order to confer a long shelf-life on an aerated food emulsion, it is generally considered necessary to generate a permanent network of emulsion droplets involving some kind of droplet aggregation mechanism [15,77–79]. Incorporating an approximately equal volume of gas bubbles into non-homogenized dairy cream of moderately high fat content (>30 wt%) converts the concentrated viscous oil-in-water emulsion into a stiff solid-like structure with a significant yield stress. A key structural requirement for efficient shear-induced aeration of dairy cream is that the milk fat be semi-crystalline [75,77]. The aerated system is stabilized by a network of clumped fat globules produced by the application of the localized shearing forces generated during

whipping at a temperature of 5–10 °C [80]. The underlying mechanism, known as partial coalescence, relies on the formation of liquid fat bridges between two or more droplets due to the penetration of solid fat crystals from one droplet into the liquid oil region of another [81]. This means that there should be sufficient liquid fat present to facilitate the shear-induced aggregation of droplets into a rigid network. At the same time, to prepare a whipped cream of the required texture and stability, the liquid fat content should not be so great as to cause excessive clumping of droplets, which would trigger emulsion phase inversion (churning into butter). Food technologists have successfully produced various kinds of imitation aerated emulsions having similar textural properties to whipped dairy cream. These typically contain other ingredients such as vegetable fats, added emulsifiers, hydrocolloids, and nondairy proteins. To develop such products effectively requires a proper understanding of the various physicochemical factors controlling the orthokinetic destabilization of emulsions [82]. These factors include the hydrodynamic shearing conditions [83], the colloidal interactions between the droplets [84,85], the state of crystallization of the fat [86–91], and the presence of small-molecule emulsifiers [77,92–96]. During shear-induced aeration there is some spreading of liquid fat on the bubble surfaces [97], and it is commonly inferred that this spreading increases the rate of emulsion droplet adsorption and aggregation [86,98]. Emulsifiers such as monoglycerides have two functional roles: they increase the sensitivity of emulsion droplets to shear-induced destabilization by partially displacing protein from the oil–water interface, and they optimize the morphology of dispersed fat crystals for the enhancement of partial coalescence. We may note that alternative mechanisms of stabilizing aerated emulsions apart from partial coalescence are also possible. One such strategy involves building flocculated layers of protein-coated droplets around the dispersed gas bubbles using the soft viscoelastic network structure of an acid-induced casein-stabilized emulsion gel [78,99,100]. Finally, in the context of this article, it is relevant to mention the recent increased interest amongst food researchers in the stabilization of emulsions and foams by adsorbed solid particles [101]. The underlying theoretical basis of this well-established mechanism, Pickering stabilization, has been discussed over the years by many researchers [102–105], including also Wasan and his collaborators [106]. But it is only fairly recently that the fabrication of food-grade nanoparticles and microparticles for the purposes of Pickering stabilization of food emulsions has received serious attention [107]. An essential feature of this mechanism is the formation of a coherent structural barrier of adsorbed particles at the surfaces of dispersed droplets (or bubbles). These Pickering particles may

Please cite this article in press as: E. Dickinson, J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.09.080

E. Dickinson / Journal of Colloid and Interface Science xxx (2014) xxx–xxx

take the form of protein aggregates or emulsion droplets. Hence protein-coated emulsion droplets may function as stabilizers of gas bubbles [101], and small emulsion droplets as stabilizers of larger oil droplets [108]. To enhance the stability and to control the overall rheological behaviour, these various dispersed entities are commonly aggregated into a gel-like network [15,109]. What this means in practice is that only a small proportion of the stabilizing particles are directly located at the oil–water or air–water interface. So any separation into discrete roles for adsorbed and non-adsorbed particles becomes a rather artificial concept in the context of a gel-like emulsion-stabilized foam. The ongoing challenge for the food scientist is to understand how the combined structuring of the mixed colloidal ingredients produces a food product having the desirable rheology, texture and shelf-life. 7. Future outlook The focus of this article has been on the structuring of colloidal particles at liquid interfaces and the relationship to emulsion and foam stability. It has been shown how valuable insight is derivable from statistical thermodynamic theories and computer simulations of model systems of hard spherical particles under equilibrium (or quasi-equilibrium) conditions. Building on this firm foundation, there is a strong recognition of the need to develop an equivalent level of understanding of the dynamics of structural change at liquid interfaces over a wide range of timescales. While it is now well established that the surface rheology of adsorbed layers plays a major role in determining the stability of colloidal systems containing dispersed droplets and bubbles, there is a need to establish, in a more quantitative way, how the mechanical properties of structured particle-based layers are influenced by the chemical nature of the constituent particles and by the strength and range of the interparticle interactions. Furthermore there is a clear requirement to move beyond the simple hard-sphere models considered here in order to represent more realistically the properties of interfaces in complex systems like foods, which typically contain irregularly shaped and deformable particles such as liquid crystals, surfactant micelles, liposomes, microgels, or protein fibrils. An informed appraisal of the recent expanding literature in the general area of soft matter physics and chemistry would appear to suggest that reasonably good progress towards this ambitious goal is beginning to be made. References [1] E. Dickinson, G. Stainsby, Colloids in Food, Applied Science, London, 1982. [2] W. Xu, A. Nikolov, D.T. Wasan, A. Gonsalves, R.P. Borwankar, J. Food Sci. 63 (1998) 183. [3] S.K. Bindal, G. Sethumadhavan, A.D. Nikolov, D.T. Wasan, AIChE J. 48 (2002) 2307. [4] D.T. Wasan, A.D. Nikolov, F. Aimetti, Adv. Colloid Interface Sci. 108–109 (2004) 187. [5] A. Nikolov, D. Wasan, Adv. Colloid Interface Sci. 206 (2014) 207. [6] R. Zsigmondy, Z. Anal. Chem. 40 (1901) 696. [7] E.A. Hauser, Food Technol. 2 (2) (1948) 144. [8] E.J.W. Vervey, J.Th.G. Overbeek, Theory of Stability of Lyophobic Colloids, Elsevier, Amsterdam, 1948. [9] E. Dickinson, Colloids Surf. B 81 (2010) 130. [10] E. Dickinson, Colloids Surf. 42 (1989) 191. [11] E. Dickinson, S.R. Euston, in: E. Dickinson (Ed.), Food Polymers, Gels and Colloids, Royal Society of Chemistry, Cambridge, UK, 1991, p. 132. [12] L.A. Pugnaloni, E. Dickinson, R. Ettelaie, A.R. Mackie, P.J. Wilde, Adv. Colloid Interface Sci. 107 (2004) 27. [13] D.G. Dalgleish, Food Hydrocoll. 20 (2006) 415. [14] E. Dickinson, Food Hydrocoll. 25 (2011) 1966. [15] E. Dickinson, J. Sci. Food Agric. 93 (2013) 710. [16] P. Jenkins, M. Snowden, Adv. Colloid Interface Sci. 68 (1996) 57. [17] E. Dickinson, M. Golding, Food Hydrocoll. 11 (1997) 13. [18] K. Koczo, A.D. Nikolov, D.T. Wasan, R.P. Borwankar, A. Gonsalves, J. Colloid Interface Sci. 178 (1996) 694. [19] W. van Megen, I. Snook, Adv. Colloid Interface Sci. 21 (1984) 119.

7

[20] B.H. Bijsterbosch, M.T.A. Bos, E. Dickinson, J.H.J. van Opheusden, P. Walstra, Faraday Discuss. 101 (1995) 51. [21] M. Whittle, E. Dickinson, Mol. Phys. 90 (1997) 739. [22] M. Whittle, E. Dickinson, J. Chem. Phys. 107 (1997) 10191. [23] E. Dickinson, J. Colloid Interface Sci. 225 (2000) 2. [24] E. Dickinson, Adv. Colloid Interface Sci. 199–200 (2013) 114. [25] E. Dickinson, J. Colloid Interface Sci. 356 (2011) 196. [26] E. Dickinson, V.J. Pinfield, D.S. Horne, F.A.M. Leermakers, J. Chem. Soc., Faraday Trans. 93 (1997) 1785. [27] E.L. Parkinson, R. Ettelaie, E. Dickinson, Biomacromolecules 6 (2005) 3018. [28] R. Ettelaie, A. Akinshina, E. Dickinson, Faraday Discuss. 139 (2008) 161. [29] C.M. Wijmans, E. Dickinson, Langmuir 14 (1998) 7278. [30] R. Ettelaie, L.A. Pugnaloni, E. Dickinson, J. Colloid Interface Sci. 287 (2005) 401. [31] R. Ettelaie, L.A. Pugnaloni, E. Dickinson, Colloids Surf. B 31 (2003) 149. [32] R.W. Zwanzig, J. Chem. Phys. 22 (1954) 1420. [33] E. Dickinson, An Introduction to Food Colloids, University Press, Oxford, 1992. [34] S. Asakura, F. Oosawa, J. Chem. Phys. 22 (1954) 486. [35] D. Henderson, J. Colloid Interface Sci. 121 (1988) 486. [36] W. Xu, A. Nikolov, D.T. Wasan, A. Gonsalves, R.P. Borwankar, J. Colloid Interface Sci. 191 (1997) 471. [37] E. Dickinson, Food Hydrocoll. 17 (2003) 25. [38] E. Dickinson, C. Ritzoulis, M.J.W. Povey, J. Colloid Interface Sci. 212 (1999) 466. [39] S.J. Radford, E. Dickinson, Colloids Surf. A 238 (2004) 71. [40] E. Dickinson, J. Dairy Sci. 80 (1997) 2607. [41] E. Dickinson, Int. Dairy J. 9 (1999) 305. [42] E. Dickinson, M. Golding, M.J.W. Povey, J. Colloid Interface Sci. 185 (1997) 515. [43] E. Dickinson, M. Golding, J. Colloid Interface Sci. 191 (1997) 166. [44] M. Srinivasan, H. Singh, P.A. Munro, Food Hydrocoll. 14 (2000) 497. [45] E. Dickinson, M. Golding, Colloids Surf. A 144 (1998) 167. [46] A. Ye, H. Singh, Food Hydrocoll. 15 (2001) 195. [47] E. Dickinson, M.G. Semenova, L.E. Belyakova, A.S. Antipova, M.M. Il’in, E.N. Tsapkina, C. Ritzoulis, J. Colloid Interface Sci. 239 (2001) 87. [48] C. Eliot, E. Dickinson, Int. Dairy J. 13 (2002) 153. [49] D.A. Edwards, H. Brenner, D.T. Wasan, Interfacial Transport Processes and Rheology, Butterworth-Heinemann, Stonheam, MA, 1991. [50] A. Prins, in: E. Dickinson (Ed.), Food Emulsions and Foams, Royal Society of Chemistry, London, 1987, p. 30. [51] A.D. Nikolov, D.T. Wasan, J. Colloid Interface Sci. 133 (1989) 1. [52] A.D. Nikolov, D.T. Wasan, Langmuir 8 (1992) 2985. [53] X.L. Chu, A.D. Nikolov, D.T. Wasan, J. Chem. Phys. 103 (1995) 6653. 105 (1996) 4892. [54] G.N. Sethumadhavan, A.D. Nikolov, D.T. Wasan, V.J. Srivastava, J.J. Kilbane II, T.D. Hayes, Ind. Eng. Chem. 42 (2003) 2634. [55] D.T. Wasan, A.D. Nikolov, L.A. Lobo, K. Koczo, D.A. Edwards, Prog. Surf. Sci. 39 (1992) 119. [56] A.D. Nikolov, D.T. Wasan, N.D. Denkov, P.A. Kralchevsky, I.B. Ivanov, Prog. Colloid Polym. Sci. 82 (1990) 1. [57] P.A. Kralchevsky, A.D. Nikolov, D.T. Wasan, I.B. Ivanov, Langmuir 6 (1990) 1180. [58] G.N. Sethumadhavan, A.D. Nikolov, D.T. Wasan, J. Colloid Interface Sci. 240 (2001) 105. [59] D.T. Wasan, A.D. Nikolov, P.A. Kralchevsky, I.B. Ivanov, Colloids Surf. 67 (1992) 139. [60] A.D. Nikolov, P.A. Kralchevsky, I.B. Ivanov, D.T. Wasan, J. Colloid Interface Sci. 133 (1989) 13. [61] A. Trokhymchuk, D. Henderson, A. Nikolov, D.T. Wasan, Langmuir 17 (2001) 4940. [62] A. Trokhymchuk, D. Henderson, A. Nikolov, D.T. Wasan, J. Phys. Chem. B 107 (2003) 3927. [63] F.A. Husband, P.J. Wilde, J. Colloid Interface Sci. 205 (1998) 316. [64] X.L. Chu, A.D. Nikolov, D.T. Wasan, Langmuir 10 (1994) 4403. [65] X.L. Chu, A.D. Nikolov, D.T. Wasan, Langmuir 12 (1996) 5004. [66] W. Xu, A. Nikolov, D.T. Wasan, AIChE J. 43 (1997) 3215. [67] D. Henderson, D.T. Wasan, A. Trokhymchuk, Mol. Phys. 102 (2004) 2081. [68] E. Dickinson, R. Parker, M. Lal, Chem. Phys. Lett. 79 (1981) 578. [69] P.G. Bolhuis, D.A. Kofke, Phys. Rev. E 54 (1996) 634. [70] E.E. Phan, W.B. Russel, J.X. Zhu, P.M. Chaikin, J. Chem. Phys. 108 (1998) 9789. [71] D.A. Kofke, P.G. Bolhuis, Phys. Rev. E 59 (1999) 618. [72] N.B. Wilding, P. Sollich, J. Chem. Phys. 133 (2010) 224102. [73] A.D. Nikolov, D.T. Wasan, Colloids Surf. A 128 (1997) 243. [74] A. Prins, in: E. Dickinson, G. Stainsby (Eds.), Advances in Food Emulsions and Foams, Elsevier Applied Science, Barking, 1988, p. 91. [75] E. Dickinson, in: B.P. Binks, T.S. Horozov (Eds.), Colloidal Particles at Liquid Interfaces, University Press, Cambridge, UK, 2006, p. 298. [76] K. Koczo, L.A. Lobo, D.T. Wasan, J. Colloid Interface Sci. 150 (1992) 492. [77] B.M.C. Pelan, K.M. Watts, I.J. Campbell, A. Lips, J. Dairy Sci. 80 (1997) 2631. [78] K.E. Allen, E. Dickinson, B.S. Murray, LWT Food Sci. Technol. 39 (2006) 225. [79] D.T. Wasan, W. Xu, A. Dutta, A. Nikolov, in: S.E. Friberg, K. Larsson, J. Sjöblom (Eds.), Food Emulsions, fourth ed., Marcel Dekker, New York, 2004, p. 221. [80] M. Anderson, B.E. Brooker, in: E. Dickinson, G. Stainsby (Eds.), Advances in Food Emulsions and Foams, Elsevier Applied Science, Barking, 1988, p. 221. [81] K. Boode, P. Walstra, A.E.A. de Groot-Mostert, Colloids Surf. A 81 (1993) 139. [82] S.A. Vanapalli, J.N. Coupland, in: S.E. Friberg, K. Larsson, J. Sjöblom (Eds.), Food Emulsions, fourth ed., Marcel Dekker, New York, 2004, p. 327.

Please cite this article in press as: E. Dickinson, J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.09.080

8

E. Dickinson / Journal of Colloid and Interface Science xxx (2014) xxx–xxx

[83] W. Xu, A. Nikolov, D.T. Wasan, J. Food Eng. 66 (2005) 97. [84] H. Casanova, J. Chen, E. Dickinson, B.S. Murray, P.V. Nelson, M. Whittle, Phys. Chem. Chem. Phys. 2 (2000) 3861. [85] M.G. Semenova, J. Chen, E. Dickinson, B.S. Murray, M. Whittle, Colloids Surf. B 22 (2001) 237. [86] D.F. Darling, J. Dairy Res. 49 (1982) 695. [87] K. Boode, C. Bisperink, P. Walstra, Colloids Surf. 61 (1991) 55. [88] S.A. Vanapalli, J.N. Coupland, Food Hydrocoll. 15 (2001) 507. [89] H.D. Goff, J. Dairy Sci. 80 (1997) 2620. [90] J.N. Coupland, Curr. Opin. Colloid Interface Sci. 7 (2002) 445. [91] J.C. Arboleya, M.J. Ridout, P.J. Wilde, Food Hydrocoll. 23 (2009) 1358. [92] E. Davies, E. Dickinson, R.D. Bee, Food Hydrocoll. 14 (2000) 145. [93] E. Davies, E. Dickinson, R.D. Bee, Int. Dairy J. 11 (2001) 827. [94] W. Xu, A. Nikolov, D.T. Wasan, J. Food Eng. 66 (2005) 107. [95] A. Bazmi, A. Duquenoy, P. Relkin, Int. Dairy Sci. 17 (2007) 1021. [96] H.J. Kim, A. Bot, I.C.M. de Vries, M. Golding, E.G. Pelan, Food Res. Int. 53 (2013) 342.

[97] L. Lobo, D.T. Wasan, Langmuir 9 (1993) 1668. [98] N.E. Hotrum, T. van Vliet, M.A. Cohen Stuart, G.A. van Aken, J. Colloid Interface Sci. 247 (2002) 125. [99] K.E. Allen, E. Dickinson, B.S. Murray, Food Hydrocoll. 22 (2008) 690. [100] K.E. Allen, B.S. Murray, E. Dickinson, Int. Dairy J. 18 (2008) 1011. [101] E. Dickinson, Curr. Opin. Colloid Interface Sci. 15 (2010) 40. [102] Th.F. Tadros, B. Vincent, in: P. Becher (Ed.), Encyclopedia of Emulsion Technology, vol. 1, Marcel Dekker, New York, 1983, p. 129. [103] S. Levine, B.D. Bowen, S.J. Partridge, Colloids Surf. 38 (1989). 325 and 345. [104] B.P. Binks, Curr. Opin. Colloid Interface Sci. 7 (2003) 21. [105] S. Tcholakova, N.D. Denkov, A. Lips, Phys. Chem. Chem. Phys. 10 (2008) 1608. [106] N.D. Denkov, I.B. Ivanov, P.A. Kralchevsky, D.T. Wasan, J. Colloid Interface Sci. 150 (1992) 589. [107] E. Dickinson, Trends Food Sci. Technol. 24 (2012) 4. [108] A. Ye, X. Zhu, H. Singh, Langmuir 29 (2013) 14403. [109] E. Dickinson, Food Hydrocoll. 28 (2012) 224.

Please cite this article in press as: E. Dickinson, J. Colloid Interface Sci. (2014), http://dx.doi.org/10.1016/j.jcis.2014.09.080

Structuring of colloidal particles at interfaces and the relationship to food emulsion and foam stability.

We consider the influence of spherical colloidal particles on the structure and stabilization of dispersions, emulsions and foams. Emphasis is placed ...
1MB Sizes 0 Downloads 4 Views