Modern methods of microscopic and spectroscopic deep insight Read more ...


Modern microscopy and spectroscopic techniques have made possible deep insights into the process of film formation from aqueous dispersions of latex particles and the evolution of the mechanical properties of these films. This knowledge is important for the design of high performance coatings that are friendly to the environment. Although all aspects of the mechanism of film formation have received active attention over the past several years, some of the most important recent advances have occurred in three areas. First, there have been detailed studies of the drying process, from which we have a deeper understanding of the ways in which water evaporates from a wet latex dispersion. Second, further information is available about the compaction process, leading to the formation of a void free film. Finally, there have been several new studies of the polymer diffusion process, particularly in films formed from structured latex.


Department of Chemistry and Erindale College, University of Toronto, 80 St George St, Toronto, Ontario, Canada M5S 3H6; e-mail: این آدرس ایمیل توسط spambots حفاظت می شود. برای دیدن شما نیاز به جاوا اسکریپت دارید

Current Opinion in Colloid & Interface Science 1997, 2:192-199

Electronic identifier: 1359-0294-002-00192

© Current Chemistry_Ltd ISSN 1359-0294


AFM   atomic force microscopy

MFT   minimum film-forming temperature

PBA   poly(butyl acrylate)

PBMA   poly(butyl methacrylate)

PMMA   poly(methyl methacrylate)

PS   polystyrene

SANS   small angle neutron scattering

SEM    scanning electron microscopy

TEM   transmission electron microscopy

VOC     volatile organic solvent

T9   glass transition temperature


The process of film formation from latex dispersions underlies much of the technology of water-borne coatings. Whereas the subject of latex film formation has been studied for more than 50 years, it is only over the past 10 years, through the application of modern instrumental methods, that a fundamental understanding of this process has begun to emerge. A major factor promoting these investigations is the need for sufficient understanding to permit the development of high performance coatings that are friendly to the environment.

When water evaporates from a dispersion of latex particles, a transparent film is formed if the temperature is sufficiently high. At lower temperatures, turbid, cracked or powdery films are formed. The minimum film-forming temperature (MFT) is a crude measure of the deformability of the particles during the time of drying, and is related to the glass transition temperature (Tg) of the latex polymer.

To conceptualize the problem, the film formation process is commonly divided into a number of stages as indicated in Figure 1. As water evaporates from the dispersion, the polymer particles become more concentrated and come into proximity. When the forces accompanying drying exceed the modulus of the particles, particle deformation occurs to yield a void-free film that is still mechanically weak. In the final stage of film formation, polymer diffusion occurs across the interparticle boundary to provide the entanglements that give strength to the film. These stages are not always well separated, and within each stage subtle features operate that often differ from system to system. For example, all latex particles have polar or ionic groups at their surface that provide colloidal stability. In some systems, this polar layer is sufficiently thick that it forms a continuous membrane in the freshly formed film [1). Break-up of the membrane becomes an important step in bringing the particle cores into intimate contact so that interdiffusion can occur [2]. Some factors (uniform particle size ,low ionic strength) promote ordering of the dispersion in the liquid state [3] even at relatively low particle concentrations. In disordered dispersions, as the volume fraction of the particles exceeds 50%, packing effects cause particle mobility to be restricted. In this review I will use the term 'high solids' to refer to latex particle concentrations near and above 50 vol%. It is likely, as drying proceeds, that order or disorder in the dispersion at high solids determines particle packing in the film.

Another feature that makes understanding film formation difficult is that different experiments that sample different aspects of the process are, in their interpretation, sometimes overgeneralized. For example, many latex films dry as a moving front in the plane of the substrate. Scientists who use optical or scattering experiments to study the drying stage of film formation often focus on the last spot to dry. In this way, they are able to monitor the system over the longest possible time scale. There is evidence, however, that this spOt is atypical. Recent atomic force microscopy (AFi\1) measurements suggest that the compression associated with the propagating drying front promotes ordering of particles in this region and squeezes salts and surfactant from the dispersion into this zone so th at their local concentration is elevated over that in the rest of the film [4°°].


A number of reviews, focusing on different aspects of film formation, have recently appeared [5,6,7°,8°0 ] . Also, a book containing the papers presented at an ACS symposium in August 1995 [9··] entitled Film Formation in Waterborne Coatings has recently been published. My own detailed review has just been published as a chapter in a book on emulsion polymers edited by Lovell and El-Aasser [10··].


The mechanism of water loss from latex dispersions is surprisingly complex. It is this stage of film formation that is, at present, the least understood. Most latex dispersions dry as a moving front in the plane of the substrate. Drying begins at the edges of the wet film. A transition zone separates the transparent dry regions from the turbid wet dispersion. Anyone who has painted a wall has observed the shrinking of the wet zone on the wall as the paint dries. In ordered latex dispersions, the films tend to dry uniformly, and in some systems a skin of coalesced polymer forms at the surface of the wet zone in the film. Early experiments on drying kinetics monitored weight loss as a function of time. These kinetics are deceptively simple: over most of the drying process, water is lost at a uniform rate. This would be expected if the area of the wet dispersion did not change as the sample dried. In fact, one observes a constant drying rate, even as the wet

zone shrinks in size [10··,11-13]. Feng and I [14··] showed recently that dispersions of hard particles (T >~IFT) dry measurably faster than dispersions of soft particles. Particle size effects on the drying rate are small, but dispersions containing mixtures of hard and soft latex lose water at a substantially slower rate than dispersions of either of the individual components. This suggests that in systems that dry as a front moving in the plane of the substrate, water evaporation takes place preferentially at the dry/wet boundary surrounding the wet zone of the film. A picture of this process is shown in Figure 2. Scriven and coworkers [15,16··] examined rapidly frozen dip-coated films by cryo-scanning electron microscopy (SE~I) to observed particle packing in the boundary region. They observe a compaction front, and comment on pore emptying which accompanies densification of the films at the outer edge of the boundary region. The forces which operate in this zone, and which affect particle packing morphology, have been investigated by Nagayama and Ivanov, their coworkers [17].

Representation of the drying front for a latex dispersion. Drying commences at the edge of the liquid droplet. The dry film at the edge is separated from the wet dispersion by a transition or boundary region. The wet edge of this boundary has been called the compaction front [15,16"1. Water evaporates more rapidly from the boundary region than from the water surface above the wet dispersion, thus creating a flux of water and particles from the wet portion to the edge [171. As drying proceeds, the dry edge grows in size and the wet dispersion contracts in area. The drying front propagates in the direction opposite to that of the water flux. Voids persist at the dry edge of the boundary [15,19"1 and, at times, even in the optically clear dry film itself. The time scale for full densification of the dry film can be hours or days.

Traditional models of water loss from latex dispersions treat the area of the wet dispersion as constant. According to these models, the entire surface of the dispersion dries at a constant rate until the particles come into contact and the water layer drops below the level of particles. Virtually all of these models invoke a drying front which moves from the surface of the film toward the substrate (18). This type of model may be appropriate in cases where a skin coating the wet zone dominates the drying, or where particle ordering suppresses liquid fluxes during drying.


Spherical particles in the dispersion form polyhedral cells in the nascent film. If the cells are isotropically deformed and ordered in a face centered cubic array, the cells will be regular rhombic dodecahedra [1,2]. Randomly packed particles lead to random Wigner-Seitz cells. Four questions dominate the discussion of particle deformation. First, do the particles deform before they touch? Second, what are the forces that drive deformation? Third, is particle deformation biaxial or isotropic? Finally, what forces resist deformation? There is clear evidence from the environmental SEl\l measurements of Keddie el 01. [19--], and from the wide angle neutron scattering measurements of Crowley et 01. [20], that particle deformation can precede contact. A similar conclusion was drawn from observations of the drying process of ordered dispersions of soft particles by small angle neutron scattering (SANS) [1,3], and from osmotic compression measurements on dispersions of soft particles at high solids [21]. In these last two sets of experiments, one observes a continuous decrease in the interparticle spacing and a continuous increase in the osmotic pressure at solids contents beyond that for close packing of hard spheres. In these systems in which the particles are not in contact, deformation occurs because the capillary and osmotic pressures are able to overcome the internal stresses in the latex.

Nevertheless, if drying is sufficiently rapid, or if the drying temperature is close to the l\lFT, then drying can precede deformation and still yield transparent films [22]. This process may resemble dry sintering, driven by the surface tension of the polymer. Sintering and deformation of polymer microspheres have been reviewed by Mazur [23--]. Small deformations operating over short times are elastic, whereas larger deformations require polymer flow (the terminal region of the viscoelastic response) to achieve compaction [24]. A somewhat different explanation has been proposed by Lin and Meier [25--,26--] for the densification of seemingly dry films. This group used AFl\I to study the deformation of latex rnonolayers on a mica substrate, and calculated that capillary forces associated with an annulus of water bridging the equators of adjacent particles are strong enough to deform the particles. In some circumstances, flocculated particles can densify even under water [27J. The plastifying effect of water can lower the l\IFT well below the Tg of the dry polymer [28]. Biaxial deformation leads to polymer cells in the nascent film which occupy the same cross-sectional area as the original particles. Isotropic deformation leads to particles with smaller cross-sectional dimensions. Distinguishing these processes by microscopy or scattering measurements is difficult, bur the latter process should lead to shrinking of the films on the substrate. van Tent and coworkers [29-31] studied the drying process by monitoring the UV-Vis transmission of the wet films as a function of weight loss. Diffraction effects led to a transmission peak that shifted to shorter wavelengths as the film dried. From their data, they inferred that particle deformation is biaxial and not isotropic. Latex films dried close to the t\IFT exhibit mud-cracking, indicating that macroscopic flow is not sufficient to relax the stresses associated with shrinking. This topic deserves more attention.


The morphology of latex films is traditionally studied by scanning and transmission electron microscopy (TEl\I). Recently, freeze-fracture [32,33] TEl\I, environmental SEl\I [18,19--], and AFl\I [34,35,36-,37J have provided powerful insights into film morphology. AFt\I has been particularly useful for the study of the evolution upon aging or annealing [25--,26--,38,39J of film surface morphology.

It is likely that the morphology of latex films is determined by the order or arrangement of particles in the dispersion at high solids. Single-component dispersions that form a colloidal crystalline phase yield face centered cubic (fcc) ordering in the film [32,33]. Under some circumstances, other ordered structures are found [40]. In films formed from blends of particles differing either" in size or composition, the situation is more complex. Particle segregation in the film is almost certainly determined by phase separation in the dispersion [41-].

Soft latex particles form void-free films upon drying that are soft, tacky, and have poor mechanical properties. Good mechanical properties are obtained from films formed from latex particles with a T g above room temperature, but these particles normally do not deform and form films when dried at room temperature. The traditional approach to obtaining useful films from hard latex particles is to add a coalescing aid (a volatile organic solvent [VOCJ to the dispersion. The solvent acts as a plasticizer, lowering the modulus of the particles in the dispersion so that they deform upon drying. The solvent also promotes polymer diffusion in the film (see below). Subsequent solvent evaporation raises the Tg of the film above room temperature, giving a hard, tack-free film. Concern about the environment is driving the search for solvent-free latex coatings. One approach is to build crosslinking chemistry into low-Tg latex so that the films cure after drying into tough rubbery films. If the functional groups that react to crosslink the polymer are built into separate particles, an important consideration is the mixing of the two functional latex in the film. Whereas one anticipates random mixing for particles of similar size with similar groups at the surface, blends of colloids of different size can demix in the dispersion. In this context, the study of the morphology of films prepared from latex blends has special importance. Examples involving blends of hard and soft latex particles are discussed below in the section on the mechanical properties of latex films.

Healing of the interfaces

Voyurskii [42J predicted that healing of the interfaces between adjacent cells in a polymer film occurs via polymer diffusion. Techniques are now available to follow this diffusion. One type of experiment employs SANS to monitor the dimensions of deuterated microspheres introduced in small amounts into a dispersion of similar but unlabeled particles. This technique was originally employed by Hahn et al. [43] at BASF for films of poly(butyl methacrylate) (PBMA), and by Klein, Sperling and coworkers at Lehigh University, USA [44,45] to study interdiffusion in melt-pressed films of polystyrene (PS) microspheres. More recently, Klein and Sperling [46] examined films prepared from PS particles comprising polymers of narrow molecular weight distribution prepared by a mini-emulsification technique. This allows interdiffusion rates for polymers of different chain lengths to be examined. Their most impressive result was to demonstrate, as predicted by theory [47], that full mechanical strength in the films is achieved when the diffusion length is comparable to the radius of gyration of the latex polymers [48,49,5000 ] .

Another approach that my coworkers and I developed [51], uses nonradiative energy transfer (ET) measurements on latex films comprised of two types of latex particles, one labeled with a donor dye (D), and the other with an acceptor (A). Polymer interdiffusion brings the 0 and A groups into proximity, leading to a strong increase in E'E This technique is rather versatile, because the labeled particles are easy to prepare. This technique has been used to study temperature and molecular weight effects [52], polymer composition [53], the influence of coalescing aids [54,55] and nonionic surfactanrs [56°°), and the influence of latex structure (particularly core-shell structures) on the interdiffusion process [57,58°°,59°]. There has been a significant discussion of data analysis in this type of experiment. If one wishes to obtain absolute, rather than relative, values of the polymer diffusion coefficient, one must take into account the details of the concentration profile at the interface, generated by diffusion [60,61,62°-65°]. The distribution of diffusion coefficients characterizing polydisperse systems has been considered explicitly by Liu et al. [66]. In films prepared from a blend of two latex with different chemical compositions, a sharp interface can be obtained and the efficiency of ET measures the amount of interface in the film [67°].

One of the interesting issues recently addressed is the role of membranes, formed from polar material at the latex surface, on the fusion and interdiffusion process. The Rhone-Poulenc group in France employed a variety of experiments, the most powerful being rehydration of newly dried films with 0 20 coupled with SANS measurements to characterize the structure of the membranes and to monitor their breakup as the films were annealed [1-3].

It is very important to understand how this membrane phase affects polymer interdiffusion. For example, Kim et al. [57,580 0 ] showed that in latex films prepared from PBi\IA core-shell latex containing up to 9 mol% poly(methacrylic acid) in the shell, polymer molecules were able to diffuse into neighboring cells when the films were annealed at lOO·C. This system was characterized by a broad distribution of diffusion constants, indicating the fastest diffusing species dominated the early time diffusion. The issue of polymer diffusion across a polar membrane has recently been examined in a more rigorous manner in a latex film system (styrene-butyl acrylate latex) with a much thinner membrane composed of poly(acrylic acid). joanicot et al. [68°°] used SANS to monitor simultaneously polymer interdiffusion rates and the break-up of the membrane structure as the .latex films were annealed. They found that the mechanical properties of the films are determined by the state of the cell walls. In humid environments the membranes take up water and lose their cohesive strength. Annealing the films at temperatures where the membranes remain intact produced massive diffusion of low molar mass polymer across the membranes, but this had no effect on the mechanical strength of the film. Only annealing at ternperautres that caused fragmentation of the membranes caused interdiffusion of the high molar mass constituents of the latex. At this stage, the films became resistant to humidity.

Mechanical properties

The ultimate use of latex coatings depends upon the mechanical properties that develop in the film. Dynamic mechanical analysis (Or-.IA) measures the linear response of the film to small deformations. Toughness, tensile strength, and strain-to-break are measures of the response of the film to large amplitude deformations. Permeability measurements assess the resistance of the film to penetrants. As mentioned above, for full tensile strength to develop, polymers near the cell surface must diffuse across the cell boundary to a depth comparable to their radius of gyration RG. Crosslinking reactions have a profound effect on the properties of low Tg latex films. For example, if a significant amount of interparticle polymer diffusion preceeds crosslinking, tough, rubbery films are obtained. If, however, crosslinking occurs prior to interdiffusion, then each particle is converted into a microgel. Only the dangling chain ends can diffuse across the cell boundaries in the latex film. As Zosel and Ley [69] have demonstrated, little mechanical strength can develop in such films.

Over the past five years, Richard and coworkers [70-72] have reported a number of important studies of the linear dynamic mechanical properties of latex films, formed from acrylate, vinyl acetate, and styrene-butadiene latex dispersions. They have shown how membranes formed from polar polymer at the particle surface act as reinforcement agents, enhancing the elastic modulus at temperatures above the Tg temperature of the film [70-73]. When such films are deformed uniaxially at T>Tg, the array of membranes stretches uniformly if the system is dry. Moisture can permeate these films. If deformed in the wet state, spatial correlations are lost in the direction perpendicular to the stretch direction [74""]. A recent report describes a combination of SANS and Dl\IA measurements on a series of SB latex (microgels) with different surface substituents and crosslink densities. Systems with high crosslink densities retained the ceIl structure in the film, had persistent membranes, and were characterized by slower local chain dynamics [75""].

One of the most intriguing aspects of the mechanical properties of latex films are the properties of films containing hard and soft polymer. If the hard component is the membrane, it provides mechanical reinforcement of the film only as long as the film remains dry; and if these films are heated or aged above Tg, the membrane breaks up to form polar occlusions in the film. l\Iechanical reinforcement is also found in films prepared from structured latex [76] (e.g. ones in which a high Tg polymer core is embedded in a low Tg shell), or from binary blends of high and Jow Tg latex [14"",77""]. Film formation (deformation) in these systems is controlled by the modulus of the low Tg component but the film formed has a much higher modulus because of the presence of the hard particles. Under some circumstances in the blended films, the high Tg particles will form a percolation network [78]. When films prepared from hard-soft latex blends or from hard-soft core-sheIl latex are heated above the Tg of the hard component, this component fuses and bicontinuous blends can be formed [79,80"",81].

Latex dispersions, particularly structured latex, are now being used for pressure-sensitive adhesives. This application appears to depend critically on the viscoelastic properties of the film produced [82,83",].

Coatings friendly to the environment

Concern for the environment is driving major changes in the coatings industry [84]. One trend is the replacement of solvent-borne coatings with water-based alternatives. These new coatings still contain volatile organic solvents (VOC, ca 15 wt%) to promote coalescence, but a further effort is devoted to the development of zero-VOC coatings which match or improve on the properties of traditional coatings. In this effort, several trends are apparent. One is the use of hard-soft latex blends or hard-soft core-shell latex to facilitate coalescence at room temperature. To improve the toughness of these films, and films produced from low Tg latex themselves, attention is being given to crosslinking chemistry, particularly ambient cure chemistry. l\Iy impression is that this is currently one of the hottest topics in the coatings industry. The key insight to appreciate is that to obtain optimum properties, polymer diffusion across the intercellular boundary in the film must precede crosslinking, otherwise there will be only weak adhesion between adjacent cells in the film. In multicomponent films, this ambient cure will permit the morphology developed during film formation to be Jocked in place.

An alternative approach involves hybrid systems. These include blends of latex with, for example, water-borne urethane dispersions [85], or hybrid particles containing both urethane and acrylate components [86].

Another important step in the development of new coatings is the control of the fate of the surfactant in the film. _New polymerizable surfactants are coming on the market. These surfacrants become confined to the cell surface in the film [87-89]. It is not yet clear whether this always leads to an improvement in all film properties. Ordinary surfactants have an important influence on the permeability of latex films [90-92].


The study of latex film formation is in a very active state, with parallel activity in both academic and industrial communities. Major strides have been taken toward understanding the various mechanisms involved in the transformation of a dispersion of particles in water into a mechanically rough film. It is easy to identify areas where more knowledge is needed, both from a mechanistic point of view and to design effective environmentally friendly coatings.

One would like a deeper understanding of the drying mechanism for particle dispersions, particularly about how the drying process affects film morphology. Hybrid systems involving a mixture of latex and other water-dispersible polymers (polyurethanes, polyesters) are becoming more widely used. Our understanding of these blended systems is much less developed than for acrylic latex. Polymerizable surfactants are becoming widely used in the coatings industry. We need to know about their fate in the latex film and their effect on film properties.