Paul Steward's Home Page | Other pages on the subject of Polymer Latices.

Literature Review of Polymer Latex Film Formation and Particle Coalescence.

By Paul A. Steward.

Please Note:

This file is quite large and takes a few moments to load. Please be patient and allow it to fully load.

If your browser allows, I recommend the version of this page which uses frames for navigating the references. Click Here For Framed Version.

Your browser should be able to show super & subscripts and, ideally, requires access to a symbol font (to show Greek letters, etc.) of comparable character set to that available in MS Windows. The page can generally be understood without such a font, but please be aware that the presence of odd or out of place characters may result from this problem!

Note on Navigation. To view a reference, click on the name of the author. This will move you to the bottom of the document. After reading the reference, click on return to re-position back in the body text. Note: if more than one reference is cited by a single link (eg, if the link looks like: Steward [1995], Hearn [1984]), then be sure to use the return link from the final reference. If you do not do this, you may end up at the wrong position in the document.

Disclaimer.

Contents:

1
Introduction.
2 Preparation of polymer (latex) films.
2.1 Casting substrates.
2.2 Film Formation.
2.2.1 Particle coalescence and film drying stage II.
2.2.2 Particle fusion (film drying stage III), film structure, and aging.
2.2.3 Solvent-casting of a film.
2.2.3.1 Volatile organic components in aqueous latices.
3 Film morphology.
3.1 Heterogeneous latex films.
3.2 Film opacity.
4 Latex film additives.
4.1 Additives in latex paints: the critical pigment volume concentration.
5 References..

UpGo to Top of Page.

Literature Review of Polymer Latex Film Formation and Particle Coalescence.

1 Introduction.

The physical and mechanical properties of thin polymer films are important from both an academic and an industrial point of view. These properties are affected not only by the nature of the polymer, but also by the method of film preparation and conditioning. In many cases, films are obtained from commercial sources, where they are manufactured by processes such as calendering or extrusion. On a smaller scale, films are prepared by solvent-casting or compression moulding. All of these techniques start with the bulk polymer, and are obviously therefore unsuitable for the direct preparation of films from latex dispersions.

UpGo to Table of Contents.

2 Preparation of polymer (latex) films.

2.1 Casting substrates.

Paint and coating technologists have long been interested in the preparation of films from aqueous-based and oil-based formulations. For the coatings technologist, adhesion to (and in the case of waterborne coatings, wetting of) the substrate will be one of the main requirements [Nicholson and Wassen (1990)]. Such surfaces, as required for film formation from an academic point of view, are far removed from the typical everyday surfaces that the developed end-coatings will typically be used on. In contrast to commercial polymer coatings, the academic study of films typically requires that they be free of any substrate, and ease of removal of the film from its casting substrate is thus one of the main requirements. Several methods for assisting with this problem have been devised. These include: casting onto photographic paper and removing the film by soaking in warm water to dissolve the gelatine; casting on to aluminium foil followed by amalgamation with mercury; and casting onto silanised plate glass and into PTFE dishes, in which case the film is removed by gently peeling it from the substrate.

Chainey et al. (Thesis & 1985 paper) evaluated several film preparation techniques during an investigation of the transmission properties of films formed from surfactant-free polymer latices, including casting onto a mercury surface, on to photographic paper, onto PTFE, into silicone rubber dishes, and on to silanised glass. After extensive trials, all were rejected on the grounds that either the substrate concerned contaminated the film, or that the minimum thickness of film that it was possible to cast was at least an order of magnitude greater than that required. The method of film preparation eventually adopted was developed from the flash coating method used for tablet coating, which is widely used in the pharmaceutical industry. The aim was to form the film so quickly that it could not disjoin, and this was achieved by spraying the latex onto a heated block, coated with PTFE, at temperatures exceeding 120° C (393 K). Multiple (20 to 50) passes of the spray gun were employed and the film was allowed to return to room temperature before it was removed from the block, and in some cases was cooled to near the polymer Tg using an appropriate solid-CO2 slush bath. Spitael and Kinget (1977), however, found that sprayed solvent-cast films exhibited a higher degree of porosity than solvent dish-cast films, due to their droplet-like nature, during spraying, which remained apparent in the final film structure.

Roulstone (for PhD Thesis) cast PBMA films onto Pyrex glass plates, from which they could be removed by soaking in hot water, or in the case of additive-present films, cast onto nylon plates from which the films could be removed without soaking.

Yaseen and Raju (1982) have reviewed the full range of film preparation techniques, finding pros and cons for each depending on the intended application of the film.

UpGo to Table of Contents.

2.2 Film formation.

The formation of a latex film arises from the coalescence of the individual latex particles - which are normally held apart by stabilising forces (electrostatic and/or steric) resulting from the charged polymer chain end groups or surfactant. These forces (and others, see later) are overcome by the evaporation of the continuous phase (water).

The formation of a continuous film (ie, transparent and crack-free) is then dependent on the minimum film formation temperature (MFFT) of the polymer, which in turn is dependent on the elastic modulus (resistance to particle deformation), and to a lesser extent, the viscosity of the polymer. If the film is cast above its MFFT, then coalescence of the latex particles can occur. However, if the film is below its MFFT, then a friable discontinuous film or powder compact may form, which is typically opaque due its structured nature and, hence, its ability to diffract light. The more desirable outcome (in the context of this study) of film-formation is something of a compromise since the tendency of the spheres to flow and fuse into a continuous film can, in the extreme, also result in a permanently tacky film that is more suited to adhesive applications [Talen (1959)].

The formation of films at temperatures slightly lower than the MFFT, has been studied by Myers and Schultz (1964) using an ultrasonic impedance technique. The results, with respect to the formation of a continuous film, were found to be dependent on the rate of drying and, hence, the rate of relief of stresses within the film. At a temperature within 6° C below the MFFT, and with a drying rate sufficiently slow, a certain amount of creep was able to occur permitting the formation of a film due to the stresses being at a level insufficient to fracture the film. As the rate of drying was increased, the creep mechanism was not fast enough to relieve the stress such that initially the films became crazed, and at the highest drying rates, cracked.

The MFFT, although tending to be close to the Tg of a given polymer, has been reported, for various polymers, to be above or below the Tg [eg, Eckersley (1990) and Jensen (1991)]. Both the MFFT and Tg are influenced by the same molecular features (eg, the inclusion of a softer polymer will lower both the Tg and MFFT). Ellgood (1985) showed, for a series of vinylidene chloride (VDC)/ethyl acrylate copolymers, that both the Tg and MFFT peaked with increasing VDC content, but not at the same composition. Below 55% VDC content, the Tg was found to be greater than the MFFT. A 15° C difference was found between the Tg and MFFT at the extremes, and different surfactants were also found to alter the MFFT and its relationship to the Tg of the copolymer. The method of feed of a second stage monomer (eg, seeded growth or blend) can lead to variation in the MFFT of the resultant latex due to the change in particle morphology [Cao (1993)]. The effect of core-shell morphology on the MFFT has, however, been found to depend on shell thickness [Devon (1990)]: thin soft shells on hard cores requiring higher drying temperatures than thicker soft shells due to the necessity to deform the core of the former in order to form a film.

Brodnyan and Konan (1964) and Kast (1985) note that comonomers that impart hydrophilic properties (eg, methyl and ethyl acrylates, etc.) into a polymer may reduce the MFFT to below the Tg, in the case of the wet film (as opposed to these properties being measured for the dry polymer) by allowing water to act as a plasticiser. Similarly, surfactant that is compatible with the polymer may also plasticise the polymer, lowering both the Tg and/or the MFFT [Eckersley (1993), Vijayendran (1980), and Vijayendran (1982)].

Eckersley and Rudin (1990), Jensen and Morgan (1991) and Sperry (1994) et al. each found the MFFT to be related to latex particle size, although this is not always the case [Brodnyan and Konan (1964)] (cf. the theories of film formation (see section on particle coalescence theory). Eckersley found the MFFT to be dependent on latex particle diameter and even in the case of a series of polydisperse copolymer latices, the results suggested that the MFFT be proportional to the number average particle diameter. However, the increase in MFFT between a 150 nm latex and 1200 nm latex was only 5° C. Jensen and Morgan found that as the (monodisperse) latex particle size decreased by a factor of seven, the MFFT was reduced by ca 10° C. Sperry found the time dependent dry MFFT (ie, the transition from a cloudy to clear film in a latex pre-dried below its MFFT) increased with increasing particle size, and concluded that this was due to a simple viscous flow model which accounted for the larger interstitial voids which were present between larger particles and the longer time required for them to be filled (by particle deformation) to give a transparent film.

Following the evaporative drying process (typically by gravimetric methods, although Cansell et al. (1990) described an alternate method using dielectric measurements), from beginning (ie, wet latex) to end (ie, film) leads to a sigmoidal curve, which can be divided into a number of stages for analysis. Poehlein (1975) and Vanderhoff (1973) et al. describe three stages in the drying process (Figure 1), studying drying with and without the aid of a 'windtunnel' to remove the humidity of the evaporating water. The drying process may be complicated, however, by virtue of it being non-uniform (ie, different areas of the film may dry at different rates) and, hence, quantitative evaluations of the rate of drying typically involve the use of estimates of the size of, for example, dried areas of film, or the use of averages to give a mean value for the film as a whole. Despite this, attempts have been made to mathematically model the evaporative process [eg, Pramojaney et al. (1980)].

Plot showing 3 stages of drying
Figure 1 Schematic plot of the water loss occurring on latex drying.

Stage I. Water evaporates from the latex surface, concentrating the latex: the rate of evaporation has been determined by a number of workers [eg, Croll (1986), Sheetz (1965), and Vanderhoff (1973)] as being the same as the rate of evaporation from water alone, or of water from a dilute solution of surfactant + electrolyte, ie, such as that which constitutes the aqueous phase of a latex. This stage is the longest of the three, and lasts until the polymer has reached ca 60 - 70% volume fraction, F, (dependent on the stability of the latex) (cf. 74% for close packed spheres) or until the surface area of the latex's liquid-air interface starts to decrease as a result of, for example, disjoining or solid film formation. Initially the particles move with Brownian motion, but this ceases as the electrical double layers undergo significant interaction, once a critical volume of the water has evaporated.

Stage II. This starts from the time at which the particles first come into irreversible contact, and iridescence may be observed on the latex surface. The rate of evaporation per unit area of open wet latex remains constant, but the overall rate of evaporation decreases greatly during this stage. (Reducing the rate of evaporation can lead to a better quality film by allowing the particles more time to pack into an ordered structure before flocculation occurs. Casting at high temperatures gives the particles sufficient energy to overcome their mutual repulsion and the films are formed before the particles are ordered [List and Kassis (1982)].) Particle deformation occurs in soft latices, as the particles start to fill inter-particular capillary channels, driven by interfacial forces (see paragraph 2.2.1).

Stage III. This stage starts with the initial formation of a continuous film. The remaining water leaves the film initially via any remaining interparticle channels and by diffusion through the polymer itself, but the rate of evaporation eventually slows to (asymptotically) approach that of diffusion alone. It is during this stage that a soft latex becomes more homogeneous and gains its mechanical properties. The rate of water removal may be decreased by film additives that are impermeable or hydrophilic.

Croll [2 references: (1986), and (1987)], however, describes the process of film formation in just two stages as an evaporation front moves into the coating leaving behind a 'dry' layer, containing no continuous water, with ahead of it a transition layer losing water to the dry layer above and supplied with water from the wet latex below. The rapid rate is maintained for as long as wet latex remains at the substrate, then the rate progressively decreases.

Bierwagon (1979) considered film formation in relation to the same three regimes as proposed by Vanderhoff, discussing drying in terms of film thickness and latex solids content. Eg, a film of low solids content could dry faster than one of high solids content despite the lower quantity of water to be removed from the latter, which however, reaches the diffusion-controlled stage (ie, surface closure) sooner, and then loses water more slowly. As the film dries from the surface down, a fixed film area is then subject to contraction in the z-plane, thereby producing stress in the x-y plane. If polymer elasticity is insufficient, then the stress can be overcome by slippage between the coalesced layer and the fluid beneath giving rise to the 'mud-cracked' surface effect.

Hwa (1964) studied the non-uniformity of film drying to determine that, as the aqueous phase evaporated, three distinct regions could be observed, eg, a dry region, a wet (latex) region, and an intermediate region of flocculated latex (such that the film as a whole embodied all three periods of Vanderhoff's drying regime). In Hwa's circular films these regions formed concentric bands, and the films dried from the outside inwards. It was noted that these rings differed, dependent on the Tg of the polymer. In the case of a low Tg polymer, the flocculated and dry regions were both continuous, whereas for hard polymers, fine radial cracks were apparent, due to the relief of stresses, and the dry region was more opaque (due to cracks) than the flocculated region. Hwa was able to show that the flocculated region had some mechanical strength (ie, it was not washed away in a stream of water, as was the wet latex) presumed to result from van der Waals' forces, and that the particles were not close packed (the volume fraction, F, was between 0.49 & 0.62) such that the region was porous. The flocculation was found to be dependent, to some extent, on the nature of the surfactant used: easily desorbed soaps were proposed to be squeezed away from the points of particle-particle contact to form micelles in the particle interstices and, hence, aid flocculation, compared to the surfactant-free latices. Surfactant which was not easily desorbed, however, delayed the onset of flocculation to high volume fractions.

The fact that latices of differing stability will flocculate at different particle-particle separations (ie, at different rates) has been used to advantage by Okubo and He (1991) in the preparation of asymmetric films from latex blends. Such films showed side-dependent (ie, polymer-substrate or polymer-air interface) variation in properties such as film tackiness and permeability.

The constant rate (rc) period of drying was studied by Poehlein et al. (1975) with respect to latex particle size, using an equation to account for the non-uniform drying. If the three regions described previously (by Hwa (1964)) are expressed in terms of film areas (eg, AI = area of wet latex; AIII = area of dry film (the rate of evaporation in this region was assumed to be negligible); and AII = flocculated region area, then:

(eqn 1)

where: W = gross film weight; t = time.

It was shown that the rate, rc, increased with increasing particle size (although this was not so apparent in still air, and was not predicted by a theoretical equation based on heat transfer theory). Poehlein et al. speculated that the rate increase with particle size was a result of the differences in water content in the double layer, or the particle surface area available for surfactant adsorption, although the latter was minimised by attempted adjustment of the surface tension of the latices to a constant value.

Armstrong and Wright (1992) noted that the films prepared from latices of a relatively large particle size (750 nm) were of a poorer quality (ie, poorer corrosion resistance due to greater porosity) than those films prepared from latices of a smaller particle size (105 nm). This was ascribed to the larger particles showing less coalescence, but it was not clear whether the poor quality of the film was the result of differing rates of drying resulting from the differing particles sizes, or simply due to the larger interparticle voids that would be found for the larger sized uncoalesced particles.

UpGo to Table of Contents.

2.2.1 Particle coalescence and film drying stage II.

Over the years a number of theories regarding the formation of polymer films, from the fusion of latex particle spheres, have been considered. These include:

Dry sintering is driven by the polymer-air surface tension. Dillon et al. discuss the coalescence processes in terms of the viscous flow of the polymer. This viscous flow results from the shearing stresses caused by the decrease in the polymer-particle surface area, and the resultant decrease in polymer surface energy, as the film is formed. By appraising values relevant to the typical polymer latex, Dillon considered the forces acting on the interparticle 'holes' (of radius R), as denoted by the Young-Laplace equation:

(eqn 2)

where: Pi = internal pressure (see Figure 2); Pe = external pressure; g = polymer-air interfacial tension; R = radius of curvature of the sphere; and concluded that the pressure acting on the void could not be contained, as a result of the polymer being permeable to the trapped air and water vapour. The extent of coalescence (as determined from the half-angle of coalescence) was then related to an equation developed by Frenkel:

(eqn 3)

where: g = surface tension; t = time; h = polymer viscosity coefficient; r = particle radius; q = angle seen in Figure 2, ie, the half-angle of contact.

Schematic of coalescing particles
Figure 2 A cross-section of sintered latex particles, and a plan view showing the interparticle capillary.

Dillon showed that a plot of q2 as a function of 1/r, as measured by electron microscopic observation of shadowed particle pairs, gave the expected linear relationship.

Brown discussed wet sintering, driven by the polymer-water interfacial tension, leading to the deformation of particles, during drying. He considered the forces acting both for and against the coalescence of the latex particles, with the conclusion that for coalescence to occur, an inequality must exist in which the capillary force, FC, (resulting from the surface tension of the interstitial water, caused by the formation of small radii of curvature between the particles as the water evaporated) must overcome the forces of resistance to deformation, FG, of the latex spheres: ie, FC > FG. These forces, Brown presumed, were proportional to the relevant pressures, with the area over which they act as the constant of proportionality, and hence: PC > PG. From Laplace's equation, Brown derived the capillary pressure, PC for the cylinder of radius R, between three contiguous latex particles (see Figure 2), in terms of the latex particle radii, r:

(eqn 4)

where: gw = polymer-water interfacial surface tension;

Note:

(eqn 5)

By treating the particles as elastic bodies, the pressure on the area of contact was also calculated, in terms of the elastic shear modulus, G, of the polymer and, hence, Brown derived an expression for coalescence:

(eqn 6)

where: Gt = time dependent shear modulus (necessary because the viscoelastic particles are treated as elastic); r = latex particle radius.

Sheetz later amended Brown's equation for capillary pressure to account for the situation when the contact angle, J, between the polymer and water did not equal zero:

(eqn 7)

Sheetz also formulated his own theory of latex coalescence. In qualitative terms, as the latex becomes concentrated by evaporation of the water, flocculation occurs as the repulsive forces of the particles are overcome. Particles at the latex-air interface are then subject to the forces of capillarity and therefore coalescence, leading to compaction and deformation of the particles under the surface. Water in the film's interior must then diffuse through the upper layers to escape and this generates a further, vacuum-like, compressive force acting normal to the film's surface. The mechanism is therefore seen to be based on Brown's wet sintering mechanism and diffusion. Scheetz analysed the thermodynamics of the system and showed that the source of the energy for the particles' fusion was the heat in the environment - converted for film formation by the evaporation of the water. In evidence for the diffusion being involved in the coalescence mechanism, Sheetz cited the facts that a film containing a water-permeable polymer dried at a rate faster than one that was less water permeable; and that a film in which capillary action was prevented (by means of a thin solvent-cast film) could form a continuous film, whilst the same polymer without the solvent-cast deposit formed a discontinuous film.

Although Dobler et al. (1992) generally agreed with the mechanism of Scheetz, they believed, from observations of iridescence, that the surface of the latex closes (ie, complete surface iridescence followed, presumably, by skin formation) long before the particles become close packed in the bulk latex (as they do in Sheetz's theory).

Like Sheetz, Mason (1973) also identified a number of erroneous assumptions and points of error in Brown's work. Firstly, in converting his forces (capillary and deformation) to pressures, Brown assumed the same constant of proportionality (an undefined area, A, over which the capillary and contact pressures are exerted) for each, eg, for equation 4:

(eqn 8)

Mason points out that these areas (over which FC and FG act) are not necessarily identical, and repeated the analysis using corrected values for the areas such that the condition Brown quoted in equation 6 became:

(eqn 9)

Mason also criticises the fact that Brown assumed that the capillary pressure remained constant whilst the latex particles coalesced, and derived a new equation for the capillary pressure based on the deformation of the spheres. From this, the criterion for film formation moved yet further from Brown's inequality, to give:

(eqn 10)

It has also been noticed [Kan (1993) and Sperry (1994)] that Brown's work has been criticised for using a polymer modulus which was obtained for the dried polymer, rather than a polymer in an aqueous environment.

Despite Brown's differences with Dillon over the role of the evaporating water phase in latex coalescence mechanisms, both research groups presumed that the forces of coalescence were inversely proportional to the latex particle radii. Vanderhoff et al. (3 references: 1966, 1970 and 1970) indicated that the pressures for coalescence, resulting from the works of Dillon and Brown, were insufficient to cause the coalescence of particles greater than 1 µm (1 micrometre) in diameter and extended the theories accordingly. These extensions to the theory were again based on determining the forces acting to cause coalescence. Vanderhoff proposed that as the water evaporated, the particle coalescence was initially hindered by repulsion of their stabilising layers. Further evaporation then resulted in particle-particle contact, and the force increased due to (i) the polymer-water interfacial tension, and (ii) the small radii of curvature in the region of coalescence (r1 and r2 as seen in Figure 2). Hence:

(eqn 11)

where: Pi = internal pressure; Pi' = internal pressure in the region of coalescence; r = particle radius, and r1 & r2 are radii as seen in the diagram; cf. equation 2.

As with the work of Brown, Mason criticised Vanderhoff's work for mathematically confusing force with pressure. However, whereas the error in Brown's model led to the error being less that an order of magnitude, Mason claimed Vanderhoff's error completely invalidated the work.

The aforementioned theories assumed that the latex particles acted as viscous fluids or elastic spheres. Kendall and Padget (1982) noted that particles are not truly viscous (as depicted by the Dillon model) from the fact that latex films can show a residual particulate structure, and that they were elastic with a viscous component. Similarly, Kendall reasoned that the fact that a film does not attain its full strength until after the point when all of the water is removed implied modifications were necessary to Brown's model. Accordingly, he introduced a new model based on the theory [Johnson and Kendall et al. (1971)] of adhesion and interfacial surface energy between elastic spheres - again, its applicability is limited by its foundation on elastic spheres. For a hexagonal close packed array of spheres, Padget states that voids are eliminated (ie, opaque film to clear film) when:

(eqn 12)

ie, for film formation:

(eqn 13)

where: g = energy released when two spheres of unit area move into contact; u = Poisson's ratio (ie, the lateral contraction per unit breadth/longitudinal extension per unit length); E = Young's modulus - Note that:

(eqn 14)

The dimensionless ratio on the left of equation 12 is described by Kendall as the "crumble number." If this number is high, then coalescence will not occur; if the crumble number is low, then a transparent non-porous film will form. The transition was said to occur at a crumble number of "around ten." Kan (1993) indirectly investigated the deforming forces during latex coalescence at the MFFT, by measuring the moduli of water saturated latices. The magnitude of the results showed good agreement with the Kendall theory.

Sperry (1994) has investigated the role of water in film formation using MFFT measurements on latices pre-dried at temperatures below their MFFT (which shows the MFFT by a transition from an opaque to a clear film) and also wet latices. In the case of a hydrophobic polymer, the dry MFFT was virtually identical to the wet MFFT indicating the presence of water and, hence, capillary forces contributed little to film formation. Plasticisation by water was said to be the cause of hydrophilic polymers yielding wet MFFTs which were lower than the dry MFFTs - by up to ca 10° C. The dry MFFTs decreased linearly with log[time] which was attributed to (Williams-Landel-Ferry type) viscoelastic relaxation resulting from van der Waals' attractive forces/polymer-air surface tension and the collapse of interstitial voids.

Lamprecht (1980) studied film formation by treating the latex particles as viscoelastic bodies, and included dynamic factors such as creep deformation in relation to the rate of water removal. Eckersley and Rudin (1990) indicated a mistake in Lamprecht's calculation, and derived a corrected criterion for film formation assuming viscoelastic spheres:

(eqn 15)

where: t´ = the time when the two spheres are in closest proximity to one another; Jc(t) = the polymer's time dependent creep compliance (= time dependent modulus, G(t), in the linear viscoelastic region).

Eckersley went on to develop her own (time-dependent) viscoelastic model of film formation based on the polymer-water interfacial tension forces acting together with the capillary force to cause coalescence. The theoretical results for particle deformation from the model agreed with experimental measurements by SEM

The use of neutron scattering experiments (SANS), together with SEM/TEM has permitted the 'observation' of latex particle structure during film drying; allowing the three-dimensionally deformed shape of the latex particles to be studied. Joanicot et al. (1990) and Chevalier et al. (1992) have studied the particle packing of latices consisting of a soft hydrophobic polystyrene-poly(butyl acrylate) copolymer core, stabilised by a hydrophilic shell (either neutralised poly(acrylic acid) or a monolayer of zwitterionic {ie, both positive and negative ionic parts such that overall neutral} surfactant). Such a stable configuration allowing the close packing of the particles before coalescence occurs.

By the use of very thin films, the SANS scattering patterns, produced by latices as the dispersions concentrated (to ca 20 - 50%, by evaporation) were studied, and interpreted in terms of each of the monodisperse particles being surrounded by 10 - 12 nearest neighbours, and forming face-centred cubic crystalline-like particle packing. Such ordering of the particles was lost (as shown by the replacement of the scattering pattern by Debye-Scherer rings) either when the dispersion was too dilute (latex volume fraction < 20%) or, if salt was present to screen the interparticle electrostatic forces, or when the latex was polydisperse.

It was originally deduced by Lissant (1966), when investigating high-internal-phase-ratio emulsions, that above the maximum packing volume available to hard spheres (74%, in a face centre cubic configuration), deformation of a sphere to fill all of the space (whilst keeping the same volume as the original sphere) will give rise to a rhomboidal dodecahedron. Joanicot et al. (1990) showed such polyhedra, by staining the hydrophilic membrane (ie, particle interfaces - which on particle deformation defined the polyhedra), when a latex had lost most of its water - comparing the structure to that of a foam.

Chevalier et al. (1992) used the SANS evidence to describe particle coalescence. Coalescence was observed to move through the drying surfactant-coated latex in the form of a 'coalescence front' progressing from the periphery of the film inwards: the progress of which could be reversibly halted by stopping the evaporation of the water. At the interface between dry film, and wet latex (in which the packing fraction was < 0.74), a phase inversion occurred, from a polymer in water to a water in polymer system. Coalescence was said to only occur if the hydrophilic layer was breached, and this was seen to differ depending whether the hydrophilic layer was the surfactant or the acrylic acid. The films stabilised by poorly adsorbed surfactant formed a continuous matrix of polymer containing hydrophilic domains, whilst those latices stabilised by the polymer retained the foam-like structure: as indicated by a collapse of the diffraction peak in the case of the former, but a peak which could be reinstated by swelling the film with water in the latter case. This work thus differentiated between coalescence, which was defined as the breakup of the hydrophilic layer, and polymer chain interdiffusion, which occurred in a later stage.

UpGo to Table of Contents.

2.2.2 Particle fusion (film drying stage III), film structure, and aging.

The aforementioned theories of coalescence considered film formation as a process in which, after the intermediate stage, the deformed latex particles were held together by physical forces. This was countered by Voyutskii [3 references: 1958, 1963 & 1964] who, from studies on adherence of rubber coated fabric strips, judged that these forces alone were insufficient to account for the mechanical and physical properties of such a film. Voyutskii discussed film formation from a polymer dispersion in the typical three stages of: evaporation to a concentrated 'solution'; followed by polymer 'globules' coming into close contact, with deformation, due to capillary forces; and then in the final stage, the film gaining its mechanical strength by virtue of the fact that solvated stabiliser molecules migrated to between the polymer globules, as a result of the loss of water, allowing the self-diffusion of polymer chains and leading to an increase in the homogeneity of the polymer. This process of mutual interdiffusion between polymer chain ends is termed autohesion, and although not an innovative idea (the mechanism being based on the work of Josefowitz and Mark (1942), Voyutskii was the first to apply it to the coalescence of latex particles.

By virtue of an ongoing process, similar to autohesion, Bradford and Vanderhoff (1966) studied the changes in structure occurring in a continuous, transparent, film as a function of film age. A styrene-butadiene copolymer film, which was replicated and viewed by TEM within hours of casting, showed the vestiges of the original latex particles. However, over a fourteen-day period, the particle contours on the film surface eventually disappeared, accompanied by the exudation of material from within the film: presumed to be stabiliser (visible as blister-like eruptions), and in some cases, electrolyte (visible as crystallites). Investigation involving a series of non-ionic surfactants showed that it was the long poly(oxyethylene) (ie, -CH2-CH2-O-CH2-CH2-) chained stabiliser which the film exuded, and not the short chained oil-soluble stabiliser. It was concluded that it was additive which was incompatible with the polymer that was secreted. The exuded surfactant was itself observed to change in appearance with time, being initially smooth but later becoming rough. This was attributed to oxidation, which was later shown could cause scission of the poly(oxyethylene) chain thus making it more compatible and, hence, allowing it to diffuse back into the polymer.

Bradford and Vanderhoff (2 references: 1966 & 1972) went on to study the aging process, which they termed further gradual coalescence, in detail, and Vanderhoff (2 references: both 1970) has discussed the results in a number of reviews of film formation. They showed that it occurred at the film-substrate interface as well as at the film-air interface (albeit at possibly different rates), and also in the film's interior where stabiliser was exuded into "pockets." Porosity due to the leaching of surfactant was reduced if the film was aged before the surfactant was leached. The further gradual coalescence was independent of the casting substrate, with the exception of a mercury substrate which it was proposed induced different stresses into the film when compared to casting on a solid substrate. The proportion of oxygen in the atmosphere also affected the rate of further gradual coalescence - particle contours disappearing rapidly in a high oxygen atmosphere (attributed to oxidative softening of the copolymer) whilst remaining unchanged in a nitrogen atmosphere. Bradford and Vanderhoff also found that reducing the polymer molecular weight (by using a chain transfer agent {t-dodecyl mercapton; ie, CH3-(CH2)10-CH2-SH}), increased the rate of further gradual coalescence, and the exudations appeared earlier, whilst an increase in the three-dimensional nature of the polymer (by crosslinking with divinyl benzene {ie, H2C=CH-C6H4-CH=CH2}), reduced the rate of aging but did not halt it completely.

Bradford and Vanderhoff observed aging in other polymers, including ethyl acrylate-methyl methacrylate copolymer, natural rubber, and vinylidene chloride-vinyl chloride-ethyl acrylate copolymer. (NB. Vinylidene chloride: H2C=CCl2; vinyl chloride: H2C=CHCl.) Roulstone (1991) has observed the disappearance of particle contours within films of surfactant-free PBMA. Bradford and Vanderhoff did not, however, find evidence of aging in poly(vinyl acetate) (PVAc) films, even after 280 days. This they attributed to the surface hydrolysis of the acetate to poly(vinyl alcohol) (PVOH) which prevented autohesion. The presence of PVOH in such films was later shown by Kast (1985) (using osmium tetroxide staining of the hydroxyl groups, and electron microscopy) as a continuous network throughout the film, separating the PVAc core particles.

Although theories for the interdiffusion of polymer chains at an interface exist [eg, Kim (1983)] (and have been shown [Yoo (2 references: 1990 & 1991)] to be applicable to latex particle film formation), it was the introduction of SANS techniques that has, since the mid-1980s, provided a means of investigating the magnitude of the polymer chain interdiffusion and, hence, the reason for a film's strength. Some of this work has recently been reviewed by Sperling et al. (1990).

Hahn et al. (2 references: 1986 & 1988) used SANS experiments on PBMA to show that latex particle coalescence was as a result of the "massive" interdiffusion of polymer from different latex particles. The diffusion coefficients, D, during further gradual coalescence at ca 80° C (353 K) were 10-16 cm2 s-1 and demonstrated a thirty-fold increase as a result of a 20° C rise in temperature. A factor of two change in molecular weight changed the diffusion coefficient six-fold. Yoo et al. (2 references: 1990 & 1991) used similar methods to Hahn to investigate the interdiffusion penetration-depth at which the film attained its full tensile strength. Results were found to be dependent on the polymer molecular weight, the spatial distribution and location of chain ends at the polymer-polymer interface, and the ratio of the polymer chains' radii of gyration to the radius of the latex particle.

Linné et al. (1988) found, also by SANS, that for small (ca 38 nm Ø) high molecular weight (6*106) polystyrene, that the molecules were fourfold constrained in the lattices, and that a retarded non-Fickian diffusion relaxation at 10-14 cm2 s-1 occurred by a modified reptation model (that accounted for the polymer chains' charged ends {and, hence, repulsion}), cf. 10-16 cm2 s-1 expected for translational diffusion at 170° C (443 K). Energy release also occurred during Tg determination with the second run giving a value 6° C lower than the first, and close to the expected value for polystyrene. Linné claimed that polymer chain interdiffusion over 110 Å was sufficient to give a tough coherent film, ie, diffusion across the particle boundary of 50 to 60 Å; ca 0.15 * Ø. The chain end diffusion of high molecular weight (eg, > 7.6*104) polymer may not follow Fickian or reptation models due to polymer chain entanglement [Wang (1993)].

Kim et al. (1993) studied particle coalescence using "direct mini-emulsified" latex particles, which were claimed to show a narrower molecular weight, and particle size distribution, when compared to conventionally emulsion-polymerised particles, and also contained no ionic chain-end groups. (Particles were prepared by anionic polymerisation to give hydrogen chain ends, which were distributed randomly throughout the particles, as opposed to being predominantly on the particle surface as in conventional ionic chain end groups.) The rates of inter-particle chain-interdiffusion were found to be greater for Kim's particles when compared to conventional latex particles: a feature ascribed to the smaller chain end groups (compared to, for example, sulphate end groups) and the fact that the need to overcome polar repulsions was not required.

The effects on polymer interdiffusion of varying amounts carboxyl chain end groups on the surface of a PBMA latex (prepared to give a PBMA-core, carboxylated-shell morphology) has been investigated by Kim et al. (1994) using fluorescence direct energy transfer measurements. Whilst not preventing chain interdiffusion, the presence of acid end groups did retard it. Problems of data interpretation arose, however, due to particle surface-composition heterogeneity, and it was presumed that the polymer chains of lowest molecular weight and lowest carboxyl content interdiffused earlier than those of higher molecular weight/carboxyl content: the latter, however, showed increased contribution to the energy transfer at longer annealing times. (Interdiffusion was possibly between miscible phases, eg, PBMA with PBMA, and PBMA-co-MAA with PBMA-co-MAA.)

Kim et al. (1994) also investigated the effects of neutralisation of (carboxyl and sulphate) acid end groups using a range of mono and divalent bases. Like the aforementioned carboxylated polymer, interdiffusion of the neutralised polymer showed an initial quick rate, followed by a longer lasting constant rate period. The presence of an ionomeric shell reduced the rates of interdiffusion further than carboxylation, in the order of NH 4+ < Na+ < Ba2+. Ie, monovalent ions retard diffusion to a lesser extent than divalent ions. Kim was unable to distinguish the contribution of Tg effects during this work.

Despite their effects on chain interdiffusion, carboxyl groups have been shown to contribute to the viscoelastic cohesive strength of particles in a film due to interfacial crosslinking [Richard (1992)] - via either hydrogen bonding or ionic dipolar interactions when neutralised. (If the neutralising molecule was sufficiently large {eg, diaminopolyoxyethylene}, then coiled lamella could be formed which had the opposite effect {when compared to neutralised carboxyl groups} of plasticisation and therefore reduced cohesive strength.)

Unsurprisingly, Sperling et al. (1990) conclude that the rate of coalescence depends on where the polymer chain ends lie with respect to the particle surface, and that films form faster when the ends lie on the particle surface. The rate of coalescence of latex particles, as determined by Padget and Moreland (1983) using TEM and an AC impedance technique, increased with increasing concentration of block copolymer surfactant. This effect was attributed to plasticisation of the (dry) latex polymer by the surfactant. (When wet, the poly(ethylene oxide) groups are hydrated and were relatively incompatible with the polymer, such that the MFFT was not changed whilst water was still present.) Exudations of the surfactant were only significant at concentrations above surface coverage.

The effect of coalescing aids on the interdiffusion of polymer chains during polymer annealing was also investigated by Wang and Winnik et al. (2 references: 1990 and 1992) Organic solvents, typically used as coalescing aids (eg, glycol derivatives), were found to provide a constant degree of enhancement in inter-diffusion rates (ie, as a result of plasticisation by the coalescing aid) during annealing, and the effect was described by a modified Williams-Landel-Ferry equation. Deviations from the model (ie, the degree of enhancement was not constant during the course of annealing) were as a result of evaporation of the coalescing aid from the film, by a Fickian-type process.

Distler and Kanig (1978) assumed that latex-cast films would always retain some form of inhomogeneity due to the distribution of stabilising charge (eg, polymer chain end groups or grafted surfactant). Citing latex particle stability as proof of the particles having an exterior more hydrophilic than their interior, they reasoned that, on deformation of the particles into a film, these surface boundary layers of hydrophilicity would interdiffuse to form an interconnected 'honeycomb,' but which would then act to inhibit further interdiffusion by virtue of the incompatibility between the hydrophilic and hydrophobic polymer. In evidence of this, Distler and Kanig point to the fact that a normally transparent film may turn opaque, or even show Bragg diffraction iridescence, when swollen with water: both of which require latex-particulate sized features to cause the necessary difference in refractive index (between the polymer and water) and crystalline structure, respectively, and hence, were testimony of the ability of water to penetrate the hydrophilic honeycomb and expose such features.

The (i) increased water absorbency, (ii) increased difficulty to leach surfactant, and (iii) reduced tendency of a copolymer latex film of vinyl acetate and a vinyl ester of Versatec 10 acid to whiten (swell) in water were attributed, by Aten and Wassenburg (1983), to the redistribution of surfactant molecules from the surfaces of the latex particles (where they were sited on latex polymerisation) to a more even distribution throughout the film, following a period of secondary 'drying' above the polymer Tg. Such redistribution was ascribed to the increased polymer chain mobility above the polymer Tg, and was not apparent in films annealed below the Tg.

Distler and Kanig (1978) also used TEM, combined with staining techniques, to observe polyhedra shaped particle boundaries within their films. In a butyl acrylate-acrylonitrile-methylol acrylamide-acrylic acid copolymer, uranyl ions (ie, salts of uranium nitrate (UO2(NO3)2: the ion is UO22+, and compounds typically fluoresce) were bound to carboxylic acid groups, to clearly show the particulate network, which was also apparent in similar films contrasted with hydrazine/osmium tetroxide [Kanig and Neff (1975)]: the 'particles' in the films showing a good degree of correlation of size compared to the original latex size. Such contrast-forming staining techniques also showed the web-like particulate structure in films formed from surfactant-free acrylate latices [Distler and Kanig (2 references: 1978 & 1980)] indicating that such particles have a high degree of surface hydrophilicity (functionality).

Joanicot et al. (1993) investigated the effect of thermal annealing on the cellular network ("membrane") of hydrophilic particle boundaries, which act to separate the hydrophobic cores of the latex particles in a film: observing the fragmentation of the membranes by the use of SANS and TEM Fragmentation of the hydrophilic membrane arose due to the high hydrophilic-hydrophobic interfacial surface area, such that the dispersion was not in a state of lowest free energy. Joanicot discusses the fragmentation of the hydrophilic membrane in comparison with the evolution (droplet maturation) of oil-in-water emulsions. In such an emulsion, the thermal motion of the stabilising surfactant (surrounding the emulsion droplets) can lead to surfactant bridging through the aqueous phase, between droplets, which can in turn allow the oil droplets to combine. For such combination to occur, requires the attainment of a critical thickness of water between the particles. The thickness of membranes between latex particles is thin (ca 2 nm), and their fragmentation is thickness-independent, but requires thermal treatment for the attainment of a threshold beyond which the hydrophilic and hydrophobic parts of the film segregate. Joanicot determined that membrane fragmentation was a function of: i, mobility of the polymer from which the membrane was made (restricted by crosslinking, etc.); ii, the mobility of the particle core (ie, a function of polymer Tg, etc.); and iii, the anchoring (ie, due to surface functionality) of the membrane polymers to the core. Once fragmentation had occurred, the hydrophilic polymer was found to form irregular clusters within the hydrophobic polymer. The irregularity was ascribed to the method of cluster formation in which the hydrophilic membranes were said to be 'attached' to cores with a distribution of strengths (ie, heterogeneous latex particle surface) and, hence, released at differing times: clusters of hydrophilic polymer occurring (in the hydrophobic polymer) at points where the membrane was released initially. This was again compared to the evolution of high internal phase ratio emulsions whereby large droplets capture smaller droplets to yield a distribution of small droplets and much larger ones: 'heterogeneous growth' (cf. homogeneous growth: the combination of equally sized smaller droplets to give a monodisperse emulsion which grows continuously).

In the aforementioned work of Joanicot et al. (1990) and Chevalier et al. (1992), it had been shown that latices formed a foam-like structure of rhombic dodecahedra upon coalescence. These actual structures have been observed [Roulstone et al. (1991) and Wang et al. (1992)] in latex films of PBMA, by electron microscopy after fracturing under liquid nitrogen. Wang et al. (1992) noted differences in structure between large and small particles (337 nm and 117 nm, respectively). Whereas the larger particles formed ordered structures, the smaller particles (for which films were formed in the presence of surfactant), underwent random packing. This was also apparent for the larger particles if the film was prepared in the presence of surfactant. Surfactant could be seen in the fracture cross-sections, and this was attributed to it being concentrated in the coalescent front as it moved through the film.

With respect to coalescence and the effect of repulsions between particles due to their surface group functionality, Nicholson and Wassen (1990) conclude that coalescence mechanisms may be divided into two groups: (i) those dependent on sintering or capillarity processes, which dominate when there are polar repulsions present; and (ii) those dependent on chain interdiffusion, when there is very little repulsion between particles. Whilst ionogenic groups may lead to polymer chain stiffening, this may be countered during latex film casting by increased water hydration leading to plasticisation.

UpGo to Table of Contents.

2.2.3 Solvent-casting of a film.

Whilst the formation of films from latices and from polymers in solution may seem to be fundamentally different processes, there are aspects of similarity: eg, macromolecules in solution behave hydrodynamically as though they are molecular dispersions having solvent-impermeable cores and peripheral solvent-permeable segments [Napper and Gilbert (1990)]; a difference mainly of scale compared with uncharged sterically stabilised latex particles. Whether solvents will necessarily deposit pore-free films of the maximum density and lowest permeability is uncertain. Different outcomes are predicted, depending upon the solvent power, when high concentrations are reached in the later stages of drying. Funke and Zorll (1975) suggested, based upon evidence from freeze-dried extracts of films in the gel stage, that good solvents would produce more open, porous structures, and Nicholson and Wassen (1990) cite Gould (1964) with a similar prediction that compact molecules in the solution will remain compact in the film state. Kesting (1985), however, suggests that polymer coils in good solvents will interpenetrate to a more compact structure than in poorer solvents which will favour earlier polymer segment-segment contacts. Spitael and Kinget (1980) concluded that the formation of a gel was the most important stage of solvent-cast film formation, and that solvents that did not gel yielded poorer films which exhibited poorer transparency.

Distler and Kanig (1978) dissolved freeze-dried samples of poly(ethyl acrylate) and poly(n-butyl acrylate) in tetrahydrofuran, diluted to such an extent (0.1%) that the polymer chains were dispersed with no "felting", ie, the resultant solution was completely homogeneous. In films formed from such solutions, it was found, from analysis by EM, that there remained a residual particulate structure, with particles of approximately the same size as those of the original latex. This they attributed to sufficient self-crosslinking (ie, acrylates have a labile tertiary hydrogen [Billmeyer (1971)]) during the polymerisation process to allow the particles to simply swell (ie, gel) rather than truly dissolve. This hypothesis of crosslinking was tested by repeating the experiment using PBMA, which lacks the ability to self-crosslink: such films cast from tetrahydrofuran showed no sign of residual structure.

The rate of evaporation of solvent from solvent-cast films depends upon (time)½ as expected for a process controlled, or limited, by diffusion, ie, Fickian desorption to the surface through a homogeneous solution of increasing concentration [Blandin (1987)]. However, the removal of the final traces of solvent from solvent cast films is a problem: attributed to the fact that the polymer may be plasticised by the solvent. Elevated temperature (to assist diffusion of the solvent in the polymer), good vacuum and long drying times are used to overcome the problem, eg, 50° C (323 K) at 3 Torr for 96 hrs [Vezin (1981)]. The removal of the final traces of solvent can be important with regard to toxicity (in the case of pharmaceutical coatings) and also the permeability properties of solvent-cast coatings. List and Laun (1980) found the water vapour permeability of isopropyl alcohol-cast Eudragit® L films to be markedly increased by residual solvent. However, the effects of residual solvent were minimised by secondary drying, and the solvent could be almost completely removed in a very short time (8 hrs) if the film was held above its Tg.

UpGo to Table of Contents.

2.2.3.1 Volatile organic components in aqueous latices.

It is common, for example in the paint industry, to add volatile organic components to latices (eg, as coalescing aids - to lower the elastic modulus and provide temporary plasticisation to promote polymer chain motion and, hence, provide a better film finish). Sullivan (1975) investigated the removal of such volatile organic compounds from latices, concluding that the ease of removal was dependent on their molecular size and polarity: the more polar a compound, the more likely it was to partition into the hydrophilic network of the evaporating water (and, hence, find less diffusive resistance), and less residual solvent would remain in the film. Sullivan found that the initial (steady-state) rate of evaporation of water was unaffected by the solvent but may ultimately be slowed if the additive is hygroscopic or interacts with the water, forming hydrogen bonds (eg, as do ethylene {ie, OHCH2CH2OH} and propylene glycol {ie, OHCH2CH2CH2OH}): in which case the evaporation will not be diffusion-controlled, as is the usual case.

Hansen (1974) also discusses the evaporation of various volatile organic components from aqueous latices. Comparing film formation by the evaporation from true solutions (ie, internal diffusion-controlled {characterised by a concentration gradient across the film thickness, with a concentration of zero at the film surface}) to the system in which the volatile is present as an additive, it was found that, in the latter case, no concentration gradient (of volatile) was apparent in the film implying that the evaporation of the volatiles was controlled at the film surface. As in the work of Sullivan (1975), however, the evaporation rate of the aqueous phase was independent of the organic phase.

In the case of diffusion-controlled evaporation, the ratio of the weight of volatiles to polymer can be related to a dimensionless reduced time variable, T, given by:

(eqn 16)

where: D = diffusion coefficient; t = time; L = film thickness.

Hence, in the case of diffusion-controlled (desorption-controlled) resistance:

(eqn 17)

where: W = weight of (subscripted) component.

Whereas for surface-controlled resistance:

(eqn 18)

The effects on polymer chain interdiffusion of small amounts of organic solvents, added to PBMA latices, has been investigated by Juhué and Lang (1994). Results depended on the nature of the solvent's properties (eg, evaporation time, level of plasticisation) with respect to the polymer. Strong plasticisation was found not to be a desirable feature (to achieve high film strength by enhancing interdiffusion) if the solvent was not easily evaporated, and similarly neither was weak plasticisation and quick evaporation. Desirable properties to enhance film formation were therefore moderate evaporation rate and moderate plasticisation.

UpGo to Table of Contents.

3 Film morphology.

Roulstone (for PhD Thesis) observed that in the case of both solvent-cast and latex-cast films, drying occurred from the upper, open surface, downwards to the substrate. This appeared to result in the solvent-cast film having a higher density skin on the top surface in comparison to the substrate side. Similarly, a difference in appearance between the two surfaces was apparent for latex-cast films, with matt upper surfaces and glossy lower (substrate) faces. Replicas of the two surfaces subjected to the resolution of a TEM showed slippage of the layers of latex particles in the upper surface, presumably arising from mud-cracking.

Monodisperse spherical latices are known to form ordered structures at 0.74 apparent volume fraction when extended diffuse double layers interact, which diffract (ie bend light {eg, at a slit, cf. refraction: bend light due to a change in refractive index; and reflection: to mirror}) light and produce opalescence. This behaviour was observed by Roulstone as the model latex-cast films dried, producing 'islands' of iridescence, and may possibly contribute to the surface irregularities. Film asymmetry was also manifest in the permeability towards both water vapour and 4-nitrophenol - which varied according to the film orientation - for both solvent-cast and latex-cast films. When the face towards the donor (permeant containing) compartment was the upper surface, it gave the greater permeability to water vapour, whilst the lower face gave the greater rate of permeability to the 4-nitrophenol. This was explained in terms of desorption being the rate controlling process, and the difference between the permeants arising from their relative affinities for the film: 4-nitrophenol being much more soluble in the film. Opposite trends were found for solvent-cast films in which it was the upper surface that had the greater density, in contrast to the latex-cast films. Abdel-Aziz (in a PhD Thesis) reported a similar effect for solvent-cast Eudragit® films, although the density was found to vary dependent on the casting solvent used, and the films could be seen, under the SEM, to contain pores when cast from acetone containing 5% W/W ethanol [Abdel-Aziz and Anderson (1976)]. Other researchers [Yaseen and Raju (1982)] have found the polarity of the casting substrate to influence the polarity of the lower surface of a film inducing an asymmetric effect.

Although it is generally assumed that latex particles exhibit sufficient colloidal stability to form a close packed-array of particles prior to coalescence, it has been reported by Okubo et al. (1981) that, in the case of a surfactant-free latex, a porous, flocculated layer was apparent at the film interfaces. For a latex of only marginal stability, the film-air interface was observed to be porous. Addition of electrolyte to the latex, before casting, further destabilised the latex such that the film-substrate interface was also found to be porous. However, addition of surfactant to increase the latex stability led to a close packed, non-porous structure in agreement with the findings of Isaacs (1966).

Both Bradford and Vanderhoff (2 papers: 1966 and 1972) and Chainey et al. (Thesis and 1985 paper)], but not Roulstone (1989 Thesis), noted changes in the surface of latex films upon aging. Vanderhoff found increased coalescence prior to surfactant exudation to the surface, whilst Chainey found that spray-cast surfaces which appeared rough under the SEM smoothed to the appearance of a solvent-cast film upon aging for one month. Roulstone et al. (1991), using a freeze-fracture replication TEM technique (in which the sample is frozen at the temperature of liquid nitrogen and fractured in-situ, so as to reduce any thermal effects of the fracture process) showed film structures in PBMA, of close packed spheres deformed into rhombic dodecahedra, with the clarity of the interparticle regions dependent on the temperature of preparation. Only at very high casting temperatures (ie, 132° C {405 K}) was there no evidence of particulate structure, such that the film then resembled the appearance of the fracture section of a solvent-cast film. It was suggested, from visual observations and DSC evidence, that PBMA underwent a melting or softening transition in the region of 120° C (393 K). Long exposure to temperatures lower than 120° C (393 K) were not investigated (eg, 6 hrs at 95° C {368 K} was the most extreme treatment where particle positions, if not the interparticle boundaries, could still be identified). Wood remarks that whilst polymers are ductile at their Tg they require heating to 40° - 50° C above their Tg to produce a processable 'melt'.

Wang et al. (3 references: 1990, 1992, and 1993) performed similar work to Roulstone et al. (1991), investigating the affect of annealing on surfactant-free PBMA films. Results showed a decrease in particle boundaries on annealing at temperatures lower than those investigated by Roulstone. Wang et al. (2 references: 1992, and 1993) also investigated the degree of chain interdiffusion using fluorescent dyes and direct energy transfer measurements. From this it was concluded that there was extensive polymer interdiffusion: at 90° C (363 K) for 2 hrs, an interpenetration depth of 6 nm (in particles of 337 nm Ø) was calculated, and this was shown in the fracture cross-sections as a change from fracturing along the interparticle boundaries (in the un-annealed film) to fracturing through the particles. A decrease in the rate of polymer diffusion with annealing time was attributed to polymer polydispersity: low molecular weight polymer dominating diffusion at early times [Wang et al. (1990)] and high molecular weight after long annealing times. List and Kassis (1982) attributed the fracture of particles (as opposed to particle boundaries) to the increased strength of the boundary resulting from the interaction of polar groups (situated predominantly on the particles' surfaces).

The existence of porosity in both solvent-cast and latex-cast films has been cited by a number of authors [Balik et al. (1989) and Okor (1980)]. The structure of a solvent cast film is dependent on the solvent used. Greater enmeshment and density may result from the use of a good solvent (see paragraph 2.2.3) compared with a theta solvent. The compacted spherical molecules in the latter somewhat parallel the case of film formation from latices.

Roulstone et al. (Thesis and 1991 reference) used mercury porosimetry and krypton BET surface area measurements on both solvent-cast and latex-cast films of PBMA. He attributed the apparent uptake of mercury at high pressures to the compressibility of both film types, rather than porosity, but, krypton BET surface areas were greatly in excess (by a factor of ca 37 times) of the geometric areas of the films used, and similar values resulted from the two types of film. The excess krypton adsorption could be attributed to microporosity at the low temperature (-196° C {77 K}) of the experiment. Mercury porosimetry, when performed on a harder polymer, did demonstrate the presence of porosity in core-shell (hard polymer core; soft shell) latex films designed to be porous, and also showed its removal when thermal treatment took the film above the Tg of the cores.

Okor (1980 Thesis) cited fracture cross-section SEM evidence for mesoporosity in plasticised solvent-cast Eudragit® films. Roulstone et al. (Thesis and 1991 reference), however, using freeze-fracture TEM on solvent-cast, and latex-cast, PBMA films found no evidence of mesoporosity. (I.U.P.A.C. defines pore size according to: micropores, < 20 Å; mesopores, 20 to 500 Å; and macropores, > 500 Å {where 1 Å = 1*10-10 m}.) Nitrogen desorption analysis of the low levels of mesoporosity found to exist in such films is not practically feasible due to the low uptakes involved. Balik et al. (1989) considered microporosity to be present in his terpolymer (vinyl chloride, butyl acrylate and vinyl acetate) latex films at 40° C above their Tg, but whether this excess carbon dioxide sorption would be better described in terms of free-volume in the rubbery polymer is a matter of semantics. Solvent cast films he considered to be structureless and to furnish a baseline performance for permeability.

UpGo to Table of Contents.

3.1 Heterogeneous latex films.

In terms of film formation, structure and properties, copolymer films may normally be considered as being effectively homogeneous in nature (assuming that the polymer composition within individual particles is uniform). However, the same is not necessarily true for copolymer films formed from polymers prepared by a shot-growth type mechanism (or any other mechanism intended to give a core-shell type particle), or when one of the comonomers is more hydrophilic in nature such that it may end up forming a (full or partial) 'shell' around the core of the more hydrophobic comonomer. Kast (1985) describes such particles as being heterogeneous or composite, and cites the aforementioned work of Distler and Kanig (1978), which showed the vestiges of the carboxylated particle boundaries within a film, as being evidence of a heterogeneous network. In similar films of either an acrylic acid-free or acid-containing butylacrylate copolymer, differences in the storage modulus/Tg were cited as experimental confirmation of heterogeneity.

The process of film formation from a heterogeneous particle will be governed [Kast (1985)] by the polymer of lowest Tg, unless that polymer is of low volume fraction ( < 30 to 40%), or surrounded by a shell of high Tg polymer (in which case the polymer may contain voids or will require heating to the MFFT of the higher Tg polymer). The result is described, with experimental evidence from the polymer's viscoelastic properties, as a multidomain polymer with the high Tg polymer dispersed in a matrix of the lower Tg polymer, although this morphology may be inverted by heating above the Tg of the hard polymer.

Permeation, which may provide useful information on structure, through such heterogeneous latex films has not been widely studied, although the situation in which a solid is dispersed in a continuous polymer phase is not unusual, eg, pigmented or filled films. Core-shell latex particles in which the core is below its Tg (at the film casting temperature), whilst the shell is soft, may provide the basis for an ideal model filled film since the uniform distribution of the 'filler' core is promoted by its non-deformation and, hence, uniform packing in the latex films. As such, the theoretical treatments of Maxwell (particle interactions neglected), Rayleigh (cubic arrangement of identical spheres), [see Barrer (1968) for details about Maxwell's and Rayleigh's theories] and Higuchi (1958) (spheres treated as points, but including particle interaction), for permeability should be applicable.

Peterson (1968) has used blends of 28% (volume fraction) of latex polymers, (poly(styrene), poly(methyl methacrylate), and poly(vinylidene chloride)), in a soft continuous phase of poly(vinyl acetate) latex, and found good agreement with the Higuchi theory for oxygen permeability. Chainey et al. (Thesis & 1985 paper) studied the helium permeability of flash-cast films formed from hard core-soft shell polymers prepared by a shot growth technique. The latices consisted of particles composed of a poly(styrene) core-PBMA shell, of varying coating thickness. The films aged in the same manner as for free PBMA films, but the extent of the permeability reduction was less, and decreased with decreasing coating thickness. This was attributed to graft copolymer inhibiting further gradual coalescence. When the solvent-cast poly(styrene) permeability value was used for the core, and the aged value for the PBMA at low volume fraction of core, or the initial value at high volume fractions of core, then reasonable agreement with the Higuchi equation was obtained, although experimental results tended to be high. For a PBMA core-poly(ethyl acrylate) shell combination, aging was observed but could not be attributed to shell aging since poly(ethyl acrylate) homopolymer did not age. Unfortunately, for this combination of polymer, the core polymer aged to the same permeability coefficient as the shell polymer, and thus this similarity gave an insensitive test of the Higuchi treatment. Good agreement for core-shell, and copolymers was found but not for blends, and this was attributed to imperfect film structure arising from incompatibility of the components.

Roulstone (for 1989 PhD Thesis) investigated film aging relative to shell thickness, for 4-nitrophenol permeation through dish-cast PBMA-poly(methyl acrylate) core-shell latex films, and found results conforming with Chainey. As in Chainey's study, there was some agreement with the Higuchi theory, but experimental results were higher than for aged films. Rather better agreement was found when water vapour was the permeant.

UpGo to Table of Contents.

3.2 Film opacity.

It has been stated previously that non-film-forming latices dry to form an opaque, white, friable powder. Such opacity can be a desirable property of a coating, for example, a paint. Typically, an additive such as TiO2 is used to achieve the opacity. The opacity results from light being scattered at an interface between substances of different refractive index (eg, the interface between air and polymer, or air and TiO2) - dependent on the laws of reflection and refraction [Berrie (1973)]. Such scattering and, hence, opacity can therefore also be achieved if a polymer film contains large numbers of voids (ie, due to the high number of polymer-air boundaries) such as in 'microvoid coatings.' The degree of opacity is dependent on (i) microvoid concentration, (ii) microvoid size and size distribution, and (iii) the aforementioned refractive index ratio. There exist a number of patents for the preparation of such coatings, and the methods have been reviewed by Seiner (1978). By utilising a method of optimising microvoid size using a model system developed by El-Aasser et al. (1976), Durbin et al. (1984) was able to prepare a latex paint containing microvoids, in combination with a small film-forming latex, which used 50% less TiO2 to yield comparable hiding power without loss of desirable physical properties (eg, scrub resistance).

An effect similar to that found for microvoid coatings can be achieved through the use of hollow spheres. Such spheres have a large polymer-air interface (requisite for light scattering) as a result of the internal chamber [Hemenway et al. (1985)].

Film turbidity and light scattering/interference measurements, combined with a knowledge of spherical particle-packing structures in latex films has been used by Van Tent and te Nijenhuis (2 references, both 1992) to provide models which allow visible light transmission measurements to be used to characterise the geometrical packing of latex particles during the process of latex film formation.

UpGo to Table of Contents.

4 Latex film additives.

Polymer latex films may contain a number of additives, ranging from stabilising surfactant (endogenous to the reaction, or post-added) to plasticisers to aid film formation. This is in addition to additives added to enable the film to perform a function: eg, colourants in latex paints.

Typically, a plasticiser or coalescing aid may be added to ease the deformation of the latex particles so that a non-film-forming polymer can become film-forming at a given temperature. The plasticiser may then be required to soften the film or, more frequently, be fugitive to allow the film to harden. In such fugitive plasticisers which partition between the polymer and the aqueous phase, there is a compromise between the necessity of the plasticiser to remain in the latex for sufficient time for it to perform its function, but not to remain so long as to soften the film: a factor dependent on the plasticisers volatility. In addition to its usual role of plasticisation, several authors have added plasticiser to various types of polymer films in an attempt to control the permeability of the film Okor (1982), or to provide sustained release pharmaceutical devices as a means of control of the release rate [Goodhart et al. (1984) and Muhammad (1991)]. Goodhart et al. (1984) found that the addition of plasticiser (triethyl citrate or dibutyl sebacate) to Aquacoat® (pseudo-latex) coated drug cores (phenylpropanolamine HCl) changed the flux of the drug such that release rates were inversely proportional to the plasticiser concentration. Addition of plasticiser up to a concentration where it aided coalescence decreased the drug release rate, whilst above a certain level (not stated, but between 24% and 40% W/W) of addition, the solute permeability increased due to the increased solute diffusivity caused by the plasticiser's water solubility. Above the higher loading (40% W/W), an increase in film casting temperature did not effect the rate of drug release, whilst at a loading of 24%, an increase in casting temperature led to reduced release rates.

Hoy (1973) provided experimental evidence which accredited the efficiency of various coalescing aids (including a number of acetates and ethoxy alcohols, as used in latex paints) to their degree of partition between the aqueous and polymer phases, and their Tg. (The ability of the plasticiser to undergo hydrogen bonding had a great effect on the results due to the tendency for the plasticiser to remain in the aqueous phase.) Casting (ie, painting) on porous substrates was found to aid the removal of plasticiser that partitioned favourably into the aqueous phase, decreasing its efficiency.

Additionally, Hoy gave theoretical equations for predicting the action of a coalescing aid on the film's MFFT - based on the assumption that the MFFT was directly proportional to the Tg, where the constant of proportionality was a 'reduced film forming temperature,' K:

(eqn 19)

Using a variation of the Fox equation (for determining the Tg of copolymers) which accounted for the fact that the plasticiser was external to the polymer (correction factor, ) and, hence, had greater mobility, the polymer's plasticised glass transition temperature, Tgp was given by:

(eqn 20)

where: Vp = volume fraction of polymer; aVA = volume fraction of coalescing aid; Tgp = pure polymer glass transition temperature; TgA = glass transition temperature of the coalescing aid.

The volume fraction of coalescing aid, VA, was then modified to account for its distribution in the latex:

(eqn 21)

where: A = weight of coalescing aid (parts per hundred of latex); S0 = initial percentage solids of latex; rA = density of coalescing aid; rp = density of polymer; fp = fraction of coalescing aid contained in the polymer.

Knowing that Vp = 1 - VA, and combining the above three equations:

(eqn 22)

Roulstone et al. (Thesis and 1992 reference) studied the water vapour permeability of films cast from PBMA surfactant-free latices in the presence of post-polymerisation additives that included surfactants, polymers and inorganic electrolytes. The surfactants had a common C12 backbone, but either anionic (sodium dodecyl sulphate, or SDS {ie, CH3(CH2)11SO4-Na+}), cationic (dodecylethyldimethylammonium bromide, or DEDAB {ie, CH3(CH2)11N+(C2H5)(CH3)2Br-}), or non-ionic (dodecyl tetraoxyethylene glycol monoether, or C12E4 {ie, CH3(CH2)11(OCH2CH2)4OH}) polar head groups. A minimum in film permeability was found at monolayer coverage of SDS, and this was interpreted as a consequence of improved particle packing order due to enhanced charge stabilisation or, more tentatively, as due to surface plasticisation by SDS, since somewhat greater coalescence was apparent in freeze-fracture TEM replicas of such films. Below monolayer coverage, permeability decreased with increased SDS concentration, and tended to lower values on film aging. Durbin (Thesis), however, argued that at this level of addition SDS decreased the particle packing order since the increased ionic strength outweighed the effects of increased charge density. At higher SDS concentrations film permeabilities increased, and 'islands' of SDS were visible in micrographs as aggregate defects, with smaller aggregates in the particle interstices. Particle boundaries could be observed and the films did not age to give lower permeabilities.

For the cationic surfactant, at concentrations sufficient to give charge reversal and to maintain stability, a minima in permeability was observed at approximately monolayer coverage. However, permeabilities were higher than for the SDS additions, and higher than for the surfactant-free films. This increased permeability was attributed to the DEDAB first forming a salt with the anionic surface groups in the interparticle regions, and then adsorbing tail first on to the surface - with multi-layers forming at higher concentrations. These films, like those with the higher concentration of SDS, also did not age to lower permeabilities.

Significantly different behaviour was observed for C12E4, which plasticised the films making them more flexible, but not tacky, and giving increased coalescence. Slightly reduced permeability at surface coverage, but then raised permeabilities at higher concentrations were also observed. Being well coalesced, the films did not age to lower permeabilities. Vanderhoff (1970) found compatibility between non-ionic surfactants of this type, and the polymer, with no surface exudations. However, when the ethylene oxide chain length was increased, incompatibility with the polymer did lead to surface exudations. Padget and Moreland (1983), using a non-ionic block copolymer surfactant, found increased coalescence at surface coverage, with surface exudation at higher levels of addition.

Atomic Force Microscopy (AFM) was used by Juhué and Lang (1993) to measure the peak to valley distances in PBMA films as a function of their concentration of (post-polymerisation added) surfactant (C9H19-C6H6-0-(CH2CH2-O)25-SO3-Na+ or sodium nonylphenolpolyglycolether-sulphate). Surface coverage was again significant, providing a minimum in the peak to valley distance: attributed to a favourable delay in flocculation, due to enhanced particle stability and, hence, the greatest degree of particle close-packing at the film surface. (Below surface coverage by surfactant, particle-particle repulsions are weak and flocculation occurs early during film formation, whilst at high surfactant concentrations, electrostatic repulsions are screened by the large concentration of sodium ions in the latex dispersion - again leading to early flocculation.) Additionally, it was found [Juhué and Lang (1994)] by using a similar, but non-ionic, surfactant (ie, the SO3-Na+ being replaced by OH) that depletion effects, as well as steric stabilisation (and, in the case of the ionic surfactant, electrostatic repulsion) contributed to the optimal packing of particles during film formation. (Polymer depletion regions {see Juhué and Lang (1994) for references} occur for colloids in the presence of non-adsorbing polymers. When two particles approach within a distance of less than twice the polymer radius of gyration, exclusion of the polymer from the space between the particles leads to development of an osmotic pressure which flocculates the particles.) The development of micelles at high surfactant concentrations, above surface coverage, contributes to the formation of poor quality films due to depletion flocculation.

Zhao et al. (3 references 1987, 1989 and 1989) used surface analytical techniques including X-ray Photoelectron Spectroscopy (XPS), SIMS, and Fourier Transform Infra­Red-Attenuated Total-internal Reflection (FTIR-ATR) spectroscopy to show that material that was incompatible with the polymer was exuded to the film surface. They revealed, by FTIR-ATR spectroscopy, that SDS was exuded from methyl methacrylate-butyl acrylate latex films mostly during drying, but also during further gradual coalescence, to give a concentration gradient which increased towards both interfaces with a higher concentration at the film-air interface than at the film-substrate interface. The same trends were observed, but to a reduced degree, for a more compatible surfactant. Urban and Evanson (1990) found surfactant exudation during film formation to be initially dependent on the water flux, but ultimately on the difference in surface tension between the polymer and substrate. It was found that mechanical stretching of the film (as may be encountered during removal from a substrate) affected the distribution of surfactant throughout the film: as the surface area is increased (by stretching), the surface concentration of surfactant is decreased and the resultant increase in surface tension is henceforth reduced by the diffusion of surfactant to the interface. Analogous to the ability of a liquid to wet a solid of higher surface energy; casting the film on a substrate of lower surface tension (eg, PTFE {= 18.5 mN m-1}) than the polymer (surface tension typically 25 to 60 mN m-1), prevents the film from wetting the substrate and, hence, surfactant diffuses to the polymer-substrate interface to lower the surface tension.

Roulstone et al. (Thesis and 1992 reference) used two samples of poly(vinyl pyrrolidone) (PVP, ie, PVP molecule ) of molecular weights 44,000 and 360,000 at 0.16 g g-1 (cf., monolayer coverage 0.1 g g-1 and 0.15 g g-1 for the two molecular weights, respectively [Kellaway and Najib (1980)]). The PVP was deposited between the particles, which showed good packing but very limited evidence for coalescence and deformation. Limited aging to a lower permeability was observed. List and Kassis (1982) suggested that high molecular weight stabilisers immobilised particles in films and inhibited aging. Bondy and Coleman (1970) also noted retarded coalescence on film formation in the presence of stabilisers. For post added KCl, at below the CCC of PBMA latex, Roulstone found a reduced packing order coupled with an increased permeability, with most of the inorganic material being exuded from the interstices to the surface layers: a tendency that increased on aging.

Pochan (1987) found, by measuring the dielectric properties of a PBMA latex film containing a water soluble polymer additive (a carboxylated styrenic terpolymer {25% by weight}), that when cast at ambient temperature, the additive was present in the interparticle interstices forming a continuous network. This morphology was, however, said not to be the equilibrium state, and thermal annealing of the film led to the additive forming isolated domains within the film if the annealing temperature was above the Tg of the additive.

Bindschaedler et al. (1987) note a broad diversity of works, and conclusions, concerning film additives in the literature with five possible fates for the emulsifier, as shown in Table 1.

Table 1. Possible fates of emulsifier added to latex films.
Fate of Emulsifier
Polymer Compatibility
Migrates and dissolves in the polymer
Compatible
Exudes towards film surface
Incompatible
Forms independent islets, possibly corresponding to interstitial voids between particles
Incompatible
Forms a continuous network embedding globules
Incompatible
Adsorbs at the particle surface
Incompatible

UpGo to Table of Contents.

4.1 Additives in latex paints: the critical pigment volume concentration.

The pigment volume concentration (PVC) (or 'heterogeneous {discontinuous} phase volume concentration' of a 'random composite' [Bierwagon (1992)] embedded in a polymer matrix!) is a measure used in the paint industry to describe the volumetric percentage of hiding pigment present (+ extenders), VP, in the total volume of non-volatile vehicle solids content (polymer matrix), VB, of a paint:

(eqn 23)

Asbeck and Van Loo (1949) posted an explanation as to why the properties of a coating underwent a dramatic change within a narrow band of the PVC The properties (eg, gloss, permeability, rust prevention) of a series of oil-based coatings of increasing values of PVC, but ground to the same degree of dispersion, showed a sudden break (eg, the permeability increased sharply; the glossiness {but also blistering} decreased markedly; and the degree of rusting increased markedly) at a value he termed the critical pigment volume concentration (CPVC). Asbeck showed that the CPVC was dependent not only on the type of pigment, and its degree of dispersion, but also on its physical relationship with the other components of the paint and, hence, the agglomeration of the system: the higher the degree of agglomeration, the lower the CPVC, and noting the fact that systems with high CPVC's have a smoother texture. From work using different (oil-based) binders, it was reasoned that the degree of colloidal stability imparted to the particles resulted in changes to the packing density, and that this too could influence the CPVC (in addition to the packing properties resulting from the shape of the particles themselves). Further work allowed the CPVC to be defined as: 'the point in a pigment-vehicle system at which there is just sufficient binder to completely fill the interstices between randomly packed pigment particles after the volatilisation of the thinner, such that it represents the densest degree of packing of the pigment particles commensurate with the degree of dispersion of the system' [Asbeck (1977)]. The change in properties of the paint at the CPVC was ascribed to the formation of air voids due to insufficient binder to fill the interstices between the most densely packed particles. Thus, below the CPVC, the pigment packing fraction (Fp) is equal to the PVC (see equation 23), whilst above the CPVC, if the total volume of air in the coating is given by VA, then Fp is given by [Rasenberg and Huisman (1985)]:

(eqn 24)

Various experimental methods of CPVC determination are mentioned by Anwari et al. (1991) However, the determination of an accurate method for the calculation of the CPVC (without preparing a formulation and investigating its properties) has proved elusive. Bierwagon (1972) describes a mathematical model based on the ideal random packing of spheres, their size distribution, and adsorbed layer thickness, etc.: the results, however, were described (by Bierwagon) as "good" for alkyd-type (acid + alcohol resin) paints, when compared to experimental results, but only as "fair" in the case of latex paints which was presumed to result from pigment-polymer interactions.

Rasenberg and Huisman (1985) describe a method which utilises porosity measurements (determined by mercury porosimetry) and the fact that Fp remains constant, above the CPVC (because it represents the highest possible pigment packing density) to determine the value of the CPVC from a knowledge of the pigment weight fraction (and the values of the densities of pigment and binder). By defining the weight fraction of the pigment as c:

(eqn 25)

where: m = mass; subscript p refers to the pigment; subscript B the binder.

Then the volume ratio, Fp, can be determined in terms of c, VA, and the respective densities of the binder and pigment (rB and rp) if all properties are defined per unit mass (kg) of coating:

(eqn 26)

which can be re-arranged:

(eqn 27)

This is thus an equation for a straight line in which the term in the square bracket is a constant (because Fp is constant). A plot of porosity (VA) versus the pigment weight fraction (c) will yield the critical pigment concentration (CPC) as the intercept on the concentration axis, when VA = 0. The CPC can then be converted to a CPVC if the densities of the pigment and binder are known.

Rasenberg and Huisman (1985) note that if Fp is constant (and the densities are known: in order to give the intercept, rB-1), then the CPVC may be determined from a single point (taken above the CPVC and extrapolated). However, this method of Rasenberg and Huisman also requires a film be prepared, rather than being a theoretical method.

Citing a previous paper on the use of a polystyrene latex as a filler, Floyd and Holsworth (1990) discussed a number of anomalies which may affect the CPVC The fact that various properties of the film appeared to show differing values for the CPVC, was attributed to the fact that the polystyrene particles were shown to sinter at points of contact (despite it being non-film forming) and, hence, provide the film with some mechanical integrity at values of the PVC above the typical CPVC Floyd also discussed certain factors that may change the concept of the CPVC: eg, clustering (ie, the association of the dispersed phase {pigment particles} as the concentration increases {even at low volume fractions, eg, < 10%} leading to the formation of doublets, triplets, multiplets, etc.: a random process and not the same as flocculation due to colloidal instability) and percolation (which commences when the cluster size is sufficient to span the film). Floyd speculated that it was this latter feature which gave rise to the change in the film's permeability properties due to the formation of pathways, and also led to the decrease in film gloss by the introduction of light scattering. At high pigment concentrations, a phase inversion can occur as the primary phase changes from being one that is polymer rich to one that is pigment rich. (This can have important connotations with regard to, for example, the electrolytic properties of a film, eg, zinc-rich paints for corrosion protection [Bierwagon (1992)], or any type of film requiring a continuous network of additive.)

Although the results of Asbeck's work have been shown to be equally applicable to latex paints, the CPVC is offset to a lower value by a fixed fractional amount which is termed the binder index [Floyd and Holsworth (1990)] of a given latex. Floyd and Holsworth note that air voids are present in all latex paints, for all values of PVC (and that their concentration increases as the CPVC is approached) due to the fact that particle deformation is hindered by the pigment. The voids are likened to a second dispersed phase, and can undergo clustering and percolation as their concentration increases with increasing PVC Because the pigment is typically impermeable, it is percolating voids that give the increased permeant flux. The CPVC of latex paints is thus attributed to a phase inversion from a bi-continuous (polymer-air) system in which the polymer is the primary phase to one in which the primary phase is air.

Typically, a dispersing or wetting agent (surfactant) is added to a paint to prevent agglomeration of the pigment. For a latex (paint), it is not only important that the pigment itself does not destabilise the latex, but also that the pigment does not remove surfactant for its own stabilisation in detriment to the latex. In ensuring this, the CPVC of a latex paint is usually equal to the ultimate pigment volume concentration (UCPVC) [Asbeck (1977)], ie, the CPVC where all the pigment particles are completely separated and act independently, whilst the polymer matrix acts as a continuous fluid [Bierwagon (1992)]. Factors that affect the ability of the pigment particles to act independently (such as clustering and flocculation, etc.) all serve to decrease the CPVC, when compared to a more stable system.

Adapted from: Modification of the Permeability of Polymer Latex Films., Nottingham Trent University PhD Thesis, 1995.
Copyright © Paul Steward, 1995.
Disclaimer.

UpGo to Table of Contents
or
UpGo to Top of Page.

References.

Nicholson J.W., Wasson E.A., Film Spreading and Film Formation by Waterborne Coatings., in Surface Coatings, Ch. 2, Vol. 3, 91-123, 1990. Return.

Chainey M., Permeation Through Polymer Latex Films, CNAA PhD Thesis, Trent Polytechnic, 1984.

Chainey M., Wilkinson M.C., Hearn J., J. Appl. Polym. Sci., 30, 4273-4285 1985. Return.

Spitael J., Kinget R., Pharm. Acta Helv., 52 (3), 47-50, 1977. Return.

Roulstone B.J., Permeation Through Polymer Latex Films, CNAA PhD Thesis, Trent Polytechnic, 1989. Return.

Yaseen M., Raju V.S.N., Prog. in Org. Coatings, 10, 125-155, 1982. Return.

Talen H.W., Hover P.F., Deutsche Farben Z, 13, 50-55, 1959. (CBDE Translation.) Return.

Myers R.R., Schultz R.K., J. Appl. Polym. Sci., 8, 755-764, 1964. Return.

Eckersley S.T., Rudin A., J. Coatings Technol., 62 (780), 89-100, 1990.

Jensen D.P., Morgan L.W., J. Appl. Polymer Sci., 42, 2845-2849, 1991. Return.

Ellgood B., J. Oil & Colour Chem. Assoc., 68, 164-169, 1985. Return.

Cao T., Xu Y., Wang Y., Chen X., Zheng A., Polym. Int., 32, 153-158, 1993. Return.

Devon M.J., Gardon J.L., Roberts G., Rudin A., J. Appl. Polym. Sci., 39, 2119-2128, 1990. Return.

Brodnyan J.G., Konen T., J. Appl. Polym. Sci., 8, 687-697, 1964. Return.

Kast H., Makromol. Chem., Suppl. 10/11, 447-461, 1985. Return.

Eckersley S.T., Rudin A., J. Appl. Polym. Sci., 48, 1369-1381, 1993.

Vijayendran B.R., Bone T., Gajria C., In Some Studies on Vinyl Acrylic Latex- Surfactant Interactions., Emulsion Polym. Vinyl Acetate Pap. Symp. J., Meeting 1980, Ed., 253-283, (1980).

Vijayendran B.R., Bone T., J. Dispersion Sci. Technol., 3 (1), 81-97, 1982. Return.

Eckersley S.T., Rudin A., J. Coatings Technol., 62 (780), 89-100, 1990.

Jensen D.P., Morgan L.W., J. Appl. Polymer Sci., 42, 2845-2849, 1991.

Sperry P.R., Snyder B.S., O'Dowd M.L.m Lesko P.M., Langmuir, 10, 2619-2628, 1994. Return.

Brodnyan J.G., Konen T., J. Appl. Polym. Sci., 8, 687-697, 1964. Return.

Cansell F., Henry F., Pichot C., J. Appl. Polym. Sci., 41, 547-563, 1990. Return.

Poehlein G.W., Vanderhoff J.W., Witmeyer R., J. Polym. Preprints, 16 (1), 268-273, 1975.

Vanderhoff J.W., Bradford E.B., Carrington W.K., J. Polym. Sci., Polym. Symp. 41, 155-174, 1973. Return.

Pramojaney N., Poehlein G.W., Vanderhoff J.W., Drying '80, 2, 93-100, 1980. Return.

Croll S.G., J. Coatings Tech., 58 (734), 41-49, 1986.

Sheetz D.P., J. Appl. Polym. Sci., 9, 3759-3773, 1965.

Vanderhoff J.W., Bradford E.B., Carrington W.K., J. Polym. Sci., Polym. Symp. 41, 155-174, 1973. Return.

List P.H., Kassis G., Acta Pharmaceutica Technologica, 28 (1), 21, 1982. (CBDE Translation.) Return.

Croll S.G., J. Coatings Tech., 58 (734), 41-49, 1986.

Croll S.G., J. Coatings Tech., 59 (751), 81-92, 1987. Return.

Bierwagon G.P., J. Coatings Technol., 51 (658), 117-129, 1979. Return.

Hwa J.C.H., J. Polym. Sci., A (2), 785-796, 1964. Return.

Okubo M., He Y., J. Appl. Polym. Sci., 42 (8), 2205-2208, 1991. Return.

Poehlein G.W., Vanderhoff J.W., Witmeyer R., J. Polym. Preprints, 16 (1), 268-273, 1975. Return.

Armstrong R.D., Wright J.D., Corrosion Sci., 33 (10), 1529-1539, 1992. Return.

Dillon R.E., Matheson L.A., Bradford E.B., J. Colloid Sci., 6, 108-117, 1951.

Frenkal J., J. Phys. (USSR), 9, 385, 1943. Return.

Brown G.L., J. Polym. Sci., 22, 423-434, 1956. Return.

Sheetz D.P., J. Appl. Polym. Sci., 9, 3759-3773, 1965. Return.

Kendall K., Padget J.C., Int. J. Adhesion and Adhesives, 2, 149-154, 1982. Return.

Dobler F., Pith T., Lambla M., Holl Y., J. Colloid and Interface Sci., 152, 1, 1992. Return.

Mason G., Brit. Polym. J., 5, 101-108, 1973. Return.

Kan C.S., Advaced Coating Fundamentals, TAPPI Notes, 101-107, 1993.

Sperry P.R., Snyder B.S., O'Dowd M.L.m Lesko P.M., Langmuir, 10, 2619-2628, 1994. Return.

Vanderhoff J.W., Tarkowski H.L., Jenkins M.C., Bradford E.B., J. Macromol. Chem., 1 (2), 361-397, 1966.

Vanderhoff J.W., Br. Polym. J., 2, 161-172, 1970.

Vanderhoff J.W., Paint and Varnish Prod., 25-37, Dec. 1970. Return.

Kendall K., Padget J.C., Int. J. Adhesion and Adhesives, 2, 149-154, 1982. Return.

Johnson K.L., Kendall K., Roberts A.D., Proc. R. Soc. Lond. A., 324, 301-313, 1971. Return.

Kan C.S., Advaced Coating Fundamentals, TAPPI Notes, 101-107, 1993. Return.

Sperry P.R., Snyder B.S., O'Dowd M.L.m Lesko P.M., Langmuir, 10, 2619-2628, 1994. Return.

Lamprecht J., Colloid & Polymer Sci., 258 (8), 960-967, 1980. Return.

Eckersley S.T., Rudin A., J. Coatings Technol., 62 (780), 89-100, 1990. Return.

Joanicot M., Wong K., Maquet J., Chevalier Y., Pichot C., Graillet C., Lindner P., Rios L., Cabane B., Prog. in Colloid & Polym. Sci., 81, 175-183, 1990.

Chevalier Y., Pichot C., Graillat C., Joanicot M., Wong K., Maquet J., Lindner P., Cabane B., Colloid and Polym. Sci., 270 (8), 806-821, 1992. Return.

Lissant K.J., J. Colloid & Interface Sci., 22, 462-468, 1966. Return.

Joanicot M., Wong K., Maquet J., Chevalier Y., Pichot C., Graillet C., Lindner P., Rios L., Cabane B., Prog. in Colloid & Polym. Sci., 81, 175-183, 1990. Return.

Chevalier Y., Pichot C., Graillat C., Joanicot M., Wong K., Maquet J., Lindner P., Cabane B., Colloid and Polym. Sci., 270 (8), 806-821, 1992. Return.

Voyutskii S.S., J. Polym. Sci., 32 (125), 528-530, 1958.

Voyutskii S.S., In Autohesion & Adhesion of High Polymers., Polymer Rev. 4, Pub. Wiley Intersci., N.Y., 1963.

Voyutskii S.S., Vakula V.L., Rubber Chem. Technol., 37, 1153-1177, 1964. Return.

Josefowitz D., Mark H., India Rubber World, 106, 33, 1942. Return.

Bradford E.B., Vanderhoff J.W., J. Macromol. Chem., 1, 335, 1966. Return.

Bradford E.B., Vanderhoff J.W., J. Macromol. Chem., 1, 335, 1966.

Bradford E.B., Vanderhoff J.W., J. Macromol. Sci. Phys., B6, 671-694, 1972. Return.

Vanderhoff J.W., Br. Polym. J., 2, 161-172, 1970.

Vanderhoff J.W., Paint and Varnish Prod., 25-37, Dec. 1970. Return.

Roulstone B.J., Permeation Through Polymer Latex Films, CNAA PhD Thesis, Trent Polytechnic, 1989.

Roulstone B.J., Wilkinson M.C., Hearn J., Wilson A.J., Polym. Int., 24, 87-94, 1991. Return.

Kast H., Makromol. Chem., Suppl. 10/11, 447-461, 1985. Return.

Kim Y.H., Wool R.P., Macromolecules, 16, 1115-1120, 1983. Return.

Yoo J.N., Sperling L.H., Glinka C.J., Klein A., Macromolecules, 23, 3962-3967, 1990.

Yoo J.N., Sperling L.H., Glinka C.J., Klein A., Macromolecules, 24, 2868-2876, 1991. Return.

Sperling L.H., Klein A., Yoo J.N., Kim K.D., Mohammadi N., Polym. Adv. Technol., 1 (3-4), 263-273, 1990. Return.

Hahn K., Ley G., Schuller H., Oberthür R., J. Colloid Polym Sci., 264, 61092-1096, 1986.

Hahn K., Ley G., Oberthür R., J. Colloid Polym Sci., 266 (7), 631-639, 1988. Return.

Yoo J.N., Sperling L.H., Glinka C.J., Klein A., Macromolecules, 23, 3962-3967, 1990.

Yoo J.N., Sperling L.H., Glinka C.J., Klein A., Macromolecules, 24, 2868-2876, 1991. Return.

Linné M.A., Klein A., Miller G.A., Sperling L.H., Wignall G.D., J. Macromol. Sci. Phys., b27 (2 & 3), 217-231, 1988. Return.

Wang Y., Winnik M.A., J. Phys. Chem., 97, 2507-2515, 1993. Return.

Kim K.D., Sperling L.H., Klein A., Macromolecules, 26 (17), 4624-4631, 1993. Return.

Kim H-B., Wang Y., Winnik M.A., Polymer, 35 (8), 1779-1786, 1994. Return.

Kim H-B., Winnik M.A., Macromolecules, 27, 1007-1012, 1994. Return.

Richard J., Maquet J., Polymer, 33 (19), 4164-4173, 1992. Return.

Sperling L.H., Klein A., Yoo J.N., Kim K.D., Mohammadi N., Polym. Adv. Technol., 1 (3-4), 263-273, 1990. Return.

Padget J.C., Moreland P.J., J. Coat. Technol., 55, 39-51, 1983. Return.

Wang Y., Winnik M.A., Macromolecules, 23 (21), 4731-4732, 1990.

Winnik M.A., Wang Y., Haley F., J. Coatings Technol., 64 (811), 51-61, 1992. Return.

Distler D., Kanig G., Colloid & Polymer Sci., 256, 1052-1060, 1978. (CBDE Translation) Return.

Aten W.C., Wassenburg T.C., Plastics and Rubber Proc. & App., 3 (2), 99-104, 1983. Return.

Distler D., Kanig G., Colloid & Polymer Sci., 256, 1052-1060, 1978. (CBDE Translation) Return.

Kanig G., Neff H., Colloid & Polym. Sci., 253, 29, 1975. (CBDE Translation.) Return.

Distler D., Kanig G., Colloid & Polymer Sci., 256, 1052-1060, 1978. (CBDE Translation)

Distler D., Kanig G., Org. Coat. Plast. Chem., 43, 606-610, 1980. Return.

Joanicot M., Wong K., Richard J., Maquet J., Cabane B., Macromolecules, 26 (12), 3168-3175, 1993. Return.

Joanicot M., Wong K., Maquet J., Chevalier Y., Pichot C., Graillet C., Lindner P., Rios L., Cabane B., Prog. in Colloid & Polym. Sci., 81, 175-183, 1990.

Chevalier Y., Pichot C., Graillat C., Joanicot M., Wong K., Maquet J., Lindner P., Cabane B., Colloid and Polym. Sci., 270 (8), 806-821, 1992. Return.

Roulstone B.J., Wilkinson M.C., Hearn J., Wilson A.J., Polym. Int., 24, 87-94, 1991.

Wang Y., Kats A., Juhué D., Winnik M.A., Langmuir, 8, 1435-1442, 1992. Return.

Wang Y., Winnik M.A., Macromolecules, 23 (21), 4731-4732, 1990.

Wang Y., Kats A., Juhué D., Winnik M.A., Langmuir, 8, 1435-1442, 1992. Return.

Nicholson J.W., Wasson E.A., Film Spreading and Film Formation by Waterborne Coatings., in Surface Coatings, Ch. 2, Vol. 3, 91-123, 1990. Return.

Napper D.H., Gilbert R.G., Emulsion Polymerization: The Mechanisms of Latex Particle Formation and Growth., in An Introduction to Polymer Colloids., Eds F. Candau, R.H. Ottewill, Pub. Kluwer Academic Publishers, 159-185, 1990. Return.

Funke W., Zorll U., Defazet, 29 (4), 146-153, 1975. (CBDE Translation.) Return.

Nicholson J.W., Wasson E.A., Film Spreading and Film Formation by Waterborne Coatings., in Surface Coatings, Ch. 2, Vol. 3, 91-123, 1990. Return.

Gould R.F., In Contact Angle, Wettability and Adhesion, Advances in Chemistry., Series 43, A.C.S., 1964. Return.

Kesting R.E., Membrane Polymers., Ch. 4, in Synthetic Polymeric Membranes:- A Structural Perspective, 2nd Ed., 106-179, 1985. Return.

Spitael J., Kinget R., Pharm. Acta Helv., 55 (6), 157-160, 1980. Return.

Distler D., Kanig G., Colloid & Polymer Sci., 256, 1052-1060, 1978. (CBDE Translation) Return.

Billmeyer jr. F.W., In Textbook of Polymer Science., 2nd Ed., Pub. John Wiley & Son, New York, 1971. Return.

Blandin H.P., David J.C., Vergnaud J.M., Illien J.P., Malizewicz M., J. Coatings Tech., 59 (746), 27-32, 1987. Return.

Vezin W.R.,Florence A.T., European Polym. J., 17, 93-99, 1981. Return.

List P.H., Laun G., Pharm. Ind., 42 (4), 399-401, 1980. Return.

Sullivan D.A., J. Paint Technol., 47 (610), 60-67, 1975. Return.

Hansen C.M., Ind. Eng. Chem., Prod. Res. Dev., 13 (2), 150-152, 1974. Return.

Sullivan D.A., J. Paint Technol., 47 (610), 60-67, 1975. Return.

Juhué D., Lang J., Macromolecules, 27, 695-701, 1994. Return.

Roulstone B.J., Permeation Through Polymer Latex Films, CNAA PhD Thesis, Trent Polytechnic, 1989. Return.

Abdel-Aziz S.A.M., Studies in Permeability in Polymeric Films., PhD Thesis, University of Strathclyde, 1976. Return.

Abdel-Aziz S.A.M., Anderson W., J. Pharm. Pharmac., 28, 801-805, 1976. Return.

Yaseen M., Raju V.S.N., Prog. in Org. Coatings, 10, 125-155, 1982. Return.

Okubo M., Takeya T., Tsutsumi Y., Kadooka T., Matsumoto T., J. Polymer Sci., Polymer Chem. Ed., 19, 1-8, 1981. Return.

Isaacs P.K., J. Macromol. Chem., 1 (1), 163-185, 1966. Return.

Bradford E.B., Vanderhoff J.W., J. Macromol. Chem., 1, 335, 1966.

Bradford E.B., Vanderhoff J.W., J. Macromol. Sci. Phys., B6, 671-694, 1972. Return.

Chainey M., Permeation Through Polymer Latex Films, CNAA PhD Thesis, Trent Polytechnic, 1984.

Chainey M., Wilkinson M.C., Hearn J., J. Appl. Polym. Sci., 30, 4273-4285 1985. Return.

Roulstone B.J., Permeation Through Polymer Latex Films, CNAA PhD Thesis, Trent Polytechnic, 1989. Return.

Roulstone B.J., Wilkinson M.C., Hearn J., Wilson A.J., Polym. Int., 24, 87-94, 1991. Return.

Wood D.A., In Polymeric Materials used in Drug Delivery Systems., 71-123, Date not known. Return.

Wang Y., Winnik M.A., Macromolecules, 23 (21), 4731-4732, 1990.

Wang Y., Kats A., Juhué D., Winnik M.A., Langmuir, 8, 1435-1442, 1992.

Wang Y., Winnik M.A., J. Phys. Chem., 97, 2507-2515, 1993. Return.

Roulstone B.J., Wilkinson M.C., Hearn J., Wilson A.J., Polym. Int., 24, 87-94, 1991. Return.

List P.H., Kassis G., Acta Pharmaceutica Technologica, 28 (1), 21, 1982. (CBDE Translation.) Return.

Balik C.M., Said M.A., Hare T.M., J. Appl. Polym. Sci., 38, 557-569, 1989.

Okor R.S., Study of Certain Factors Which Influence Solute Permeability of Polymer Films., PhD Thesis, University of Strathclyde, 1980. Return.

Roulstone B.J., Permeation Through Polymer Latex Films, CNAA PhD Thesis, Trent Polytechnic, 1989.

Roulstone B.J., Wilkinson M.C., Hearn J., Wilson A.J., Polym. Int., 24, 87-94, 1991. Return.

Okor R.S., Study of Certain Factors Which Influence Solute Permeability of Polymer Films., PhD Thesis, University of Strathclyde, 1980. Return.

Roulstone B.J., Permeation Through Polymer Latex Films, CNAA PhD Thesis, Trent Polytechnic, 1989.

Roulstone B.J., Wilkinson M.C., Hearn J., Wilson A.J., Polym. Int., 24, 87-94, 1991. Return.

I.U.P.A.C. Manual of Symbols & Terminology., Appendix 2, Part 1, Colloid & Surface Chemistry., Pure and Applied Chem., 31, 578, 1972. Return.

Balik C.M., Said M.A., Hare T.M., J. Appl. Polym. Sci., 38, 557-569, 1989. Return.

Kast H., Makromol. Chem., Suppl. 10/11, 447-461, 1985. Return.

Distler D., Kanig G., Colloid & Polymer Sci., 256, 1052-1060, 1978. (CBDE Translation) Return.

Kast H., Makromol. Chem., Suppl. 10/11, 447-461, 1985. Return.

Barrer R.M., Diffusion and Permeation in Heterogeneous Media., Ch. 6, in Diffusion in Polymers., Eds. Crank J., Park G.S., 165-217, Pub. Academic Press, 1968. Return.

Higuchi W.I., J. Polym. Chem., 62, 649-653, 1958. Return.

Peterson C.M., J. Appl. Polym. Sci., 12, 2619-2667, 1968. Return.

Chainey M., Permeation Through Polymer Latex Films, CNAA PhD Thesis, Trent Polytechnic, 1984.

Chainey M., Wilkinson M.C., Hearn J., Makromol. Chem. Suppl., 10/11, 435-446, 1985. Return.

Roulstone B.J., Permeation Through Polymer Latex Films, CNAA PhD Thesis, Trent Polytechnic, 1989. Return.

Berrie A.H., Pigments and Their Use in Polymers., Ch. 2 in The Chemist in Industry 1. Fine Chemicals for Polymers., E.S. Stern (Ed.), Clarendon Press, Oxford, 22-41, 1973. Return.

Seiner J.A., Ind. Eng. Chem. Prod. Res. Dev., 17, 302-317, 1978. Return.

El-Aasser M.S., Iqbal S., Vanderhoff J.W., Colloid and Interface Sci., Vol. V, Academic Press N.Y., 381, 1976. Return.

Durbin D.P., El-Aasser M.S., Vanderhoff J.W., Ind. Eng. Chem. Prod. Res. Dev., 23, 569, 1984. Return.

Hemenway C.P., Latimer J.J., Young J.E., Tappei, 68 (5), 102, 1985. Return.

Van Tent A., te Nijenhuis K., J. Colloid and Interface Sci., 150 (1), 97-114, 1992.

Van Tent A., te Nijenhuis K., Prog. Org. Coat., 20, 459-470, 1992. Return.

Okor R.S., Int. J. Pharm., 11 (1), 1-9, 1982. Return.

Goodhart F., Harris M.R., Murphy K.S., Nesbitt R.U., Pharm. Technol., 8 (4), 64-71, 1984.

Muhammad N.A., Boisvert W., Harris M.R., Weiss J., Drug Dev. & Ind. Pharm., 17 (18), 2497-2509, 1991. Return.

Hoy K.L., J. Paint Technol., 45 (579), 51-56, 1973. Return.

Roulstone B.J., Permeation Through Polymer Latex Films, CNAA PhD Thesis, Trent Polytechnic, 1989.

Roulstone B.J., Wilkinson M.C., Hearn J., Polym. Int., 27, 43-50, 1992. Return.

Durbin P.D., PhD Thesis, Lehigh University, Microfilm No. 8108193. Return.

Vanderhoff J.W., Br. Polym. J., 2, 161-172, 1970. Return.

Padget J.C., Moreland P.J., J. Coat. Technol., 55, 39-51, 1983. Return.

Juhué D., Lang J., Langmuir, 9, 792-796, 1993. Return.

Juhué D., Lang J., Colloids & Surfaces A: Physico. Chem. Eng. Aspects, 87, 177-185, 1994. Return.

Zhao C.L., Holl Y., Pith T., Lambla M., Colloid and Polym. Sci., 256 (9), 823-829, 1987.

Zhao C.L., Dobler F., Pith T., Holl Y., Lambla M., J. Colloid and Interface Sci., 128 (2), 437-449, 1989.

Zhao C.L., Holl Y., Pith T., Lambla M., Brit. Polym. J., 21, 155-160, 1989. Return.

Urban M.W., Evanson K.W., Polym. Comm., 31, 279-282, 1990. Return.

Roulstone B.J., Permeation Through Polymer Latex Films, CNAA PhD Thesis, Trent Polytechnic, 1989.

Roulstone B.J., Wilkinson M.C., Hearn J., Polym. Int., 27, 43-50, 1992. Return.

Kellaway I.W., Najib N.M., Int. J. of Pharm., 6, 285-294, 1980. Return.

List P.H., Kassis G., Acta Pharmaceutica Technologica, 28 (1), 21, 1982. (CBDE Translation.) Return.

Bondy C., Coleman M.M., J. Oil & Colour Chemists' Assoc., 53, 555-577, 1970. Return.

Pochan D., Morphology of Films Dried From Latexes Containing Water-Soluble Polymers: Dielectric Relaxation-Phenomena., Water-Borne & Higher-Solids Coatings Symposium., No. 14, 155-166, 1987. Return.

Bindschaedler C., Gurney R., Doelker E., J. Appl. Polym. Sci., 34, 2631-2647, 1987. Return.

Bierwagon G.P., J. Coatings Technol., 64 (806), 71-75, 1992. Return.

Asbeck W.K., Van Loo M., Ind. Eng. Chem., 41, 1470-1475, 1949. Return.

Asbeck W.K., J. Coatings Technol., 49 (635), 59-70, 1977. Return.

Rasenberg C.J.F.M., Huisman H.F., Prog. in Org. Coat., 13, 223-235, 1985. Return.

Anwari F., Carlozzo B.J., Chokshi K., Chosa M., DiLorenzo M., Heble M., Knauss C.J., McCarthy J., Rozick P., Slifko P.M., Stipkovich W., Weaver J.C., Wolfe M., J. Coatings Technol., 63 (802), 35-46, 1991. Return.

Bierwagon G.P., J. Paint Technol., 44 (574), 45-55, 1972. Return.

Rasenberg C.J.F.M., Huisman H.F., Prog. in Org. Coat., 13, 223-235, 1985. Return.

Rasenberg C.J.F.M., Huisman H.F., Prog. in Org. Coat., 13, 223-235, 1985. Return.

Floyd F.L., Holsworth R.M., Polym. Sci. Eng., 63, 180-185, 1990. Return.

Bierwagon G.P., J. Coatings Technol., 64 (806), 71-75, 1992. Return.

Floyd F.L., Holsworth R.M., Polym. Sci. Eng., 63, 180-185, 1990. Return.

Asbeck W.K., J. Coatings Technol., 49 (635), 59-70, 1977. Return.

Bierwagon G.P., J. Coatings Technol., 64 (806), 71-75, 1992. Return.

UpGo to Top of Page.

Copyright © Paul Steward 1995.

Created by Paul Steward.
Last Revised: Thursday, September 17, 1998 01:56 PM
Email: paul.steward@initium.demon.co.uk
Disclaimer.