Copolymer adsorption and the effect on colloidal stability.

The main aim of the work described in this thesis is to study the effect of different types of copolymers on the stability of aqueous oxide dispersions. Such dispersions are a major component in water-borne paints. In order to obtain a better insight in steric stabilisation we first investigated the relation between the adsorbed amount and layer thickness, and paid attention to the effect of the type of copolymer on the adsorbed amount. We also studied the adsorption kinetics as these are relevant for industrial purposes.An introduction on steric stabilisation is given in Chapter 1. For block copolymers the solvent may be non-selective or selective. In a non-selective solvent both blocks are solvated and the polymer molecules are likely to be in a non-aggregated conformation. However, in a selective solvent the molecules form micelles in which the non-soluble blocks are clustered together, surrounded by a layer of solubilised chains. The adsorption kinetics are expected to be affected by the existence of such micelles. Another important feature for the adsorption of block copolymers is the selectivity of the surface. When only one of the blocks has affinity for the surface this will give rise to selective adsorption. On the other hand, the adsorption of a block copolymer in which both blocks have affinity for the surface is non-selective. The resultant polymer layer will differ for both cases. In thesis we studied selective and non-selective adsorption from a selective and a non-selective solvent. As the architecture of the copolymers is also relevant we paid attention to the adsorption of both block copolymers and graft copolymers.In Chapter 2 we describe the properties of spread monolayers of polystyrene-poly(ethylene oxide) (PS-PEO) diblock copolymers at the air-water interface. The surface pressure and the thickness of the layer were measured as a function of the adsorbed amount. The thickness was determined with neutron reflectivity measurements.Upon compression of the polymer monolayer the surface pressure increases over the entire experimental range of compression. At low coverage the adsorbing PEO block forms a flat "pancake" structure at the surface. When the surface area per molecule is decreased the PEO is pushed out of the surface layer into the solution to form a "cigar" or "brush" structure, which is firmly anchored by the PS block. Some scaling analysis have suggested that this desorption occurs as a first-order surface phase transition. When the polymer layer is compressed further, so that the surface density σincreases, the chains stretch and the thickness H of the layer increases too. Theories predict that H scales as Nσ <sup>1/3</SUP>, where N is the number of monomers per polymer chain. This is confirmed by our results. However, our experimental data do not show the first-order surface phase transition between pancake and brush. Numerical self-consistent-field calculations also show a gradual transition rather than a first-order phase transition.In Chapter 3 we present a study on the non-selective adsorption of two series of diblock copolymers, poly(vinyl methyl ether)-poly(2-ethyl-2-oxazoline) and poly(2-methyl-2oxazoline)- poly(ethylene oxide), from aqueous solution on a macroscopically flat silicium oxide surface. The adsorbed amounts in this study, and in that of Chapters 4 and 5, were measured with an optical reflectometer in an impinging jet flow cell. The hydrodynamic layer thickness was determined by dynamic light scattering.The different blocks in the copolymers all have affinity for the silica surface. In all cases there is a small difference between the segmental adsorption energies of the two blocks, giving rise to non-selective adsorption of the block copolymers. For the two types of block copolymers used in this study, the adsorbed amount as a function of block copolymer composition shows a shallow maximum; at this maximum the longest block is also the more strongly adsorbing block. The same trend is found for the hydrodynamid layer thickness. These findings differ from theoretical predictions concerning selective adsorption, where a pronounced maximum is found for a short anchor block. With numerical self-consistent field calculations we demonstrate that the same trends as in our experimental findings can be predicted by theory. In non-selective adsorption of diblock copolymers, with a small difference between the adsorption energies of the blocks, both blocks compete for the same adsorption sites on the surface. When the blocks are incompatible they try to avoid each other, which promotes an anchor-buoy structure. These factors then give rise to a maximum in the adsorbed amount as a function of the block copolymer composition. At this maximum the longest block is also the more strongly adsorbing block. The adsorbed layer has the typical anchor-buoy structure which is necessary for an effective steric stabilisation, but this structure is less pronounced than for selective adsorption.The kinetics of adsorption of diblock copolymers can be very slow if the polymers form micelles in solution. In Chapter 4 we compare the experimental adsorption rates on silica and titania with the theoretical flux of copolymer molecules towards the surface for four poly(dimethyl siloxane)-poly(2-ethyl-2-oxazoline) diblock copolymers with the same block length ratio but different molar masses. In aqueous solution these block copolymers form large polydisperse micelles with a very low critica <strong>l</strong> micellisation concentration (lower than 2 mg 1-1).On both surfaces the adsorption behaviour is governed by the anchoring of the hydrophobic siloxane blocks The adsorption kinetics are affected by the exchange rate of free polymer molecules between micelles and solution. For the three smallest molar masses the exchange rate is fast compared to the time a micelle needs to diffuse across the diffusive layer. Before the micelles arrive at the surface they have already broken up into free polymers. Because the cmc is very low, the experimental adsorption rate is determined by the diffusion of micelles towards the surface. For the longest polymer this is not the case: the exchange of polymer molecules between micelles and solution is now relatively slow. As the micelles do not adsorb directly, the adsorption rate is retarded by the slow exchange process. We were able to make an estimate of the micellar relaxation time, i.e., the time a micelle needs to break up. For the largest polymer the relaxation time is of the order of a few tens of seconds. The other polymers have a micellar relaxation time that is shorter than roughly one second.The adsorption increases linearly as a function of time, up to very high adsorbed amounts where it reaches a plateau. Such high adsorbed amount is expected for strongly (and selectively) adsorbing diblock copolymers with a relatively short anchor block. The adsorbed amount on silica is considerably higher than on titania. The reason is probably that the hydrophobic block is more strongly anchored to a silica surface than to titania, so that the density of the adsorbed layer can become higher on silica.In Chapter 5 we investigate the interfacial behaviour of graft or comb copolymers. We compare the adsorption of graft copolymers with an adsorbing backbone and nonadsorbing side chains to the reverse situation of adsorbing side chains and a nonadsorbing backbone. Two high- molar-mass poly(acryl amide)-graft-poly(ethylene oxide) copolymers with different side chain densities were used in this study.On titania only the backbone of these polymers adsorbs and the side chains do not. The adsorbed amount is then about the same as that found for the homopolymer without side chains. On the other hand, on silica the side chains adsorb and the backbone does have no affinity for the surface. For both polymer samples we observe a maximum in the adsorbed amount as a function of time ("overshoot"), after which the adsorbed amount decreases and a plateau is reached. The plateau adsorbed amount on silica is much higher than on titania and also much higher than for both types of homopolymers. Upon adsorption the graft copolymers initially adopt a conformation in which only part of the side chains are adsorbed. Following the overshoot, the graft copolymers show a decrease in the total adsorbed amount. The overshoot depends on the polymer concentration, which suggests that it is not caused by conformational changes in the adsorbed layer but by an exchange process between surface and solution.Differences in graft distribution and graft density in the polymer sample are probably responsible for the displacement of adsorbed chains by polymer molecules from solution. The average number of grafts per molecule is rather low in our polymer samples. On statistical grounds there is probably an appreciable polydispersity in graft distribution and in graft density. Molecules in which the grafts are clustered to some extend can displace molecules with more regularly separated grafts, and molecules with a high graft density can displace those with a lower number of side chains. The newly arriving molecules can then adsorb in a flatter conformation with a lower adsorbed amount as the extra loss in conformational entropy is compensated by the gain in adsorption energy.The effect of the polymers used in Chapters 3 to 5 on the stability of an aqueous silicium oxide dispersion is described in Chapter 6. The time-dependent increase of the average hydrodynamic radius of silicium oxide aggregates in the presence of electrolyte was measured. The increase of this radius with time is a measure of the aggregation rate of the dispersion. The effect of polymers on the stability of a dispersion was studied by adding polymer to the dispersion and recording the effect in the aggregation rateComparison of the aggregation rate of this "protected" silica with that of uncovered silica particles gives then an indication of the steric stabilisation by the adsorbing polymers.Four different series of diblock and graft copolymers were used in these stability measurements. For two series of non-selectively adsorbing diblock copolymers, poly(vinyl methyl ether)-poly(2-ethyl-2-oxazoline) and poly(2-methyl-2-oxazoline)poly(ethyiene oxide), we find a good correlation between the adsorbed amount and the stabilising effect. A higher adsorbed amount provides a better steric stabilisation. Nevertheless, for these polymers the adsorbed amounts are not high enough (up to about 1.2 mg M <sup>-2</SUP>) to protect the dispersion completely against aggregation. A series of amphiphilic diblock copolymers of poly(dimethyl siloxane)-poly(2-ethyl-2-oxazoline) with very high adsorbed amounts (between 3.5 and 8 mg M <sup>-2</SUP>) give excellent steric stabilisation of the dispersion. Adsorbed layers of the two graft copolymers of poly(acryl amide)-poly(ethylene oxide), with a non-adsorbing backbone and adsorbing side chains, are also effective in preventing the silica from aggregating. Even though the adsorbed amount of these graft copolymers is only around 1.3 mg M <sup>-2</SUP>, which is much lower than that of the amphiphilic polymers, aggregation is completely prevented.The best steric stabilisation is found for those systems in which either the surface or the solvent is selective. In practical aqueous systems, however, it is difficult to synthesise diblock copolymers in which both blocks are soluble and where only one of the blocks has affinity for the surface. We have shown that copolymers with a different architecture, graft copolymers, also can provide good steric stabilisation and may be a good alternative to diblock copolymers. Very good steric stabilisers are amphiphilic diblock copolymers in a selective solvent. However, it is important that the hydrophobic blocks are flexible enough for fast adsorption kinetics and that they completely wet the surface. Which copolymer should be chosen for the steric stabilisation of a practical colloidal system depends largely on the nature of the particles and the solvent, and on the availability of suitable copolymers