(665g) Development of Particle Packed Layer in the Drying Process of Latex Paints

In order to minimize the amount of volatile organic compound (VOC) in the coating industry, latex paints are frequently used as the substitution for VOC based paints. In latex paints, submicron sized resin particles are dispersed in water and, the resin particles are packed during drying and finally formed a film over the coated material after drying. One of the important physical properties of latex paints is a glass transition temperature, Tg, of dispersed resin particles. That is to say, the resin particles will produce uniform film when drying at the temperature higher than Tg, while the film becomes brittle at a lower drying temperature. As for the packing process of dispersed resin particles, the resin particles are first accumulated in the region near the drying surface and the packing of particle layer developed in the depth direction from the surface. This process of the particle packing layer formation have been directly observed using CryoSEM technique by Scriven or Francis. On the contrary, it is well known in the drying process of particulate dispersion system that the boundary of highly accumulated wet region and porous particle packed region can be clearly observed as a drying front and the drying front proceeds from the outer surface to the internal part of the material to be dried. The transport phenomena in the porous particle packed layer significantly affects the drying process of wet material inside of the drying front. So, we have experimentally investigated the drying process of latex paints focusing on the development of dried particle packed layer.

The latex paints used in this study contain two kinds of dispersed resin particles having different Tg value. One is Tg = -11Cº, which means the dispersed particles are flexible under a room temperature, producing uniform film after drying. The other is Tg = 108Cº, which means solid like resin particles are dispersed in water, resulting in a brittle dried film. The concentration of resin particles in each latex paint was adjusted to 23vol%. We have carried out direct observation of one-dimensional drying process of latex paint in a one-side opened square cell composed of two glass plates with the separation of 80um. Three sides of a square thin glass plate (18mm x 18mm) were stuck onto the other glass plate using double-side tapes, the thickness of which is 80um. The latex paint inserted into the thin space having the dimensions of 15mm x 15mm surrounded by top and bottom glass plates and surrounded by the double side tapes will be dried only from the open side. Then, the drying process of the latex paint was considered as uni-directional drying. The test cell placed on a transparent glass heater and then the drying process of the latex paint was observed using digital microscope with dark field observation mode. The temperature of the heater was controlled as 25 and 40Cº. In this observation, we could clearly observed that the drying front proceeded uniformly in the direction normal to the open side of the test cell and measured the thickness of the particle packed layer as a distance between the open side to the drying front. The moving speed of the drying front was also calculated from the time variation of particle layer thickness.

The moving speed of the drying front was initially very large and then gradually decreased with time. In the initial stage of drying process until the thickness of the particle layer became much larger than the glass plate separation, the drying process could not considered as uni-directional. Additionally, the air flow near the open side significantly enhanced the evaporation of water because of thin particle packed layer. As for the effect of Tg on the drying process, the moving speed of the drying front for the latex paint of Tg = 108Cº was much larger than that of Tg = -11Cº. We also have carried out rheological measurement of latex paint in order to elucidate the difference in particle aggregation. As a result, the latex paint of Tg = 108Cº were well dispersed showing a small storage modulus in a linear viscoelastic regime and a lower viscosity, while more aggregated latex paint of Tg = -11Cº indicated a high viscosity at relatively small shear rates and a larger storage modulus at an oscillatory shear application with small shear strains. Therefore, since the aggregated structure of resin particles hold water in the structure and deteriorates water evaporation from the latex paint, the moving speed of the drying front at Tg = -11Cº became relatively slow. In the following drying stage, the gradual decrease in the moving speed of the drying front was reasonably described by the development of particle packing layer. That is to say, as increasing in the thickness of particle packed layer, the water evaporated at the drying front required long time to diffuse out of the particle layer.

Since the development process of particle packer layer was governed by the diffusion of water vapor through porous particle packed layer, the moving speed of the drying front should be described by the one-dimensional Fick’s diffusion equation in a porous material. Additionally, the permeation of water vapor though a particle packed layer was modeled by Cozeny-Karman equation introducing void fraction of particle layer. Therefore, we have obtained the formulation where the product of the moving speed of drying front and the thickness of particle layer is constant, and the void fraction in the particle layer can be determined from the constant value. As a result, we found that the decreasing behavior in the moving speed of drying front have been successfully correlated by the model established for each latex paint at different drying temperatures. Also, the packing density of the particle packed layer at Tg = -11Cº was much larger than that for the packing density of maximum packing layer of solid spheres, which indicates flexible resin particles are deformed and then packed in the drying process. The higher packing density at the higher drying temperature indicates the resin particles seemed to have been well deformed and more tightly packed. On the other hand, the packing density of solid-like resin particles of Tg = 108Cº was much smaller than the maximum packing density of solid sphere compensating for large clack formation in a particle packed layer.