Predicting Properties of Architectural Coatings by Modeling Data Generated from a Designed Mixture Experiment

By Steven De Backer , Michael P. Diebold, Steven P. Bailey

Titanium dioxide (TiO₂) is a well-known pigment used in the coatings industry. It is also often one of the more expensive in-gredients in the formula. Therefore, it should be of no surprise that a lot of materials are proposed by a variety of suppliers to reduce the amount of TiO₂, often claiming to have no influence on the properties of the paint. In this study, we seek a better understanding of the impact of a (partial) replacement of the TiO₂ with a selection of these alternative materials on the basic properties of the paint (e.g., hiding power, mechanical strength, color, etc.).

Since many architectural coatings are formulated above the critical pigment volume concentration (CPVC), we focus in this study on this formulation area. The traditional approach for evaluating the effect of a new ingredient in a paint formula is to add the ingredient to the formula, eventually compensating by (partially) leaving out another ingredient and then testing the new paint for a set of properties.¹

The formulator appears to build, with only a few experiments, a feel for the potential of this new ingredient. Although the trial-and-error approach can be very useful for initial screening, it can be very time-consuming to optimize paints because a lot of parameters have to be balanced and very often compromises have to be made. With this study, we will use an experimental mixture design approach for the formulation of paints. ^2-4 In this approach, the boundaries for the composition paint are determined in advance (i.e., minimum and maximum values of all the ingredients).