CATALYST INDUCED DRYING

Alkyd based coatings harden by an autoxidation reactions. This autoxidation process is catalyzed by organometallic compounds. In many cases cobalt is used as a catalyst. However, some studies have shown that cobalt compounds can be carcinogenic. Therefore, manufactures are searching for possible alternatives. To achieve this knowledge about the precise working principle is required. Using the MRI setup, described on the other pages, the exact effect of the catalyst can be visualized. We have looked at the concentration dependence of cobalt, the effect of manganese as a catalyst, as well as the effect of the addition of secondary driers.

Fig 1. Squared front position against the concentration of a cobalt based catalyst

A clear concentration dependence of the catalyst concentration can be found, see figure 1. In all cases the drying is oxygen diffusion limited. However, the process is clearly affected by the concentration. Based on the equations derived, the change in curing speed can be addressed to three variables, the oxygen solubility at the surface of the coating, a change in diffusion constant, or a change in oxygen present in the final network structure.
 

Fig. 2. Drying observed when manganese is used as a catalyst. Left figure shows NMR measurements. Right figure shows computer simulations.

When manganese is used as a catalyst the drying of the alkyd coating is more homogeneous. As is clearly visible in figure 2. Oxygen diffusion is no longer a limiting factor. A reaction-diffusion model can be used to explain the drying properties, indicating a higher diffusion constant and lower reaction rate. However, this might largely affect the final hardness. A study has shown a decreased final hardness in coatings using manganese as a catalyst.

Fig 3. This figure shows the oxygen diffusion model, as well as the curing observed in the uncured region.

When we add Ca and Zr as secondary driers, the drying process speeds up. Since the process is already diffusion limited, change in the reaction rate is not a parameter that can explain this change. Again the change in curing speed can only be addressed to a change in three variables, the oxygen solubility at the surface of the coating, a change in diffusion constant, or a change in oxygen present in the final network structure. Which is the proper explanation we have no clue. Another effect observed in solvent borne alkyd coatings is that in the deeper layers also curing takes place. We can model this by diffusion of less reactive compounds, formed at the cross-linking front, into the deeper layers of the coating. As a result the coating thickness is one of the important variables that increases this effect. Experiments have confirmed the model.