Eugenio Bonetti - A Cellular Automata model for the release of corrosion inhibitors from primer coatings

Presentation - pdf

Eugenio Bonetti a, Flor R. Siperstein a, Peter Visser d, Simon Gibbon c
a
Department of Chemical Engineering and Analytical Science, The University of Manchester, Manchester, UK

b AkzoNobel, Sassenheim, The Netherlands
c AkzoNobel, Felling, UK

Primer coatings are commonly employed to provide active corrosion protection through the release of corrosion inhibitors upon exposure to moisture. Experimental studies have highlighted the key role of the microstructure of the coatings in the process of leaching, as the pigment particles form a network of clusters that provides the pathway for the transport and release of the active species. However, a clear relationship between the characteristics of the microstructure and the observed leaching behaviour has not been established. Understanding and controlling this relationship will lead to the possibility to design better coatings with improved protection performance.

For this purpose, the aim of this work is the development of a model to investigate the relationship between microstructure of coatings and release of corrosion inhibitors. The modelling approach consists of an algorithm to generate virtual coatings with specific formulation parameters and a method based on a Cellular Automata model to simulate the dissolution and transport of inhibitors. The results of simulations show the changes in the release of inhibitors which result from changes in the microstructure of the coating. This proves that modelling and simulation techniques can be valid tools to assist the research and help improving the understanding of the effect of the formulation on the protection properties of organic coatings.

Q&A

From  Simon Gibbon : Could you speculate on what is missing in the model?
From  Philip Gill : What do you think is the main understanding missing from them model?
Answer: I think as I said, the characterisation of the microstructure is quite important. This work is based on the reconstruction of a pixel generation of the material. It is not based on tomographic data of a coating to be used directly as input of the simulation. One thing that is missing is the validation of the microstructure. It is necessary to be sure that the virtually generated microstructure reproduces the main features of the real microstructure. Also regarding the process of release it is important to better understand the parameters of the real process. For example, it is considered that the fillers are inert and do not play any role in the leaching and corrosion protection, but recent work has shown that fillers actually dissolve and leach out of the coating which means that they are not as inert as they thought to be. It is important to better identify the processes that occur during the leaching.

 From  Stefan Bon : How does the geometry (aspect ratio e.g.)of the particles influence the percolation threshold?
Answer: the particle shape is another parameter that affects the structure and the arrangement. I noticed that in this work I used a random shape particles that is the shape of the pigments in the coatings formulation we considered. For example, strontium chromate has more elongated shape, like needle and this result in decrease in the percolation threshold, and the formation of a large cluster forms at low PVC than in the formulations we considered.

From  Beatriz Rocha De Moraes : What type of pigment did you study? can they affect in the analyzes performed?
Answer: I based the work on lithium carbonate, magnesium oxide, barium sulphate, and titanium dioxide, it is mainly the same formulation that Ander presented.

From  Koray Yıldırım :  I have a question regarding the simulation principle, is the only leaching interface where the electrolye is in touch with the inhibitor salts? was the transport through polymeric media considered in this model?
Answer: No, I didn’t considered the transport in the polymer phase. The presence of the ions was not detected in the polymer during experiments. Possible some degree of transport happens, but the transport through the polymer is negligible for the overall release. The mechanism that provides the most release is through the cavity. It is possible that some transport occurs through the polymer but it is not the main mechanism.

From  Simon Gibbon : WRT Eugenio/Ander computationally you know PSD / shape / etc., experimentally very difficult to measure - this limits our abilities to predict / control properties - how do we move this forward?  Is the issue of particle characterisation similar in other formulated products?
Answer: Quantitative data for validation is not available, but you need something from the real system to show that the model does represent the real coating. The model does not have interaction between particles, it can be included but the complexity of the model will increase, but then the requirement of parameters as input may increase, as you would need to quantify these interactions.

From  Anabelle Legrix : Many thanks Eugenio, very interesting modelling work with great insight into the leaching mechanism and the effect of particle size/packing etc. Modelling of packing depending on size and shape is indeed difficult with broad psd and various shapes.

From Simon: Does anybody knows from other formulations have similar challenges where knowing what happens of particles in liquids allow you to model what happens in the formulation? It seems that particle dispersion and what happens in the fluid stage is important, where characterisation is a limitation. It seems that we can model more than what we can characterise.
Answer: we can introduce into the model a lot more complexity.
From  Philip Gill : Simon - Yes for novel mixing methods (solids in polymers/liquids) such as resonant acoustic mixing, we are really limited on what we can measure.  So modelling of particle distribution/mixing is key for the future.