Flux algorithm might miss important adsorbates

[rwest]

Just found this in an old email thread. Thought I'd post it here before archiving the email, so we can deal with it later.

I have a question for you guys about a consequence of the flux constraint and exclusion of edge species. Did you guys discuss the possibility that an edge species has a low rate of formation, but a much lower rate of consumption that would lead to a slow, but important accumulation of a species on the catalyst surface? For example, if your thermodynamic estimates suggest that an edge species would have a significant coverage at the equilibrium conditions of the last iteration, then should this species continue to be included in the core? It seems possible at first thought that this could lead to entropic or just site competition effects that could be missed by the flux constraint and that could become important over the lifetime of the intended reaction – but maybe you guys discussed this and decided that it wasn’t a concern?

The short answer is: yes, it is a concern.
Because there are inherent differences between homogeneous and heterogeneous phenomena, I think that RMG-Cat will require further modifications to the core algorithm(s). Catalyst poisoning is likely to be one of them, since site-blockage doesn't really have a neat gas-phase analogue. Richard and I have tossed around various ideas, including some test for equilibrium or steady state, etc., but it needs to be implemented and tested.