(522g) Establishing Discrete Ising Model for Zeolite Deactivation: Inspiration from the Game of Go

Cai, D., Tsinghua University
Zeolite catalysis is now playing important roles both in modern catalysis studies and in industrial applications. A variety of coking processes not only lead to the deactivation, but also turn valuable feedstocks into useless coke In this regard, determining the coking relationships in order to get rid of it has become one of the final subjects in zeolite catalysis. Among these processes, Methanol-to-olefins (MTO) process has become a hot topic and attracted many attentions.

If we considered zeolite coking as phase transition in Ising model, under the high acid density conditions, the coke blocked interconnection of zeolite channels and then trapped feedstocks or products in the channels. Based on it, we illustrated a new model named Discrete Ising Model for Deactivation, inspired by the Game of Go.

Deactivation process happens in zeolite cages stochastically, and the connectivity among cages change accompany with the deactivation process. During the deactivation process, blocked sites may suddenly appear and result in the waste of active sites with trapped feedstocks or products in it. Systematically, our model was built based on the following hypothesis: Firstly, deactivation is a stochastic process, with a complete independence among different cages with acid sites. Coke species appear and grow up randomly in the same cage.

As the reaction proceeds, the number of reachable sites decreases, in agreement with our common sense. However, while the reaction and deactivation proceed to a certain extent where ~60% of the cages are deactivated, a slight increase of coke will lead to a gradually climb of blocked site. That is, the increase of acid density will lead to a significantly nonlinear increase of coke yield and deactivation of catalyst. A potential danger in industrial application can be revealed by our result. During the manufacture in industry, the sensitivity between reaction rate and deactivation percentage at a certain point is the origin of its fragility.

In conclusion, we developed the Discrete Ising Model for Deactivation (DIMD) considering the changing connectivity among channels and cages. With the help of numerical simulation, we captured the nonlinear change along with time and sensitivity to the acid percentage. Based on analysis above, we can predict some critical point at carbon percentage of 60% in reaction process and acid percentage of 57% in catalyst synthesis. We also get the carbon distribution that is consist to the experimental result. Generally, this model sheds lights on the study of deactivation in zeolites. An unexpected nonlinear change of blocked percentage is observed and illustrates complexity in zeolite networks.