Black Box Discovery and Optimisation of New Catalysts Using an Evolutionary Strategy

Black Box Discovery and Optimisation of New Catalysts Using an Evolutionary Strategy 

Recently, the evolutionary strategy (ES) has been successfully applied for the iterative optimisation of catalysts in a series of reactions [7, 10, 15, 27-30]. For all the case studies reported, the optimisation process was found to converge rapidly after typically four to six generations. 

For the present case study (Strategy WGS 1), seven generations of 72 catalytic materials each have been prepared and tested. CO conversion at standard conditions for the ten best compositions of each generation is shown in Fig. 10.7. 

From the chemical viewpoint, some essential elements and support materials were identified, such as Cu, Ce and Mn, as having the highest frequency in the catalytic materials (Fig. 10.8). More precisely, by checking synergistic and antago- 

This methodology was also applied in the same WGS domain, but with added new objective functions such as non-pyrophoricity of the systems, which was simply defined by excluding large amounts of copper. Again, new promising formulas were discovered, which are now in the process of being scaled-up [13]. 

---