Construct a pkl file for new adsorbates?

Thanks so much for your fantastic work on Open-Catalyst-Dataset.
I noticed that you used the pkl file to obtain the adsorbate object in the .

If I want to use your model to predict the adsorption energy of some new adsorbents (such as benzene) on catalysts, how can I construct a new pkl file? Or is there another way?

BTW, if I want to construct a new bulk, is there any method?

Finally, how to use the ML relaxations(ocpmodels) way to predict the energy?


Hi -

The database of adsorbates and bulk materials are found here - Open-Catalyst-Dataset/ocdata/databases/pkls at main · Open-Catalyst-Project/Open-Catalyst-Dataset · GitHub. If you want to add new species you can directly add it to these and use these databases in your code directly. This also gives you an idea of the format it expects your new species to be in.

To compute the adsorption energies, the following tutorial can help walk you through the steps - AdsorbML/tutorials/adsorbml_walkthrough.ipynb at main · Open-Catalyst-Project/AdsorbML · GitHub. Let us know if you have any questions.

Thank you for your answer. I think I understand.