Currently, most of the computational workflows are implemented to work with a specific Density Functional Theory (DFT) code and workflow management system. This means that users have either to learn many different DFT codes or to convert the workflow from one code to another. To accelerate the materials discovery, it is necessary to break this paradigm. In this talk, I will describe what are the advantages and limitations to implement interoperable workflows using a recently implemented workflow to discover intercalation battery cathode materials [1] and made it interoperable across several DFT softwares (VASP, CATSTEP, Quantum Espresso) and queuing systems (MyQueue, Aiida, SimStack, PipelinePilot). To allow interoperability between the implementations the initial input for each code must be standardized. We have therefore developed a JSON data-structure template containing all information on the crystal electrode materials, xc-functional, k-points, smearing, supercell scaling etc. necessary to run the entire workflow. This JSON structure allows for easy initialization of an implementation and highlights applicability to embed such workflows within API frameworks.
We have selected five cathode materials (Li and Mg as charge carriers) which will be run by all implementations and in the end report a value for the low, average, and high state of charge for all materials. To ensure the usability and searchability of the produced results ontological concepts will be embedded in the output. This will enable a semantic search of the data and increase the findability.

References
[1] F. T. Bölle, Batteries and Supercaps, 2020, voume 3, 488-498