In recent decades, computer simulations have been playing an ever-increasing role in improving our understanding of the world around us. Advancements in molecular-level simulations of adsorption phenomena for example, have shown excellent matches between experimental measurements and predicted data. This has led us to wonder if we could use super-computers to discover top-performing adsorbents before going to the laboratory to synthesize and test them.
We have recently been investigating the use of metal-organic framework adsorbents, or "MOFs", in an emergency healthcare environment, specifically looking at the benefits MOFs could provide for oxygen storage tanks. A MOF is a porous material, which can adsorb oxygen enabling it to be stored at much lower pressures. By utilising high-throughput computational screening we were able to identify MOFs with the most promising oxygen-storage properties, comparing them to a database of thousands of already-synthesised structures.
To assist with the analysis of the large generated datasets, we also developed an interactive 5D visualisation tool for producing 1000s of unique structure-property plots to convey the full information obtained on oxygen storage for all the structures under study. This is freely accessible at http://aam.ceb.cam.ac.uk/mof-explorer. This tool allows users to visualise oxygen gravimetric and volumetric uptakes with respect to different combinations of structural properties such as void fraction, pore and window size, heat of adsorption and surface area to better understand their role in oxygen adsorption performance, in up to 5 dimensions simultaneously.
The result of this work is that we have discovered a MOF which now holds the world-record for oxygen storage. This MOF, known as UMCM-152, exceeds the next-best known material by 22.5% and 15% in its gravimetric and volumetric deliverable capacities respectively. Utilising UMCM-152's record-breaking properties, we can make oxygen tanks safer, easier to handle and more cost-effective than current high-pressure alternatives. What is more, we have proven the principle of large-scale computational screening to guide material synthesis for rapid discovery in other applications - the possibilities now are endless!
The paper in Nature Communications is here: https://go.nature.com/2IGJDTN