Gradient Boosted Machine Learning Model to Predict H2, CH4, and CO2 Uptake in Metal-Organic Frameworks Using Experimental Data

Predictive screening of metal-organic framework(MOF) materialsfor their gas uptake properties has been previously limited by usingdata from a range of simulated sources, meaning the final predictionsare dependent on the performance of these original models. In thiswork, experimental gas uptake data has been used to create a GradientBoosted Tree model for the prediction of H-2, CH4, and CO2 uptake over a range of temperatures and pressuresin MOF materials. The descriptors used in this database were obtainedfrom the literature, with no computational modeling needed. This modelwas repeated 10 times, showing an average R (2) of 0.86 and a mean absolute error (MAE) of & PLUSMN;2.88 wt % acrossthe runs. This model will provide gas uptake predictions for a rangeof gases, temperatures, and pressures as a one-stop solution, withthe data provided being based on previous experimental observationsin the literature, rather than simulations, which may differ fromtheir real-world results. The objective of this work is to createa machine learning model for the inference of gas uptake in MOFs.The basis of model development is experimental as opposed to simulateddata to realize its applications by practitioners. The real-worldnature of this research materializes in a focus on the applicationof algorithms as opposed to the detailed assessment of the algorithms.