(520b) A New Approach to Predict Adsorption in Metal-Organic Frameworks with Unsaturated Metal Sites

Recent years have seen an exponential increase in the number of computational studies aiming to predict adsorption in MOFs. A particularly promising trend is the use of large-scale computational screening of MOFs for adsorption applications, so as to identify the most promising candidates for further experimental study among a wide pool of available materials. These approaches have tremendous potential to reduce the time and money currently spent on experimental testing of MOFs, fast-tracking their translation into commercial applications. However, a crucial but often overlooked fact is that conventional molecular models for adsorption in MOFs neglect important physics of the process, and thus can lead to spectacularly inaccurate predictions. This is particularly the case for MOFs that possess coordinatively unstaturated sites (CUS), or open metal sites – the specific nature of the coordination bonds formed by these sites with particular adsorbate molecules is not captured by standard force fields. The impact of neglecting these interactions can be huge, particularly considering that MOFs that contain CUS are among the most promising materials for gas storage and separation applications, precisely due to these strong and selective binding sites [1].

Our communication presents a generally applicable and transferable approach for incorporating specific coordination-type interactions between CUS and adsorbate molecules into standard molecular models, based on a combination of quantum mechanical calculations with classical Monte Carlo simulations. Crucially, our approach makes use of only a small number of QM calculations in the vicinity of the CUS, thus keeping computational expense low, while making use of conventional classical approaches to describe interactions away from the CUS [1, 2]. We demonstrate generality by applying our model to describe adsorption of both polar and non-polar adsorbates (paraffins, olefins, nitrogen and carbon monoxide) [3], and demonstrate transferability by predicting adsorption in different MOFs with the same type of metal unit (copper paddlewheel) [4]. In all cases, excellent agreement with experimental adsorption isotherms is obtained. This means that the procedure is amenable for incorporation into large-scale screening efforts, and has the potential to greatly improve the accuracy of computational screening of MOFs for adsorption applications. Overall, our new model provides detailed insight into the molecular level adsorption mechanisms on MOFs with CUS, and constitutes a useful tool to design new materials for challenging separations.

[1] Fischer, M.; Gomes, J. R. B.; Jorge, M. Mol. Simul. 40 (2014) 537-556.

[2] Fischer, M.; Gomes, J. R. B.; Froba, M.; Jorge, M. Langmuir 28 (2012) 8537-8549.

[3] Campbell, C.; Gomes, J. R. B.; Fischer, M.; Jorge, M. submitted for publication.

[4] Campbell, C.; Ferreiro-Rangel, C. A.; Fischer, M.; Gomes, J. R. B.; Jorge, M. J. Phys. Chem. C 121 (2017) 441–458.