Two-dimensional magnetism and the SCAN functional

Abstract

Lower dimensional magnetism has gained a great deal of attention in the growing field of spintronics. This is due to the low energy costs of manipulating the spin degree of freedom and heightened sensitivity to external perturbations. There is however the major challenge that, since the synthesis of such materials is very difficult, studying their properties via experiments is infeasible. A way to circumvent this problem is to use computational tools to study lower-dimensional magnetic materials. Density functional theory is the most widely used framework in materials simulations for it provides a healthy compromise between accuracy and computational resources needed. Its performance however depends on how one treats the many-body effects, the physics of which is captured in a functional called the exchange-correlation functional.

The aim of this thesis is to study the performance of the SCAN and r2SCAN exchange-correlation functional when studying lower-dimensional magnetism. There are two reasons for this - (1) SCAN and r2SCAN, being newly developed functionals, are yet to be tested extensively, and (2) If they are shown to accurately predict the properties of lower-dimensional ferromagnets then they would be extremely useful in circumventing the problem mentioned in the first paragraph. To this end, we have picked the few lower-dimensional ferromagnets whose mono-layer and few-layer forms have been experimentally isolated (CrI3, CrBr3, CrCl3, CrGeTe3, and CrTe2 ) and test if SCAN and r2SCAN can reproduce their determined properties. Our results show that both SCAN and r2SCAN do very well in predicting the structural and magnetic properties of the considered lower-dimensional ferromagnetic materials. Interestingly, both functionals seem to do this without needing to be supplemented with an empirical +U term. While SCAN provides better estimates of cell parameters, the r2SCAN functional seems to give more accurate predictions for the Curie temperatures. We expect the work done in the thesis to improve confidence in the performance of these two functionals and encourage one to use them in predicting the existence and properties of lower-dimensional ferromagnets which are yet to be experimentally observed.