Atomistic material modelling

I personally find materialistic modelling promising. Not only because it is publishable but also its a field where physics is applied for practical applications. It seems weird that Malaysians do not do much atomistic material
modelling. This is a field most easily accessed, cheap, and can supply much applications. It's the best bet for developing country like us. Thailand, Vietnam all now jump into the band wagon madly.

Doing a bit of literature review on materialistic modelling you will find that there are definitely a huge avenue to pursue. Modelling can be done at many scales and levels, from ab initio up to continuum limit. There are expansive and accurate methods like quantum monte carlo (QMC), quantum molecular dynamics (QMD) - or known as Carr-Parinello MD (CPMD), DFT, quantum chemistry methods such as CI and Hartree-Fock. Then there are semiempirical methods such as tight binding methods. Then there is classical molecular dynamics method. Almost any phenomena process related to material can be modelled and calculated (accuracy and correctness are another issue). You just got to know the methods which is not so difficult (merely tedious).


I have gradually formed a small group of students with other research collaborators. We are slowly picking up the know-how in MD using a software called LAMMPS. Meanwhile, we are also venturing into other software to model different types of systems as well, e.g., biomolecule MD. When another master student join in (if he ever turn up) early next year, I will get him to do DFT calculations for ferroelectrics. 


Our current plan is to understand the step-by-step procedure to obtain force fields for a new material using existing software packages. A lot of very technical details involved e.g., translating one type of file format of a software package into another, fine-tuning the input parameters etc. But once we have gone through that, I expect to acquire the necessary know-how to calculate the physical properties of any arbitrary crystal (meaning any suitable material with experimentally measured data). It's kind of 'cheap' in a sense, as these are work done mostly by material scientists instead of by physicists. 


At the moment we have already acquired some very minor achievement where we have used LAMMPS to reproduce some mechanical and thermodynamics properties of a well studied material SrTiO3. My master student improved the force field and reproduced a results that suit experiments better. I told him that now our agenda is to master the knowledge to generate force fields for materials which are not yet available in the literature. I reckon that obtaining the force field for a material not been published itself is a good achievement that could have real impact to the real world. Another beginning student is modelling the growth of graphene on a silicon carbide substrate by properly annealing it, also using LAMMPS. This is a project in collaboration with Prof Lai, a Malaysian who is a very senior professor in national central university in Taiwan. He have provided much helpful and unreserved advice to us. So grateful to Prof. Lai. 


There is a lot of room in computational modelling of materials, and I am half-way in exploring it. Hopes remain to go deeper into it.