Tuning the speed of hydrogen in amorphous materials

Dnr:

NAISS 2023/5-239

Type:

NAISS Medium Compute

Principal Investigator:

Gunnar Pálsson

Affiliation:

Uppsala universitet

Start Date:

2023-06-01

End Date:

2024-06-01

Primary Classification:

10304: Condensed Matter Physics

The goal of this project is to increase our understanding of mechanisms underpinning thermodynamic and kinetic properties of hydrogen in amorphous metals to create hydrogen super highways and robust roadblocks. This will be implemented through a theoretical screening method to efficiently identify promising material combinations of elements and an experimental optical technique to measure hydrogen transport.
Amorphous materials can be viewed as representing a new bronze-age of materials science since; By overcoming the restrictions imposed by crystal symmetry we can nowadays place atoms in three-dimensional space in ways that have never been achieved in nature. We attain the ability to create materials with new properties previously not possible, such as increased hardening, resistance to failure, increased hydrogen uptake, to name a few.
Amorphous metal hydrides exhibit a distribution of energies, which can be tailored by composition, local strain, and finite size. Both thermodynamic and kinetic properties are influenced by the hydrogen-hydrogen interaction energy, which can be determined from local and global volume changes in the material. The ultimate diffusion constant is determined by the electronic structure, which depends on the arrangement of atoms. For crystalline materials, the structure imposes lower symmetries on the electronic structure and restricts the motion of hydrogen to well defined directions. With amorphous materials this restriction is lifted, and the diffusive behavior approaches the ultimate limit.
We will use Density Functional Theory (DFT) and kinetic Monte Carlo to computationally discover suitable amorphous alloying compounds that exhibit either very fast or very slow hydrogen diffusion. The starting point will be binary combinations based on effective medium theory, using the electron density of the crystalline materials as a guide to the selection process. For instance, zirconium has a lower electron density than palladium, and thus hydrogen will reside closer to the zirconium than the palladium atoms in the amorphous structure. The second step is to create candidate ground-state structures using the stochastic quenching method. Once this method produces a candidate, transition state theory within density functional theory is used to calculate barrier heights. These are then put into a kinetic Monte Carlo (KMC) code that we have developed recently, whereby we simulate an amorphous material, by drawing site energies and barrier heights from distributions. The distributions themselves will come from the transition state calculations. The KMC is then used to compute the mean square displacement and the diffusion constant. The advantage of this method is that KMC is orders of magnitude faster compared to DFT and can also be used as a testbed for screening what kind of distributions and coordination yields desirable diffusion constants. We will innovate the theoretical method by also calculating the hydrogen-hydrogen interaction by evaluating the hydrogen-induced local force field, using DFT, and adding this to the Monte Carlo simulations in the mean field approximation.