Combining experimental and computational physics helps solve global problems

A lone scientist who is laboring day and night in the laboratory to find fundamental laws of nature is long gone and in the modern world, it is all about optimizing, collaborating, and embracing new developments and technologies into research. These innovations include instrument development, for example attosecond laser technology, which was the topic of this year’s Nobel prize in physics, but also advances in computer-based modeling, such as machine learning algorithms, artificial intelligence tools, and atomistic modeling of large and complex systems. Often these developments go hand-in-hand: for example, experimental breakthroughs in solid state physics and nanolithography allow production of even smaller computer chips, which in turn enable larger computing power.
XPS machine used for analysing materials
Materials science samples in a vacuum chamber ready to be transferred to X-ray analysis.

One of the most successful computational tools in physics is density functional theory (DFT). Its foundation is in quantum mechanics, which is a theory describing the behavior of the smallest constituents of matter, that is electrons. Essentially, the ground state energy of a material system is calculated from the ground state electron density, from first principles or ab initio, meaning no empirical values are used.
DFT-based methods are especially used to describe the electronic structure of matter, i.e., how electrons reside around nuclei, behave, and interact, with each other and also with ions. Electronic structure and its dynamics govern many of the macroscopic properties of matter from conductivity to color of materials, but also their chemical reactivity.

Materials science is booming: in order for us to meet the needs of increasing global energy consumption while battling against climate change, more sustainable materials and processes are urgently looked for. Currently, we have 118 different elements in the periodic table of elements. Only about 80 of them have stable or sufficiently long-lived isotopes for materials science purposes, however, this also opens up many more combinations than is ever possible to synthesize or test in laboratories. This is where DFT can come to help provide screening for suitable materials which are then synthesized in the laboratory. On the other hand, DFT provides insight for processes. Once we understand chemical reactions at a fundamental level, it allows us to tailor the reactants or e.g., catalysts so that the reaction is more efficient.

In our recent research carried out at Nano and molecular systems research unit, atomistic modeling techniques revealed interesting phenomena, and can guide real time experiments. In an article published in the ACS Omega journal (Asikainen et al., ACS Omega 8, 45056-45064 (2023)), we found that heterojunctions made from two dimensional (2D) semiconductors reveal semiconducting or metallic characteristics at the interface region depending on the atomic termination. This is explained on the difference in charge transfer happening at the interface and the following figure illustrates this. In the figure two heterojunctions formed by semiconducting materials TiO2 and MoSSe are shown and at the interface depending on whether S or Se atoms are present, the charge density patterns are changing indicating their different conducting properties. Using modeling techniques, it is easy to visualize the electron gain and loss occurring across semiconductor heterojunctions and atomistic calculations such as these can serve as a guide to experiments during the device fabrication to tune the materials according to specific technological needs.

Graphic illustration showing atomic charge density differences.

Figure: The figure show charge density difference of (a) TiO2/MoSSe and (b) TiO2/MoSeS. The yellow isosurface refers to electron gain, and blue refers to electron loss.

Modeling tools can not only predict new materials and processes, but also explain scientific phenomena experimentalists struggle to reveal. At times, it is also possible to execute ‘virtual’ experiments which take place in extreme conditions or are simply too dangerous or time-consuming to carry out. Computational approaches allow us to explain what happens in the core of a nuclear explosion, follow the time-evolution of our universe, and calculate the pressure and temperature in earth’s core. Thus, it is right to say that computational physics is extremely important to lead us to where no man has gone before – and to warn us about where not to go.


Profile photo
Docent/Adjunct Professor
Materials and Mechanical Engineering
University of Oulu

S. Assa Aravindh is a computational physicist with expertise in condensed matter theory. She uses first-principles simulations to investigate physico-chemical properties of materials, for example steel, battery materials and catalysts.

Associate Professor (Tenure Track)
Nano and Molecular Systems Research Unit
University of Oulu

Minna Patanen is an experimental physicist using X-ray-based techniques to investigate physico-chemical properties of materials.