Growing up, I wondered how a field as abstract as mathematics could be so useful in our daily lives in solving so many problems in our society. On the other hand, I remember learning the laws of motion in high school and not understanding how important they are on a daily basis. The examples used to represent the importance of these fields were always short in my head. Nobody would care about the exact time a ball takes to drop if it didn’t have any practical application. The answers I found back then to those questions were always short compared to my curiosity. I wanted to know more, so when I finished high school I enrolled in Physics Engineering at the University of Aveiro.
The first sense of choice in my undergraduate studies was definitely my bachelor’s project choice, and that decision could not be easier. Among photonics and optoelectronics, advanced materials, instrumentation, and theoretical and computational physics seemed to be the areas that suited me the most by far. I have always had an interest in challenging my capabilities. My project, submitted as “Numerical Study of Heat Conduction in a Bar (1D),” allowed me to combine my analytical and numerical skills. The heat propagation equation was deduced and implemented in MatLab using implicit/explicit Euler and Crank-Nicolson methods for several boundary conditions and materials.
In my master’s, I knew I would like to continue into the branch of theoretical and computational physics. Density functional theory (DFT) tools, such as the Vienna Ab initio Simulation Package (VASP) and the Munich SPR-KKR code, were employed to calculate microscopic properties relevant to the study of magnetic materials. The ultimate goal was to perform simulations of magnetic systems based on transition metals such as simple cubic, body-centered cubic, face-centered cubic, and possibly model Heusler compounds. With the estimated exchange energies, Monte Carlo methods would allow the estimation of critical thermodynamic properties. However, as I was not familiar with the UNIX environment or density functional theory, it took me a while to get used to it. Moving forward from density functional theory to Monte Carlo calculations in MatLab was no longer an option. Consequently, I had to readjust the initial work plan. This setback gave me an excellent opportunity not only to explore DFT but also VASP deeper. I had time to perform convergence studies (cutoff energies and k-points) and test different functionals, pseudopotentials, crystal structures, and magnetic properties of Fe, Co, and Ni. As experimental properties are known, these kinds of calculations help get things started. My supervisors always encouraged me to explore the theory and packages available. Beyond computational resources, I had access to several scripts in Python and C. However, I wrote my scripts in MatLab even when libraries and tools were available, such as Atomic Simulation Environment (ASE) to determine the Birch-Murnaghan equation coefficients. Despite external calculations written in MatLab, the research group recommended shell scripting to automate a set of computations such as structural optimization no matter the tool employed. The evolution of the Curie temperature of the Fe\(_2\)MnSi\(_{1-x}\)Ge\(_x\) system was studied computationally after the advice was accepted and followed. Computational tools were employed to gain insight into the Curie Temperature (T\(_{C}\)) versus Ge content (x) behaviour, separating the structural contribution (change in lattice parameters) from the chemical contribution (solid solution composition). The computational results were then compared to recent experimental work from the group of Prof. Mario Reis from the Institute of Physics, Fluminense Federal University, recently published in Journal of Alloys and Compounds 893 (2021) 162236.
During the time I was working on my master’s dissertation, I had the opportunity to apply for a research grant for bachelor holders, within the scope of the FCT project entitled “PORTUGAL at ISOLDE: Research in Materials Physics and Nuclear with Isotopes and Radioactive Techniques” (CERN/FIS-PAR/0005/2017). The project allowed me to expand my knowledge of the density of states and band structures. The calculations were designed to target hexagonal and orthorhombic LuFeO\(_3\) experimental structures, as well as Lu\(_3\)Fe\(_5\)O\(_{12}\) experimental compounds. Besides calculations, the physics of hyperfine interactions and the experimental methods that make use of hyperfine interactions were studied in the widely distributed online course by professor Stefaan Cottenier. This project allowed me to explore a high-throughput density functional theory branch, exploring the values of the Hubbard and Hund parameters and verifying their agreement with experimental band gaps.
Computational simulations and multiscale modelling of ceramics and multifunctional ferroic nanostructures allowed me not only to strengthen my physical domain but also to expand my knowledge of computational resources and their limits. The time consumption optimization of those calculations with 8 or 16 CPU cores or Random Access Memory overhead started to trigger me into developing more curiosity, which I had not experienced when I was studying electrical circuits, electronics, electronic instrumentation for physics, optoelectronics, previously in my course. Although I previously had to use the Arduino IDE to blink LEDs or build an oximeter in optoelectronics, it never got me so excited about embedded systems. However, the development of my problem-solving personality into learning shell scripting, python, and a bit of C and C++, constantly challenging my boundaries and building bridges between the knowledge acquired in the past, is what drives me the most.
After concluding my studies, my interest in technologies such as hydrogen storage solutions, CO\(_2\) capture, and batteries increased a lot. Definitely, my academic years gave me another perspective on those technologies, and I am proud of that.