“A plasma is difficult to study,” says Luca Vialetto. “A lot of mechanisms are coupled together such as flow, power and interaction between the particles.” The third-year PhD candidate developed an innovative method to simulate the behaviour of electrons in a low temperature plasma. “With my model, we can study how electrons drive the reaction and how to efficiently improve the dissociation of CO2.”
Fast and accurate
Following an idea of his supervisor, plasma chemistry expert Paola Diomede, Luca modified the Monte Carlo Flux method, a much overlooked method from the nineties. This is a smart combination of Monte Carlo (MC) calculations - accurate but computing-time-consuming - and the electron transport equation which is usually solved numerically by implying many assumptions and approximations.
Luca explains: “First, I do an MC simulation of the motion of the electrons but only for a very short time. Then I use a deterministic method, that means I solve a system of linear equations to obtain the full distribution of the electrons. Since I use no approximations, it has the same accuracy as MC but it is much faster, minutes instead of days.”
The researchers checked their results by comparing them with experimental data from literature and showed that they are in excellent agreement with alternative solutions, obtained under two-term or multi-term approximations.
Then, Luca worked with a colleague who is simulating heavy particles such as molecules. “With our models coupled together we can study what is happening during the experiments. They can measure electron densities, composition, gas temperature, these are all inputs for the model. We compare our model to the experiments and vice versa. The aim is to describe exactly what happens in an experiment.”
Start from scratch
“What I like about my project is that we started from zero. Normally, you would use an existing software package to run your simulations and compare with experiments. I wrote my own code from the start, which allowed me to build up knowledge about the correct rules and assumptions so that the code is computationally efficient and also very accurate. Eventually, we will release the code as an open source package. Others can then use and modify it for their own experiments.”
Luca's work requires great knowledge about plasma physics and great skill in programming. Still, he says that there is an even more difficult skill that he is learning during his time at DIFFER, which is teaching students. “I think it is important to communicate with students that are a few years younger but really smart. We need their ideas. To explain my work in a way that is understandable but precise, this was one of the most difficult challenges. But I enjoy it so much that I am now sure I want to continue in academia after my PhD.”
This interview was published in the DIFFER Annual Report 2020, available online.