Taming Complexity Together
Mission Statement - Who, What, and Why
Who are we?
Our team consists of experimental and theoretical physicists, chemists, and chemical engineers from two universities in Vienna. Our group includes recognized leaders in surface science, nano-catalysis, and computational modeling.
What are we doing?
We tightly interweave experiment and theory: we apply multi-technique, multi-environment experiments with a pathway from well-defined single crystals to epitaxial thin films to powders, from UHV to high-pressure/electrochemical conditions to liquid phase. We use selected spectroscopic techniques as handshake methods to bridge different environments. In computations, we use machine-learned potentials to accelerate first principles calculations by many orders of magnitude. Machine Learning (ML) is also employed to predict experimental observables from structures, and to determine simple descriptors that relate to complex experimental properties (i.e. selectivity).
Why are we doing that?
Multicomponent materials and their surfaces and interfaces are essential for converting chemicals and for storing energy. Despite many decades of research, we are still far from a microscopic understanding; unraveling the many interwoven dependencies is simply overwhelming. ML is currently revolutionizing the way we deal with such complexity and with large amounts of data.
Our main goal is to understand – to “tame” – the complexity of multi-cation oxides through ML. Oxides are earth abundant and belong to the most-promising materials in photo-, electro-, and heterogeneous catalysis. Their physical and chemical properties are easily altered by doping, by exposing different surface facets, or by changing the environment. Our objective is to develop microscopic models for their behavior in the bulk and at the surface – in ultra-high vacuum, under gas pressure, and in contact with aqueous solutions. In the long term, we will develop descriptors that are easy to measure or calculate, and correlate with durability, selectivity, and catalytic turnover rates. Ultimately this should allow optimizing materials for technical applications.