The first two years of the project focused on the development of advanced computational methodologies for materials discovery under extreme conditions and their application to nitride, carbo‑nitride, and hydride systems. During the first phase, substantial effort was devoted to creating scalable symmetry‑aware structure generation and screening methods, enabling the systematic exploration of previously inaccessible regions of chemico-structural space. A highly optimized workflow was established that allows the symmetry‑restricted exploration of crystal structures at unprecedented scale, allowing to investigate billions of candidate structures per day on modern GPU architectures.
In parallel, new scientific insights were obtained through close collaboration of theory with experimental synthesis. Several nitride and carbo‑nitride systems were investigated, revealing recoverable high‑pressure phases with complex electronic behaviour and even layered motifs.
To investigate hydrogen‑rich systems, a finite‑temperature finite-pressure structure relaxation workflow was developed, allowing anharmonic and thermal effects to be incorporated efficiently under symmetry constraints. This enables reliable structural refinement at finite-temperature —including quantum effects (no tunnelling)— at significantly lower computational cost than fully quantum approaches. In addition, a new benchmark methodology for machine‑learned interatomic potentials was established based on symmetry‑restricted energy landscapes, addressing a critical gap in evaluating these models for unknown crystal prototypes. The project also identified the need for advanced analysis tools, leading to the development of an open‑source, web‑based platform for interactive visualization and comparison of crystal structures.
