EDIT 27.03.2024, NOTE: We have already received more applications than allocated space, and have increased the number of total participants. It has been decided that for the remaining few spots the applications will remain open for a short period, after which a selection process will be done.

EDIT2 03.04.2024, NOTE: We will close the application on 12.04.24.

The long-time evolution of nanoscale and multiscale systems, e.g., heterogeneous catalysis [1,2], surfaces [3] and metallic alloys [4], typically involves morphological changes of the atomic structure, and chemical reactions. The method of choice to tackle such evolution on the mesoscopic time scale is the kinetic Monte Carlo (kMC) under multiple variations. In kMC, the evolution of a system is described as a series of jumps between metastable states. These states represent minima on a potential energy surface (PES), and the jumps are rare, independent events. Such evolution forms a Markov process between the metastable states, with an appropriate estimation of the rates of transition from one to the other. The Transition State Theory (TST) [5,6] provides means of computing the transition rates among known configurations (minima and saddle points) of the PES.

For a physically meaningful atomistic kMC simulation, it is crucial to identify all relevant minima and transition states on the PES [7,8], and compose a catalog of possible events for any given configurations encountered during the evolution. At a more mesoscopic level, evolution can focus on objects, moving on longer time and length scales. Through its numerous variations, the kMC framework offers flexibility both in terms of execution and in the physical modeling of the system. For instance, the event catalog can be composed in advance, or on-the-fly [9,10]; the atoms can be positioned on a discrete lattice or in a continuous space. These choices, however, have consequences on practically all aspects of the simulation.

In this advanced school, we will introduce the fundamentals of the theoretical background, and elucidate the working principles of the long time-scale simulations, focusing on kMC simulations. The school will consist of a series of theoretical lectures, and hands-on practices. Theoretical topics will cover the basics of TST and beyond [5,6], and explorations of PES surfaces [7,8]. The hands-on practical sessions will include extracting reaction pathways from PES in classical potentials or DFT, using approaches such as the ci-NEB [11], ARTn [12]; constructing kMC simulations at two important levels of approximation (on-lattice with kmos3 [13] and off-lattice with EON [9] and kART [10]), and analysis to obtain relevant thermodynamic properties from outputs, such as diffusion coefficients. The atomic-scale data can be transformed into macroscopic transport coefficients, for which the analytical techniques, and the associated KineCluE [14] software will be introduced.

During the CECAM workshop “SSbench: Saddle point Search algorithms…” held in Toulouse, France, 27-29.06.23, it has become evident that the community is aware of the importance and sustainability of code/package development, which is often done by graduate students or postdocs. This workshop will thus devote a section to educate, and discuss good-practices in scientific software management. The goal is to induce a higher level of software produced by young scientists in the long run.

Throughout the five-day school, the attendees will develop a comprehensive understanding of the long-time scale simulations, from searching for transition states, to running kMC simulations. We also provide hands-on exercises for practical experience with edge-cutting simulation tools. The coverage of basic theory keeps the school inclusive and welcoming to all levels of researchers interested to work on long time scale phenomena. This school will also initiate collaboration chances between researchers from various research fields.