Skyrmion Control and Phase Transitions in Cu2OSeO3

Magnetic skyrmions are nanometric and non-trivial spin textures with non-zero topological charge. Their robustness against perturbations and the possibility to control them using external stimuli make them ideal candidates for future spintronic applications. In particular the magnetoelectric skyrmion host Cu2OSeO3 holds a lot of promise for low power devices since skyrmions in this compound can be controlled by electric fields alone. Using Lorentz transmission electron microscopy to perform real space and real time biasing experiments on thin lamellas of Cu2OSeO3 in a geometry that is most suitable for technological applications, we observe reproducible creation and annihilation of skyrmions. For a more quantitative analysis, we develop new feature detection algorithms to reliably extract skyrmion positions even in noisy images. We further produce Due to its low pinning, Cu2OSeO3 allows for the formation of large and well-arranged triangular skyrmion lattices. This makes this compound a perfect testbed to study the evolution of skyrmion configurations under external stimuli. Experiments are carried out again using Lorentz transmission electron microscopy on thin lamellas of Cu2OSeO3. We investigate how defects in skyrmion lattices are arranged at grain boundaries and develop algorithms to extract them and to directly visualize their alignment. These defects are at the core of the melting of skyrmion lattices in this system. We show that a controlled magnetic field ramp can induce skyrmion ensembles in Cu2OSeO3 to transition from a two-dimensional solid through a thus far unknown ordered liquid phase called the hexatic phase, to a liquid. We find that this transition is a topological defect-induced two-step process as predicted by the Kosterlitz-Thouless-Halperin-Nelson-Young (KTHNY) theory. Finally, we go beyond equilibrium phenomena to explore the effect of quenching the system from its liquid phase to its solid phase using different quench rates and find first evidence that our system belongs to the Kibble-Zurek universality class.