Chandan Kumar, S. Mathur, Sourav Mandal, Anjan Barman
Three-dimensional (3D) nanoarchitectures are emerging as key platforms for high-density data storage, high-speed information processing, and charge-free logic operations in spintronic and magnonic technologies. Within this landscape, 3D artificial spin ice (3D-ASI) systems are particularly important because their highly frustrated nanomagnet arrangements give rise to complex magnetic charge states that cannot be achieved in planar structures. Their intricate 3D interactions, curvature- and shape-induced effects, unconventional switching pathways, and emergent phenomena make them highly promising for neuromorphic computing, reconfigurable magnonics, and other next-generation applications. In this work, we use micromagnetic simulations to explore the direction-dependent magnetisation switching in a 3D network of magnetic nanowires forming an icosahedral nanoarchitecture. Our results reveal a previously unexplored 3D-ASI behaviour in this geometry, where the magnetisation reversal proceeds through a devil’s staircase-like sequence of discrete plateaus. The detailed examination of the magnetic microstate at each plateau of the hysteresis curve reveals a unified mechanism governing the stabilisation and transport of high-magnetic-charge vertices (Q = ±3Q or ±5Q), which are energetically expensive. These findings position the icosahedral 3D-ASI as a promising platform for neuromorphic computing and sustainable reconfigurable magnonic technologies.