Over the last several years, magnon-based computing has been intensely investigated as a promising candidate to complement and surpass CMOS-based technologies [1]. Many of the basic logic devices, such as a magnon transistor [2], various types of gates [3] or adders [4] have already been demonstrated experimentally. They make use of spin waves, or their quasiparticles – magnons, as energy-efficient information carriers, which can perform Boolean or even non-Boolean operations. So far, no complex magnonic circuit was experimentally realized, because the steering of spin waves, which is essential for two-dimensional chip design, is a major challenge [5]. Also, one of the elementary premises of complex magnonic networks is a need for operation in the absence of an external magnetic field. If an external magnetic field is used to stabilize magnetization in the plane of a thin magnetic waveguide, even a basic circuit element such as a spin-wave turn shows a large dispersion mismatch for regions before and after the turn, as magnons in in-plane magnetized media have anisotropic dispersion, and thus exhibits huge energy losses.We tackled these challenges by locally-controlled effective field, which stabilizes the magnetization of different parts of the magnonic circuit in the desired direction. This enables to match both, spin-wave group velocity and frequency and prevents dispersion mismatch, energy dissipation, and reflections from boundaries [6]. In the presented experiments we use corrugated magnetic thin films and waveguides to locally control the effective field. A sinusoidally modulated substrate on which magnetic film is deposited is prepared by focused-electron-beam-induced deposition. The surface curvature in the films locally modifies the contributions of dipolar and exchange energies and can be described as an effective uniaxial anisotropy term that is perpendicular to the corrugation direction [7]. The direction of the anisotropy axis can be spatially controlled, and arbitrary magnetization landscapes can be created on demand. Using this approach, one can design non-trivial magnonic circuits functioning without the need for an external magnetic field and pave the way towards complex magnonic circuits and all-magnon data processing.[1] A. V. Chumak et al., IEEE Trans. Magn. 58, 6 (2022)
[2] A. V. Chumak et al., Nat. Commun. 5, 4700 (2014)
[3] T. Schneider et al., Appl. Phys. Lett. 92, 022505 (2008)
[4] U. Garlando et al., IEEE Trans. Emerg. Top. Comput. (2023) (Early Access)
[5] K. Vogt et al., Appl. Phys. Lett. 101, 042410 (2012)
[6] L. Flajšman et al., Phys. Rev. B 101, 014436 (2020)
[7] I. Turčan et al., Appl. Phys. Lett. 118, 092405 (2021)
