AEPSWS: Electrical Spin-Wave Spectroscopy

• AEPSWS is an electrical transduction method that excites and detects spin waves in nanostructured magnetic systems using lithographically defined microwave antennas.
• It employs vector network analyzer-based S-parameter measurements and signal-processing routines to extract dispersion relations, group velocities, and attenuation lengths.
• The technique enables dynamic control through current-induced spin Hall effects and voltage-tunable magnetoelastic coupling, informing advances in magnonic circuits and nonreciprocal devices.

All-electrical propagating spin-wave spectroscopy (AEPSWS) is an experimental technique enabling quantitative, phase-resolved measurement of spin-wave propagation in nanostructured magnetic systems using purely electrical means. AEPSWS combines lithographically-defined microwave antennas for localized spin-wave excitation and detection, vector network analyzer (VNA)-based S-parameter measurements, and signal-processing routines for the extraction of propagation parameters including dispersion relations, group velocities, attenuation lengths, non-reciprocity, and interaction with external control fields, all without optical access.

1. Principles, Device Architectures, and Materials

AEPSWS is fundamentally based on electrical transduction: microwaves delivered to inductive antennas on a magnetic film generate localized oscillating Oersted fields that excite spin waves of defined wavevectors. The same or similar antennas, situated at a controlled distance, transduce the time-varying stray fields from propagating spin waves into a measurable electrical signal. This inductive scheme allows for on-chip, broadband, and phase-sensitive probing of spin-wave dynamics in various materials and geometries.

Typical device structures include:

• Bilayer metallic waveguides: e.g., Ni$_{80}$Fe$_{20}$ (permalloy, Py, $t_{Py}=15$ nm) with Pt ($t_{Pt}=10$ nm) overlayers patterned into strips ($W=10\,\mu$m, $L\sim 20\,\mu$m) (Gladii et al., 2016).
• Nanoscale magnetoelectric heterostructures: Pb(Zr$_x$Ti$_{1-x}$)O$_3$/CoFeB composite waveguides ($w=700$ nm) for voltage-driven actuation (Narducci et al., 2023).
• YIG nanowaveguides: 20 nm thick, $2.5\,\mu$m wide, 50-parallel-channel films for coherent propagation studies (Collet et al., 2016).
• Sub-micron spin-wave logic circuits: inline majority gates and wavevector-diverse devices with CoFeB or permalloy at widths down to 850 nm (Talmelli et al., 2019).

The antenna geometry and associated metal stack (e.g., Au, Ti/Au, CPW or U-shaped designs) determine the excited wavevector spectrum, usually peaking at $k_{peak}\sim\pi/w$ for antenna width $w$. Materials with low Gilbert damping (α down to $10^{-4}$ for epitaxial YIG; $10^{-2}$ typical for Py, CoFeB) and engineered magnetic anisotropies (e.g., via Ga-doping in YIG (Carmiggelt et al., 2021)) are preferred for long propagation distances and high spectral resolution.

2. Electrical Excitation, Detection, and Spectroscopy Methodology

In AEPSWS, microwave excitation and detection channels are provided by a VNA, using coplanar waveguide probes or wire-bonded RF lines to contact on-chip antennas. The process is as follows:

1. Excitation: Antenna 1 receives an RF signal, generating an Oersted field $h_\mathrm{rf}$ localized under the antenna, with field symmetry determined by the antenna design. This field excites coherent spin waves in the underlying magnetic conduit.
2. Spin-wave propagation: Spin waves propagate along the magnetic waveguide, acquiring amplitude decay and phase shift proportional to their group velocity $v_g$ and attenuation length $L_{att}$.
3. Detection: Antenna 2 senses the spin-wave-induced dynamic magnetic flux, and the resulting voltage is recorded as $S_{21}(f)$ (or $\Delta L_{21}(f)$ after appropriate calibration).
4. Phase-sensitive spectroscopy: Phase oscillations in $S_{21}(f)$ correspond to spin-wave propagation, with distinct oscillatory periods and envelope decays that are analyzed to extract $k(f)$, $v_g$, and $L_{att}$.

AEPSWS is compatible with a wide frequency range (typically 4–20 GHz, dictated by materials and geometry), and supports multi-mode detection as evidenced by oscillatory features in both amplitude and phase data.

3. Signal Processing and Quantitative Data Analysis

The extraction of spin-wave parameters from AEPSWS data is grounded in signal-processing protocols and analytic/numerical modeling:

• Wavevector determination: The spatial Fourier transform of the antenna field profile, e.g., $h_x(k_x)\propto \mathrm{sinc}(k_xL_{eff}/2)$, selects the dominant excited $k_x$. The frequency dependence of $S_{21}(f)$ reveals interference fringes whose spacing ($\Delta f$) yields $k$ via $v_g=\Delta f \cdot D$.
• Attenuation measurement: The amplitude envelope $A(D)$ decays as $A(D)\propto \exp[-(D+D_\mathrm{eff})/L_{att}]$. Log-linear fits of amplitude versus antenna spacing $D$ extract $L_{att}$.
• Dispersion and group velocity: The spin-wave dispersion, typically modeled using the Kalinikos–Slavin formalism, is fitted to extracted $k(f)$ data. Group velocity is calculated as $v_g=\partial\omega/\partial k$ either analytically or from finite differences.
• Nonreciprocity and directionality: Measuring $S_{21}$ and $S_{12}$ resolves propagation direction and quantifies frequency or amplitude nonreciprocity, enabling determination of Dzyaloshinskii–Moriya interaction (DMI) constants or spin-transfer torque effects (Gladii et al., 2016Lucassen et al., 2019).

Advanced protocols address cross-talk and background subtraction (field-derivative processing, time-of-flight gating (Thiancourt et al., 2024)), as well as the careful inclusion of correction terms for antenna width and decay length in the phase analysis of monotonic dispersion branches.

4. Functional Extensions: Active Control, Electric-Field and Spin-Orbit Effects

AEPSWS, with purely electrical control, offers unique access to dynamic manipulation of spin-wave propagation via additional external stimuli and internal interactions:

• Spin Hall effect-driven amplification/attenuation: Application of a DC current in a heavy metal overlayer (e.g., Pt in Py/Pt bilayers) generates a spin current that injects a Slonczewski spin-transfer torque in the waveguide. This modulates the spin-wave relaxation rate $\Gamma$ linearly in current density ($\Delta\Gamma\propto \theta_{SHE}J_c$), achieving up to 14% change in $L_{att}$ for $2.3\times10^{11}$ A/m$^2$ in Pt (Gladii et al., 2016).
• Voltage-tunable magnetoelastic coupling: In Pb(Zr,Ti)O$_3$/CoFeB systems, an applied bias alters the effective field $H_{me}$ through magnetostrictive interaction, tuning the resonance frequency by up to 300 MHz, and providing a coupling coefficient $\alpha_{me}$ up to 1.69 mT/V (Narducci et al., 2023).
• Chiral and nonreciprocal effects: Through stacking order and interface engineering (Pt/Co/Ir vs. Ir/Co/Pt), AEPSWS quantitatively separates DMI, surface, and volume anisotropy contributions in nonreciprocal frequency shifts (Lucassen et al., 2019).
• Temperature-dependent regimes: AEPSWS is operable over a wide range of temperatures, including millikelvin, facilitating studies of quantum magnonics and the influence of substrate magnetization on spin-wave propagation (Knauer et al., 2022).

5. Device Engineering, Optimization, and Antenna Design

Device performance and measurement fidelity in AEPSWS depend critically on the geometry and layout of both magnetic conduits and antennas:

• Antenna design: U-shaped or single-k meander antennas suppress parasitic crosstalk and allow selective excitation of desired $k$-bands. CPW and GSG structures enable multi-mode or single-mode operation depending on line/gap periodicity (Talmelli et al., 2019Lucassen et al., 2019).
• Waveguide lithography: Precise control of strip width, edge definition, and thickness is required. Arrays of parallel nanowaveguides (e.g., 50 YIG strips) support coherent multi-channel propagation and robustness to fabrication variations (Collet et al., 2016).
• Antenna spacing and width: The antenna width ($w$) sets $k_{max}\sim2\pi/w$, and the antenna separation ($D$) tunes the phase and amplitude response, dictating the setup’s $k$-resolution and sensitivity to damping/attenuation.
• Parasitic coupling and de-embedding: For wide strips and closely spaced antennas, direct electrical coupling obfuscates the spin-wave response. Sophisticated de-embedding and device miniaturization minimize these effects (Lucassen et al., 2019).

Practical guidelines include minimizing $w$ for high-$k$ access, tuning $D$ for optimal fringe visibility versus signal-to-noise, and using time-gated processing to suppress cross-talk and field-independent backgrounds (Thiancourt et al., 2024).

6. Applications and Impact: Magnonic Circuits, Nonreciprocal Devices, and Future Directions

AEPSWS delivers a fully electrical, quantitative platform for characterization and active manipulation of propagating spin waves at the sub-micron scale, pivotal for advancing magnonics and hybrid spintronic systems:

• Magnonic logic and majority gates: AEPSWS directly validates the operation of inline majority gates using sub-micron CoFeB waveguides, extracting logic states from phase-resolved transmission and supporting frequency-division multiplexing (Talmelli et al., 2019).
• Nonreciprocal and chiral magnonics: Studies of DMI and interfacial engineering leverage AEPSWS to realize unidirectional spin-wave channels and spin-wave diodes via amplitude/frequency nonreciprocity (Wang et al., 2019Takagi et al., 2017Thiancourt et al., 2024Devolder, 2023).
• Quantum and cryogenic magnonics: The technique’s compatibility with ultralow temperatures and integrated devices facilitates the development of magnon-based quantum technologies, magnon–qubit coupling, and on-chip calibration tools (Knauer et al., 2022).
• Voltage and current-controlled components: Demonstrated electrical gating, amplification, and damping of spin-wave signals establish AEPSWS as a foundation for active magnonic circuit elements—analogous to transistor action—essential for future magnonic, neuromorphic, and hybrid computing architectures (Gladii et al., 2016Narducci et al., 2023).
• Extension to new materials and regimes: The method is generalizable to heterogeneous stacks, PMA multilayers, synthetic antiferromagnets, bulk chiral crystals, and systems exhibiting topological or quantum spin-wave phenomena (Sushruth et al., 2020Thiancourt et al., 2024).

AEPSWS, by uniting scalable, CMOS-compatible device engineering with high-precision, broadband, and robust spectroscopic capability, forms a central pillar for both fundamental workflows in magnonic materials research and the rapid prototyping of application-oriented spin-wave devices.