Tunable laser-generated GHz surface acoustic waves during magnetostructural phase transition in FeRh thin films

TL;DR

This study demonstrates tunable, laser-generated GHz surface acoustic waves in FeRh thin films by exploiting a first-order photoinduced AFM→FM phase transition that accompanies lattice expansion. The observed SAWs (center frequency ~3.1 GHz) are predominantly produced by lattice transformation during PIPT, with amplitude controlled by temperature relative to the transition and pump fluence; a thermodynamic FeRh-based model links the generation to a ~95 ps lattice change, while non-equilibrium kinetics play a minor role. Interferometric and photoelastic measurements confirm a stronger SAW generation in the AFM state than in FM, enabling phase-dependent on-chip acoustic control. The findings pave the way for optically activated, phase-switchable SAW emitters in magnonic and neuromorphic devices, where FeRh can serve as a tunable, noncontact transducer.

Abstract

Laser-generated surface acoustic waves (SAW) facilitate an efficient information processing in modern spintronics and magnonics. The ability to tune SAW parameters is crucial to achieve an acoustic control over magnonic properties. Such tunability can be achieved in phase-changing magnetic materials accommodating both spin waves and SAWs. A promising material is FeRh alloy, a metallic antiferromagnet at room temperature undergoing a phase transition into ferromagnetic state accompanied by a crystal lattice expansion at 370 K. This transition can also be induced by femtosecond laser pulses. In this paper we use the phase transition in 60 nm Fe49Rh51 film to optically generate pulses of Gigahertz quasi-Rayleigh SAWs. We detect them via photoelastic effect and show that the lattice transformation during phase transition is a dominant strain-generation mechanism for above-threshold excitation. The weight of this contribution rises as the sample is heated closer to AFM-FM transition temperature and 'switches off' when heated above it allowing to control the SAW amplitude. A model based on thermodynamical parameters of Fe49Rh51 shows that the lattice transformation occurring within 95 ps effectively contributes to SAW generation happening at a comparable timescale, while non-equilibrium fast kinetics of the phase transition does not.

Figures (4)

• Figure 1: Scheme of experiment and sample characterization. (a) Scheme of sample and experiment; (b) Thermal hysteresis of magnetization for external in-plane field $\mu_0H$=100 mT; (c) Sketch of photoelastic SAW detection. WP is a Wollastone prism, $\lambda$/4 is a quarter-wave plate; (d) Transient reflectivity change $\Delta R(t)$ for pump fluences 0.5 mJ/cm$^2$ (black) and 3.4 mJ/cm$^2$ (red) at 295 K. Dotted black curve is the scaled version of solid black one. Shaded area is a phase transition contribution to reflectivity $\Delta R_{PT}$. Inset: fluence dependence of maximal $\Delta R_{PT}$ value used to find threshold and saturation fluences at 295 (blue squares) and 330 K (green circles).
• Figure 2: SAW pulses in FeRh/MgO (001) detected by photoelastic effect below and above phase transition temperature. (a)-(c) 2D spatial maps of $\Delta R_s/R$ for $t$=3 ns after excitation for initial sample temperatures (a) $T_0$=295 K$<$$T_{PT}$, (b) $T_0$=330 K$<$$T_{PT}$ and (c) $T_0$=430 K$>$$T_{PT}$. Pump fluence is above saturation: $W$=10 mJ/cm$^2>W_s$. (d)-(f) 1D scans along $x$ direction for the same set of temperatures as (a)-(c). Dots are experimental data, solid lines are S-spline interpolation, dotted lines are envelops used to determine signal amplitudes; (g) Normalized FFTs of the SAW pulses from (d)-(f).
• Figure 3: SAW pulses in FeRh/MgO (001) detected by interferometry below and above AFM-FM transition temperature. (a) Sketch of interferometer Klokov_diamond_2021 used to detect quasi-Rayleigh SAW. $\lambda$/4 are quater-wave plates; (b)-(c) 2D spatial maps of $\Delta \varphi$ for $t$=3 ns after excitation with above-threshold fluence $W$=13 mJ/cm$^2$ for initial sample temperatures (b) $T_0$305 K$<$$T_{PT}$ and (c) $T_0$430 K$>$$T_{PT}$.
• Figure 4: Laser-induced strain generation during PIPT in FeRh. Calculation results (lines) and experimental data (symbols) for initial sample temperatures $T_0$=295 K (blue, squares), 330 K (green, circles) and 430 K (red, triangles). The calculations performed in accordance with Eq. (\ref{['eq:calc']}). The dashed lines show calculation result with $\varepsilon_{PT}=0$;