1.    SCIENTIFIC PROPOSAL

1.1 Background and state of the art

Coupling the waves of an electromagnetic signal—a wifi signal—with strain waves in a micrometric device is what Surface Acoustic Waves (SAW) filters do, based on the piezoelectric effect. This coupling has enabled a remarkable miniaturization of wireless communication technologies because it allows for the conversion of long wavelengths, 15 centimeters for a wifi wave, into much shorter wavelengths, few micrometers for acoustic waves (all working at the GHz range).

The interaction between “strain” and magnetization—magnetoelasticity—is receiving an increasing interest because it offers an alternative control of magnetization at the nanoscale that avoids electrical currents. An example is the efficient coupling through strain between ferroelectric and magnetic orders in multiferroics [1]. Strain is used in spintronics as an additional degree of freedom to adjust magnetic properties. The field of ‘Straintronics’ is emerging [2- JPCondMatter2019] and strain-control of functionalities might not be limited to magnetic properties.

Our project, SoundSpin, studies the coupling between surface acoustic waves (SAW) and magnetization dynamics according to the following two lines:

Line 1: Acoustic manipulation of magnetization dynamics

Magnons are the fundamental excitations (quasiparticles) of ordered magnetic systems (ferro-, ferri- and antiferromagnets), which in their collective, synchronized form are referred to as spin waves. Recent interest for spin waves is motivated by the possibility of their integration in nanoscale devices for high-speed and low-power signal processing [3,4]. However, generation of spin waves with high amplitudes—and their detection—is challenging due to the mismatch of wavelengths with electromagnetic waves in free space, which are of the order of several centimeters whereas the spin-wave wavelengths are of the order of hundreds of nanometers.

All magnetic materials might display spin waves with a frequency and wavelength determined by the strength of magnetic interactions—mostly dipolar and exchange interactions. These collective magnetic excitations are traditionally excited via magnetic fields generated by antennas or striplines with an electrical current [5]. The spin-transfer-torque effect [6,7], generated from spin currents or ultrashort laser pulses [8,9], provide promising alternative pathways towards the control of dynamic magnetic states at the nanoscale without using magnetic fields. In metallic ferromagnets the damping of spin waves results in low amplitude and/or strongly localized excitations, limiting the applicability in devices. A promising strategy for handling magnetization variation at the nanoscale together with low-power dissipation is the use of strain. The magnetoelastic (ME) effect is the change of material’s magnetic properties under a mechanical deformation—strain. A change in atomic distances caused by strain modifies magnetic interactions, resulting in a ME-induced anisotropy. SAW are strain waves that travel at the surface of a material and can be generated with oscillating electric fields in a piezoelectric material [10,11]. Today, radiofrequency filters and delay lines based on SAW are a standard technology used in mobile phones because of their ability to convert a centimeter wave in free space into a micrometer wave in a chip. The working frequencies range from hundreds of MHz to a few GHz with associated wavelength of a few micrometers.

The interaction between SAW and magnetization dynamics is receiving an increasing interest. There are multiple studies reporting changes in SAW propagation caused by the back action of magnetization dynamics [12-15]—we have recently shown how to obtain nonreciprocal SAW propagation through the interaction with magnetization dynamics [15-PRAppl2020]. Conversely, changes in magnetic states caused by SAW have been also reported [16-22]—including our direct observation of moving domain walls in nanostructures [18-NatCom2017, 21-MRS2018] and the generation of large angle spin waves [22-PRL2020]. Additionally, a recent study from Dr. M.V. Costache (a research team member) built a hybrid magnetometer based on the coupling between spin waves in a ferromagnetic insulator and the strain modes of a glass microsphere deposited on top [23-PRL2020].

In line 1 we will explore and exploit the coupling between acoustic waves and magnetic waves in different magnetic materials, focusing on its generation and modulation. Our proposed work consists, on the one hand, in generating large angle spin waves and sizable magnetization dynamics with SAW in different materials including Heusler alloys and 2D-magnets, and, on the other hand, in controlling spin wave propagation with SAW, aiming at the creation of functional magnonic device—a device capable to create frequency band gaps or nonreciprocal spin-wave propagation controlled by SAW.

Line 2: Acoustic generation of spins

The ability of spin-polarized currents to manipulate and interact with magnetic states has strongly impacted the development of computation technologies. This effect is exploited by read-heads since the end of the past century [24] and more recently by spin transfer torque writing in magnetic random-access memories (STT-MRAM) [25]. Today, an active field in spintronics aims at the generation of pure spin currents, which differ from spin-polarized currents in not involving net charge flow (Spin insulatronics, Spin Caloritronics or Spin Cavitronics). Thus, they maintain the capability to alter magnetic states while Joule heat losses are radically suppressed. The latter aspect is crucial for the future development of highly-performant and energy-efficient electronic nano-devices since it would reduce the energy consumption by several orders of magnitude.

Currently, the main strategies to generate pure spin currents are based on i) the spin-pumping effect [26-PRL2006,27,28], ii) the spin Hall effect [29], and iii) the spin Seebeck effect [30,31]. The latter strategy is particularly interesting, since it involves temperature gradients to create spin currents—and heat is an unwelcome byproduct of many operations in electronic devices, which could be utilized to produce useful spin currents. However, the downside is that heat dissipation is a non-directional phenomenon, spreading slowly in all directions.

Our recent studies on SAW-generation of large angle spin waves [23-PRL2020] and SAW-generation of fast and localized heat gradients [32] suggest the possibility of using hybrid magnetic/piezoelectric devices to generate pure spin currents with SAW.

In line 2 we will study the generation of pure spin currents induced by SAW patterns. Our proposed work in this line consists in optimizing SAW generation to obtain, on the one hand, maximal heat gradients (fast and localized) and, on the other hand, large angle magnetization precession at high frequencies. In both cases we might generate spin currents either through the spin Seebeck effect or through the spin pumping effect.

1.2 Initial hypotheses and general objectives

During the last four years and within the framework of the ACME project [33] we, the Grup de Magnetisme de la Universitat de Barcelona (GMUB), working at the Dept. de Física de la Matèria Condensada (FMC), have demonstrated the possibility of modifying magnetic states with SAW [18,21,22] and, conversely, we have produced non-reciprocal acoustic wave propagation in hybrid magnetic/semiconductor systems controlled by the magnetic state [15]. On the one hand, we have participated in the development at ALBA Synchrotron of a technique based on X-ray photoemission electron microscopy (X-PEEM) capable of imaging and quantifying SAW [34-JSynchRad2019]. We have resolved the standing and propagating components of SAW from a superposition of waves, providing a method to create standing or propagating SAW. Using the mentioned set up we have measured i) the evolution of strain and magnetization dynamics in nanostructures at the picosecond time scale [18-NatCom2017, 21-MRS2020] (see Fig. 1A), ii) large-angle spin waves created by SAW over distances up to 6 millimeters [22-PRL2020] (see Fig. 1B), and iii) the effect of SAW patterns on the promotion of catalytic reactions [35-AngewChem2020]. On the other hand, we have explored the modulation of SAW propagation through the interaction with spin waves in hybrid devices of magnetic/semiconductor and obtained a non-reciprocal propagation of SAW [15-PRApp2020].

Figure 1. Schematic plot of the experimental setup we use to obtain simultaneously images of SAW and magnetization. Circularly polarized X-rays illuminate the sample in the form of 20 ps pulses with a repetition rate of 500 MHz. The interdigital transducers, IDTs, receive an AC electric signal of the same frequency, which is phase-locked to the synchrotron repetition rate, generating a piezoelectric SAW that propagates through the LiNbO₃ substrate and interacts with the magnetic material. The piezoelectric voltage at the surface sample is probed with X-PEEM, as well as the magnetization contrast along the X-ray propagation direction arising from XMCD. In panel A we show a sample with nickel squares (2×2 micrometers) where we image their magnetic configuration [18-NatCom2017,21-MRS2019], and in panel B we show an extended nickel film where we image large-angle magnetization waves [22-PRL2020].

Most of our work on the coupling between SAW and magnetic excitations has been done in polycrystalline metallic ferromagnets deposited on the piezoelectric insulator LiNbO₃. All the imaging experiments we have done used a frequency of 500 MHz and its subharmonics because they matched with ALBA synchrotron repetition rate of light bunches and worked well with the synchronization technique we used (higher frequencies were not available). In this project we aim at studying the coupling of SAW with magnetization in other materials, such as Heusler alloys, which have a very low magnetic damping, and 2D-magnets, which have a strong magnetoelectric coupling. Our goal is also to work at higher frequencies (at least until 5 GHz that is where current 5G technologies are)

We have tentatively explored within the ACME project [33] the possibility of modulating the spin-wave propagation by SAW. However, the chosen approach of using Brillouin Light Scattering did not work because we were unable to disentangle phononic and magnetic components in the light signal. One of the goals of the project is to create functional magnonic devices [3,4], where magnetic wave propagation is controlled and adjusted by SAW. We propose two different approaches to study the effect of SAW patterns on spin waves: i) measurements of ferromagnetic resonance (FMR) in the presence of SAW (see Fig. 2A), and ii) measurements of spin wave propagation based on a two-antenna system in the presence of SAW (see Fig. 2B).

Figure 2. Schematic plots of measuring setups for functional magnonic devices consisting of a heterostructure piezoelectric/ferromagnetic. Panel A shows a single coplanar waveguide for FMR measurements in presence of SAW patterns. Panel B shows a similar structure with two antennas to create and detect magnetic excitations in the presence of SAW patterns.

We have also found recently that using SAW we can locally and quickly heat a single nano-element, or a part of it, up to one degree with a pulse of less than one millisecond (see Fig. 3A) [32]. Thus, we believe that engineering the creation of SAW, which is an energy-efficient technology, well established in commercial applications, we could produce a considerable increase of heat gradients along the directions perpendicular to the SAW path. Thus, SAW patterns may be envisioned to create controllable temperature gradients that serve to generate pure spin currents through the spin Seebeck effect (see Fig. 3B).

Additionally, it is possible to generate spin currents through spin pumping, by driving a ferromagnet into a magnetic resonance (see Fig. 3C). Our recent development on the generation of magnetization dynamics with SAW [18-NatCom2020,21-MRS2019,22-PRL2020,32]—and many previous results from other groups [36-38]—suggest that the SAW–generation of spin currents through spin pumping is feasible. We aim at measuring the corresponding spin accumulation in a nearby layer having strong spin-orbit interaction, both electrically through inverse spin Hall effect and by direct imaging using X-PEEM.

Figure 3. Panel A shows a schematic plot of the SAW that propagates along the piezoelectric substrate having a well-defined acoustic path. Measurements of scanning thermal microscopy (SThM) are done at the surface of the piezoelectric, either within the acoustic path (green) or outside the acoustic path (orange). Panels B and C show respectively schematic plots of the spin Seebeck effect and the spin pumping effect.

Bibliography (in bold all references from our group and relevant to the project)

[1] Multiferroics: progress and prospects in thin films, R. Ramesh and N. A. Spaldin, Nature Mater. 6, 21-29 (2007).

[2] Preface to special issue on magneto-elastic effects, M. Foerster and F. Macià, J. Phys.: Condens. Matter 31, 190301 (2019).

[3] Magnonics: spin waves connecting charges, spins and photons, A. V. Chumak and H. Schultheiss, J. Phys. D: Appl. Phys. 50, 300201 (2017).

[4] Move over, spintronics: Here comes magnonics to the rescue of electronics, D. Johnson, News article published in spectrum.IEEE.org (04/17/2017).

[5] Current-induced spin-wave Doppler shift, V. Vlaminck and M. Bailleul, Science 322, 410 (2008).

[6] Spin transfer torques, D. C. Ralph and M. D. Stiles, J. Magn. Magn. Mater. 320, 1190 (2008).

[7] Current-induced torques in magnetic materials, A. Brataas et al., Nature Mater. 11, 372 (2012).

[8] All-optical control of ferromagnetic thin films and nanostructures, C.-H. Lambert et al., Science 345, 1337 (2014).

[9] Generation of spin waves by a train of fs-laser pulses: a novel approach for tuning magnon wavelength, I. V. Savochkin et al., Sci. Rep. 7, 5668 (2017).

[10] B. A. Auld, Acoustic fields and waves in solids, vol. 2 (John Wiley and Sons, NY 1973).

[11] M. F. Lewis, Rayleigh-wave. Theory and application (Springer Berlin, Heidelberg 1985).

[12] Elastically driven ferromagnetic resonance in nickel thin films, M. Weiler et al., Phys. Rev. Lett. 106, 117601 (2011).

[13] Traveling surface spin-wave resonance spectroscopy using surface acoustic waves, P. G. Gowtham et al., J. Appl. Phys. 118, 233910 (2015).

[14] Effect of magnetoelastic film thickness on power absorption in acoustically driven ferromagnetic resonance, D. Labanowski et al., Appl. Phys. Lett. 108, 022905 (2016).

[15] Large Nonreciprocal propagation of surface acoustic waves in epitaxial ferromagnetic / semiconductor hybrid structures, A Hernández-Mínguez et al., Phys. Rev. Appl. 13, 044018 (2020).

[16] Magnetization dynamics triggered by surface acoustic waves, S. Davis et al., Appl. Phys. Lett. 97, 232507 (2010).

[17] Precessional magnetization switching by a surface acoustic wave, L. Thevenard et al., Phys. Rev. B 93, 134430 (2016).

[18] Direct imaging of delayed magneto-dynamic modes induced by surface acoustic waves, M. Foerster et al., Nature Commun. 8, 407 (2017).

[19] Resonant magneto-acoustic switching: influence of Rayleigh wave frequency and wavevector, P. Kuszewski et al., J. Phys.: Cond. Matter 30, 244003 (2018).

[20] Voltage-driven, local, and efficient excitation of nitrogen-vacancy centers in diamond, D. Labanowski et al., Sci. Adv. 4, eaat6574 (2018).

[21] Subnanosecond magnetization dynamics driven by strain waves, M. Foerster et al., MRS Bulletin 43, 854-859 (2018).

[22] Generation and imaging of magnetoacoustic waves over millimeter distances, B. Casals et al., Phys. Rev. Lett. 124, 137202 (2020).

[23] Ferromagnetic resonance assisted optomechanical magnetometer, M.F. Colombano et al., Phys. Rev. Lett. 125, 147201 (2020).

[24] Giant magnetoresistance of (001)Fe/(001)Cr magnetic superlattices, M. Baibich et al., Phys. Rev. Lett. 61, 2472 (1988).

[25] Current-driven excitation of magnetic multilayers, J. C. Slonczewski, J. Magn. Magn. Mater. 159, L1 (1996).

[26] Electrical Detection of Spin Pumping due to the Precessing Magnetization of a Single Ferromagnet, MV Costache et al., Phys. Rev. Lett. 97, 216603 (2006)

[27] Spin pumping and magnetization dynamics in metallic multilayers, Y. Tserkovnyak et al., Phys. Rev. B 66, 224403 (2002).

[28] Quantifying spin Hall angles from spin pumping: experiments and theory, O. Mosendz et al., Phys. Rev. Lett 104, 046601 (2010).

[29] Perpendicular switching of a single ferromagnetic layer induced by in-plane current injection, I. M. Miron et al., Nature 476, 189 (2011).

[30] Observation of the spin Seebeck effect, K. Uchida et al., Nature 455, 778 (2008).

[31] Observation of longitudinal spin-Seebeck effect in magnetic insulators, K. Uchida et al., Appl. Phys. Lett. 97, 172505 (2010).

[32] Results obtained by the group and not published yet (see Fig. 3A).

[33] ACME project (MINECO) MAT2015-69144-P.

[34] Quantification of propagating and standing surface acoustic waves by stroboscopic X-ray photoemission electron microscopy, M. Foerster et al., J. Synchrotron Rad. 26, 184-193 (2019).

[35] On the promotion of catalytic reactions by surface acoustic waves, B. von Boehn et al., Angew. Chem. Int. Ed. 59, 20224–20229 (2020).

[36] Spin pumping with coherent elastic waves, M. Weiler et al., Phys. Rev. Lett. 108, 176601 (2012).

[37] Acoustic spin pumping: Direct generation of spin currents from sound waves in Pt/Y3Fe5O12 hybrid structures, K. Uchida et al., J. Appl. Phys. 111, 053903 (2012).

[38] Acoustic ferromagnetic resonance and spin pumping induced by surface acoustic waves, J. Puebla et al., J. Phys. D: Appl. Phys. 53, 264002 (2020).