betaFeedback

WorksPhysics and Astronomy

Quantitative evaluation method for magnetoelastic coupling between surface acoustic waves and spin waves using electrical and optical measurements

Abstract

Coupling and hybridization of different elementary excitations leads to new functionalities. In phononics and spintronics, magnetoelastic coupling between Rayleigh-type surface acoustic wave (SAW) and spin wave (SW) has recently attracted much attention. Quantitatively evaluating and comparing the coupled system are essential to develop the study of the magnetoelastic SAW-SW coupling. So far, previous studies of SAW-SW coupling have employed a quantity called coupling strength. However, it is still challenging to compare the coupling strength values among studies fairly because the quantity depends on the device geometry and the applied magnetic field angle, which are not unified among the previous studies. Here, we focus on a practical constant composed of a magnetoelastic constant and a strain amplitude that depends only on the material properties. We demonstrate a versatile evaluation technique to evaluate the practical constant by combining electrical measurements and optical imaging. An essential part of the technique is an analysis that can be used under off-resonance conditions where SAW and SW resonance frequencies do not match. Existing analysis can only handle the case under on-resonance conditions. Our analysis makes it possible to observe the magnetoelastic couplings between SAW with resonance frequencies that can be imaged optically and SW with resonance frequencies in the gigahertz range. Our demonstrated technique, which uses electrical and optical measurements under off-resonance conditions, can significantly advance research on SAW-SW coupled systems.

Introduction

Different elementary excitations localized at the same location may couple. Rayleigh-type surface acoustic wave (SAW) and spin wave (SW) are known to couple via magnetoelastic interactions in magnetic thin films on substrates [1, 2, 3, 4, 5, 6, 7, 8]. This coupled system has attracted considerable attention due to its potential as an information medium and functionality [9, 10, 11, 12, 13, 14, 15, 16].

It is essential to quantitatively evaluate and compare each coupled system to advance research on magnetoelastic SAW-SW coupling. Previous studies used the coupling strength obtained from electrical measurements to evaluate the coupled system [17, 18, 19]. However, it is difficult to compare the coupling strength among different studies quantitatively because it depends on the spatial distribution of the SAW, the magnetic field angle, and the size of the magnetic film. Here, we developed a method to evaluate a more practical constant that characterizes the coupled system by combining electrical and optical measurements. The practical constant, bDbD, is the quantity defined by the product of the magnetoelastic constant bb and strain amplitude DD, and it depends mainly on the material. Furthermore, we developed a method to separate the constant bDbD into two types, longitudinal strain-derived and shear strain-derived.

SAW optical imaging is required to evaluate the constant bDbD. Accurate imaging can be performed when the optical spot size is small relative to the wavelength of the SAW [20, 21]. For SAW on LN substrates, which is used in many related studies, the upper limit of the SAW frequency at which accurate optical imaging is possible is about 1 GHz. When using analytical methods of the magnetoelastic coupling under off-resonance conditions, where the resonance frequencies of SAW and SW match, as used in previous studies [1, 2], the evaluation of bDbD is possible when SAW and SW frequencies are the same and below 1 GHz. In general, the resonance frequency of SW in ferromagnets is in the order of gigahertz. Thus, the number of magnetic materials that support SW below 1 GHz is limited. To overcome this limitation, we developed an analytical method to detect the magnetoelastic coupling even under off-resonance conditions. We demonstrated that the practical constant bDbD can be evaluated by detecting the magnetoelastic coupling between a SAW with a frequency of ∼800\sim800 MHz and a SW with a frequency of several gigahertz and then using the SAW’s optical imaging results.

Chapter 2 briefly describes a theoretical model of the SAW-SW coupling system. Then, in Chapter 3, we characterize the SAW using optical measurements and evaluate the magnetoelastic coupling strength using electrical measurements. Finally, the practical constants are estimated by combining these results.

Theoretical Model of coupled system

In this section, we model a system in which the SAW and SW are coupled via magnetoelastic coupling in a magnetic thin film placed in a Fabry-Perot-type SAW resonator. We then deduce the electrical reflection coefficient’s model function, the magnetoelastic coupling strength that has been evaluated in previous studies [17, 18, 19], and the practical constants we focus on in this paper.

Figure 1(a) shows the architecture of a system in which a single SAW mode a^\hat{a} is coupled to a single SW mode c^\hat{c} and an itinerant microwave mode a^i\hat{a}_{i}. We model the SAW, SW, and microwave coupling in a device consisting of a SAW resonator, a magnetic thin film, an interdigital transducer (IDT) shown in Figure 1(b) with the architecture in Figure 1(a). The system is build on the two coupling Hamiltonians HmeH_{\mathrm{me}} and HpH_{\mathrm{p}}. Here, HmeH_{\mathrm{me}} is the coupling between the SAW resonator mode a^\hat{a} and the SW mode c^\hat{c}, given by

Figure 1. (a) Architecture of the SAW-SW coupled system. The system consists of two coupled harmonic oscillator modes, a SAW resonator mode a^\hat{a}, whose energy is specified by ℏωp\hbar\omega_p, and spin wave mode c^\hat{c}, whose energy is specified by ℏωm\hbar\omega_m, and these are coupled at a rate gg. An input (output) itinerant microwave field mode a^i\hat{a}_i (a^o\hat{a}_o) is coupled to the coupled system through the SAW resonator mode at a rate κe\kappa_e. κp\kappa_p and κm\kappa_m are rates of the intrinsic energy dissipation for the SAW resonator mode and the SW, respectively. (b) Microscope image and design of a Fabry-Perot-type SAW resonator and a Ni film in Device 1 with a coordinate system. The SAW resonator consists of a 0.5-mm-thick YZ-cut LiNbO3_3 substrate, Bragg mirrors, and interdigitated transducer (IDT). Bragg mirrors and IDT are made by depositing 5 nm of Ti and 80 nm of Au. Two Bragg mirrors composed of 240 fingers of 100-μ\mum-long with a 1-μ\mum line and space are placed at a distance of 48 μ\mum. The IDT composed of ten fingers of 100-μ\mum-long with a 1-μ\mum line and space are placed 2.5 μ\mum apart from the edge of the Bragg mirror. A Ni(50 nm) film with a length of 21 μ\mum and a width of 20 μ\mum is deposited at the center between Bragg mirror and IDT. The distance between the center of the film and the right Bragg mirror is 35 μ\mum. A magnetic field HDCH_{\mathrm{DC}} is applied at an angle θ\theta in the magnetic film plane. A magnetoresistance in the Ni film is measured by a sourcemeter (SM). The SAW resonator is electrically characterized by microwave reflection measurement using Port 1 of a vector network analyzer (VNA) through IDT. Optical imaging of the SAW resonator is performed using light with a wavelength of 660 nm focused at the surface. The light is incident in the zz-axis direction and the optical path modulation added to the reflected light in the coaxial direction is observed. The reflected light is received by a multi-mode fiber (MMF) and analyzed by a high-speed photodetector (HPD) connected to Port 2 of the VNA.

Hme=ℏg(a^†c^+c^†a^),(1)H_{\mathrm{me}}=\hbar g\left(\hat{a}^{\dagger}\hat{c}+\hat{c}^{\dagger}\hat{a}\right), \tag*{(1)}

with a coupling rate gg. HpH_{\mathrm{p}} describes the coupling between an itinerant microwave mode a^i(ω)\hat{a}_{i}(\omega) and the SAW resonator mode a^\hat{a}, given by

Hp=−iℏκe∫−∞∞dω2π(a^†a^i(ω)−a^i†(ω)a^),(2)H_{\mathrm{p}}=-i\hbar\sqrt{\kappa_{e}}\int_{-\infty}^{\infty}\frac{d\omega}{2\pi}\left(\hat{a}^{\dagger}\hat{a}_{i}(\omega)-\hat{a}_{i}^{\dagger}(\omega)\hat{a}\right), \tag*{(2)}

with a coupling rate κe\kappa_{e}.

Electrical reflection coefficient S11

The total interaction Hamiltonian Ht=Hme+HpH_{\mathrm{t}}=H_{\mathrm{me}}+H_{\mathrm{p}} with the intrinsic dissipations represented by rates κp\kappa_{p} and κm\kappa_{m}, for the SAW mode and the SW mode, respectively, defines the dynamics of the variables in the coupled system. For the SAW mode operator a^\hat{a} we have the following Fourier-domain relation from the equation of motion:

a^(ω)=χp(ω)(−κea^i(ω)−i(g)c^(ω)),(3)\hat{a}(\omega)=\chi_{p}(\omega)\left(-\sqrt{\kappa_{e}}\hat{a}_{i}(\omega)-i(g)\hat{c}(\omega)\right), \tag*{(3)}

where the susceptibility χp(ω)\chi_{p}(\omega) is defined as

χp(ω)=(−i(ω−ωp)+κp+κe2)−1.(4)\chi_{p}(\omega)=\left(-i\left(\omega-\omega_{p}\right)+\frac{\kappa_{p}+\kappa_{e}}{2}\right)^{-1}. \tag*{(4)}

Here and hereafter the thermal and quantum noise terms are omitted. For the SW mode operator c^\hat{c} we have

c^(ω)=χm(ω)(−ga^(ω)),(5)\hat{c}(\omega)=\chi_{m}(\omega)\left(-g\hat{a}(\omega)\right), \tag*{(5)}

where the susceptibility χm(ω)\chi_{m}(\omega) is defined as

χm(ω)=(−i(ω−ωm)+κm2)−1.(6)\chi_{m}(\omega)=\left(-i\left(\omega-\omega_{m}\right)+\frac{\kappa_{m}}{2}\right)^{-1}. \tag*{(6)}

Solving the algebraic Eqs. (1) and (2) with the boundary condition a^o(ω)=a^i(ω)+κea^(ω)\hat{a}_{o}(\omega)=\hat{a}_{i}(\omega)+\sqrt{\kappa_{e}}\hat{a}(\omega) [22], the electrical reflection coefficient can be obtained as

S11(ω)=⟨a^o(ω)a^i(ω)⟩(7)S_{11}(\omega)=\left\langle\frac{\hat{a}_{o}(\omega)}{\hat{a}_{i}(\omega)}\right\rangle \tag*{(7)}
=i(ω−ωp)−κp−κe2+g2i(ω−ωm)−κm/2i(ω−ωp)−κp+κe2+g2i(ω−ωm)−κm/2.(8)=\frac{i\left(\omega-\omega_{p}\right)-\frac{\kappa_{p}-\kappa_{e}}{2}+\frac{g^{2}}{i\left(\omega-\omega_{m}\right)-\kappa_{m}/2}}{i\left(\omega-\omega_{p}\right)-\frac{\kappa_{p}+\kappa_{e}}{2}+\frac{g^{2}}{i\left(\omega-\omega_{m}\right)-\kappa_{m}/2}}. \tag*{(8)}

Magnetoelastic coupling strength g and practical constant bD

The classical magnetoelastic Hamiltonian HmeH_{\mathrm{me}} can be given by [23]

Hme=∫d3r∑i,j,k,l=x,y,zBi,j,k,lmimjMS2ekl,(9)H_{\mathrm{me}}=\int d^{3}r\sum_{i,j,k,l=x,y,z}B_{i,j,k,l}\frac{m_{i}m_{j}}{M_{S}^{2}}e_{kl}, \tag*{(9)}

where, Bi,j,k,lB_{i,j,k,l} is the magnetoelastic constant, mim_{i} is the component of magnetization, MSM_{S} is the saturation magnetization of magnetic material, and ekle_{kl} is the component of the strain tensor. Since only the longitudinal strain exxe_{xx} and shear strain ezxe_{zx} contributions are finite for Rayleigh-type SAW in in-plane magnetized films, Eq. (9) becomes

Hme=∫d3r[b1mx2MS2exx+2b2mzmxMS2ezx],(10)H_{\mathrm{me}}=\int d^{3}r\left[b_{1}\frac{m_{x}^{2}}{M_{S}^{2}}e_{xx}+2b_{2}\frac{m_{z}m_{x}}{M_{S}^{2}}e_{zx}\right], \tag*{(10)}

where, b1=Bx,x,x,xb_{1}=B_{x,x,x,x} and b2=Bz,x,z,xb_{2}=B_{z,x,z,x}. The first and second terms represent the magnetoelastic coupling derived from longitudinal strain and shear strain, respectively. As described in Appendix A, after extracting the time-dependent term of magnetization from Eq. (10) and quantizing it, we obtain Eq. (1) where gg is expressed as

g=γkVSWρvpMSVSAW(b1Dlsin⁡2ϕ)2+(2b2Dscos⁡ϕ)2(11)g=\sqrt{\frac{\gamma kV_{\mathrm{SW}}}{\rho v_{p}M_{S}V_{\mathrm{SAW}}}}\sqrt{(b_{1}D_{l}\sin2\phi)^{2}+(2b_{2}D_{s}\cos\phi)^{2}} \tag*{(11)}
=(glcsin⁡2ϕ)2+(gsccos⁡ϕ)2,(12)=\sqrt{(g_{l}^{c}\sin2\phi)^{2}+(g_{s}^{c}\cos\phi)^{2}}, \tag*{(12)}

where

glc=b1DlγkVSWρvpMSVSAW,(13)g_{l}^{c}=b_{1}D_{l}\sqrt{\frac{\gamma kV_{\mathrm{SW}}}{\rho v_{p}M_{S}V_{\mathrm{SAW}}}}, \tag*{(13)}

and

gsc=2b2DsγkVSWρvpMSVSAW.(14)g_{s}^{c}=2b_{2}D_{s}\sqrt{\frac{\gamma kV_{\mathrm{SW}}}{\rho v_{p}M_{S}V_{\mathrm{SAW}}}}. \tag*{(14)}

Here, γ\gamma is the gyromagnetic ratio of magnetic material, kk is the wavenumber of the SAW, ρ\rho is the mass density of the substrate, vpv_p is the velocity of the SAW, VSAWV_{\mathrm{SAW}} is the volume of the SAW, VSWV_{\mathrm{SW}} is the volume of the magnetic thin film, and ϕ\phi is the relative angle of the magnetic field with respect to the direction of propagation of the SAW. glcg_l^c (gscg_s^c) is the longitudinal (shear) strain-derived magnetoelastic coupling strength, and DlD_l (DsD_s) is the dimensionless quantity representing the amplitude of the longitudinal (shear) strain in the quantized displacement field. The relations between the strains exxe_{xx} and ezxe_{zx} in the classical displacement field and DlD_l and DsD_s are shown in Eqs. (22) and (23), respectively.

Previous studies mainly evaluated the coupling strength gg in Eq. (11) [17, 18, 19], which includes many parameters. To enable a fair and accurate comparison of the magnetoelastic SAW-SW coupling across different material combinations, a normalization is crucial. The b1Dlb_1D_l and b2Dsb_2D_s in Eqs. (13) and (14) are the most suitable and practical quantities to characterize the magnetoelastic SAW-SW coupling in each system because they depend only on the material. By determining other values in Eqs. (13) and (14) from literature, design, electrical measurements, and optical imaging results, we can evaluate the constants, b1Dlb_1D_l and b2Dsb_2D_s. Note that attempts to estimate b1,2b_{1,2} were made in previous studies with assumptions [19, 24, 25]. Also, note that gscg_s^c is suggested to be mixed with an effect of the magneto-rotational coupling, which has the same angular dependence [8, 24, 26, 27].

Experiments

Device preparation

Figure 1(b) shows a microscope image of a SAW-SW device. The part that appears to be gold is a Fabry-Perot-type SAW resonator consisting of two Bragg mirrors and interdigitated transducer (IDT) composed of Ti(5) and Au(80) by magnetron sputtering on a 0.5-mm-thick YZ-cut LiNbO3_3 substrate, where the values in parentheses are the thicknesses in nanometers. The lower panel in Figure 1(b) shows the design drawing of the SAW-SW device. The SAW resonator is fabricated so that the SAW propagation direction is along the crystalline ZZ-axis of the LiNbO3_3 monocrystal. The SAW resonator mode is coherently excited by driving the IDT with a microwave. In the resonator’s center, we deposit a rectangular Ni(50) film by magnetron sputtering.

To investigate the device geometry dependence of the magnetoelastic coupling, we fabricate two distinct SAW-SW devices, Device 1 and Device 2. This approach allows us to compare and analyze the effects of different device geometries on the coupling strength, glcg_l^c and gscg_s^c in Eqs. (13) and (14). The length in the yy-direction of each Ni film is 20 μ\mum in common, and the length in the xx-direction is 21 μ\mum for Device 1 and 13 μ\mum for Device 2. The center position of the Ni film relative to the SAW resonator is the same for the two devices. Note that although the only difference in design is the Ni film size, in practice, the spatial distribution of SAW resonator modes can differ from device to device. As described in Sec. III B, the actual SAW mode volumes are different.

Optical imaging

To determine the SAW resonator modes in both devices, we first acquire two-dimensional plots of the amplitude and phase of the surface slope associated with the SAW. The optical path modulation method, which we reported [20, 21], is employed as the measurement technique. The method diagnoses SAW using path modulation of reflected light caused by the surface slope associated with SAW. As shown in Figure 1(b), light with a wavelength of 660 nm is launched down onto the device surface along the zz direction and focused to a diameter of about 1 μ\mum at the surface. The reflected light is input to a high-speed photodetector (HPD), and then the electrical signal from the HPD is input to Port 2 of a vector network analyzer (VNA). The microwave from Port 1 of the VNA excites the SAW through the IDT, and the excited SAW coherently modulates the path of the reflected light. By acquiring the transmitted signal S21S_{21} while scanning the position of the optical spot, the SAW can be imaged.

Figure 2(a) and Figure 2(b) show the two-dimensional plots of the amplitude and the phase of the optical path modulation signal while the optical spot position is scanned in the xx axis every 0.5 μ\mum and in the yy axis every 2 μ\mum in a region (20 μ\mum ×\times 100 μ\mum) around center of the Ni film in Device 1. The interval between the peak position of ∣S21∣\lvert S_{21}\rvert in Figure 2(a) is about 2 μ\mum, half the wavelength λSAW\lambda_{\mathrm{SAW}} of the excited SAW, and the phase in Figure 2(b) changes by exactly π\pi at each peak. As expected, these results indicate that the SAW is a standing wave in the SAW resonator. Furthermore, in the Ni film region in Figure 2(a) and Figure 2(b), there are no nodes in the yy direction in both phase and amplitude. The result suggests that the fundamental SAW mode is excited in the Ni region. In the same way, the excited SAW mode in Device 2 is also confirmed to be the fundamental mode (see Appendix § 1.2.3 for details). Note that, in contrast to the Ni region, the signal in the LiNbO3_3 region may include intensity modulation due to photoelastic effect [28]. It is not easy to discuss the cause of the phase shift between the Ni and LiNbO3_3 regions here.

Figure 2. Imaging results of the SAW resonator in Device 1 by optical path modulation measurement. Color map of (a) amplitude and (b) phase acquired by the VNA at a frequency ω/2π=803\omega/2\pi= 803 MHz. (c)(d) Position dependence of optical signals in the Bragg mirror. The x=0x = 0 position represents the positon of the opposite ends of the Bragg mirrors. The blue dots represent the optical signal within Bragg mirror on the (c) left side and (d) right side. The red curves represent the results of exponential fittings of the envelopes. Note that here the SAW is excited through the IDT connected to Port 1 of the VNA.

We then experimentally determine the mode volume of the SAW using optical imaging. The effective mode volume is defined as VSAW=w×l×dV_{\mathrm{SAW}} = w \times l \times d, where ww is the width of the IDT, ll is the effective length in the SAW propagation direction, and dd is the effective depth of the SAW resonator mode. Since the SAW is localized on the order of SAW wavelength in the depth direction, we assume d∼λSAWd \sim\lambda_{\mathrm{SAW}}. To experimentally determine the effective length ll, we perform optical imaging of the SAW seeping into the Bragg mirror at y=0y = 0 μ\mum in Figure 2(a). Figure 2(c) and Figure 2(d) show the xx-position dependence of the optical path modulations signal. The position x=0x = 0 is the opposite end of the two Bragg mirrors. The blue dots represent the measured S21S_{21} amplitudes, and the red lines represent the exponential fitting results. The result show that the SAW seeps about 260 μm260\ \mu\mathrm{m} (270 μm270\ \mu\mathrm{m}) into the left (right) Bragg mirror, namely, l=48+260+270=578 μml = 48 + 260 + 270 = 578\ \mu\mathrm{m}. For the SAW resonator in Device 1 we have VSAW=2.2×10−15 m3V_{\mathrm{SAW}} = 2.2 \times10^{-15}\ \mathrm{m}^{3}. Similarly, for the SAW resonator in Device 2 VSAW=1.7×10−15 m3V_{\mathrm{SAW}} = 1.7 \times10^{-15}\ \mathrm{m}^{3} (see Appendix C for details).

Electrical measurements

Magnetic domain

To determine how much magnetic field to apply to the Ni film to make it single magnetic domain, we measure the magnetic field dependence of the resistivity change due to the anisotropic magnetoresistance effect using the experimental setup in Figure 1(b). Figure 3(a) shows the magnetoresistance curve under the magnetic field direction of θ=90∘\theta= 90^\circ. The results of magnetoresistance curve measured at various θ\theta suggest that the Ni film is a single magnetic domain when μ0HDC=10 mT\mu_0 H_{\mathrm{DC}} = 10\ \mathrm{mT} or more (blue background) is applied at an arbitrary angle θ\theta. In a single magnetic domain, we can assume that a SW mode coupled to the SAW resonator mode has the same wavenumber as the SAW resonator mode.

Figure 3. (a) Anisotropic magnetoresistance measurement on Device 1 in θ=90∘\theta= 90^\circ. ΔR\Delta R is obtained by subtracting the resistance for each magnetic field from the resistance at μ0HDC=200 mT\mu_0 H_{\mathrm{DC}} = 200\,\mathrm{mT}. (b) Magnetic field dependence of the resonance frequency of SW with the same wavenumber as SAW with a resonance frequency of 803 MHz. Green and red lines represent the resonance frequencies of SW at magnetic field angles θ=90∘\theta= 90^\circ and θ=0∘\theta= 0^\circ, respectively. (c) Amplitude and (d) phase spectra (blue dots) of the SAW resonator mode acquired by microwave reflection measurement S11S_{11} at Port 1 of the VNA under magnetic field μ0HDC=15 mT\mu_0 H_{\mathrm{DC}} = 15\,\mathrm{mT} and angle θ=45∘\theta= 45^\circ. Red curves represent the fitting results using Eq. (8). The dip around 803 MHz (black dash line) in (c) corresponds to the SAW resonator mode. (e)(f) Color maps of evaluated coupling strength gg in (e) Device 1 and (f) Device 2 as a function of magnetic field HDCH_{\mathrm{DC}} and magnetic field angle θ\theta.

SW resonance frequency

Figure 3(b) shows a magnetic field dependence of calculated resonance frequencies of the SW with the same wavenumber k=2π/λSAW=2π×2.5×105 m−1k = 2\pi/\lambda_{\mathrm{SAW}} = 2\pi\times2.5 \times10^{5}\ \mathrm{m}^{-1} as the SAW resonator mode in Ni film in Device 1 assuming no magnetic anisotropy [29]. The green and red lines are the dependence at θ=90∘\theta= 90^\circ and θ=0∘\theta= 0^\circ, while other angle cases lie between them. Figure 3(b) implies that SW modes in Ni film are off-resonance with SAW resonator mode in the magnetic field region 10 mT−20 mT10\ \mathrm{mT} - 20\ \mathrm{mT} (blue background) where we perform experiments.

Evaluation of coupled system parameters

Here, we first summarize experimentally achieved parameters relevant to the SAW-SW coupled system shown in Figure 1(a), such as the coupling rate gg and κe\kappa_e appearing in Eqs. (1) and (2), respectively, and the intrinsic dissipations represented by the rates κp\kappa_p and κm\kappa_m for the SAW resonator mode and the SW mode. The experimental setup used in evaluating these parameters is shown in Figure 1(b). An itinerant microwave field generated from Port 1 of the VNA drives the system consisting of the SAW resonator and the SW mode via the IDT.

To evaluate the parameters ωp\omega_p, κp\kappa_p, and κe\kappa_e related to the SAW resonator, we measure the reflection coefficient S11S_{11} from the SAW resonator in the condition where the SAW and SW are far-off-resonance (μ0HDC=200 mT\mu_0 H_{\mathrm{DC}} = 200\ \mathrm{mT}). Under such condition, the SAW and the SW are not coupled, and the term including gg in Eq. (8) can be neglected, and the reflection coefficient of the SAW resonator S11S_{11} is simplified to

S11(ω)=i(ω−ωp)−κp−κe2i(ω−ωp)−κp+κe2.(15)S_{11}(\omega) = \frac{i(\omega-\omega_p)-\frac{\kappa_p-\kappa_e}{2}}{i(\omega-\omega_p)-\frac{\kappa_p+\kappa_e}{2}}. \tag*{(15)}

By fitting the obtained result using Eq. (15), we obtain the SAW resonator-related parameters as ωp/2π=803 MHz\omega_p/2\pi= 803\ \mathrm{MHz}, κp/2π=0.61 MHz\kappa_p/2\pi= 0.61\ \mathrm{MHz}, and κe/2π=7.81 kHz\kappa_e/2\pi= 7.81\ \mathrm{kHz}.

When the magnetic field μ0HDC\mu_0 H_{\mathrm{DC}} is reduced from 200 mT200\ \mathrm{mT} to 15 mT15\ \mathrm{mT}, the S11S_{11} signal will contain SW information shown in Eq. (8) because the frequency difference between the SAW and SW will become smaller, as shown in Figure 3(b). In other words, the term containing gg in Eq. (8) cannot be ignored in this situation. Figure 3(c) and Figure 3(d) show the electrical reflection spectra S11(ω)S_{11}(\omega) of the SAW resonator for Device 1 at μ0HDC=15 mT\mu_0 H_{\mathrm{DC}} = 15\ \mathrm{mT} and θ=45∘\theta= 45^\circ. From the fitting in Figure 3(c) and Figure 3(d) based on Eq. (8), we obtain g/2π=5.67 MHzg/2\pi= 5.67\ \mathrm{MHz}, ωm/2π=3.61 GHz\omega_m/2\pi= 3.61\ \mathrm{GHz}, and κm/2π=1.07 GHz\kappa_m/2\pi= 1.07\ \mathrm{GHz}. Similarly, by fitting the S11S_{11} spectra obtained at each magnetic field HDCH_{\mathrm{DC}} and field angle θ\theta, we obtain two-dimensional-color-maps of the evaluated coupling strength gg for Device 1 and Device 2, in Figure 3(e) and Figure 3(f), respectively. There is no apparent dependence on the magnetic field. This feature is consistent with the fact that the coupling strength gg in Eq. (11) is independent of the magnetic field. Note that as the magnetic field angle θ\theta is varied, the parameters ωp\omega_p, κp\kappa_p, and κe\kappa_e, related to the SAW resonator also vary due to magnetostriction (see Appendix B for details). When fitting at an angle θ\theta, we use the parameters evaluated under far-off-resonance conditions at the θ\theta.

Next, in preparation for estimating the practical constants b1,2D1,sb_{1,2}D_{1,s}, we decompose evaluated gg into glcg_l^c and gscg_s^c using Eq. (12). For instance, Figure 4(a) shows a cross section of Figure 3(e) at the magnetic field μ0HDC=15 mT\mu_0 H_{\mathrm{DC}} = 15\ \mathrm{mT} (white dash line). From the fitting Figure 4(a) based on Eq. (12), we obtain glc/2π=5.4 MHzg_l^c/2\pi= 5.4\ \mathrm{MHz} and gsc/2π=3.1 MHzg_s^c/2\pi= 3.1\ \mathrm{MHz}. Figure 4(b) shows the magnetic field dependence of glcg_l^c and gscg_s^c obtained from the fitting of Figure 3(e) and Figure 3(f) at each magnetic field. The blue and red colors represent glcg_l^c and gscg_s^c, and the pattern fill or not corresponds to Device 1 or Device 2. The four lines represent the average of the respective values. Each average value is noted in Table I and II.

Figure 4. (a) Cross section of Figure 3(e) at a magnetic field of 15 mT (indicated by white dash line) with fitting curve (red line) based on Eq. (12). (b) Magnetic field dependence of glcg_l^c and glsg_l^s for each device evaluated using Eq. (12) from the results in Figure 3(e) and Figure 3(f). The four lines represent the average of the respective values.

Here, to validate the evaluation method in this paper, we confirm the device geometry dependence of glcg_l^c and gscg_s^c. From Eqs. (13) and (14), the ratio of glcg_l^c and gscg_s^c between each device depends only on VSAWV_{\mathrm{SAW}} and VSWV_{\mathrm{SW}}, and can be written together as Rl,s=gl,sc(1)/gl,sc(2)=VSW(1)VSAW(2)VSW(2)VSAW(1)R_{l,s} = g_{l,s}^c(1)/g_{l,s}^c(2) = \sqrt{\frac{V_{\mathrm{SW}}(1)V_{\mathrm{SAW}}(2)}{V_{\mathrm{SW}}(2)V_{\mathrm{SAW}}(1)}}, where the number of the device is written in parentheses. The reason why values other than

ParameterValue
glc/2πg_l^c/2\pi5.58±0.025.58 \pm0.02 MHz
gsc/2πg_s^c/2\pi2.95±0.042.95 \pm0.04 MHz
SAW mode volume: VSAWV_{\mathrm{SAW}}2.26×103 μm32.26 \times10^{3}\ \mu\mathrm{m}^{3}
Ni film volume: VSWV_{\mathrm{SW}}2.1 μm32.1\ \mu\mathrm{m}^{3}
SAW velocity: vp=(ωp/2π)λSAWv_p = (\omega_p/2\pi)\lambda_{\mathrm{SAW}}3.2×103 m/s3.2 \times10^{3}\ \mathrm{m/s}
SAW wavenumber: k/2πk/2\pi2.5×107 m−12.5 \times10^{7}\ \mathrm{m}^{-1}
Ni gyromagnetic ratio: γ/2π\gamma/2\pi30.59 GHz/T30.59\ \mathrm{GHz/T} [30]
Ni saturation magnetization: MSM_{\mathrm{S}}5.25×105 A/m5.25 \times10^{5}\ \mathrm{A/m}
LiNbO3_3 mass density: ρ\rho8.9×103 kg/m38.9 \times10^{3}\ \mathrm{kg/m}^{3} [31]

Table 1. Parameters of Device 1.

VSAWV_{\mathrm{SAW}} and VSWV_{\mathrm{SW}} can be regarded as common regardless of devices is that both devices are fabricated on the same substrate by the same process. From the design values of VSW(1,2)V_{\mathrm{SW}}(1, 2) and the optically evaluated VSAW(1,2)V_{\mathrm{SAW}}(1, 2) in Sec. III B, Rl,sR_{l,s} is obtained as 1.09. Whereas, from Figure 4(b), the experimentally evaluated RlR_l and RsR_s are 1.20 and 1.11, respectively. All values are close, confirming the validity of the method.

Finally, we estimate the practical constants b1Dlb_1D_l and b2Dsb_2D_s, that depends only on the material and fabrication process. Substituting the values of glcg_l^c and gscg_s^c for each device in Figure 4(b) and the known values summarized in Table I into Eqs. (13) and (14), we obtain b1Dl=8.1×105 J/m3b_1D_l = 8.1 \times10^{5}\ \mathrm{J/m}^{3} and b2Ds=4.3×105 J/m3b_2D_s = 4.3 \times10^{5}\ \mathrm{J/m}^{3} for Device 1 and b1Dl=7.4×105 J/m3b_1D_l = 7.4 \times10^{5}\ \mathrm{J/m}^{3} and b2Ds=4.3×105 J/m3b_2D_s = 4.3 \times10^{5}\ \mathrm{J/m}^{3} for Device 2. Here, the saturation magnetization MSM_{\mathrm{S}} in Table I is obtained by SQUID measurements. Each constant is very well matched across devices. This is evidence that our method can estimate the practical constant b1,2Ds,lb_{1,2}D_{s,l} that characterizes the magnetoelastic coupling.

Discussion

Our method demonstrated in this paper, which combines optical imaging and electrical measurements, is helpful for accurate understanding SAW-SW coupling systems. Since SAW modes can be easily affected by the magnetic thin film on the substrate and by the structure and quality of the electrodes, it is essential to confirm the SAW mode and to evaluate the SAW mode volume by optical imaging. For SAW resonators on a LiNbO3_3 substrate, the upper limit of the SAW resonance frequency at which optical imaging is possible is about 1 GHz. However, in general, the resonance frequency of SW in magnetic thin films under an applied magnetic field that makes them a single magnetic domain is several gigahertz. Hence, only a few materials were used in magnetoelastic coupling studies, and light was used sparingly. The analysis presented in this paper, which can also be used under off-resonance conditions, allows the introduction of optical imaging and a more comprehensive selection of magnetic materials. Our method will significantly contribute to the field, promising a more accurate evaluation of the magnetoelastic coupling.

Furthermore, our method provides an experimental way to evaluate the most practical constant bDbD that characterizes the SAW-SW coupled system. First, bb is the magnetoelastic coupling constant determined by the magnetic thin film’s physical properties. This value can vary depending on the substrate materials, surface conditions, and the fabrication processes. Second, DD is the constant that determines the magnitude of the strain created by the SAW, which is determined by the substrate’s physical properties. This quantity is also sensitive to the magnetic thin film and metallic structures placed on top. In other words, these constants bb and DD are challenging to estimate analytically. On the other hand, their product bDbD can be determined experimentally using our method and is directly comparable among studies. Experimental evaluation and comparison of bDbD is essential.

In this study, we assume that the SW is excited by the SAW and has the same wavenumber as the SAW. Previous X-ray study [32], simultaneous detection of SAW and SW under on-resonance conditions in a setup similar to that in this paper, which verified the assumption. Future studies with X-ray and light [33, 34, 35, 36] will verify the above assumption under off-resonance conditions.

Conclusion

In conclusion, we demonstrated the novel method for evaluating the most practical constant bDbD characterizing magnetoelastic SAW-SW coupling using electrical and optical measurements. The constant bDbD is directly comparable among studies. The analysis that can be used even under off-resonance conditions is essential to the method’s development. The analysis enables the observation of the magnetoelastic couplings between SAW with resonance frequencies that are optically imageable and SW with resonance frequencies in the gigahertz range. Our method will further advance the study of coupled SAW-SW systems.

Acknowledgments

This work was supported by JSPS KAKENHI (grant no. JP21K18145, JP22K14589, JP22KJ1995, JP23KJ1209, JP23KJ1159, JP24H00007, JP24H02232), JST (grant no. JPMJFS2123, JPMJPR200A), MEXT Initiative to Establish Next-generation Novel Integrated Circuits Centers (X-NICS) (grant no. JPJ011438), and the Collaborative Research Program of the Institute for Chemical Research, Kyoto University.

Derivation of magnetoelastic coupling strength gg

magnetoelastic coupling strength gg

This Appendix describes how to obtain the theoretical expression of gg in Eq. (11).

A magnetization vector m\mathbf{m} can be expressed as follows, considering the situation where the magnetization is saturated at an angle ϕ\phi.

m=(mxmymz)=(MScos⁡ϕ−mζsin⁡ϕMSsin⁡ϕ+mζcos⁡ϕmξ).(16)\mathbf{m} = \begin{pmatrix} m_x \\ m_y \\ m_z \end{pmatrix} = \begin{pmatrix} M_{\mathrm{S}}\cos\phi- m_{\zeta}\sin\phi\\ M_{\mathrm{S}}\sin\phi+ m_{\zeta}\cos\phi\\ m_{\xi} \end{pmatrix}. \tag*{(16)}

Where mξm_{\xi} and mζm_{\zeta} are the components perpendicular to the quantization axis, oriented in-plane and out-of-plane, respectively. They are time-dependent. Substituting Eq. (16) into Eq. (10) and extracting only the first-order term of the time-dependent magnetization, Eq. (10) becomes

Hme=∫d3r[b1mζMSsin⁡(2ϕ)exx+4b2mξMScos⁡(ϕ)ezx].(17)H_{\mathrm{me}}=\int d^{3}r\left[b_{1}\frac{m_{\zeta}}{M_{\mathrm{S}}}\sin(2\phi)e_{xx}+4b_{2}\frac{m_{\xi}}{M_{\mathrm{S}}}\cos(\phi)e_{zx}\right]. \tag*{(17)}

Next, we represent the strain eije_{ij} using boson operators where i,j∈{x,y,z}i,j\in\{x,y,z\}. Here, we consider plane waves located on the xyxy-plane and propagating in the xx-axis direction. Displacement and momentum density are written using the boson operators aμka_{\mu k}, aμk†a_{\mu k}^{\dagger} as follows [37, 38, 39],

ui=ℏ2ρVSAW∑k,μϵiμ(k)ωμk−1/2fμk(z)(aμk†e−ikx+aμkeikx)(18)u_i=\sqrt{\frac{\hbar}{2\rho V_{\mathrm{SAW}}}}\sum_{k,\mu}\epsilon_{i\mu}(k)\omega_{\mu k}^{-1/2}f_{\mu k}(z)\left(a_{\mu k}^{\dagger}e^{-ikx}+a_{\mu k}e^{ikx}\right) \tag*{(18)}

and

ρu˙i=ρℏ2VSAW∑k,μiϵiμ(k)ωμk1/2fμk(z)(aμk†eikx−aμke−ikx).(19)\rho\dot{u}_i=\sqrt{\frac{\rho\hbar}{2V_{\mathrm{SAW}}}}\sum_{k,\mu}i\epsilon_{i\mu}(k)\omega_{\mu k}^{1/2}f_{\mu k}(z)\left(a_{\mu k}^{\dagger}e^{ikx}-a_{\mu k}e^{-ikx}\right). \tag*{(19)}

where μ={x,y,z}\mu=\{x,y,z\}, kk is wavenumber, ϵiμ(k)\epsilon_{i\mu}(k) is unitary polarization vector, ωμk\omega_{\mu k} is angular frequency of the acoustic waves, and fμk(z)f_{\mu k}(z) is dimensionless decay term that depends on the zz-position and can be expressed as fμk(z)=∑μAμe−Bμkzf_{\mu k}(z)=\sum_{\mu}A_{\mu}e^{-B_{\mu}kz}. Here AμA_{\mu} and BμB_{\mu} are dimensionless constants determined solely by the material constants. From Eqs. (18) and (19), exxe_{xx} and ezxe_{zx} can be quantized as follows,

exx=−ikℏ2ρVSAW∑kωk−1/2fxk(z)(axk†e−ikx−axkeikx)(20)e_{xx}=-ik\sqrt{\frac{\hbar}{2\rho V_{\mathrm{SAW}}}}\sum_k\omega_k^{-1/2}f_{xk}(z)\left(a_{xk}^{\dagger}e^{-ikx}-a_{xk}e^{ikx}\right) \tag*{(20)}

and

ezx=12ℏ2ρVSAW∑kωk−1/2{(axk†e−ikx+axkeikx)∂fxk(z)∂z−ik(azk†e−ikx−azkeikx)fzk(z)}.(21)e_{zx}=\frac{1}{2}\sqrt{\frac{\hbar}{2\rho V_{\mathrm{SAW}}}}\sum_k\omega_k^{-1/2}\left\{\left(a_{xk}^{\dagger}e^{-ikx}+a_{xk}e^{ikx}\right)\frac{\partial f_{xk}(z)}{\partial z}-ik\left(a_{zk}^{\dagger}e^{-ikx}-a_{zk}e^{ikx}\right)f_{zk}(z)\right\}. \tag*{(21)}

Note that the frequency ωμ\omega_{\mu} is assumed to be constant.

Now consider the SAW resonator mode. At the surface, i.e., z=0z=0, the decay terms can be described as fxk(z)=D1f_{xk}(z)=D_{1}, ∂fxk(z)∂z=kD2\frac{\partial f_{xk}(z)}{\partial z}=kD_{2}, and fzk(z)=D3f_{zk}(z)=D_{3} with the constants D1D_{1}, D2D_{2}, and D3D_{3}, which are determined only by the material properties. For standing waves (k=±k0k=\pm k_{0}), exxe_{xx} and ezxe_{zx} are written as

exx=2ℏ2ρVSAWvpk0D1cos⁡(k0x)(axk0†+axk0)(22)e_{xx}=2\sqrt{\frac{\hbar}{2\rho V_{\mathrm{SAW}}v_{\mathrm{p}}k_{0}}}D_{1}\cos(k_{0}x)\left(a_{xk_{0}}^{\dagger}+a_{xk_{0}}\right) \tag*{(22)}

and

ezx=ℏ2ρVSAWvpk0cos⁡(k0x){D2(axk0†+axk0)+D3(azk0†+azk0)},(23)e_{zx}=\sqrt{\frac{\hbar}{2\rho V_{\mathrm{SAW}}v_{\mathrm{p}}k_{0}}}\cos(k_{0}x)\left\{D_{2}\left(a_{xk_{0}}^{\dagger}+a_{xk_{0}}\right)+D_{3}\left(a_{zk_{0}}^{\dagger}+a_{zk_{0}}\right)\right\}, \tag*{(23)}

where ωk0=ω−k0=vpk0\omega_{k_{0}}=\omega_{-k_{0}}=v_{\mathrm{p}}k_{0}.

Similarly, we also represent the micro-magnetization mkm_k using the boson operators. Defining circularly polarised components as m±=mζ±imξm^{\pm}=m_{\zeta}\pm im_{\xi}, m±m^{\pm} is expressed using the magnon operators ck,ck†c_k,c_k^{\dagger} as follows [39],

m+=ℏγMS2VSW∑keikrck.(24)m^{+}=\sqrt{\frac{\hbar\gamma M_{\mathrm{S}}}{2V_{\mathrm{SW}}}}\sum_k e^{ikr}c_k. \tag*{(24)}

Using this expression, we get

mζ=12ℏγMS2VSW∑k(eikrck+e−ikrck†)(25)m_{\zeta}=\frac{1}{2}\sqrt{\frac{\hbar\gamma M_{\mathrm{S}}}{2V_{\mathrm{SW}}}}\sum_k\left(e^{ikr}c_k+e^{-ikr}c_k^{\dagger}\right) \tag*{(25)}

and

mξ=12iℏγMS2VSW∑k(eikrck−e−ikrck†).(26)m_{\xi}=\frac{1}{2i}\sqrt{\frac{\hbar\gamma M_{\mathrm{S}}}{2V_{\mathrm{SW}}}}\sum_k\left(e^{ikr}c_k-e^{-ikr}c_k^{\dagger}\right). \tag*{(26)}

Substituting Eqs. (22), (23), (25), and (26) into Eq. (17), HmeH_{\mathrm{me}} is quantized as follows

Hme=ℏγk0ρvpMS∫d3rcos⁡2(k0r)VSAWVSW[b1sin⁡2ϕ(ck0+ck0†)(axk0+axk0†)−2ib2cos⁡ϕ(ck0−ck0†){(axk0+axk0†)D2+(azk0+azk0†)D3}].(27)H_{\mathrm{me}}=\hbar\sqrt{\frac{\gamma k_{0}}{\rho v_{\mathrm{p}}M_{\mathrm{S}}}}\frac{\int d^{3}r\cos^{2}(k_{0}r)}{\sqrt{V_{\mathrm{SAW}}V_{\mathrm{SW}}}}\left[b_{1}\sin2\phi\left(c_{k_{0}}+c_{k_{0}}^{\dagger}\right)\left(a_{xk_{0}}+a_{xk_{0}}^{\dagger}\right)-2ib_{2}\cos\phi\left(c_{k_{0}}-c_{k_{0}}^{\dagger}\right)\left\{\left(a_{xk_{0}}+a_{xk_{0}}^{\dagger}\right)D_{2}+\left(a_{zk_{0}}+a_{zk_{0}}^{\dagger}\right)D_{3}\right\}\right]. \tag*{(27)}

To simplify Eq. (27), we further introduce a new operator,

atk=C(b1D1sin⁡(2ϕ)cxk−2ib2D2cos⁡(ϕ)cxk−2ib2D3cos⁡(ϕ)czk),(28)a_{tk}=C\left(b_{1}D_{1}\sin(2\phi)c_{xk}-2ib_{2}D_{2}\cos(\phi)c_{xk}-2ib_{2}D_{3}\cos(\phi)c_{zk}\right), \tag*{(28)}

where C=1/(b1D1sin⁡2ϕ)2+{2b2(D2+D3)cos⁡ϕ}2C=1/\sqrt{(b_{1}D_{1}\sin2\phi)^{2}+\{2b_{2}(D_{2}+D_{3})\cos\phi\}^{2}} and ctkc_{tk} is boson operator (t∈{x,y,z}t\in\{x,y,z\}). Applying the rotational wave approximation to Eq. (27), HmeH_{\mathrm{me}} can be written as

ParameterValue
glc/2πg_l^c/2\pi4.65±0.02 MHz4.65 \pm0.02\ \mathrm{MHz}
gsc/2πg_s^c/2\pi2.66±0.02 MHz2.66 \pm0.02\ \mathrm{MHz}
SAW mode volume: VSAWV_{\mathrm{SAW}}1.70×103 μm31.70 \times10^{3}\ \mu\mathrm{m}^{3}
Ni film volume: VmagV_{\mathrm{mag}}1.3 μm31.3\ \mu\mathrm{m}^{3}

Table 2. TABLE II. Parameters of Device 2.

Hme=ℏγk0VSWρvpMSVSAW(b1Dlsin⁡2ϕ)2+(2b2Dscos⁡ϕ)2(atk0ck0†+atk0†ck0)(29)H_{\mathrm{me}} = \hbar\sqrt{\frac{\gamma k_{0} V_{\mathrm{SW}}}{\rho v_{\mathrm{p}} M_{\mathrm{S}} V_{\mathrm{SAW}}}} \sqrt{\left(b_{1}D_{l}\sin2\phi\right)^{2}+\left(2b_{2}D_{s}\cos\phi\right)^{2}} \left(a_{t k_{0}} c_{k_{0}}^{\dagger}+a_{t k_{0}}^{\dagger}c_{k_{0}}\right) \tag*{(29)}
=ℏg(atk0ck0†+atk0†ck0),(30)= \hbar g\left(a_{t k_{0}} c_{k_{0}}^{\dagger}+a_{t k_{0}}^{\dagger}c_{k_{0}}\right), \tag*{(30)}

where

g=γk0VSWρvpMSVSAW(b1Dlsin⁡2ϕ)2+(2b2Dscos⁡ϕ)2.(31)g = \sqrt{\frac{\gamma k_{0} V_{\mathrm{SW}}}{\rho v_{\mathrm{p}} M_{\mathrm{S}} V_{\mathrm{SAW}}}} \sqrt{\left(b_{1}D_{l}\sin2\phi\right)^{2}+\left(2b_{2}D_{s}\cos\phi\right)^{2}}. \tag*{(31)}

Here, Dl=D1D_{l}=D_{1}, Ds=D2+D3D_{s}=D_{2}+D_{3}, and the integral interval is the entire thin magnetic film. Replacing k0=kk_{0}=k, atk0=a^a_{t k_{0}}=\hat{a}, ck0=c^c_{k_{0}}=\hat{c}, we obtain Eq. (11).

Magnetic field angle dependence of SAW resonator characteristics

This appendix describes the effect of the applied magnetic field on the SAW resonator through magnetostriction. Figure 5 shows the magnetic field angle dependence of the resonance frequency ωp\omega_{p} of the SAW resonator modes for Device 1 and Device 2 at a magnetic field of 200 mT. The other parameters, κp\kappa_{p} and κe\kappa_{e}, related the SAW resonator similarly depend on the magnetic field angle. These changes are due to the magnetostriction of the Ni thin film.

Figure 5. Magnetic field angle dependence of the SAW resonance frequency ωp\omega_p in Device 1 and Device 2 at μ0HDC=200\mu_0 H_{\mathrm{DC}} = 200 mT.

Measurement results of Device 2

This Appendix summarises the measurement results of Device 2.

Figure 6(a) shows an optical microscope image of Device 2. Device 1 and 2 use the same SAW resonator design. The center position of the Ni thin film in Device 2 is the same as Device 1, but the xx-axis length of the Ni thin film is 13 μm13\ \mu\mathrm{m}.

Figure 6. (a) Microscope image of Device 2. (b) AMR measurement on Device 2 in θ=0∘\theta= 0^\circ. ΔR\Delta R is obtained by subtracting the resistance for each magnetic field from the resistance at μ0HDC=200\mu_0 H_{\mathrm{DC}} = 200 mT. (c) Imaging results of the SAW resonator in Device 2. Color map of amplitude and phase at a frequency ω/2π=803\omega/2\pi= 803 MHz. (d) Position dependence of optical signals in the Bragg mirror.

Figure 6(b) shows the AMR measurement results when an external magnetic field is applied in the xx axis. The result show that the resistance is saturated above 10 mT, indicating that the Ni film in Device 2 is single magnetic domain.

Next, we show the optical measurement results of Device 2 and discuss the SAW mode and SAW mode volume. Figure 6(c) shows two-dimensional plots of the amplitude and phase of the detected optical path modulation signal by scanning the optical spot in the xx axis every 0.5 μm0.5\ \mu\mathrm{m} and in the yy axis every 2 μm2\ \mu\mathrm{m} in a region (20 μm×100 μm20\ \mu\mathrm{m} \times100\ \mu\mathrm{m}) around center of the Ni film in Device 2. The phase distribution in Figure 6(c) shows that the SAW excited in the Ni film is a fundamental mode with a wavelength of 4 μm4\ \mu\mathrm{m} in Device 2 as in Device 1.

We next obtain the effective SAW mode volume in Device 2. The difference from Device 1 is the length of the seepage ll into the Bragg mirrors. Figure 6(d) shows that the SAW seeps about 220 μm220\ \mu\mathrm{m} (160 μm160\ \mu\mathrm{m}) into the left (right) Bragg mirror, namely, l=48+220+160=422 μml=48+220+160=422\ \mu\mathrm{m}.

Finally, we estimate the constants b1,2Dl,sb_{1,2}D_{l,s} of Device 2. Table II shows the parameters of Device 2. Note that vpv_{p}, kk, γ\gamma, MsM_{s}, and ρ\rho are the same as in Device 1 in Table I. Substituting these parameters into Eqs. (13) and (14), we obtain b1Dl=7.4×105 J/m3b_{1}D_{l}=7.4\times10^{5}\ \mathrm{J/m^{3}} and b2Ds=4.3×105 J/m3b_{2}D_{s}=4.3\times10^{5}\ \mathrm{J/m^{3}}.

References

  1. [1]M. Weiler, L. Dreher, C. Heeg, H. Huebl, R. Gross, M. S. Brandt, and S. T. B. Goennenwein, “Elastically Driven Ferromagnetic Resonance in Nickel Thin Films”, Physical Review Letters 106, 117601 (2011).
  2. [2]L. Dreher, M. Weiler, M. Perpeintner, H. Huebl, R. Gross, M. S. Brandt, and S. T. B. Goennenwein, “Surface Acoustic Wave Driven Ferromagnetic Resonance in Nickel Thin Films: Theory and Experiment”, Physical Review B 86, 134415 (2012).DOI
  3. [3]P. G. Gowtham, T. Moriyama, D. C. Ralph, and R. A. Buhrman, “Traveling Surface Spin-Wave Resonance Spectroscopy Using Surface Acoustic Waves”, Journal of Applied Physics 118, 233910 (2015).DOI
  4. [4]X. Li, D. Labanowski, S. Salahuddin, and C. S. Lynch, “Spin Wave Generation by Surface Acoustic Waves”, Journal of Applied Physics 122, 043904 (2017).DOI
  5. [5]M. Kuss, M. Heigl, L. Flacke, A. Hefele, A. Horner, M. Weiler, M. Albrecht, and A. Wixforth, “Symmetry of the Magnetoelastic Interaction of Rayleigh and Shear Horizontal Magnetoacoustic Waves in Nickel Thin Films on LiTaO₃”, Physical Review Applied 15, 034046 (2021).DOI
  6. [6]Y. Li, C. Zhao, W. Zhang, A. Hoffmann, and V. Novosad, “Advances in Coherent Coupling between Magnons and Acoustic Phonons”, APL Materials 9, 060902 (2021).DOI
  7. [7]N. K. P. Babu, A. Trzaskowska, P. Graczyk, G. Centala, S. N. Mieszczak, H. Glowinski, M. Zdunek, S. Mielcarek, and J. W. Kłos, “The Interaction between Surface Acoustic Waves and Spin Waves: The Role of Anisotropy and Spatial Profiles of the Modes”, Nano Letters 21, 946 (2021).DOI
  8. [8]K. Yamamoto, M. Xu, J. Puebla, Y. Otani, and S. Maekawa, “Interaction between Surface Acoustic Waves and Spin Waves in a Ferromagnetic Thin Film”, Journal of Magnetism and Magnetic Materials 545, 168672 (2022).DOI
  9. [9]R. Sasaki, Y. Nii, Y. Iguchi, and Y. Onose, “Nonreciprocal Propagation of Surface Acoustic Wave in Ni/LiNbO3”, Physical Review B 95, 040407(R) (2017).DOI
  10. [10]H. Matsumoto, T. Kawada, M. Ishibashi, M. Kawaguchi, and M. Hayashi, “Large Surface Acoustic Wave Non-reciprocity in Synthetic Antiferromagnets”, Applied Physics Express 15, 063003 (2022).DOI
  11. [11]J. Puebla, Y. Hwang, S. Maekawa, and Y. Otani, “Perspectives on Spintronics with Surface Acoustic Waves”, Applied Physics Letters 120, 220502 (2022).DOI
  12. [12]M. Kuss, M. Albrecht, and M. Weiler, “Chiral Magnetoacoustics”, Frontiers in Physics 10, 981257 (2022).DOI
  13. [13]P. Rovillain, J. Y. Duquesne, L. Christienne, M. Eddrief, M. G. Pini, A. Rettori, S. Tacchi, and M. Marangolo “Impact of Spin-Wave Dispersion on Surface-Acoustic-Wave Velocity”, Physical Review Applied 18, 064043 (2022).DOI
  14. [14]D. Hatanaka, M. Asano, H. Okamoto, and H. Yamaguchi, “Phononic Crystal Cavity Magnomechanics”, Physial Review Applied 19, 054071 (2023).DOI
  15. [15]P. J. Shah, D. A. Bas, A. Hamadeh, and M. Wolf, “Symmetry and Nonlinearity of Spin Wave Resonance Excited by Focused Surface Acoustic Waves”, Advanced Electronic Materials 9, 2300524 (2023).DOI
  16. [16]L. Liao, J. Puebla, K. Yamamoto, J. Kim, S. Maekawa, Y. Hwang, Y. Ba, and Y. Otani, “Valley-Selective Phonon-Magnon Scattering in Magnetoelastic Superlattices”, Physical Review Letters 131, 176701 (2023).arxiv.org/abs/2307.03367
  17. [17]D. Hatanaka, M. Asano, H. Okamoto, K. Kunihashi, H. Sanada, and H. Yamaguchi, “On-Chip Coherent Transduction between Magnons and Acoustic Phonons in Cavity Magnomechanics”, Physical Review Applied 17, 034024 (2022).DOI
  18. [18]H. Matsumoto, I. Yasuda, M. Asano, Y. Todaka, T. Kawada, M. Kawaguchi, D. Hatanaka, and M. Hayashi, “Magnon-Phonon Coupling of Synthetic Antiferromagnets in a Surface Acoustic Wave Cavity Resonator”, Nano Letters 24, 5683 (2024).DOI
  19. [19]Y. Hwang, J. Puebla, K. Kondo, C. Gonzalez-Ballestero, H. Isshiki, C. S. Munoz, L. Liao, F. Chen, W. Luo, S. Maekawa, and Y. Otani, “Strongly Coupled Spin Waves and Surface Acoustic Waves at Room Temperature”, Physical Review Letters 132, 056704 (2024).DOI
  20. [20]R. Hisatomi, K. Taga, R. Sasaki, Y. Shiota, T. Moriyama, and T. Ono, “Quantitative Optical Imaging Method for Surface Acoustic Waves Using Optical Path Modulation”, Physical Review B 107, 165416 (2023).DOI
  21. [21]K. Taga, R. Hisatomi, Y. Ohnuma, R. Sasaki, T. Ono, Y. Nakamura, and K. Usami, “Optical Polarimetric Measurement of Surface Acoustic Waves”, Applied Physics Letters 119, 181106 (2021).DOI
  22. [22]A. A. Clerk, M. H. Devoret, S. M. Girvin, F. Marquardt, and R. J. Schoelkopf, “Introduction to quantum noise, measurement, and amplification”, Review of Modern Physics 82, 1155 (2010).DOI
  23. [23]C. Kittel, “Interaction of Spin Waves and Ultrasonic Waves in Ferromagnetic Crystals”, Physical Review 110, 836 (1958).DOI
  24. [24]M. Kuss, M. Heigl, L. Flacke, A. Horner, M. Weiler, M. Albrecht, and A. Wixforth, “Nonreciprocal Dzyaloshinskii-Moriya Magnetoacoustic Waves”, Physical Review Letters 125, 217203 (2020).DOI
  25. [25]M. Kuss, M. Heigl, L. Flacke, A. Horner, M. Weiler, A. Wixforth, and M. Albrecht, “Nonreciprocal Magnetoacoustic Waves in Dipolar-Coupled Ferromagnetic Bilayers”, Physical Review Applied 15, 034060 (2021).DOI
  26. [26]S. Maekawa and M. Tachiki, “Surface Acoustic Attenuation due to Surface Spin Wave in Ferro- and Antiferromagnets”, in AIP Conference Proceedings 29, 542 (1976).DOI
  27. [27]M. Xu, K. Yamamoto, J. Puebla, K. Baumgaertl, B. Rana, K. Miura, H. Takahashi, D. Grundler, S. Maekawa, and Y. Otani, “Nonreciprocal Surface Acoustic Wave Propagation via Magneto-Rotation Coupling”, Science Advances 6, eabb1724 (2020).DOI
  28. [28]T. Saito, O. Matsuda, M. Tomoda, and O. B. Wright, “Imaging gigahertz surface acoustic waves through the photoelastic effect”, Journal of the Optical Society of America B 27, 2632 (2010).DOI
  29. [29]B. A. Kalinikos and A. N. Slavin, “Theory of Dipole-Exchange Spin Wave Spectrum for Ferromagnetic Films with Mixed Exchange Boundary Conditions”, Journal of Physics C: Solid State Physics 19, 7013 (1986).DOI
  30. [30]A. J. P. Meyer and G. Asch, “Experimental g and g Values of Fe, Co, Ni, and Their Alloys”, Journal of Applied Physics 32, S330 (1961).DOI
  31. [31]A. F. Mills, “Basic Heat and Mass Transfer” (1999).
  32. [32]B. Casals, N. Statuto, M. Foerster, A. Hernandez-Minguez, R. Cichelero, P. Manshausen, A. Mandziak, L. Aballe, J. M. Hernandez, and F. Macia, “Generation and Imaging of Magnetoacoustic Waves over Millimeter Distances”, Physical Review Letters 124, 137202 (2020).DOI
  33. [33]M. Kraimia, P. Kuszewski, J.-Y. Duquesne, A. Lemaitre, F. Margaillan, C. Gourdon, and L. Thevenard, “Time-and Space-Resolved Nonlinear Magnetoacoustic Dynamics”, Physical Review B 101, 144425 (2020).DOI
  34. [34]C. Zhao, Z. Zhang, Y. Li, W. Zhang, J. E. Pearson, R. Divan, Q. Liu, V. Novosad, J. Wang, and A. Hoffmann, “Direct Imaging of resonance Phonon-Magnon Coupling”, Physical Review Applied 15, 014052 (2021).DOI
  35. [35]M. Geilen, A. Nicoliou, D. Narducci, M. Mohseni, M. Bechberger, M. Ender, F. Ciubotaru, B. Hillebrands, A. Müller, C. Adelmann, and P. Pirro, “Fully resonance magneto-elastic spin-wave excitation by surface acoustic waves under conservation of energy and linear momentum”, Applied Physical Letters 120, 242404 (2022).DOI
  36. [36]Y. Kunz, M. Kuss, M. Schneider, M. Geilen, P. Pirro, M. Albrecht, and M. Weiler, “Coherent Surface Acoustic Wave-spin Wave Interactions Detected by Micro-Focused Brillouin Light Scattering Spectroscopy”, Applied Physical Letters 124, 152403 (2024).DOI
  37. [37]C. Kittel, “Quantum Theory of Solids” (John Wiley and Sons, New York, 1963).
  38. [38]C. Cohen Tannoudji, B. Diu, and F. Laloe, “Quantum Mechanics”, United States, Wiley (1999).
  39. [39]S. C. Guerreiro and S. M. Rezende, “Magnon-Phonon Interconversion in a Dynamically Reconfigurable Magnetic Material”, Physical Review B 92, 214437 (2015).DOI

Paper details

vlnc:W_8rdvg69q
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents