betaFeedback

WorksPhysics and Astronomy

Hybrid magnonic spintronic system for tunable broadband signal filtering and microwave generation

Abstract

Non-conventional beyond-the-state-of-the-art signal processing schemes require parallelism, scalability, robustness and energy efficiency to meet the demands of complex data-driven applications. With further research, magnonic and spintronic circuits can potentially help to fulfill these requirements. We present an experimental proof-of-concept of a hybrid device that can employ broad deteriorated microwave signals to excite and detect low energy propagating spin waves (SWs). For this, we use the output signal of a spin-transfer torque nano-oscillator (STNO) and connect it to a RF filter based on a magnonic delay-line. The STNO serves as a tunable nano-scaled signal generator with a broad output linewidth. Its RF output is fed as input into the magnonic delay-line circuit. Tuning the magnetic field solely at the magnonic circuit, we demonstrate the capability to selectively filter a broad RF input, obtaining a spin-wave output signal with a much narrower linewidth. This allows to tune the frequency of the RF signal at the output simply by tuning the magnetic field. Our findings are a first step towards a versatile, energy-efficient and compact wave-based filter with high sensitivity. Such a device can use even low-power, degraded signals and convert them into tunable SW outputs, effectively reducing the need for charge-based signal processing.

The rise of complex data-driven applications such as artificial intelligence, has driven a surge in demand for unconventional signal processing methods that allow for parallelism while adhering to stringent energy constraints[1, 2]. With the size of complementary metal-oxide-semiconductor (CMOS) transistors approaching the atomic scale, effects such as Joule heating and current leakage due to quantum tunneling become more critical as they can potentially hinder performance. Spin-wave (SW) based data processing is one promising route which can help extend the envelope of CMOS technology. Magnonic systems – which manipulate magnons (quanta of SWs) – exhibit intrinsic non-linearity, element of memory and cascadability[3, 4, 5, 6]. The latter enable programmable and reconfigurable circuits, the reason these systems are pursued for applications in magnon computing[7, 8, 9, 10, 11] and magnonic RF processing[12, 13, 14, 15, 16].

Another promising route in the context of "beyond CMOS" technologies is the use of spintronic elements (those that aim at exploiting both the charge and the spin of an electron) to overcome or compensate some of the limitations of CMOS technology[17]. One prominent spintronic device that is under intensive research and prototyping in this regard is the spin-transfer torque nano-oscillator (STNO)[18, 19]. STNOs are intrinsically non-linear devices that operate via electric currents based on the principle of Spin-transfer torque (STT)[20, 21, 22, 23]. Typical STNOs consist of a tri-layer stack whereby two ferromagnetic layers, namely the polarizing and the free layers, are separated by a non-magnetic spacer. Depending on whether the spacer is metallic or insulating, a STNO is referred to as a giant magneto-resistance valve[24, 25] or a magnetic tunnel junction (MTJ)[26], respectively. Injecting a dc current into the STNO leads to the spin-polarization of its electrons due to the polarizing layer. The spin-polarized current then transfers spin angular momentum onto the free layer, thus exerting torque on its magnetization. The precessing magnetization generates a time-varying magneto-resistance, leading to a microwave voltage signal[27, 28, 29]. In the context of microwave generation, a STNO can be thought of as a micro-scaled tunable microwave generator, with typical generation powers lying in the sub-microwatt range and frequencies ranging from a few hundreds of MHz to several tens of GHz[30, 31, 32]. Although still challenging, merging spintronic and magnonic approaches can open promising perspectives for hy-

brid devices capitalizing on strengths from both worlds. In this work, we propose to integrate a STNO with a magnonic circuit to realize a tunable, broadband signal filter. The presented approach is a fully experimental proof-of-concept for the device combining a STNO and a RF filter based on a magnonic delay line aiming to harness advantages from both spintronic and magnonic elements, namely microwave generation and wave-based circuitry, respectively. The STNO is used as a broadband nano-scaled microwave generator, whereby it outputs signal due to noise amplification induced by the injected current. That signal is detected by the SW system and selectively filtered by using the frequency (wave vector) dependent SW excitation efficiency. Then that excited SW signal is detected again at the output end of the circuit amplifying the filtering effect further.

To realize the feasibility of the proof-of-concept, the two circuits will be connected electrically. First steps, however, involve studying each circuit separately before the actual connection between both is carried out. In Figure 1, we present the spintronic circuit comprising the connections to and from the STNO.

Schematic of the circuit constituting the STNO layer

Figure 1. Schematic of the circuit constituting the STNO layer. DC current ISTNOI_{\mathrm{STNO}} is injected into the STNO via a bias-T, which in turn is used to direct back the RF signal generated by the STNO through an amplifier and onto a spectrum analyzer. The STNO is placed between the poles of an electric magnet that generates an in-plane magnetic field BSTNOB_{\mathrm{STNO}}.

The orange dots highlight the over-scaling of STNO with respect to the contact electrodes. The STNO consists of a multilayered stack, the diameter of which is 300 nm. The stack’s composition is as follows: IrMn/CoFe30/Ru/CoFe40B20/MgO/CoFe40B20/Ta/NiFe19 of respective thicknesses: 6 nm/2.6 nm/0.85 nm/1.8 nm/1 nm/2 nm/0.21 nm/7 nm. The same stack has been previously used in[33], where sputtering along with ion beam and optical etching techniques were employed to deposit and fabricate the material stack.

The STNO is placed in an electro-magnet that generates an in-plane magnetic field BSTNOB_{\mathrm{STNO}}. A source-meter is used to supply the STNO with dc current ISTNOI_{\mathrm{STNO}}. The dc current is directed through the dc port of a bias-T, and injected into the structure via a RF probe. The RF output is directed back through the RF port of the bias-T and detected by a spectrum analyzer after being passed through an amplifier (working range 3-8 GHz with an amplification of 30 dB). First, the STNO’s output signal is investigated as a function of BSTNOB_{\mathrm{STNO}} for fixed ISTNOI_{\mathrm{STNO}}. Figure 2(a) shows the power spectrum as a function of field BSTNOB_{\mathrm{STNO}} and detection frequency. We fix ISTNOI_{\mathrm{STNO}} at 3 mA, which provides a sufficiently large microwave output signal. The STNO’s free layer is first saturated at high fields above 500 mT. Thereafter, we sweep BSTNOB_{\mathrm{STNO}} from 52 mT to 9 mT at 3 mA. The field at which the STNO is later operated throughout the experiment is marked with the dotted black line. Three modes are observed, two that increase their frequencies sharply with any field increase, and one whose frequency exhibits a steady linear increase within the swept BSTNOB_{\mathrm{STNO}} range. Based on their spectral distribution, these modes do not match the standard ones expected for such systems[34, 35]. The main mode at the lowest frequencies derives from noise-driven excitation of the uniform mode (FMR), while the other two may be either attributed to picked-up signals from the synthetic anti-ferromagnet, or to non-uniform modal excitations. These studied STNO stacks were designed specifically for experiments investigating vortex-based MTJs. The same device stack has been studied operating in the vortex state in[36]. Nevertheless, due to technical limitations imposed by the Yttrium-Iron-Garnet (YIG) delay line (see rationale in the section below), high frequencies in the GHz range rather than in the MHz one are required. The latter is the reason why the devices were not studied in the vortex state, for typically the dynamics lie in the sub-GHz range[28, 31, 37, 38]. In the framework of this study, we are interested in rf output of the STNO regardless of its origin, as the aim of this study does not evolve around the characterization of STNO devices.

2D color-plot and power spectrum of the STNO output

Figure 2. (a) 2D color-plot of the frequency dependence of the STNO’s output signal on the applied magnetic field BSTNOB_{\mathrm{STNO}} at ISTNO=3I_{\mathrm{STNO}} = 3 mA. The dotted black line represents the chosen field at which the STNO will be operated throughout the experiment. (b) Power spectrum of the STNO’s microwave output at fixed BSTNOB_{\mathrm{STNO}} and ISTNOI_{\mathrm{STNO}}.

Figure 2 (b) represents the STNO output spectrum at fixed BSTNOB_{\mathrm{STNO}} and ISTNOI_{\mathrm{STNO}}. These values are thereafter held constant at 21 mT and 3 mA, respectively throughout the study. Only frequencies between 3 GHz and 5 GHz are shown to match the frequency range of the magnonic circuit, as the latter is limited by the electro-magnet.

In Figure 3 (a), we present the circuit constituting the SW system. To excite and detect SWs, Coplanar Waveguide (CPW) antennas are fabricated on a 100 nm thick YIG chip. The widths of the current carrying lines are 2 μ\mum for the middle line and 1 μ\mum the outer lines with a gap of 2.5 μ\mum in between them. The magnonic circuit is placed between the poles of a second electro-magnet with the generated magnetic field BSWB_{\mathrm{SW}} being perpendicular to the direction of SW propagation. The CPW antennas on both ends, separated by 30μ\mum, are used to respectively excite and detect SWs. The basic functioning of such a spin wave transducer has already been described in numerous other publications[39, 40, 41]. As depicted in Figure 3 (a), for a first characterization, a signal generator connected to one of the CPWs is used to excite SWs at a frequency matching the excitation spectrum of the CPW at the given magnetic field. This excited spin-wave signal can be then detected at the second CPW, from which it is directed to an amplifier, then to a spectrum analyzer. Figure 3 (b) presents the spectrum of the detected SW signal as excited by the microwave signal generator. The signal generator’s power is set to the minimum value (PP = 100 nW). To demonstrate the filtering capability of the magnonic circuit, the magnetic field is swept from 0 mT to 88 mT after applying a microwave signal to the input CPW at a fixed frequency. To excite/detect SWs at 3.75 GHz, a magnetic field BSWB_{\mathrm{SW}} = 64 mT is required (see yellow peak in Figure 3). SW transmission at higher frequencies requires higher magnetic fields, e.g. a frequency of 4.5 GHz can be transmitted using a magnetic field of BSWB_{\mathrm{SW}} = 87 mT. The excitation efficiency of the CPW is wavevector specific, which makes it necessary that the applied magnetic field BSWB_{\mathrm{SW}} is adjusted such that the wavevector specificity matches the chosen microwave frequency via the SW dispersion of the material.

Schematic of the magnonic circuit and detected spin-wave signal

Figure 3. (a) Schematic of the magnonic circuit where Coplanar waveguide antennas are used to excite and detect spin waves. The magnetic field BSWB_{\mathrm{SW}} is applied perpendicular to the SW propagation direction. (b) The detected SW signal, as excited by the microwave signal generator, at the resonant Fields BSWB_{\mathrm{SW}}.

cro­wave frequency via the SW dispersion of the material. For this reason, the SW dispersion for this film is calculated whilst highlighting the CPW’s excitation spectrum (see Figure 4 (a)). Essentially, the shape of the transmission spectrum depends on the dispersion of the propagating spin waves and the excitation efficiency of the antennas used. The CPW’s excitation efficiency spectrum, described by the wave vector spectrum of their magnetic field distribution, is shown in Figure 4 (a). Due to the periodicity of the current distribution, the efficiency shows a maximum at a finite wave vector of 0.7 rad/μ\mum, followed by a second small maximum at higher frequencies. Here, the excitation efficiency describes how efficiently a single antenna excites and detects spin waves of fixed kk. Using the dispersion relation, the wave vector-dependent efficiency is translated into frequency domain. Further, the group velocity determines the influence of the attenuation of the spin waves during propagation. In Figure 4 (a) the calculated SW dispersion for the 100 nm thick YIG chip at different magnetic fields for the Damon–Eshbach geometry (magnetic field applied perpendicular to the direction of propagation of spin waves) is shown. For all magnetic fields shown, the dispersions flatten out towards higher wave vectors. This, in combination with the efficiency, leads to a peak of maximum signal followed by a very weak second peak towards higher frequencies. An exemplary measurement highlighting this behavior for different magnetic fields is shown in Figure 4 (b). For this, a vector network analyzer is used to measure the S21S_{21} entry of the scattering matrix. This measurement is carried-out for fields ranging from 40 mT to 98 mT. If the external field is increased, the frequency of the spin waves increases for a fixed k-vector. As a result, increasing the field in Figure 4 (b), the position of the maximum shifts towards higher frequencies. One key point to address here is the reduced transmitted signal amplitude at lower frequencies. These increased insertion losses are related to the increase in impedance mismatch at lower frequencies, where the impedance of the magnonic device is significantly lower than the 50 Ω\Omega impedances of the connected devices. For this reason, we apply a sufficiently large in-plane field to operate the STNO in the uniform mode regime at frequencies that match those of the magnonic system in the range with small insertion losses (3 GHz and above). In the in-plane oscillating regime, the lowest frequency mode is on the order of GHz, and hence can yield a better SW frequency matching for the studied magnonic system. For detailed information and quantitative measurements on the characterization of the SW excitation efficiencies for this film but also other thicknesses, CPW antenna dimensions and impedance, in addition to determining the loss channels and the energy dissipated through those channels, refer to[16].

Calculated spin-wave dispersion curves and forward transmission spectra

Figure 4. (a) Calculated spin wave dispersion curves for the utilized Yttrium-Iron-Garnet chip for fields ranging from 40 mT up to 89 mT. The antenna’s excitation efficiency is shown on the right-hand side. Maximum efficiency lies around wave-vectors close to 0.7 rad/μ\mum (b) Spectra of the forward transmission (component S21S_{21} of the scattering matrix) as obtained from a vector network analyzer for fields ranging from 40 mT up to 98 mT. The black line traced at the base of the transmission spectra is the electromagnetic cross-talk between the two CPWs at zero field.

Figure 5 (a) is a schematic of the merged system combining the STNO and the SW circuits. The same aforementioned approach regarding RF signal generation is followed for the STNO, however its output is now plugged-in as an input into the magnonic system. Simply put, the STNO hereby and forth replaces the signal generator used to excite SWs (see Figure 3). As for the remainder of the SW circuit, it is preserved as the SWs now excited by the STNO’s input are detected by means of the second CPW antenna. The RF signal is then amplified and passed onto the spectrum analyzer. Thus, the entire detected signal has been transmitted through the SW domain. First, we fix the conditions at the STNO with BSTNO=21B_{\mathrm{STNO}} = 21 mT and ISTNO=3I_{\mathrm{STNO}} = 3 mA (the exact same modes as in Figure 2 (b) are excited). Then, BSWB_{\mathrm{SW}} is swept and the frequency variation versus BSWB_{\mathrm{SW}} is presented in Figure 5 (b). A similar behavior as previously exhibited in Figure 3 (b) is observed even though the signal used to excite the SWs is much broader with a bandwidth of approximately 1 GHz. This means that the SW layer can output narrowband signals regardless of the specificity and the quality of the source used to generate these signals. In other words, it is possible to selectively filter a (deteriorated) input signal just by tuning BSWB_{\mathrm{SW}}. In Figure 5 (c), the filtering of the broad input signal (in black) is clearly presented as a function of BSWB_{\mathrm{SW}}. Filtering is indeed possible for frequencies outside the presented range as well, however those happen to either lie outside the amplification range and/or the accessibility of the electro-magnet at the SW layer. Within the studied range of BSWB_{\mathrm{SW}} in Figure 5 (c), only a few filtered signals are shown for presentation purposes. Setting the appropriate field (BSWB_{\mathrm{SW}}), the desired pass band can be chosen. The selectivity of the filtering (i.e. the bandwidth of the filtered signal) depends on the dispersion relation in the YIG and the wavevector range excited by the CPW, which is about 100 MHz in this case. Additionally, in section I of Supplementary Material, we evidence the independence of this SW filtering on the conditions at the STNO. We show that changing the field solely on the magnonic part (BSWB_{\mathrm{SW}}) is what defines the frequency at which the broad input is filtered. As can be noticed from the filtered spectra in Figure 5 (c), the envelope formed by the maxima of the SW signal does not coincide with those of the STNO input. The reason behind this is impedance mismatch, as impedance on the SW system is frequency dependent, and can thus be further/closer to that of the STNO at different frequencies. This results in a frequency dependent efficiency of power transfer from the source (STNO) into the SW system. At frequencies where the impedance mismatch is larger, a more significant portion of the STNO’s input into the SW system is reflected back, hence leading to weaker transmitted SW signal. The impedance mismatch hence causes a change in the ratio of the transmitted power to the output power (by the STNO), which results in a different trend if one is to consider the S21S_{21} parameter in this case (compared to Figure 4 (b) where a VNA was used).

Proposed concept of the magnonic filter and filtered signals

Figure 5. (a) The proposed concept of the magnonic filter consisting of the source (STNO circuit) and the filtering device (SW circuit). (b) 2D color-plot of the frequency dependence of the SW’s output signal on the applied magnetic field BSWB_{\mathrm{SW}} at fixed STNO conditions: BSW=21B_{\mathrm{SW}} = 21 mT and ISTNO=3I_{\mathrm{STNO}} = 3 mA. (c) Signals filtered by the SW layer for different BSWB_{\mathrm{SW}} values ranging from 43 mT to 70 mT in reference to the broad input signal from STNO (black curve).

Thus, we are able to demonstrate the detection/emission of SWs excited by a STNO, and find means of use for broad output STNOs. In addition, we suggested a direct application to employ the proposed device-concept as an RF source with an integrated magnonic filter that offers both frequency tunability and selectivity. This introduces an approach to RF signal processing where a nanoscaled spintronic oscillator can directly drive a magnonic circuit, opening possibilities for compact, tunable microwave filters in magnon-based computing or sensing applications. Our work constitutes a promising first step in that direction with room for further development. Optimization techniques should address better impedance matching between the STNO and the magnonic system, increasing the output power of the STNOs to avoid the need of amplification. A way to realize this would be by reconfiguring or re-engineering the material stack accordingly, in addition to employing a larger number of coupled oscillators in the form of an array to generate larger output powers, and exhibit a higher resiliency to external conditions[42, 43]. On the other hand, it is also possible to optimize the CPW designs to get a more selective excitation with a narrower linewidth in wave-vector space. Ultimately, the long-term goal is to fabricate the STNOs directly on the SW chip. This eliminates the need for electrical connection between the circuits as SWs are then excited by the dynamic magnetic stray fields generated by the STNOs. The system will use an on-chip STNO as the source for an on-chip magnonic filter. This integration promises benefits for RF systems, such as eliminating the need for a high quality LC filter or external synthesizer in the accessible frequency range, for example. Such a system has been previously investigated via micromagnetic simulations in[32]. Additionally, the STNO system can be designed to operate at zero bias field[28]. For the magnonic system, integrated on-chip permanent magnets can be used to manipulate and tune SW dispersion[16, 44, 45]. In section II of Supplementary Material, we present a qualitative energy-efficient approach to manipulating the filtered signal’s frequency, whereby we shift this frequency by means of a small permanent block magnet suspended above the YIG delay line. The altitude of the latter above the magnetic material influences the frequency at which SWs are emitted/detected via a tunable additional field component.

In conclusion, we demonstrate a hybrid system where a spintronic nano-oscillator (with a broad GHz output) directly excites a magnonic circuit. By tuning the magnetic field on the magnonic side, we achieve selective filtering of the broad input—delivering a narrow-band SW output at a chosen frequency, irrespective of the input’s linewidth. This tunable magnonic filter concept could allow detection and cleanup of deteriorated RF signals in a field-programmable manner, reducing reliance on charge-based components.This

work provides the experimental proof that a STNO can drive a magnonic filter, paving the way for integrated magnonic-spintronic circuits for microwave signal processing.

Supplementary Material

For an additional measurement demonstrating filtering achieved by sweeping ISTNOI_{\mathrm{STNO}} instead of BSWB_{\mathrm{SW}}, and evidencing the importance of BSWB_{\mathrm{SW}} in tuning the filtered frequencies, please refer to the Supplementary Experimental Measurement I.

For an additional measurement demonstrating the tunability of the filtered signal by means of a permanent block magnet, please refer to the Supplementary Experimental Measurement II.

Acknowledgements

This work has been supported by the European Research Council within the Starting Grant No. 101042439 "CoSpiN", the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) - TRR 173 - 268565370" (project B01) and the EU Horizon Europe research and innovation program within the project “MandMEMS” (Grant No. 101070536).

References

  1. [1]K. Filz, Y. Mahmood, and L. Heim, RAND Research Reports (2025).
  2. [2]L. de Roucy-Rochegonde and A. Buffard, Ifri Papers (2025), iSBN: 979-10-373-1000-2.
  3. [3]S. Dutta, S.-C. Chang, N. Kani, D. F. Nikonov, S. Manipatruni, I. A. Young, and A. Naeemi, Scientific reports 5, 9861 (2015).
  4. [4]T. Fischer, M. Kewenig, D. Bozhko, A. Serga, I. Syvorotka, F. Ciubotaru, C. Adelmann, B. Hillebrands, and A. V. Chumak, Applied Physics Letters 110 (2017).
  5. [5]G. Csaba, Á. Papp, and W. Porod, Physics Letters A 381, 1471 (2017).
  6. [6]A. N. Mahmoud, F. Ciubotaru, F. Vanderveken, C. Adelmann, S. Cotofana, and S. Hamdioui, IEEE Transactions on Circuits and Systems II: Express Briefs 71, 919 (2023).
  7. [7]S. Klingler, P. Pirro, T. Brächer, B. Leven, B. Hillebrands, and A. V. Chumak, Applied Physics Letters 106 (2015).
  8. [8]A. V. Chumak, P. Kabos, M. Wu, C. Abert, C. Adelmann, A. O. Adeyeye, J. Åkerman, F. G. Aliev, A. Anane, A. Awad, et al., IEEE Transactions on Magnetics 58, 1 (2022).
  9. [9]R. Nakane, A. Hirose, and G. Tanaka, Physical Review Applied 19, 034047 (2023).
  10. [10]W. Namiki, D. Nishioka, T. Tsuchiya, and K. Terabe, Neuromorphic Computing and Engineering 4, 024015 (2024).
  11. [11]G. Gubbiotti, A. Barman, S. Ladak, C. Bran, D. Grundler, M. Huth, H. Plank, G. Schmidt, S. Van Dijken, R. Streubel, et al., Journal of Physics: Condensed Matter (2024).
  12. [12]K. Vogt, F. Y. Fradin, J. E. Pearson, T. T Sebastian, S. D. Bader, B. Hillebrands, A. Hoffmann, and H. Schultheiss, Nature communications 5, 3727 (2014).
  13. [13]Q. Wang, P. Pirro, R. Verba, A. Slavin, B. Hillebrands, and A. V. Chumak, Science advances 4, e1701517 (2018).
  14. [14]R. Erdélyi, G. Csaba, L. Maucha, F. Kohl, B. Heinz, J. Greil, M. Becherer, P. Pirro, and Á. Papp, Scientific Reports 15, 9806 (2025).
  15. [15]K. Dávidková, K. Levchenko, F. Bruckner, R. Verba, F. Majcen, Q. Wang, M. Lindner, C. Dubs, V. Vlaminck, J. Klíma, et al., Physical Review Applied 23, 034026 (2025).
  16. [16]F. Kohl, B. Heinz, Á. Papp, G. Csaba, and P. Pirro, arXiv preprint arXiv:2505.08656 (2025).arxiv.org/abs/2505.08656
  17. [17]I. Žutić, J. Fabian, and S. D. Sarma, Reviews of modern physics 76, 323 (2004).
  18. [18]A. Slavin and V. Tiberkevich, IEEE Transactions on Magnetics 45, 1875 (2009).
  19. [19]A. Sharma, A. A. Tulapurkar, and B. Muralidharan, Physical Review Applied 8, 064014 (2017).
  20. [20]J. C. Slonczewski, Journal of Magnetism and Magnetic Materials 159, L1 (1996).
  21. [21]L. Berger, Physical Review B 54, 9353 (1996).
  22. [22]M. Tsoi, A. Jansen, J. Bass, W.-C. Chiang, M. Seck, V. Tsoi, and P. Wyder, Physical Review Letters 80, 4281 (1998).
  23. [23]E. Myers, D. Ralph, J. Katine, R. Louie, and R. Buhrman, Science 285, 867 (1999).
  24. [24]M. N. Baibich, J. M. Broto, A. Fert, F. N. Van Dau, F. Petroff, P. Etienne, G. Creuzet, A. Friederich, and J. Chazelas, Physical review letters 61, 2472 (1988).
  25. [25]G. Binasch, P. Grünberg, F. Saurenbach, and W. Zinn, Physical review B 39, 4828 (1989).
  26. [26]M. Julliere, Physics letters A 54, 225 (1975).
  27. [27]T. Wada, T. Yamane, T. Seki, T. Nozaki, Y. Suzuki, H. Kubota, A. Fukushima, S. Yuasa, H. Maehara, Y. Nagamine, et al., Physical Review B—Condensed Matter and Materials Physics 81, 104410 (2010).
  28. [28]A. Hamadeh, A. Koujok, S. Perna, D. R. Rodrigues, A. Riveros, V. Lomakin, G. Finocchio, G. de Loubens, O. Klein, and P. Pirro, IEEE Transactions on Nanotechnology (2024).
  29. [29]A. A. Hamadeh, A. Koujok, D. R. Rodrigues, A. Riveros, V. Lomakin, G. Finocchio, G. De Loubens, O. Klein, and P. Pirro, Communications Physics 8, 85 (2025).
  30. [30]A. M. Deac, A. Fukushima, H. Kubota, H. Maehara, Y. Suzuki, S. Yuasa, Y. Nagamine, K. Tsunekawa, D. D. Djayaprawira, and N. Watanabe, Nature Physics 4, 803 (2008).
  31. [31]A. Dussaux, B. Georges, J. Grollier, V. Cros, A. Khvalkovskiy, A. Fukushima, M. Konoto, H. Kubota, K. Yakushiji, S. Yuasa, et al., Nature communications 1, 1 (2010).
  32. [32]A. Hamadeh, D. Breitbach, M. Ender, A. Koujok, M. Mohseni, F. Kohl, J. Maskill, M. Bechberger, and P. Pirro, Journal of Applied Physics 132 (2022).
  33. [33]N.-T. Phan, N. Prasad, A. Hakam, A. S. E. Valli, L. Anghel, L. Benetti, A. Madhavan, A. S. Jenkins, R. Ferreira, M. D. Stiles, et al., Physical Review Applied 21, 034063 (2024).
  34. [34]J.-V. Kim, in Solid State Physics, Vol. 63 (Elsevier, 2012) pp. 217–294.
  35. [35]S. Urazhdin, V. Demidov, H. Ulrichs, T. Kendziorczyk, T. Kuhn, J. Leuthold, G. Wilde, and S. Demokritov, Nature nanotechnology 9, 509 (2014).
  36. [36]I. Bendjeddou, M. J. Garcia, A. S. El Valli, A. Litvinenko, V. Cros, U. Ebels, A. Jenkins, R. Ferreira, R. Dutta, D. Morche, et al., IEEE Transactions on Microwave Theory and Techniques 71, 2771 (2023).
  37. [37]N. Locatelli, A. Hamadeh, F. Abreu Araujo, A. D. Belanovsky, P. N. Skirdkov, R. Lebrun, V. V. Naletov, K. A. Zvezdin, M. Muñoz, J. Grollier, et al., Scientific reports 5, 17039 (2015).
  38. [38]A. Koujok, A. Riveros, D. R. Rodrigues, G. Finocchio, M. Weiler, A. Hamadeh, and P. Pirro, Applied Physics Letters 123 (2023).
  39. [39]M. Bailleul, D. Olligs, and C. Fermon, Applied Physics Letters 83, 972 (2003).
  40. [40]V. Vlaminck and M. Bailleul, Phys. Rev. B 81, 014425 (2010).
  41. [41]F. Kohl, B. Heinz, Ádam Papp, R. Erdélyi, G. Csaba, and P. Pirro, “Identification and minimization of losses in microscaled spin-wave transducers,” (2025), arXiv:2505.08656 [physics.app-ph].arxiv.org/abs/2505.08656
  42. [42]T. Qu and R. Victora, Scientific reports 5, 11462 (2015).
  43. [43]R. Lebrun, S. Tsunegi, P. Bortolotti, H. Kubota, A. Jenkins, M. Romera, K. Yakushiji, A. Fukushima, J. Grollier, S. Yuasa, et al., Nature communications 8, 15825 (2017).
  44. [44]M. Cocconcelli, S. Tacchi, R. Erdélyi, F. Maspero, A. Del Giacco, A. Plaza, O. Koplak, A. Cattoni, R. Silvani, M. Madami, et al., Physical Review Applied 22, 064063 (2024).
  45. [45]M. Cocconcelli, F. Maspero, A. Micelli, A. Toniato, A. Del Giacco, N. Pellizzi, A. Plaza, A. Cattoni, M. Madami, R. Silvani, et al., Advanced Materials, 2503493 (2025).

Paper details

vlnc:W_krs4qbcm
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