betaFeedback

WorksPhysics and Astronomy

Broad-band THz emission by Spin-to-Charge Conversion in Topological Material – Ferromagnet Heterostructures

Abstract

Terahertz spintronic devices combine ultrafast operation with low power consumption, making them strong candidates for next-generation memory technologies. In this study, we use time-domain terahertz emission spectroscopy to investigate spin-to-charge conversion (SCC) in bilayer heterostructures comprising topological insulators (TIs) or Weyl semimetals (WSMs) with ferromagnetic metals (FMs). SCC is studied in TI materials \ce{Bi2Se3}, Pb-doped \ce{Bi2Se3}, and (Bi1−x_{1-x}Sbx_x)2_2Te3_3, and the WSM NbP. Our results reveal that the dependence of SCC on TI thickness varies with interface quality, indicating that thickness dependence alone is not a reliable criterion for distinguishing between inverse spin Hall effect and the inverse Rashba–Edelstein effect mechanisms. We find efficient SCC in TIs depends on both in-situ growth to prevent surface oxidation and proper composition. In NbP∣\vertFM bilayers, we observe THz emission with efficiency and bandwidth comparable to that of TIs, highlighting the broader potential of topological materials. Finally, broadband spectral measurements demonstrate that both TIs and WSMs can generate THz pulses with frequencies extending up to 8,THz. These findings underscore the promise of topological materials as efficient platforms for ultrafast, broadband spintronic applications.

Introduction

Spintronic devices have attracted considerable attention due to their potential for low power consumption and high-speed operation[1, 2]. Central challenges in this field include the generation, manipulation, and detection of spin currents. Among these, the interconversion between spin and charge currents plays a critical role in both the creation and readout of spin signals[3, 4]. Spin–charge interconversion is typically realized via mechanisms such as the spin Hall effect (SHE), which occurs in the bulk of materials with strong spin–orbit coupling (SOC)[5], and the Rashba–Edelstein effect (REE), which emerges at interfaces with broken inversion symmetry[6], along with their reciprocal processes—the inverse spin Hall effect (ISHE) and the inverse Rashba–Edelstein effect (IREE)[4, 7].

The discovery of topological insulators (TIs) has introduced a new class of materials with high spin–charge conversion (SCC) efficiency due to strong SOC[8, 9]. Multiple mechanisms have been proposed to account for SCC in these systems. On the one hand, the topological surface states (TSS), characterized by helical spin textures and spin–momentum locking, can contribute to SCC[10, 11, 12, 13, 14, 15]. On the other hand, topologically trivial bulk states and surface two-dimensional electron gas (2DEG) have also been shown to contribute to SCC[10, 16, 17]. As a result, the relative contributions of surface (REE) and bulk (ISHE) effects remain an active subject of debate. A unifying perspective grounded in bulk–surface correspondence has even suggested that the roles of surface and bulk states cannot be separated and should instead be regarded as equivalent[18].

Experimental investigations of SCC in TI typically employ heterostructures, in which a magnetic layer is placed adjacent to the TI layer. The role of the magnetic layer depends on the measurement technique: it acts as a spin current source in THz emission experiments[9, 19] and spin pumping measurements[12, 14, 16, 17, 19, 20], or as a spin-torque detector in spin-torque ferromagnetic resonance measurements[10, 11, 13, 20]. Among these techniques, THz emission spectroscopy offers a unique advantage as a contact-free, ultrafast probe of SCC[7, 21]. In this approach, an ultrashort laser pulse excites the ferromagnetic metal (FM) layer, generating a spin voltage and hence a spin-polarized current, which is subsequently converted into a transverse charge current in the adjacent TI layer[22, 23]. The resulting sub-picosecond charge current emits a transient electromagnetic pulse in the THz frequency range. The amplitude of this THz signal provides a direct measure of SCC efficiency. Previous THz emission studies have reported that a Bi2Se3(10 QL)∣Co(3 nm)\mathrm{Bi_2Se_3}(10\ \mathrm{QL})|\mathrm{Co}(3\ \mathrm{nm}) bilayer exhibits SCC efficiency comparable to that of the widely used Pt(6 nm)∣Co(3 nm)\mathrm{Pt}(6\ \mathrm{nm})|\mathrm{Co}(3\ \mathrm{nm}) system[9], and have attributed this performance to the contribution of the TSS. However, the influence of the FM layer on the TSS has often been neglected. Theoretical studies suggest that the presence of a metallic over-layer (e.g., Co or Ni) induces charge transfer at the interface, which shifts the TI Fermi level into the conduction band, pushing the TSS away from the Fermi level and may even destroy the helical spin structure[24, 25, 26, 27, 28, 29]. These effects introduce ambiguity in determining whether SCC originates primarily from surface or bulk states in TI|FM heterostructures.

In addition to TIs, another class of topological materials—Weyl semimetals (WSMs)—offers exciting potential for SCC, but remains comparatively less explored. Like TIs, WSMs exhibit band inversion and strong SOC[30, 31, 32]. Their topologically nontrivial band structures feature Weyl points that act as monopoles of Berry curvature in momentum space. This gives rise to a large intrinsic spin Hall effect[33, 34, 35, 36], suggesting that WSMs could serve as efficient platforms for SCC and broadband THz emission.

In this work, we employ all-optical time-domain THz emission spectroscopy[21] to investigate SCC in TI|FM and WSM|FM heterostructures on a sub-picosecond timescale (Figure 1a). We choose amorphous CoFeB as the FM layer due to its high spin polarization[37] and resistance to oxidation[38]. We begin by investigating Bi2_2Se3_3|CoFeB bilayers, identifying two distinct contributions to the THz signal: one from SCC and another from a shift current. By isolating these components, we find that they exhibit different dependencies on TI thickness.

Figure 1. a. Bilayer structure in ferromagnet (FM)|topological material (TM) films. b. Schematic of the THz emission setup.

exhibit different dependencies on TI thickness. Importantly, we show that in-situ growth is crucial for achieving strong THz emission. We further study a series of (Bi1−x_{1-x}Sbx_x)2_2Te3_3|CoFeB heterostructures with different compositions. The THz emission efficiency reaches a minimum when x=0.9x = 0.9, where the chemical potential is close to the charge neutrality point in pristine (Bi,Sb)2_2Te3_3 single layer. A similar suppression is observed in Pb-doped Bi2_2Se3_3|CoFeB samples where bare Pb-doped Bi2_2Se3_3 has insulating bulk. Finally, we demonstrate that a WSM|FM bilayer composed of NbP and permalloy (Py) can also produce strong broadband THz emission. Our spectral analysis confirms that both TI and WSM heterostructures emit radiation up to 8 THz. These results highlight the potential of topological materials for next-generation ultrafast spintronic applications.

Sample preparation and experiment setup

The TI bilayer sample growth proceeds in two stages. First, Bi2_2Se3_3, (Bi,Sb)2_2Te3_3, and Pb-doped Bi2_2Se3_3 are synthesized on 10×1010 \times10 mm2^2 c-plane (0001) sapphire (Al2_2O3_3) substrates via molecular beam epitaxy (MBE). The films are then transferred from the MBE chamber to a sputtering system for the second stage, where a 4 nm amorphous CoFeB layer is deposited on top of the TI film. During the transfer, some TI films are left unprotected and thus exposed to air; we refer to these as ex-situ grown samples. Other TI films are protected by a Se (for Bi2_2Se3_3 and Pb-doped Bi2_2Se3_3) or Te (for (Bi,Sb)2_2Te3_3) capping layer. After transfer, the capping layer is removed by heating in the sputtering system, yielding the in-situ grown samples. The WSM bilayer sample NbP(5 nm)|Py(6 nm) is grown in-situ. The single-crystalline NbP layer is grown on a MgO (100) substrate by MBE[39].

Figure 1b illustrates the THz emission spectroscopy setup used for all measurements except for Figure 5b. An ultrafast laser (center photon energy of 1.55 eV, pulse duration of 35 fs, repetition rate of 1 kHz) is split into pump and probe beams. On the pump side, an optical parametric amplifier (OPA) converts the photon energy to 0.47–1.55 eV, and subsequent difference-frequency generation (DFG) yields photon energies in the range of 0.20–0.48 eV. The pump beam is focused onto the sample at normal incidence with a spot size of approximately 1 mm and an average power of 7 mW. Its polarization lies along the x^\hat{x} direction, while a static magnetic field of 0.02 T is applied along y^\hat{y} using a permanent magnet. The emitted THz wave is collected by one off-axis parabolic mirror (OAP) and then focused by a second OAP onto an electro-optical (EO) crystal. We use 1 mm ZnTe as the EO crystal. The probe beam co-propagates with the THz field into the EO crystal, where the THz electric field induces birefringence, thereby changing the polarization of probe beam. This polarization change is detected in the birefringence measurement using a quarter-wave plate (QWP), a Wollaston prism, and a balanced photodetector. All measurements are performed at room temperature in a dry-air environment with relative humidity maintained below 3% to minimize water absorption. A THz wire-grid polarizer (not shown) is placed in the THz path to select the THz electric field component along the x^\hat{x} direction. By scanning a motorized delay stage in the pump line, we record the time-domain profile of the THz electric field. Another similar setup using 250 μ\mum GaP is used for broadband THz detection for Figure 5b. More details about the broadband setup are included in the Supplementary Material. We also measure the THz transmission of the films and calculate the THz impedance to extract the charge current. More details about the THz transmission setup can be found in our previous paper[34, 40, 41].

Figure 5. a. Time-domain THz waveforms for NbP(5 nm)|Py(6 nm) sample under opposite magnetic field directions. Pump photon energy: 0.8 eV. b. Broadband THz emission spectra of different heterostructures. Pump photon energy: 1.55 eV.

Results

We first present the THz emission results obtained from the in-situ grown Bi₂Se₃(16 QL)|CoFeB(4 nm) sample using 0.8 eV pump. By applying an external magnetic field in the ±y^\hat{y} direction using a permanent magnet, we observe a sign change in the emitted THz signal, confirming its magnetic origin (Figure 2a). We decompose the measured signals into field-symmetric (B-even) and field-antisymmetric (B-odd) components (Figure 2b). The field-antisymmetric term is half of the difference between the two traces in Figure 2a while the field-symmetric term is half of the sum of the two time traces. The B-odd contribution is two orders of magnitude larger

Figure 2. a. THz emission from an in situ grown Bi₂Se₃(16 QL)|CoFeB(4 nm) sample under opposite magnetic field directions. b. Decomposition of the THz emission into field-symmetric (even) and field-antisymmetric (odd) components. The B-even curve is magnified by a factor of 10 for clarity. The dashed line shows the THz emission from a single-layer Bi₂Se₃(16 QL) film. c. Azimuthal angle dependence of the peak THz signal with the external field applied in +B^\hat{B} and −B^\hat{B} directions, as well as in zero-field (0 T). The circles represent experimental data, while the curves are fits using Eqs. (1) and (2). d. SCC and shift-current contributions to the THz signal (top) and extracted charge current (bottom) in Bi₂Se₃(dd QL)|CoFeB(4 nm) samples with varying TI thicknesses (dd = 6, 16, 32, 64 QL). Data points are averaged over two samples grown under identical conditions; error bars indicate the standard deviation. Pump photon energy: 0.8 eV.

two traces in Figure 2a while the field-symmetric term is half of the sum of the two time traces. The B-odd contribution is two orders of magnitude larger than the B-even one, indicating that the magnetism-driven THz emission dominates. We attribute the B-odd component to SCC, and the B-even component to the shift current from the bare Bi₂Se₃ surface[42, 43]. For comparison, we also show the THz emission from a single-layer Bi₂Se₃(16 QL) film in Figure 2b. Since SCC is absent in the single-layer sample, the observed signal arises purely from the surface shift current. Its amplitude is roughly an order of magnitude smaller than the SCC contribution in the bilayer and exhibits a three-fold symmetry imposed by the crystal lattice. This confirms that the 4 nm CoFeB layer serves effectively as the spin current source, even in the absence of a protective capping layer[44].

The coexistence of SCC and shift current also explains the sample azimuthal angle ϕ\phi dependence shown in Figure 2c. Under the external field B⃗=By^\vec{B} = B\hat{y}, the spin current JsJ_s flows along z^\hat{z} with the spin polarization σ⃗=σy^\vec{\sigma} = \sigma\hat{y} (see Figure 1a). The SCC-induced THz emission direction (∝j⃗s×σ⃗∝z^×y^∝x^\propto\vec{j}_s \times\vec{\sigma} \propto\hat{z} \times\hat{y} \propto\hat{x}) is then fixed along x^\hat{x}, and thus exhibits no azimuthal dependence[7]. Meanwhile, the shift current originating from the Bi2_2Se3_3 surface follows a three-fold symmetry that depends on the crystal orientation[42, 43]. Consequently, under ±B\pm B (red and blue circles in Figure 2c), the peak signal S±B(ϕ)S_{\pm B}(\phi) can be described by

S±B(ϕ)=±SSCC+Sshiftsin⁡[3(ϕ−ϕ0)],(1)S_{\pm B}(\phi) = \pm S_{\mathrm{SCC}} + S_{\mathrm{shift}}\sin[3(\phi-\phi_0)], \tag*{(1)}

where SSCCS_{\mathrm{SCC}} is the SCC contribution whose sign changes with BB polarity, SshiftS_{\mathrm{shift}} is the shift current contribution, and ϕ0\phi_0 is an offset between the crystal axes and the laboratory frame. When the external field is removed, the magnetization M⃗\vec{M} and spin current polarization σ⃗\vec{\sigma} rotates with the sample azimuthal angle ϕ\phi,

σ⃗=σ[cos⁡(ϕ−ϕ0′)x^+sin⁡(ϕ−ϕ0′)y^].\vec{\sigma} = \sigma[\cos(\phi-\phi'_0)\hat{x}+\sin(\phi-\phi'_0)\hat{y}].

Following the same analysis above, at 0 T (black circles in Figure 2c), the peak signal S0T(ϕ)S_{0T}(\phi) can be described as

S0T(ϕ)=SSCCsin⁡(ϕ−ϕ0′)+Sshiftsin⁡[3(ϕ−ϕ0)],(2)S_{0T}(\phi) = S_{\mathrm{SCC}}\sin(\phi-\phi'_0)+S_{\mathrm{shift}}\sin[3(\phi-\phi_0)], \tag*{(2)}

where ϕ0′\phi'_0 is an additional offset capturing the remanent magnetization direction. The fitting results are plotted with solid lines in Figure 2c which agrees with the data.

We apply this analysis to Bi2_2Se3_3(dd QL)|CoFeB(4 nm) samples with varying TI thicknesses, as shown in Figure 2d. Here, the signal refers to the measured emitted terahertz electric field, while the current denotes the extracted transverse charge sheet current within the heterostructure, taking into account its impedance[7]. As dd increases from 6 QL to 64 QL, the SCC component exhibits a nonmonotonic trend, peaking around d=16d = 16 QL. In contrast, the shift current shows no clear thickness dependence, consistent with its origin from inversion symmetry breaking at the surface. For thin samples (6 QL and 16 QL), SCC clearly dominates the THz emission; at d=32d = 32 QL, SCC remains approximately twice as large as the shift current contribution; and at d=64d = 64 QL, the two components become comparable. A similar trend is observed for the extracted sheet current, reinforcing this thickness-dependent behavior. Our results differ from previous observations in Bi2_2Se3_3(dd QL)|Co(4 nm) bilayers, where the THz signal decreases monotonically beyond d=6d = 6 QL[9]. The weak thickness dependence of the shift current, in contrast to the pronounced nonmonotonicity of SCC, suggests that the SCC in our samples is not dominated by a surface-driven mechanism.

To investigate the impact of surface condition on THz emission, we prepared ex-situ grown Bi2_2Se3_3(16 QL)|CoFeB(4 nm) films for comparison. The key difference between ex-situ and in-situ grown samples lies in whether the TI surface was exposed to air prior to capping, and thus potentially oxidized. While in-situ grown samples consistently exhibit high THz emission efficiency, their ex-situ counterparts display significantly reduced signal amplitudes, as illustrated in Figure 3a for representative Bi2_2Se3_3(16 QL)|CoFeB(4 nm) samples. Notably, the two ex-situ samples were grown in separate batches and exhibit large variations in

Figure 3. a. Time-domain THz waveforms for two in situ and two ex situ Bi2_2Se3_3(16 QL)|CoFeB(4 nm) films and one Pt(16 nm)|CoFeB(4 nm) reference. b. SCC and shift-current contributions to the THz signal (top) and extracted current (bottom), obtained from the azimuthal-angle dependence in ex situ grown Bi2_2Se3_3(dd QL)|CoFeB(4 nm) samples with varying TI thickness dd. The data point at d=16d = 16 QL is averaged over the two samples shown in a; the corresponding error bar indicates their standard deviation. Other thicknesses are represented by a single sample and thus do not include error bars. Pump photon energy: 0.8 eV.

signal strength, likely due to different durations of air exposure. Measurements of THz transmission and impedance confirm that both ex-situ grown samples are comparable in out-coupling current to THz emission (see Supplementary Material, Fig. S1), ruling out these factors as the cause of the observed discrepancy in emission. Although topological surface states are theoretically protected from oxidation[45, 46, 47, 48], the suppressed emission observed in ex-situ grown samples may result from reduced spin-current injection efficiency across the oxidized FM/TI interface. As shown in Figure 3b, this variability introduces significant uncertainty and makes it difficult to extract consistent thickness-dependent trends[9]. In conclusion, our results demonstrate that the observed thickness dependence of THz emission is highly sensitive to sample growth conditions. This sensitivity compromises the reliability of using thickness-dependent measurements alone to distinguish between ISHE and IREE mechanisms in Bi2Se3∣CoFeB\mathrm{Bi}_{2}\mathrm{Se}_{3}|\mathrm{CoFeB} heterostructures⁹.

Next, we present the THz emission results obtained from in-situ grown (Bi1−xSbx)2Te3(16 QL)∣CoFeB(4 nm)(\mathrm{Bi}_{1-x}\mathrm{Sb}_{x})_{2}\mathrm{Te}_{3}(16\ \mathrm{QL})|\mathrm{CoFeB}(4\ \mathrm{nm}) samples. In the absence of the FM layer, the chemical potential in (Bi1−xSbx)2Te3(\mathrm{Bi}_{1-x}\mathrm{Sb}_{x})_{2}\mathrm{Te}_{3} can be tuned via Sb doping, approaching the Dirac point near x≈0.9x \approx0.9⁴⁹–⁵³. Figure 4a shows the time-domain THz emission from samples with x=0x = 0, 0.50.5, 0.750.75, 0.90.9, and 11 under 1.55 eV1.55\ \mathrm{eV} pump photon energy. We find that the THz emission amplitude reaches a minimum at x≈0.9x \approx0.9, consistent with the charge neutrality point of the bare TI layer. This composition dependence is summarized in Figure 4b. Notably, the stoichiometric Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3} sample exhibits THz emission efficiency comparable to that of Bi2Se3\mathrm{Bi}_{2}\mathrm{Se}_{3} (see Figure 2). These results suggest that the SCC efficiency is strongly composition dependent. However, the precise correlation with the chemical potential remains unclear, as the presence of the FM layer may shift the Fermi level due to interfacial charge transfer²⁴–²⁹. Whether the observed minimum in THz emission at x≈0.9x \approx0.9 coincides with the intrinsic charge neutrality point of the TI layer or is merely coincidental requires further investigation. Repeating the measurement under a lower pump photon energy of 0.2 eV0.2\ \mathrm{eV} yields the same trend (data not shown), suggesting the robustness of the observed composition dependence.

Figure 4. a. Time-domain THz waveforms for (Bi1−xSbx)2Te3(16 QL)∣CoFeB(4 nm)(\mathrm{Bi}_{1-x}\mathrm{Sb}_{x})_{2}\mathrm{Te}_{3}(16\ \mathrm{QL})|\mathrm{CoFeB}(4\ \mathrm{nm}). b. Sb composition (xx) dependence of the THz emission from (Bi1−xSbx)2Te3(16 QL)∣CoFeB(4 nm)(\mathrm{Bi}_{1-x}\mathrm{Sb}_{x})_{2}\mathrm{Te}_{3}(16\ \mathrm{QL})|\mathrm{CoFeB}(4\ \mathrm{nm}). c. Time-domain THz waveforms for Bi2Se3(16 QL)∣CoFeB(4 nm)\mathrm{Bi}_{2}\mathrm{Se}_{3}(16\ \mathrm{QL})|\mathrm{CoFeB}(4\ \mathrm{nm}) and Pb-Bi2Se3(16 QL)∣CoFeB(4 nm)\mathrm{Bi}_{2}\mathrm{Se}_{3}(16\ \mathrm{QL})|\mathrm{CoFeB}(4\ \mathrm{nm}). Pump photon energy: 1.55 eV1.55\ \mathrm{eV}.

In Figure 4c, we compare Bi2Se3(16 QL)∣CoFeB(4 nm)\mathrm{Bi}_{2}\mathrm{Se}_{3}(16\ \mathrm{QL})|\mathrm{CoFeB}(4\ \mathrm{nm}) and Pb-doped Bi2Se3(16 QL)∣CoFeB(4 nm)\mathrm{Bi}_{2}\mathrm{Se}_{3}(16\ \mathrm{QL})|\mathrm{CoFeB}(4\ \mathrm{nm}). Without the FM layer, as-grown Bi2Se3\mathrm{Bi}_{2}\mathrm{Se}_{3} is bulk-conducting, while Pb doping makes it bulk-insulating. Similar to the (Bi,Sb)2Te3(\mathrm{Bi},\mathrm{Sb})_{2}\mathrm{Te}_{3} case, we observe a significant reduction in THz emission from the Pb-doped sample, indicating suppressed SCC efficiency. Additional mechanisms, such as the surface warping effect⁵⁰,⁵⁴,⁵⁵, and interfacial Rashba spin splitting⁵⁶–⁵⁹ should also be considered for further investigation.

So far, we have focused on the TI|FM structure. We finally explore the potential of SCC in WSM³⁰–³² using NbP. NbP is a type-I WSM, known to exhibit Weyl nodes and associated Fermi arcs in both bulk crystals and thin films, as confirmed by experimental studies³⁹,⁶⁰–⁶². We measure a bilayer composed of 5 nm5\ \mathrm{nm} NbP and 6 nm6\ \mathrm{nm} permalloy (Py). The time-domain THz signals under opposite magnetic fields are displayed in Figure 5a. We find that the THz emission efficiency of this WSM|FM bilayer is about one-fifth of that in in-situ grown Bi2Se3(16 QL)∣CoFeB(4 nm)\mathrm{Bi}_{2}\mathrm{Se}_{3}(16\ \mathrm{QL})|\mathrm{CoFeB}(4\ \mathrm{nm}). In Figure 5b, we present the THz electric field spectra obtained by performing a Fourier transform (FT) on the time-domain signals, taking into account the response function of a 250 μm250\ \mu\mathrm{m}-thick GaP EO crystal⁶³,⁶⁴. We compare the spectra of the TI|FM samples Bi2Se3(16 QL)∣CoFeB(4\mathrm{Bi}_{2}\mathrm{Se}_{3}(16\ \mathrm{QL})|\mathrm{CoFeB}(4 nm), (Bi0.1Sb0.9)2Te3(16 QL)∣CoFeB(4 nm)(\mathrm{Bi}_{0.1}\mathrm{Sb}_{0.9})_{2}\mathrm{Te}_{3}(16\ \mathrm{QL})|\mathrm{CoFeB}(4\ \mathrm{nm}), and the WSM|FM sample NbP(5 nm)|Py(6 nm). We find that the majority of the spectral weight lies between 0.2 THz and 8.3 THz, with a cutoff around 8 THz set by GaP absorption. These broadband THz emission spectra of topological materials including TI and WSM are comparable to that from established spin Hall emitters. This underscores the potential of topological materials for efficient spintronic THz generation. We further emphasize the critical role of the metallic FM layer. It serves not only as the source of the spin-polarized current which boosts the THz emission, but also modifies the surface of the topological material, which may suppress SCC efficiency. Further studies are required to elucidate the interfacial mechanisms and optimize heterostructure design for maximized THz performance.

Summary

We have conducted a comprehensive study of THz emission in topological material|ferromagnetic metal (TM|FM) bilayer thin films. Our findings highlight the critical role of the TI layer condition without an exposure to the atmosphere in achieving high THz emission efficiency. In particular, preserving the TI surface from oxidation and optimizing its composition both significantly enhance the THz generation from spin to charger conversion. We further observe that the dependence on TI thickness varies with interface quality, indicating that thickness alone is not a reliable criterion for distinguishing between ISHE and IREE mechanisms. It is likely that both processes contribute to SCC in these heterostructures. Moreover, we demonstrate that WSM|FM bilayers can also facilitate efficient SCC, though the underlying mechanism requires further investigation. Finally, our broadband spectral measurements reveal that both TIs and WSMs can generate THz pulses with spectral bandwidths over 8 THz, underscoring the promise of topological materials for future ultrafast spintronic applications.

Data and materials availability: All data needed to evaluate the conclusions are present in the paper. Additional data related to this paper may be requested from the authors.

Acknowledgments X.H. and L.W. acknowledge the support by Air Force Office of Scientific Research under award No. FA955022-1-0410. X.Y. and S.O. were sponsored by the Army Research Office and was accomplished under Grant Number W911NF-20-2-0166. X.H. acknowledges the Gordon and Betty Moore Foundation EPiQS Initiative through QuantEmX Travel Grant No. GBMF5305 to perform experiments in Freie Universität Berlin. The development of the THz emission spectroscopy was partially sponsored by the Army Research Office and was accomplished under Grant Number W911NF-25-2-0016. GB and TK acknowledge funding by the DFG Collaborative Research Center SFB TRR 227 “Ultrafast spin dynamics” (project ID 328545488, projects A05 and B02) and the ERC-2023 Advanced Grant ORBITERA (grant no. 101142285). A.B.-P. acknowledges support by the Generalitat Valenciana (CIDEGENT/2021/005).

Author Contributions: L.W. conceived the project. X.H., Z.N. and L.W. built the THz emission setup. X.H. performed most of the THz emission experiments and analyzed the data with L.W. G.B. and C.I. performed the broadband THz emission experiments under the supervision of T.K.. X.Y. and S.O. grew the TI films. T.R.T. and S.H. grew the FM films. A.B.-P. and S.S.P.P. grew the NbP bilayer film.

Competing interests: The authors declare that they have no competing interests.

References

References

  1. [1]1 I. Žutić, J. Fabian, and S. D. Sarma, Reviews of modern physics 76, 323 (2004).
  2. [2]2 B. Dieny, I. L. Prejbeanu, K. Garello, P. Gambardella, P. Freitas, R. Lehndorff, W. Raberg, U. Ebels, S. O. Demokritov, J. Akerman, et al., Nature Electronics 3, 446 (2020).
  3. [3]3 W. Han, Y. Otani, and S. Maekawa, npj Quantum Materials 3, 27 (2018).
  4. [4]4 J. R. Sánchez, L. Vila, G. Desfonds, S. Gambarelli, J. Atjané, J. De Teresa, C. Magén, and A. Fert, Nature communications 4, 2944 (2013).
  5. [5]5 J. Sinova, S. O. Valenzuela, J. Wunderlich, C. Back, and T. Jungwirth, Reviews of modern physics 87, 1213 (2015).
  6. [6]6 V. M. Edelstein, Solid State Communications 73, 233 (1990).
  7. [7]7 T. Seifert, S. Jaiswal, U. Martens, J. Hannegan, L. Braun, P. Maldonado, F. Freimuth, A. Kronenberg, J. Henrizi, I. Radu, et al., Nature photonics 10, 483 (2016).
  8. [8]8 L.-G. Zhu, B. Kuberak, K. Fai Mak, and J. Shan, Scientific reports 5, 10308 (2015).
  9. [9]9 X. Wang, L. Cheng, D. Zhu, Y. Wu, M. Chen, Y. Wang, D. Zhao, C. B. Boothroyd, Y. M. Lam, J.-X. Zhu, et al., Advanced Materials 30, 1802356 (2018).
  10. [10]10 Y. Wang, D. Zhu, Y. Wu, Y. Yang, J. Yu, R. Ramaswamy, R. Mishra, S. Shi, M. Elyasi, K.-L. Teo, et al., Nature communications 8, 1364 (2017).
  11. [11]11 M. A. Noyan, X. Zhang, J. Moon, E. Cobas, M. Lohmann, Q. Zou, L. Li, M. Weinert, B. T. Jonker, O. M. van ‘t Erve, et al., Nano Letters (2025).
  12. [12]12 Y. Shiomi, K. Nomura, Y. Kajiwara, K. Eto, M. Novak, K. Segawa, Y. Ando, and E. Saitoh, Phys. Rev. Lett. 113, 196601 (2014).
  13. [13]13 A. R. Melnik, J. Lee, A. Richardella, J. L. Grab, P. J. Mintun, M. H. Fischer, A. Vaezi, A. Manchon, E.-A. Kim, N. Samarth, et al., Nature 511, 449 (2014).
  14. [14]14 H. Wang, J. Kally, J. S. Lee, T. Liu, H. Chang, D. R. Hickey, K. A. Mkhoyan, M. Wu, A. Richardella, and N. Samarth, Phys. Rev. Lett. 117, 076601 (2016).
  15. [15]15 J. S. Lee, A. Richardella, D. R. Hickey, K. A. Mkhoyan, and N. Samarth, Physical Review B 92, 155312 (2015).
  16. [16]16 P. Deorani, J. Son, K. Banerjee, N. Koirala, M. Brahlek, S. Oh, and H. Yang, Physical Review B 90, 094403 (2014).
  17. [17]17 M. Jamali, J. S. Lee, J. S. Jeong, F. Mahfouzi, Y. Lv, Z. Zhao, B. K. Nikollic, K. A. Mkhoyan, N. Samarth, and J.-P. Wang, Nano letters 15, 7126 (2015).
  18. [18]18 H. Wang, J. Kally, C. Sahin, T. Liu, W. Yanez, E. J. Kamp, A. Richardella, M. Wu, M. E. Flatté, and N. Samarth, Physical review research 1, 012014 (2019).
  19. [19]19 P. Sahu, Y. Yang, Y. Fan, H. Jaffrès, J.-Y. Chen, X. Devaux, Y. Fagot-Revurat, S. Migot, E. Rongione, T. Chen, et al., ACS Applied Materials & Interfaces 15, 38592 (2023).
  20. [20]20 S. Shi, E. Liu, F. Hu, G. Shi, A. Manchon, and H. Yang, Physical Review B 111, 094433 (2025).
  21. [21]21 Z. Ni, K. Wang, Y. Zhang, O. Pozo, B. Xu, X. Han, K. Manna, J. Paglione, C. Felser, A. G. Grushin, et al., Nature communications 12, 1 (2021).
  22. [22]22 R. Rouzegar, L. Brandt, L. Nävdorník, D. A. Reiss, A. L. Chekhov, O. Gueckstock, C. In, M. Wolf, T. S. Seifert, P. W. Brouwer, et al., Physical Review B 106, 144427 (2022).
  23. [23]23 E. Beaurepaire, G. Turner, S. Harrel, M. Beard, J.-Y. Bigot, and C. Schmuttenmaer, Applied Physics Letters 84, 3465 (2004).
  24. [24]24 J. Zhang, J. P. Velev, X. Dang, and E. Y. Tsymbal, Physical Review B 94, 014435 (2016).
  25. [25]25 D. Culcer, E. H. Hwang, T. D. Stanescu, and S. Das Sarma, Phys. Rev. B 82, 155457 (2010).
  26. [26]26 S. V. Eremeev, V. N. Men’shov, V. V. Tugushev, P. M. Echenique, and E. V. Chulkov, Phys. Rev. B 88, 144430 (2013).
  27. [27]27 W. Luo and X.-L. Qi, Phys. Rev. B 87, 085431 (2013).
  28. [28]28 Y.-T. Hsu, K. Park, and E.-A. Kim, Phys. Rev. B 96, 235433 (2017).
  29. [29]29 C. D. Spataru and F. Léonard, Physical Review B 90, 085115 (2014).
  30. [30]30 B. Yan and C. Felser, Annual Review of Condensed Matter Physics 8, 337 (2017).
  31. [31]31 X. Wan, A. M. Turner, A. Vishwanath, and S. Y. Savrasov, Physical Review B—Condensed Matter and Materials Physics 83, 205101 (2011).
  32. [32]32 N. Armitage, E. Mele, and A. Vishwanath, Reviews of Modern Physics 90, 015001 (2018).
  33. [33]33 Y. Sun, Y. Zhang, C. Felser, and B. Yan, Physical Review Letters 117, 146403 (2016).
  34. [34]34 X. Han, A. Markou, J. Stensberg, Y. Sun, C. Felser, and L. Wu, Physical Review B 105, 174406 (2022).
  35. [35]35 K. Manna, L. Muechler, T.-H. Kao, R. Stinshoff, Y. Zhang, J. Gooth, N. Kumar, G. Kreiner, K. Koepernik, R. Car, J. Kübler, G. H. Fecher, C. Shekhar, Y. Sun, and C. Felser, Phys. Rev. X 8, 041045 (2018).
  36. [36]36 E. Liu, Y. Sun, N. Kumar, L. Muechler, A. Sun, L. Jiao, S.-Y. Yang, D. Liu, A. Liang, Q. Xu, et al., Nature physics 14, 1125 (2018).
  37. [37]37 S. Huang, T. Chen, and C. Chien, Applied Physics Letters 92 (2008).
  38. [38]38 S. Jen, Y. Yao, Y. Chen, J. Wu, C. Lee, T. Tsai, and Y. Chang, Journal of applied physics 99 (2006).
  39. [39]39 A. Bedoya-Pinto, A. K. Pandeya, D. Liu, H. Deniz, K. Chang, H. Tan, H. Jan, J. Kostaonsky, and S. S. Parkin, ACS nano 14, 4405 (2020).
  40. [40]40 X. Han, H. T. Yi, S. Oh, and L. Wu, Nano Letters 24, 914 (2024).
  41. [41]41 J. Stensberg, X. Han, S. Lee, S. A. McGill, J. Paglione, I. Takeuchi, C. L. Kane, and L. Wu, Physical Review Letters 130, 096901 (2023).
  42. [42]42 L. Braun, G. Mussler, A. Hruban, M. Konczykowski, T. Schumann, M. Wolf, M. Münzenberg, L. Perfetti, and T. Kampfrath, Nature communications 7, 1 (2016).
  43. [43]43 J. Stensberg, X. Han, Z. Ni, X. Yao, X. Yuan, D. Mallick, A. Gandhi, S. Oh, and L. Wu, Phys. Rev. B 109, 245112 (2024).
  44. [44]44 S. Jen, Y. Yao, Y. Chen, J. Wu, C. Lee, T. Tsai, and Y. Chang, Journal of applied physics 99 (2006).
  45. [45]45 X. Wang, G. Bian, T. Miller, and T.-C. Chiang, Phys. Rev. Lett. 108, 096404 (2012).
  46. [46]46 L. V. Yashina, J. Sánchez-Barriga, M. R. Scholz, A. A. Volykhov, A. P. Sirotina, S. Neudachina, Vera, M. E. Tamm, A. Varykhalov, D. Marchenko, G. Springholz, et al., Acs Nano 7, 5181 (2013).
  47. [47]47 J. Yang, B. Zheng, Z. Chen, W. Xu, R. Wang, and H. Xu, The Journal of Physical Chemistry C 124, 6253 (2020).
  48. [48]A. J. Green, S. Dey, Y. Q. An, B. O’Brien, S. O’Mullane, B. Thiel, and A. C. Diebold, Journal of Vacuum Science & Technology A 34 (2016).

Paper details

vlnc:W_fvu8mqzk
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