betaFeedback

WorksPhysics and Astronomy

Observation of a dynamic magneto-chiral instability in photoexcited tellurium

Abstract

In a system of charged chiral fermions driven out of equilibrium, an electric current parallel to the magnetic field can generate a dynamic instability by which electromagnetic waves become amplified. Whether a similar instability can occur in chiral solid-state systems remains an open question. Using time-domain terahertz (THz) emission spectroscopy, we detect signatures of what we dub a ``dynamic magneto-chiral instability" in elemental tellurium, a structurally chiral crystal. Upon transient photoexcitation in a moderate external magnetic field, tellurium emits THz radiation consisting of coherent modes that amplify over time. An explanation for this amplification is proposed using a theoretical model based on a dynamic instability of electromagnetic waves interacting with infrared-active oscillators of impurity acceptor states in tellurium to form an amplifying polariton. Our work not only uncovers the presence of a magneto-chiral instability but also highlights its promise for THz-wave amplification in chiral materials.

Magnetic fields can induce anomalous, parity-breaking phenomena in chiral plasmas, i.e., systems with an inequivalent number of left- and right-handed chiral particles. A well-known example of such phenomena is the chiral magnetic effect (CME), where the existence of a chiral anomaly generates an electric current jj parallel to an applied magnetic field B\mathbf{B}, j=σMBj = \sigma_{\mathrm{M}}\mathbf{B} [1]. Here, σM\sigma_{\mathrm{M}} is a pseudoscalar that is odd under inversion and even under time reversal, reflecting the symmetries of the current jj and the magnetic field B\mathbf{B}. The CME is predicted in quark-gluon plasmas and can lead to a dynamic instability that amplifies magnetic fields as a function of time [2–4]. Such a chiral plasma instability has garnered interest across various research fields such as early universe physics [5], supernovas [6], turbulence-driven dynamos in the post-merger phase of binary neutron star coalescence [7], and quark-gluon plasma studies [8]. In solid state systems, the CME is evidenced in transport measurements on Weyl semi-metals [9, 10], and the associated dynamic instability is predicted to manifest as an anomalous reflectance of light from the surface of a time-reversal-breaking Weyl semi-metal prepared in a non-equilibrium steady state [11].

Going beyond the CME in chiral plasmas, we propose that magneto-chiral currents j=σMBj = \sigma_{\mathrm{M}}\mathbf{B} in structurally chiral crystals driven out of equilibrium may also give rise to a dynamic instability. σM\sigma_{\mathrm{M}}, which transforms like a pseudoscalar (or an electric toroidal monopole), is symmetry-allowed in chiral crystals [12]. Although in equilibrium σM\sigma_{\mathrm{M}} must vanish according to band theory [13], a finite-frequency response is still possible [14]. Furthermore, if an excitation generates an out-of-equilibrium unequal occupation of left-handed and right-handed electronic states in a transient state that is sufficiently long-lived, a magneto-chiral current analogous to the CME can arise. Note that the imbalance between left-handed and right-handed electronic states can exist in chiral crystals even without the presence of emergent Weyl fermions at low energies. As we show in this work, in the presence of electromagnetic waves, the out-of-equilibrium crystal with nonzero σM\sigma_{\mathrm{M}} can display a dynamic magneto-chiral instability manifested as an amplified response.

We use time-domain THz emission spectroscopy in a magnetic field to reveal signatures of the proposed magneto-chiral dynamic instability in elemental tellurium, a nonmagnetic structurally-chiral crystal. Tellurium features helical atomic chains along its chiral axis c\mathbf{c} [see Fig. 1(a-b)], allowing nonreciprocal transport [15] and current-induced magnetization [16]. We photoexcite tellurium in an external magnetic field B0\mathbf{B}_0 using an ultrafast near-infrared (NIR) pulse with energy 1.2 eV (greater than the 0.32 eV band-gap of semiconducting tellurium [17]). The photoexcitation generates a transient current and coherent modes that radiate THz waves, the electric field of which we resolve in the time domain using standard electro-optic sampling (Methods).

We find that the radiated THz polarized along B0B_0 contains coherent modes that unexpectedly increase in amplitude over time, consistently across varying magnetic fields and temperatures. We conjecture that this amplification is a signature of a dynamic instability in photoexcited tellurium at frequencies well below the band gap and approaching the hydrodynamic limit for charge carriers. To justify this conjecture, we develop a theoretical model in which the magneto-chiral instability is manifested in the polariton consisting of THz light coupled with an IR-active oscillator associated with a tellurium impurity level to form a polariton, leading to an amplifying, radiative instability that explains the observed THz amplification.

Figure 1a,b show the side view and top view of the experimental geometry, respectively. The helical axis c\mathbf{c} of tellurium lies in the sample plane, and the NIR pulse

is incident at $45^\circ$ to the sample normal n^\hat{\mathbf{n}}. The emitted THz is collected at 90∘90^\circ away from the incident NIR. Figure 1c shows the measured time profiles of the THz electric field (ETHz(t)E_{\mathrm{THz}}(t)) emitted from Te upon photoexcitation at a temperature of T=7T = 7 K and an external magnetic field B0∥cB_0 \parallel c of ±6\pm6 T. Here the incident NIR pump is either ss-polarized (electric field component E~s\widetilde{E}_s along B0B_0) or pp-polarized (electric field component E~p\widetilde{E}_p perpendicular to B0B_0). Note that experiments were performed for two sample orientations: B0∥cB_0 \parallel c (shown in Figure 1) and B0⊥cB_0 \perp c. To study the response that is odd or even with respect to the external magnetic field, we show the B0B_0-antisymmetrized THz emission [(+6(+6 T) data subtracted by (−6(-6 T) data] and the B0B_0-symmetrized [sum of (+6(+6 T)$ and (−6(-6 T) data], together with the 0 T THz emission in Figure 1d. The sharp signal at t=0t = 0 ps lasting for about 1 ps observed in B0B_0-symmetrized and the 0 T THz emission scan can be attributed to a photo-Dember effect from electron-hole mobility differences near the sample surface [18]. Strikingly, the B0B_0-antisymmetrized data shows coherent oscillations that persist for a much longer time (shown for up to 17.5 ps in Figure 1d) than typical THz emission due to just a photo-Dember effect. In contrast note that the 0 T and B0B_0-symmetrized data show no THz emission after about 2 ps (time-window indicated by the gray arrow in Figure 1d). The long-lasting signal in the B0B_0-antisymmetrized data is also clearly evident in the raw THz emission from ss-polarized pump (top panel, Figure 1c). We further note that the B0B_0-antisymmetrized emission at all times scales nearly linearly with the applied external magnetic field B0B_0: scaling the emitted B0B_0-antisymmetrized time profiles with B0B_0 results in all curves nearly overlapping, as shown in Figure 1e.

Figure 1. THz emission from tellurium in an external magnetic field. (b) The side (top) view of the sample geometry. The helical axis cc is oriented parallel to the sample surface, which corresponds to the natural cleavage plane (101ˉ0)(10\bar{1}0) of tellurium. The near-infrared (NIR) pump is incident onto the sample surface at 45∘45^\circ. The THz emission is collected at 90∘90^\circ from the incident NIR pump and is detected as either EsE_s (∥B0\parallel B_0) or EpE_p (⊥B0\perp B_0). The sample surface normal is denoted with a dashed arrow n^\hat{\mathbf{n}}, the rotation around which allows access to B0∥cB_0 \parallel c (shown in a, b) and B0⊥cB_0 \perp c sample orientations. c EsE_s and EpE_p for B0=±6B_0 = \pm6 T at a temperature of 17 K. d The B0B_0-symmetrized and B0B_0-antisymmetrized THz emission, along with the THz emission at 0 T. The gray arrows mark the time window in which the 0 T and B0B_0-symmetrized emission data are almost negligible but the B0B_0-antisymmetrized emission data still show coherent oscillations. e B0B_0-antisymmetrized data re-scaled by dividing the magnetic field amplitude ∣B0∣\lvert B_0\rvert.

Possible reasons for the increase in THz emission with increasing B0B_0 include a radiating Hall current or rotation of a surface dipole radiation by an external magnetic field [19, 20]. These effects will enhance the THz emission only in the Ep(⊥B0)E_p(\perp B_0) direction [19, 20], as the Lorentz force is always perpendicular to B0B_0. For the remainder of the work, we thus focus on the Es(∥B0)E_s(\parallel B_0) anti-symmetrized emission (referred to as SoddS_{\mathrm{odd}} in the remainder of the text), which is not affected by the transport Hall current.

Figure 2a shows the Fourier transform of SoddS_{\mathrm{odd}} for B0∥cB_0 \parallel c. This frequency domain data consists of five main peaks, whose frequencies do not change significantly with magnetic field. Based on this observation, we apply a non-iterative time-domain linear prediction algorithm [21] to the time-domain THz data SoddS_{\mathrm{odd}}. Linear prediction decomposes the data into a sum of harmonic oscillators with exponential factors ∑iAieβitcos⁡(ωit+ϕi)\sum_i A_i e^{\beta_i t}\cos(\omega_i t+\phi_i). This procedure requires no initial fit parameters and uniquely fits noise-free sinusoidal data with exponential envelopes (SI Sec. A), whereas for noisy data, it reliably identifies sinusoidal components above the noise level, as shown in NMR [22] and pump-probe spectroscopy [23]. We find that constraining the fitting algorithm for SoddS_{\mathrm{odd}} to allow only decaying exponentials (β<0\beta<0) does not fit the data well - there are significant residuals with amplitudes increasing with time (SI Fig. S1). In contrast, allowing for growing exponentials (β>0\beta>0) accurately describes both the magnetic field and temperature dependent SoddS_{\mathrm{odd}} data as shown in Figure 2b-Figure 2c. The time-domain linear prediction fits (colored solid lines) closely match the data (empty circles), with negligible residuals (blue solid lines). We performed the same analysis for the SoddS_{\mathrm{odd}} signal measured for B0⊥cB_0 \perp c, and find that allowing β>0\beta>0 is necessary for obtaining fits that accurately describe the data, as shown in Figure 2e-Figure 2f. From the fitting, we extract five distinct sinusoidal modes for both sample orientations. The extracted frequencies of these modes are shown in the right panels of Figure 2a and Figure 2d. As expected from the scaling behavior of the time-domain data (Figure 1e), there is negligible dependence of the mode frequencies on the external magnetic field. We further note two caveats to the extracted modes shown in Figure 2d: the small splitting of the <0.1< 0.1 THz mode is attributed to mode chirp (SI Sec. A); and the amplitude of the highest-frequency mode is close to the noise level. These caveats do not affect the main conclusions of this work.

Figure 2. Temperature and field-dependence of the EsE_s polarized THz emission SoddS_{\mathrm{odd}}. a-c Data for the B0∥cB_0 \parallel c sample orientation: a Fourier transform of the EsE_s emission SoddS_{\mathrm{odd}}, with the ∣B0∣\lvert B_0\rvert dependence of the extracted mode frequencies summarized in the right panel. Modes are color-coded based on their frequency values. b Fit to the temperature-dependent EsE_s emission SoddS_{\mathrm{odd}} at 6 T. Colored circles represent the data, colored lines are the time-domain fits, and blue traces show the residuals. Black lines represent the B0B_0-symmetrized emissions, providing a reference for t=0t=0. c Fit to the ∣B0∣\lvert B_0\rvert-dependent EsE_s emission SoddS_{\mathrm{odd}} at 17 K. d-f Data for the B0⊥cB_0 \perp c sample orientation: d Fourier transform of the EsE_s emission SoddS_{\mathrm{odd}}, and the ∣B0∣\lvert B_0\rvert dependence of the extracted mode frequencies. e Fit to the temperature-dependent EsE_s emission data at 6 T. f Fit to the ∣B0∣\lvert B_0\rvert-dependent EsE_s emission SoddS_{\mathrm{odd}} at 12 K. In a and d left panels, gray bars denote the linear-prediction-extracted mode frequencies averaged over different fields, with the one dashed bar indicating uncertainty due to the mode being near the noise level.

Having developed a reliable method to extract the dynamics of the modes contributing to SoddS_{\mathrm{odd}}, we next determine their possible origin. Since these modes radiate in the far-field, they must have a net infrared-dipole moment and, given that the crystal structure of tellurium breaks inversion symmetry, suggesting the modes may originate from optical phonons. However, the frequencies of the observed modes are much smaller than any of tellurium’s optical phonons [18, 24]. Moreover, the absence of coherent oscillations in a 1030 nm pump - 515 nm transient reflectivity probe measurement under B0∥cB_0 \parallel c (SI Sec. B) indicates that they are not Raman-active. Instead, the energies of the observed modes are close to the energies of acceptor states in tellurium due to impurities (e.g. selenium or group V elements) [25, 26, 27, 28] as measured previously in infrared magneto-transmission experiments. Those results also showed negligible B0B_0-field dependence of the energies [25, 29, 30], consistent with our observations (SI Sec. C). The mode frequencies differ between the two sample orientations (gray bars in Figure 2b-c), likely due to the anisotropic nature of tellurium. We next focus on the modes in SoddS_{\mathrm{odd}} that are described by exponentially growing sinusoids (β>0\beta> 0) and investigate their temperature and field dependence. We isolate the β>0\beta> 0 modes by subtracting the time-domain fits of all the other modes. In the B0∥cB_0 \parallel c measurement geometry, Figure 3a-b show the results for the 0.37 THz β>0\beta> 0 mode. In the B0⊥cB_0 \perp c geometry, Figure 3c-d show the results for the 0.32 THz β>0\beta> 0 mode and Figure 3e-f for the 0.39 THz β>0\beta> 0 mode. The isolated mode trajectories (empty circles) overlap relatively well with the subtracted trajectories (solid lines). The dependence of the extracted β\beta on the applied field and temperature is summarized in Figure 3g-j, which shows that β>0\beta> 0 consistently for these modes except at high temperature.

Figure 3. The amplified modes in the EsE_s THz emission SoddS_{\mathrm{odd}}. a-b EsE_s-polarized SoddS_{\mathrm{odd}}, with B0∥cB_0 \parallel c. a Temperature dependence of the 0.37 THz mode at 6 T. b Same as a, but showing the ∣B0∣\lvert B_0\rvert dependence at 17 K. c-f EsE_s-polarized SoddS_{\mathrm{odd}} with B0⊥cB_0 \perp c. c Temperature dependence of the 0.32 THz mode at 6 T. d Same as c, but showing ∣B0∣\lvert B_0\rvert dependence at 12 K. e Temperature dependence of the 0.39 THz mode for SoddS_{\mathrm{odd}} taken at 6 T. f Same as e, but showing ∣B0∣\lvert B_0\rvert dependence at 12 K. In a-f, colored circles represent the data after subtracting time-domain fit components of all other modes, and colored lines show the fits for the modes using parameters obtained from the linear prediction model. g, h Temperature dependence of the amplification rate β\beta for B0∥cB_0 \parallel c and B0⊥cB_0 \perp c, respectively. Color-coding of the modes for a given sample orientation is consistent with symbols on the upper right of a-f and with the color-coding in Figure 2 a and d. For B0⊥cB_0 \perp c, 0.32 THz (0.39 THz) is denoted with dark green (light green) symbols. i,j ∣B0∣\lvert B_0\rvert dependence of the amplification rate β\beta for B0∥cB_0 \parallel c and B0⊥cB_0 \perp c, respectively. The error bars in panels g-j are obtained from an additional least square fit using the output of the linear prediction model as initial parameters.

We now provide an interpretation of the amplifying components in the SoddS_{\mathrm{odd}} emissions by developing a model based on magneto-chiral currents. In general, the free-electron current density jj can be written as follows, assuming linear response on the Fourier domain of time and space:

j⃗(ω,k)=σM(ω)B⃗(ω,k)+σE(ω)E⃗(ω,k)(1)\vec{j}(\omega,k) = \sigma_{\mathrm{M}}(\omega)\vec{B}(\omega,k) + \sigma_{\mathrm{E}}(\omega)\vec{E}(\omega,k) \tag*{(1)}

where E⃗\vec{E} and B⃗\vec{B} refer to the total electric and magnetic fields, respectively, as a function of the angular frequency ω\omega and wave-vector kk, and σE\sigma_{\mathrm{E}} is the electric conductivity. The pseudoscalar σM\sigma_{\mathrm{M}} describes the frequency-dependent magneto-chiral conductivity. Importantly, σM(ω→0)\sigma_{\mathrm{M}}(\omega\to0) vanishes in a near-equilibrium situation but can be nonzero for a non-equilibrium carrier distribution between electronic states with opposite chirality [14]. Thus, in the k→0k \to0 limit, we assume that σM(ω)\sigma_{\mathrm{M}}(\omega) for a crystal driven out of equilibrium includes both adiabatic and frequency-dependent responses,

σM(ω)=e24π2ℏ2[μDC+μACωτωτ+i],(2)\sigma_{\mathrm{M}}(\omega) = \frac{e^2}{4\pi^2\hbar^2}\left[\mu_{\mathrm{DC}}+\mu_{\mathrm{AC}}\frac{\omega\tau}{\omega\tau+i}\right], \tag*{(2)}

where μAC\mu_{\mathrm{AC}} and μDC\mu_{\mathrm{DC}} are real-valued parameters with dimensions of energy. This form satisfies causality in linear response. The frequency dependence of the μAC\mu_{\mathrm{AC}} term can be derived from the Boltzmann equation in the single relaxation-time approximation [14], for which τ\tau denotes the elastic scattering lifetime of free carriers. We emphasize that small perturbations near equilibrium do not give rise to a finite μDC\mu_{\mathrm{DC}}, which instead represents a far-from-equilibrium effect. By assuming constant μDC\mu_{\mathrm{DC}} and μAC\mu_{\mathrm{AC}}, we implicitly consider the system to be in a long-lived non-equilibrium transient state. This assumption is justified since, in our experiment, photoexcitation of tellurium causes a long-lived non-equilibrium carrier distribution in tellurium’s conduction and valence bands, persisting over the time window shown in Figure 1-Figure 3. Note that beyond this time-window (>30> 30 ps), the THz emission data indicates that σM\sigma_{\mathrm{M}} decays (i.e., μAC\mu_{\mathrm{AC}} and μDC\mu_{\mathrm{DC}} decay) between 20-30 ps (SI Sec. D), consistent with the carrier recombination time [31].

Given that the energies of the observed modes correspond to energies of impurity acceptor states, we model σE\sigma_{\mathrm{E}} as a sum of the DC conductivity σE0\sigma_{\mathrm{E0}} and a contribution from the bound current,

σE(ω)=σE0−iωCω02−ω2−iγω.(3)\sigma_{\mathrm{E}}(\omega) = \sigma_{\mathrm{E0}} - \frac{i\omega C}{\omega_0^2-\omega^2-i\gamma\omega}. \tag*{(3)}

The second term is the response of an infrared (IR)-active oscillator with intrinsic angular frequency ω0\omega_0, damping rate γ\gamma, and C=ne2/mC = ne^2/m where nn is the impurity concentration and mm the oscillator mass [32].

Inserting Eqs. (1)-(3) into Maxwell’s equations and using a plane-wave ansatz for electromagnetic fluctuations, E⃗,B⃗∝ei(k⃗⋅r⃗−ωt)\vec{E},\vec{B} \propto e^{i(\vec{k}\cdot\vec{r}-\omega t)}, we calculate the dispersion relations ω(k)\omega(k) numerically. We use physically reasonable parameters: σE0=50 Ω−1m−1\sigma_{\mathrm{E0}} = 50\ \Omega^{-1}\mathrm{m}^{-1} [33], ω0=1 ps−1\omega_0 = 1\ \mathrm{ps}^{-1} (on the order of the observed mode frequencies), n=1016 cm−3n = 10^{16}\ \mathrm{cm}^{-3}, μDC=0.5 eV\mu_{\mathrm{DC}} = 0.5\ \mathrm{eV}, μAC=−0.5 eV\mu_{\mathrm{AC}} = -0.5\ \mathrm{eV}, τ=1\tau= 1 ps, and the relative permittivity ϵr=30\epsilon_r = 30. Note that μAC\mu_{\mathrm{AC}} and μDC\mu_{\mathrm{DC}} approximate the energy difference between the pump laser photon (1.2 eV) and tellurium’s band gap (0.32 eV). The results are shown in Figure 4a for the real part Re⁡[ω(k)]\operatorname{Re}[\omega(k)] and in Figure 4b for the imaginary part Im⁡[ω(k)]\operatorname{Im}[\omega(k)], with both depending only on k=∣k⃗∣k = |\vec{k}| due to the model isotropy. Two unstable solutions (Im⁡[ω(k)]>0\operatorname{Im}[\omega(k)] > 0, red and blue) are present, implying that these electromagnetic modes grow as eIm⁡[ω(k)]te^{\operatorname{Im}[\omega(k)]t}. We associate these modes to the magneto-chiral dynamical instability. The stable modes (Im⁡[ω(k)]<0\operatorname{Im}[\omega(k)] < 0, gray lines in Figure 4a) are the conventional damped polariton arising due to the coupling between the IR-active mode and the electromagnetic wave [34]. The slight degeneracy lift is due to helicity-breaking σM\sigma_{\mathrm{M}}. One of the two instabilities shown in Figure 4 (red) is non-radiative with vanishing group velocity and Poynting vector. This instability resembles the chiral plasma instability discussed in Ref. [2] and occurs due to μDC≠0\mu_{\mathrm{DC}} \ne0. Adding a Hall current under an external magnetic field enables propagation of this red instability, reproducing the chiral magnetic instability noted in Refs. [35] [11]. However, an unrealistically large value of μDC\mu_{\mathrm{DC}} would be required to observe this red instability in the THz frequency range of the oscillators in Sodd\mathrm{S}_{\mathrm{odd}} (SI Sec. E).

Figure 4. Calculated dispersions of the polariton illustrating the magneto-chiral dynamic instability. a The Re⁡[ω(k)]\operatorname{Re}[\omega(k)] of the plane-wave solutions. Gray: Stable modes (Im⁡[ω(k)]<0\operatorname{Im}[\omega(k)] < 0) with an avoided crossing, similar to the conventional polariton resulting from an IR-oscillator coupling with a photon. Red: A non-propagating (Re⁡[ω(k)]=0\operatorname{Re}[\omega(k)] = 0) instability of the electromagnetic wave. Blue: The amplifying polariton with Re⁡[ω(k)]\operatorname{Re}[\omega(k)] pinned to ω0\omega_0, assuming γ=0\gamma= 0. The Re⁡[ω(k)]\operatorname{Re}[\omega(k)] dispersions are mostly independent of γ\gamma. b The Im⁡[ω(k)]\operatorname{Im}[\omega(k)] of the solutions. Red: Im⁡[ω(k)]\operatorname{Im}[\omega(k)] of the non-radiating instability. Blue: Im⁡[ω(k)]\operatorname{Im}[\omega(k)] of the amplifying polariton assuming different γ\gamma. From dark blue to light blue, γ\gamma increases from 0 to 2.5×10−2 ps−12.5 \times10^{-2}\ \mathrm{ps}^{-1} in increments of 5×10−3 ps−15 \times10^{-3}\ \mathrm{ps}^{-1}.

In contrast, the second magneto-chiral instability (blue in Figure 4) is radiative (Re⁡[ω(k)]≠0\operatorname{Re}[\omega(k)] \ne0) for the physically reasonable parameters noted above. Moreover, the angular frequency is pinned at Re⁡[ω(k)]≈ω0=1 ps−1\operatorname{Re}[\omega(k)] \approx\omega_0 = 1\ \mathrm{ps}^{-1} for kk larger than 25 mm−1^{-1}, which is the value that maximizes Im⁡[ω(k)]\operatorname{Im}[\omega(k)]. This frequency pinning arises from the polaritonic coupling, where the IR-active oscillator gains instability through its interaction with the electron fluid via electromagnetic waves. Such polaritonic coupling via the magneto-chiral instability can be generalized to other IR-active oscillators (e.g. optical phonons). We note that the generation of collective modes with large wave vectors up to k∼100k \sim100 mm−1^{-1} is allowed given the short laser penetration depth (300 nm [36], considerably shorter than the emitted THz wavelength). While we do not track the time evolution of μDC\mu_{\mathrm{DC}} and μAC\mu_{\mathrm{AC}} in the model, it is reasonable to expect that the energy to power the radiative instability due to a finite μAC\mu_{\mathrm{AC}} is supplied by the decay of the non-zero μDC\mu_{\mathrm{DC}}, which represents the system relaxing from a far-out-of-equilibrium state.

As shown in Figure 4b, without damping (γ=0\gamma= 0), the radiative instability’s Im⁡[ω(k)]\operatorname{Im}[\omega(k)] of approximately 10−210^{-2} ps−1^{-1} aligns within an order of magnitude with the observed β\beta. As γ\gamma increases, the amplification rate Im⁡[ω(k)]\operatorname{Im}[\omega(k)] of the instability decreases. γ\gamma is sensitive to disorder, phonon scattering, and local potentials in cyclotron motions, and it typically increases with temperature and decreases with magnetic field under lower fields [37]. Thus, we expect the amplification rate to decrease (increase) with temperature (magnetic field at lower fields), which is indeed observed in the experimentally extracted β\beta, see Figure 3g-Figure 3j. Note that if γ\gamma exceeds 3×10−23 \times10^{-2} ps−1^{-1}, then the radiative instability vanishes (Im⁡[ω(k)]\operatorname{Im}[\omega(k)] is no longer >0> 0), which provides a viable explanation for the experimental observation that only a subset of the IR-active modes are amplified over time (those with smaller γ\gamma).

We next discuss the crystal symmetry requirements for generating a magneto-chiral current with photoexcitation. In a crystal, the coefficient σM\sigma_{\mathrm{M}} is replaced by a rank-2 tensor σM,ij\sigma_{\mathrm{M},ij} that relates the current component jij_i with the magnetic field component BjB_j, as discussed in Ref [14]. The non-zero elements of this tensor for a given crystal structure can be derived via group theory. Using the facts that Te has the chiral space group P3121\mathrm{P}3_{1}21 with corresponding point group D3D_3, and that the out-of-plane (in-plane) components of both the current and the magnetic field transform as the same non-trivial irreducible representation A2A_2 (EE) of D3D_3, it follows that σM,ij\sigma_{\mathrm{M},ij} is a diagonal tensor with σM,xx=σM,yy≠σM,zz\sigma_{\mathrm{M},xx} = \sigma_{\mathrm{M},yy} \ne\sigma_{\mathrm{M},zz} [38]. This reflects the fact that a pseudoscalar transforms as the trivial irreducible representation, as expected for any of the 11 enantiomorphic point groups [39]. However, symmetry alone is insufficient to ensure a finite current, since μDC\mu_{\mathrm{DC}} is only non-zero in a far-from-equilibrium situation where electronic states with opposite chirality have different occupations. In our experiment, we expect that after the initial photoexcitation, charge fluctuations that preserve the crystalline symmetries (i.e., that transform as the A1A_1 irreducible representation) will be the long lived ones [40, 41]. Since the pseudoscalar also transforms as the A1A_1 irreducible representation, the photoexcitation can potentially generate a long-lived non-equilibrium distribution of carriers with a net chirality.

We note that the microscopic origin of μDC\mu_{\mathrm{DC}} and μAC\mu_{\mathrm{AC}} depends on the properties of the underlying band structure, as discussed in [14]. While chiral-imbalanced Weyl nodes in tellurium have been previously reported to lead to a CME [10], which can be described with a μDC\mu_{\mathrm{DC}} term, the Weyl nodes are located a few hundreds of meV away from the middle of the band gap [42, 43, 44, 45] [10]. Future studies of the non-equilibrium carrier distributions (e.g., using time-resolved ARPES experiments combined with DFT calculations) can provide further insights into the microscopic origin of the magneto-chiral dynamic instability in tellurium. Importantly, we emphasize that our analysis relies only on the chiral symmetry of the crystal structure of tellurium, and not on the precise microscopic mechanisms at play. Our findings highlight the potential to manipulate THz-range electromagnetic waves in far-out-of-equilibrium chiral materials, including wave amplification.

Acknowledgments: We thank Enrico Speranza, Rong Zhang, Yingkai Liu, Taylor Hughes, and Thomas P. Devereux for useful discussions. This work was supported by the Quantum Sensing and Quantum Materials, an Energy Frontier Research Center funded by the U.S. Department of Energy (DOE), Office of Science, Basic Energy Sciences (BES), under Award No.DE-SC0021238 (project conception, data acquisition and analysis, and manuscript preparation). F.M., Y.L., A.M., and C.L. acknowledge support from NSF Career Award No. DMR-2144256 for the development of the experimental setup. F.M. acknowledges support from the EPIQS program of the Gordon and Betty Moore Foundation, Grant GBMF11069. Y.H. acknowledges support from the IQUIST (Illinois Quantum Information Science and Technology Center) Postdoctoral Fellowship. P.Z was supported by the Center for Emergent Materials, an NSF MRSEC, under award number DMR-2011876. R.M.F. (theoretical model) was supported by the Air Force Office of Scientific Research under Award No. FA9550-21-1-0423.

Author contributions: Y.H. performed sample characterization, the THz emission experiments, and the corresponding data analysis. Y.L., A.M., Y.H., C.L., and F.M. designed and built the THz emission setup compatible with a magnetic field. D.P. and J.Y.Y. assisted with the THz emission measurements without the echo. N.A., Y.H., P.Z., R.M.F., and J.N. developed the theoretical understanding and modeling of the magneto-chiral instability. E.A.P. performed preliminary synthesis reactions and assisted Y.H. with structural characterization, supervised by D.S.P., D.C., and P.A. provided insights into the impurity levels in tellurium. Y.H., N.A., P.Z., R.M.F., J.N., and F.M. wrote the manuscript with input from all the authors. F.M. conceived and supervised this project.

Competing interests: The authors declare no competing interests.

Methods: THz emission measurements were performed with our custom-built time-domain THz spectroscopy setup based on a Yb:KGW amplifier laser (PHAROS, Light Conversion). The fundamental laser pulse wavelength is 1030 nm with a pulse duration of ∼160\sim160 fs. The fundamental beam is split into a pump and a probe, with the pump incident onto the sample at a 45∘45^\circ angle-of-incidence while the probe is used for electro-optic sampling (EOS) of the emitted THz field. The sample is mounted at the center of a magneto-optic cryostat (OptiCool, Quantum Design) with TPX windows to allow THz transmission. The THz field radiated by the sample is collected and collimated at 90∘90^\circ from the pump beam and at a 45∘45^\circ angle relative to the sample normal, using a THz-transmissive TPX lens mounted inside the cryostat. The emitted THz is then focused onto a (110)-cut CdTe crystal. The EOS probe beam is made to spatially and temporally overlap with the emitted THz on the CdTe crystal. The THz field E(t)\mathbf{E}(t) is thus measured by scanning the time delay of the 1030 nm electro-optic sampling beam relative to the emitted THz field [46].

The detection bandwidth of the CdTe crystal decreases above 1 THz [47]. The experiments were performed with two prototypical sample orientations, B0∥c\mathbf{B}_0 \parallel c and B0⊥c\mathbf{B}_0 \perp c, which can be accessed by rotating the sample. To selectively detect the THz emission component EsE_{\mathrm{s}} parallel to the BB field or the component EpE_{\mathrm{p}} perpendicular to the BB field, we rotated four optical components: the THz wire grid polarizer (WGP) with an extinction ratio 10−310^{-3}, the half-wave plate, and polarizer for the 1030 nm probe, and the CdTe crystal for EO sampling. The WGP has an extinction ratio of better than 10−310^{-3}.

Data availability: The data in this manuscript is available at Illinois Data Bank [48].

Corresponding authors: Correspondence to Yijing Huang ([email protected]) and/or Fahad Mahmood ([email protected]).

References

References

  1. [13]Y. Chen, S. Wu, and A. A. Burkov, Axion response in Weyl semimetals, Phys. Rev. B 88, 125105 (2013).DOI
  2. [14]S. Zhong, J. E. Moore, and I. Souza, Gyrotropic magnetic effect and the magnetic moment on the fermi surface, Phys. Rev. Lett. 116, 077201 (2016).DOI
  3. [15]F. Calavalle, M. Suárez-Rodríguez, B. Martín-García, A. Johansson, D. C. Vaz, H. Yang, I. V. Maznichenko, S. Ostanin, A. Mateo-Alonso, A. Chuvilin, I. Mertig, M. Gobbi, F. Casanova, and L. E. Hueso, Gate-tunable and chirality-dependent charge-to-spin conversion in tellurium nanowires, Nature Materials 21, 526 (2022).DOI
  4. [16]T. Furukawa, Y. Shimokawa, K. Kobayashi, and T. Itou, Observation of current-induced bulk magnetization in elemental tellurium, Nature Communications 8, 954 (2017).DOI
  5. [17]R. S. Caldwell and H. Y. Fan, Optical properties of tellurium and selenium, Phys. Rev. 114, 664 (1959).
  6. [18]T. Dekorsy, H. Auer, H. J. Bakker, H. G. Roskos, and H. Kurz, THz electromagnetic emission by coherent infrared-active phonons, Phys. Rev. B 53, 4005 (1996).pubmed.ncbi.nlm.nih.gov/9983955
  7. [19]M. B. Johnston, D. M. Whittaker, A. Corchia, A. G. Davies, and E. H. Linfield, Simulation of terahertz generation at semiconductor surfaces, Phys. Rev. B 65, 165301 (2002).
  8. [20]J. Shan, C. Weiss, R. Wallenstein, R. Beigang, and T. Heinz, Origin of magnetic field enhancement in the generation of terahertz radiation from semiconductor surfaces, Optics Letters 26, 849 (2001).pubmed.ncbi.nlm.nih.gov/18040471
  9. [21]H. Barkhuijsen, R. de Beer, W. Bovéé, and D. van Ormondt, Retrieval of frequencies, amplitudes, damping factors, and phases from time-domain signals using a linear least-squares procedure, Journal of Magnetic Resonance (1969) 61, 465 (1985).DOI
  10. [22]J. J. Led and H. Gesmar, Application of the linear prediction method to NMR spectroscopy, Chemical Reviews 91, 1413 (1991).DOI
  11. [23]Y. Huang, S. Teitelbaum, S. Yang, G. De la Peña, T. Sato, M. Chollet, D. Zhu, J. L. Niedziela, D. Bansal, A. F. May, A. M. Lindenberg, O. Delaire, M. Trigo, and D. A. Reis, Nonthermal bonding origin of a novel pho toexcited lattice instability in snse, Phys. Rev. Lett. 131, 156902 (2023).
  12. [24]B. Torrie, Raman spectrum of tellurium, Solid State Communications 8, 1899 (1970).DOI
  13. [25]Y. Couder, M. Hulin, and H. Thomé, Cyclotron resonance in tellurium, Phys. Rev. B 7, 4373 (1973).DOI
  14. [26]O. Betbeder-Matibet and M. Hulin, A. Semi-Empirical Model for the Valence Band Structure of Tellurium, Physica Status Solidi (B) 36, 573 (1969).
  15. [27]T. Tani and S. Tanaka, Pressure effect on the impurity state and impurity conduction in tellurium, in The Physics of Selenium and Tellurium: Proceedings of the International Conference on the Physics of Selenium and Tellurium, Königstein, Fed. Rep. of Germany, May 28–31, 1979 (Springer, 1979) pp. 142–152.DOI
  16. [28]K. Natori, T. Ando, M. Tsukada, K. Nakao, and Y. Uemura, The acceptor states in tellurium, Journal of the Physical Society of Japan 34, 1263 (1973).DOI
  17. [29]D. Hardy, C. Rigaux, J. P. Vieren, and N. H. Hau, Impurities and intervalence band magneto-optical transitions in tellurium, Physica Status Solidi (B) 47, 643 (1971).
  18. [30]H. Shinno, R. Yoshizaki, S. Tanaka, T. Doi, and H. Kamimura, Conduction band structure of tellurium, Journal of the Physical Society of Japan 35, 525 (1973).DOI
  19. [31]G. Jnawali, Y. Xiang, S. M. Linser, I. A. Shojaei, R. Wang, G. Qiu, C. Lian, B. M. Wong, W. Wu, P. D. Ye, et al., Ultrafast photoinduced band splitting and carrier dynamics in chiral tellurium nanosheets, Nature Communications 11, 3991 (2020).DOI
  20. [32]M. Dressel and G. Grüner, Electrodynamics of Solids: Optical Properties of Electrons in Matter (Cambridge University Press, Cambridge ; New York, 2002).
  21. [33]V. E. Bottom, The hall effect and electrical resistivity of tellurium, Science 115, 570 (1952).pubmed.ncbi.nlm.nih.gov/17759180
  22. [34]K. Huang, On the interaction between the radiation field and ionic crystals, Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences 208, 352 (1951).
  23. [35]T. Amitani and Y. Nishida, Dynamical chiral magnetic current and instability in weyl semimetals, Phys. Rev. B 107, 014302 (2023).
  24. [36]S. Tutihasi, G. Roberts, R. Keezer, and R. Drews, Optical properties of tellurium in the fundamental absorption region, Physical Review 177, 1143 (1969).
  25. [37]M. Hopkins, R. Nicholas, D. Barnes, M. Brummell, J. Harris, and C. Foxon, Temperature dependence of the cyclotron-resonance linewidth in GaAs-Ga₁₋ₓAlₓAs heterojunctions, Physical Review B 39, 13302 (1989).pubmed.ncbi.nlm.nih.gov/9948231
  26. [38]W.-Y. He and K. T. Law, Magnetoelectric effects in gyrotropic superconductors, Phys. Rev. Res. 2, 012073 (2020).DOI
  27. [39]E. M. Anastassakis, Morphic effects in lattice dynamics, in Dynamical properties of solids, Vol. 4 (Elsevier, 1980) p. 357.DOI
  28. [40]J. J. Li, J. Chen, D. A. Reis, S. Fahy, and R. Merlin, Optical probing of ultrafast electronic decay in Bi and Sb with slow phonons, Phys. Rev. Lett. 110, 047401 (2013).pubmed.ncbi.nlm.nih.gov/25166198
  29. [41]S. M. O’Mahony, F. Murphy-Armando, E. D. Murray, J. D. Querales-Flores, I. Savić, and S. Fahy, Ultrafast relaxation of symmetry-breaking photo-induced atomic forces, Phys. Rev. Lett. 123, 087401 (2019).DOI
  30. [42]G. Chang, B. J. Wieder, F. Schindler, D. S. Sanchez, I. Belopolski, S.-M. Huang, B. Singh, D. Wu, T.-R. Chang, T. Neuper, et al., Topological quantum properties of chiral crystals, Nature materials 17, 978 (2018).DOI
  31. [43]J. Ma, B. Cheng, L. Li, Z. Fan, H. Mu, J. Lai, X. Song, D. Yang, J. Cheng, Z. Wang, C. Zeng, and D. Sun, Unveiling weyl-related optical responses in semiconducting tellurium by mid-infrared circular photogalvanic effect, Nature Communications 13, 5425 (2022).DOI
  32. [44]J. Chen, T. Zhang, J. Wang, L. Xu, Z. Lin, J. Liu, C. Wang, N. Zhang, S. P. Lau, W. Zhang, et al., Topological phase change transistors based on tellurium weyl semiconductor, Science advances 8, eabn3837 (2022).DOI
  33. [45]M. Hirayama, R. Okugawa, S. Ishibashi, S. Murakami, and T. Miyake, Weyl node and spin texture in trigonal tellurium and selenium, Phys. Rev. Lett. 114, 206401 (2015).
  34. [46]P. C. Planken, H.-K. Nienhuys, H. J. Bakker, and T. Wenckebach, Measurement and calculation of the orientation dependence of terahertz pulse detection in ZnTe, JOSA B 18, 313 (2001).DOI
  35. [47]X. Ropagnol, M. Matoba, J. E. Kneck, F. Blanchard, E. Isgandarov, J. Yumoto, and T. Ozaki, Efficient terahertz generation and detection in cadmium telluride using ultrafast ytterbium laser, Applied Physics Letters 117, 181101 (2020).DOI
  36. [48]Y. Huang and M. Farahad, Observation of a magneto-chiral instability in photoexcited tellurium (2025), available from https://doi.org/10.13012/B2IDB-1409842_V1.

Paper details

vlnc:W_9b2wkpqj
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