Introduction

The strong interest in topological Dirac and Weyl semimetals is rooted both in their fundamental attraction as model systems for experimenting with theories of particle physics and in their unique electronic properties, such as the suppressed backscattering, peculiar surface states, chiral and spin-polarized transport, and novel responses to applied electric and magnetic fields controlled by topological invariance, which are promising for technological applications1,2,3,4. Recently, significant attention gained theoretical ideas of how to prepare these phases on demand by photoexcitation and periodic driving by external stimuli (Floquet state engineering)5,6,7,8. Much less explored are the pathways towards experimental realizations of these theoretical proposals. The first strides in this direction are made by the very recent results on WTe2 and MoTe2, where metastable crystal symmetry change leading to a topologically distinct phase was induced by THz light pulses9,10. Here, we discover a transition to a topologically distinct electronic phase which does not rely on the change of the macroscopic symmetry of the crystal lattice and therefore can be photoinduced without an accompanying crystallographic phase transition. This discovery is significant because it provides a pathway towards tuning the band structure topology that is decoupled from complexities, domain size limitations, and relaxation phenomena associated with a bulk phase transition between different crystalline phases11.

Zirconium pentatelluride, ZrTe5, is a remarkable topological material. It was theoretically predicted to be at the phase boundary between a weak topological insulator (TI) and a strong topological insulator12 and initially was proposed as a rare example of 3D Dirac semimetal (DSM)13,14,15. Recent high-resolution ARPES16,17,18 and transport19 measurements suggest that ZrTe5 is a weak topological insulator, but with a very small band gap (~20 meV) at the Γ (q = 0) point and a tiny Fermi surface hosting charge carriers with very small mass14,15. The WTI nature was corroborated by observing the topological edge states at the steps of ZrTe5 crystalline surface20. The bandgap of the insulating state closes under pressure and at 6.2 GPa ZrTe5 becomes a (possibly topological) superconductor21. Recently, evidence for strain-tuned topological phase transition from weak to strong topological insulator with the Dirac semimetal state at the critical point was reported in ZrTe522. Here, we discover that a topological phase transition from TI to DSM can be induced transiently, by ultrafast optical pumping.

An indication that ultrafast optical control of the topological electronic band structure in ZrTe5 is possible has recently been obtained in surface-sensitive time-resolved (tr)-ARPES experiments, suggesting potential for technological applications in ultrafast optoelectronics23. A downshift of the valence energy band was observed following photoexcitation with a sub-picosecond, 1.55 eV photon pulse, which was attributed to transient heating after optical perturbation. The corresponding electronic equilibration times in the valence and conduction bands, 1.6 ps and 0.8 ps, respectively, were consistent with the conventional electron-lattice relaxation24,25. Similarly short, few-ps phonon-mediated relaxation times were observed for hot carrier dynamics in ZrTe5 from time-resolved optical reflectivity using similar optical pumping, with even shorter time (~0.25 ps) for electron-electron thermalization26. Here, we use bulk-sensitive MeV UED to study the temporal dynamics of the photoinduced structural changes in ZrTe5. We discover a transient metastable atomic structure with a markedly longer thermalization time scale (~100–200 ps) consistent with SOC-controlled relaxation27 and deduce the corresponding electronic band structure as the DSM state using DFT calculations.

Results

Experimental procedure and results

ZrTe5 is a layered material, where each layer is formed by quasi-one-dimensional trigonal prismatic ZrTe3 chains with non-equivalent apical (Te1) and dimer (Te2) tellurium atoms running along the crystalline a-axis, which are bound together along the c-axis via pairs of Te (Te3) atoms forming zigzag chains [Fig. 1a]. The prismatic chains and the zigzag Te3 chains form two-dimensional sheets of ZrTe5 in the ac plane, stacking along the b-axis via weak van der Waals interaction. Nevertheless, electronically ZrTe5 is a 3D material, with a closed, albeit strongly anisotropic Fermi surface (\(k_{\mathrm{F}}^c/k_{\mathrm{F}}^a\sim k_{\mathrm{F}}^c/k_{\mathrm{F}}^b\sim 5\), where \(k_{\mathrm{F}} = \left( {k_{\mathrm{F}}^a,k_{\mathrm{F}}^b,k_{\mathrm{F}}^c} \right)\) is Fermi wavevector) and strongly anisotropic dispersion and electron mass3,15,16,17,28. We use the orthorhombic lattice notation [space group Cmcm (#63)] with lattice parameters a = 3.9943 Å, b = 14.547 Å, and c = 13.749 Å at 300 K and two formula units per unit cell12,29 for the steady state of ZrTe5. The atomic Wyckoff positions are shown in Supplementary Table 1.

Fig. 1: Probing photo-induced dynamics of ZrTe5.
figure 1

a Schematics of the crystal structure. Different colors highlight the symmetrically non-equivalent apical (Te1, green), dimer (Te2, blue), and zigzag (Te3, red) tellurium atoms forming the ZrTe3 chains and the ZrTe5 sheets. Chains of ZrTe3 prisms, consisting of Te1 and Te2 atoms extend along the a-axis and are connected by Te3 atoms along the c-axis to form ZrTe5 layers in the a−c plane. These layers are van der Waals bonded in the b-axis direction. Wiggly arrow illustrates selective photoexcitation of Te3 atoms; thick arrows mark the direction of atomic movement after photoexcitation. b UED pattern before photoexcitation. The diffraction is mainly from a single crystal flake with [010] as its plane normal and a smaller twisted flake with [1–10] normal, the latter contributes to the (11L) and (−1−1L) reflection rows as indicated by the horizontal arrows. c Difference pattern between before and after (at an ≈800 ps delay) photoexcitation. d Simulated diffraction pattern based on Bloch-wave method and using our structural refinement shown in Fig. 3 (a = 878, b = 2.5e8, see Methods), which agrees with the structure model from ref. 29. The calculation includes contributions of (H0L) zone with sample thickness of 80 nm and (HHL) zone with sample thickness of 51 nm. The diffraction patterns are represented as color maps with color scales shown in the insets. The intensity is shown in arbitrary units corresponding to the readout of the CCD camera.

Full size image

In our work, thin single crystals of ZrTe5 exfoliated from high quality bulk crystal samples3,28 are excited by 1.55 eV, 60 fs laser pulses with the polarization in the ac crystal plane and probed via diffraction of 100 fs, 4.0 MeV electron pulses. Optical properties of the material in the photon energy region around 1.55 eV are obtained with ellipsometry measurements. The imaginary part, 〈ε2〉, of the pseudo dielectric function, 〈ε〉 = 〈ε1〉 + i〈ε2〉 at 300 K is shown in Supplementary Fig. 1; there is weak anisotropy of 〈ε2〉 in this range. The 1.55 eV energy is close to the inter-band transition, which is centered around 1.3 eV. Inter-band transitions can lead25 to lattice deformation if the intensity of the excitation is sufficiently high. Experiments were performed at temperatures of 55 K and 27 K, with an incident excitation fluence of 3.5 mJ/cm2. No significant difference in dynamical lattice behavior at these two temperatures was observed.

A typical UED diffraction pattern obtained prior to the photoexcitation is shown in Fig. 1b and the difference pattern after and before the photoexcitation in Fig. 1c. Here, the diffraction intensity measured at a large time-delay, t ≈ 800 ps (averaged within the time window [707, 1000] ps), where the lattice dynamics reach quasi-equilibrium, is subtracted from the intensity at a negative time delay, t < 0, before arrival of the pump pulse (averaged within the 160 ps time window). A remarkable change in Bragg reflections is observed, which is most clear for the (00L) peaks with systematic increase in intensity for most of the reflections. The intensity of (00L) Bragg reflections is sensitive to intra-unit-cell atomic displacements, δzv, along the c-axis direction (v indexes atom at position zv and with atomic scattering factor fv). In the frozen-phonon model of a static distortion of the crystal lattice, such displacements lead to an intensity modulation of (00L) peaks, ΔI00L ~ |∑vfv(L·δzv)\(e^{-i(2\pi{L{\cdot}z_v})}\)|230, which is roughly consistent with the observed photoinduced change of the Bragg diffraction pattern. These observations clearly indicate an involvement of the displacements of Te2 and Te3 atoms [Fig. 1a], which reside at low-symmetry 8f Wyckoff positions in the equilibrium ZrTe5 structure29 (Supplementary Table 1), with zTe2 = 0.151 and zTe3 = 0.434, respectively, along the z (c-axis) direction. While the above formula for intensity change induced by displacement obtained in kinematic, single-scattering (Born) approximation is very useful for gaining qualitative understanding of atomic displacements, it can rarely (as in ref. 9), if at all, be used for the quantitative analysis of electron diffraction where the interaction of the scattering electrons with the system is strong and multiple scattering effects cannot be neglected. Here, we quantitatively analyze the lattice structure of ZrTe5 and refine zTe2 and zTe3 positions from the UED patterns before and after photoexcitation using the Bloch-wave dynamical scattering theory, which accurately accounts for multiple scattering effects. The corresponding calculated electron diffraction pattern for the unperturbed equilibrium structure29 is shown in Fig. 1d.

Temporal dynamics of scattering

An important further insight is provided by the quantitative analysis of the temporal dynamics of Bragg peak intensities encoding the photoinduced atomic displacements, which is presented in Fig. 2. At the early stage of the photoexcitation, the dynamics is reflected in intensity transfer from Bragg peaks to thermal diffuse scattering (TDS), Fig. 2a. The decrease of Bragg peak intensities is due to the lattice deformation and the increased atomic vibration as a result of energy transfer from the photoexcited electrons to the lattice. The observed time constant that describes these initial dynamics of both TDS and total Bragg scattering is τ1 ≈ 3 ps, which is consistent with phonon-phonon relaxation and is only slightly slower than the time scale of electron-phonon coupling measured by ARPES23. After the initial fast drop, the total Bragg scattering starts increasing, but on a very much slower time scale, τ2 ≈ 150 ps. This increase is accompanied by a similarly slow further growth of diffuse scattering intensity, which indicates that both signals arise from the same underlying physics.

Fig. 2: Lattice dynamics of ZrTe5 observed from photon pump and electron probe.
figure 2

a Photo-induced dynamics of Bragg scattering averaged over (0 0 12), (0 0 14), (0 0 16), (3 3 5), (3 3 6), (6 0 0), (1 1 13), (1 1 14), (2 0 12), (2 0 16), (4 010) and their equivalent Bragg peaks (black circles) and that of thermal diffuse scattering (TDS) magnified by a factor of five (red circles) at short time delays, τ < 100 ps. b Dynamics of TDS (red circles) and transmitted central beam (blue circles) at time delays up to 1 ns. Solid lines in a, b show fits to two-time exponential relaxation with τ1 = 3 ps and τ2 = 150 ps (see text). c Photo-induced intensity of the (00L) peaks as a function of time delay. Open circles are experimental data and solid lines are fits to two-time relaxation describing the initial (τ1 = 3−6 ps) intensity drop via phonon mechanism, which is concurrent with the increasing TDS, and the long-time increase (τ2 = 160(30) ps) associated with the photo-induced modification of the electronic band structure. d Lattice dynamics of the (odd odd 0) Bragg peaks, which are insensitive to the Te displacements associated with the electronic band reconstruction and only show the phonon-related short-time intensity drop. The inset zooms in at the initial drop of the intensity of these peaks and compares it with that of (0 0 12) peak at short time delays. The error bars in all panels show one standard deviation.

Full size image

In our experiment, the intensity of the transmitted central beam is monitored by a separate detector (see “Methods” for details). This allows us to measure total scattering, which provides an important consistency validation. We notice that the intensity of the central beam has a downward trend within the 1 ns window, which becomes very clear after averaging over the multiple measured time-series, Fig. 2b. The dynamics of the central beam can be described with a single exponent with time constant ≈150 ps, which is consistent with the slow recovery of Bragg peaks and a concomitant increase in thermal diffuse scattering. This indicates that the depletion of the central beam is consequential to the increase of scattering governed by the slow relaxation process.

The individual Bragg peaks also reveal the same two-time-scales dynamics, with an initial fast drop governed by the time constant τ1 of few ps and a slow recovery within 1 ns window with a much longer relaxation time, τ1 ~ 100−200 ps. While the intensity of the majority of Bragg peaks recovers after the initial drop, the level to which it recovers is strongly q-dependent [Fig. 2c, d]. Some peaks, notably those having non-zero component along (00L) direction, including (00L) peaks in Fig. 2c, reach higher intensity than in their initial, unperturbed state, which can be explained by the presence of atomic displacements along z (c-axis) lattice direction in the transient metastable state. On the other hand, (HH0) peaks show significantly different dynamics, with no recovery at all, Fig. 2d. This indicates the absence of photoinduced atomic displacements in this direction, or a possibility that the relevant displacements are scattered in phase, yielding cancellation of their contributions to the observed Bragg intensity.

The two distinct relaxation times, τ1 and τ2, obtained by fitting the intensity of Bragg peaks at different wave vectors, q, to two-time relaxation dynamics, ΔI = A(1 − \(e^{-t/{\tau_1}}\)) + B(1 − \({e^{t - {\tau _2}}}\)), are presented in Fig. 3a. While the short relaxation time of a few ps, τ1, is consistent with the electron-phonon and phonon-phonon scattering mechanisms24,25, the long time, τ2 = 160(30) ps, which is roughly two orders of magnitude larger, indicates that conventional electron-phonon scattering is suppressed and implicates a different mechanism, involving much weaker interactions compared to the Coulomb forces that govern electron and phonon scattering. This is entirely consistent with the involvement of SOC, which is broadly recognized to be important for determining the electronic properties of ZrTe53,17,18. SOC is a relativistic interaction that in atomic systems has a degree of smallness described by the fine structure constant, α ≈ 1/137, i. e. is typically two orders of magnitude weaker compared to Coulomb interactions. SOC-controlled electronic spin relaxation times of the order of hundreds of picoseconds to few nanoseconds are indeed observed in semiconductor quantum wells27,31,32 and in bulk Ge33 and graphene34. The long relaxation times result from the weakness of SOC compared to the Coulomb interaction that governs conventional charge relaxation, combined with the small density of conduction electrons. In the case of Dirac electrons in graphene, as well as in the present case of ZrTe5, the latter is a consequence of the vanishing density of states near the Dirac point35.

Fig. 3: Two distinct time scales of structural response and refinement of Te positions from UED.
figure 3

a Two distinct time constants (τ1 and τ2) as a function of reciprocal vector, q, observed in ZrTe5 in our pump-probe experiments. The thick solid lines show the average values of τ1 = 5.5 ps (green) and τ2 = 160 ps (magenta). The inset illustrates the structural model of ZrTe5 with the arrows indicating the displacement direction of Te3 atoms after photoexcitation. b Measured UED intensity profiles (open circles) of the (00L) reflections before (t < 0, top) and after (t > 700 ps, bottom) photoexcitation. The small peak on the left side of (0 0 6) is a contamination from the (1 1 4) reflection of the small flake and was discarded in the refinement. The error bars show one standard deviation. The solid blue line in the top panel is the calculated intensity profile based on the structure model in Ref. 29 and the DW factor B = 0.46 refined from our data. The red solid line in the bottom panel is calculated based on the same model but with B = 0.52 and zTe3 = 0.432, both refined from the data. We first determine the sample geometry (thickness and bending) by fitting the intensity profile before photoexcitation (top) and then determine the displacement of zTe2 and zTe3 from the intensity profile after-photoexcitation (bottom) based on the goodness of fit (χ2). The (002) reflections, which are at the edge of the beam stopper cannot be reliably measured (the leading edge in both panels is visibly truncated) and are excluded from the refinement. c The value of χ2 quantifying the goodness of fit as a function of the Wyckoff position of Te3 and Te2 along c direction for the intensity profile after-photoexcitation. The green and red circles show χ2 as a function of Te2 position with Te3 fixed at zTe3 = 0.434 (green) and zTe3 = 0.432 (red), respectively. The black circles show χ2 as a function of Te3 position with Te2 fixed at zTe2 = 0.151. The solid lines are the second order polynomial fits used to determine the minimum χ2. The global minimum, χ2 ≈ 1.03, is reached for zTe2 = 0.151 and zTe3 = 0.432. The horizontal dotted line shows the level of χ2 = 1.11, which corresponds to 5% statistical significance test and determines the absolute accuracy of our refinements, ±0.001.

Full size image

Bloch-wave dynamical scattering structure refinement

In order to quantitatively analyze the photoinduced changes of ZrTe5 crystalline lattice and determine the positions of Te atoms in the transient metastable photoexcited state, we conducted a structural refinement by comparing the experimental electron diffraction patterns with the ones calculated using the Bloch-wave dynamical scattering method. We focus on the (00L) reflections, of which sufficient number (up to |L| = 20) are present in our data and which show clear and systematic changes in the difference pattern [Fig. 1c]. As discussed above, this allows us to refine the positions, zTe3 and zTe2, of Te3 and Te2 atoms at the low-symmetry crystallographic 8f Wyckoff site and Debye-Waller (DW) factor, which quantifies the degree of crystallographic disorder and reflects the strength of TDS. We note that displacement of Te1 and Zr atoms along the c-axis alters the crystal symmetry of ZrTe5 structure, which was not observed in our UED experiment.

The results of the refinement are presented in Fig. 3b, which shows the (00L) Bragg intensity profiles measured by UED before (top) and after (bottom) the photoexcitation along with the best fits obtained using Bloch-wave ED calculations and varying the zTe3 and zTe3 atomic positions and DW factor. We used χ2 analysis (see “Methods”) to evaluate the goodness of the fit. We first refine the diffraction pattern measured at the steady state, before the photoexcitation. In addition to the Te3 and Te2 atomic positions, in this refinement we also determine the sample geometry, including sample thickness and minor crystal tilt and bending, and DW factor, B = 0.46. The values of the atomic positions obtained in this refinement, zTe2 = 0.151 and zTe3 = 0.434, are in agreement with the results of ref. 29 (Supplementary Table 1). We then keep the sample geometry unchanged and analyze the UED patterns in the photo-excited metastable state at long-time delays, t ≈ 800 ps. Here, we use an iterative procedure where we first keep Te3 atoms at their steady-state position and vary the Te2 position (green circles in Fig. 3c) to obtain zTe2 = 0.151(1) for the minimum χ2 ≈ 1.24. We then vary the position of Te3 at this fixed position of Te2 [black circles in Fig. 3c] to determine zTe3 = 0.432(1) and B = 0.52, with significantly improved goodness of fit (χ2 ≈ 1.03; for 430 free parameters in the fit, the 5% test of statistical significance corresponds to χ2 ≈ 1.11). The increased value of DW factor is in agreement with the observed increase of TDS in the photoexcited state [Fig. 3a, b]. We then again vary the Te2 atom position using the newly obtained zTe3 = 0.432 [red circles in Fig. 3c] to verify that the global minimum is achieved. We thus conclude that the observed marked changes of the Bragg reflection pattern in the transient photoexcited state induced by a femtosecond laser pulse result from a small but clearly identifiable displacement of the Te3 atoms and a much smaller, barely identifiable displacement of the Te2 atoms, which we were able to refine in our analysis.

Discussion

In order to understand the physics underlying the observed transient structural change, its relation to the electronic properties of ZrTe5 and the role of SOC, we performed the first principles DFT calculations of the electronic band structure, which are presented in Fig. 4. The photoinduced structural modification in ZrTe5 is characterized by the displacement of Te3 atoms in the 8f Wyckoff positions along the c-axis direction, from zTe3 = 0.434 to zTe3 = 0.432, which corresponds to an elongation of the Te3–Te3 bond by ~1% [Fig. 1a and inset of Fig. 3a]. Such a selective response of Te3 atoms to the laser pump agrees well with the electronic density of states (DOS) where the dominant contribution to the DOS at and above the Fermi level is provided by a large peak derived from 5p orbitals of the Te3 atoms [Fig. 4a]. The 1.55 eV incident photons excite electrons from the valence bands to the conduction bands, preferentially populating the Te3 derived states contributing to this peak. The resulting increased Coulomb repulsion pushes Te3 atoms apart, thus weakening the Te3–Te3 bond. Since the DOS at the Fermi level is predominantly contributed by the Te3 atoms’ 5p orbitals, the seemingly small change in the structural parameters could affect the low-energy electronic structure significantly.

Fig. 4: The narrow-gap TI and Dirac semimetal electronic band structure of ZrTe5 before and after photoexcitation calculated by DFT.
figure 4

a The density of states (DOS) at time zero, for Te3 position zTe3 = 0.434. The arrows illustrate the photoexcitation process where electrons from valence bands are excited to a Te3-dominated DOS peak near the Fermi level in the conduction band. b The direct gap at Γ point as a function of zTe3 position for zTe2 = 0.151 (black) and of zTe2 position for zTe3 = 0.434 (red) and zTe3 = 0.433 (green). c The weak topological insulator band structure of ZrTe5 in the equilibrium state at time zero (zTe3 = 0.434); a small direct gap of ≈25 meV at the Γ point is clearly seen. d The band structure in the photoexcited state (zTe3 = 0.434) has gap closed at the Dirac point, revealing a photoinduced phase transition from topological insulator to Dirac semimetal. e, f Zoom in at the band structure near the Γ point (marked by the black rectangles) in c, d, respectively.

Full size image

To explore the evolution of photoexcitation modified band structure of ZrTe5, we performed a series of the first-principles calculations for zTe3 = 0.430−0.437 (see Supplementary Discussion and Supplementary Figs. 2–5 for details). The total energy minimum occurs at zTe3 = 0.434 (Supplementary Fig. 2), in good agreement with the experiment. The corresponding calculated band structure reveals that the equilibrium state of ZrTe5 has a small direct gap of ≈25 meV at the Γ-point and a smaller indirect gap between Γ and M points [Fig. 4c], also in good agreement with the ARPES results. In sharp contrast, at zTe3 = 0.432 both gaps are closed and Dirac point emerges at the Brillouin zone center (Γ point), manifesting a photoinduced Lifshitz-type phase transition from topological insulator to Dirac semimetal [Fig. 4(d),(e)]. Wannier functional analysis of the electronic structure reveals a charge transfer of ~0.6% from Te1 to Te3 atoms accompanying this displacement (such displacement corresponds to the 15–19 meV Ag Raman-active modes of lattice vibrations21). This is consistent with the picture that the 1.55 eV laser illumination weakens the Te3–Te3 bond by selectively moving Te3 atoms apart. The calculated total energy per formula unit for the transient photoexcited state at t ≈ 800 ps (zTe3 = 0.432) is only 12 meV higher than of the equilibrium structure at time zero (zTe3 = 0.434). Remarkably, band structure calculations in the absence of SOC show that the gap is not closed, and the Dirac semimetal state is not formed neither at time zero, nor in the photoexcited structure at t ≈ 800 ps [Supplementary Fig. 4]. Hence, it is the strong SOC in the ZrTe5 crystal that locks electronic spins and momenta, imparting electronic charge quasi-particles with chiral nature3,36 that is fully responsible for the formation of the transient photoinduced DSM state.

Strong SOC in ZrTe5 has wide-ranging experimental consequences. In magnetotransport, SOC is manifested in a very large Lande factor, g ≈ 15.4−24.3, observed from Zeeman splitting13,14,37. A circular dichroic (CD-)ARPES directly reveals the SOC-controlled chiral band structure, where electronic states near the Fermi level are spin-polarized36. This is consistent with the chiral magnetic effect observed in longitudinal magnetotransport3. Chirality protection can also explain the exceptional quasiparticle coherence, which is revealed by record-narrow ARPES spectra17 and remarkably robust quantum transport underpinning a number of unusual and fascinating novel phenomena, such as vanishing quantum oscillations (zero-spin effect)19, 3D quantum Hall effect28, the anomalous Hall effect38, and two-particle resonant states39. The exceptionally long, 160 ± 30 ps equilibration time of the photoexcited transient state in ZrTe5, which DFT calculations show to be DSM observed in our present work, is very naturally explained by a spin-relaxation time of the SOC-governed chiral electronic structure. As we show in Supplementary Figs. 4 and 5, the closure of Dirac gap in ZrTe5 band structure is governed by SOC. Hence, the newly emergent electronic Dirac quasiparticles are chiral (helical)40, with spin-momentum locking that is distinct from that in the static TI phase. The electronic spin re-equilibration that is invariably required for the formation of the transient Dirac state relies on weak magnetic interactions and therefore must involve a correspondingly long SOC controlled spin relaxation time, such as observed in semiconductor quantum wells27,31,32, bulk Ge33, and graphene34, and as we observe here. The electron-phonon relaxation mechanisms, on the other hand, are suppressed due to the vanishing density of states in the Dirac electronic spectrum35.

In conclusion, the simultaneously acquired temporal evolution of Bragg intensity of many reflections enables us to observe the subtle changes in atomic structure of complex electronic materials after pulsed laser excitation11,41,42. Here, we report a photoinduced deformation of the local structure in ZrTe5, which selectively weakens specific Te-Te bond and underlies electronic topological phase transition to a Dirac semimetal. The optical excitation with a femtosecond laser pulse induces metastable transient state whose energy is only slightly above the ground state. The change in atomic positions in the photoinduced state is quantitatively determined from the Bragg intensities of UED through structural refinement using Bloch-wave dynamical scattering theory. The DFT band structure calculations show that in the presence of spin-orbit coupling this photoinduced meta-stable state features chiral Dirac quasiparticles, which emerge at the Γ-point. The equilibration time to the transient Dirac semimetal observed in this work is about two orders of magnitude longer than phonon-mediated thermalization times, consistent with the vanishing density of states in Dirac spectrum and slow spin re-polarization of SOC-governed chiral electronic state. We note that similar considerations might apply to equally long relaxation times of the photoexcited coherent atomic displacements observed in Bi, another well-known topological material with strong SOC43. Our results open exciting new opportunities for photon-induced band engineering of topological DSM states and call for a number of follow-up studies. Further support to our findings can be obtained by time-resolved CD-ARPES experiments23,36, which can directly probe the relaxation times of the chiral band structure. The coherent atomic displacements observed in our work indicate connection to a particular optic phonon mode, thus suggesting a possibility of energy-selective tuning of the band structure in ZrTe5 by populating this mode using THz excitation.

When our manuscript was under review, we became aware of the optical pump-probe work by Vaswani et al.44, which fully corroborates our observations. There, a THz-pump-field induced metastable phase with unique, Raman phonon-assisted topological switching dynamics and with lifetime in excess of 100 ps was observed in ZrTe5. Using first-principles modeling similar to ours, the authors identify a mode-selective Raman coupling driving the system from strong to weak topological insulator with a Dirac semimetal phase established at a critical atomic displacement. In their case, the topological transition is controlled by the phonon coherent pumping, as suggested above.

Methods

Single crystalline samples

Single crystalline ZrTe5 samples were prepared via a Te-flux method, as described in3. High purity Zr and Te elemental mixture Zr0.0025Te0.9975 were sealed under vacuum in a double-walled quartz ampule and first melted at 900 °C in a box furnace and fully rocked to achieve homogeneity for 72 h. The melt was followed by slow cooling and rapid heating treatment between 445 °C and 505 °C for 21 days, in order to re-melt crystals with small sizes. The resultant single crystals were typically ~0.1 × 0.3 × 20 mm3. Crystals were chemically and structurally analyzed by powder X-ray diffraction, scanning electron microscopy with energy dispersive X-ray spectroscopy, and transmission electron microscopy. The samples for UED measurements were exfoliated using Scotch tape, yielding typical lateral crystal size around 50–100 um. The flakes have an average thickness around 100 nm as determined with electron energy loss spectroscopy.

Pump-probe diffraction experiments

Pump-probe diffraction experiments were performed at SLAC using the 4.0 MeV-UED facility with 3.5 mJ/cm2 excitation fluence (1.55 eV pump) at 55 K and 27 K. The exposure time of a single shot for excitation and probing were 60 fs and 100 fs, respectively; 900 single-shot images were integrated at each time step. The time interval between measured points varied from 0.25 ps in the region of fast dynamics to 26 ps is the region of slow dynamics, as reflected by the density of points in Fig. 2. In the experimental setup the center beam was monitored by a separate detector. For quantitative intensity analysis and minimizing the effect of long-term electron count fluctuation due to the possible beam drifts intensities of the diffraction patterns in all the analysis were normalized by the corresponding value of the central beam intensity.

Structure refinement

Structure refinement was carried out by quantitatively comparing the intensities of Bragg reflections before and after photoexcitation with the ones calculated using the Bloch-wave dynamical scattering method. We used \(\chi^{2} = \frac{1}{N}{\sum \nolimits_i} {\left( {\frac{{{\boldsymbol{I}}_i^{{\mathrm{exp}}} - a - b{\boldsymbol{I}}_i^{{\mathrm{cal}}}}}{{\sigma _i}}} \right)^2}\) to evaluate the goodness of fit, where superscript “exp” refers to experimentally measured intensity and “cal” refers to calculated intensity, the subscript i numbers data points in the (00L) scan, and N = 430 is the total number of data points. σi are the standard deviations, which were obtained as mean square deviations from the average for the set of diffraction patterns collected within the respective time window and a and b are fitting parameters corresponding to the background and intensity scaling constant. The calculated intensity accounts for diffraction from a single crystal flake with its plane horizontal and [010] parallel to the incident electron beam and a smaller twisted flake with [1–10] along the incident beam [see Fig. 1c]. The refinement of the diffraction patterns before photoexcitation was used to determine the sample geometry, which included sample thickness, small sample tilt, and bending. From the fitting, the thickness of the [010] single crystal flake was determined to be 80 nm and the thickness of the smaller [1–10] flake was 51 nm. The minimum χ2 obtained in refinement before and after photoexcitation were 1.02 and 1.03, respectively.

First-principles calculations

The first-principles calculations were performed by using the WIEN2K (version 18.2) implementation45 of the full potential linearized augmented plane-wave method in the generalized gradient approximation46 of density functional theory with the SOC treated within the second variational method. The basis size was determined by RmtKmax = 7 and the primitive Brillouin zone was sampled with a regular 24 × 24 × 6 mesh containing 628 irreducible k points to achieve energy convergence of 1 meV. The band structure was plotted in a dense 721 k-point path to show the opening and closing of the small gap at the Brillouin zone center.