Appearance of T ∗ d phase across the T d -1T ′ phase boundary in Weyl semimetal MoTe 2
Yu Tao, John A. Schneeloch, Chunruo Duan, Masaaki Matsuda, Sachith E. Dissanayake, Adam A. Aczel, Jaime A. Fernandez-Baca, Feng Ye, Despina Louca
AAppearance of T ∗ d phase across the T d –1T (cid:48) phase boundary in Weyl semimetal MoTe Yu Tao, John A. Schneeloch, Chunruo Duan, Masaaki Matsuda, Sachith E. Dissanayake, ∗ Adam A. Aczel,
2, 3
Jaime A. Fernandez-Baca, Feng Ye, and Despina Louca † Department of Physics, University of Virginia, Charlottesville, Virginia 22904, USA Neutron Scattering Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, USA Department of Physics and Astronomy, University of Tennessee, Knoxville, Tennessee 37996, USA
Using elastic neutron scattering on single crystals of MoTe and Mo − x W x Te ( x < ∼ . ∗ d phase, present between the low temperatureorthorhombic T d phase and high temperature monoclinic 1T (cid:48) phase, is explored. The T ∗ d phaseappears only on warming from T d and is observed in the hysteresis region prior to the 1T (cid:48) transition.This phase consists of four layers in its unit cell, and is constructed by an “AABB” sequence oflayer stacking operations rather than the “AB” and “AA” sequences of the 1T (cid:48) and T d phases,respectively. Though the T ∗ d phase emerges without disorder on warming from T d , on cooling from1T (cid:48) diffuse scattering is observed that suggests a frustrated tendency toward the “AABB” stacking. Many layered materials have structure-property rela-tionships that depend on their layer stacking. For ex-ample, the transition metal dichalcogenide MoTe is re-ported to be a type-II Weyl semimetal in its orthorhom-bic T d phase (with the non-centrosymmetric P nm space group) [1, 2], but not in its monoclinic 1T (cid:48) phase(with the centrosymmetric space group P /m ). The twophases have nearly-identical layers and differ mainly byin-plane displacements. Though there is much interest ininvestigating Weyl semimetals, the properties of MoTe are not completely understood. For instance, there ismuch debate on the origin of the extreme magnetoresis-tance observed at low temperatures [3–5], the numberand location of Weyl points in the T d phase [6], and thetopological nature of the observed surface Fermi arcs thatare a necessary but not sufficient condition for a Weylsemimetal [7]. Structural distortions have been known tooccur, such as stacking disorder during the phase transi-tion, evidenced by the presence of diffuse scattering ob-served in neutron [8] and X-ray [9] experiments, and hys-teresis effects that extend far beyond the transition re-gion, as seen in resistivity measurements along the ther-mal hysteresis loop [10]. These effects have been largelyignored, though one of the surface Fermi arcs was notedto persist to ∼
90 K above the transition temperature andto have a history-dependent appearance [6]. In general,structural phase transitions that involve in-plane trans-lations of layers resulting from changes in temperatureor pressure have been neglected, but many materials fallin this category, including Ta NiSe [11], In Se [12, 13], α -RuCl [14], CrX (X=Cl, Br, I) [15], and MoS [16–18]. A better understanding of these types of transitionswould not only elucidate these material properties, butcould also lead to the discovery of new phases.The T d and 1T (cid:48) phases can be constructed from astacking pattern of “A” and “B” operations, as shown in ∗ Present address: Duke University, Dept. of Physics, Durham,NC 27708 † Corresponding author. Email: [email protected]
Fig. 1(a). The A operation maps one layer of T d to thelayer below it, so T d can be built from repeating “AA”sequences. The B operation is the same as for A but fol-lowed by a translation of ± (cid:48) can be builtfrom repeating “AB” sequences. We previously reportedthat diffuse scattering observed in the H L scatteringplane on cooling from 1T (cid:48) towards T d (in particular, thelow intensity along (60 L )) is consistent with a disorderedA/B stacking pattern [8]. How the stacking changes withtemperature has not been closely examined, though anexplanation for the relative stability of the T d and 1T (cid:48) phases via free energy calculations was earlier proposed[19]. Understanding the nature of layer stacking will pro-vide useful insight into how Weyl nodes disappear acrossthe phase boundary.We performed elastic neutron scattering as a functionof temperature to study the mechanism of the struc-tural phase transition between the 1T (cid:48) and T d phases inMoTe . On warming, the recently discovered T ∗ d phase[20] was observed, having a pseudo-orthorhombic struc-ture and a four-layer unit cell, rather than the two-layerunit cells of 1T (cid:48) and T d . The stacking sequence of T ∗ d canbe described by “AABB”, as shown in Fig. 1(a). Uponwarming, the T d → T ∗ d transition is not accompanied bydisorder. Diffuse scattering is observed on further warm-ing from T ∗ d to 1T (cid:48) . On the other hand, on cooling from1T (cid:48) to T d , the T ∗ d phase is absent and only diffuse scat-tering is observed that suggests a frustrated tendencytoward the “AABB” layer order.Elastic neutron scattering was performed at Oak RidgeNational Laboratory, on the triple axis spectrometersHB1, CG4C, and HB1A at the High Flux Isotope Re-actor; and on the time-of-flight spectrometer CORELLIat the Spallation Neutron Source [21]. Though the crys-tals are monoclinic at room temperature, for simplicity,we use orthorhombic coordinates, with a ≈ . b ≈ . c ≈ . (cid:48) -40 (cid:48) -S-40 (cid:48) -120 (cid:48) for HB1 and CG4C, and 40 (cid:48) -40 (cid:48) -S-40 (cid:48) -80 (cid:48) for HB1A.Incident neutron energies were 13.5 meV for HB1, 4.5meV for CG4C, and 14.6 meV for HB1A. Resistance mea- a r X i v : . [ c ond - m a t . m t r l - s c i ] A ug surements were performed in a Quantum Design PhysicalProperty Measurement System. All crystals were grownin excess Te flux, including the two used for neutron scat-tering, “MT1” and “MT2”. MT1 has the compositionMoTe , while MT2 has the composition Mo − x W x Te with x < ∼ .
01 as estimated by energy dispersive X-rayspectroscopy and the c -axis lattice constant. Details canbe found in the Supplemental Materials. (c)(b)(d) (e)Cooling ac T d AAAA T d *ABBA 1T’ABAB (a) == MoTe (20¯3) D T ′ (202) D T ′ (205) T * d (206) T * d Warming
MT2 (f) (g) Te MT1 (203) T d (202) T d (20¯3) D T ′ (202) D T ′ (203) T d (202) T d Figure 1. (a) The stacking patterns of 1T (cid:48) , T ∗ d , and T d . Rect-angles show cells centered on points of inversion symmetry foreach layer. Neutron scattering intensity maps for (b,c) MT1and (d,e) MT2 as a function of temperature along the (2 , , L )line on cooling (left) and warming (right). Data were takenon HB1A for (b,c) and HB1 for (d,e). (f,g) Intensity plotsalong (2 , , L ) showing diffuse scattering in MT1 on cooling(f) and warming (g). In Figures 1(b-e), neutron scattering intensity scansalong (2 , , L ) are combined for many temperatures oncooling and warming through the hysteresis. In the 1T (cid:48) phase, the (202) D T (cid:48) and (20¯3) D T (cid:48) Bragg peaks are ob-served near L = 2 . L = 2 .
7, respectively; D D (cid:48) twins. (Since MT1 couldnot be warmed fully into 1T (cid:48) in Fig. 1(c) for technical rea-sons, diffuse scattering was present on subsequent coolingfrom 300 K in Fig. 1(b).) At low temperatures, T d -phaseBragg peaks at L = 2 and L = 3 are observed, as indi-cated in the figure. On warming from T d past ∼
260 K, apeak appears at L = 2.5, indicating the onset of T ∗ d . Thepresence of this peak at half-integer L indicates an out- of-plane doubling of the unit cell, so we label this peak(205) T ∗ d (Fig. 1(e)). With additional warming, a gradualtransformation into the 1T (cid:48) phase occurs, accompaniedby diffuse scattering indicating stacking disorder. Exam-ples of the diffuse scattering can be seen in the individualplots of intensity along (2 , , L ) in Fig. 1(f), where 1T (cid:48) is transitioning into T d ; and in Fig. 1(g), where T ∗ d istransitioning into 1T (cid:48) . For MT2, we measured throughthe hysteresis twice and found the same pattern of diffusescattering at the same temperatures along the hysteresis,suggesting that the appearance of the diffuse scatteringthrough the hysteresis is reproducible. (c)(d)(b)(a)(e) Figure 2. Neutron scattering intensity maps (a,c) and simu-lated data (b,d) in the 0 KL (a,b) and 2 KL (c,d) scatteringplanes in the T ∗ d phase, from the MT1 crystal measured atCORELLI. Data taken on warming at 300 K. (e) Ratio ofselected Bragg peak intensities between 300 K and 240 K.Intensities are from Gaussian fits from data averaged within ± H - and K -directions. The T ∗ d phase structure can be deduced from the fol-lowing observations: First, T ∗ d appears to be orthorhom-bic, but has additional peaks at half-integer L values rel-ative to T d , indicating a four-layer unit cell. Second, the1T (cid:48) and T d phases can be built from A/B stacking se-quences, so we presume the same is true for T ∗ d . Thereare only two possible pseudo-orthorhombic stacking se-quences: “AABB” and “ABBA”, which are twins of eachother. Since this structure, with highest possible sym-metry P2 /m, appears to have an orthorhombic unit cellbut has atomic positions incompatible with orthorhombicspace groups, we refer to it as pseudo-orthorhombic.To verify the predicted AABB stacking structure of T ∗ d ,we carried out single-crystal neutron diffraction measure-ments on the MT1 crystal on CORELLI, and the data areshown in Figures 2(a,c). The data were taken on warm-ing to 300 K, and the presence of peaks at half-integer L in the 2 KL plane in Fig. 2(c) confirm the presence ofthe T ∗ d phase. The diffuse scattering streaks along L arefrom stacking disorder that was already present on warm-ing from 240 K, possibly due to not cooling sufficientlyinto T d beforehand. (There is a discrepancy between thedetection of T ∗ d in MT1 at 300 K on CORELLI and up to ∼
280 K on HB1A. The cause of the temperature discrep-ancy is unknown, but may be related to the presence ofstacking disorder before warming.) Figures 2(b,d) showsimulated intensity maps. To match the data, it was nec-essary to consider a 47.8% volume fraction of T d as wellas 28.2% and 24.0% volume fractions of the two T ∗ d twins.The volume fractions were obtained by fitting the inten-sities of Bragg peaks within − ≤ H ≤ − ≤ K ≤ − ≤ L ≤
20 with the calculated peak intensitiesof the ideal “AA”, “AABB”, and “ABBA” stacking se-quences of T d and the two T ∗ d twins, respectively. Thesestructures were built from layers having the coordinatesin Ref. [22]. As can be seen in Figures 2(a-d), the pat-terns of peak intensities in these scattering planes matchthose arising from our model.Stringent constraints on possible T ∗ d structures followfrom the lack of change in (00 L ) and (01 L ) peak inten-sities between the T d phase at 240 K and the T ∗ d phaseat 300 K (as seen from the near-unity intensity ratiosin Fig. 2(e). For context, intensity ratios for (20 L ) and(30 L ) peaks are also included.) A lack of change in 0 KL peak intensities implies a lack of change in atomic posi-tions along the b - and c -directions, but is consistent withlayer displacements along the a -direction, as is the casebetween 1T (cid:48) and T d [8]. The AABB structure should becentrosymmetric, since it can be transformed from thecentrosymmetric AB-stacked 1T (cid:48) phase by a centrosym-metric series of translations (see Supplementary Materi-als). Inversion symmetry centers for the AABB struc-ture are depicted in Fig. S2 in the Supplemental Ma-terials. Barring small non-centrosymmetric distortions,which are unlikely given that first-principles calculationshave shown that MoTe layers isolated from the non-centrosymmetric T d environment tend to become cen-trosymmetric [23], we conclude that T ∗ d is centrosym-metric with P2 /m symmetry. A structural refinementassuming P2 /m symmetry was performed (see Supple-mentary Materials), with rough agreement between therefined and ideal coordinates, though the absence of vis-ible 0 KL peaks in our data (apart from those with even K + L ) indicates that the true T ∗ d structure is closer tothe ideal AABB stacking than our refined coordinates.For a closer look at how the transition proceeds, inFig. 3(a,b) we plot intensities for four Bragg peaks as afunction of temperature. The integrated intensities wereobtained from fits of the data shown in Figures 1(d) and1(e). On cooling below ∼
280 K (Fig. 3(a)), there is asteady decrease in the intensity of the 1T (cid:48) peaks in thehysteresis region. The peaks eventually become difficultto resolve from the diffuse scattering, and fitting was not
D1D2 (a) (b)(c)(e)
MT2 (d)(f)
MT2
Figure 3. (a,b) Bragg peak intensities plotted as a functionof temperature on warming and cooling for MT2. Red bandsdenote regions where fitting was poor. Closed symbols denotefits to the same hysteresis loop (with cooling data measuredbefore warming). Open symbols correspond to a previoushysteresis loop. (c) Plots of intensity integrated within (2, 0,2 . ≤ L ≤ .
61) for MT2 taken through two different hys-teresis loops. Data taken on CG4C for the narrow hysteresis(black), and on HB1 for the wide hysteresis (green; from samedata as Fig. 1(d,e).) Curves normalized to their largest val-ues. (d) Neutron scattering intensity along (2 , , L ) for MT2,with data taken on CG4C at various temperatures, verticallydisplaced for clarity. (e) Resistance of a MoTe crystal, mea-sured through two hysteresis loops that begin on warmingfrom 200 K. (f) The derivative dR/dT of the data shown in(e). done within the region indicated by the pink shaded bar.On further cooling, the T d peaks appear.In contrast, on warming the intensities of the T d peaksin Fig. 3(b) remain constant until a sudden change isobserved around 260 K. At this temperature, the T ∗ d peaks appear (magenta symbols) at the expense of theT d peaks. On further warming, the T d and T ∗ d peak in-tensities both decrease and disappear by 280 K. Again,diffuse scattering is observed (pink shaded region) priorto the crystal transforming fully into 1T (cid:48) . Though a co-herent, long-range T ∗ d phase only appears on warming,the intensity shift toward (2 , , .
5) on cooling as seenin Fig. 1(b) and 1(d) suggests a tendency toward theAABB stacking, though frustrated and not resulting inan ordered structure. On both cooling into T d or warm-ing into 1T (cid:48) , there is a gradual increase in the intensityof the T d and 1T (cid:48) peaks, which occurs with a decrease indiffuse scattering (see Fig. S3 of the Supplemental Ma-terials). This lingering diffuse scattering is probably re-lated to the long residual hysteresis commonly observedin the resistivity measurements (e.g., in Ref. [24], or Fig.S4 in the Supplemental Materials.)To investigate the boundary between the T d and T ∗ d re-gions, in Fig. 3(c), intensity integrated near (2 , , .
5) isplotted for two different thermal hysteresis loops for theMT2 crystal. The narrow hysteresis loop (black sym-bols) corresponds to the sample warming into T ∗ d , thencooling back to T d without entering 1T (cid:48) . Fig. 3(d) isa plot of the data used to calculate the narrow hystere-sis loop intensities. The (205) T ∗ d peak intensity rises andfalls through the hysteresis loop. Diffuse scattering is notpresent even at a temperature a few Kelvin below the T ∗ d peak’s disappearance on cooling. Thus, T ∗ d → T d likelyproceeds without disorder. In contrast, a wider hystere-sis loop (green symbols) is observed when the sample isallowed to warm into 1T (cid:48) . This is coupled to the sub-stantial diffuse scattering present on cooling, as shownearlier in Fig. 1(d,e). Nevertheless, for both narrow andwide hysteresis loops, a sudden drop of intensity near(2 , , .
5) appears on cooling below 255 K, though moregradually for the wide hysteresis loop.A similar pattern can be seen in the resistance dataof Fig. 3(e), taken on a MoTe crystal with residual re-sistance ratio ∼
460 through consecutive narrow (black;200 to 265 K) and wide (green; 200 to 350 K) hysteresisloops. On cooling, the resistance decreases quickly andin a symmetric manner (for cooling vs. warming) for thenarrow hysteresis loop, but more slowly and asymmetricfor the wide hysteresis loop. Even so, the temperatureat which both loops begin to bend on cooling is similar,as seen from dR/dT in Fig. 3(f), though slightly lowerfor the wide hysteresis loop. The kink seen on warming(near 258 K) is likely the onset of T ∗ d and not 1T (cid:48) , judg-ing from the temperature and the similarities betweenthe resistance and neutron scattering hysteresis loops.We next discuss how these structural transitions pro-ceed and the kinds of interlayer interactions that maybe responsible, beginning from the observation that theonset to T d occurs at a similar temperature whether cool-ing from the ordered T ∗ d phase, or from the frustrated T ∗ d region accessed on cooling from 1T (cid:48) . Since the onset tem-perature to T d does not appear to vary substantially withoverall stacking disorder, we suggest that short-range rather than long-range interlayer interactions determinethe onset temperatures (into 1T (cid:48) or T ∗ d as well as T d ).(Though we use the term “interlayer interactions”, weemphasize that these are effective interactions. Whetheran interlayer boundary shifts from A → B depends on thefree energy, which depends on the surrounding environ-ment, which is specified by the A/B stacking sequence.“Interlayer interactions” represent the dependence of aninterlayer boundary’s contribution to the free energy onthe surrounding stacking, and can be indirect, involvingchanges to band structure, phonon dispersion, etc.)In contrast, long-range interlayer interactions may gov-ern the gradual decrease in diffuse scattering and in-crease in Bragg peak intensities on warming into 1T (cid:48) or cooling into T d . What kind of stacking faults caus-ing this diffuse scattering persist on cooling into T d ,even when short-range interlayer interactions favor anordered phase? At twin boundaries, shifts of A → B orB → A (e.g., AAA A BBB... → AAA B BBB...) would notchange the short-range environment, and could only beinduced by changes in long-range interlayer interactions.The decrease in diffuse scattering in T d on cooling can beexplained by the annihilation of these twin boundaries,either by joining in pairs or by exiting a crystal surface.The lack of change on subsequent warming can be ex-plained by the relaxation of conditions that, on cooling,had driven twin boundaries to annihilate.Previous studies on MoTe should be re-examined inlight of the existence of the T ∗ d phase. First, the hys-teresis loop in resistivity (first reported in Ref. [25]) hasbeen interpreted as indicating the transition between T d and 1T (cid:48) , but in view of the current data, most of thechange in the resistance occurs between T d and T ∗ d onwarming. Second, second harmonic generation (SHG) in-tensity measurements, expected to be zero for inversionsymmetry and nonzero otherwise, show abrupt (within < (cid:48) occursgradually, and since T ∗ d appears to be centrosymmet-ric, the abrupt warming transition seen in SHG may bedue to the T d → T ∗ d rather than T d → (cid:48) transition. Theabrupt transition on cooling is harder to explain, but itis possible that the loss of inversion symmetry on coolinginto T d occurs suddenly even as the transition proceedswith disorder. Our findings may also inform proposedapplications, such as the photoinduced ultrafast topolog-ical switch in Ref. [27]; since the T d → T ∗ d → T d transitionoccurs without disorder and with only a ∼ ∗ d appears to be centrosymmetric, a topologi-cal switch may more efficiently use T ∗ d rather than 1T (cid:48) .In conclusion, using elastic neutron scattering, wemapped the changes in stacking that occur in the ther-mal hysteresis between the T d and 1T (cid:48) phases in MoTe .On warming from the orthorhombic T d , T ∗ d arises with-out diffuse scattering and corresponds to an “AABB”sequence of stacking operations. Diffuse scattering ispresent on further warming from T ∗ d to 1T (cid:48) , and on cool-ing from 1T (cid:48) to T d , where a frustrated tendency towardthe “AABB” stacking is seen. Thus, the 1T (cid:48) -T d transi-tion has complex structural behavior and deserves furtherstudy. ACKNOWLEDGEMENTS
This work has been supported by the Department ofEnergy, Grant number DE-FG02-01ER45927. A portionof this research used resources at the High Flux IsotopeReactor and the Spallation Neutron Source, which areDOE Office of Science User Facilities operated by OakRidge National Laboratory. [1] Y. Sun, S.-C. Wu, M. N. Ali, C. Felser, and B. Yan,“Prediction of Weyl semimetal in orthorhombic MoTe ,”Phys. Rev. B , 161107(R) (2015).[2] K. Deng, G. Wan, P. Deng, K. Zhang, S. Ding, E. Wang,M. Yan, H. Huang, H. Zhang, Z. Xu, et al. , “Experimen-tal observation of topological fermi arcs in type-II Weylsemimetal MoTe ,” Nature Physics , 1105 (2016).[3] D. Rhodes, R. Sch¨onemann, N. Aryal, Q. Zhou, Q. R.Zhang, E. Kampert, Y.-C. Chiu, Y. Lai, Y. Shimura,G. T. McCandless, J. Y. Chan, D. W. Paley, J. Lee, A. D.Finke, J. P. C. Ruff, S. Das, E. Manousakis, and L. Bali-cas, “Bulk Fermi surface of the Weyl type-II semimetalliccandidate γ -MoTe ,” Phys. Rev. B , 165134 (2017).[4] Q. Zhou, D. Rhodes, Q. R. Zhang, S. Tang, R. Sch¨one-mann, and L. Balicas, “Hall effect within the colos-sal magnetoresistive semimetallic state of MoTe ,” Phys.Rev. B , 121101(R) (2016).[5] S. Thirupathaiah, R. Jha, B. Pal, J. S. Matias, P. K.Das, P. K. Sivakumar, I. Vobornik, N. C. Plumb, M. Shi,R. A. Ribeiro, and D. D. Sarma, “MoTe : An uncom-pensated semimetal with extremely large magnetoresis-tance,” Phys. Rev. B , 241105(R) (2017).[6] A. P. Weber, P. R¨ußmann, N. Xu, S. Muff, M. Fanciulli,A. Magrez, P. Bugnon, H. Berger, N. C. Plumb, M. Shi,S. Bl¨ugel, P. Mavropoulos, and J. H. Dil, “Spin-ResolvedElectronic Response to the Phase Transition in MoTe ,”Phys. Rev. Lett. , 156401 (2018).[7] N. Xu, Z. W. Wang, A. Magrez, P. Bugnon, H. Berger,C. E. Matt, V. N. Strocov, N. C. Plumb, M. Radovic,E. Pomjakushina, K. Conder, J. H. Dil, J. Mesot,R. Yu, H. Ding, and M. Shi, “Evidence of a Coulomb-Interaction-Induced Lifshitz Transition and Robust Hy-brid Weyl Semimetal in T d -MoTe ,” Phys. Rev. Lett. , 136401 (2018).[8] J. A. Schneeloch, C. Duan, J. Yang, J. Liu, X. Wang, andD. Louca, “Emergence of topologically protected statesin the MoTe Weyl semimetal with layer-stacking order,”Phys. Rev. B , 161105(R) (2019).[9] R. Clarke, E. Marseglia, and H. P. Hughes, “A low-temperature structural phase transition in β -MoTe ,”Philos. Mag. B , 121–126 (1978).[10] T. Zandt, H. Dwelk, C. Janowitz, and R. Manzke,“Quadratic temperature dependence up to 50?K of theresistivity of metallic MoTe ,” J. Alloys Compd. Proceed-ings of the 15th International Conference on Solid Com-pounds of Transition Elements, , 216–218 (2007).[11] A. Nakano, K. Sugawara, S. Tamura, N. Katayama,K. Matsubayashi, T. Okada, Y. Uwatoko, K. Mu-nakata, A. Nakao, H. Sagayama, R. Kumai, K. Sug-imoto, N. Maejima, A. Machida, T. Watanuki, andH. Sawa, “Pressure-induced coherent sliding-layer tran-sition in the excitonic insulator Ta NiSe ,” IUCrJ (2018), 10.1107/S2052252517018334.[12] F. Ke, C. Liu, Y. Gao, J. Zhang, D. Tan, Y. Han, Y. Ma,J. Shu, W. Yang, B. Chen, H.-K. Mao, X.-J. Chen,and C. Gao, “Interlayer-glide-driven isosymmetric phasetransition in compressed In Se ,” Appl. Phys. Lett. ,212102 (2014).[13] Jinggeng Zhao and Liuxiang Yang, “Structure Evolutionsand Metallic Transitions in In Se Under High Pressure,”J. Phys. Chem. C , 5445–5452 (2014). [14] A. Glamazda, P. Lemmens, S.-H. Do, Y. S. Kwon, andK.-Y. Choi, “Relation between Kitaev magnetism andstructure in α -RuCl ,” Phys. Rev. B , 174429 (2017).[15] M. A. McGuire, H. Dixit, V. R. Cooper, and B. C. Sales,“Coupling of Crystal Structure and Magnetism in theLayered, Ferromagnetic Insulator CrI ,” Chem. Mater. , 612–620 (2015).[16] L. Hromadov´a, R. Martoˇn´ak, and E. Tosatti, “Structurechange, layer sliding, and metallization in high-pressureMoS ,” Phys. Rev. B , 144105 (2013).[17] Z.-H. Chi, X.-M. Zhao, H. Zhang, A. F. Goncharov, S. S.Lobanov, T. Kagayama, M. Sakata, and X.-J. Chen,“Pressure-Induced Metallization of Molybdenum Disul-fide,” Phys. Rev. Lett. , 036802 (2014).[18] A. P. Nayak, S. Bhattacharyya, J. Zhu, J. Liu, X. Wu,T. Pandey, C. Jin, A. K. Singh, D. Akinwande, andJ.-F. Lin, “Pressure-induced semiconducting to metallictransition in multilayered molybdenum disulphide,” Nat.Commun. , 3731 (2014).[19] H.-J. Kim, S.-H. Kang, I. Hamada, and Y.-W. Son, “Ori-gins of the structural phase transitions in MoTe andWTe ,” Phys. Rev. B , 180101(R) (2017).[20] S. Dissanayake, C. Duan, J. Yang, J. Liu, M. Matsuda,C. Yue, J. A. Schneeloch, J. C. Y. Teo, and D. Louca,“Electronic band tuning under pressure in MoTe topo-logical semimetal,” npj Quantum Mater. , 1–7 (2019).[21] F. Ye, Y. Liu, R. Whitfield, R. Osborn, andS. Rosenkranz, “Implementation of cross correlation forenergy discrimination on the time-of-flight spectrometerCORELLI,” J. Appl. Crystallogr. , 315–322 (2018).[22] Y. Qi, P. G. Naumov, M. N. Ali, C. R. Rajamathi,W. Schnelle, O. Barkalov, M. Hanfland, S.-C. Wu,C. Shekhar, Y. Sun, V. S¨uß, M. Schmidt, U. Schwarz,E. Pippel, P. Werner, R. Hillebrand, T. F¨orster, E. Kam-pert, S. Parkin, R. J. Cava, C. Felser, B. Yan, and S. A.Medvedev, “Superconductivity in Weyl semimetal candi-date MoTe ,” Nat. Commun. , 11038 (2016).[23] C. Heikes, I-L. Liu, T. Metz, C. Eckberg, P. Neves,Y. Wu, L. Hung, P. Piccoli, H. Cao, J. Leao, J. Paglione,T. Yildirim, N. P. Butch, and W. Ratcliff, “Mechani-cal control of crystal symmetry and superconductivity inWeyl semimetal MoTe ,” Phys. Rev. Mater. , 074202(2018).[24] T. Zandt, H. Dwelk, C. Janowitz, and R. Manzke,“Quadratic temperature dependence up to 50 k of theresistivity of metallic MoTe ,” J. Alloys Compd. ,216–218 (2007).[25] H. P. Hughes and R. H. Friend, “Electrical resistivityanomaly in β -MoTe ,” J. Phys. C: Solid State Phys. ,L103 (1978).[26] H. Sakai, K. Ikeura, M. S. Bahramy, N. Ogawa,D. Hashizume, J. Fujioka, Y. Tokura, and S. Ishiwata,“Critical enhancement of thermopower in a chemicallytuned polar semimetal MoTe ,” Sci. Adv. , e1601378(2016).[27] M. Y. Zhang, Z. X. Wang, Y. N. Li, L. Y. Shi, D. Wu,T. Lin, S. J. Zhang, Y. Q. Liu, Q. M. Liu, J. Wang,T. Dong, and N. L. Wang, “Light-Induced Subpicosec-ond Lattice Symmetry Switch in MoTe ,” Phys. Rev. X9