betaFeedback

WorksPhysics and Astronomy

Symmetry breaking and structure instability in ultra-thin 2H-TaS2 across charge density wave transition

Abstract

Ultra-thin 2D materials have shown complete paradigm shift of understanding of physical and electronic properties because of confinement effects, symmetry breaking and novel phenomena at nanoscale. Bulk 2H-TaS2 undergoes an incommensurate charge density wave (I-CDW) transition temperature, TI-CDW - 76 K, however, onset of CDW in atomically thin layers is not clear. We explored the evidence of CDW instability in exfoliated atomically thin 2H-TaS2 using low temperature Raman spectroscopy. We have emphasized on CDW associated modes, M1 - 125 cm-1, M2 -158 cm-1, and M3 -334 cm-1, with thickness - 3 nm (one unit-cell). The asymmetric (Fano) line shape of M2 suggests evidence of strong electron-phonon coupling, which mainly drives the CDW instability. Our observations provide key evidence that the CDW can persists even in one-unit cell with a TI-CDW well above - 200 K, which is higher than bulk 2H-TaS2.

Atomically thin two dimensional (2D) transition metal dichalcogenides (TMDCs)[1] have gained significant attentions due to their wide range of electronic, optical, chemical and thermal properties and potential applications field effect transistors[2], optoelectronics[3], photo detectors, energy harvesting and quantum sensing and metrology.[4] Among these 2D materials, the ultra-thin layers of TiSe2_2[5], NbSe2_2[6], TaSe2_2[7], VSe2_2[8], and TaS2_2[9] have exotic physical properties due to symmetry breaking with reduced dimensionality and thus showcase the emerging quantum correlated phenomena coupled with the atomic arrangements.[10, 11, 12, 13, 14] The quantum size of these material strongly affects the electron correlations, for instance, tunable band gap, charge density wave (CDW), superconductivity, and metal-insulator transitions, which are important to understand the many body physics of electron-electron and electron-phonon interactions,[15] and improved electronic transport.[16] In general, CDW is a periodic modulation of the electronic charge density accompanying with lattice distortion in the material, which is associated with the formation of superlattice and new type of electronic band structure because of the breaking of the lattice symmetry in the crystal. [17, 18] The energy imbalance of CDW is connected with the opening of an electronic band gap within the first Brillouin zone involving both electronic and lattice ordering of the materials. Various theoretical and experimental studies like inelastic neutron and X-ray scattering[19], time resolved optical pump-probe spectroscopy[20] and Raman spectroscopy[21] shows that CDW can be rooted from fermi surface nesting and electron-phonon (e-phe\text{-}ph) interactions. [17, 21, 22, 23] CDW can be classified based on the periodicity of the distorted lattice (CDW Wave vector, q⃗CDW\vec{q}_{CDW}), as commensurate (C-CDW), nearly commensurate (NC-CDW) and incommensurate (I-CDW).[24] Here, the C-CDW has q⃗CDW\vec{q}_{CDW} as an integer multiple whereas I-CDW has non-integer multiple of reciprocal lattice vector of the undistorted lattice.[25]

Among layered materials, bulk 2H-TaS22\mathrm{H}\text{-}\mathrm{TaS}_2 is metallic in nature and exhibits I-CDW at transition temperature (TI−CDWT_{I-CDW}) between 75-80 K and undergoes superconducting transition at TCT_C ~ 0.8 K.[26] Similar to 2H polymorphs of TMDCs, crystalline unit cell of 2H-TaS22\mathrm{H}\text{-}\mathrm{TaS}_2 contains two tri-atomic layers (SS-TaTa-SS) separated from each other with weak van der Waals (vdWs) gap.[27] Monolayer of TaS2\mathrm{TaS}_2 is mainly consisting of three atomic planes, where TaTa is sandwiched between two SS layers. Previous studies of 1T polymorph of TaS2\mathrm{TaS}_2 have demonstrated the effect of reduced dimensionality on the CDW, however a very few studies available for CDW instability in exfoliated 2H-TaS22\mathrm{H}\text{-}\mathrm{TaS}_2. In this regards, the origin mechanism and TI−CDWT_{I-CDW} of the CDW in ultra-thin layer of 2H-TaS22\mathrm{H}\text{-}\mathrm{TaS}_2 is not clear. A recent study revealed that CDW phase cannot exist below a critical thickness (dd) of ~ 3 nm,[28] however, another report declares the existence of CDW in monolayer of 2H-TaS22\mathrm{H}\text{-}\mathrm{TaS}_2[29]. Hence, the CDW instability in thin layers of 2H-TaS22\mathrm{H}\text{-}\mathrm{TaS}_2 is highly debatable research, which needs to be explored in detail. The CDW transition temperature (TI−CDWT_{I-CDW}) can be tuned by doping, reduced dimensionality[12], electrostatic gating[14] or strain engineering[30, 31] which is already reported for TMDCs like NbSe2\mathrm{NbSe}_2[32], VSe2\mathrm{VSe}_2[33], TaSe2\mathrm{TaSe}_2[34], TiSe2\mathrm{TiSe}_2[35] and 1T-TaS21\mathrm{T}\text{-}\mathrm{TaS}_2. [26]

Fundamentally, the CDW state in 2H-TaS22\mathrm{H}\text{-}\mathrm{TaS}_2 is stabilized by accompanying a periodic lattice distortion majorly through e-phe\text{-}ph interactions, therefore, the investigation of the phonons is very important for CDW transition in atomically thin flakes.[8] To address this discrepancy, there are various experiments such as charge transport,[36] transmission electron microscopy[28] and inelastic X-ray and neutron scattering[22], have been used so far. Considering the involvements of e-phe\text{-}ph interaction in CDW, Raman spectroscopy is an easy but a very strong tool to probe the dynamics of phonon instability in thin exfoliated layer of 2H-TaS22\mathrm{H}\text{-}\mathrm{TaS}_2.[8]

In this work we report on the thickness (dd) dependent phonon anomaly and structural instability in ultra-thin 2H2\mathrm{H}-TaS2\mathrm{TaS}_{2} CDW material. The atomically thin flakes are mechanically exfoliated on SiO2SiO_2/Si substrate from a single crystal of 2H2\mathrm{H}-TaS2\mathrm{TaS}_{2}, which is grown by chemical vapor transport technique.[8] The experimental detail regarding Raman, AFM, and transport can be found in the Supporting Information (SI). The TI−CDWT_{\mathrm{I-CDW}} for bulk 2H2\mathrm{H}-TaS2\mathrm{TaS}_{2} is estimated using low temperature electrical transport (Figure S1a) and heat capacity measurements (Figure S1b) in SI. Thus, the bulk TaS2\mathrm{TaS}_{2} has an expected I-CDW below TI−CDWT_{\mathrm{I-CDW}} ∼\sim 76 K. The bulk crystal and thin flakes are characterized by low temperature Raman and polarization spectroscopy to explore the phonon responses during the CDW instability in the exfoliated thin layer of 2H2\mathrm{H}-TaS2\mathrm{TaS}_{2}. Thickness, dd, of the exfoliated flakes have been estimated from the atomic force microscopy (AFM) and corresponding optical images are shown in Figure 1. The high dd (∼\sim 100 nm) in Figure 1 (a) has been considered as the bulk crystal. Atomically thin layers of the mechanically exfoliated 2H2\mathrm{H}-TaS2\mathrm{TaS}_{2} have been first identified by optical microscopy, represented in Figure 1b-d, followed by dd identifications using topographic images by AFM (insets of Figure 1e-g). The corresponding AFM height profiles have been shown for ∼\sim 10 nm (Figure 1e), ∼\sim 7 nm (Figure 1f), ∼\sim 5 nm and ∼\sim 3 nm as marked in Figure 1g. We have also analyzed the larger flakes and shown in SI, for instance, ∼\sim 27 nm (Figure S2a), ∼\sim 16 nm (Figure S2b).

Optical images and AFM height profiles of bulk and mechanically exfoliated thin flakes

Figure 1. Optical images of (a) 2H2\mathrm{H}-TaS2\mathrm{TaS}_{2} single crystal and mechanically exfoliated thin flakes (b) ∼\sim 10 nm (c) ∼\sim 7 nm (d) ∼\sim 5 nm and ∼\sim 3 nm on Si/SiO2Si/SiO_2 substrate. Height profiles of (e) ∼\sim 10 nm (f) ∼\sim 7 nm (g) ∼\sim 5 nm and ∼\sim 3 nm thin flakes and corresponding topographic AFM images are shown in respective insets.

The 2H-TaS2_2 crystallizes in hexagonal phase, wherein a unit cell is composed of two tri-atomic layers (SS-TaTa-SS) stacked with weak (vdWs gap (Figure 2a). The tri-atomic layer (called monolayer) is made of covalent bonded SS-TaTa-SS in the trigonal prismatic structure, where TaTa atomic layer is sandwiched between layers of SS atoms.1 Thus, primitive cell of 2H-TaS2_2 has three atoms. Correspondingly, there are nine vibrational modes at the center of the Brillouin zone, three of which are acoustic and six are optical modes. According to the group theory analysis, these optical modes can be represented as irreducible representation; Γ=A1g+E1g+2E2g+A2u+E1u\Gamma= A_{1g}+E_{1g}+2E_{2g}+A_{2u}+E_{1u}. [29, 37] Here, A1gA_{1g}, E1gE_{1g}, and E2gE_{2g} are Raman active while A2uA_{2u}, and E1uE_{1u} are infra-red active modes. Figure 2b has the normalized Raman spectra of the bulk 2H-TaS2_2 showing various modes at ∼21\sim21 cm−1^{-1} (E2g2E^{2}_{2g}), ∼180\sim180 cm−1^{-1} (2ph2ph), ∼236\sim236 cm−1^{-1} (E2g1E^{1}_{2g}), and ∼400\sim400 cm−1^{-1} (A1gA_{1g}), and symmetries are assigned in consistent with earlier reports.[21, 36] The schematics of the corresponding atomic displacements of Raman-active modes are represented in inset of Figure 2b. The A1gA_{1g} symmetry has atomic displacements along the out of plane while the E2gE_{2g} has the displacements along the in-plane direction of the flakes. The mode at ∼180\sim180 cm−1^{-1} is a characteristic mode of CDW materials and defined as two phonon mode (2ph2ph), which is originated from the second order scattering process of two acoustic phonons.[21, 38] Generally, 2ph2ph mode is a second order scattering process of two acoustic phonons having frequency approximately twice of Kohn anomaly near CDW wavevector (q⃗CDW=2/3ΓM\vec{q}_{CDW} = 2/3\Gamma\mathrm{M}). The mode at ∼360\sim360 cm−1^{-1}, is a arising from the background.[21] A strong dimensionality effect is observed in the Raman spectra of flakes with varying dd, at 300 K, (Figure 2c). For the ∼3\sim3 nm flakes, the broad modes at ∼88\sim88 cm−1^{-1}, ∼228\sim228 cm−1^{-1}, 380 cm−1^{-1} are related to the substrate (Si/SiO2_2).[21] The contribution from the substrate is ubiquitous for atomically thin layers on the substrate and is also observed in Raman spectra of monolayers of TMDCs such as MoS2_2[39], WS2_2[40], WSe2_2[41], and TaSe2_2[42]. A systematic variation of phonon frequency (ω\omega) and integerated area (IAIA) of the E2g1E^{1}_{2g} and A1gA_{1g} with dd are summarized in Figure 2d and 2e. A clear dd dependence on the Raman modes is observed, where the the E2g1E^{1}_{2g} mode softens(red shifts) with increasing dd (Figure 2d). The softening with dd is because of the enhancement of the dielectric screening of the long-range coulomb interactions in bulk .[43, 44] On the other hand the A1gA_{1g} mode hardens (blue shifts) for layer dd up to ∼7\sim7 nm and then drops abruptly for ∼10\sim10 nm and followed by hardening at higher dd. The abrupt softening of A1gA_{1g} mode in the ∼10\sim10 nm has been observed in earlier reports.[44] The hardening of A1gA_{1g} mode is understood with the classical harmonic oscillators coupled through vdWs forces as the effective restoring force is increasing with addition of layers.[45] Interestingly, the IAIA of both E2g1E^{1}_{2g} and A1gA_{1g} modes (Figure 2e), decreases with the increase in the dd (except for ∼3\sim3 nm), that indicates the lower cross-section area in thick flakes.[39] The anomalous behavior of IAIA for ∼3\sim3 nm is because of the involvements of substrate contributions in estimating the IAIA.

Crystal structure and Raman spectra of layered 2H-TaS$_2$

Figure 2. (a) Crystal structure of layered 2H-TaS2_2, where Ta and S atoms are shown with green and magenta color, respectively, and box represents the unit cell. (b) Normalized room temperature Raman spectra of bulk (inset shows the schematic representation of E2g2E^{2}_{2g}, E2g1E^{1}_{2g}, and A1gA_{1g} Raman modes), Thickness dependent (c) Raman spectra, (d) ω\omega, and (e) IAIA of E2g1E^{1}_{2g}, and A1gA_{1g} Raman modes.

Figure 3a shows the Raman spectra, at 10 K, where the bulk crystal has expected 2ph2_{ph}, E2g1E_{2g}^{1} and A1gA_{1g} symmetry modes, while the exfoliated flakes shows new Raman modes M1_1 ∼ 125 cm−1^{-1}, M2_2 ∼ 158 cm−1^{-1}, M3_3 ∼ 334 cm−1^{-1}, arising due to dimensionality effect. To the best of our understanding these modes are observed because of CDW transition but not discussed in details, so far. For bulk crystal, the intensity of the A1gA_{1g} modes for spectra at 10 K have increased in compared with spectra at 300 K (Figure 2b).

Raman spectra and thickness dependence of Raman modes

Figure 3. (a) Raman spectra at 10 K for bulk, ∼ 10 nm, ∼ 7 nm, and ∼ 3 nm flakes. Thickness dependence ω\omega, and IAI_A (b-c) of E2g1E^{1}_{2g}, A1gA_{1g}, (d-e) M1_1, and M2_2 Raman modes, respectively.

Apart from the E2g1E_{2g}^{1}, A1gA_{1g} and 2ph2_{ph} modes, no signatures of new modes (M1_1, M2_2, and M3_3) observed in the spectra of bulk sample (TCDW_{\mathrm{CDW}} ∼ 76 K). Thus, M1_1, M2_2 and M3_3 modes are characteristic of thin flakes (≤ 16 nm) and are associated with the lower dimensionality, symmetry breaking, and structural instabilities associated with CDW. The detailed analysis of the new modes is done in following texts. Similar to the dd dependence at 300 K, both the E2g1E_{2g}^{1} and A1gA_{1g} modes have shown the softening and hardening along with IAI_A at 10 K as well (Figure 3b and Figure 3c), which shows this nature is intrinsic to the flakes. For the one-unit cell, (∼ 3 nm), the E2g1E_{2g}^{1} is very broad and the data could not be shown in Figure 3b and Figure 3c. Symmetry of the M1_1, M2_2 and M3_3 modes is confirmed by the polarized Raman spectroscopy (Figure S3).[8, 46] The M2_2 mode is present in parallel polarization, whereas M1_1 and M3_3 are present for both parallel as well as perpendicular polarization. Thus, M2_2 belong to A symmetry while M1_1 and M3_3 are of E symmetry.[46] The modes M1_1 and M2_2 exist between the frequency range of 120-160 cm−1^{-1} and can associated with mechanism of 2ph2_{ph} mode which is observed for bulk crystal. Remarkably, M1_1 and M2_2 exhibit distinct symmetries but show a significant softening with the increase in dd of the flakes (Figure 3d). The ω\omega of M1_1 is constant below ~ 7 nm and softens for thick flakes, which follows nature of E2g1E^{1}_{2g}. However, in the case of M2_2, which is of AA-symmetry, the softening with dd is anomalous and surprising.

Polarized Raman spectra with varying incident-light polarization

Figure S3. Polarized Raman spectra of ~ 7 nm thin layer with the variation of polarization of incident light from 0° to 180° in backscattering geometry.

The line shape analysis of M2_2 show an asymmetry towards low energy indicating the existence of Fano resonance (Figure S4), which is also observed in the graphene[47], Bi2_2Se3_3[45] and nanowires[48]. Fano resonance is a quantum interference effect which depends primarily on the ee-phph coupling strength of the material.[45, 49] Thus, the anomalous response of M2_2 and the asymmetric line shape provides significant evidence of the involved ee-phph coupling in thin flakes of 2H2\mathrm{H}-TaS2\mathrm{TaS}_2.[21] For 2H2\mathrm{H}-TaS2\mathrm{TaS}_2, the ee-phph coupling is the major cause of CDW. Similar to the dd dependence for IAIA of E2g1E_{2g}^{1} and A1gA_{1g} at 300K (Figure 2e) and 10K (Figure 3c), the IAIA of both M1_1 and M2_2 also decreases with increasing dd (Figure 3e). Recently, Han-Chun Wu et.al has observed new Raman mode ~ 155 cm−1^{-1} in monolayer of 2H2\mathrm{H}-TaS2\mathrm{TaS}_2, which is known to be degenerate CDW breathing and wriggling mode.[29] In bulk 2H2\mathrm{H}-TaS2\mathrm{TaS}_2, CDW has been well understood using transport, angle resolved photoemission spectroscopy, density functional theory.[21] To explore the cause and effect of CDW in thin layers and to elucidate the observation of new CDW modes, a detailed temperature dependent studies of ultra-thin 2H2\mathrm{H}-TaS2\mathrm{TaS}_2 have been performed (Figure 4). For bulk and ~ 27 nm thick flake (Figure S5), the M1_1, M2_2, and M3_3 have not observed even in the low temperature range, while for flakes with dd below ~ 16 nm (Figure S5), these modes are observed clearly. The presence of M1_1, M2_2, and M3_3 indicates the existence of structural instability, which appears at temperature, TI−CDWT_{\mathrm{I-CDW}}, where the I-CDW sets up. For bulk crystal, only CDW collective modes are observed below 80 K,[36] however the structural instability and CDW gap (ΔCDW\Delta_{\mathrm{CDW}}) has also been reported at ~ 240 K using angle-resolved photoemission spectroscopy and density functional theory analysis.[21] For resistance measurement on monolayer of 2H2\mathrm{H}-TaS2\mathrm{TaS}_2, the reported TI−CDWT_{\mathrm{I-CDW}} is ~ 93 K, which is quite high as compared to bulk 2H2\mathrm{H}-TaS2\mathrm{TaS}_2 i.e. ~ 76 K.[29] Hence, the appearance of the new modes in thin layer can be inferred due to the structural instability or formation of the CDW even at higher temperatures in few-layered than relative to the bulk of 2H2\mathrm{H}-TaS2\mathrm{TaS}_2. The increased CDW transition temperature in thin layer is due to the reduced dimensionality, which enhances ee-phph coupling and has been observed in other 2D CDWs as well.

Temperature dependent Raman spectra and intensity color maps for bulk and flakes of varying thickness

Figure 4. Temperature dependent Raman spectra of (a) bulk (b) ~ 10 nm (c) ~ 7 nm (d) ~ 3 nm. Here, rectangle higlights the Raman modes asociated with I-CDW transition. The corresponding intensity color map is shown for (e) bulk, (f) ~ 10 nm (g) ~ 7 nm and (h) ~ 3 nm flakes.

Thickness-dependent Fano line shapes for three thin-layer thicknesses

Figure S4. Thickness dependence Fano line shape of M2M_2 mode for (a) ~ 10 nm (b) ~ 7 nm, and (c) ~ 3 nm.

Temperature-dependent Raman spectra for three thin-layer thicknesses

Figure S5. Temperature dependent Raman spectra of (a) ~ 27 nm, (b) ~ 16 nm and (c) ~ 5 nm thin layer.

From the appearance of the M1_1, M2_2 and M3_3 modes, the TI−CDWT_{\mathrm{I-CDW}} of the flakes has been estimated. Corresponding temperature dependent color map of Raman intensity has also been shown for bulk (Figure 4e), ~ 10 nm (Figure 4f), ~ 7 nm (Figure 4g), ~ 3 nm (Figure 4h). For all samples, the intensity of the modes increases at low temperatures. In case of ~ 3 nm, intensity map little noisy because of the substrate effects. Evidence of ee-phph coupling and commensurability of CDW can be strongly identified by the variation of ω\omega, fwhmfwhm, and IAIA of new modes, M1_1 and M2_2.

Figure 5 shows the temperature dependence of the ω\omega, and IAIA of CDW modes, M1M_{1} and M2M_{2}, for ∼10\sim10 nm, ∼7\sim7 nm and ∼3\sim3 nm thin flakes. In general the Raman modes softens at higher temperatures because of thermal and volume expansion with the temperatures.[8, 46, 51, 52] The anharmonic behavior of temperature dependence of A1gA_{1g} ∼400\sim400 cm−1^{-1} mode is shown in Figure 5a and the detailed analysis of ω\omega, fwhmfwhm, and IAIA of A1gA_{1g} mode is shown in Figure S6. On the other hand, the ω\omega and IAIA, (fwhmfwhm in Figure S7), of CDW modes (M1M_{1} and M2M_{2}) show anomalous behavior with rise in temperature (Figure. Figure 5b-Figure 5g). The anomalous behavior in the Raman spectra is a signature of the structural instabilities associated with the CDW in the flakes below TCDWT_{CDW} and is related with the formation of CDW superlattices and strong ee-phph coupling of new CDW modes. Noticeably, the IAIA for all exfoliated flakes has strong thickness dependence and reaches maxima at certain temperature (ordering at T1T_{1}) followed by a decrease at lower temperatures. This behavior is indication of interaction of density of electronic states with phonons and is related to the structural ordering at low temperature, which is also observed in VSe2_2 by Pandey. et.al.[8] The temperature dependence of the ω\omega, fwhmfwhm, and IAIA of CDW modes, M1M_{1} and M2M_{2}, for ∼16\sim16 nm, ∼5\sim5 nm thick and thin flakes, has also been shown in SI (Figure. Figure S8). Based on the temperature dependence of the IA of the M1_1 and M2_2 (Figure 5e-Figure 5g and Figure S8-Figure S8), two regions have been defined, (i) onset of the I-CDW (TI−CDWT_{I-CDW}, green color) and (ii) onset of ordered state (T1T_1, pink color). The dd dependence of TI−CDWT_{I-CDW} and T1T_1, has been summarized in Figure 5h, indicates that both transition temperature decreases with the increase in the dd of the exfoliated flakes. The observation of anomalous behavior of CDW modes (M1_1, M2_2, and M3_3) provide the substantial evidence of ee-phph coupling, which is suggesting that the CDW is strongly affected by the nanoscale dimensionality of atomically thin 2H-TaS2_2.

Temperature dependence of Raman modes and transition temperatures

Figure 5. Temperature dependence ω\omega of (a) A1gA_{1g}, and of M1M_{1} and M2M_{2} for (b) ∼10\sim10 nm, (c) ∼7\sim7 nm, and (d) ∼3\sim3 nm thin layer. Temperature dependence IAIA (e-f-g), for ∼10\sim10 nm, ∼7\sim7 nm, and ∼3\sim3 nm, respectively. (h) Thickness dependence TI−CDWT_{I-CDW} and T1T_{1} of thin layer 2H-TaS2_2. Solid red line is the fitted data with anharmonic model.

Temperature-dependent fwhm and IA of the $A_{1g}$ mode

Figure S6. Temperature dependent of (a) fwhm, and (b) IA of A1gA_{1g} ~ 400 cm−1^{-1} mode.

Figure S7. Temperature dependence of fwhm for (a) ~ 10 nm, (b) ~ 7 nm, and (c) ~ 3 nm.

Figure S8. Temperature dependence of ω\omega (a-b), fwhm (c-d), and IA (e-f) for ~ 5 nm, and ~ 16 nm, respectively.

Conclusion

In summary the confinement effects and symmetry breaking at nanoscale shows a higher TI−CDWT_{I-CDW} ~ 200 K and presence of CDW modes, M1_1 ~ 125 cm−1\mathrm{cm}^{-1} (E1E_1), M2_2 ~ 158 cm−1\mathrm{cm}^{-1} (A1A_1), and M3_3 ~ 334 cm−1\mathrm{cm}^{-1} (E2E_2), in one unit cell of exfoliated 2H-TaS2_2. The symmetries and nature of the CDW associated modes is understood from low-temperature and polarized Raman spectroscopy. The evidence of strong electron-phonon coupling has been understood from the asymmetric (Fano) line shape of M2_2, which may be giving the CDW instabilities. We observed that the TI−CDWT_{I-CDW} has inverse relation with the dd of the flakes, which suggest existence of CDW instability above TI−CDWT_{I-CDW} of bulk (~ 76 K) 2H-TaS2_2.

Associated Content

Supporting Information is available free of charge. It includes the details of synthesis, characterization, Raman measurements, AFM imaging, and additional results and discussions complementary to the main texts.

ACKNOWLEDGEMENT: We would like to acknowledge IIT Mandi for the research facility.

Symmetry breaking and structure instability in ultra-thin 2H2H-TaS2\mathrm{TaS}_2 across charge density wave transition

Divya Rawat; Aksa Thomas; Ajay Soni

Growth and characterization of single-crystal 2H2H-TaS2\mathrm{TaS}_2

Single crystal of 2H2H-TaS2\mathrm{TaS}_2 was grown by chemical vapor transport technique using iodine as a transporting agent.[8] [53] Thin flakes of crystals were mechanically exfoliated and transferred to SiO2\mathrm{SiO}_2/Si substrate after piranha cleaning. The optical images were captured using the Olympus microscope attached to the Raman spectrometer. The thickness identification of the exfoliated flakes was done by atomic force microscopy (AFM) using Bruker Dimension icon AFM. The estimated thickness of flakes is lying within an error bar of ∼1\sim1 nm. Low temperature resistance and heat capacity measurements were performed by the Quantum Design make physical properties measurement system. Raman spectroscopy was performed using Jobin Vyon LabRAM HR Evolution Raman spectrometer with Czerny-turner grating (1800 gr/mm) and Peltier cooled CCD detector in a back-scattering configuration. A 532 nm laser was used and 50x long working distance objective was used for focusing laser on the sample. Ultra-low frequency filter was used to access low-frequency Raman modes. To control the polarization state, a λ/2\lambda/2 half-waveplate and an analyzer was used in front of the objective lens and spectrometer respectively to select the desired polarization component of the incident and scattered light. Temperature-dependent Raman measurement was performed using a Montana instrument make closed-cycle cryostation. All the Raman spectra were fitted by Lorentzian and Fano model to evaluate full width at half maximum (fwhm), phonon frequency (ω\omega), and integrated area (IA) of optically active Raman modes.

Resistance and Heat-capacity measurements of bulk 2H2H-TaS2\mathrm{TaS}_2.

Evidence of the charge density wave (CDW) in bulk 2H2H-TaS2\mathrm{TaS}_2 has been observed through resistance (RR) and heat capacity (CpC_p) measurement. The broad hump in RR curve (Figure S1a), indicates the CDW phase transition and the TI-CDWT_{I\text{-}CDW} has been estimated from the dR/dTdR/dT (inset. Figure S1a). The sharp peak in the CpC_p data (Figure S1b), also strongly supports the evidence of CDW at ∼76\sim76 K.

Optical and AFM imaging with height profile of exfoliated thick flakes (∼27\sim27 nm and ∼16\sim16 nm)

Thick films of 2H2H-TaS2\mathrm{TaS}_2 having thickness ∼27\sim27 nm, and ∼16\sim16 nm have also been obtained through mechanical exfoliation. Optical and AFM images with corresponding height profile have been represented in Figure S2.

Optical images and height profiles of mechanically exfoliated thin flakes

Figure S2. Optical images of mechanically exfoliated thin flakes (a) ~ 27 nm (b) ~ 16 nm on Si/SiO₂ substrate. The height profiles of (c) ~ 27 nm (d) ~ 16 nm thin flakes and corresponding topographic AFM images are shown in respective insets.

Polarized Raman spectroscopy of ∼7\sim7 nm thin layer of 2H2H-TaS2\mathrm{TaS}_2

Low temperature Raman spectra at 10 K is intense in case of ∼7\sim7 nm (Figure 2a in manuscript). Therefore, further ∼7\sim7 nm thin flake is chosen to perform the polarization Raman spectroscopy for the identification of the symmetry of new Raman modes (M1M_1, M2M_2, and M3M_3) in thin layer of 2H-TaS2_2. Here, the laser is incident normal to the abab plane and polarization of incident laser is rotated (θ\theta) from 0∘0^\circ to 180∘180^\circ using λ/2\lambda/2 half wave-plate. In polarized Raman spectroscopy, two different configurations have been defined, parallel (XXXX) and perpendicular (XYXY) configuration.[8] In XXXX configuration (θ=0∘\theta= 0^\circ) the incident and the scattered light are polarized along the same direction, while for the XYXY, they are orthogonal to each other. Among all new Raman modes, M2M_2 is absent in XYXY configuration, while M1M_1 and M3M_3 are present (Figure S3), indicates that M1M_1 and M3M_3 are of EE- and M2M_2 is of A-symmetry Raman modes.[46]

Analysis of asymmetric (Fano) line shape of M2M_2 at 10 K.

The line-shape of M2M_2 has asymmetric shape towards lower energy showing the presence of an interesting Fano resonance, that can provide vital information about the interaction of the electronic continuum with a discrete phonon state, electron-electron interaction and electron-plasmon interaction.[45] The asymmetric line shape for all thin flakes ∼10\sim10 nm (Figure S4a), ∼7\sim7 nm, (Figure S4b) ∼3\sim3 nm (Figure S4c) (∼10\sim10 nm, ∼7\sim7 nm, and ∼3\sim3 nm) was fitted using Breit Wigner Fano (BWF) function, which can be defined by the mathematical expression, F(ω)=I0(q+ε)2(1+ε2)F(\omega) = I_0 \frac{(q+\varepsilon)^2}{(1+\varepsilon^2)}, where ε=(ω−ω0)/Γ\varepsilon= (\omega-\omega_0)/\Gamma and $1/q$ is defined as asymmetric parameter.[49] The 1/q1/q is commonly interpreted as a measure of the electron-phonon (ee-phph) coupling strength and it sign indicate the location of charge continuum, whether near valence band or conduction band. Existence of the Fano line-shape, itself provide the significant evidence of elel-phph coupling$[45], which can be interpreted as the interference of a continuum of electronic excitations with the phonon line and plays an important role behind the existence of the new modes.

Temperature dependent Raman spectra of ∼27\sim27 nm, ∼16\sim16 nm, and ∼5\sim5 nm.

To indicate the evolution of new modes (M1M_1, M2M_2, M3M_3) in thick flakes, temperature dependent Raman spectroscopy have been performed on other exfoliated flakes also (∼27\sim27 nm, ∼16\sim16 nm, and ∼5\sim5 nm), represented in Figure S5. The existence of new modes have been observed for ∼16\sim16 nm thick layer (Figure S5b) also, but absent in case of ∼27\sim27 nm (Figure S5a). From the temperature dependent Raman spectra, we can accomplish that new modes are present only in thin layer ≤16\leq16 nm. The region of temperature, where the new modes exist have been shown by the dashed rectangle.

Variation of fwhm and IA of A1gA_{1g} Raman mode for all thin flakes.

The temperature dependence of ω\omega of A1gA_{1g} mode for thin layer (∼10\sim10 nm, ∼7\sim7 nm, and ∼3\sim3 nm), represented in Figure 5a (Main manuscript) has been fitted with the anharmonic decay model using the equation: ω(T)=ωo−Aeℏωo2kBT−1\omega(T) = \omega_o - \frac{A}{e^{\frac{\hbar\omega_o}{2k_B T}}-1}; where ωo\omega_o is the phonon frequency at 0 K and AA is the anharmonic constant.[51] Experimental data points are quite well fitted with the anharmonic model (Figure 5a), indicates the purely anharmonic contribution in usual phonon, which mainly arise because of thermal and volume expansion with the temperature.[8, 51, 52] The temperature dependence of fwhm (Figure S6), IA (Figure S6) also indicates the normal anharmonic behavior because of the thermal and volume expansion.[51]

Variation of fwhm of M₁ and M₂ for ~ 10 nm, ~ 7 nm, and ~ 3 nm thin flakes.

The temperature dependence of fwhm of M₁ and M₂ for ~ 10 nm (Figure S7), ~ 7 nm (Figure S7), and ~ 3 nm (Figure S7), shown the anomalous behavior with respect to normal phonon modes, which indicates the existence of e-ph coupling.

Variation of ω\omega, fwhm, and IA of M₁ and M₂ for ~ 16 nm, ~ 5 nm thin flakes.

Figure S8 shows the temperature dependence of the ω\omega, fwhm, and IA of CDW modes, M₁ and M₂, for ~ 16 nm, and ~ 5 nm thin flakes. The TI-CDWT_{I\text{-CDW}} has been estimated from the appearance of CDW modes and TIT_I has been estimated from the temperature dependence of IA which show anomalous behavior below certain temperature. Decrease in the IA below certain temperature, indicates the decrease in the electronic density that will correspond to the existence of e-ph coupling and commensurability of the formed CDW superlattice.[8]

References

  1. [1]Vogel, E. M.; Robinson, J. A. MRS Bulletin 2015, 40, (7), 558-563.
  2. [2]Liao, W.; Zhao, S.; Li, F.; Wang, C.; Ge, Y.; Wang, H.; Wang, S.; Zhang, H. Nanoscale horizons 2020, 5, (5), 787-807.
  3. [3]Tian, H.; Chin, M. L.; Najmaei, S.; Guo, Q.; Xia, F.; Wang, H.; Dubey, M. Nano Research 2016, 9, (6), 1543-1560.
  4. [4]Jeong, G. H.; Sasikala, S. P.; Yun, T.; Lee, G. Y.; Lee, W. J.; Kim, S. O. 2020, 32, (35), 1907006.
  5. [5]Goli, P.; Khan, J.; Wickramaratne, D.; Lake, R. K.; Balandin, A. A. Nano Lett. 2012, 12, (11), 5941-5945.
  6. [6]Xi, X.; Zhao, L.; Wang, Z.; Berger, H.; Forró, L.; Shan, J.; Mak, K. F. Nature nanotechnology 2015, 10, (9), 765-769.
  7. [7]Sammakay, R.; Wickramaratne, D.; Pope, T.; Lake, R.; Salguero, T.; Balandin, A. Nano Lett. 2015, 15, (5), 2965-2973.
  8. [8]Pandey, J.; Soni, A. Phys. Rev. Res. 2020, 2, (3), 033118.
  9. [9]Hall, J.; Ehlen, N.; Berges, J.; van Loon, E.; van Efferen, C.; Murray, C.; Rösner, M.; Li, J.; Senkovskiy, B. V.; Hell, M.; Rolf, M.; Heider, T.; Asensio, M. C.; Avila, J.; Plucinski, L.; Wehling, T.; Grüneis, A.; Michely, T. ACS Nano 2019, 13, (9), 10210-10220.
  10. [10]Kyashnin, Y.; VanGennep, D.; Mito, M.; Medvedev, S.; Thiyagarajan, R.; Karis, O.; Vasiliev, A.; Eriksson, O.; Abdel-Hafiez, M. Phys. Rev. Lett. 2020, 125, (18), 186401.
  11. [11]Grasset, R.; Gallais, Y.; Sacuto, A.; Cazayous, M.; Mañas-Valero, S.; Coronado, E.; Méasson, M.-A. Phys. Rev. Lett. 2019, 122, (12), 127001.
  12. [12]Lin, D.; Li, S.; Wen, J.; Berger, H.; Forró, L.; Zhou, H.; Jia, S.; Taniguchi, T.; Watanabe, K.; Xi, X.; Bahramy, M. S. Nature Communications 2020, 11, (1), 2406.
  13. [13]Navarro-Moratalla, E.; Island, J. O.; Manas-Valero, S.; Pinilla-Cienfuegos, E.; Castellanos-Gomez, A.; Quereda, J.; Rubio-Bollinger, G.; Chirolli, L.; Silva-Guillén, J. A.; Agrait, N. Nature Communications 2016, 7, (1), 11043.
  14. [14]Yu, Y.; Yang, F.; Lu, X. F.; Yan, Y. J.; Cho, Y.-H.; Ma, L.; Niu, X.; Kim, S.; Son, Y.-W.; Feng, D. Nature nanotechnology 2015, 10, (3), 270-276.
  15. [15]Jin, C.; Kim, J.; Suh, J.; Shi, Z.; Chen, B.; Fan, X.; Kam, M.; Watanabe, K.; Taniguchi, T.; Tongay, S.; Zettl, A.; Wu, J.; Wang, F. Nature Physics 2017, 13, (2), 127-131.
  16. [16]Butler, S. Z.; Hollen, S. M.; Cao, L.; Cui, Y.; Gupta, J. A.; Gutiérrez, H. R.; Heinz, T. F.; Hong, S. S.; Huang, J.; Ismach, A. F.; Johnston-Halperin, E.; Kuno, M.; Plashnitsa, V. V.; Robinson, R. D.; Ruoff, R. S.; Salahuddin, S.; Shan, J.; Shi, L.; Spencer, M. G.; Terrones, M.; Windl, W.; Goldberger, J. E. ACS Nano 2013, 7, (4), 2898-2926.
  17. [17]Grüner, G. Reviews of Modern Physics 1988, 60, (4), 1129-1181.
  18. [18]Wilson, J.; Di Salvo, F.; Mahajan, S. Phys. Rev. Lett. 1974, 32, (16), 882.
  19. [19]Moncton, D. E.; Axe, J.; DiSalvo, F. Phys. Rev. B 1977, 16, (2), 801.
  20. [20]Anikin, A.; Schaller, R.; Wiederrecht, G.; Margine, E.; Mazin, I.; Karapetrov, G. Phys. Rev. B 2020, 102, (20), 205139.
  21. [21]Joshi, J.; Hill, H. M.; Chowdhury, S.; Malliakas, C. D.; Tavazza, F.; Chatterjee, U.; Hight Walker, A. R.; Vora, P. M. Phys. Rev. B 2019, 99, (24), 245144.
  22. [22]Diego, J.; Said, A. H.; Mahatha, S. K.; Bianco, R.; Monacelli, L.; Calandra, M.; Mauri, F.; Rossnagel, K.; Errea, I.; Blanco-Canosa, S. Nature Communications 2021, 12, (1), 598.
  23. [23]Chan, S.-K.; Heine, V. Journal of Physics F: Metal Physics 1973, 3, (4), 795.
  24. [24]Hill, H. M.; Chowdhury, S.; Simpson, J. R.; Rigosi, A. F.; Newell, D. B.; Berger, H.; Tavazza, F.; Hight Walker, A. R. Phys. Rev. B 2019, 99, (17), 174110.
  25. [25]Tsutsumi, K. Phys. Rev. B 1982, 26, (10), 5756.
  26. [26]Zhao, X.-M.; Zhang, K.; Cao, Z.-Y.; Zhao, Z.-W.; Struzhkin, V. V.; Goncharov, A. F.; Wang, H.-K.; Gavriliuk, A. G.; Mao, H.-K.; Chen, X.-J. Phys. Rev. B 2020, 101, (13), 134506.
  27. [27]Meetsma, A.; Wiegers, G.; Haange, R.; De Boer. Acta. Cryst. 1990, 46, (9), 1598-1599.
  28. [28]Shi, J.; Wang, X.; Zhang, S.; Xiao, L.; Huan, Y.; Gong, Y.; Zhang, Z.; Li, Y.; Zhou, X.; Hong, M. Nature communications 2017, 8, (1), 958.
  29. [29]Zhang, D.; Wu, Y.; Su, Y.-H.; Hsu, M.-C.; Ó Coileáin, C.; Cho, J.; Choi, M.; Chun, B. S.; Guo, Y.; Chang, C.-R. IUCrJ 2020, 7, (5), 913-919.
  30. [30]Zhang, D.; Ha, J.; Baek, H.; Chan, Y.-H.; Natterer, F. D.; Myers, A. F.; Schumacher, J. D.; Cullen, W. G.; Davydov, A. V.; Kuk, Y. Phys. Rev. Mat. 2017, 1, (2), 024005.
  31. [31]Wu, X.; Cai, Y.; Bian, J.; Su, G.; Luo, C.; Yang, Y.; Zhang, G. RSC Advances 2020, 10, (28), 16718-16726.
  32. [32]Xi, X.; Berger, H.; Forró, L.; Shan, J.; Mak, K. F. Phys. Rev. Lett. 2016, 117, (10), 106801.
  33. [33]Fumega, A. O.; Diego, J.; Pardo, V.; Blanco-Canosa, S.; Errea, I. Nano Lett. 2023, 23, (5), 1794-1800.
  34. [34]Lawan Adam, M.; Buba Garba, I.; Alhaji Bala, A.; Aji Suleiman, A.; Muhammad Gana, S.; Lawan Adam, F. Phys. Rev. B 2023, 107, (10), 104510.
  35. [35]Kolekar, S.; Bonilla, M.; Diaz, H. C.; Hashimoto, M.; Lu, D.; Batzill, M. Advanced Quantum Technologies 2018, 1, (3), 1800070.
  36. [36]Zhang, K.; Cao, Z.-Y.; Chen, X.-J. Applied Physics Letters 2019, 114, (14), 141901.
  37. [37]Sugai, S.; Murase, K.; Uchida, S.; Tanaka, S. Solid State Communications 1981, 40, (4), 399-401.
  38. [38]Klein, M. Phys. Rev. B 1981, 24, (8), 4208.
  39. [39]Lee, C.; Yan, H.; Brus, L. E.; Heinz, T. F.; Hone, J.; Ryu, S. ACS Nano 2010, 4, (5), 2695-2700.
  40. [40]Berkdemir, A.; Gutiérrez, H. R.; Botello-Méndez, A. R.; Perea-López, N.; Elías, A. L.; Chia, C.-I.; Wang, B.; Crespi, V. H.; López-Urías, F.; Charlier, J.-C.; Terrones, H.; Terrones, M. Scientific Reports 2013, 3, (1), 1755.
  41. [41]Li, Z.; Wang, Y.; Jiang, J.; Liang, Y.; Zhong, B.; Zhang, H.; Yu, K.; Kan, G.; Zou, M. Nano Research 2020, 13, (2), 591-595.
  42. [42]Hajiyev, P.; Cong, C.; Qiu, C.; Yu, T. Scientific Reports 2013, 3, (1), 2593.
  43. [43]Liang, L.; Puretzky, A. A.; Sumpter, B. G.; Meunier, V. Nanoscale 2017, 9, (40), 15340-15355.
  44. [44]Valencia-Ibáñez, S.; Jiménez-Guerrero, D.; Salcedo-Pimienta, J. D.; Vega-Bustos, K.; Cárdenas-Chirivi, G.; López-Manrique, L.; Herrera-Vasco, E.; Galvis, J. A.; Hernández, Y.; Giraldo-Gallo, P. Advanced Electronic Materials 2022, 8, (10), 2200457.
  45. [45]Zhang, J.; Peng, Z.; Soni, A.; Zhao, Y.; Xiong, Y.; Peng, B.; Wang, J.; Dresselhaus, M. S.; Xiong, Q. Nano Lett. 2011, 11, (6), 2407-2414.
  46. [46]Rawat, D.; Singh, A.; Singh, N. K.; Soni, A. Phys. Rev. B 2023, 107, (15), 155203.
  47. [47]Malard, L. M.; Pimenta, M. A.; Dresselhaus, G.; Dresselhaus, M. S. Physics Reports 2009, 473, (5), 51-87.
  48. [48]Gupta, R.; Xiong, Q.; Adu, C. K.; Kim, U. J.; Eklund, P. C. Nano Lett. 2003, 3, (5), 627-631.
  49. [49]Zhang, W. L.; Li, H.; Xia, D.; Liu, H. W.; Shi, Y. G.; Luo, J. L.; Hu, J.; Richard, P.; Ding, H. Phys. Rev. B 2015, 92, (6), 060502.
  50. [51]Singh, N. K.; Rawat, D.; Dey, D.; Elsukova, A.; Persson, P. O. Å.; Eklund, P.; Taraphder, A.; Soni, A. Phys. Rev. B 2022, 105, (4), 045134.
  51. [52]Pandey, J.; Ghoshal, D.; Dey, D.; Gupta, T.; Taraphder, A.; Koratkar, N.; Soni, A. Phys. Rev. B 2020, 102, (20), 205308.

Paper details

vlnc:W_eq6e7krb
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