The controllability over strongly correlated electronic states promises unique electronic devices. A recent example is an optically induced ultrafast switching device based on the transition between the correlated Mott insulating state and a metallic state of a transition metal dichalcogenide 1T-TaS2. However, the electronic switching has been challenging and the nature of the transition has been veiled. Here we demonstrate the nanoscale electronic manipulation of the Mott state of 1T-TaS2. The voltage pulse from a scanning tunnelling microscope switches the insulating phase locally into a metallic phase with irregularly textured domain walls in the charge density wave order inherent to this Mott state. The metallic state is revealed as a correlated phase, which is induced by the moderate reduction of electron correlation due to the charge density wave decoherence.

, superconductivity [6][7][8][9] , and discommensurations [10][11][12][13][14][15] . Intercalation of other transition metal ions between the MC 2 layers gives rise to distinct superstructures, leading to significant changes in crystallographic structures and physical properties. Fe-intercalated TaS 2 shows highly anisotropic ferromagnetism at low temperatures [16][17][18][19][20] . . Magnetic hysteresis curves of the crystals were obtained using a Quantum DesignMagnetic Property Measurement System, and the real Fe compositions were estimated from the saturation magnetic moments in the magnetic hysteresis curves with the assumption that each Supplementary Information, section S1). The distinct feature between the 2a×2a and √3a×√3a superstructures in Fe x TaS 2 is the different stacking sequence of the 2D supercells along the c-axis. Specifically, the 2a×2a superstructure consists of identically stacked 2D supercells (i.e., AA-type stacking), while the √3a×√3a superstructure contains shifted 2D supercells with AB-type stacking, as shown in Fig. 1b and Fig. 1f, -5 -respectively. These different stacking sequences result in the centrosymmetric P6 3 /mmc and noncentrosymmetric and chiral P6 3 22 space groups for the 2a×2a and √3a×√3a superstructures, respectively.We found complicated configurations of antiphase domains in the dark-field images of There is an extinction rule for the dark-field images of antiphase boundaries in the 2a×2a superstructure. For example, the antiphase boundary between the BB-type and CC-type antiphase domains appears in the S1=( /2 00) (Fig 2a) and S2=(0 1/2 0) (Fig. 2b) dark-field images, but disappears in the S3=(1/2 /2 0) dark-field image of Fig. 2c (see also Supplementary Information, section S3). Each antiphase boundary becomes invisible in a dark-field image taken using one out of three superlattice spots (namely, S1, S2, or S3), when no antiphase shift at the boundary exists along a certain superlattice modulation wave vector. This absence of antiphase shifts at the antiphase boundary leads to the extinction rule for the antiphase boundaries in superlattice dark-field images. This rule is summarized in Fig. 3, showing the local structures near boundaries between two antiphase domains. The boundaries are highlighted with yellow bands, and the three directions of the superlattice modulation wave vectors are denoted by S1, S2, and S3, respectively, as shown in Fig. 1a. The red, yellow, blue, and green circles correspond to -6 -AA-, BB-, CC-, and DD-type superstructures, respectively, which are associated with four possible origins of the 2a×2a Fe superstructure. It is evident that the superlattice modulation along only one out of three equivalent crystallographic directions does not show any antiphase shift; this is indicated by light green dashed lines (along the S1 direction), light blue dashed lines (along the S2 direction), and pink dashed lines (along the S3 direction). For example, the antiphase boundary between BB-type and CC-type (or AA-type and DD-type) antiphase domains ha...

We investigate the interplay of the electron-electron and electron-phonon interactions in the electronic structure of an exotic insulating state in the layered dichalcogenide 1T-TaS2, where the charge-density-wave (CDW) order coexists with a Mott correlation gap. Scanning tunneling microscopy and spectroscopy measurements with high spatial and energy resolution determine unambiguously the CDW and the Mott gap as 0.20 -0.24 eV and 0.32 eV, respectively, through the real space electron phases measured across the multiply formed energy gaps. An unusual local reduction of the Mott gap is observed on the defect site, which indicates the renormalization of the on-site Coulomb interaction by the electron-phonon coupling as predicted by the Hubbard-Holstein model. The Mott-gap renormalization provides new insight into the disorder-induced quasi-metallic phases of 1T-TaS2.PACS numbers: 71.10. Hf, 71.20.Be, 71.27.+a, 71.30.+h Metal-insulator transitions in low dimensional condensed matter systems are driven by various interactions between relevant degrees of freedom, such as spin, charge, orbital, and lattice. For example, charge density waves (CDW) [1] and Mott insulators [2] are paradigmatic examples of electron-phonon (e-ph) and electronelectron (e-e) couplings, respectively. These couplings are often entangled to yield exotic states of electrons as recently discussed for high-temperature superconductors [3]. On the other hand, the interplay of e-ph and e-e couplings has been discussed for a while for the CDW transition [4] in a layered transition metal dichalcogenide of 1T-TaS 2 [5]. Upon decreasing temperature, it undergoes a series of transitions from a metallic phase through incommensurate and nearly commensurate CDW to commensurate (C)-CDW [4]. The ordered CDW superstructures could be directly observed using scanning tunneling microscopy (STM) in real space [6,7]. The C-CDW superstructure splits the broad metallic band into a manifold of narrow subbands but cannot account for the most important gap opening at the Fermi level (E F ) with one half-filled band left [8]. The insulating gap was then explained by introducing the on-site Coulomb repulsion, which is enhanced by the substantial narrowing of the band width due to the CDW formation [5]. Thus, the multiple gap structure in this system itself is a hallmark of the interplay between e-ph and e-e couplings.However, while this material has been investigated for a long time, the multiple gap structure is not fully characterized yet by experiments. The gap structure was investigated by angle resolved photoemission spectroscopy (ARPES) [9], inverse-ARPES [10] and scanning tunneling spectroscopy (STS) [7] below the critical temperature (T c ∼180 K). ARPES probed only the occupied part of the gap and inverse-ARPES for the empty part did not provide proper energy resolution. The previous STS study only focused on the gap at E F without any clear information on the multiple gaps. Moreover, these spectroscopic works reported substantially larger gaps than those measure...