We investigate atomic and electronic structures of the intriguing low temperature phase of IrTe2 using high-resolution scanning tunneling microscopy and spectroscopy. We confirm various stripe superstructures such as ×3, ×5, and ×8. The strong vertical and lateral distortions of the lattice for the stripe structures are observed in agreement with recent calculations. The spatial modulations of electronic density of states are clearly identified as separated from the structural distortions. These structural and spectroscopic characteristics are not consistent with the charge-density wave and soliton lattice model proposed recently. Instead, we show that the Ir (Te) dimerization together with the Ir 5d charge ordering can explain these superstructures, supporting the Ir dimerization mechanism of the phase transition.Strong spin-orbital coupling (SOC) has been widely recognized as the source of new physics in condensed matter systems such as various topological phases [1][2][3][4][5][6][7][8][9][10][11] and the SOC-induced Mott insulating state [12]. As a general strategy, it would be interesting to study how various quantum phases change under the influence of a strong SOC. In this respect, IrTe 2 is a very attractive materials. First of all, Ir has a strong SOC. It was shown to have a unique quasi 1D charge-density-wave-(CDW)-like ground state and the superconductivity emerges through the electron doping with a possibility of the quantum critical behavior [13]. It is natural to expect unusual behaviors in the CDW and superconducting states. Indeed, the recently observed superconductivity of Pt-or Pd-doped IrTe 2 was suspected as being non-conventional or topological [13,14] but a very recent scanning tunneling spectroscopy (STS) study showed the conventional s-wave pairing behavior [15]. While further investigations are called for the nature of the superconductivity, the experimental evidence has been accumulated to indicate the unusual nature of the CDW-like phase transition (T c ≈ 260 K without doping) with a wave vector of (1/5, 0, 1/5) (the ×5 phase, hereafter) [16]. Most importantly, the band gap opening does not occur but the strong restructuring of Fermi surfaces and band structures exists [17][18][19]. Extra intriguing aspects were also revealed such as the charge ordering in Ir 5d orbitals [20], the dimer-like distortions in both Ir and Te layers, and the depolymerization in the Te interlayer bondings [21,22]. At present, what is clear is the importance of the interplay of various different degrees of freedoms, structures for both Ir and Te layers, charges, orbitals, and bonds and to pin down the major driving force of this phase transition remains as a challenging task.The charge-order or CDW nature of the ground state can best be checked by scanning tunneling microscopy and spectroscopy (STM/STS) studies since STM/STS can directly probe the spatial modulation of the local density of states (LDOS). Indeed, a very recent STM study identified various stripe superstructures of ×3, ×5, ×8, and ×11 [23] in t...