-(BEDT-TTF)2I3 is a layered organic material, and an exotic material which has massless Dirac fermion dispersion at the Fermi level similarly to graphene. Its main difference from graphene is that it is a three-dimensional bulk material and that its Dirac cone is markedly tilted. Such nature shows interesting physical properties, which have been studied through experiments and theories. In organic conductors, the semi-empirical extendedHuckel approximation usually gives good parameters in the tight binding model at ambient pressure. One of the reasons why it can give a good Fermi surface is that organic conductors usually have simple and large Fermi surfaces. For -(BEDT-TTF)2I3, however, the electronic structure is very delicate, and small difference can change the electronic structure near the Fermi level greatly. Another possible source of error in the semi-empirical approximation is the fact that the correctness of the pressure dependence has not been well studied. First-principles study does not have such shortcomings. The superiority of the dispersion given by the first-principles theory to that given by the semi-empirical theory has been confirmed by Kobayashi et al. to explain the transport properties. In our previous study, we have shown that the conducting plane is the a b -plane and that the Dirac fermion dispersion exists in the P 1 symmetry near ambient pressure, where experiments have observed the charge ordered state in the P1 symmetry at low temperatures. The main difference between the phases of the P 1 and P1 symmetry is that the charges on the A and A0 molecules are not equivalent in the latter phase, which inevitably affects the electronic structures near the Fermi level. Therefore, we must confirm whether or not the Dirac fermion dispersion appears at about 2GPa, where theoretical analyses predict its existence. The pressure dependence of the Dirac fermion dispersion is also demanded in order to analyze the physical properties of such dispersion. Our calculation employs the exchange–correlation function in the generalized gradient approximation (PW91) in the density functional theory. Technical parameters in the calculation are written in refs. 15, 20, and 21. We calculate electronic structures in two cases; we use the unit cells from the X-ray measurement, and we determine the unit cells theoretically with the constraint of the P 1 symmetry. We list the parameters of the dispersion in the experimental unit cell, but we mainly discuss them in the theoretical unit cell. We stress that theoretical pressure is hydrostatic. We briefly compare theoretical cells with experimental cells. The theoretical lattice constants a, b, and c at ambient pressure are slightly larger than the experimental ones (3, 2, and 1%, respectively), while the differences in the angles , , and are less than 1%. The situation is also the same under pressure. The changes in a, b, and c are 5:5, 5:0, and 2:8% in the experiment, while they are 5:1, 4:6, and 3:5% in the theory. A question is sometimes posed whether or not the experimentally applied pressure is really hydrostatic. Considering this dependence on pressure, the experimental change is almost the same as the theoretical change. Therefore, we conclude that the experimental pressure is also hydrostatic. Next, we analyze the Dirac fermion dispersion. Previous calculations have shown that the electronic structure of this material is almost two-dimensional. Therefore, we concentrate on the electronic structure on the conducting a b -plane. (The axis with the superscript means the reciprocal axis.) Following Kobayashi et al. and Morinari et al., the tilted Dirac fermion dispersion is given by diagonalizing Hðkx; kyÞ 1⁄4 ðv0k0 þ v y 0kyÞ1þ vxkx x þ vyky y, where 1 is a unit 2 2matrix and x and y are Pauli matrices, kx ky 1⁄4 cosð 0Þ sinð 0Þ sinð 0Þ cosð 0Þ k0 x k0 y !
Read full abstract