Photoemission near the Fermi energy in one dimension

Abstract

Angle-resolved photoemission spectra taken with high energy resolution on the one-dimensional cornpound (TaSe&)zI confirm the suppression of the spectral intensity at the Fermi edge for one-dimensional systems and lend support to correlation effects as its most probable cause.The photoemission spectrum of nearly-onedimensional materials like (TaSez)2I is a very interesting problem in condensed-matter physics. Several experiments have suggested a very low spectral intensity near the Fermi energy EF. ' Dardel et al. examined this striking result with very high energy resolution, concluding that the effect is most likely not related to a low quasiparticle density of states near EF, but rather to a significant difference between the excitation spectrum and the density of states.Several mechanisms were proposed to explain the suppression of spectral intensity near EF, but no final evidence in favor of one of them could be extracted from angle-integrated photoemission data. %e did perform high-energy-resolution, angle-resolved studies of (Tase4)2I. The results strongly favor one of the hypotheses formulated in Ref. 2: the suppression of the photoemission intensity near EF is due to correlation effects that drive an infrared catastrophe, as suggested by Schulz in the Luttinger model framework.The samples studied in our experiments were similar to those of Ref. 2; the most significant differences in the procedure were the use in our study of monochromatized synchrotron radiation, and a high angular resolution of +0. 5'. The energy resolution in the different runs of our experiments ranged from 25 to 50 meV. Typical experimental results are shown in Figs. 1 and 2. In Fig. 1 we see a series of high energy and angular resolution spectra, taken at a constant photon energy of 22 eV, and at different azimuthal angles. The spectra exhibit two important features: first, a dispersive peak that does not cross the Fermi energy, but reaches its maximum energy, approximately 0.5 eV below EF, at a k vector corresponding to the border of the first Brillouin zone (ki =0.245 A ' in the direction of the crysta1).Second, we also see that the photoemission spectral intensity is very weak near EF, as was observed in the angle-integrated spectra of Ref. 2. Note that the intensity is weak, but not zero. This is clear from Fig. 2, which shows the spectral region near EF. For all directions in Fig. 2, we do see detectable signal up to the Fermi level, but never an indication of a band crossing a Fermi edge. Several photon energies were used in the experiments, in the range 15 -45 eV. The results of these tests indicate that the suppression of the near-edge spectral signal is not a trivial final-state effect. Similarly, experiments performed with different photon polarizations rule out polarization effects as a possible cause of the low intensity near 46 13 624

One-dimensional Fermi liquids

“…As in the organic conductors, the single-particle density of states shows no significant weight at the Fermi energy in the blue bronze K 0.3 MoO 3 , and appreciable spectral weight is only found a sizable fraction of an eV below [293]. Angle-resolved studies on related materials either show a fading away of the signal as the Fermi surface is approached [294], or weight dispersing some finite energy below E F [295]. Clearly, there have been speculations about 1D correlations at the origin of the mysterious behaviour.…”

Surface characterization and surface electronic structure of organic quasi-one-dimensional charge transfer salts

“…Indeed, ͑quasi-͒1D metals were found to display marked deviations from conventional metallic behavior using ͑AR͒PES. [2][3][4][5][6][7][8][9][10][11][12][13] Basically all 1D materials studied so far show no clear Fermi cutoff. Only recently did we obtain convincing evidence for spin-charge separation in the charge-transfer salt TTF-TCNQ based on an analysis within the 1D Hubbard model.…”

Spectral functions for the tomonaga-luttinger and luther-emery liquids

“…ARPES measurements on the Bechgaard salts 12 , the transition metalchalcogenides 13 , the blue bronzes 14 and the cuprates 15 have provided evidence in favor of correlated one-dimensional physics although full agreement with the predicted theoretical picture is still lacking.…”