Investigation of renormalization effects in high temperature cuprate superconductors

While in conventional superconductors coupling between electrons and phonons is known to be responsible for the electron pairing, for the high temperature superconductors the pairing media remains under debates. Since the interactions of electrons with other degrees of freedom (phonons, magnetic excitations, etc) manifest themselves by an additional renormalization in the electronic dispersion, they can be investigated by means of Angle Resolved Photoelectron Spectroscopy. In the work renormalization in two families of high Tc cuprates have been studied. Along the diagonal of the two-dimensional BZ, the renormalization effects are represented by an unusual band dispersion that develops a so-called ‘‘kink’’. In the vicinity of the (pi, 0) point of the BZ, where the order parameter reaches its maximum, the renormalization is noticeably stronger and makes itself evident even in the shape of a single spectral line measured for a fixed momentum. It was shown that for the Bi-2212 samples substitution of Cu atoms in Cu-O plane changes renormalization features in ARPES spectra both in nodal and antinodal parts of the Brillouin zone. The smearing of the dip in the in the spectral line shape measured at (pi; 0) point can be well explained by coupling of electrons to the magnetic resonance mode. The effect of Zn and Ni substitution on the antinodal ARPES spectra was shown to be in good agreement with the influence of these impurities on magnetic resonance mode seen in inelastic neutron scattering experiments. This, in addition to the previous ARPES studies of temperature and doping dependence of peak-dip-hump structure, mass renormalization near antinodal region and a kink in the nodal part of Brillouin zone, provides further evidence that the coupling to magnetic excitations, rather than to phonons, is responsible for the observed unusual renormalization. Unlike the well studied Bi-2212 family of cuprates, photoemission on YBCO-123 turns out to be much more complicated. The observed spectra have a strong contribution from a heavily overdoped surface component with the hole doping level of about x~0.30, which is weakly dependent on the sample stochiometry. Absence of any signs of superconductivity in the spectra of the overdoped component was argued to result from the unusually high doping level. This conclusion is supported by the fact that the overdoped bands give rise to the Fermi surface and band structure consistent with the predictions of the LDA calculations, as well as, by the dependence of the photoemission matrix element on the excitation energy, which closely follows that of the superconducting bulk component. Specific experimental geometry was used to enhance the signal coming from the superconducting component. In particular, experiments with circularly polarized light bundled with simple theoretical considerations enabled better separation of the surface and the bulk components. This type of experiments also suggests that the overdoped component is mainly localized in the topmost CuO2 bilayer, while the next bilayers in the YBCO-123 structure already represent bulk properties and retain superconductivity. Using partially Ca substituted samples it was possible to obtain spectra with a suppressed overdoped component. The likely reason for the suppression is a shift of the most probable cleavage plane from the Ba–O interface to the Y layer. Spectra from the Ca substituted sample clearly reveal a sizable superconducting gap, and strong renormalization effects in the vicinity of the antinodal point. The fact that the renormalization vanishes above Tc and has strong momentum dependence, diminishing away from the (pi; 0)/(0; pi) point, strongly suggests that the reason for this renormalization in YBCO-123 is coupling of the electronic subsystem to spin resonance, similar to the case of Bi-2212