Effect of hole doping on the electronic structure ofNd1−xSr

Abstract

Photoelectron spectroscopy has been employed to follow the development of the electronic structure in Nd, "Sr"TiO3 as a function of x. Resonant effects at the Ti 3p threshold were used to highlight Ti 3d contributions in the valence bands. Particular attention is focused on Ti 3d intensity within -3 eV of the Fermi level (Ez). The total intensity in this region is found to correlate linearly with composition, as expected. Changes in the shape of this spectral structure are discussed in terms of the one-electron spectral function. For x (0.25 (insulating or semiconducting compositions), only incoherent intensity associated with the lower Hubbard band is present. Additional intensity, attributed to a coherent quasiparticle contribution, appears at the composition of the metal-insulator transition x -0.25. The relative intensities of these two components are determined as a function of x. Ti 3d intensity also occurs in the higher binding-energy region of the predominantly 0 2p band due to Ti 3d -0 2p hybridization.Composition-dependent changes in this region are analyzed in order to determine changes in the Ti 3d -O 2p hybridization with composition and structure.

Photoemission study of the metallic state of lightly electron-doped SrTiO3

“…From an analysis of optical-conductivity spectra [5,6] and transport and magnetic measurements [7] of Sr 1Àx La x TiO 3 (0 6 x 6 1), it was concluded that this system behaves as a Fermi liquid for small doping (x < 0:5). Even for very small x (down to x ¼ 0:1), however, this conclusion contradicts most previous photoemission results on metallic Sr 1Àx R x TiO 3 (R ¼ La [8,9] and Nd [10,11]) and metallic Ba 1Àx R x TiO 3Àd (R ¼ Y, La, and Nd [12,13]). It was shown that a prominent state induced by doping electron carriers appears at the binding energy of about 1.5 eV below the Fermi level (E F ), which cannot be explained by a simple picture of d-band filling or by the Fermi liquid behavior as suggested by the optical spectra.…”

“…From a comparison between the UPS spectra for the fractured surface and those for the scraped surface, it was shown that the spectral intensity of the metallic state is considerably suppressed by scraping the surface. Because of the extreme sensitivity to the surface preparations, most previous photoemission studies on the scraped surfaces have not shown the metallic state clearly [8][9][10][11][12][13].…”

Coherent and Incoherent Excitations of Electron-DopedSrTiO3

“…To this end, we adopt a local description of the electronic structure: the coherent states and IGSs originate from Ti sites initially having nominally-d 1 configuration (the same initial state to that of the Ti 3+ peak in the Ti 2p 3/2 spectrum). This interpretation is supported by the observation [8,14] that the spectral weight of the region including the coherent states and the IGSs scales with x, which is similar to the Ti 3+ -component weight in the Ti 2p 3/2 spectra scaling with x [15]. Then, the mixed character of p and d in the IGSs can be explained as the final state composed of d 1 L (L denotes a hole in the O 2p ligand orbital) and d 0 , since the former has p character while the latter has d character.…”

“…The Ti 2p 3/2 spectrum of STO:Nb showed a weak Ti 3+ peak at ∼2 eV below the main Ti 4+ peak at E B = 459.5 eV. From the intensity ratio of Ti 3+ /Ti 4+ peaks [14,15], we deduced x = 0.06±0.02. This value was consistent with x < 0.1 estimated from the Ti 2p XAS line-shape broadening [ Fig.…”

“…However, a Hubbard model, which produces incoherent states through electron correlation as shown in Fig. 1(b), is insufficient to explain the IGSs [7,14,15] since they are observed already at x ∼ 0.05 [10] where electron correlation is expected to be small [6]. The origin of the IGSs at x ∼ 0 was attributed to chemical disorder [11], polaron effect [16], or partly to donor levels [9], but no consensus has been reached yet (see p. 1171 of [17]).…”