Surface electronic structure for various surface preparations of Nb-doped SrTiO₃ (001)

Abstract

Difference in electronic structure between tetragonal and cubic Sr Nb O 2 N J. Appl. Phys. 98, 043706 (2005); 10.1063/1.2032612High-resolution synchrotron-radiation photoemission characterization for atomically-controlled SrTiO 3 ( 001 ) substrate surfaces subjected to various surface treatments High-resolution angle-resolved photoemission spectroscopy (ARPES) was used to study the surface electronic structure of Nb-doped SrTiO 3 (STO) single crystals prepared using a variety of surface preparations. ARPES measurements show that simple degreasing with subsequent anneal in vacuum is not an adequate surface preparation of STO, rather, preparations consisting of etching with buffered HF or HCl, and to a lesser extent, simple water leaching resulted in surfaces with much less disorder. A non-dispersing, mid-gap state was found $800 meV above the top of the valence band for samples which underwent etching. This mid-gap state is not present for vacuum-annealed and water-leached samples, as well as for STO thin films grown using molecular beam epitaxy. Theoretical modeling using density functional theory suggests that this mid-gap state is not related to the SrO-and TiO 2 -terminated surfaces, but rather, is due to a partial hydrogenation of the STO surface that occurs during etching. V C 2013 AIP Publishing LLC.

Final-state effect on x-ray photoelectron spectrum of nominallyd1andn-dopedd0transition-metal oxides

“…The 2p 1/2 peak is significantly broader than the 2p 3/2 peak due to Coster-Kronig transitions [28]. [31][32][33]. In all these n-doped SrTiO 3 , a small shoulder located about 1.5 eV lower than the Ti 4+ peak emerges, and is typically interpreted as a Ti…”

“…n core = 0), all TM d levels are shifted down by U dc to screen the core hole. By fitting to published experimental data from XPS of SrTiO 3 , ellipsometry, and angle-resolved photoemission spectroscopy (ARPES) [15,32,[41][42][43], we take ǫ d (e g ) = 2.0 eV, ǫ d (t 2g ) = 0 eV, ǫ p (Γ) = −3.0 eV, V (e g ) = 2.5eV, V (t 2g ) = −1.3 eV, U = 6.0 eV, and U dc = 8.0 eV [44]. This problem can be solved exactly by the technique introduced by Gunnarsson and Schönhammer [12,45], and the details are provided in Appendix B.…”

Electron Correlation in Oxygen Vacancy inSrTiO3

“…The OV in SrTiO 3 is particularly intriguing. On the one hand, the OV concentration is roughly proportional to that of the carriers [14,20,21], strongly suggesting that vacancies are electron donors; on the other hand, an OV produces an in-gap signal, peaked approximately 1.0 eV below the conduction band, in angle resolved photoemission spectroscopy (ARPES), even at temperatures as low as 20 K [14,19,22]. Within the single-particle description, the ARPES spectrum implies that an OV results in a deep impurity level and therefore is not likely to be the electron donor.…”

“…3. By fitting the ARPES bands [19,40], we choose ε d − ε p = 4 eV (ε p ≡ −0.8 eV so that the valence band top is at zero energy), t pd = 1.5 eV, and t pp = 0.2 eV. This choice leads to the experimental band gap of 3.2 eV.…”

Surface-hydrogen-induced metallization and rumpling in thin BaTiO3 films

“…33,34 In Figure 4a, we see that the charge obtained from the adsorbed H is present mostly on the surface Ti state, which has been lowered from the middle of the conduction band and now forms its bottom; the reason for this drop in energy will be discussed in Section IV. Although all of the surface Ti d states are occupied from the local density of states (LDOS), the majority of the charge is located in the d xy band, with a lesser amount in the d yz and d xz bands, and a smaller amount still in the e g bands.…”