Electronic properties of bulk and thin filmSrRuO3: Search for the metal-insulator transition

Abstract

We calculate the properties of the 4d ferromagnet SrRuO 3 in bulk and thin film form with the aim of understanding the experimentally observed metal-to-insulator transition at reduced thickness. Although the spatial extent of the 4d orbitals is quite large, many experimental results have suggested that electron-electron correlations play an important role in determining this material's electronic structure. In order to investigate the importance of correlation, we use two approaches which go beyond the conventional local-density approximation to density-functional theory ͑DFT͒: the local spin-density approximation+ Hubbard U ͑LSDA+ U͒ and the pseudopotential self-interaction correction ͑pseudo-SIC͒ methods. We find that the details of the electronic structure predicted with the LSDA do not agree with the experimental spectroscopic data for bulk and thin film SrRuO 3 . Improvement is found by including electron-electron correlations, and we suggest that bulk orthorhombic SrRuO 3 is a moderately correlated ferromagnet whose electronic structure is best described by a 0.6 eV on-site Hubbard term, or equivalently with corrections for the self-interaction error. We also perform ab initio transport calculations that confirm that SrRuO 3 has a negative spin polarization at the Fermi level, due to the position of the minority Ru 4d band center. Even with static correlations included in our calculations we are unable to reproduce the experimentally observed metal-insulator transition, suggesting that the electronic behavior of SrRuO 3 ultrathin films might be dominated by extrinsic factors, such as surface disorder and defects, or due to dynamic spin correlations which are not included in our theoretical methods.

Ab initioand photoemission study of correlation effects in SrRuO3thin films

“…14 which demonstrates that, although the localized states are not present in DFT calculations, a significantly broad component (∼40% of total intensity) arises in the convoluted Ru 4d PDOS at BE corresponding the reported data of the incoherent peak position. 9,14,25 It can be concluded that both the finite admixture of O 2p states in the dominant Ru 4d region and the asymmetric shape of the coherent Ru 4d feature, also obtained in some previous theoretical studies, 14,25,28,30 may contribute to the experimentally observed broad incoherent feature at ∼1.2-1.5 eV, thus suppressing the relative importance of localized surface states.…”

“…The sharp peak near the Fermi level (A) and the broad one centered at ∼1.2 eV (B) are usually assigned to the coherent and incoherent parts of the spectral function, respectively. 9,14,25 The peaks D and E can be attributed to the mixture of Ru 4d and O 2p orbitals. The finite emission at 8-11 eV (negligible peak F) is due to the contaminations.…”

“…Moreover, Siemons et al 19 have concluded that a tiny change in ruthenium stoichiometry may have a fairly considerable effect on the PES spectrum, indicating a highly stoichiometry-dependent degree of correlation for this material. Maiti et al 9 have proposed that the evidence of strong correlation effects in SrRuO 3 might be attributed to the change of symmetry at the surface, whereas Rondinelli et al 25 have suggested that electronic behavior of SrRuO 3 thin films might be affected by extrinsic factors, such as surface roughness or defects. In addition, one should also take into account the fact that physical properties of this perovskite are sensitive to the strain 26 and thin film thickness.…”

Designing a robustly metallic noncenstrosymmetric ruthenate oxide with large thermopower anisotropy

“…Nonetheless, the weakly dispersed t 2g states persist, and (Sr,Ca)Ru 2 O 6 remains metallic [N(E F )B4.36 eV per Ru atom] consistent with SOI calculations on bulk SrRuO 3 (ref. 35), where spin-orbit coupling also does not greatly alter the electronic structure.…”

First-Principles Study of the Thermoelectric Properties of SrRuO3

“…13 Theoretically, the thermoelectric properties of SrTiO 3 /SrRuO 3 superlattices have been studied 14 and a metallic non-censtrosymmetric ruthenate double-perovskite oxide SrCaRu 2 O 6 with large thermopower anisotropy has been proposed. 15 Although the electronic, structural, magnetic and lattice dynamical properties of bulk SRO have already been extensively explored theoretically, [16][17][18][19][20][21] to the best of our knowledge, the thermoelectric aspects has not been reported. In spite of the fact that the thermoelectric performances of SRO appear relatively modest and do not make it a very promising candidate for technological applications, it remains a very important magnetic oxide compound.…”