Search

We have employed hard x-ray photoemission (HAXPES) to study a delta-doped SrTiO3 layer that consisted of a 3-nm thickness of La-doped SrTiO3 with 6% La embedded in a SrTiO3 film. Results are compared to a thick, uniformily doped La:SrTiO3 layer. We find no indication of a band offset for the delta-doped layer, but evidence of the presence of Ti3+ in both the thick sample and the delta-layer, and indications of a density of states increase near the Fermi energy in the delta-doped layer. These results further demonstrate that HAXPES is a powerful tool for the non-destructive investigation of deeply buried doped layers. [ABSTRACT FROM AUTHOR]

The chemical and electronic modifications induced by the electroforming process on the Ti/HfO2/TiN-based resistive switching devices were investigated by non-destructive hard x-ray photoelectron spectroscopy (HAXPES). The results indicate an increase of the titanium top electrode oxidation at the interface with HfO2 after the electroforming process. Additionally, the binding energy values of the HAXPES peaks of the electroformed sample increased as compared to the as-prepared sample. The observed changes between both samples are attributed to the creation of n-type defects, such as oxygen vacancies, in the HfO2 layer near the Ti interface during the electroforming process. [ABSTRACT FROM AUTHOR]

A key benefit of angle-resolved photoelectron spectroscopy (ARPES) in the X-ray range is the significant increase of the information depth, thanks to the large inelastic mean-free-path of the escaping photoelectrons. In practice hard X-ray ARPES (HARPES) faces severe challenges by low cross sections, large photon momentum transfer, and in particular strong phonon scattering and photoelectron diffraction effects. Here, we show that these challenges can be overcome by extending ultra-efficient time-of-flight momentum microscopy into the hard X-ray regime. Phonon scattering destroys the initial momentum distribution but subsequent diffraction at the lattice imprints a pronounced Kikuchi-type pattern on the background signal. Moreover, the pattern of the valence electrons is modulated by diffraction as well. For the examples of the medium-weight element materials Mo and layered TiTe2, we demonstrate how comprehensive valence-band and core-level photoemission data taken under identical conditions can be used to effectively remove photoelectron diffraction effects in HARPES band maps. The study of electronic structure of new materials has benefited from more widely available angle-resolved photoelectron spectroscopy (ARPES) at synchrotron sources, but hard X-ray ARPES, capable of mapping at a depth of some tens nanometres is still of limited access. The authors report on a method to obtain bulk electronic structure using hard X-rays ARPES combined with an effective data processing background removal strategy capable of revealing the valence band electronic dispersion of metal and semiconductor surfaces. [ABSTRACT FROM AUTHOR]

Abstract: We used hard X-ray photoelectron spectroscopy (HAXPES) with 8keV X-rays to investigate the 1s emission of carbon. We recorded spectra extending from the peak of the C 1s electrons (“elastic” line) to electrons with up to 110eV energy-loss. Using two samples side by side, we could compare the inelastic mean free paths (IMFPs) of the electrons of almost 8keV in diamond and graphite and find them to be practically identical despite about 50% difference in densities. Published extrapolations of their IMFP calculations at lower energies are in good agreement with this result. We show that information from the almost structureless region of overlapping multiple extrinsic energy-losses can be used to quantify the fraction of photoelectrons experiencing intrinsic energy-losses (those due to the sudden creation of the hole). We find that this fraction is 58% of the primary excited C 1s electrons for diamond and is practically the same for graphite. This is at first sight an unexpected result since hole-screening should differ in a semimetal from that in an insulator. The observation can be accounted for by dynamic screening in contrast to static screening. [Copyright &y& Elsevier]

Abstract: The multilevel resistive switching effect in Pt/HfxAl1−xOy/TiN stacks has been studied by hard X-ray photoelectron spectroscopy (HAXPES). Atomic layer deposition was used to grow HfxAl1−xOy films with graded Al depth profile, where the engineered Al concentration across the film facilitates the desired oxygen vacancies' profile. The method to derive the electrical potential profile across dielectric from the HAXPES spectra is proposed. By combining the information on the chemical state at both interfaces with the extracted potential distribution, a qualitative model for the resistive switching effect in the Pt/HfxAl1−xOy/TiN metal–insulator–metal stack is proposed. According to this model, the conductive filaments controlling the resistivity are formed by the electric field driven condensation of the charged oxygen vacancies in the oxide. [Copyright &y& Elsevier]

Highlights: [•] Electronic conditions at interfaces of functional multilayers probed by HAXPES. [•] In situ monitoring of chemical changes in MIM stacks undergoing resistance switching. [•] Potential barrier profile in functional FTJ experimentally determined by HAXPES. [•] VO present in ultrathin BaTiO3 layer affects band line-up at the interface. [ABSTRACT FROM AUTHOR]

We present a detailed study of the electronic structure and chemical state of high-quality stoichiometric EuO and O-rich Eu1 O1+x thin films grown directly on silicon without any buffer layer using hard x-ray photoemission spectroscopy (HAXPES). We determine the EuO oxidation state from a consistent quantitative peak analysis of 4f valence band and 3d core-level spectra. The results prove that nearly ideal, stoichiometric, and homogeneous EuO thin films can be grown on silicon, with a uniform depth distribution of divalent Eu cations. Furthermore, we identify the chemical stability of the EuO/silicon interface from Si 2p core-level photoemission. This work clearly demonstrates the successful integration of high-quality EuO thin films directly on silicon, opening up the pathway for the future incorporation of this functional magnetic oxide into silicon-based spintronic devices. [ABSTRACT FROM AUTHOR]

Abstract: The spatial distribution of charges in a Pt/HfO2/Si stack has been manipulated by applying a cyclic bias voltage ±2.5V in combination with moderate (T ∼630K) heating. The modifications were monitored in situ by room temperature capacitance–voltage (C–V) and current–voltage (I–V) measurements and analyzed ex situ by hard X-ray photoelectron spectroscopy which additionally provides information on possible chemical changes at the interfaces. The experimental data on the charge/potential distributions resulting from the different steps of bias-temperature stress (BTS) are consistent with the model that additional oxygen vacancies, which are generated in HfO2 and positively charged by charge transfer across the interface with a high work function metal (Pt), are driven across the HfO2 layer. These vacancies ultimately control the observed growth/dissolution of SiO x at the bottom interface upon negative/positive BTS, respectively. [Copyright &y& Elsevier]

Abstract: A method is presented for analysing the energy loss part (dominated by contributions from extrinsic, intrinsic and surface plasmon excitations) of electron spectra excited from free-electron-like solids by X-rays. It is an improved version of the model originally proposed by Steiner, Höchst and Hüfner and it accounts for contributions of electrons suffering multiple energy losses of different origin. Applied for analysing a Ge 2s photoelectron spectrum excited from a thin Ge layer, the applicability of the method to separate contributions to the inelastic background of the spectra due to intrinsic, extrinsic and surface plasmons is demonstrated. The aim of this work to give a detailed description of the simple statistical model applied in ref. [M. Novák, S. Egri, L. Kövér, I. Cserny, W. Drube, W.S.M. Werner, Surf. Sci. 601 (2007) 2344] for practical analysis of photoelectron spectra. The effects of the application of energy loss distributions (for single bulk and single surface excitations) derived by using different methods are also investigated. [Copyright &y& Elsevier]