Enhancement of the Seebeck coefficient (S ) without reducing the electrical conductivity (sigma) is essential to realize practical thermoelectric materials exhibiting a dimensionless figure of merit (ZT=S2 x sigma x T x kappa-1) exceeding 2, where T is the absolute temperature and kappa is the thermal conductivity. Here, we demonstrate that a high-density two-dimensional electron gas (2DEG) confined within a unit cell layer thickness in SrTiO(3) yields unusually large |S|, approximately five times larger than that of SrTiO(3) bulks, while maintaining a high sigma2DEG. In the best case, we observe |S|=850 microV K-1 and sigma2DEG=1.4 x 10(3) S cm-1. In addition, by using the kappa of bulk single-crystal SrTiO(3) at room temperature, we estimate ZT approximately 2.4 for the 2DEG, corresponding to ZT approximately 0.24 for a complete device having the 2DEG as the active region. The present approach using a 2DEG provides a new route to realize practical thermoelectric materials without the use of toxic heavy elements.

In many subfields of chemistry and physics, numerous attempts have been made to accelerate scientific discovery using data-driven experimental design algorithms. Among them, Bayesian optimization has been proven to be an effective tool. A standard implementation (e.g., scikit-learn), however, can accommodate only small training data. We designed an efficient protocol for Bayesian optimization that employs Thompson sampling, random feature maps, one-rank Cholesky update and automatic hyperparameter tuning, and implemented it as an open-source python library called COMBO (COMmon Bayesian Optimization library). Promising results using COMBO to determine the atomic structure of a crystalline interface are presented. COMBO is available at https: //github.com/tsudalab/combo.

Determining both cation and oxygen sublattices of grain boundaries is essential to understand the properties of oxides. Here, with scanning transmission electron microscopy, electron energy-loss spectroscopy, and first-principles calculations, both the Ce and oxygen sublattices of a (210)Σ5 CeO(2) grain boundary were determined. Oxygen vacancies are shown to play a crucial role in the stable grain boundary structure. This finding paves the way for a comprehensive understanding of grain boundaries through the atomic scale determination of atom and defect locations.

Theoretical calculations of electron energy-loss near-edge structure (ELNES) and x-ray absorption near-edge structure (XANES) of selected wide-gap materials including TiO 2 , AlN, GaN, InN, ZnO, and their polymorphs are performed using a first-principles method. Calculations of 39 K and L 3 ͑L 2,3 ͒ edges are made using large supercells containing 72 to 128 atoms. A core hole is included in the final state, and the matrix elements of the electric dipole transition between the ground state and the final state are computed. Structures of some metastable crystals are optimized by a plane-wave basis pseudopotential method. Spectral differences in ELNES and XANES among polymorphs are quantitatively reproduced in this way. The origin of the spectral differences is pursued from the viewpoint of chemical bondings. Crystallographic orientation dependence of ELNES and XANES is also examined both by experiment and theory. The dependence is found to be much larger in K edges than that in L 3 ͑L 2,3 ͒ edges.

Determining the atomic structures of oxide surfaces is critical for understanding their physical and chemical properties but also challenging because the breaking of atomic bonds in the formation of the surface termination can involve complex reconstructions. We used advanced transmission electron microscopy to directly observe the atomic structure of reduced titania (TiO2) (110) surfaces from directions parallel to the surface. In our direct atomic-resolution images, reconstructed titanium atoms at the top surface layer are clearly imaged and are found to occupy the interstitial sites of the TiO2 structure. Combining observations from two orthogonal directions, the three-dimensional positioning of the Ti interstitials is identified at atomic dimensions and allows a resolution of two previous models that differ in their oxygen stoichiometries.