Supporting Information

Transcription

1 Supporting Information High-Mobility Bismuth-based Transparent P-Type Oxide from High-throughput Material Screening Amit Bhatia 1,, Geoffroy Hautier 2,,*, Tan Nilgianskul 1, Anna Miglio 2, Jingying Sun 3, Hyung Joon Kim 4, Kee Hoon Kim 4, Shuo Chen 3, Gian-Marco Rignanese 2, Xavier Gonze 2, Jin Suntivich 1,* 1 Materials Science and Engineering Department, Cornell University, Ithaca, New York 14850, United States of America 2 Institute of Condensed Matter and Nanosciences, Université catholique de Louvain, Louvain-la-Neuve 1348, Belgium 3 Department of Physics, and Texas Center for Superconductivity, University of Houston, Houston, Texas, 77204, United States of America 4 Center for Novel States of Complex Materials Research, Department of Physics and Astronomy, Seoul National University, Seoul, , South Korea These authors contribute equally to this work * Correspondence: geoffroy.hautier@uclouvain.be, jsuntivich@cornell.edu S1

2 Experimental Methods Ab initio calculations. The high-throughput screening was performed following the methodology reported previously 1,2. Density functional theory (DFT) in the generalized gradient approximation (GGA) 3 with the VASP software 4 was used to assess hole effective masses. We used the materials project database 5,6 as our starting data set focusing on quaternary oxides. Follow-up computations using the more accurate single-shot GW method was performed to compute the material s band-gap. GW computations were performed using ABINIT and normconserving pseudopotentials 7. Oxide Synthesis. Ba 2-x K x BiTaO 6 (with x ranging from 0 to 0.7) was prepared using a solid state synthesis method. Briefly, stoichiometric amounts of BaCO 3 (99.95%, Alfa Aesar), Bi 2 O 3 (99.999%, Alfa Aesar) and Ta 2 O 5 (99.993%, Alfa Aesar) were thoroughly ground with agate mortar and pestle. Then, the mixture was pressed into a pellet using a hydraulic press (Grimco) and heat treated in a furnace at 1100 C for 12 hours under flowing argon (Ultrahigh purity, Airgas) atmosphere. Thin Film Deposition. Pulsed laser deposition (PLD, Neocera) was used to deposit thin films of Ba 2-x K x BiTaO 6 onto (100) MgO single-crystalline substrates (MTI). A pulsed laser beam from a KrF excimer laser (λ = 248 nm) was focused onto the target that was formed from the re-ground and re-pressed the oxide powders. The fluence energy and repetition rate of the laser were adjusted to achieve a deposition rate of ~ 5 nm/min. The substrate temperature and oxygen pressure were maintained at 650 C and torr. After the deposition, the films were annealed at 400 C for 5 hours in an oxygen pressure of 1 torr. S2

3 Structural and Optical Characterizations. Crystal structures of the powder samples and films were characterized by X-ray Diffraction (Scintag Theta-Theta X-ray Diffractometer, Bruker General Area Detector Diffraction System, and Rigaku SmartLab X-ray Diffractometer) using Cu Kα radiation (λ = Å). Optical transmission measurements were performed using UV-Vis-NIR Spectrometer (Shimazdu). Transport Measurement. Powder samples of BKBT were pressed into pellets of ~1 mm thickness, which was then cut into square bar specimens (~7 7 mm) using diamond saw. The conductivity was measured using van der Pauw technique (Keithley) on contacts made from Agink covered Ga-In eutectics (Sigma-Aldrich). Hole carrier density (n p ) and Hall mobility (µ H ) were determined using Physical Property Measurement System (PPMS, Quantum Design) with Au/Ti electrodes at room temperature. Transmission Electron Microscopy (TEM) Characterization. Powder samples of BKBT were first dispersed in ethanol by ultrasonication. Then one droplet of each dispersion was placed onto individual TEM grid for observation with a JEOL 2010F TEM. The images and diffraction patterns were acquired with an acceleration voltage of 200 kv. S3

4 Valence Band Positions vs. Vacuum Level To assess the typical nature of the doping, we performed a slab computation on the (100) orientation of the rhombohedral Ba 2 BiTaO 6 (BBT) structure. We constructed the slab with the procedure presented in Sun et al. S8. The BBT slab computation was performed within GGA-PBE and we added the valence and conduction band correction from single-shot GW. Valence and conduction band edges can be used to assess dopability S9. Comparing BBT's valence band edge to typical p-type TCOs (e.g., CuAlO 2 and SnO) and n-type TCOs (e.g., In 2 O 3 and ZnO), BBT is expected to behave more as a p-type than n-type material. Computations on a (110) slab gave similar results. Figure S1. Conduction band and valence band levels vs vacuum for BBT compared to typical n- type TCOs (In 2 O 3 and ZnO) as well as p-type TCOs (CuAlO 2 and SnO). S4

5 Orbital overlap analysis of the rhombohedral Ba 2 BiTaO 6 structure We performed an orbital overlap analysis in the COHP framework using the lobster software S10-12 on our VASP GGA-PBE results. Figure S2 shows that the valence band is formed by antibonding between Bi 6s and O 2p states. Figure S2. (Left) Density of state from a projection on localized basis sets using the lobster software on GGA-PBE results, (center) COHP analysis, and (right) schematic molecular orbital diagram for the Bi 6s O 2p hybridization. S5

6 Additional TEM images for Ba2BiTaO6 and Ba1.3K0.7BiTaO6 Figure S3. (a) Low magnification TEM image of Ba2BiTaO6 (BBT). (b) HRTEM image and (c) the selected area electron diffraction indexed along the rhombohedral crystallographic axis. (a) Low magnification TEM image of Ba1.3K0.7BiTaO6. (b) HRTEM image and (c) the selected area electron diffraction indexed along the rhombohedral crystallographic axis. S6

7 Indirect optical transition in Ba 2-x K x BiTaO 6 We observe a systematic reduction in the BKBT band-gap (both direct, Fig. 3B, and indirect) with increasing K with the Tauc plot analysis. Figure S4. Plot of a square root of an absorption coefficient with photon energy, showing a decreasing indirect optical transition with K + substitution. Computed absorption coefficient for Ba 2 BiTaO 6 We computed the absorption coefficient from the frequency dependent dielectric matrix within the random phase approximation (RPA) in VASP within GGA-PBE. The k-point grid was 16x16x16. A scissor operator was applied to reproduce the GW band gap (3.8 ev). The computation shows no forbidden transition and an absorption onset present at 3.8 ev. S7

8 Figure S5. Computed absorption coefficient vs energy. The computation is done within RPA and with GGA-PBE with a scissor shift to 3.8 ev electronic band gap. Hall measurement on Ba 1.3 K 0.7 BiTaO 6 Figure S6. Hall voltages as a function of magnetic field. The slope was used to estimate the carrier concentration, which was calculated to be ~5 x cm -3 (yielding the hole mobility of ~33 cm 2 /V s). In a separate measurement on a sample synthesized independently, the carrier concentration was calculated to be ~9 x cm -3 (yielding the hole mobility of ~35 cm 2 /V s). S8

9 References for the Supporting Information 1. Hautier, G., Miglio, A., Ceder, G., Rignanese, G. M., & Gonze, X., Identification and design principles of low hole effective mass p-type transparent conducting oxides. Nat. Commun. 4, (2013). 2. Hautier, G., Miglio, A., Waroquiers, D., Rignanese, G. M., & Gonze, X., How Does Chemistry Influence Electron Effective Mass in Oxides? A High-Throughput Computational Analysis. Chem. Mat. 26, (2014). 3. Perdew, J. P., Burke, K., & Ernzerhof, M., Generalized gradient approximation made simple. Phys. Rev. Lett. 77, (1996). 4. Kresse, G. & Furthmuller, J., Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 54, (1996). 5. Jain, A. et al., Commentary: The Materials Project: A materials genome approach to accelerating materials innovation. APL Mater. 1, (2013). 6. Ong, S. P. et al., Python Materials Genomics (pymatgen): A robust, open-source python library for materials analysis. Comput. Mater. Sci. 68, (2013). 7. Gonze, X. et al., ABINIT: First-principles approach to material and nanosystem properties. Comput. Phys. Commun. 180, (2009). 8. Sun, W. H. & Ceder, G., Efficient creation and convergence of surface slabs. Surf. Sci. 617, (2013). 9. Robertson, J. & Clark, S. J., Limits to doping in oxides. Phys. Rev. B 83, (2011). 10. Dronskowski, R. & Blochl, P. E., Crystal orbital Hamilton populations (COHP): energyresolved visualization of chemical bonding in solids based on density-functional calculations. J. Phys. Chem. 97, (1993). 11. Deringer, V. L., Tchougreeff, A. L., & Dronskowski, R., Crystal Orbital Hamilton Population (COHP) Analysis As Projected from Plane-Wave Basis Sets. J. Phys. Chem. A 115, (2011). 12. Maintz, S., Deringer, V. L., Tchougreeff, A. L., & Dronskowski, R., Analytic Projection From Plane-Wave and PAW Wavefunctions and Application to Chemical-Bonding Analysis in Solids. J. Comput. Chem. 34, (2013). S9