Session E2: Computational Methods for Improved Transparent Conducting Oxides

Structure and Properties of Amorphous Transparent Conducting Oxides

Driven by technological appeal, the research area of amorphous oxide semiconductors has grown tremendously since the first demonstration of the unique properties of amorphous indium oxide more than a decade ago. Today, amorphous oxides, such as a-ITO, a-IZO, a-IGZO, or a-ZITO, exhibit the optical, electrical, thermal, and mechanical properties that are comparable or even superior to those possessed by their crystalline counterparts, pushing the latter out of the market. Large-area uniformity, low-cost low-temperature deposition, high carrier mobility, optical transparency, and mechanical flexibility make these materials appealing for next-generation thin-film electronics. Yet, the structural variations associated with crystalline-to-amorphous transition as well as their role in carrier generation and transport properties of these oxides are far from being understood. Although amorphous oxides lack grain boundaries, factors like (i) size and distribution of nanocrystalline inclusions; (ii) spatial distribution and clustering of incorporated cations in multicomponent oxides; (iii) formation of trap defects; and (iv) piezoelectric effects associated with internal strains, will contribute to electron scattering. In this work, ab-initio molecular dynamics (MD) and accurate density-functional approaches are employed to understand how the properties of amorphous ternary and quaternary oxides depend on quench rates, cation compositions, and oxygen stoichiometries. The MD results, combined with thorough experimental characterization, reveal that interplay between the local and long-range structural preferences of the constituent oxides gives rise to a complex composition-dependent structural behavior in the amorphous oxides. The proposed network models of metal-oxygen polyhedra help explain the observed intriguing electrical and optical properties in In-based oxides and suggest ways to broaden the phase space of amorphous oxide semiconductors with tunable properties.   

Quasiparticle energies, excitonic effects, and dielectric screening in transparent conducting oxides

Using the power of high-performance super computers, computational materials scientists nowadays employ highly accurate quantum-mechanical approaches to reliably predict materials properties. In particular, many-body perturbation theory is an excellent framework for performing theoretical spectroscopy on novel materials including transparent conducting oxides, since this framework accurately describes quasiparticle and excitonic effects.\\ We recently used hybrid exchange-correlation functionals and an efficient implementation of the Bethe-Salpeter approach to investigate several important transparent conducting oxides. Despite their exceptional potential for applications in photovoltaics and optoelectronics their optical properties oftentimes remain poorly understood: Our calculations explain the optical spectrum of bixbyite indium oxide over a very large photon energy range, which allows us to discuss the importance of quasiparticle and excitonic effects at low photon energies around the absorption onset, but also for excitations up to 40 eV. We show that in this regime the energy dependence of the electronic self energy cannot be neglected. Furthermore, we investigated the influence of excitonic effects on optical absorption for lanthanum-aluminum oxide and hafnium oxide. Their complicated conduction band structures require an accurate description of quasiparticle energies and we find that for these strongly polar materials, a contribution of the lattice polarizability to dielectric screening needs to be taken into account. We discuss how this affects the electron-hole interaction and find a strong influence on excitonic effects.\\ The deep understanding of electronic excitations that can be obtained using these modern first-principles techniques, eventually will allow for computational materials design, e.g.\ of band gaps, densities of states, and optical properties of transparent conducting oxides and other materials with societally important applications.   

Limits of transparency of transparent conducting oxides

A fundamental understanding of the factors that limit transparency in transparent conducting oxides (TCOs) is essential for further progress in materials and applications. These materials have a sufficiently large band gap, so that direct optical transitions do not lead to absorption of light within the visible spectrum. Since the presence of free carriers is essential for conductivity and thus for device applications, this introduces the possibility of additional absorption processes. In particular, indirect processes are possible, and these will constitute a fundamental limit of the material. The Drude theory is widely used to describe free-carrier absorption, but it is phenomenological in nature and tends to work poorly at shorter wavelengths, where band-structure effects are important. \\ \\ We will present calculations of phonon- and defect-assisted free-carrier absorption in a TCO completely from first principles [1]. We will focus in detail on SnO$_2$, but the methodology is general and we will also compare the results obtained for other TCO materials such as In$_2$O$_3$. These calculations provide not just quantitative results but also deeper insights in the mechanisms that govern absorption processes, which is essential for engineering improved materials to be used in more efficient devices.\newline\newline [1] H. Peelaers, E. Kioupakis, and C.G. Van de Walle, Appl. Phys. Lett. 100, 011914 (2012).   

High-throughput search for improved transparent conducting oxides

High-throughput methodologies are a very useful computational tool to explore the space of binary and ternary oxides. We use these methods to search for new and improved transparent conducting oxides (TCOs). TCOs exhibit both visible transparency and good carrier mobility and underpin many energy and electronic applications (e.g. photovoltaics, transparent transistors). We find several potential new n-type and p-type TCOs with a low effective mass. Combining different ab initio approaches, we characterize candidate oxides by their effective mass (mobility), band gap (transparency) and dopability. We present several compounds, not considered previously as TCOs, and discuss the chemical rationale for their promising properties. This analysis is useful to formulate design strategies for future high mobility oxides and has led to follow-up studies including preliminary experimental characterization of a p-type TCO candidate with unexpected chemistry. G. Hautier, A. Miglio, D. Waroquiers, G.-M. Rignanese, and X. Gonze, “How Does Chemistry Influence Electron Effective Mass in Oxides? A High-Throughput Computational Analysis”, Chem. Mater. 26, 5447 (2014). G. Hautier, A. Miglio, G. Ceder, G.-M. Rignanese, and X. Gonze, “Identification and design principles of low hole effective mass p-type transparent conducting oxides”, Nature Commun. 4, 2292 (2013).   

Hybrid functional studies of defects and hole polarons in oxides.

Transparent conducting oxides (TCOs) are ubiquitous, appearing in windows, flat-panel displays, solar cells, solid-state lighting, and transistors that all exploit TCOs' combination of high electrical conductivity and optical transparency. Thanks to this large and growing list of applications, there has been a surge of interest in the science of these materials, focusing on the fundamental properties and doping opportunities in traditional TCOs as well as the exploration of promising new candidate materials. Hybrid density functional theory has proven instrumental in elucidating the physics of TCOs. One example is the study of dopants and defects that determine the conductivity. Accurate formation energies and charge-state transition levels can now be obtained thanks to the accurate electronic structure provided by a hybrid functional. This allows us to address the origins of unintentional conductivity: for SnO$_{\mathrm{2}}$, In$_{\mathrm{2}}$O$_{\mathrm{3}}$, and Ga$_{\mathrm{2}}$O$_{\mathrm{3}}$, we demonstrate that this is \textit{not} due to native defects such as oxygen vacancies, but must be attributed to unintentional incorporation of impurities. We can also provide guidelines for achieving higher doping levels, suggesting several impurities as candidate donors with high solubility. Limitations on doping due to the formation or incorporation of compensating centers are addressed as well. Hybrid functional calculations also overcome the shortcomings associated with traditional local or semi-local functionals, which do not properly describe charge localization. Hybrid functionals accurately describe polaron formation, i.e., the self-trapping of holes when $p-$type doping of the oxide materials is attempted. Consequences of polaron formation for optical characterization of the material will be discussed. This work was performed in collaboration with Anderson Janotti and Chris G. Van de Walle, and was in part under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.