Session Z10: Insulators and Dielectrics: Dielectric, Thermal, and Transport Properties

Dielectric properties of cluster-deposited TiO$_{2}$ nanocomposites

TiO$_{2}$-polymer nanocomposites are expected to have a high dielectric permittivity of TiO$_{2}$ and large breakdown strength of the polymer, resulting in high energy density suitable for energy storage devices. Since chemically prepared nanocomposites tend to have poor film quality and inhomogeneities due to agglomeration, cluster deposition technique was used to prepare monodispersed TiO$_{2}$--paraffin nanocomposite films. TiO$_{2}$ clusters were coated in-flight with paraffin wax having comparatively better dielectric strength (7.9 -- 11.8 MV/m) using a thermal evaporation source in prior to deposition. Bare TiO$_{2}$ clusters with average particle size ranging from 8 to 12 nm having a maximum dielectric permittivity of 54 were obtained. The structural and dielectric properties of these nanocomposites with varying volume fractions will be discussed. This research is supported by ONR and NCMN.   

Lattice Dielectric and Thermodynamic Properties of Yttria Stabilized Zirconia (YSZ) Solids

A study of lattice dielectric and thermodynamic properties of Yttria Stabilized Zirconia (YSZ) solids as a function of yttria concentration is reported. Within the local density approximation (LDA) and the harmonic approximation, we find excellent agreement between calculated and experimental specific heat and dielectric constants. From the variation of the specific heat of YSZ with yttria composition, we propose a simple additivity rule that estimates the dependence of the specific heat of YSZ on yttria concentration. Whereas for the dielectric constants of YSZ, the values are bounded by the dielectric constants of the cubic and amorphous zirconia.   

Space-Charge-Limited Conduction Under Trap Density Gradient Exhibiting Bulk-Limited Diode

Space-charge-limited (SCL) conduction in the presence of a trap density gradient is studied theoretically, which is compared with experiments in detail. Under this condition, the current ($J)$ -- voltage ($V)$ characteristics at low voltage are ohmic and symmetric with respect to the bias polarity. At high voltage, \textit{JV} characteristics follow the $J \propto \quad V^{m}$ (1 $\le \quad m \quad \le $ 2) law at both polarities and are asymmetric with respect to the bias polarity. These characteristics have not reported in the previous studies and agree well with experiments. This agreements verify that a bulk-limited conduction can exhibit rectifying, i.e., diode-like \textit{JV} characteristics without relying on Schottky barriers and diffusion currents. The theory is presented in easily tractable algebraic recurrence formulae and reproduces experimental \textit{JV }characteristics excellently using three free parameters, only one more than that used in standard SCL theory.   

Optical second-harmonic generation measurements of porous low-k dielectric materials

Low-$k$ dielectric materials based on porous carbon-doped oxides, with relative dielectric constants as low as 2.1, are widely used as thin insulating films in the microelectronics industry. Knowledge of these materials' basic electronic properties, such as energy gaps, barrier heights, and trap states, is essential for modeling their electrical leakage and stability characteristics. We use femtosecond laser pulses to probe the dynamics of charge-carrier transfer processes across Si/LKD interfacial barriers by optical second harmonic generation (SHG). Larger electric fields from multiphoton injection can be developed in Si/LKD systems compared to Si/SiO$_{2}$, indicating a significantly higher density of traps in the LKD. This is consistent with previously reported measurements of trap density by photoinjection techniques$^{\ast }$. We will also discuss results on the dynamics of discharging and on the dependence of charging phenomena on layer thickness. $^{\ast }$J. M. Atkin, D. Song, T. M. Shaw, E. Cartier, R. B. Laibowitz, and T. F. Heinz, \textit{J. Appl. Phys.} \textbf{103}, 094104 (2008).   

Energy-band parameters of atomic-layer-deposited Al$_{2}$O$_{3}$ and HfO$_{2}$ on InxGa$_{1-x}$As

X-ray photoelectron spectroscopy (XPS) combined with reflection electron energy loss spectroscopy (REELS) were used to determine the energy-band parameters, valence-band offsets $\Delta $E$_{V}$, conduction-band offsets $\Delta $E$_{C}$, and energy-band gaps E$_{g}$, of the atomic layer deposited (ALD) high k dielectrics of Al$_{2}$O$_{3}$ and HfO$_{2}$ on InxGa$_{1-x}$As (x=0, 0.15, 0.25, and 0.53). Using REELS, Eg values of the ALD-Al$_{2}$O$_{3}$ and -HfO$_{2}$ were estimated to be 6.77 and 5.56 $\pm$0.05 eV, respectively. The $\Delta $E$_{C}$'s and $\Delta $E$_{V}$'s are larger than 1.5 and 2.5 eV, respectively, for all the ALD-oxide/In$_{x}$Ga$_{1-x}$As samples. The $\Delta $Ec values obtained from the HR-XPS and REELS analyses are in good agreement with those estimated from the electrical measurement according to Fowler-Nordheim tunneling. The results are valuable to the understanding and modeling of the III-V high k MOS devices.   

Resonant photo-ionization of charged oxygen vacancy defects in Si/SiO2/HfO2 film stacks observed by second-harmonic generation.

The semiconductor industry recently achieved a historic milestone with the introduction of high-k gate dielectrics. However, much effort continues to focus on characterizing defects in these materials that promote charge trapping. Using internal multi-photon photoemission (IMPE) and time-dependent electrostatic field-induced second harmonic (TD-EFISH) generation, we probe the charge trapping kinetics in Si/SiO2/HfO2 gate stacks. During IMPE charging, the TD-EFISH response of the HfO2 film uniquely increases and then decreases when irradiated at the characteristic three-photon energy of 4.71 eV. The decrease in TD-EFISH is explained by resonant three-photon excitation of HfO2-induced negatively charged oxygen vacancy defects in the SiO2 interfacial layer, and subsequent removal of its negative charge by tunneling to the Si substrate. This interpretation is supported by spectroscopic ellipsometry and electron spin resonance measurements, and ab initio calculations, which identify the 4.71 eV transition with this defect. Photo-ionization of this defect also explains hysteresis in the TD-EFISH response when charges trapped at the surface are quenched between successive IMPE charging cycles. The results demonstrate that second harmonic generation can potentially be used as an in situ, real time monitor of charge trapping kinetics prior to device fabrication.   

Combined theoretical and experimental study of thin hafnia films

Hafnia-based dielectric films have replaced silica as a gate dielectric in field effect transistors. We present a joint experimental and theoretical study of ultra thin hafnia films grown on Si (001) by atomic layer deposition. Using density functional theory we investigate the surface energy of monoclinic and tetragonal hafnia films in search for thermodynamic means of controlling the film microstructure. Our calculations of the surface phase diagram reveal that in the absence of hydrogen (111), and ($\bar {1}11)$ are the lowest energy surface terminations of monoclinic hafnia under a wide range of chemical environment. On the other hand, the structural analysis, indicates films with thickness of 4 nm or less to be polycrystalline, predominantly monoclinic with the texture axis being the normal to the ($21\bar {1})$, ($11\bar {2})$. Our calculations suggest that under oxygen rich conditions the ($11\bar {2})$ termination can be stabilized. Furthermore, we discuss the effects of the hydroxylation on the thermodynamics of the hafnia film grown by ALD and provide a new perspective into the dynamics of the film growth.   

Resonant Localized Nanoplasma in Oblique Far Infrared Reflectivity of Transition Metal Granular Films

We report on near normal and angle dependent specular infrared reflectivity of transition metal and SiO$_{2}$ cosputtered nanogranular $\sim $550 nm thick films in the insulating regime. Their reflectivity is characterized by well defined vibrational bands, an overdamped Drude contribution, due to carriers denoting the existence of conducting critical paths not yet truncated, and a distinctive band at $\sim $1450cm$^{-1}$ originating in electron promotion, localization, and polaron formation. (TM) P-polarized oblique reflectivity, as from globally insulating Co$_{0.38}$(SiO$_{2})_{0.62}$, Fe$_{0.34}$(SiO$_{2})_{0.66 }$ or Ni$_{0.28}$(SiO$_{2})_{0.72}$, reveals a remarkable resonance at the $\sim $1450cm$^{-1}$ band threshold. Its maximum intensity is reached at the radiation tangential component null condition allowing for a collective electronic excitation induced as localized plasma. It is attributed to carriers that are not able to overcome the metal-dielectric rough interface. As the angle of incidence increases the longitudinal highest frequency vibrational band merges with the P-polarized resonance inducing broadening and softening reminiscent to lattice modes undergoing strong electron-phonon interactions.   

Phase Equilibrium of Size-Dispersed Colloid Systems with soft pair interactions: A Monte Carlo Study

We have studied the solid-fluid coexistence for systems of polydisperse soft spheres that interact via power-law potentials. We employed isobaric semi-grand ensemble simulations. Gibbs-Duhem integration traced the coexistence pressure as a function of the variance of the imposed activity distribution. Fluid-solid coexistence densities were determined to be monotonically increasing functions of the breadth of particle size dispersity. Our simulations testified to the existence of terminal diameter dispersity, i.e a dispersity above which there can be no amorphous-crystalline phase coexistence. At the terminus, size-dispersity increases from 5.8\% to 6.7\% for the crystalline phase as pair interactions soften. The crystalline phases' allowable size dispersity is enhanced by potential softness, while the dispersity of the amorphous (liquid) phases follows an opposite trend. Overall, the 2 phase region shrinks monotonically and substantially as colloid-colloid interactions soften. Furthermore, amorphous phases accommodate, on average, smaller particles than those of the ordered (fcc) phase. Crystalline phases composed of size disperse particulates exhibit a higher degree of local order than their mono-disperse counterparts, admittedly at differing thermodynamic conditions. GSRT-ENTER   

First-principles simulations of extended phosphorus oxynitride structures in LiPON glasses

The thin film electrolyte LiPON, having the composition of Li$_{3+x}$PO$_{4-y}$N$_z$ with $x=3z-2y$, was developed at Oak Ridge National Lab in the 1990's for use in solid state batteries and related applications.\footnote{ N. J. Dudney, {\em{Interface}} {\bf{17}}:3, 44 (2008) and listed references.} In an effort to understand and to optimize properties of this electrolyte material, we expanded previous studies of isolated defects in crystalline Li$_3$PO$_4$\footnote{Y. A. Du and N. A. W. Holzwarth, {\em{Phys. Rev. B}} {\bf{78}}, 174301 (2008).} to focus on more complicated phosphate structures based on combinations of tetrahedral P$-$O bonds and bridging P$-$O$-$P bonds. For example, crystalline LiPO$_3$\footnote{E. V. Murashova and N. N. Chudinova, {\em{Cryst. Rept.}} {\bf{46}}, 942 (2001).} and P$_2$O$_5$\footnote{E. H. Arbib and co-workers, {\em{J. Solid State Chem.}} {\bf{127}}, 350 (1996).} are composed of phosphate structures with linear and branched chains, respectively. Both these and related structures derived from substituting O with N and adjusting mobile Li ion concentrations approximate components found in LiPON films.$^{\rm{2}}$ In the simulated structures, we find that N is energetically more stable at bridging bond sites than at tetrahedral sites by 2-3~eV and that the Li ion migration energies are 0.5$-$0.6~eV, similar to values measured in LiPON films.   

Systematic Size-Dependence of Electrical Resistivity Profiles in Bi$_{0.9}$Sb$_{0.1}$ Crystals

Recently, Fu and Kane[1] predicted that surface states with a odd-$Z_{2}$ topopological signature exist in the alloy Bi$_{1-x}$Sb$_{x}$ over a broad rangeof $x$. Hsieh \textit{et al}. [2] have confirmed by ARPES that, in Bi$_{0.9}$Sb$_{0.1}$, an odd number of surface states cross the chemical potential. To investigate the surface-state transport properties, we have examined how the resistivity-temperature (\textit{$\rho $-T}) profiles change as the transverse widths of crystals are varied from 20 to 1000 $\mu $m. In large crystals, the bulk conductance dominates, whereas in the small sample limit, surface conductance may be observable. Measurements on a large number of crystals reveal a reproducible systematic change in the $\rho $-T profiles. We find that the conductance at 4 K converges to a well-defined value when expressed as sheet conductance per square$ G$. The value is roughly 1000 times the quantum of conductance $e^{2}$/$h$. We discuss the interpretation of these results in the context of surface state conduction. 1. L. Fu and C. L. Kane, Phys. Rev. B 76, 045302 (2007). 2. D. Hsieh et al., Nature 452, 970 (2008).   

Low temperature C/T$^{3}$ peak in dipole disordered Bi$_{2}$Ti$_{2}$O$_{7}$

The presence of a low temperature peak in C/T$^{3}$ vs T, indicating excess entropy above the Debye contribution, is practically ubiquitous among both crystalline and amorphous materials. We present specific heat measurements on crystalline Bi$_{2}$Ti$_{2}$O$_{7}$, which is known to have incoherent ionic displacements leading to dipole disorder, and other related crystalline materials, including Bi$_{2}$NbInO$_{7}$, Bi$_{4}$Ti$_{3}$O$_{12}$, Y$_{2}$Ti$_{2}$O$_{7}$, and PbTiO$_{3}$, all of which exhibit excess low temperature entropy. We find that the C/T$^{3}$ peak for these samples agree roughly with a proposed scaling relation, with the peak temperature for Bi$_{2}$Ti$_{2}$O$_{7}$ being among the lowest reported for any crystalline material. We discuss our results in the context of understanding the evolution of crystalline degrees of freedom to glassy degrees of freedom through our investigations of dipole disordered Bi$_{2}$Ti$_{2}$O$_{7}$.   

Understanding the Material Thermodynamics of Two-Step Solar Thermochemical Water-Splitting Cycles

Metal oxide materials may be used in two-step solar thermochemical water-splitting cycles to renewably produce hydrogen: At high temperature, the oxide material is reduced, and at a lower temperature, the material re-oxidizes upon contact with water vapor producing hydrogen gas. Here, we present the first completely general analysis of the equilibrium thermodynamics of a two-step metal oxide water splitting cycle. We determine the temperature and pressure regimes in which both steps of the cycle are thermodynamically favorable in terms of the enthalpy and entropy of oxide reduction. Armed with this thermodynamic construct, we apply computational methods, such as density functional theory (DFT) and CALPHAD modeling, to assess many proposed oxide cycles. Using CALPHAD thermodynamic data, we survey a large number (more than 100) binary oxide redox couples, and show that none have both thermodynamically favorable reduction and oxidation steps. In an effort to find more thermodynamically favored redox couples, we investigate the utility of DFT calculations to screen materials' thermodynamics. In addition, we identify several driving forces which could enable more efficient two-step cycles, including a large positive solid-state entropy of reduction of the oxide.   

Using High-Voltage Direct Current in Removing Coke from a Zeolite Catalyst Grain

Zeolite catalysts are commonly used in petroleum refining processes. Over a period of time, these catalysts lose their activity due to gradual deposition of carbonaceous materials, called coke. The coked catalysts are usually reactivated by combusting the coke at elevated temperatures in presence of an oxygen-enriched gas. But the elevated temperatures cause damages to the structure of the catalyst which result in reduced activity. Normally, the catalyst is reactivated 3 or 4 times before it must be returned to the manufacturer for reclamation of the valuable platinum and/or rhenium content. This study is an attempt to come up with a new procedure to remove coke from a zeolite catalyst grain using high-voltage direct current. It is found that the process is self-terminated due to the loss of electrical conductivity of the grain. In addition, it is found that there an optimum current range for which up to 90{\%} of the coke can be removed. Higher or lower currents result in much smaller removal of the coke.