We demonstrate the power of modern first-principles theory by presenting examples of successful knowledge-based design of novel materials belonging to a family the so-called MAX phases. The MAX phases are a comparatively new class of nanolaminated compounds with a unique combination of metallic and ceramic properties. They have a common formula M_n+1AX_n (n = 1-3), where M is an early transition metal,A is an A-group element, and X is carbon or nitrogen. However, no magnetic MAX phases were known prior to our work. Using first-principles calculations, we predicted a series of thermodynamically stable magnetic MAX phases, e.g. (Cr_1-xMn_x)₂AlC [1], (Cr_1-xMn_x)₂GeC [2], and Mn₂GaC [3]. The materials have been synthesized [2,3] and have shown a variety of intriguing magnetic behavior. Furthermore, we have investigated theoretically a series of two-dimensional materials synthesized by selective etching of the A element from the MAX phases, known as MXenes, and have discovered MXenes with Dirac electrons [4]. The Dirac MXenes possess twelve conical crossings in the 1^st Brillouin zone with giant spin-orbit splitting. Our findings indicate that the 2D band structure of MXenes is protected against external perturbations and it is preserved even in multilayer phases. These results, together with experimental advances in MXene phases, presents a novel entire class of two-dimensional materials that may exhibit Dirac fermions with corresponding properties.
[1]. M. Dahlqvist, B. Alling, I. A. Abrikosov, J. Rosen, Phys. Rev. B 84, 220403(R) (2011).
[2]. A. S. Ingason, A. Mockute, M. Dahlqvist, F. Magnus, S. Olafsson, U. B. Arnalds, B. Alling, I.A. Abrikosov, B. Hjörvarsson, P. O. Å. Persson, and J. Rosen, Phys. Rev. Lett. 110, 195502 (2013).
[3] A. S. Ingason, A. Petruhins, M. Dahlqvist, F. Magnus, A. Mockute, B. Alling, L. Hultman, I. A. Abrikosov, P. O. Å. Persson, and J. Rosen, Mater. Res. Lett. 2, 89 (2014).
[4] H. Fashandi, V. Ivády, P. Eklund, A. Lloyd. Spetz, M. I. Katsnelson, and I. A. Abrikosov, arXiv:1506.05398 [cond-mat.mtrl-sci]

Scanning probe induced mechanical switching has recently emerged as an alternative method for domain manipulation due to their highly localized and electrically erasable characteristics. [1] Therefore the fundamental understanding of the role of local strains to the large effective piezoelectric and ferroelectric response of ferroelectric thin films is imperative. Here, we developed a self-consistent phase-field model to investigate the tip pressure (600nN) induced domain reorganization in 50nm thick (001) oriented lead zirconate titanate (Pb(Zr_1-xTi_x)O₃) films of rhombohedral (R) (x=0.2), tetragonal (T) (x=0.8) and morphotropic phase boundary (MPB) compositions (x=0.47). We found that at T phase composition c(001) domains is switched into a₁(100) and a₂(010) lateral domains and their growths are blocked by a/c twin domain walls, while at R phase composition the initial upward r₁⁺ (111) and r₄⁺ (1-11) domains are vertically switched into downward r₁^-(11-1) and r₄^-(1-1-1) domains respectively. Both switchings occur in limited 10nm deep regions from the top surface where the tip is located. At MPB composition R-T switching occurs, confirming that a transition between the R and T phase is ferroelastically active and can be directly induced by externally applied pressure. The domain switching at MPB is found to penetrate the entire film thickness, highlighting that the MPB composition is indeed softer than either pure R and T compositions. Our simulation results substantiate the experimental observation of change in hysteretic piezoresponse and domain structure caused by externally applied pressure. The phase-field model, combined with strain-mediated spectroscopy approach shows capability of exploring switching behaviors in a spatially resolved manner which would be otherwise difficult to resolve.
This research was sponsored by the Division of Materials Sciences and Engineering, Basic Energy Sciences, Department of Energy (YC, SJ, SVK). Research was conducted at the Center for Nanophase Materials Sciences, which also provided support and which is a DOE Office of Science User Facility. NBG gratefully acknowledges funding from the US National Science Foundation under grant number DMR-1255379. LQC gratefully acknowledges funding from DOE under Award No. DE-FG02-07ER46417.
[1] H. Lu, C.-W. Bark, D. Esque de los Ojos, J. Alcala, C. B. Eom, G. Catalan and A. Gruverman, Science 336, 59 (2012)

The physics of perovskite superlattices is well-known to be rich and varied. A feature of particular interest is that the properties of a superlattice can be enhanced over or even distinct from those of the constituent compounds, depending furthermore on the choice of layer thicknesses, termination and epitaxial strain. In this talk I describe recent work on first-principles-based modeling of polarization, dielectric response and piezoelectric response in ferroelectric-ferroelectric BaTiO₃/PbTiO₃ superlattices and of a first-order coupled magnetic nonpolar-polar metal-insulator transition in epitaxially strained SrCrO₃/SrTiO₃. The use of the first-principles models to interpret and interpolate first-principles results forms an essential part of a guided-sampling first-principles high-throughput approach allowing the identification of superlattice materials with optimal functional behavior targeted to specific technological applications and potentially the discovery of further novel physical phenomena in these systems.

Semiconductors obviously cover very wide applications in electronics, optoelectronics, and photovoltaics, and suggestions for novel useful semiconducting materials are always welcome. In particular, semiconductors that consist of earth-abundant and non-toxic elements only are especially attractive; stable production is possible because short supply of key elements is less likely to happen and these materials are likely to be environmentally friendly. Nitride semiconductors are appealing because the anion is nitrogen, a very common element, but those currently commercialized are mostly limited to GaN and related alloys. This motivates us to investigate ternary nitride compounds using high-throughput first principles calculations to identify promising semiconductors. Our algorithm involves looking at thermodynamic and lattice dynamic stability as well as electronic properties. The algorithm identifies previously reported nitride ternaries. ZnSnN₂ is a relatively new material that is recently gaining interest because this compound consists of earth-abundant elements only [1]. Its theoretical direct band gap is around 1.4 eV, which makes it an excellent candidate for potential applications in thin film photovoltaics. Other examples of previously reported ternary nitrides include LiZnN [2], ZnSiN₂ [3] and ZnGeN₂ [3]. Unreported and promising compounds are also found through calculations.
[1] L. Lahourcade et al., Adv. Mater. 25, 2562 (2013). [2] K. Kuriyama et al., Phys. Rev. B 49, 4511 (1994). [3] T. Endo et al., J. Mater. Sci. Lett. 11, 424 (1992).

Rare-earth pyrochlores, commonly exhibiting anomalously low lattice thermal conductivities, are considered as promising topcoat materials for thermal barrier coatings. However the structural origin underlying their low thermal conductivities remain unclear. In the present study, we investigated the phonon properties of two groups of RE pyrochlores, Ln₂Zr₂O₇ (Ln = La, Nd, Sm, Gd) and Gd₂T₂O₇ (T = Zr, Hf, Sn, Pb) employing density functional theory and quasi harmonic approximation. Through the relaxation time approximation (RTA) with Debye model, the thermal conductivities of those RE pyrochlores were predicted, showing good agreement with experimental measurements. The low thermal conductivities of RE pyrochlores were shown to largely come from the interference between the low-lying optical branches and acoustic branches. The structural origin underlying the low-lying optical branches was then clarified and the competition between scattering processes in transverse and longitude acoustic branches was discussed.

The number of component elements increased steadily in the 60-year development of photovoltaic semiconductors, i.e., from silicon in 1950s, to GaAs and CdTe in 1960s, CuInSe₂ in 1970s, Cu(In,Ga)Se₂ in 1980s, Cu₂ZnSnS₄ in 1990s and more recently Cu₂ZnSn(S,Se)₄ and CH₃NH₃PbI₃. The increased number of elements makes the material properties more flexible, however, it also causes the dramatic increase of possible point defects in the lattice, which can significantly influence the optical and electrical properties and thus the photovoltaic performance of these multinary compound semiconductors. Whether they can work as ideal solar cell absorber material depends on the behavior of their intrinsic defects. Through ab initio calculations, we can predict the dominant defects in new semiconductors and determine whether there are high concentration of deep-level defects that may act as electron-hole recombination centers. Then the semiconductors free of detrimental defects can be identified, which will accelerate the discovery of new photovoltaic semiconductors with high energy-conversion efficiency. I will discuss such ab initio screening in four classes of semiconductor systems, including the quaternary I₂-II-IV-VI₄ (Cu₂ZnSnS₄, Cu₂ZnSnSe₄), the ternary I-V-VI₂ (CuSbS₂ and CuSbSe₂), the binary Sb₂Se₃ as well as the halide perovskites (CsSnI₃, CH₃NH₃SnI₃ and CH₃NH₃PbI₃), which were all proposed as the candidate photovoltaic materials with high efficiency. Based on the calculated formation energies (concentration) and transition energy levels of possible defects, I will discuss the influence of the chemical component and growth conditions on the defect formation/ionization and thus the electrical and optical properties of the samples, which will help us understand the related experiments and also judge whether the defects impose any intrinsic limit to the efficiency of these photovoltaic semiconductors.
References:
(1) S. Chen, A. Walsh, X.-G. Gong, S.-H. Wei, Adv. Mater. 25, 1522 (2013)
(2) P. Xu, S. Chen, H.-J. Xiang, X.-G. Gong, S.-H. Wei, Chem. Mater. 26, 6068 (2014)
(3) B. Yang, L. Wang, J. Han, Y. Zhou, H. Song, S. Chen, J. Zhong, L. Lv, D. Niu, J. Tang, Chem. Mater. 26, 3135 (2014)
(4) A. Walsh, D. Scanlon, S. Chen, S.-H. Wei, X.-G. Gong, Angew. Chem. Int. Ed. 54, 1791 (2015)
(5) Y. Zhou, L. Wang, S. Chen, S. Qin, X. Liu, J. Chen, D. Xue, M. Luo, Y. Cao, Y. Cheng, E. Sargent, J. Tang, Nature Photonics 9, 409 (2015)

It is anticipated that coupling the intrinsic magnetism and porosity within the same material will offer an ideal medium for applications in magnetic separation, magnetic sensing, or low-density magnets. Beyond the adsorption and separation applications, some MOFs also exhibit some unique magnetic properties. Therefore, a clear understanding of the origin of magnetism in metalminus;organic frameworks (MOFs) would provide useful insight for tuning the electromagnetic properties of MOFs and finding new applications. Motivated by the manipulation of magnetic ordering with metal-doping methods, we chose M-MOF-74 (M=Mn, Fe, Co, Ni) to systematically investigate the electromagnetic behavior in MOFs by means of a density functional theory (DFT+U) method because isostructural MOF-74 containing various paramagnetic metal centers have been successfully synthesized.
Our calculated results show that the open paramagnetic metal sites in three-dimensional porous magnets M-MOF-74 (M = Ni, Co, Fe, Mn) favor high-spin electronic arrangement. Fe- and Co-MOF-74 exhibit ferromagnetic (FM) features and significantly distinct energy gaps between spin-up and spin-down channels in metastable states. After replacement of the Co center with a Ni ion, the FM feature is exhibited for the stable state since the “extra” valence electron is filled in the spin-down 3d bands to shift the Fermi level to higher energy. In contrast, after removal of one valence electron (i.e., replacement of the Fe center with Mn atoms), the energy gap is significantly enlarged and an antiferromagnetic (AFM) feature will be discerned. Our studies may enhance the understanding of the origin of magnetism in MOFs and provide useful insight for tuning the electromagnetic properties of MOFs and designing low-magnetic materials.

The cell walls in plant tissues are made up of just four basic building blocks: cellulose, the main structural fiber of the plant kingdom, hemicellulose, lignin and pectin. Although the microstructure of plant cell walls varies in different types of plants, broadly speaking, cellulose fibers reinforce a matrix of hemicellulose and either pectin or lignin. The cellular structure of plants varies from the honeycomb-like cells of wood to the closed-cell, liquid-filled foam-like parenchyma cells of apples and potatoes. The arrangement of the four basic building blocks plant cell walls and the variations in cellular structure give rise to a remarkably wide range of mechanical properties: the Young&’s moduli span 4 orders of magnitude while the compressive strengths span nearly 3 orders of magnitude. Here, we review the microstructure of both the cell wall and the cellular structure in four plant materials (wood, arborescent palm stems, bamboo and parenchyma) to explain the wide range in mechanical properties in plants.

Nano structure of Si has drawn much attention owing to their unique electronic and optical properties. Recently, two dimensional silicon nano sheets (Si-NSs) were successfully synthesized [1,2] and extensively investigated on the possible applications to novel devices. It is common for the nanostructured materials to be under high level of stressed state due to the significant surface or interfacial effect. In the present work, we focused on the effect of lattice strain induced by surface oxidation of Si-NSs on their electronic band structures. We employed the density functional theory for the electronic structure calculation of the oxidized Si-NSs simulated by reactive molecular dynamics simulation. Si-NSs exhibit either direct or indirect band gap depending on the growth orientation while the bulk Si shows only indirect band gap. Moreover, applying bilateral strain to the Si-NSs induces the change in the conduction band minimum of the electronic band structure and leads to the band gap transition. For the (111) surface orientation, Si-NSs undergoes the indirect-to-direct transition by tensile strain. Meanwhile, (001)/(110) orientation Si-NSs undergoes the direct-to-indirect transition by compression. We suppose that the structural deformation due to the strain leads to the adjustment of the conduction band minimum energy according to its bonding or antibonding character. The present work revealed a general tendency that increasing the bond length along the lateral dimension enhanced the Si-NSs to possess the direct band gap characteristics. This finding would be very important for the strain engineering of the Si-NSs, because a slight change in the surface state such as surface oxidation can affect significantly the physical properties of the materials.
[1] U. Kim et al, “Synthesis of Si Nanosheets by a Chemical Vapor Deposition Process and Their Blue Emissions”, ACS Nano, 5, 2176 (2011).
[2] S. Kim et al, “Two-dimensionally Grown Single-Crystal Silicon Nanosheets with Tunable Visible-Light Emissions”, ACS Nano, 8, 6556-6562 (2014).

Materials that undergo metal-insulator transitions (MITs) are under intense study because the transition is scientifically fascinating and technologically promising for various applications. Among these materials, VO₂ has served as a prototype due to its favorable transition temperature. While the physical underpinnings of the transition have been heavily investigated experimentally and computationally, quantitative modeling of electronic transport in the two phases has yet to be undertaken. In this work, we establish a density-functional-theory-based (DFT) approach to model electronic transport properties in VO₂ in the semiconducting and metallic regimes, focusing on band transport using Boltzmann transport equation. Free parameters in the model are calibrated using experimentally measured transport quantities. We find that the approach can efficiently model the metallic and semiconducting phases. Using this methodology, we performed high throughput DFT calculations to investigate effects of doping on VO₂ from mechanical, thermodynamic and electronic transport aspects.

In this talk, I will present our recent simulations of nanostructure systems using linear scaling three dimensional fragment method (LS3DF) and nonadiabatic molecular dynamics (NMD). The LS3DF method is used to study the effects of Moire's pattern of a double layer MoS2 and MoSe2, and it is found that the structural Moire's pattern will localize the electronic states in such systems. The LS3DF method is also used to study the effect of electrostatic potential fluctuations in hybrid perovskite (CH3NH3)PbI3, and found that the random orientation of the organic molecule will cause carrier localization in the system, which will have significant impact on the carrier transport in the system. NMD is used to study the carrier transport in a monolayer of organic molecule, and it is found that the dynamics disorder play an important role in the carrier transport of such systems. The same method is also used to study carrier transport in the hybrid perovskite system.

Fully parameter-free ab initio methods based on density functional theory are steadily gaining popularity. Their atomistic view on physical processes and the access to chemical trends are attractive for a knowledge-driven development of novel electronic, biological or engineering materials. However, the apparent restriction of the method to ground state properties is a severe challenge for the understanding and prediction of many materials properties, as, e.g., heat capacities and phase diagrams, for which finite temperature effects are decisive. Over the last years we have developed a large range of multi-physics tools to go beyond this limitation. A key challenge was and is the accurate determination of free energies including all relevant excitation mechanisms individually as well as non-adiabatic coupling effects stabilizing certain phases. Within this talk we will discuss some of our recent methodological developments that foster a large-scale screening of material properties even at temperatures where anharmonic lattice vibrations and magnetic disorder become important. The finite-temperature ab initio methods will finally be applied to the calculation of martensitic and intermartensitic phase transitions in Ni-Mn-X Heusler alloys. The predicted chemical trends are important for a tailored design of their magnetocaloric and shape-memory properties.

We present a simple and efficient approximation to the quantum mechanical formulation of the electron relaxation time induced by electron-phonon interaction. The approximation is suited for the high temperature regime providing a practical procedure for high-throughput screening of thermoelectric materials. The method is applied to calculate the electronic transport coefficients and to determine the optimal carrier concentration which maximizes the thermoelectric figure of merit ZT at a target temperature. Computational screening of optimized ZT values is performed in the compositional space of half-Heusler compounds selected from materials databases and consisting of cheap earth-abundant elements. The method is validated by demonstrating agreement with the results of the exact electron-phonon calculations and with the experimental data for the currently used compounds. Several new compounds promising for thermoelectric applications are identified, one of which was successfully synthesized by our experimental collaborators [1]. The computational results are examined for the presence of correlations between the ZT values and the material properties of half-Heusler compounds. In particular, we find deviations from the Wiedemann-Franz law in these compounds at low carrier concentrations and high temperatures by directly computing electrical and the electronic part of the thermal conductivities.
[1] G. Joshi et al., Energy Environ. Sci. 7, 4070 (2014)

The incorporation of hydrogen in metals can strongly affect the physical properties of the host matrix and can lead to irreversible damages. Therefore, the solubility of hydrogen in metals is a fundamental data to design new protections preventing hydrogen embrittlement and safety materials. However, the apparent solubility can be influenced by the presence of crystalline defects such as point defects, dislocations or grain boundaries. In particular, it has been suggested that hydrogen promotes the formation of new vacancies, which participates to the mechanisms of degradation of metals (Fukai, 2003, J. Alloys Comp.). In this study, we determine the vacancy and hydrogen concentrations in nickel at thermodynamic equilibrium up to the melting point from a combination of first principles calculations and statistical physic.
First, we determine the solution enthalpy and entropy of the incorporation of H in a perfect crystal. We extend the calculations of these thermodynamic quantities to the formation of H-vacancy clusters where H atom is located in the displacement field generated by the vacancy. We limit the study to the interstitial sites inside and tangent to the vacancy core where the displacement field leads to a significant deformation of the sites (Metsue et al., 2014, Phil. Mag.). The calculations are performed up to the melting point for a wide range of H₂ chemical potential where the Gibbs free energy is expressed as a sum of vibration and electronic contributions from the computation of the phonon dispersion curve and the electronic density of states. The latter contribution is required to get a realistic vacancy concentration (Metsue et al., 2014, J. Chem. Phys.) The solution enthalpies and entropies calculated for a H₂ fugacity of 1 bar are in good agreement with previous experimental data (Eichenuaer et al., 1965, Z. Metall., Stafford et al., 1974, Acta Met.). In addition, we found that H-rich octahedral sites inside the vacancy core are destabilized at high H₂ chemical potential, similarly to fcc Fe (Nazarov et al., 2010, Phys. Rev. B).
Then, we calculate the solubility and the vacancy concentration from the solution and H-vacancy clusters formation Gibbs free energy. We take into account an additional configuration part to the total entropy related to the distribution of H in the interstitial sites and the H-vacancy clusters. A close agreement is found between experiments and the hydrogen concentration calculated for a H₂ fugacity of 1 bar. In particular, we show that the electronic excitations lead to a positive deviation in the Arrhenius plot of the solubility at temperatures close to the melting point. The vacancy concentration is calculated for a wide range of H₂ chemical potential and we discuss its implications on the deformation of the lattice.

The macroscopic structure of hcp zirconium alloys - that are used as cladding materials in nuclear reactors - is affected under irradiation, as these alloys experience a stress-free dimensional change. This irradiation growth is linked to the large amount of point defects created under irradiation, that evolve towards larger clusters, leading to prismatic dislocation loops ( loops), both of interstitial and vacancy type. At large irradiation dose, a growth enhancement is observed, correlated to the appearance of basal vacancy loops (faulted loops). Experiments have suggested that hydrogen - coming from the oxidation of the rods - promotes the formation of loops, thus explaining the higher irradiation growth when the H content is increased. A proper modeling of kinetic evolutions of the microstructure under irradiation requires first to understand the relative stability of the different defects, including the influence of H, from the small scale of defect sizes that are accessible by ab initio to the larger scale of extended defects that are observable by TEM.
We propose here a hybrid approach, relying both on atomistic calculations and on continuous models of defect energetics. Ab initio calculations, focusing on the elementary interactions between point defects, (i) establish how to build the small vacancy clusters and (ii) evidence a strong binding of H to these clusters. The energetics of extended vacancy clusters is then modeled using continuous laws, that are first validated on empirical potential calculations and second parametized on ab initio data. The easier formation of loops is seen to be explained by their higher stability [1]. Finally, a thermodynamic modeling of H segregation on the various faults shows that these faults are enriched in hydrogen, leading to a decrease of the stacking fault energies. This is consistent with the trapping of H by vacancy loops observed experimentally. The strongest trapping, and thus the strongest stabilization, is obtained for vacancy loops lying in the basal planes, i.e. the loops responsible for the breakaway growth observed under high irradiation dose [2].
[1] C. Varvenne, O. Mackain, and E. Clouet, Acta Mater. 78 (2014) 65-77
[2] C. Varvenne, O. Mackain, and E. Clouet, submitted, (2015)

The magnetic critical order-disorder temperature T_c of a material is a parameter that sets an upper operational limit for magnetic devices. First-principles based methods hold the promise of significantly speeding up the search for new magnetic materials by predicting T_c and thus guide experiments.
In this work, we employ and critically evaluate a first-principles approach based on supercell calculations for predicting T_c. Supercell calculations have the benefit of allowing for straightforward incorporation of structural or vibrational disorder effects on the magnetic interactions. As a model material we use the recently discovered nanolaminate Mn₂GaC,¹ which belongs to the family of so-called MAX phases (where M is an early transition metal, A an A-group element, and X is C and/or N).
First, we derive the exchange interaction parameters J_ij between pairs of atoms i and j of the bilinear Heisenberg Hamiltonian using the novel magnetic direct cluster averaging method (MDCA).² In this method, the J&’s are calculated through an averaging procedure over a large number of supercells in which the magnetic state has been randomly generated with respect to all magnetic atoms except for those of atoms i and j, for which J_ijis sought. We compare the J&’s from the MDCA calculations to the same parameters calculated using the Connolly-Williams structure inversion method, and show that the two methods yield closely matching results.
Secondly, we use the MDCA-derived J's as input parameters in Monte Carlo simulations to calculate the magnetic energy, total magnetic moment, specific heat, magnetic susceptibility as well as T_c. For Mn₂GaC, we find T_c = 660 K. The uncertainty in the calculated value of T_c caused by uncertainties in the J&’s is discussed and exemplified in our case by a calculation of the standard deviation, which is 133 K.
¹ A. S. Ingason, A. Petruhins, M. Dahlqvist, F. Magnus, A. Mockute, B. Alling, L. Hultman, I. A. Abrikosov, P. O. Å. Persson, and J. Rosen, Materials Research Letters 2, 89 (2014).
² A. Lindmaa, R. Lizárraga, E. Holmström, I. A. Abrikosov, and B. Alling, Physical Review B 88, 054414 (2013).

Uranium dioxide UO₂ is used as a standard fuel in pressurized water reactors (PWR). During fission reactions of uranium intragranular bubbles of xenon are generated. The presence of these bubbles modifies the thermomechanical properties of the fuel. Challenges in terms of safety due to the presence of these bubbles led to extensive work both from experimental and theoretical points of view, in order to better understand the changing of the physical properties and behavior of the fuel.
It is known from the literature that these bubbles are microfaceted, with (111) and (100) surfaces. We then study here simplified systems of xenon on semi infinite UO₂ surfaces. In a first step, we assess the relative stability of UO₂ surfaces according to their orientation and then to their polarity, by mixing thermostatistical relaxation and analytic formulations within a simple electrostatic model. The main result is that, whereas the (111) surface appears stable by construction and does not involved major reorganization, the polar (100) one is only stabilized through drastic rearrangement of the surface region.
In a second step, we proceed to xenon adsorption on these relaxed surfaces through Monte Carlo simulations in the grand canonical ensemble within semi-empirical potentials. The most striking feature revealed by the simulation is the existence of a phase transition from a dilute xenon phase towards a dense one, in which coexist the FCC, BCC and HCP structures, by increasing the chemical potential. Otherwise, the xenon density is found to increase with the temperature for a given chemical potential.
In a third step, the pressure inside the xenon bubble and in the UO₂ matrix has been investigated. In the former case, we show that whatever the xenon structure may be, the pressure increases with the xenon density, but not with the temperature for a fixed density. In what concern the matrix, we will present the pressure profile before and after xenon adsorption. The next step will be to introduce these so obtained results in a micromechanical model, which will allow us to propose a thermomechanical behavior law for the porous UO₂.

A novel Cluster Dynamics (CD) model that describes the nucleation and evolution of defect clusters in oxides systems has been developed. The model has been used to predict clustering of vacancies and interstitials into voids and dislocation loops, respectively, in irradiated UO₂. The model reproduces well a range of experimental data on nucleation and growth behavior and its temperature dependence. A very important feature of this model is its ability to predict the off-stoichiometry (or composition) of defect clusters, allowing, in turn, for the tracking of off-stoichiometry of the matrix. The effect of migration energy of point defects on the concentration and average size of voids has been studied. Also, the effect of irradiation conditions such as irradiation temperature, irradiation dose and dose rate on clusters concentration and composition has been investigated. The preliminary results show that Frenkel defects, as opposed to Schottky defects, dominate the nucleation process in irradiated UO₂. Vacancy clusters tend to grow mainly by absorbing oxygen vacancies and the migration energy of uranium vacancies is the rate limiting energy in nucleation and growth of voids. The results also show that, in a stoichiometric UO₂ under irradiation, vacancy clusters (voids) tend to have both hypo- and hyper-stoichiometric composition with a higher fraction of hyper-stoichiometric composition clusters. Hyper-stoichiometric cluster compositions indicate that the matrix would become oxygen rich even if the initial state is perfectly stoichiometric.

Complex oxides are being extensively studied as superionic materials (in which ionic conductivity is large) for application as fuel cell and battery materials. Experimental studies have revealed that the ionic conductivity in A2B2O7 pyrochlores is sensitive to the chemical composition of the material, with, in particular, B=Ti pyrochlores exhibiting relatively low ionic conductivities and B=Zr pyrochlores having much higher conductivities. This difference in conductivity has been linked to the tendency of the material to exhibit cation disorder (the swapping of A and B cations). The disordering tendency of the material depends on its chemical composition, such that a change in chemistry is often accompanied by a concurrent change in the level of disorder, leaving some doubt as to the true origin (chemistry or disorder) of the enhancement in conductivity. Meanwhile, cation disorder can be controllably introduced into these materials via irradiation and much work has been done examining the disordering that occurs in pyrochlores as a function of the irradiation dose, as these materials are also proposed for nuclear waste forms. However, the two properties - ionic conductivity and radiation-induced disordering - have not been explicitly connected.
Here, in support of ongoing experimental studies, we examine the independent contributions of chemistry and disorder on ionic conductivity in two pyrochlores, Gd2Ti2O7 and Gd2Zr2O7. Using molecular dynamics and accelerated molecular dynamics, we isolate the role that the B-site chemistry and the A-B disorder have on oxygen ionic conductivity. We examine both structural defects, i.e. vacant oxygen sites relative to the basic fluorite structure, as well as intrinsic defects (interstitials and vacancies) that might be present after irradiation. We find contrasting effects of both disorder and chemistry on the mobility of each type of defect. In particular, disorder always increases the conductivity associated with the structural defects, but low levels of disorder can inhibit the motion of intrinsic defects. We contrast this behavior with that found in AB2O4 spinels, where disorder decreases the mobility of oxygen defects. These results highlight the non-trivial and contrasting effects associated with ionic conductivity in these complex materials.

We report on non-adiabatic aspects of the initial stages of radiation damage, when ions and electrons get out of mutual equilibrium, and energy is exchanged between them, determining the ways radiation energy is dissipated. The hypothesis behind this work is that by understanding the mechanisms of energy dissipation by electrons and ions, we provide a means to manipulate them to reduce the defect production, and quench the damage. Our work describes quantitatively the energy transfer process from electrons to atoms via ionization, a process known as Coulomb explosion.
Recently, we reported on a correlation between nuclear and electronic stopping power that revealed the effects of ionization on the ions dynamics [Correa et al. Phys. Rev. Lett. 108 (2012)069901]. In this work, we use the same TD-DFT methodology to determine the time dependent forces experienced by Ni target atoms as a Ni projectile moves along its trajectory. Similar to our previous observation, we find large differences in the magnitude of momentum transfer as compared to the value given by the adiabatic approximation. This effect is the quantitative signature of the Coulomb explosion (CE). With these results in mind, we discuss the impact CE has on the modification of the energy dissipation pathways in the early stage of defect formation.
Implications of these results in more accurate computer simulations of radiation damage are expected, as both the nuclear and electronic aspects of the problem are both included in the atomic-scale simulations of radiation damage.
Work supported by the Energy Dissipation to Defect Evolution Center (EDDE), an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science.

Iron-chromium alloys are the basis for ferritic and ferritic-martensitic steels that will be used in future fission (generation IV) and fusion nuclear reactors. With Cr contents between typically 8 to 12%, or even 14% in the matrix of some oxide dispersion-strengthened steels, one expects a precipitation of a Cr-rich phase (α&’), leading to the hardening and embritshy;tlement of the materials. For such compositions and at temperatures of interest, the kinetics of precipitation during a thermal ageing is usually very slow, but it can be strongly accelerated under irradiation, due to point defect supersaturation. We present a modeling of the α&’ precipitation under irradiation, based on atomistic kinetic Monte Carlo simulations, with parameters fitted on ab initio calculations. The Monte Carlo simulations are combined with a cluster dynamics model to get an accurate description of the evolution of sink densities, which control the point defect supersaturation. The simulations are used to predict the acceleration factor, depending on the alloy composition, irradiation flux and temperature. They are compared to available experimental results. The possible effects of ballistic mixing (which occurs in displacement cascades) and of the strong interaction between carbon atoms and vacancies, are also discussed.

In order to better understand plasticity in alloys, it is important to describe the deformation mechanisms with accuracy, and in particular to understand how dislocations, responsible for the plastic deformation, interact with solute atoms. Long-range interactions are well described with continuum elasticity theory, whereas short-range interactions depend on atomistic mechanisms. Previous ab initio studies of the ½<111> screw dislocation in body centered cubic (bcc) metals showed that in pure metals, the dislocation core adopts a symmetrical configuration, the easy core configuration, centered on a triangle of first-neighbor <111> atomic columns where helicity is reversed with respect to the bulk. The other core configurations - the asymmetric core, the hard core where the three <111> core atomic columns are at the same altitude, and the split core centered near an atomic column - are all unstable in pure bcc metals [1].
In this work, we investigate the effect of interstitials solute atoms on the ½<111> screw dislocations in bcc metals using ab initio calculations performed using the VASP code. First considering Fe(C), our DFT calculations show that, when a row of solute atoms is added in the neighborhood of the dislocation core, both the dislocation and the solute atoms reorganize to form a low-energy configuration. We find surprisingly that the dislocation adopts a hard-core configuration with the solute atoms placed at the center of regular trigonal prisms formed by the Fe atoms inside the three <111> core atomic columns [2]. This structure is similar to the building unit of Fe₃C cementite. We obtained the same core reconstruction with other solutes (B, N, O) in Fe and in other metals (W, Mo). Within this unexpectedly stabilized hard core, the interaction energy between the dislocation and the solute atoms is strongly attractive and leads in equilibrium conditions at room temperature to a core saturation by solute atoms, even for very low carbon concentrations in bulk. Consequences on the dislocation mobility and relations to dynamical strain ageing will also be discussed.
[1] L. Dézerald, L. Ventelon, E. Clouet, C. Denoual, D. Rodney and F. Willaime, “Ab initio modeling of the two-dimensional energy landscape of screw dislocations in bcc transition metals”, Phys. Rev. B 89, 024104-13 (2014).
[2] L. Ventelon, B. Lüthi, E. Clouet, L.Proville, B. Legrand, D. Rodney and F. Willaime, “Dislocation core reconstruction induced by carbon segregation in bcc iron”, submitted to Phys. Rev. B Rapid Com. (in press)

Radiation induced segregation of alloying elements in nanostructured ferritic alloys (NFAs), a material used for fuel cladding, results in embrittlement at grain boundaries, which comprises the lifetime of the material. The study models this segregation by the creation, migration, recombination, and annihilation of point defects by electron irradiation using rate theory. A progression was used in the development of the models from simplistic pure iron to binary Fe-Y to the complex Fe-Y-O. The rate theory equations break from previous ternary alloy models in literature by modeling the movement of oxygen, an element preferentially found in the octahedral sites and travels by interstitial diffusion. The simulations from both models showed agreement with literature with enrichment of yttrium and oxygen at sinks such as grain boundaries.

We present an ab initio computational framework for the Rapid Prototyping of Phase Diagrams (RaPPD) of alloy systems. The approach is based on a pre-generated database of special quasirandom structures (SQS) including, not only simple crystal structures (e.g. fcc, bcc, hcp), but also common intermetallic phases with multiple sublattices. Software tools were developed to directly convert SQS ab initio data into standard thermodynamic database format and combine it to elemental SGTE data, thus enabling rapid and effortless generation and visualization of thermodynamic data for multicomponent systems with standard CALPHAD tools. Interface to widely used interactive graphical visualization tools are also demonstrated. A key aspect of the process is a formal method to calculate the formation energies of mechanically unstable phases (which are surprisingly common) that is consistent with SGTE data regarding such phases. The method is illustrated with applications to Rhenium-based alloys.

Liquefaction of biomass is one of the major paths in the utilization of nonfood biomass resources for renewable fuels and chemical feedstocks. However, investigations of the liquefaction of biomass have been limited to mineral acids catalysts which often with unstable compounds; many studies failed to identify the compounds in the lique#64257;ed products. Recently, our work reported a highly efficient simple strategy using 1-(1-alkylsulfonic)-3-methylimidazolium chloride ionic liquid as catalysts to liquefy cellulose in ethylene glycol at 180 °C to produce a stable lique#64257;ed oil with a well-de#64257;ned composition of merely three compounds and identi#64257;cation of a cyclopentenone derivative in a cellulose liquefaction process. The focus of this paper is on the computational modeling investigations of such acidic ionic liquid catalyzed liquefaction of cellulose in ethylene glycol. Theoretical results are qualitatively in agreement with experiments. The calculations not only explain the experimental observations but also provide the insight into the conformational, electronic structural and thermodynamic energetic profile of such interactions.

Molecular diodes based on charge transfer complexes of fullerene[60] with different porphyrins has been proposed and studied theoretically. Current-voltage characteristics and the rectification ratios (RR) of different molecular diodes were calculated using direct ab initio method at PBE/def2-SVP level of theory with D3 dispersion correction the range from -2 to +2 V.
The rectifying effect in molecular junctions of the form metal|molecule|metal, is defined in terms of the absence of inversion symmetry, I(V)ne;-I(-V), where I and V are the current and the applied voltage, respectively. The dominant factors inducing rectification are geometric asymmetry in the molecular junction and the spatial profile of the electrostatic potential.
The highest RR of 32.4 was determined for the complex of C60 with zinc tetraphenylporphyrin at 0.8 V. Other molecular diodes show lower RR, however, all complexes show RR higher than 1 at all bias voltages. The asymmetric evolutions and alignment of the molecules with the applied bias were found to be essential in generating the molecular diode rectification behavior. Metal nature in metalloporphyrins and the interaction porphyrin - electrode significantly affects RR of molecular diode. Large metal ions like Cd²⁺ and Ag²⁺ in metalloporphyrins disfavor rectification creating conducting channels in two directions, while smaller ions Zn²⁺ and Cu²⁺ favor rectification increasing interaction between gold electrode and porphyrin macrocycle.

The presence of impurities in solids is a source of many interesting effects, particularly relevant in the conductivity, optical properties and specific heat. For instance, in nanoelectronics, such effects could be useful to develop molecular devices such as novel computer architectures [1,2], chemical [3] and biomedical sensors [4]. However, the inclusion of impurities breaks the Bloch symmetry, restricting the systems that can be addressed theoretically in an exact way to those of few atoms. In this work, we present an alternative method to study the electrical conductance in real-space by means of a renormalization plus convolution method [5] applied on the Kubo-Greenwood formula for multidimensional systems of macroscopic size with site and bond impurities. The results shows that the spectral average of the conductivity depends strongly on the location of the site and bond impurities in periodic chains. Particularly, when the distance between impurities follows the Fibonacci sequence, we find that the spectral average falls following a power law as the number of atoms in the system grows [6]. In addition, we demonstrate analytical and numerically the presence of novel ballistic transport states in two and three dimensions, when respectively three lines or planes of site impurities for null chemical potential is present [7]. Finally, we analyze the impurity effects on the conductance spectra of periodic and aperiodic solids with nano to macroscopic length when the number of impurities grows.
This work has been partially supported by UNAM-IN113813. Computations were performed at Miztli of DGTIC, UNAM.
[1] C. Joachim, J.K. Gimzewski, A. Aviram, Nature408, 541(2000).
[2] J.C. Ellenbogen, J.C. Love, Proc. IEEE88, 386 (2000).
[3] Y. Cui, Q.Wei, H. Park, C.M. Lieber, Science293, 1289 (2001).
[4] F. Patolsky, G. Zheng, O. Hayden, M. Lakadamyali, X. Zhuang, C.M. Lieber, Proc. Natl. Acad. Sci. USA101, 14017 (2004).
[5] V. Sanchez and C. Wang, Phys. Rev. B70, 144207 (2004).
[6] V. Sanchez, C. Ramirez, F. Sanchez, C. Wang, Physica B449, 121 (2014).
[7] C. Wang, C. Ramirez, F. Sanchez, and V. Sanchez, Phys. Status Solidi B, 252, 1370 (2015).

Recent progress in high-performance computational environments have expanded he range of applications of computational metallurgy is expanding very rapidly [1]. We have performed the large-scale molecular dynamics (MD) simulations, which are performed on graphics processing unit (GPU) to discuss the nature of solidification and related properties [1-5]. Up to now, we have revealed the evolution of the grain boundary groove [4] and the spontaneous evolution of anisotropy in the solid nucleus in the undercooling melt of iron [5]. In this study, homogeneous nucleation from an undercooled iron melt is investigated by the statistical sampling of million-atom MD simulations. The undercooled iron melt in a cell with a size of 53.4 x 53.4 x 4.3 nm³ (1,037,880 atoms) is isothermally undercooled with various temperatures for 10000 ps under zero pressure by a NPT constant condition. The nucleation rate and the incubation time of nucleation as functions of temperature have characteristic shapes with a nose at the critical temperature. This indicates that thermally activated homogeneous nucleation occurs spontaneously in MD simulations without any inducing factor [1,6].
[1] Y. Shibuta, M. Ohno, T. Takaki, JOM, article in press. (doi:10.1007/s11837-015-1452-2).
[2] Y. Shibuta, S. Takamoto, T. Suzuki, ISIJ Int., 48 (2008) 1582.
[3] Y. Watanabe, Y. Shibuta, T. Suzuki, ISIJ Int., 50 (2010) 1158.
[4] Y. Shibuta, K. Oguchi, T. Suzuki, ISIJ Int., 52 (2012) 2205.
[5] Y. Shibuta, K. Oguchi, M. Ohno, Scripta Mater., 86 (2014) 20.
[6] Y. Shibuta, K. Oguchi, T. Takaki, M. Ohno, submitted.

The electromechanical coupling between strain gradients and ferroelectric polarization, known as the flexoelectric effect, provides a new mechanical approach to switch the polarization purely mechanically in ferroelectric thin films. This approach has been experimentally demonstrated recently by imposing a loading force via an atomic force microscope (AFM) tip onto the film surface. The critical force required for mechanical switching depends on several factors, such as film thickness, with or without top electrode, tip radius, static or moving tip, etc. Phase-field modeling serves as an effective method to study the domain structure evolution of ferroelectric films during the polarization switching process. By performing a series of phase field simulations, we study the flexoelectric polarization switching realized by AFM tip pressing against the Graphene/BaTiO₃/SrTiO₃ heterostructure. Local control of nanoscopic ferroelectric domains is realized in the ferroelectric capacitor structure due to the localization of strain gradients generated by the AFM tip. With 1 nm graphene on top of 19-nm-thick BaTiO₃ thin film, the critical force is estimated to be around 2000 nN. It is also shown that the critical force is significantly influenced by loading conditions, misfit strain, graphene thickness and built-in bias. In addition, we have set a limit for the maximal thickness of graphene top electrode in order to realize mechanical switching in the heterostructure, which is predicted to be about 3 nm. The phase-field modeling results show good agreements with experimental observations and give insights to the better control of mechanical switching in ferroelectric heterostructures.

We have performed first-principles study of electronic properties of FeCr_xSe ( x =0.0, 0.01, 0.02, 0.04) alloys using Korringa-Kohn-Rostoker Atomic Sphere Approximation within the coherent potential approximation (KKR-ASA-CPA). We found from our calculations that the excess of Cr into FeSe significantly affects the electronic structure with respect to the parent FeSe. The results have been analyzed in terms of changes in the DOS, PDOS, band structures, Fermi surface, bare Sommerfeld constant and the superconducting transition temperature of FeCr_0.01Se , FeCr_0.02Se and FeCr_0.04Se alloys respectively. Our calculations show that calculated T_c are in good agreement with the experimental results for these alloys.

Eu-doped CaAlSiN3 (CASN) is widely utilized as an efficient red phosphor; however, the high price of rare-earth metals has driven efforts toward finding non-rare-earth metal dopants. We report first-principles calculations based on density functional theory (DFT) and geared toward identifying new non-rare-earth metal dopants for use in the CASN-based phosphors. We calculated the formation energies, the electronic structures, and the optical absorption spectra of various metal dopants (Eu, Mn, Sn, and Bi) in CASN. The calculated density of states, band structures, and absorption spectra were consistent with previous experimental observations obtained from Eu- and Mn-doped CASN. The DFT calculations suggested that Sn and Bi are promising candidates as non-rare-earth metal dopants in CASN-based phosphors. Our calculations demonstrate that DFT-based first-principles calculations provide a viable tool for finding new phosphor materials.

Building on our understanding of the chemical bond [1], advances in synthetic chemistry, and large-scale computation, materials design is becoming a reality [2]. We will present our recent progress in this regard across three strands of research: (i) harnessing of meta-stable states induced by temperature, light and pressure [3]; (ii) the development of electroactive metal-organic frameworks [4]; (iii) the optimisation of emerging materials for photovoltaics, including kesterite and perovskite structured compounds [5]. A major challenge is the accurate description of finite temperature effects including lattice vibrations and structural disorder, for which we employ many body [6] and multi-scale approaches [7]. While, we note a number of successes, including the validation of the timescales of orientational disorder in methylammonium lead iodide by quasi-elastic neutron scattering [8], there remains a significant amount of work left to complete this quest.
This work has benefited from funding by the EPSRC, the ERC, and the Royal Society.
1. L. Pauling, The Nature of the Chemical Bond, Cornell University Press (1960).
2. A. Walsh, Nature Chem. 7, 274 (2015).
3. J. M. Skelton et al, CrystEngComm 17, 383 (2015).
4. C. H. Hendon and A. Walsh, Chem. Sci. 6, 3674 (2015).
5. K. T. Butler, J. M. Frost and A. Walsh, Energy Environ. Sci. 8, 828 (2015).
6. J. M. Skelton et al, APL Mater. 4, 041102 (2015).
7. J. Buckeridge et al, Chem. Mater. 27, 3844 (2015).
8. A. M. A. Leguy et al, Nature Comm. 6, 7134 (2015).

The relationship between processing, mechanical properties, and microstructure of a material is highly complex. Understanding these relationships helps researchers optimize and predict a material's performance. However, existing mathematical and physical mechanism-based models are not always capable of modeling such complex systems to a high degree of accuracy. In particular, there are some systems such as 7xxx series aluminum that are so complex that there has been little to no success using conventional modeling methods. Artificial neural networks (ANNs) offer an alternative method of modeling these relationships. These are highly accurate mathematical models that can predict the cause-effect relationships of complex systems with a small amount of computational power. To use this method, the user needs to identify the primary input and output attributes and have generated empirical data linking these attributes. The ANNs can establish these linkages from the purely empirical data and without any understanding of the physical mechanisms that control the outputs. This makes them well suited for modeling materials and their properties, and they have already found some success in this area. Here we use ANNs to predict the link between processing route attributes, such as solution-treatment temperature, quench rate, and aging-treatment, and the microstructure attributes including grain size, degree of recrystallization, and strengthening precipitate size and volume fraction, of 7050 aluminum.

Grain size has an important effect on the performance of metallic materials, and thus, controlling the grain size during processing is a key concern for process designers. Grain growth is a complicated process that is affected by many factors, including temperature, deformation, and microstructural features. In this work, we focused on developing methods to control the local grain size through a combination of model based prediction and feed-forward/feedback control. In this work, the Monte Carlo Potts model is used to predict grain size evolution, and is then coupled to a control algorithm to control the local grain size. In a Monte Carlo grain growth simulation, the physical domain is discretized into N evenly spaced sites, and in each Monte Carlo step, sampling (site selection) is randomly performed N times, with the probability of the grain orientation at a site switching related to the change in energy. Conventionally, the Monte Carlo Potts model is utilized to simulate grain growth assuming a uniform temperature field, and thus, a uniform site selection probability. However, it may be possible, and even desirable, to have temperature gradients at the grain scale. To model grain growth in the presence of a local temperature gradient, a site selection probability function that biases the Monte Carlo sampling based on the local temperature may be used. The primary biased site selection probability function in the literature for handling local temperature variations assumes a grain growth exponent n = 2, where the average grain diameter grows proportional to t^1/n, where t is time. However, there is experimental evidence that the grain growth exponent may deviate from n = 2. In this work, a more general site selection probability function is derived that allows for different growth exponents. The Monte Carlo simulation is then linked to a control algorithm to determine the temperature field history to move towards a target microstructure. This work builds an interface for industrial control, allowing strategic temperature manipulation for achieving the desired material microstructure statistics, and therefore systematic advances in reducing the materials development cycle are expected.

We investigate how the physical microstructure of electrode layers in multi layer ceramic capacitors (MLCCs) can influence the electrical microstructure using a three dimensional finite element model. Our finite element code [1] solves Maxwell&’s equations in time and space for an alternating electric voltage in the milli- to mega- Hertz frequency range across a simple configuration consisting of a ceramic microstructure with planar electrodes on either side. This allows us to compare our results with experimental impedance spectra. The strength of our finite element approach is that we can visualise and measure the distribution of electric field and current density in three dimensions providing information on which microstructural features contribute most to electrical heterogeneity.
In this work we focus on the electrode. We consider both the morphology (electrode roughness) and material heterogeneity (pore and ceramic inclusions within the electrode). Our model consists of rough electrode surfaces with granular ceramic layers. To help analyse our systems we have developed a statistical analysis of conduction pathway lengths, provided by a stream trace of the current density vector field [2]. This links the electrical microstructure and the simulated impedance spectra, allowing for a more detailed explanation of the physical meaning of impedance measurements.
We start with a simple system composed of a homogenous layer of ceramic material with a rough interface between the electrodes similar to the work by Samantaray et al [3]. We add resistive grain boundaries to this model and examine the electrical microstructure to explain changes in the impedance spectra. This approach is repeated for other microstructural features so that their contribution to the electrical response can be studied both in isolation and in combination with each other. Finding the features that dominate the electrical response will enable us to prioritise areas for further development in the optimisation of MLCC devices.
[1] J.Dean, J.H. Harding and D.C. Sinclair; J. Amer. Ceram. Soc. 97 (2014) 885.
[2] J.P. Heath, J.Dean, J.H. Harding and D.C. Sinclair; J. Amer. Ceram. Soc. 98 (2015) 1925.
[3] M.M. Samantaray, A. Gurav, E.C. Dickey and C.A. Randall; J. Amer. Ceram. Soc. 95 (2012) 257.

The 7xxx series of aluminum alloys, consisting primarily of aluminum, zinc, magnesium, and copper is of high value to the aerospace industry and accordingly requires a good balance between strength and stress corrosion cracking (SCC) resistance. This work focuses specifically on 7050 aluminum alloys and investigates the effects of 18 different processing treatments on stress corrosion cracking (SCC) susceptibility. Traditional SCC tests are time-consuming and expensive. Therefore, this work explores using potentiodymanic polarization curves as a method of high-throughput SCC testing by quickly obtaining information on corrosion potential, corrosion current density, and open circuit potential for each sample. By combining this information with the relevant characteristics of the grain boundary microstructure, obtained by SEM, the severity of the SCC can be plotted in a style similar to an Ashby Deformation Map. These maps allow for the relative quantitative comparison of SCC as a function of processing methods and their resulting microstructures.

Abstract:
High-temperature oxidation of silicon-carbide nanoparticles (nSiC) underlies a wide range of technologies from high-power electronic switches for efficient electrical grid, thermal protection of space vehicles, to self-healing ceramic nanocomposites. Our combined quantum molecular dynamics (QMD) and reactive molecular dynamics (RMD) simulations revealed unexpected formation of nanocarbon products during high-temperature oxidation of nSiC. The nanocarbon with a unique porous geometry may find various applications including supercapacitors, battery electrodes, and biomedical imaging. We first performed small QMD simulations, which were then used to train an RMD force field using a multi-objective genetic algorithm. Subsequently, 112 million-atom RMD simulations were performed to encompass large spatiotemporal scales required for the formation of enough reaction products to assess their geometry.
This research was supported by the Department of Energy (DOE), Office of Science, Basic Energy Sciences, Materials Science and Engineering Division, Grant # DE-FG02-04ER-46130.

Iron is widely used as structural material and it is well-known that strength of iron significantly increases due to alloying elements such as silicon, even though it contains only a few atomic percent alloying element. However, reason of such highly nonlinear alloying effect has not been fully clarified from atomic level. Yield strength of coarse grained bcc metals is mainly controlled by mobility of screw dislocation as a result of kink nucleation and migration in screw dislocation. Therefore, atomistic understanding of solute atom effect on kink nucleation and migration processes is crucial for understanding of alloying effect, and moreover, predicting yield strength.
In this study, a framework for predicting macroscopic yield stress of dilute Fe-Si alloy is presented. The framework is fully based on atomistic analysis of screw dislocation motion, such as kink nucleation and migration, using newly developed interatomic interaction between Fe and Si based on first-principles density functional theory. Using the interatomic interaction, we found that screw dislocation and solute Si atom have a short-ranged attractive interaction.
Firstly, activation barriers of kink nucleation and migration processes were obtained as a function of applied shear stress using nudged elastic band (NEB) method. Si significantly decreases the activation barrier of kink nucleation process when the kink approaches to the Si atom and vice versa, and its magnitude decreases with increasing applied shear stress. Si also strongly affects to the kink migration process by increasing the activation barrier of kink migration. Based on these atomistic information of activation energy and using transition state theory, we formulated frequencies of kink nucleation and migration as a function of solute concentration, temperature and applied shear stress. Then, eventually yield strength, i.e. critical resolved shear stress (CRSS), was formulated as a function of the frequencies.
Finally, predicted yield strength using this framework was compared with experimental result in various temperatures, solute concentrations and strain rates. Our framework quantitatively reproduces temperature, solute concentration and strain rate dependences of yield strength without using any empirical data.

Hydrogen has attracted much attention in iron/steel research since its presence can significantly degrade the mechanical properties of high-strength steel. One of the important issues is the trap state of hydrogen with various defects in steel, involving carbon and vacancy. In addition, the interaction between dislocations and hydrogen is considered to play an important role in hydrogen-related fractures for metals; it has been reported that hydrogen affects the dislocation stability/mobility in iron. However, the effects of carbon on the interaction of hydrogen with dislocation and the role of carbon and hydrogen on the void formation process in iron remain unclear. Hence, at first, within the atomic scale, we performed ab-initio simulations based on Density Functional Theory to investigate the stability and interactions of hydrogen and carbon around point defects in Fe-C-H systems. We found, among others, that as the number of trapped carbon increases, the accumulation of hydrogen around vacancy is reduced and vice versa. Furthermore, we have extended the system to a much larger scale by developing an interatomic potential for Fe-C-H system to model non-equilibrium processes using Molecular Dynamics. The effect of carbon on the distribution of hydrogen near defects, the binding of hydrogen with defects as well as the effects of hydrogen and carbon on the process of nucleation and void formation at various temperatures are analyzed.

Proton exchange membrane fuel cell (PEMFC) have been developed intensively for applications in residential co-generation systems or electric vehicles. Low temperature fuel cells function by converting the energy released in the oxidation of H₂ or other hydrogen containing fuels into electrical energy. If pure H₂ is used as fuel, Pt is a good anode material, but Pt anodes are deactivated by CO in the H₂ fuel. Among various anode catalysts developed, Pt-Ru alloys are still best candidates. The Pt-Ru alloy exhibit both high CO-tolerance and acceptable durability under practical operating conditions. Recently some studies reported that well mixed Pt and Ru in the Pt₂Ru₃ alloys catalysts can improve the tolerance level of CO poisoning. At the present there is a great challenge to design anode catalysts that can accommodate CO density 300 ppm ~ 500 ppm for the cost reduction and toward larger scale commercialization. CO tolerance mechanisms of monodisperse Pt-Ru nanoparticle supported on high surface area of carbon black investigated by different experimental groups. On the other hand, quantum chemical calculation, investigated the adsorption phenomena of atoms on the catalytic surface area. However, still it is not clear of CO adsorption and its relationship to the structural change of catalysts and changing it sequence. To get desirable performance of the Pt₂Ru₃ catalyst we need to know exact atomic arrangement of Pt and Ru in the Pt₂Ru₃ alloys. Therefore in this study, we have applied a quantum chemical calculation and materials informatics to elucidate the behavior of the Pt₂Ru₃ alloy catalyst against CO at atomic level.
reference
[1]ECS Transactions, 61(13), (2014) 1-6.Md. Khorshed Alam and Hiromitsu Takaba

We performed first principles spin polarized calculations to study electronic properties of Hydrogen and Fluorine terminated armchair blue phosphorene nanoribbons (ABPNRs). We considered the ABPNRs with fully passivated and partially passivated edges with H and F atoms. We found fully edge hydrogenated and fully edge flourinated ABPNRs are indirect band gap semconductors. But partially edge saturated ABPNRs with H and F atoms exhibit half metallic behavior. Partially edge passivation will introduce new electronic states around the fermi level results into spin-up and spin-down channels. Our results suggest that fully edge passivated ABPNRs have potential to use in semiconducting device applications and partially edge passivated ABPNRs can be used in spintronic applications.

Polymer nanocomposites have shown great potential as dielectric materials. It has been found that the introduction of nanoparticles can lead to a significant enhancement in the dielectric breakdown strength. Further improvement has been observed when specific functional groups that have the potential to trap electrons have been grafted to the particle surface. The optimization of these functional groups is a key part of our research effort. At present, there is no well-accepted theory that describes the mechanism behind the observed improvements, although there is general agreement that scattering and trapping of carriers is important.
We are applying quantum mechanical computation as well as material quantitative structure-property relationship (MQSPR) modeling to the optimization of nanodielectric materials. Instead of trying to model the whole process of dielectric breakdown, we have made a limited number of assumptions that are well accepted. The key assumption is that the breakdown occurs via electron avalanche. The breakdown phenomenon starts with electron intrusion, where a small number of free electrons move into the bulk material and induce more free electrons by impact ionization. The foreign species, like the functionalized nanoparticles, may provide additional low energy states, and electrons would be more likely to reside in these states so that the energy distribution of entire injected carrier population could be shifted down at equilibrium. This process could be described as follows: Charged Polymer + Neutral Particle → Neutral Polymer + Charged Particle. The effect of functional groups can be simply estimated by calculating the charge transfer energy cost from charged polymers to neutral particles. The electron affinity (EA) and ionization energy (IE) were used to measure the energy cost for negative and positive charges.
Direct quantum computation of the grafted nanoparticles requires enormous resources and was infeasible for broad studies. In this study, the quantum computation was done using the DFT method b3lyp/6-311++G** on Si(OH)₃ based functional groups with only a small contribution to the HOMO/LUMO orbitals from Si(OH)₃ observed. The machine learning model was built using partial least square with feature selection.
A clear reverse correlation was observed between breakdown strength of polymer nanocomposite (epoxy with functionalized silica) and EA+IE. MQSPR models for EA and IE were also built. The models demonstrate good predictive capability. The 10-fold cross-validation R² is 0.93 and 0.85 for EA and IE respectively.
In correlating the energy change between charged matrix polymers and functionalized nanoparticles with the dielectric breakdown strength, it was observed that the breakdown strength increased as the energy change increased. Thus, the energy change could be used as an indicator to predict the breakdown strength change in nanodielectrics due to the addition of functional groups to the nanoparticles.

Density functional theory (DFT) has been successful in describing the properties of materials, specically ground state properties, e.g., structural, mechanical, and vibrational properties. To develop and design new functional materials such as photovoltaic devices, however, understanding ground state properties is not always sufficient especially when electrons are excited in the key physical processes. The GW-Bethe Salpeter Equation (GW-BSE) approach is the fully ab initio method of choice for general materials problems involving one- and two-electron excitations (i.e., band properties and optical excitations). However, GW calculations are typically extremely expensive compared to DFT calculations which has limited the regular application of GW to a large-scale materials problem. This is in part due to the computational load typically requiring enormous numbers of FFTs for plane wave bases; separately, the memory requirements for most GW algorithms can be daunting.
We describe our collaborative efforts to develop new software and new algorithms that permit GW calculations to be performed on large-scale parallel computers efficiently. We illustrate calculations of the dielectric screening matrix which is one of the most expensive parts of a GW calculation. Our approach uses a real-space representation of the polarizability that avoids extensive use of FFTs, and we compare its behavior to the more conventional G-space GW approaches. The self-energy correction calculations via both real-space approach and eigendecomposition method of dielectric matrix are presented as well. We also summarize the capabilities of our highly scalable Car-Parinello ab initio molecular dynamics simulation package "OpenAtom" [1] which our GW software is being interfaced. We briefly describe how OpenAtom leverages the Charm++ parallel libraries [2, 3] to achieve admirable parallel scaling on large problems as well as signicantly reduce the complexity of software parallelization from the viewpoint of the scientic user.
[1] http://charm.cs.uiuc.edu/OpenAtom
[2] E. Bohm, A. Bhatele, L. V. Kale, M. E. Tuckerman, S. Kumar, J. A. Gunnels and G. J. Martyna: "Fine grained parallelization of the Car-Parrinello ab initio MD method on Blue Gene/L" IBM J. RES. & DEV. VOL. 52 NO. 1/2, 2008.
[3] G. Martyna, E. Bohm, R. Venkataraman, L. Kale, and A. Bhatele: Chapter 5: OpenAtom: Ab-initio Molecular Dynamics for Petascale Platforms, BOOK: Parallel Science and Engineering Applications: The Charm++ Approach

An understanding of the structure and properties of the aqueous interface of fats, or triacylglycerols (TAGs), is important for a number of applications including food processing, nanomedicine[1] (via the implementation of solid-lipid nanoparticles) and laundry detergents.[2] The effective removal of crystallized TAGs from fabrics and other surfaces can be expensive in both time and energy, requiring high temperatures (330K+) even with modern detergents.[2] Recently, however, it has been shown that employing nanodiamonds (NDs) in combination with an anionic surfactant (sodium dodecylbenzene sulfonate, SDBS) can lead to the removal of tristearin (TS), a model fatty acid, at room temperatures.[2] While raising significant prospects for the use NDs in cold-water laundering products, the precise mechanism for the action of NDs and surfactants at the TS-aqueous interface is still uncertain. Molecular dynamics (MD) simulations of these systems can provide important information about the interactions of the different components at the atomistic level, and give clues to the optimization of these ND/surfactant formulations for improved performance.
For the first time, the TS-aqueous interface has been characterized using MD simulations.[3] The lipids form an ordered gel phase, with the fatty acid tails packing in a hexagonal arrangement. In the presence of SDBS, the structure of the TAG phase shows no sign of structural instability, despite the ability of surfactant molecules to embed within the TS bilayer. However, in the combined presence of NDs and SDBS, the structure of the TS bilayer is disrupted. We find that the surface charge density on the NDs is a major factor in the complex interplay of the interactions between the different species. Our findings provide guidance for the development of energy-efficient solutions for the low-temperature removal of crystallized fats, as well as advancing our understanding of the important TAG-aqueous interface.
[1] Almeida, A and Souto, E; Solid lipid nanoparticles as a drug delivary system for peptides and proteins, Adv. Drug Del. Rev.,2007, 59, 478-490.
[2] Cui, X., et al.; Nanodiamond promotes surfactant-mediated triglyceride removal from a hydrophobic surface at of below room temperature, ACS Appl. Mater. Interfaces, 2012, 4, 3225-3252.
[3] Hughes, Z.E. and Walsh, T.R.; Tristearin bilayers: Structure of the aqueous interface and stability in the presence of surfactants, RSC Advances, 2015, 5, 49933-49943.

The electronic transport in solids with a large number of impurities is still an unclear issue, where the interference between the electronic wavefunction and aperiodic potentials has multiple consequences. Recently, branched nanowires with designed three-dimensional (3D) morphology have been obtained, and they have wide applications in energy conversion and storage devices [1]. Nonlinear electrical properties of branched nanowires have been also reported [2]. Since the presence of impurities avoids the use of the Bloch theorem as well as the reciprocal lattice, such system should be addressed in the real space. In this work, a renormalization plus convolution method developed for the Kubo-Greenwood formula [3] is used to calculate the electronic transport in branched nanowires. We report enhancements to the ballistic AC conductivity when periodically or quasiperiodically placed Fano-Anderson impurities are introduced to an otherwise periodic nanowires, which is connected to two semi-infinite periodic leads at its ends [4]. Moreover, the temperature variation analysis suggests the possibility to observe these resonant AC conducting peaks for specific electronic filling and applied electric field oscillating frequency even at room temperature. Experimentally, the chemical potential position in the band structure of a nanowire can be modified by an applied gate voltage [5]. Finally, the AC conductance spectra of nanotubes with Fano-Anderson impurities will also be presented.
This work has been partially supported by UNAM-IN113714. Computations were performed at Miztli of DGTIC, UNAM.
[1] C. Cheng and H.J. Fan, Nano Today7, 327 (2012).
[2] D.B. Suyatin, et al., Nano Lett.8, 1100 (2008).
[3] V. Sanchez and C. Wang, Phys. Rev. B70, 144207 (2004).
[4] V. Sanchez and C. Wang, Phil. Mag.95, 326 (2015).
[5] J. Moon, et al., Nano Lett.13, 1196 (2013).

Since, the discovery of graphene, 2D materials has attracted immense interest among the scientific community. It has given a new route to explore the carbon based 2D materials. Carbon can exists in several hybridization states, which gives it an unique ability to adopt novel atomic arrangements such as rings, sheets, tubes, and solids. In this study, we report a new sp² hybridized 2D allotrope "pentahexoctite” made out of 5-6-8 rings of carbon. This planar sheet has Cmm symmetry and phonon calculation confirms its dynamical stability. It has mechanical strength comparable with that of graphene. Interestingly, this 2D sheet is metallic and has band structure with combinations of direction dependent flat as well as dispersive bands. In addition, carbon nanotubes (CNTs) are generated out of this sheet, shows chirality-dependent electronic and mechanical properties. Along with these remarkable properties, the "pentahexoctite" sheet joins the family of 2D allotropes of carbon. [1]
1. Babu Ram Sharma, Aaditya Manjanath, and Abhishek K. Singh, Sci. Rep. 4, 7164 (2014

The tremendous development of portable electronics and sensor networks makes it an urgent requirement to develop sustainable and stable energy sources for them. Recently, triboelectric nanogenerators (TENGs) based on contact electrification and electrostatic induction have shown unique merits including large output power, high efficiency, low weight, cost effective materials, and simple fabrication. Requirements for continuing improving their output performance demand rational design and careful optimization of both materials and structures of TENGs. Thus a thorough theoretical understanding of TENGs and their systematical simulation method is completely urgent in the whole research field.
In this research, we developed the first theoretical governing equation (V-Q-x relationship) and the first lumped-parameter equivalent circuit model for TENGs. Then the first systematical simulation tool and method for TENGs is built, which includes the complex coupling of both electrostatic and circuit simulation. Utilizing this simulation tool, the load characteristics of TENGs have been clearly uncovered. When TENGs are connected with resistive loads, a “three-working-region” behavior is shown because of the impedance match between the generator and the load. There exists an optimum load resistance to maximize the generator energy output, which can be carefully controlled by the TENG structural parameter and the frequency of the external mechanical motion. Besides, when TENGs are utilized to charge a capacitor in a periodic external motion, it is equivalent to utilizing a DC voltage source with an internal resistance to charge the capacitor. Same as the resistive load, there exists an optimum load capacitance under which the maximum energy can be stored. This optimum load capacitance is proportional to both the inherent TENG capacitance and the charging cycle number.
With the developed theoretical model and simulation method, optimized performance of triboelectric nanogenerators has been reached. The attached-electrode contact-mode and sliding-mode TENGs need to maintain the minimum gap size to be much smaller than their effective dielectric thickness. The electrostatic shield effect of the primary electrode is the main design consideration of single-electrode TENGs. Contact-mode freestanding TENGs have superior linear characteristics and sliding-mode freestanding TENGs have excellent height tolerance of the moving object. For the most-complicated grating structure, an optimum number of grating units and an optimum unit aspect ratio due to the edge effect can be calculated with my model. The theory presented here is the first interpretation and analysis of the TENGs&’ working principle, clearly showing its unique operation characteristics, which will be able to serve as important guidance for rational design of the device structure as a power source in specific applications and self-powered systems. [1]
Ref:
1. S. Niu, Z. L. Wang, Nano Energy, 2015, 14, 161.

We focus on the development of novel hydrogen storage alloys of Nd-Mg-Ni system because their attractive kinetic properties promoted by catalytic elements Nd and Ni. Searching for the Mg-based multicomponent compound and synthesizing it are significant for in-situ formation of ultrafine NdH_x-MgH₂-Ni/Mg₂NiH₄ composites with excellent hydrogen storage properties. In order to find out the best composition of addition, the phase equilibria at the Mg-rich corner is constructed by Calphad-type thermodynamic calculations and verified by a series of equilibrated alloys at 400 and 500 °C. There are four ternary compounds existing in the Mg-rich corner: Nd₄Mg₈₀Ni₈, Nd₁₆Mg₉₆Ni₁₂, NdMg₅Ni and NdMg₂Ni. The crystal structure of Nd₄Mg₈₀Ni₈ and Nd₁₆Mg₉₆Ni₁₂ are characterized by the synchrotron powder X-ray diffraction (SR-PXRD). Nd₄Mg₈₀Ni₈ has the crystal structure of space group I4₁/amd with lattice parameters of a = b = 11.2743 Å and c = 15.9170 Å, while Nd₁₆Mg₉₆Ni₁₂ has the crystal structure of space group of Cmc2₁ with lattice parameters of a = 15.3422 Å, b = 21.6750 Å and c = 9.4868 Å. The hydrogenation of the compounds leads to the decomposition of this compound into nanocomposites of Mg₂NiH₄, MgH₂ and NdH₂. Among the four compounds, the Nd₄Mg₈₀Ni₈ shows the largest hydrogen capacity of 5.18 wt.% H₂ at 350 °C under 4 MPa H₂. The hydriding and dehydriding kinetic properties of those ternary compounds are enhanced greatly compared with other Mg-riched Nd-Mg-Ni alloys. Furthermore, the hydriding/dehydriding (H/D) kinetics are systematically investigated by the Chou kinetic model to compare the H/D rates and analyze the kinetic mechanism.

Despite been made of simple elements, bone tissue gets its remarkable mechanical performance from a complex hierarchical organization. Through a fine tuned structure, this nanocomposite material combines the best properties of a strong and stiff mineral phase embedded in a soft yet tough organic matrix. Whereas bone nanostructure is well documented, the mechanisms that allow bone tissue to achieve its outstanding mechanical behavior remain unclear. In this study, we developed a coarse-grained molecular model of bone tissue to explore the intricate relationship between its nanostructure and mechanics. We used this model to study the role of extrafibrillar mineralization on the tissue&’s mechanical response under both tension and compression.
A simple bead spring or “coarse-grained” model is used to represent bone tissue. The geometry of the coarse-grained model is based on previous full atomistic simulations. In the simplified coarse-grained representation, one collagen bead represents about 180 atoms. The collagen molecules are replicated in order to form a fibrils of diameter d = 40 nm. The fibrils are mineralized in silico by filling the gaps in the model with hydroxyapatite elements. The extrafibrillar mineral is made of elongated plate-like crystal of thickness t=0.8 or 2.5 nm aligned to the collagen fibrils. The crystals are surrounded by a hydrated layer. A total of four crystals is created between every fibrils.
After equilibration, the model presents the main features seen experimentally. Individual crystals can be discerned as the hydrated layer surrounding the crystals prevents them from merging. The characteristic banding pattern of the collagen fibrils is also conserved. The response of the model in compression is in very good agreement with experiments performed on bone nanopillars. Increasing the crystal thickness from 0.8 to 2.5 nm leads to significant increase in the axial elastic modulus (8.6 vs 15.7 GPa) and strength (0.19 vs 0.53 GPa). The thickness of the crystals play a role in the deformation mechanisms of the tissue. For thin crystals, the crystal phase deforms by axial buckling. By contrast, a structure with thicker crystal collapse in a catastrophic Euler type buckling. The structure presents some signs of delamination between the mineral crystals and dislocations inside the crystals. In tension, the model for thickness of 0.8 and 2.5 nm exhibit similar elastic properties as in compression (9.8 vs 15.6 GPa). Yield stress arise at 0.39 and 0.46 GPa and can be related to collagen-mineral interactions.
The model developed here shed some light on the deformation mechanisms of bone tissue. Tissue elasticity and compressive strength are directly dependent on the amount of mineral and crystal thickness. The yielding properties in tension are related to collagen-mineral and mineral-mineral interactions. Identifying bone behavior at the nanoscale is essential to developing guidelines for the engineering of mineralized tissues.

The current numerical models to predict the mechanical behaviour of materials can be classified into two approaches namely continuum models and discrete element models (DEM). While continuum models have been studied in great detail, some aspects such as incorporating the effect of microstructural features such as grains texture, cracks and other defects proves difficult. As an alternative, DEM are gaining importance.
Discrete element methods describe the material in terms of lumped masses interacting through constitutively prescribed forces. DEM are taken to describe length and time scales in between molecular dynamics (MD) and continuum models. Thus the interaction potentials between the lumped masses do not arise from physical arguments as in MD. Most commonly, the interactions have been taken to occur through linear springs. The main difficulty in this approach lies in extending the models to inelastic deformations. There has been no systematic approach to developing inelastic interaction elements.
In this work we propose a discrete element model which may be able to overcome these limitations and handle elastic as well as inelastic problems. In our approach, the domain is discretised into lumped masses and the interaction forces are derived from the continuum strain energy density function. The deformation of the lattice element is described in terms of a Lagrangian strain and corresponding forces are obtained from the continuum strain energy density function. We believe this will allow discrete element methods to be generalised to describe inelastic material behaviour. In this work, we reproduce standard results of Elasticity in order to describe and validate the approach in detail.

Accurate modelling of hydration in polymers can provide critical information, as the presence of water influences the structure, and electrical and mechanical properties. Previous studies of the hydration of polyimide have only considered systems of fully polymerized, monodisperse polyimide chains, neglecting any possible effects from residual monomers and a stable intermediate reaction step, poly(amic acid). Atomistic studies of polyimide are conducted using structures generated using a newly developed dynamic polymerization technique that combines molecular dynamics and Monte Carlo methods. This technique results in a more realistic structure and allows for tracking both the network structure and the hydration behavior as the system polymerizes. The final system is used to investigate the effects of water concentration and water model parameters on hydration behavior.

A detailed atomic understanding of how a solid-liquid interface affects a chemical reaction can greatly assist in guiding the design and engineering of next generation catalytic materials. One reaction of particular interest, due to its relation to water-splitting and aqueous proton transport, is the auto-ionization of water. This process results in water ions that will subsequently recombine to water. Here, we present the results of an ab initio molecular dynamics study aimed at characterizing the recombination reaction in the environment of an electrochemical double-layer. Previous studies have demonstrated that a perfectly ordered Pt electrode, in contact with liquid water, can support long-lived nanoscale heterogeneity in the density and mobility of water molecules at the interface. We discuss the effect of this heterogeneity on the recombination dynamics of water ions. We investigate this by using QM/MM simulations where the solid and its interactions with the water molecules are treated classically, whilst the water molecules are treated at the DFT level. The simulations compare the water ion recombination mechanism and reaction times at the interface to that of the bulk liquid.

Surface segregation refers to the phenomenon that chemical composition at the surface of alloy materials differs from the corresponding value in their bulk region. Knowledge on surface segregation is pertinent to various engineering applications such as adsorption, wetting, oxidation, corrosion, electrical contact, friction and wear, crystal growth, and catalysis. In this presentation, we reported our research on accurately predicting the influence of surface segregation on the functional properties of nanostructured alloy materials using a multiscale computation technique. The employed multiscale computational approach consists of three hierarchical components: (1) developing reliable interatomic potentials for alloys with the modified embedded atom method based on the first-principles computation data, (2) applying these atomic interaction potentials to determine the chemical composition of extended and nanoparticle surfaces of alloys using the atomistic Monte Carlo method, and (3) evaluating properties of surface-segregated nanomaterials using the first-principles and/or mesoscale computation methods. We have successfully applied our multiscale computation to design catalytic and magnetic nanoparticles.
Case I: Platinum (Pt) alloys are the most active catalysts for oxygen reduction reaction (ORR) occurring in proton exchange membrane fuel cells. The surface chemical composition in these catalysts determines the electronic structure and further their catalytic performance for ORR. In the present study, we used our multiscale computation method to elaborate the relation between the surface composition, electronic structure, reaction pathway, and reaction rate of ORR on nanosegregated Pt alloy catalysts.
Case II: For both FePt and CoPt nanoparticles, our multiscale model predicted that it was energetically favorable to exchange the surface Fe and/or Co atoms with the interior Pt atoms and thus to induce the Pt surface segregation. Comparing the magnetic properties of bulk-terminated and surface-segregated nanoparticles, we found that the surface segregation processes in the FePt and CoPt nanoparticles could cause a decrease in their total magnetic moments, a change in their (easy and/or hard) magnetization axes, and a reduction in their magnetic anisotropy.
Hence, our computational approach is a valuable tool for computational design of novel functional nanomaterials.

In this paper, we first present a general multiscale model based upon the generalized Peierls-Nabarro model to describe the interlayer dislocations in bilayer materials. In our model, the bilayer material is divided into two linear elastic 2D sheets which are used to describe each individual layer, the strains in each sheet can be relaxed by both in-plane elastic deformation and also by out-of-plane buckling; the deformation of 2D sheets can be described by classical linear elastic thin plate theory. The interface between these two sheets has a relative displacement in the presence of dislocations, and the two sheets are connected by a nonlinear potential. In our model, a 3-dimensional version of generalized stacking-fault energy (GSFE) which obtained from first principle calculation is used to describe the interlayer bonding between the two sheets. The structure and deformation of the bilayer with a dislocation is determined by the force balance between the local stresses in the sheets and the restoring force from the interlayer bonding. We apply this approach to determine the structure and energetics of four interlayer dislocations in bilayer graphene: edge, screw, 30^o, and 60^o (these angles represent the Burgers vector direction relative to the line direction). A pronounced buckling is formed at the position of partial dislocation to relax the strain induced by edge component of a partial dislocation. We determine the buckling amplitude, in-plane strain distributions, partial dislocation structures, core widths and dislocation energies. We found the dislocation core width in buckled structure decreases as the increase of edge component of its Burgers vector, which is different from the flat case. The results from our multiscale model provides an excellent quantitative match to the atomistic results. From these results, we construct a simple analytical model to describe the buckling and in-plane deformation of bilayer graphene with dislocations of arbitrary Burgers vector and demonstrate that the analytical model is in excellent agreements with the simulation results.

Biomass pyrolysis with subsequent upgrading of pyrolysis oil (also known as bio-oil) into liquid hydrocarbons is one of the major strategies of utilizing biomass for renewable energies. However, a significant challenge with bio-oil is its high oxygen content, which makes it unsuitable for combustion applications before blending with conventional transportation fuels. Hydro-deoxygenation (HDO) is a major route of upgrading bio-oil by lowering its oxygen content. Recently, Roman-Leshkov and co-workers have shown MoO₃ to be active for HDO of bio-oil model compounds at atmospheric pressures and temperatures of only 573-673 K and evidence for the occurrence of a reverse Mars van Krevelen mechanism involving oxygen vacancies has been found.
In this study, we have performed Density Functional Theory (DFT) calculations on the acetone (CH₃COCH₃) hydro-deoxygenation to propylene (CH₃CHCH₂) as a model reaction to gain insights into the thermodynamics and energetics of oxygen vacancy formation on MoO₃ surfaces and subsequent deoxygenation pathways. A perfect O-terminated α-MoO₃ (010) surface represented by a 4 x 3 x 4 supercell is reduced to generate a terminal oxygen defect site in the presence of H₂. Hydrogen dissociatively adsorbs on adjacent surface oxygen sites and combines to form a water molecule. CH₃COCH₃ adsorbs at the oxygen deficient Mo site through a O-Mo bond and dehydrogenates to form CH₃COCH₂ by transfer of a hydrogen atom to an adjacent terminal oxygen site. The hydroxyl (OH) then hydrogenates the secondary carbon atom to form CH₃CHOCH₂. The reaction pathways for the conversion of CH₃COCH₃ to CH₃CHCH₂ are investigated with a Nudged Elastic Band (NEB) method that allows us to search for unknown transition states and reaction mechanisms. This results in the discovery of a new low transition state reaction mechanism for the deoxygenation step. The deoxygenation then releases CH₃CHCH₂ into the gas phase and regenerates the terminal oxygen atom on the Mo site by completing the reoxidation of the reduced MoO₃ (010) surface back to perfect O-terminated MoO₃ (010) surface.

To meet the ever-increasing demand for large-scale integration and cost-effective technologies, basic components of devices are getting smaller, with size ranging from the nanometer-scale to atomic-scale. Quantum effects are thus important in the modeling of the systems within these scales. With rapid progress in computing power of present supercomputers in Japan, computational materials design, CMD^® techniques [1-2] with quantum effects integration, have paved way for new materials and mechanisms [3] .
In this meeting, we will present some case studies of CMD^® application on surface and interfaces which resulted to understanding of (1) switching mechanism of transition metal oxide (TMO)-based resistance random access memories (RRAMs) [4,5]; (2) origin of magnetic anisotropy of Co/Ni multilayers for magnetic recording devices [6]: (3) dynamics of hydrogen in surface and interfaces for hydrogen storage using our own computational code NANIWA series [7,8].
Furthermore, our computational modeling of systems has gone beyond drawing the mechanisms of experimentally studied/synthesized materials. Here, we present new materials out of rational design alone using CMD^® techniques. Specifically, we show less and non-precious metal-based NOx reduction catalysts using DFT and KMC simulations for automobile industry (case study 4) [9,10]. Comparison with the conventional pure precious metal catalysts and the origin of the differences will be discussed.
References:
[1] H. Kasai, H. Akai, and H. Yoshida, Keisanki Materials Design Nyumon, Osaka University Press, Osaka (2005).
[2] H. Kasai and M. Tsuda, Computational Materials Design Case Study 1: Intelligent/Directed Materials Design for Polymer Electrolyte Fuel Cells and Hydrogen Storage Applications, Osaka University Press, Osaka (2008).
[3] Patent list: http://www.dyn.ap.eng.osaka-u.ac.jp/web/patent.php.
[4] H. Kishi, A. A. A. Sarhan, M. Sakaue, S. M. Aspera, M.Y. David, H. Nakanishi, H. Kasai, Y. Tamai, S. Ohnishi,and N. Awaya, Jpn. J. Appl. Phys. 50, 071101 (2011).
[5] H. Kishi, T. Kishi, W. A. Dino, E. Minamitani, H. Akinaga, H. Nakanishi, and H. Kasai, J. Comput. Theor. Nanosci. 5, 1976 (2008).
[6] K. Kojima, W. A. Dino, M. Suzuki, T. Yasue, K. Kudo, N. Akutsu, E. Bauer, T. Koshikawa, and H. Kasai, J.Vac. Soc. Jpn. 56, 139 (2013).
[7] H. Kasai and A. Okiji, Prog. Theor. Phys. Suppl. 106, 341 (1991).
[8] H. Kasai, W. A. Dino, and A. Okiji, Surf. Sci. Rep. 43,1 (2001).
[9] H. Kishi, A. A. B. Padama, R. L. Arevalo, J. L. V. Moreno, H. Kasai, M. Taniguchi, M. Uenishi, H. Tanaka and Y. Nishihata, J. Phys.: Condens. Matter 24, 262001(2012).
[10] R. Arevalo, MCS Escaño, H. Kasai. J. Vac. Sci. & Tech. 33 021402 (2015).

In this new age of flexible electronics, highly durable (and yet flexible) printed circuit boards (PCBs) are especially critical to the operation of such bendable electronic devices. From a materials&’ perspective, the adhesive properties of the aromatic polyimide layer to the metal surface very much determines the overall long-term performance of such devices [1]. Given its high mechanical strength, and thermal and chemical resistance, poly(pyromellitic dianhydride oxydianiline) (PMDA-ODA) has been the choice polymeric substrate for the copper in these flexible PCBs [2-3]. To date, the poor adhesion in PMDA-ODA/Cu hybrid-interface has led to the limitations of its use as flexible and bendable PCB materials [4-5]. Despite of the significant demands for the improvement of PMDA-ODA/Cu hybrid-interface, there is still no clear principle for the early stage of the adsorption of organic film on metal surface.
In this work, using first-principles density-functional theory (including van der Waals force corrections), we study the fundamental physio-chemical properties of the molecular fragments of PMDA-ODA on both pristine Cu(111), as well as oxidic O/Cu(111). We make the selection of the adsorption models in the case of each fragment and surface by comparing binding energies of the molecular fragments at various adsorption sites. Henceforth we analyze the adsorption structures and the electronic structure of chosen models. Our results indeed show that the accurate picture of the adsorption models of PMDA-ODA moieties on Cu surfaces and may pave the road to explore this hybrid-interface by offering an increase understanding of the physio-chemical reaction on metal surface.
[1] B. Noh, J. Yoon, and S. Jung, Int. J. Adhes. Adhes.30, 30 (2010)
[2] M. Ramos, Vacuum 64, 255 (2002)
[3] Y. Takagi, Y. Gunjo, H. Toyoda , and H. Sugai, Vacuum83, 501 (2009)
[4] R.F. Saraf, J.M. Roldan, T. Derderian, IBM J. Res. Dev.38, 441 (1994)
[5] S. Bang, K. Kim, H. Jung, T. Kim a, S. Jeon a, J. Seol, Thin Solid Films558, 405 (2014)

Epitaxial stabilization of thin films is a powerful technique for growing novel materials with advanced functionality. While metastable structures can be kinetically stabilized, its relative thermodynamic stability also governs its ease of synthesis. The thermodynamic stability of a thin film can be modeled by a simple continuum model that incorporates energetic contributions from the substrate-film interface, film surface, and substrate-induced strain of the film. Accurate predictions of these contributions can accelerate the discovery of new thin film materials. We use density functional theory (DFT) to elucidate experimental observations of epitaxial growth of rutile and anatase TiO₂ on (Ba, Sr)TiO₃ perovskite (001), (101), and (111) surfaces. We do this by calculating epitaxial interface energies, surface energies, and substrate-induced strain energies of the film and insert these quantities into a simple continuum model to evaluate thin film stability. Our DFT calculations explain a number of observed phenomenon related the preferential growth of TiO₂ rutile or the metastable anatase phase. These phenomenon include the strong/weak epitaxial stabilization of anatase on the (Ba, Sr)TiO₃ (001)/(101) surface, respectively, as well as the strong epitaxial stabilization of rutile on the (Ba, Sr)TiO₃ (111) surface. Our results demonstrate that interface energies calculated from first principles can explain the facet-selective growth of TiO₂ on perovskite surfaces. We discuss ways to use our method to predict the epitaxial stability of new materials.

Grain boundary energies vary as a function of five microscopic parameters which describe the crystallographic alignment of neighboring grains, as well as the orientation of the interface separating these grains. Determining quantitative grain boundary (GB) energies as a function of these orientational parameters is essential in order to understand the population of GBs in polycrystalline ceramics and films, which dictates mechanical properties (ductility, toughness), dielectric properties (charge trapping, leakage current), and ionic diffusion rates in polycrystalline materials. Despite their importance, GB energies of oxide systems as a function of all five microscopic parameters remain relatively unexplored for most oxide systems, due in part to a lack of validated simulation techniques available to probe these systems.
Here, we investigate the use of reactive molecular dynamics (MD) potentials in determining interfacial energies in TiO2 (rutile), a well-investigated model oxide system. Reactive MD potentials are anticipated to better recreate interface energetics, as they can account for bond breaking and reforming, for over- and under-coordinated configurations likely to occur at free surfaces and interfaces, and for mixed valence states that are expected to occur near interfaces due to charge re-equilibration. We calculate surface and interfacial configurations and energies in the TiO2 (rutile) system, using Ti/O COMB3 and Reaxff potentials, as implemented in LAMMPS. Computational predictions from these two potential sets are compared against available experimental observations, as well as first-principles calculation for select surfaces and interfaces.
Both potential sets adhere closely to the experimentally observed relative order of surface energies. Furthermore, GB energies of arbitrary free surfaces are close to those calculated theoretically by decomposing unstable surfaces into terraced stable surfaces. We adopt an approach which allows for rigid displacement of one grain with respect to its neighbor, and report loci of energy minima for a range of twin and tilt boundaries; energy minima typically require non-zero rigid lattice displacements. Very deep GB energy cusps are calculated for (101) and (301) twins, which are the most commonly observed twin systems, as well as for (100) reflection twins.

In thin film ferroelectric and related devices the interfacial structure between the ferroelectric phase like Pb(Zr_0.5Ti_0.5)O₃ or PZT, and metal electrodes, such as Pt, is primarily dictated by the synthesis history. Specifically, the seeding and evolution of the interface is considered to dictate the resulting structure of the thin film and its properties. In general, the interface free of secondary phase is considered to be beneficial for the efficient device function. However, it is known that during synthesis of PZT films with Pt electrodes the formation and the disappearance of secondary phases like the intermetallic Pt₃Pb is observed. Nonetheless, the mechanism of Pt₃Pb formation has not yet been understood in a systematic way. Here, density functional theory (DFT) and the climbing image nudge elastic band (c-NEB) method are used to calculate the migration barriers for the diffusion of Pb across the PZT/Pt and PZT/Pt₃Pb interfaces under different environmental conditions. The migration barriers are predicted to be strongly dependent on atmospheric conditions and the crystallographic phase of the PZT (i.e., tetragonal or cubic). In particular, a reversal in the Pb diffusion direction at the PZT/Pt interface is predicted to take place in the oxygen rich environment. This prediction is confirmed by experimental in situ X-ray diffraction measurements of a PZT/Pt interface.

Benefitial physical and chemical properties of metal nanocrystals highly depend on their sizes and shapes. Numerous shapes of Ag nanocrystals have been synthesized in solution with the help of structure-directing agents (SDAs). Polyvinylpyrrolidone (PVP) is a well-known SDA for promoting the growth of Ag(100) facets in ethylene glycol (EG) solution. Interfacial properties are critical in SDA-induced seletivity, but they are difficult to probe in experiment, thus the mechanisms of how SDA works are still not well-understood. In this miniscule level, molecular dynamics (MD) can serve as a great tool to reveal the enigma.
As thermodynamic equilibrium is difficult to maintain during growth, the nanocrystal shape can be greatly affected by the kinetics of atom addition and surface diffusion. For large nanocrystals, where the corner effects are negligible, the atom deposition rate onto various facets is the major factor in determining the final shape. PVP selectively binds stronger to Ag(100), thus it may slow down the rate of atom addition to such facet. A kinetic Wulff shape construction suggests that, in a reversible cube-to-octahedron transformation, the atom deposition flux ratio F₁₁₁/F₁₀₀ needs to be greater than radic;3 to form a perfect cube at steady state. Using MD simulations, we quantify the relative flux F₁₀₀ and F₁₁₁ for different PVP lengths and concentrations using simulated deposition and potential of mean force (PMF) calculation, and build correspoding shape according to the kinetic Wulff construction. By applying adsorption kinetics, we parameterize relative flux into the ratio of rate constants k₁₁₁/k₁₀₀ and trapping coefficients p₁₁₁/ p₁₀₀. The rate constant is associated with the energy barrier reflected in the PMF profile which accounts for the barrier of an incoming Ag atom finding a deposition site, and the trapping coefficient is related to the thickness of the adsorbed PVP layer. Our results show that the rate constant ratios k₁₁₁/k₁₀₀ are consistently around 1.35 regardless of concentration and chain length, which indicates that it may only be associated with the different binding affinity of PVP on Ag(100) and Ag(111). Whereas the trapping coefficient varies with PVP concentration and conformation. When both Ag(100) and Ag(111) facets are fully covered with longer PVP chains, P₁₁₁/P₁₀₀ asymp; 1.25, which gives a relative flux F₁₁₁/F₁₀₀ asymp; 1.68, predicting nanocubes with tiny {111}-corners. Nanocube will progress to truncated cubes and to cuboctahedra as the PVP concentration decreases, because the resulted insufficient capping will cause the diminishing of the difference in trapping coefficients. By using PVP with 5 repeating units, which is comparable to its Kuhn length in EG, we demonstrate why chain lentgh < 5 does not work in experiment. Our results are consistent with expertimental observations, thus illustrate the mechanisms of how polymeric SDAs select nanocrystal shapes in solution phase syntheses.

Present study deals with the extensive analysis of the mechanical response of sodium silicate glasses. We propose a new method using atomistic simulations combined with coarse-grained analysis to calculate structural and mechanical properties at different scales.
The numerical simulations were performed on large samples, and the results were compared to experiments (neutron scattering, NMR, Brillouin scattering). We showed that the adjustment of the cutoff in the non-Coulombic part of the potential function allows us not only to establish precisely the pressure/density relation, but also to obtain realistic mechanical and structural properties as well.
The systems were tested by deforming the periodic simulation box in a homogeneous way. During compression or tension the dimensions of the simulation box was reduced by a constant displacement step while the positions of the particles were rescaled in a homogeneous way. After the box displacement a new equilibrium position was searched using the Polak-Ribiere conjugate gradient algorithm. The shear deformation was done similarly by tilting the simulation box.
Combining quasi static shear and compressive mechanical deformation, we were able to reconstruct hardening yield surfaces in the 3D stress space, parameterized by the residual shear strain and densification.
Classical molecular dynamics gives us a valuable insight to understand the effect of pressure and composition on the plastic response of the material as well as to identify the causes of failure at the submicrometer scale.
We found that sodium facilitates local plastic rearrangements which reduces the shear strength but makes glasses more ductile and therefore safer to use it in structural applications.
After identifying the smallest scale where the material could be considered homogeneous and isotropic we were able to develop continuum based material models.
The observed mechanical response was described using a pressure dependent, hardening yield surface. The functions were then implemented in a finite element code. This way the microscopic results could be compared with mesoscopic real life experiments such as nanoindentation.

Computational methods for structure and property predictions offer the potential for in silico design of functional materials for purposes such as gas-adsorption. Here we evaluate the crystal energy landscapes of porous organic materials which display a variety of inclusion solvates with radically different crystal packing arrangements and hence different porosities and gas-adsorption behavior. [1] Structures capable of solvate formation typically occur at lattice energies well above the global minimum on the crystal energy landscape (with lattice energy differences of up to 60 kJ /mol possible).[2] Coupled with the fact that the solvent is usually highly disordered, their prediction by application of standard crystal structure prediction methods presents a challenge.
Here we present a Monte-Carlo solvent insertion procedure for the generation of solvated crystal structures. We apply this technique to an organic cage molecule in order to understand the energy re-ranking that occurs in the crystal energy landscape upon inclusion of the solvent contribution to the lattice energy. From this we are able to rationalize the preferential formation of a set of experimentally observed solvent inclusion polymorphs as arising from the additional stabilization provided by the solvent. Additionally, we explore how this technique can be applied to structures coming from a crystal structure prediction for the computational prediction of novel crystal phases.
References:
[1] Jones, J. T. A. et al. On-off porosity switching in a molecular organic solid. Angew. Chem. Int. Ed. Engl.50, 749-53 (2011).
[2] Pyzer-Knapp, E. O. et al. Predicted crystal energy landscapes of porous organic cages. Chem. Sci.5, 2235 (2014).

The development of materials with increased technological and economic performance is critical for progress, especially in industries involved with aerospace, biomedical devices, and energy generation and storage. Recently, a new class of highly alloyed materials, high-entropy alloys (HEAs), has been proposed. Efforts to design, characterise and exploit HEAs are currently underway in many countries. However, the current theory of high-entropy alloy formation, described in the literature, is lacking and often misleading. Many attempts to predict their formation have been unsuccessful.
We have developed a method for rapidly predicting the formation and stability of high-entropy alloys (HEAs) across all compositions using 73 elements in the periodic table. Our model, implemented in a new piece of software, holds true for the 177 experimentally characterised HEA systems. Significantly, from the ~186,000,000 systems screened, the code has predicted a total of ~3,000 new HEA systems. The code is capable of searching through every stoichiometry of these systems to optimise the material properties, further. A prediction as to the stability of each alloy at a specific temperature can also be made, allowing precipitation temperatures to be forecast. This is the first time such a comprehensive and accessible code has been released allowing for huge advances in the design, development and discovery of new technologically important alloys.

Pure titanium is observed to undergo a transition from a hexagonally close pack structure (alpha phase) to a body center cubic structure (beta phase) at 1155 K under normal pressures. Alloying elements can raise or lower this transition temperature. Alpha-beta titanium alloys are the most extensively used due to their combination of valuable properties from both phases. Accurate atomistic modelling of this important class of alloys relies on an accurate representation of both phases. However, the most robust interatomic potentials currently available are unable to predict this transition at any temperature with the beta phase being only metastable.
We present here a new titanium potential based on the modified embedded atom method which is able to correctly demonstrate this phase transition in titanium. By varying the ground state energy of the BCC phase, the transition temperature can be tuned to match experimental data without sacrificing other ground state predictions such as elastic constants and defect energies. This new titanium potential will allow for the development of multi-element systems of interatomic potentials which can accurately model the most commonly used alloys.

The discovery of advanced materials is the key bottleneck limiting the commercialization of thermoelectric technology for waste heat recovery and refrigeration. Computationally-driven approaches can accelerate the discovery of new thermoelectric materials and provide insights into useful structure-property relations that govern thermoelectric performance. We have developed TEDesignLab (www.tedesignlab.org), an open-access thermoelectrics-focused virtual laboratory that contains calculated thermoelectric properties as well as potential thermoelectric performance rankings based on a metric (β)¹ . The computationally-tractable metric (β) combines first-principles calculations and modeled electron and phonon transport to offer a reliable assessment of the intrinsic material properties that govern the thermoelectric figure of merit (zT). Another useful component of TEDesignLab is the interactive web-based tools that enable users to mine the raw data and find new structure-property relations. The usefulness of the calculated properties available on TEDesignLab goes beyond thermoelectrics. For example, charge carrier mobilities are important in the design of photovoltaics and lattice thermal conductivities in the design of thermal barrier coatings. With TEDesignLab, we also aim to establish a close partnership between experiments and computations by providing users with resources to analyze raw experimental thermoelectric data and to provide open access through our experimental databank.
[1] J. Yan, P. Gorai, B. Ortiz, S. Miller, S. Barnett, T. Mason, V. Stevanovic, and E. S. Toberer, "Material Descriptors For Predicting Thermoelectric Performance," Energy Environ. Sci. 8 (2015) 983-994.

Technologies based on piezoelectricity, ferroelectricity and non-linear optical activity all rely on noncentrosymmetric (NCS) materials. However, we presently lack a complete recipe for the design of new NCS materials. Using first-principles density functional theory (DFT) calculations we investigate ferroelectricity in strained double-perovskite fluorides, Na₃ScF₆ and K₂NaScF₆. The experimental room temperature crystal structures of the fluoroscandates are centrosymmetric, i.e. Na₃ScF₆ (P2₁/n) and K₂NaScF₆ (Fm3m). From lattice dynamical calculations, we find that double fluoroperovskites bear some similarities to ABO₃ perovskites. Specifically, our results demonstrate that in their prototypical cubic geometry, infrared active modes are inherent, and harden as the tolerance factor approaches unity. In contrast to ABO₃ perovskites, acentric displacements of alkali metal cations (Na, K) are dominant over displacements of the transition metal cation in these double
fluoroperovskite. Biaxial strain investigations of centrosymmetric vs. polar geometries of Na₃ScF₆ and K₂NaScF₆ reveal that paraelectric phases are favored in the compressive strain region while at sufficiently large tensile strains in-plane polarizations (~5-18 mu;C/cm-2) can be stabilized. Furthermore, we observe that the stable polar structures in both chemistries exist in the presence of a single rotational distortion about the c axis that enhances the polarization. Lastly, we comment on progress towards designing new NCS materials through oxygen-fluorine substitution in double perovskite fluorides.
This material is based upon work supported by the National Science Foundation under grant no. DMR-1454688. DFT calculations were performed on the high-performance computing facilities available at the Center for Nanoscale Materials (CARBON Cluster) at Argonne National Laboratory, supported by the U.S. DOE, Office of Basic Energy Sciences (BES), DE-AC02-06CH11357.

Design of highly efficient, stable phosphorescent dopants with high triplet energies for blue emission is one of the central challenges in expanding the application of organic light-emitting diode (OLED) devices. Owing to the large design space for organic dopants, rapid evaluation and virtual screening based on structure-property relationships has been of particular interest. While first-principles quantum chemical methods are known to predict the key properties of dopants - e.g. triplet excitation (T₁) energy - accurately, the computational cost is too high for screening a large number of structures. Semi-empirical quantum mechanics (SQM) approach can be an alternative with a significant speed-up in property predictions, yet its poor accuracy often limits its applicability. In this work, we propose a new SQM-based virtual screening framework for rapid screening of phosphorescent dopants in OLED devices. The method utilizes heuristic informatics algorithms to provide predictive capability rivaling first-principles methods. We will illustrate the hybrid workflow for a focused library of Ir-based OLED dopant materials. The aim of this approach is to allow researchers to screen 10,000 or more dopants per day only using a typical modern-day workstation. Such a dramatic improvement in throughput and accuracy for virtual screening framework would have significant impact on discovery and development of novel OLED materials.

Making physically meaningful connections between atomistic-scale processes and continuum scale theories of plastic deformation remains a grand challenge in computational materials science. This connection is particularly daunting in the context of amorphous materials which lack the ordered structure against which defects are traditionally defined. Here we present our initial results from an effort to synthesize data from atomistic simulations with a continuum-scale theory of plastic localization in the context of metallic glass. Prior work has indicated that beyond characterizing the mean structure, properly characterizing structural fluctuations is critical for understanding the instabilities that lead to shear band nucleation. We extract data regarding these structural fluctuations from molecular dynamics models of metallic glass. The results of subjecting these atomistic simulations to large strains are directly compared to the results of the continuum theory starting from the atomistically generated initial conditions. Important variations in the failure mode arise both with respect to the preparation of the glass, i.e. the quench rate, and the dimensionality of the model. These results are discussed in terms of the statistics of the structural fluctuations in the glass and the mathematics of the underlying constitutive equations.

Mesoscopic numerical simulations are used to examine the dynamics of thin films of phase-separating polymers that contain nanoparticles. When phase separation occurs, the nanoparticles segregate to the interfaces between polymer phases, which locks the configuration of the phases into place. To control the spatial morphology of the thin film, an electric field is applied to align the phases and the particles perpendicular to the film. The computational approach is a novel Brownian Dynamics-Cahn Hilliard (BD-CH) hybrid model that combines particle- and grid-based simulation techniques, employed with highly parallel computing for large-scale 3D simulations. Once the desired morphologies are obtained, one of the polymer phases can be dissolved away, leaving a membrane-like structure containing pore-channels that are lined with the nanoparticles. The particles can be made on the scale of nanometers and they can be made catalytic, which gives rise to a variety of technological capabilities, including membranes for chemical separations, catalysis, and energy-related devices such as photovoltaics and light-emitting diodes. This talk will focus on the computational results that illustrate the relationships between the average channel diameter, the channel areal density, and the interfacial nanoparticle arrangements.

Physical vapor deposition (PVD) is one of many techniques used to synthesize thin films. The PVD conditions and materials employed greatly influence the topology and microstructure of the deposited film. Experimental observations are useful in determining guidelines and diagrams that relate PVD conditions to thin film microstructures but they are costly, limited and difficult for multiphase systems. As such, a predictive model for microstructure formation and evolution during PVD is much needed. The purpose of this work is to develop a phase-field model to simulate thin film growth during PVD including grain evolution within the microstructure for an isotropic single-phase polycrystalline metal. Phase-field modeling is a popular technique for predicting the evolution of microstructures in materials through interfacial kinetics. To incorporate critical aspects of the PVD process and microstructure evolution, the proposed model leverages previous modeling efforts on ballistic deposition of single-phase materials and grain orientation evolution in polycrystalline materials. In this model, the equations of motion governing PVD couple the evolution of the growing solid film to an incident vapor flux such that the film grows at the expense of the available vapor. To allow for the growth and evolution of individual grains within the growing solid film, a previously developed model that captures grain boundary motion and grain rotation is coupled to the PVD equations of motion. To illustrate the capabilities of the new model, this model is used to simulate PVD growth of a single-phase polycrystalline metal capturing solid growth and polycrystalline evolution within the thin film. Therefore, this model will allow predictive studies and discovery of film microstructures as a function of PVD conditions and initial substrate geometries for isotropic single-phase polycrystalline metals.

A 3D phase field model has been developed for investigating the grain growth process in porous solids. The model couples the grain boundary motion and pore surface motion, which enables us to capture the coevolution of grain size and pore morphology in porous polycrystalline solids. As such, the model takes into account all possible interactions between the pore and the grain boundary such as pore drag and pore breakaway, which affect strongly the overall rate of grain growth in such materials. We carried out a formal asymptotic analysis of the phase field model to match it to its sharp-interface counterpart. This analysis fixes all the model parameters in terms of the regular thermodynamic and kinetic data. We used the mesoscale simulator MARMOT developed at Idaho National Laboratory to perform 3D simulations of the model. The simulations show a profound effect of pore fraction, grain and pore size distributions, and pore configurations on the kinetics of grain growth process.

Molecular dynamics has been used to study the effect of substrate crystal structure and orientation on the homoepitaxial growth of CdS. The growth simulations were performed on CdS substrates oriented along the , [0001], [-1100], [11-20] wurtzite (WZ) and [111], [112], [110], [010], and [010] tilted °8 (representing a miscut) zinc-blende (ZB) directions. Algorithms that identify structure and dislocation were then employed to analyze the epilayers. Epilayers grown on the [-1100] WZ, [11-20] WZ, [111] ZB, [-110] ZB, [010] ZB, and [010] -°8 ZB substrates had a high degree of crystal perfection and contained only point defects. All of these substrates contained surface atoms arranged with a rectangular geometry. The sole exception was the epilayer grown on the [112] ZB substrate which contained a small concentration of stacking faults in addition to point defects. In contrast, epilayers grown on substrates with hexagonal surface geometry ( WZ and ZB) contained in a high degree of crystalline disorder consisting of polytypism (presence of WZ and ZB phases), stacking faults, and twins in addition to point defects. These results indicate that the growth orientation, and more specifically surface atomic geometry, is a critical factor to achieve highly crystalline epilayers. The polytypism observed in the epilayers grown on [0001] WZ and [111] ZB substrates is attributed to the small energy difference between WZ and ZB structures. Stacking faults and twins are mainly created by Shockley partial and WZ full dislocations. The simulation results provide an insight into CdS structures crystallinity according to the direction growth.

Molecular dynamics simulations employing a Stillinger-Webber potential have been applied to study the microstructure of CdTe thin films grown on single crystalline CdS substrates. Different substrate structure and orientations were explored including [0 0 0 1], [-1 1 0 0], [1 1 -2 0], and zinc-blende [1 1 1], [1 1 -2], [-1 1 0], [0 1 0], and [0 1 0] with a miscut of 8#8304;. Substrate surface microstructure was characterized by dangling bond density, polarity, and elemental composition. Algorithms that identify lattice structure and detect defects were employed to analyze the microstructure of the simulated epilayers in terms of grain size and dislocation line density. The analyses revealed a high dislocation density and small grain sizes at the interface, but fewer dislocations and larger grains at locations within the films away from the interface. These results are in good agreement with experimental data reported in literature. The simulated film structures also exhibit polytypism as they are composed of both zinc blende and wurtzite phases, leading to extensive stacking faults, partial and full dislocations. The dislocation line analyses revealed that some substrate orientations promote epilayer dislocations that thread across the film thickness while others localize dislocations to the interface. The results obtained provide insight into the optimal substrate surface needed to obtain low defect CdTe thin films and could help guide experimental efforts towards the fabrication of highly crystalline heteroepitaxial systems.
Acknowledgement - Sandia National Laboratories is a multiprogram laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the US Department of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000. This work was performed under DOE Project No. EE0005958.

Structural polymorphism among ceramic materials is a common phenomenon that is widely seen as a signature of kinetics-dominated behavior in even the most thermodynamically-driven synthesis methods. As such, metastable polymorphic phases are often ignored, or considered only as experimental corrections in bulk-scale first-principles phase diagrams, resulting in a significant deviation between materials theory and reality.
A common example of ceramic polymorphism are the pyrite and marcasite phases of FeS₂. While these minerals are among the most common sulfides in the Earth's crust and have been studied extensively for applications in photovoltaics and energy storage, the mechanism underlying their formation is not well understood. In our work, we develop a novel method for modeling aqueous ceramic interfaces, combining atomistic- and continuum-level theory to accurately reproduce the surface chemistry of FeS₂. We demonstrate that the hydrothermal formation of FeS₂ pyrite and marcasite can be explained by the pH dependence of the surface energies of the two phases, which gives rise to a phase transition at sizes relevant to nucleation from aqueous solution. This result suggests that bulk structural polymorphism can arise from purely thermodynamic effects that manifest themselves only at small size scales. Consequently, we propose that a first-principles understanding of finite-size phase stability in realistic synthesis environments can serve as an efficient metric of the synthetic accessibility of metastable polymorphs and as a guide to the experimental synthesis thereof.

The one-dimensional lattice defects known as dislocations are widely believed to provide short-circuit paths for diffusion in crystalline materials. In this contribution I will present recent work that we have been doing on characterizing and understanding mass transport processes along dislocations in the perovskite-type oxide SrTiO₃. I will focus on computational studies of point-defect processes at the periodic array of dislocations that constitutes a low-angle tilt grain boundary. Static lattice calculations reveal a strong driving energy for space-charge formation at dislocations. Molecular dynamics simulations indicate that oxygen diffusion along the dislocation is slower than in the bulk. I will also present complementary results obtained from experimental studies of oxygen diffusion along the dislocation array. Combining all results and literature reports, I will present a comprehensive and consistent picture of the transport properties of dislocations in SrTiO₃. Finally I will describe the consequences for memresistive devices.

Achieving shape and size controlled synthesis of nanoparticles can enable us to exploit their unique catalytic, optical, and magnetic properties. In the past decade, a number of promising solution-phase synthesis techniques have been developed to fabricate various nanostructures. A deep, fundamental understanding of the structure-directing mechanism in these syntheses would contribute towards improving the selectivity of next-generation techniques. In our study, molecular dynamics simulation is used to probe the structure-directing mechanism of colloidal Ag nanoparticles synthesis. We study the structure-directing agent polyvinylpyrrolidone (PVP), which is well-known for producing {100}-faceted Ag nanostructures such as nanowires. The proposed structure-directing mechanism of PVP is they preferentially bind to Ag(100) over Ag(111) surfaces. To confirm this hypothesis, we calculate the potential of mean force (PMF) along the distance orthogonal to the Ag surface plane to quantify the solution-phase binding energies of PVP-derived molecules on Ag(100) and Ag(111) surfaces. We also study the effect of PVP on the relative atom deposition rate on Ag(111) over Ag(100) surface. By calculating the PMF profiles, we show how PVP influence the formation of {100}-faceted Ag nanostructures.

Graphene-oxide (GO) is one of the most studied functionalized graphene structures. Its promising applications are numerous despite the issues related to the synthesis of GOs with well controlled oxygen percentage coverage and epoxide-to-hydroxyl ratios.¹ While GO nanostructures have been considered in the development of flexible electronics,² battery electrodes,³ and hydrogen storage,⁴ macroscopic GO-nanostructure-based materials, such as GO paper⁵ and GO foams,⁶ have unique material properties, such as large stiffness and negative thermal expansion, that make them particularly intriguing. The ideal tool to investigate physical and chemical properties of large GOs without losing their atomistic structural features is classical molecular dynamics (MD) simulations. In particular, the development of macroscopic materials made of nanoscopic GOs requires, at some point, atomistic methods to unravel the influence of the nanostructures on the macrostructure. Here, a first step towards the investigation of this influence is given by examining the accuracy in the predictions of structure, energy and mechanical properties of GO nanostructures in classical MD simulations. Three well known classical potentials are used and their predictions compared. Two potentials are reactive, i.e., they allow for dynamical breaking and formation of bonds, and the third is a non-reactive potential, i.e., the chemical bonds among the structure are pre-defined and remain unchanged during the simulation. The particular potentials considered are the Reactive Empirical Bond-Order Potential parameterized for systems with carbon, hydrogen and oxygen atoms (REBO-CHO);⁷ the third generation of the Charge Optimized Bond Order potential (COMB3);⁸ and the Chemistry at HARvard Macromolecular Mechanics potential (CHARMM);⁹ respectively. The physical and chemical properties investigated include the binding energy, the carbon-oxygen bond distance and the Young&’s modulus of the GO nanostructures. The results indicate that compared to DFT calculations or experimental results, the potential that yields the best quantitative agreement is COMB3. However, REBO-CHO and CHARMM are shown to be quite good and appropriate for modeling most of the GO-systems considered. We discuss the relevance of these results for modeling GO macrostructures.
[1] A. F. Fonseca, H. Zhang and K. Cho, Carbon84, 365 (2015).
[2] G. Eda, G. Fanchini and M. Chhowalla, Nature Nanotechnology3, 270 (2008).
[3] G. Eda et al., Appl. Phys. Lett.92, 233305 (2008).
[4] L. Wang et al., ACS Nano3, 2995 (2009).
[5] D. A. Dikin et al., Nature448, 457 (2007).
[6] S. Vinod et al., Nature Communications5, Article number: 4541 (2014).
[7] A. F. Fonseca et al., Phys. Rev. B84, 075460 (2011).
[8] T. Liang et al., Materials Science and Engineering R74, 255 (2013).
[9] A. D. MacKerell, et al., J. Phys. Chem. B102, 3586 (1998).

Contact resonance atomic force microscopy (CR-AFM)^1,2 is a powerful tool for mapping subsurface differences of layered (2D) materials based on their distinct mechanical properties. Few layer graphene (FLG; mono-, bi-, or trilayer thickness) on the silicon face of cubic silicon carbide (3C-SiC-111) exhibits very clean and distinct surfaces and yields high-contrast CR-AFM images showing clear subsurface structure fingerprints. However, the experimental deconvolution of stiffness contributions arising from the different sub surface layers is difficult. To interpret the contributions from surface domains with different layer thicknesses, we use density functional theory (DFT) with an explicit inclusion of van der Waals interactions to calculate atomic displacements for surface stresses acting on FLG on SiC based on radic;3×radic;3R30° and 6radic;3×6radic;3R30° interface models. The calculated moduli from these stress-strain relations then serve as input parameters for a continuum model, which describes the AFM probe-sample contact and simulates contact stiffnesses for the atomistic models from the mechanical impedance of each layer and the assumption of a Hertzian tip-substrate contact.^3-5 The experimentally measured contact stiffnesses can be used to estimate Youngs' moduli and allow for a direct comparison with the theoretical models. Further, we contrast impurity free FLG/SiC with oxygen intercalated samples and derive the mechanical structure fingerprint based on a SiC + silicate layer structure model. The experimental data as well as the simulations revealed distinct, mechanically harder oxygen-intercalated areas.^6
1 U. Rabe, S. Amelio, E. Kester, V. Scherer, S. Hirsekorn, and W. Arnold, Ultrasonics 38, 430 (2000).
² U. Rabe, M. Kopycinska, S. Hirsekorn, and W. Arnold, Ultrasonics 40, 49 (2002).
³ U. Rabe, K. Janser, and W. Arnold, Review of Scientific Instruments 67, 3281 (1996).
⁴ H. Hertz, Journal für die reine und angewandte Mathematik 92, 156 (1881).
⁵ G. G. Yaralioglu, F. L. Degertekin, K. B. Crozier, and C. F. Quate, J. Appl. Phys., 87, 7491 (2000).
⁶ Q. Tu, B. Lange, Z. Parlak, J. M. Lopes, S. Zauscher, V. Blum, Nanomechanical Subsurface Structure Fingerprint for Graphene Based Nanostructures: Contact Resonance AFM and Theory, in preparation

Amorphous silicon oxycarbide (SiOC) is of great technological interest, and its properties are closely related to its C distribution. Combining density functional theory and classical potential atomic scale calculations, we show that the clustering tendency of C atoms in SiOC can be effectively impacted by introduction of hydrogen (H): without H, the C-C interaction is attractive, leading to enrichment of aggregated SiC₄ tetrahedral units; with hydrogen, the C-C interaction is repulsive, leading to enrichment of randomly distributed SiCO₃ tetrahedral units. Our results suggest that by differing amounts of H present in the SiOC samples, we may tailor the behaviors of SiOC by controlling the C distribution.
This work was funded by the DOE Office of Nuclear Energy, Nuclear Energy Enabling Technologies, Reactor Materials program, under contract No. DE-NE0000533. Computational support was provided by DOE-NERSC and DOE-OLCF.

Metallic nanoparticles offer an attractive alternative to their bulk counterparts for many applications such as catalysis, hydrogen storage, gas sensing, etc. Their performance depends strongly on surface structure; therefore, nanoparticle coalescence can play an important role, as it determines the resultant structure of the active sites where reactions actually take place, i.e. facets, edges, vertices or protrusions.
With this in mind, we performed both classical molecular dynamics (MD) simulations and magnetron-sputtering inert gas condensation depositions of various single or hybrid metallic nanoparticles, supported by high-resolution transmission electron microscopy (HRTEM), to study the mechanisms that govern their coalescence.
Our MD simulations captured the propagation of a crystallisation wave through amorphous Pd nanoparticles, even at room temperature. In contrast, under the same conditions, Ta nanoparticles remained amorphous. Both results are in complete agreement with our experimental findings and may be explained crystallographically.
An almost-epitaxial alignment, assisted by the formation of twins and surface protrusions, was observed in crystalline nanoparticles of both species. We elucidate the protrusion creation through a shearing mechanism and by tracking the glide of temporarily emergent interface dislocations.
Based on our atomistic simulation and experimental data, we propose a rigorous yet simple analytical method which describes element-independent coalescence behaviour, emphasising only on the predominant dependency of neck formation on temperature and size-dependent nanoparticle melting points. Its simplicity makes our model a suitable starting point for the development of a meso-scale simulation technique that can describe the growth of porous films and allow for their bespoke design.
Finally, we utilise this model to predict the resultant structure of Ag-Cu mixed particle coalescence. This system has been frequently studied, due to its combination of low price, biocompatibility, and plasmonic and catalytic properties. Practically all previous studies used Ag-Cu nanoalloys as starting points; however, we demonstrate that controlled simultaneous co-deposition of pure Ag and Cu nanoparticles allows designing Ag-Cu nanoparticles of desired configurations via coalescence mechanisms.

Classical molecular dynamics with inter-atomic potentials can be quite successful at predicting the vibrations of materials, and is especially useful for handling large structural models (e.g. tens of thousands of atoms) that are intractable by quantum methods. However, to predict Raman spectra, electrons must be re-introduced to the problem via a model for the polarizability and its variation with atomic positions. The established approach for bulk semiconductors is to construct a bond-polarizability model, attributing the polarizability to cylindrically symmetrical inter-atomic bonds. Using an assumed form for the polarizability as a function of bond length, the parameters are fit to experimental data. Then, a Raman intensity can be computed for the vibrational eigenvectors from classical potentials. However, in the case of amorphous silicon, use of the existing models with vibrational eigenvectors from DFT has shown significant discrepancies from experiment [R. M. Ribeiro et al., phys. stat. sol. (c) 7, 1432 (2010)]. By contrast we find that the Raman spectrum calculated by density-functional perturbation theory for a set of 64-atom a-Si:H structures agrees well with the measured spectrum. These results suggest that the models have compensated errors in the classical vibrational frequencies with errors in the Raman tensors. To generate a more accurate and transferable bond polarizability model, we fit a general bond-polarizability model to the ab initio atomic Raman tensors in our data set to obtain parameters and functional forms that can describe the Raman intensities of vibrations in large structural models. This new Raman model can be used to study problems such as medium- and long-range order in a-Si:H, nanocrystalline Si, amorphous/crystalline interfaces, or a-Si:H nanowires at sizes that would be inaccessible for ab initio vibrational calculations, and offers the potential to apply the same methodology to other material systems.

Time-dependent density functional theory (TDDFT) is one of the most prominent and widely used methods for calculating excited states of medium-to-large molecules, and it is recognized as a powerful tool for studying electronic transitions of molecules [1]. In our calculations, the real-time and real-space (RSRT) technique is adopted in solving equation by the finite difference approach [2]. In particular, the real-space approach is suitable for large-scale parallel computing, and also allows the capture of a clear physical image. Within the framework of this approach, we can solve for the wavefunctions on the grid with a fixed domain, which encompasses the physical system of interest [3].
We have applied RSRT-TDDFT to various conjugated polymers to study their optical properties[3] as their importance are well known in the real applications, e.g., polymer light emitting diodes. The studies have so far been performed on the polymers in gas phase whereas organic light emitting diodes are in the solid-state where each of the polymers can be considered to be in a solid solution of the same or different materials. In this study, we perform the combination of “Extra Large Scale Electronic Structure Calculation (ELSES)” [4] and “RSRT-TDDFT” on K-computer to study the optical properties of conjugated polymers, especially focusing on the differences between in the solid-state and in the gas phase.
[1] E. Runge, and E. K. U. Gross, Phys. Rev. Lett. 52, 997 (1984)
[2] K. Yabana and G. F. Bertisch, Phys. Rev. B 54, 4484 (1996)
[3] Y. Zempo, N. Akino, M. Ishida, M. Ishitobi, and Y. Kurita, J. Phys. Cond. Matt. 20, 064231 (2008)
[4] http://www.elses.jp/

I will describe a 3-Dimensional particle based mesoscopic transport model that can describe all relevant multi-species (electrons, holes, singlet and triplet excitons) transport processes and interactions in the complex architecture comprising organic optoelectronic devices. This model uses the Kinetic Monte Carlo, KMC, approach to predict charge and exciton motion through a random walk simulation that keeps track of where the particles are at a given time, so allowing for interactions between the particles [1,2]. Hops of charges and excitons between localised sites are formulated so that they relate to the microscopic kinetics of the system. KMC methods devised for simulations of surface reactions were adapted at Bath to study the link between morphology and device performance in organic photovoltaics [3].
We have applied this model to simulate the electrical and optical characteristics of a stacked organic light emitting diode, OLEDs based on a copolymer blend and validated by experimental measurements. OLEDs often include an interlayer, a layer of polymer semiconductor between the hole injection layer and the light emitting polymer. We look at the influence of the interlayer on a blue emitting poly fluorene copolymer blend (LEP) of poly-(9,9'-dioctylfluorene-co-bis-N) (F8) and N'-(4-butylphenyl)-bis-N,N'-phenyl-1,4-phenylenediamine) (PFB) (95:5% mol) with a 15 nm copolymer interlayer of poly (9,9-dioctyl fluorene-N-(4-(2-butyl)phenyl)-diphenylamine) (TFB) and F8 (50:50% mol). Our model replicates the unexpected two fold increase in current in a hole only device measured on addition of the interlayer. The interlayer shifts the recombination zone away from the anode, causing a region of high exciton concentration that increases the singlet generation yield via triplet-triplet reactions that causes a 17 fold increase in the luminous efficiency. By increasing the interlayer length from 15 nm to 50 nm a linear increase in exciton formation efficiency occurs, however, due to the decrease in the outcoupling of light from the device, no appreciable change in the simulated device external quantum efficiency was observed.
1. J Bailey, E N Wright, X Wang, A B Walker, D D C Bradley, J-S Kim J Appl Phys (2014) 115 204508
2. R G E Kimber, E N Wright, S E J O'Kane, A B Walker, J C Blakesley Phys. Rev. B (2012) 86, 235206
3. P K Watkins, A B Walker, G L B Verschoor Nano Letters (2005) 5 1814

Characterizing and understanding the fundamental chemistry of natural carbonaceous materials such as kerogen is of great interest to various research fields including energy production, chemical engineering, and geochemistry. However, the vast chemical diversity of such materials across multiple length scales forbids direct simulation and characterization of their optical properties from atomistic models. Using an ensemble proxy model, we tackle this problem with a multi-scale simulation approach. The UV/Visible optical absorption property of a carbonaceous material with specific compositional chemistry can now be calculated using density functional theory from an ensemble of molecular structures, which are generated from larger scale MD simulations. The correlation between the chemical composition of kerogen and its optical absorption properties obtained from our ensemble proxy model agrees quantitatively with experimental measurements.¹
1. N. Ferralis*, Y. Liu*, K. D. Bake, A. E. Pomerantz and J. C. Grossman, Carbon 88 (0), 139-147 (2015).

Nickelate interfaces have been of recent interest due to a number of their interesting electronic properties including their orbital ordering which can be made to resemble that of cuprate superconductors. [1] Separately, nickelate thin films such as LaNiO3 exhibit a metal-insulator transition as a function of film thickness [2,3,4] which has been correlated to important surface distortions [2,4] but a first principles elucidation of this type of transition has remained elusive as typical band structure calculations, when chosen to reproduce the properties of bulk LaNiO3, invariably predict metallic systems. The role of electronic correlations in these systems remains to be fully understood. One promising approach for modeling the electronic phases of nickelates is the slave-rotor method [5,6] which correctly predicted two out of three phases in certain heterostructures [7] . We have developed a new, generalized slave particle method [8] that, in the appropriate limits, can reproduce and correct the behavior of the slave-rotor method in the low-correlation limit as well as another slave-particle method, the slave-spin method [9] that correctly predicted the quasiparticle weights in Fe-based superconductors [10]. Our method allows us to create different models in such a way as to isolate the different degrees of freedom in an extended Hubbard model in a very numerically effective way on real materials. We will show how these various models can be used to explore the effects of correlations using nickelate heterostructures or thin films.
1. A. Disa. D. Kumah, A. Malashevich, H. Chen, D. Arena, E. Specht, S. Ismail-Beigi, F. Walker, and C. Ahn,Orbital engineering in symmetry-breaking polar heterostructures. Physical Review Letters 114, 026801, 2015.DOI: 10.1103/PhysRevLetters114.026801
2. D. Kumah, A. Malashevich, A. Disa, D. Arena, F. Walker, S. Ismail-Beigi, and C. Ahn, Effect of Surface Termination on the Electronic Properties of LaNiO3 Films. Phys. Rev. Applied 2, 054004, 2014. DOI: 10.1103/PhysRevApplied.2.054004
3. King et al, Atomic-scale control of competing electronic phases in ultrathin LaNiO3, Nature Nanotechnology 9, 443-447 (2014)
4. D. P. Kumah, A. S. Disa, J. H. Ngai, H. Chen, A. Malashevich, J. W. Reiner, S. Ismail-Beigi, F.-J. Walker, and C. H. Ahn, Tuning the structure of nickelates to achieve two-dimensional electron conduction, Adv. Mater. 26, 1935 (2014)
5. S. Florens and A. Georges, Phys. Rev. B 66, 165111 (2002)
6. S. Florens and A. Georges, Phys. Rev. B 70, 035114 (2004)
7. B. Lau and A. J. Millis, Phys. Rev. Lett. 110, 126404 (2013)
8. Alexandru B. Georgescu, Sohrab Ismail-Beigi, A Generalized Slave-Particle Method, arXiv:1506.03515
9. L. De&’Medici, A. Georges, and S. Biermann, Phys. Rev. B 72, 205124 (2005)
10. L. De Medici, G. Giovannetti, and M. Capone, Phys. Rev. Lett. 112, 177001 (2014), arXiv:1212.3966

Interatomic potentials have been widely used to accelerate atomistic simulations such as molecular dynamics for a long time. Recently, frameworks to build an accurate interatomic potential were proposed, combining a systematic set of density functional theory (DFT) calculations with machine learning techniques [1,2]. Using these techniques, the accuracy of the potential energy surface (PES) is generally much better than that obtained using conventional interatomic potentials. Another advantage is their applicability to a wide range of materials, including metallic, covalent, and ionic materials. One of these frameworks is to use the compressed sensing deriving a sparse representation for the interatomic potential [3].
We have applied a method of constructing a linearized PES by elastic net regression to a wide range of elemental metals [4]. Compared with the other approach based on systematic first-principles calculations, the elastic net interatomic potential has the following main advantages. 1) A well-optimized sparse representation for the PES can be obtained, which increases the accuracy of atomistic simulations while decreasing the computational cost. 2) The accuracy can be easily controlled, i.e., the trade-off between the accuracy and computational cost is determined by a small number of parameters. 3) Information on the forces acting on atoms and stress tensors can be included in the training data in a straightforward manner. This ensures the reliability of the force and stress tensor calculation using constructed interatomic potentials.
As a result of applying the present method, we found that the energetics can be expressed by a linear relationship with simple basis functions depending only on distances between atoms. A sparse set of suitable basis functions for expressing the PES can also be easily extracted from 4,836 basis functions by elastic net regression. As a result, we have obtained a sparse PES with prediction errors ranging from 0.3 to 3.5 meV/atom. The prediction errors for the force and the stress tensor were within 0.03 eV/Å and 0.15 GPa, respectively. Also, we compared equilibrium lattice constants and phonon dispersion relationships obtained by the elastic net PES and by DFT calculation. The former were in good agreement with the latter for all ten elemental metals considering in this study.
[1] J. Behler and M. Parrinello, Phys. Rev. Lett. 98, 146401 (2007).
[2] A. P. Bartoacute;k, M. C. Payne, R. Kondor and G. Csányi, Phys. Rev. Lett. 104, 136403 (2010).
[3] A. Seko, A. Takahashi and Isao Tanaka, Phys. Rev. B 90, 024101 (2014).
[4] A. Seko, A. Takahashi and Isao Tanaka, arXiv:1505.03994.

Cell size is one of the most fundamental structural parameters for foamed polymeric materials, and its reduction is usually associated with an improvement of properties, e.g. high impact resistance, low thermal conductivity and high stiffness-to-weight ratio. The further reduction of cell size to the nanometer scale presents opportunities for even greater improvements and opens a new frontier in this type of material.
Among the different techniques available for the production of nanofoams, solid state foaming, mostly employed for microcellular materials production, is one of the most promising. This technique has been shown to produce good results for different polymers, with relatively high nucleation densities and low cell sizes. Furthermore, it presents a potentially high scalability and low ecological impact. However, there are still many aspects that require a better understanding of the fundamental science involved. Although many empirical optimization studies have been carried out, the manipulation of cell size on the nanometeric scale requires the application of a novel theoretical framework compared to that applied previously for conventional microcellular materials: current methods are reaching their limits in their applicability to structures involving ultra-small dimensions.
The work presented here shows the application of modelling strategies to the production of nanofoams with solid state foaming technique. Molecular dynamics simulations were employed to simulate some of the fundamental stages of foam production: gas dissolution, polymer stability during foaming, and cell stability. Both pure and CO₂-infused polymers mixtures, stable and unstable, were studied. Stable mixtures presented a nanocellular structure after CO₂ removal, but not all of them retained it after further geometry optimization. The results obtained allow the creation of a set of rules for the selection of polymers with improved suitability for the production of stable nanofoams via a solid state foaming process.

Developing rigorous and predictive descriptions of symmetry-breaking phenomena and anisotropic properties at the continuum scale remains a challenging problem. Many promising structural and electronic materials have high-symmetry phases at high temperatures that become mechanically unstable at low-temperature with respect to a reduced-symmetry phase. DFT predicts the zero-K energy surface to be non-convex at these high-symmetry phases, making it impossible to predict their free energies, mechanical properties, or other constitutive relationships, using common statistical mechanical approaches that rely on the harmonic approximation. Even in mechanically stable crystal phases, crystal symmetry and anisotropy can strongly influence mesoscale behavior due to strongly anharmonic or anisotropic elasticity relationships and chemo-mechanical coupling.
We have recently developed approaches that combine group theory and alloy theory to both automate and simplify the rigorous calculation of finite-temperature mesoscale materials properties from atomistic first-principles calculations. The results of these calculations can subsequently be used to inform materials design or materials simulations at larger length scales. We present these techniques for calculating symmetry-breaking free energies, anisotropic finite-temperature elastic properties, and chemo-mechanical response functions, along with examples of their application to hydride phase precipitation and chemo-mechanical coupling in Heusler alloy phases.

While solving macroscopic or coarse-scale set of evolution equations may be sufficient in many applications, in a large number of complex systems accurate models are only given at a more detailed (microscopic or fine-scale) level of description. The major drawback of modeling such fully-resolved and expensive systems is the long-term analysis of the whole course of evolution of such systems. There are numerous biological and materials systems with the above properties. For example, the process of clot formation and growth at a site on a blood vessel wall involve a number of multi-scale simultaneous processes including: multiple chemical reactions, species transport and flow. To model these processes we have incorporated advection-diffusion-reaction (ADR) of multiple species into an extended version of Dissipative Particle Dynamics (DPD) method which is considered as a coarse-grained Molecular Dynamics method. A projective integration technic is used in order to facilitate acceleration in time. In the first approach, a continuum set of ADR equations are solved on a coarser grid at the end of each DPD simulation, and the results are used to initialize another burst of fine-scale DPD simulation in the cycle. In the second approach, projective integration is performed through a reduced-order model based on proper orthogonal decomposition (POD)-assisted computational methodology in which the ADR equations are no longer solved; velocity and concentration fields are parametrized by a few POD basis functions. The resulting POD modes are computed on the fly by sampling snapshots of the fine-scale DPD simulation, and then these modes are integrated in time.

We present our multiscale modeling of colloidal nanoparticles self-assembled into many different lattices and superstructures, as observed in recent experiments. First, we discuss how superparamagnetic magnetite nanocubes self-assemble into chiral and other unique superstructures in the presence of magnetic fields [1]. To this goal, we have developed mean-field Monte Carlo codes where forces acting between nanoparticles are parametrized by known laws, molecular dynamics simulations, and estimates of coupling strengths present under dynamical conditions during the self-assembly. We also use atomistic molecular dynamics simulations to describe the self-assembly of chiral CdS (truncated tetrahedra) nanoparticles in the presence of circularly polarized light and the formation of hollow nanoparticle-based capsules and other superstructures in dependence on the used pH [2]. Finally, we model the solvation and self-assembly of different nanoparticles in bulk solvents and at electrified interfaces of ionic solutions [3]. We provide a detail analysis of the parameters that control the self-assembly processes of these nanomaterial systems.
[1] G. Singh et al., Science 345, 1149 (2014).
[2] J. Yeom et al., Nat. Mat. 14, 66 (2015) & submitted.
[3] M. K. Bera et al., Nano Lett. 14, 6816 (2014) & submitted.

In order to design materials with desirable characteristics it is important to have a good understanding of the atomic-scale chemical and physical properties of materials. Thus, with tremendous progress in computer technologies during the last decade atomistic-level simulation is rapidly becoming an essential tool in materials science. Here, we have demonstrated the simulation approach that allows us to estimate the thermodynamic stability and gas sorption/separation ability of nanoporous materials in various ranges of pressures and temperatures without resorting to any empirical parameter fittings. The proposed model accounted for multiple porous occupancy, host lattice relaxation, and the description of the quantum nature of adsorbed gases [1]. In order to evaluate the parameters of weak interactions, a time-dependent density-functional formalism and local density technique entirely in real space have been implemented for calculations of vdW dispersion coefficients for atoms/molecules within the all-electron mixed-basis approach [2].
We applied this approach to construct the phase diagrams of gas hydrates, three-dimensional hydrogen-bonded water structures in which water molecules arrange themselves in a cage-like structure around guest molecules. The thermodynamic stabilities of hydrogen and mixed hydrogen-propane, hydrogen-methane, hydrogen-ethane hydrates as well as the He hydrate based on different ice structures have been studied and obtained results are in agreement with available experimental data [3-5]. Recently, thermodynamic stability of ozone hydrates has been studied and it has been found that the stabilization of ozone in clathrate hydrates is useful for enrichment and long-time storage of ozone without usage of carbon chlorines or fluorides. The formation of CO₂, CH₄, CH₄+CO₂ and N₂+CO₂ hydrates has been investigated at different gas phase compositions, pressures and temperatures using the proposal approach. It has been found that at high nitrogen concentration in the case of CO₂/N₂ gas mixture carbon dioxide can replace methane in the hydrate phase at temperatures and pressures typical for the permafrost regions or below the seafloor since the phase stability of binary N₂+CO₂ hydrate is similar to pure CH₄ hydrate.
The proposed method is quite general and can be applied to the various nanoporous compounds with weak guest-host interactions. From this point of view, the present methodology can support experimental explorations of the novel gas storage and separation materials.
REFERENCES
[1] R.V. Belosludov et al., J. Chem. Phys. 131 (2009) 244510.
[2] R. V. Belosludov et al. in Handbook of Sustainable Engineering, ed. by K-M. Lee and J. Kauffman, Springer, New York, (2013) pp. 1215-1247.
[3] R. V. Belosludov et al.Mol. Simul.38 (2012) 773.
[4] R. V. Belosludov et al.J. Phys. Chem. C118 (2014) 2587.
[5] R. V. Belosludov et al.J. Renew. Sust. Energy,6 (2014) 053132.

In the “materials-genome” (MG) approach, the objective is to use multi-scale materials computation to by-pass time-consuming and expensive experimentation that would be needed for material development and property prediction. In the case of structural materials for load-bearing applications, the bottleneck of the MG approach lies with the development of accurate models to describe the microstructure-defect interactions at the meso-scale. In this talk, a new scheme for computational dislocation plasticity in the meso-scale is presented. This new approach is based on the dynamics of coarse-grained dislocation-density functions. Since any quantity of dislocations can be represented by a density, such an approach does not suffer from the saturation problem of molecular dynamics or discrete dislocation dynamics when the system size is too big to handle. A critical issue to address, however, is a realistic description of the interactions between dislocation densities. Our new scheme takes exact consideration of the mutual elastic interactions between dislocations, through generalization of the Hirth-Lothe formalism of such interactions and reducing the line-integral formulation involved into an algebraic form straightforward enough for efficient numerical implementation. Other features in the model include (i) the continuity nature of the movements of dislocation densities, (ii) forest hardening, (iii) generation according to high spatial gradients in dislocation densities, and (iv) annihilation. Numerical implementation is by means of the finite volume method (FVM), which is well suited for high gradients often encountered in dislocation plasticity.
As a first case study, the model is utilized to predict vibration-induced softening and dislocation pattern formation, which are known experimental phenomena in crystalline metals. The simulations reveal the main mechanism for subcell formation under oscillatory loadings to be the enhanced elimination of statistically stored dislocations (SSDs) by the oscillatory stress, leaving behind geometrically necessary dislocations with low Schmid factors which then form the subgrain walls. The oscillatory stress helps the depletion of the SSDs, by bringing reversals into their motion which then increase their chance of meeting up and annihilation. This is the first simulation effort to successfully capture the cell formation phenomenon under vibratory loadings. A second case study will concern small-scale crystal plasticity. The new model is found capable of capturing a number of key experimental features in the plasticity of micron-sized crystals, including power-law relation between strength and size, low dislocation storage and jerky deformation. These results indicate that our dislocation-density function dynamics approach is promising for predicting dislocation plasticity behaviour in real time and space scales.

In Al-Li-Cu ternary alloys, the stable T₁ phase has a low density, high elastic modulus and high specific strength. Due to this combination of properties, alloys strengthened with the T₁ phase have attracted a great deal of interest especially in aerospace applications. We wish to use atomic-scale density functional theory (DFT) calculations to better understand the effects of T₁ precipitates on the mechanical properties of Al-Li-Cu alloys. A prerequisite for this type of study involves structural information of the precipitate phase; however, even though many experimental studies have been made to discover the T₁ crystal structure, it still remains the subject of some controversy. Here, we use density functional theory (DFT) calculations to perform structural and compositional determination of the T₁ structure by comparing the energetic stability of the previously proposed atomistic models for the T₁ phase. Based on our results, we propose a new structure for the T₁ phase, specifically building upon the work of Van Smaalen et al. [1] with a modification of the position of Li atoms in the T₁ structure. To compare the relative stability of the previously proposed structures, the formation energy of the various experimentally suggested T₁ crystal structures was calculated using a special quasi-random structure (SQS) approach to describe a disordered Al-Cu sublattice. Cluster expansion (CE) calculations are also used to find a corresponding stable T₁ structure on the calculated convex hull of the Al-Li-Cu ternary system. Our calculations indicate that the stoichiometry of the T₁ stable structure at 0 K deviates slightly from Al₂LiCu stoichiometry.
Reference
1. S. Van Smaalen et al. Journal of solid state chemistry 85, 293-298 (1990)

We report on the first prediction of the famous dislocation cell structure in FCC crystals, which we have recently made using continuum dislocation dynamics theory. In this density-based framework, the dislocation density evolution is governed by the rates of transport, cross slip and reactions, which are captured by kinetic equations derived based on statistical mechanics principles. Novel finite element method has been used to solve the dislocation kinetic equations coupled with crystal mechanics. Within this framework, the cross slip process and dislocation-dislocation reactions at short range are treated using a Monte Carlo approach in the continuum frame. It is found that dislocation cells form at strains as low as 1% and that they evolve in the crystal as deformation proceeds, following the similitude law: the average size is inversely proportional to the applied stress. Our simulations show that cross slip is essential in forming dislocation cells and in the cell size evolution. Results showing crystal hardening behaviour, dislocation density evolution, slip patterns and GND density will be presented and analysed in conjunction with 3D spatially resolved X-ray data and similar discrete dislocation dynamics simulations.

Bauschinger Effect (B.E.) is defined from simple forward and reverse loading tests, and is manifested by the reverse flow curve exhibiting a reduced elastic limit, a well-rounded appearance of the initial plastic portion and a permanent softening with respect to the forward hardening curve. Investigation of B.E. is of particular interest for the understanding of cyclic deformation and for the modeling of kinematic hardening in crystal plasticity. Today, existing interpretations and modelling of the B.E. are dominated by the concepts of Backstress and long range internal stresses, which at the elementary scale are justified by the accumulation of geometrically necessary dislocations in a deformed sample. Hewever, such interpretations are in many regards incompatible with the true B.E. observed in single crystals of pure metals.
In this study, the B.E. of Ni single crystals is investigated with the help of 3D dislocation dynamics simulations. Among the different elementary features controlling the strain hardening, we show that the junction strength and the mobile dislocation mean free path are key physical parameters to understand the dislocation microstructure assymtry upon load path changes. A new model based on short-range properties observed in the simulations is proposed. This model, which neglects the possible influence of a Backstress and whose parameters are directly determined from DD simulation results, captures many details of existing experimental data.

High-entropy alloys (HEA) represent a new class of advanced materials with unique properties that cannot be achieved by microalloying processes. In particular, mechanical strength and toughness are usually seen to be improved by increasing the number N of alloy components [1]. The origins of strengthening in these alloys remain uncertain, and there have been no theories that predict the observed trends.
Here, we develop a rigorous strengthening model for multicomponent fcc solid solutions at arbitrary compositions, that is a generalization of the predictive model of Leyson et al. [2] for dilute cases. A key aspect in the development is to adopt an effective medium approach, where each elemental constituent is seen as a “solute” embedded into the effective “matrix” of the surrounding material. We validate this model against extensive MD simulations performed on model Fe_(1-x)/2Ni_(1-x)/2Cr_x alloys. The theory is then applied to predict yield stresses in the Ni-Co-Fe-Cr family of fcc HEAs using only available experimental data, demonstrating the origins of the high strength, high-temperature plateau strength, and detailed trends with composition and number of component in agreement with experiments [3]. A simplified version of the model enables clear identification of the important materials parameters in determining the strength, paving the way towards informatics-type design of HEAs. Possibility to include short-range order effects will be finally discussed.
[1] Y. Zhang, T.T. Zuo, Z. Tang, M.C. Gao, K.A. Dahmen, P.K. Liaw, Z.P. Lu, Prog. Mat. Sci. 61, 1 (2014).
[2] G.P.M. Leyson, L.G. Hector Jr, W.A. Curtin. Acta Materialia 60, 3873 (2012).
[3] C. Varvenne, A. Luque and W. A. Curtin, in preparation.

Better understanding of the nucleation of cracks in metals subject to cyclic loading rests on understanding how the dislocation microstructure evolves as the load is applied and how the surface roughness develops due to that evolution. The surface of crystals roughen as a result of dislocation escape from the crystal's bulk to the surface. In this work, the three-dimensional discrete dislocation dynamics (DDD) method is used to simulate the dislocation activity taking place in single crystal pure FCC Nickel loaded cyclically. Early stages of the surface roughness development is studied in crystals having sizes ranging from 0.5 to 5 microns with initial dislocation densities ranging between 10^11 to 10^14/m^2. The initial dislocation microstructure was randomly generated with no specific pattern prescribed. The effects of various deformation mechanisms, such as cross-slip, are explicitly accounted for. The events of dislocation escape from the crystal surface are captured and the full time dependent surface morphology due to the dislocation escape has been reconstructed. Surface roughness has been quantified using a number of different roughness metrics and the effects of crystal size, dislocation density and the number of loading cycles are presented. It was found that roughness is seldomly reversible and that bigger crystals develop roughness faster than smaller ones. Furthermore, the surface roughness evolution in crystals having a well-developed persistent slip band (PSB) is also investigated. The correlations between the dislocation density in the walls and the channels and the surface roughness in one full cycle of loading are computed. The results are finally validated by direct comparisons with experimental results. Finally, further insights on the correlation between the spatio-temporal evolution of the point defects (due to dislocation climb) concentration and surface roughness evolution are discussed.

Metals and alloys used for engineering applications often undergo multi-axial strain path changes during their processing, and under service conditions, attributing them their typical mechanical properties. Depending on the type of strain path changes, different kinds of plasticity mechanisms are activated which could result in either a lowering of yield strength (Bauschinger effect), or an increase. Understanding these multi-scale mechanical phenomena requires studying their microstructure and residual stress evolution using both experiment and modeling techniques. Mechanical tests involving in-situ diffraction have helped provide an interpretation of the macroscopic mechanical behavior based on evolution of internal lattice strains, however, these are usually performed under uniaxial loading conditions [1]. Phenomenological and crystal plasticity models are capable of predicting the local and effective response of metals undergoing strain path changes (for example [2,3]), however, hardening laws implemented in these models are typically based on uniaxial experimental data and may not appropriately capture behavior of materials subjected to multi-axial stress states.
In the present work, we propose to extend the synergy between experiments and modeling to study the role of multi-axial strain path changes on mechanical behavior of materials. Recently, in-situ neutron diffraction experiments were performed during strain path change tests on bi-axial proportionally and non-proportionally loaded cruciform shaped samples of 316L stainless steel. Finite element simulations are carried out at the macro scale to study the predictive capabilities of different phenomenological hardening laws. These simulations also underline the role of cruciform sample geometry on the mechanical response of 316L steel during uniaxial, equi-biaxial and 900 strain path change tests and can be used to guide future experiments involving different geometries. At the meso-scale, the results of macro scale simulations are used to drive an elasto-viscoplastic fast Fourier transform model to test predictive capabilities of dislocation and twin based hardening laws. Virtual diffraction profiles are generated to understand the role of type II and type III stresses on the evolution of internal lattice strains and peak broadening. Results reveal that recovery of internal stresses could play a crucial role on the mechanical response after strain path change and should be accounted for in existing hardening models.
[1] Gonzalez et al., Mat. Sci. Engg. A 546,2012,263-271
[2] Zecevic et al., Mat. Sci. Engg. A 638,2015,262-274
[3] Mánik et al., IJP 69,2015,1-2

A persistent challenge in multiscale modeling of materials is the prediction of plastic materials behavior based on the evolution of the dislocation state. An important step towards a dislocation based continuum description was recently achieved with the so called Continuum Dislocation Dynamics (CDD) [1]. CDD captures the kinematics of moving curved dislocations in flux-type evolution equations for dislocation density variables, coupled to the stress field via average dislocation velocity laws based on the Peach-Koehler force. The lowest order closure of CDD employs only three internal variables per slip system, namely the total dislocation density, the classical dislocation density tensor and a so called curvature density.
Distributions of dislocations and plastic slip obtained from strongly simplified two-dimensional CDD simulations were earlier shown to reflect salient features observed in three-dimensional discrete dislocation simulations [1]. In the current work we present a three-dimensional implementation of the lowest order CDD theory in conservative form [2] as a materials sub-routine for ABAQUS in conjunction with the crystal plasticity framework DAMASK [3]. We use this code to simulate bending of micro-beams and compare the mechanical response and the dislocation distribution on multiple slip systems to discrete dislocation results from the literature [4]. The CDD simulations reproduce size effects, dislocation pile-ups and a reduced plastic slip close to the surfaces, as observed in experiments and discrete dislocation simulations.
[1] T. Hochrainer, S. Sandfeld, M. Zaiser, P. Gumbsch (2014). Continuum dislocation dynamics: Towards a physical theoryof crystal plasticity, J. Mech. Phys. Solids, 63, 167-178
[2] A. Ebrahimi, M. Monavari, and T. Hochrainer (2014). Numerical Implementation of Continuum Dislocation Dynamics with the Discontinuous-Galerkin Method, MRS Proceedings, 1651, mrsf13-1651-kk06-05
[3] F. Roters, P. Eisenlohr, C. Kords, D.D. Tjahjanto, M. Diehl, D. Raabe (2012). DAMASK: the Düsseldorf Advanced MAterial Simulation Kit for studying crystal plasticity using an FE based or a spectral numerical solver,IUTAM Symposium on Linking Scales in Computations: From Microstructure to Macro-scale Properties, Procedia IUTAM 3 , 3-10
[4] C. Motz, D. Weygand, J. Senger, P. Gumbsch (2008). Micro-bending tests: A comparison between three-dimensional discrete dislocation dynamics simulations and experiments, Acta Materialia, 56, 1942-1955

Gallium nitride(GaN)-based LEDs have a large number of defects such as threading and misfit dislocations (in the order of 10⁸ - 10¹⁰cm^-2). These threading and misfit dislocations in the GaN film have been attributed to high lattice-mismatch between the sapphire substrate and the film. The threading dislocations in GaN films greatly reduce the optoelectronic performances as well as the life time of these devices. Therefore, it is important to understand the dislocation formation, interaction and growth mechanism in the gallium nitride film layer. In this work, we have studied dislocation formation, their interaction with other dislocations and their growth mechanism in the GaN film using a multiscale approach. The electroelastic field components of a single piezoelectric dislocation in the GaN film have been investigated analytically by using the integral method based on the Stroh formalism. The finite element modeling and analysis of a single piezoelectric dislocation in the GaN film also have been carried out to compare the analytical results of the electroelastic fields generated by the single dislocation. Dislocations in the wurtzite GaN also have been modeled by molecular dynamics. The potential model is very important in the molecular dynamics simulation in explaining the stress field around dislocations and in understanding the glide mechanisms. The Stillinger-Weber semi-empirical interatomic potential and second nearest-neighbor modified embedded atom method have been used for this purpose. The stress field obtained by molecular dynamics simulations has been compared with the analytical and finite element analysis results.
The mobilities of the edge and screw dislocations have also been analyzed in this work. Dislocations move by glide at velocities which depend on the applied shear stress, purity of crystal, temperature and the structure of dislocation. In particular, the structures of 5/7 core, 4 core a-type edge and screw dislocations have been investigated to compare the mobility of each dislocation type. The glide mechanisms of dislocations have been examined in various state of applied shear stress and temperatures. Molecular dynamics simulation of dislocations in this work can provide the stress tensor components and the velocity of dislocations as a function of temperature and applied shear stress. The results can be used in dislocation dynamics simulations to better understand the underlying mechanism of dislocation generation and movement in GaN devices.

Shear bands are the macroscopic flow defects in bulk metallic glasses (BMGs). They are thin bands composed of large localized shear strains, i.e. shear transformation zones (STZs), due to the plastic instability during deformation. The yielding, plasticity, and fracture of BMGs are essentially associated with shear bands mechanism. When a shear band arises, the increase of local temperature by the energy dissipation of STZs in the shear-band plane can be very high or low. Even though this local heating could be observed in a very short period of time, it may significantly influence the formation of shear band itself and change the deformation behavior of metallic glass. A heterogeneously random Eshelby shear transformation model [P.Y. Zhao, J. Li, and Y.Z. Wang, Int. J. Plasticity 40, 1 (2013)] based on microelasticity and kinetic Monte Carlo approach is used to investigate the possible effects by local heating, e.g., structural relaxation, on mechanical behavior of metallic glass. This model can overcome the spatial-temporal scale restrictions in the study of shear bands by molecular dynamics. In this work, the heat equation and a heat sink are introduced into the model for the variation in local temperature. As a result, the process of shear localization with the relation between inhomogeneous flow and heat evolution in metallic glass during deformation can be discussed. The work may inspire new strategies for the design of BMGs.

Rare-earth zirconate pyrochlores are considered as promising candidate materials for the thermal barrier coating (TBC). One key material parameter to assess the reliability and durability of various RE zirconate pyrochlores as TBC materials is the coefficient of thermal expansion (CTE). Thus, to facilitate material exploration, the ability to accurately predict the CTE is of critical importance. In this study, we performed first-principles calculations to investigate the thermal expansion behaviors of several RE₂Zr₂O₇ pyrochlores. Our findings show that RE₂Zr₂O₇ pyrochlores exhibit low-lying optical phonon branches that become imaginary as the volume expands, preventing correct characterization of thermal expansion behaviors. These branches were shown to correspond to RE-cation rattling vibrational modes that exhibit great anharmonic features. We proposed a tailored quasi-harmonic (QH) approximation based on phonon calculations to remedy the effects due to those rattling modes to produce the correct free energy versus volume relationship. Using this approach, the CTEs of RE₂Zr₂O₇ pyrochlores were predicted. The model predictions were then compared with available experimental data, showing excellent agreement. Meanwhile, the QH Debye model and the first-principles molecular dynamics (FPMD) simulations were also employed to evaluate the CTEs of RE₂Zr₂O₇ pyrochlores. We found that the predictions from the FPMD simulations are also in close agreement with experimental data while the QH Debye model largely overestimates the CTEs.

In the field of Solid State Lighting, all-organic compounds (metal-free) are often overlooked as potential emitters for phosphor applications since they are typically non-phosphorescent. Recently however, organic phosphorescence has been described in novel all-organic crystals. Inspired by the “heavy-atom effect”, often seen with addition of iridium or osmium, bromine has been used to induce efficient room-temperature phosphorescence in an organic crystal. However, the phosphorescence effects are lost when this molecule is placed in solution. The underlying mechanisms of this phenomenon are thought to be intermolecular halogen bonding between Bromine and Oxygen. To elucidate the governing mechanisms we explored the electronic and optical properties of a series of purely organic molecules using time-dependent density functional theory (TDDFT). These calculations are compared with the absorption, fluorescence, and phosphorescence experimental spectra in their crystals as well as solution forms to elucidate the functional phosphorescent mechanism. We illustrate how these results can be used to guide future PHOLED design.

Recent experiments aimed at probing the dynamics of excitons have revealed that semiconducting films composed of disordered molecular subunits, unlike expectations for their perfectly ordered counterparts, can exhibit a time-dependent diffusivity in which the effective early time diffusion constant is larger than that of the steady state. This observation has led to speculation about what role, if any, microscopic disorder may play in enhancing exciton transport properties. In this article, we present the results of a model study aimed at addressing this point. Specifically, we introduce a general model, based upon Förster theory, for incoherent exciton diffusion in a material composed of independent molecular subunits with static energetic disorder. Energetic disorder leads to heterogeneity in molecule-to-molecule transition rates, which we demonstrate has two important consequences related to exciton transport. First, the distribution of local site-specific hopping rates is broadened in a manner that results in a decrease in average exciton diffusivity relative to that in a perfectly ordered film. Second, since excitons prefer to make transitions that are downhill in energy, the steady state distribution of exciton energies is biased towards low energy molecular subunits, those that exhibit reduced diffusivity relative to a perfectly ordered film. These effects combine to reduce the net diffusivity in a manner that is time dependent and grows more pronounced as disorder is increased. Notably, however, we demonstrate that the presence of energetic disorder can give rise to a population of molecular subunits with exciton transfer rates exceeding those of subunits in an energetically uniform material. Such enhancements may play an important role in processes that are sensitive to molecular-scale fluctuations in exciton density field.

Interfaces such as grain boundaries and line defects are the predominant sites where mechanical failure, chemical interactions, segregation of impurities, nucleation and rapid atomic diffusion may occur. This can often be detrimental to technologically important nanomaterials and thus a fundamental understanding of how such one-dimensional defects may have an influence on its physio-chemcal properties is pivotal to optimizing their device performance. Of late, phosphrene - the new member to the two-dimensional nanomaterials, has attracted much attention its high carrier mobility and good mechanical flexibility. [1-4]
In this study, using density-functional theory, we investigate the energetics and electronic structure of a single-layered phospohrene with various line defects. We have generated different line defect models based on a fault method, rather than the classical rotating method (which is typically used for studying bulk grain boundaries). This has allowed us to generate new line defects with a much lower defect formation energy (even when compared to that found on graphene and h-BN), and we show how these low-energy line defects could well modulate the electronic band gap energies of this new two-dimensional nanomaterial. [5]
[1] E. S. Reich, Nature,506, 19 (2014)
[2] X. Peng et al., Phys. Rev. B. 90, 085402 (2014)
[3] A. Ramasubramaniam et al.,J. Appl. Phys.111, 054302 (2012)
[4] H. Liu et al.,ACS Nano, 8, 4033 (2014)
[5] Y. Liu et al., Nano Lett. 14, 6782 (2014)

Understanding the magnetic properties of a molecular spintronics device is critical for its advancement and futuristic applications in computer memory and logic units. This paper reports our Monte Carlo (MC) studies aiming to explain the experimentally observed paramagnetic molecule induced antiferromagnetic coupling between the ferromagnetic (FM) electrodes. Recently developed magnetic tunnel junction based molecular spintronics devices (MTJMSDs), exhibited molecule induced strong antiferromagnetic coupling. MTJMSDs were prepared by chemically bonding the paramagnetic molecules between the FM electrodes along the exposed side edges of magnetic tunnel junctions. Our MC studies focused on the atomic model analogous to the MTJMSD and studied the effect of molecule&’s magnetic couplings with the two FM electrodes. Simulations show that when a molecule established ferromagnetic coupling with one electrode and antiferromagnetic coupling with the other electrode then theoretical results effectively explained the experimental findings. MC and experimental studies suggest that the strength of exchange coupling between molecule and FM electrode should be ge;50% of the interatomic exchange coupling strength of the FM electrodes. This paper also discusses various forms of molecular coupling with ferromagnetic electrodes and their effect on the wide range of molecular spintronics devices.

Ceramic dielectric capacitors with high energy density and low energy loss are of ever-increasing importance in today&’s high power electric and electronic systems, from smart grids and switch power in consumer use to electric guns and direct energy weapons in military service, from oil drilling devices and spacecraft in harsh environment to magnetic resonance imaging and implantable defibrillators in medical system. Ferroelectrics, which intrinsically possess high polarization, have nowadays been explored as high energy storage candidates, while it is still challenging to reduce the hysteresis loss and enhance the ultimate breakdown strength. Researchers turned to coat the ferroelectric ceramics with some linear dielectrics of high breakdown strength to sustain dielectric breakdown through the core-shell modification of the ceramic microstructures; nevertheless, theoretical analysis or modeling of how the linear shell influences the hysteretic polarization and the energy storage behavior is still scarcely conducted. In this work, a microstructure-level model of the core-shell structured ferroelectric ceramic system was successfully developed through the voronoi tessellation random construction routine. With the assumption that the linear shell has no hysteresis in polarization and the hysteretic ferroelectric core is depicted by a modified hyperbolic tangent model, finite element method was applied to solve the composite system with simplified boundary conditions. Statistics of electric field spatial distribution indicate that increasing applied electric field leads to intensified electric field fluctuation and this effect can be alleviated by the incorporation with thicker shell. Energy storage properties were extracted from the as-calculated hysteresis loops. The charged as well as the discharged energy density possesses a maximum value with regard to the shell fraction. The influences of several parameters such as the remnant and the saturation polarization, the core initial permittivity and the shell permittivity on the energy storage behaviors were investigated. Results show that enhancing the saturation polarization and meanwhile lowering the remnant polarization provide the most effective way to obtain ceramics with high discharged energy. Reducing the core initial permittivity and the shell permittivity are helpful in the improvement of energy efficiency. Given experimental parameters, this model can guide the design and optimization of core-shell structured ferroelectric ceramics for energy storage.

Balsa, is one of the lowest density wood species in nature. Due to its superior specific stiffness and strength properties, it is commonly used as the core materials in structural sandwich panels. To better understand the mechanics of balsa, we present a multi-scale modelling approach for predicting the effective elastic properties of its cell wall material. A three-step homogenization using finite element method is used to obtain the overall elastic properties of the cell wall material. Properties of chemical constituents and their arrangements are considered at different resolution levels. For this purpose, different unit cells, representing the structure of the material at each level (i.e. nano-, micro- and meso-level) are identified. The role of cellulose crystallinity and the effect of microfibril angle (MFA) in S2 layer are highlighted. Although the focus is on modelling the cell wall material of balsa, the application of the model in estimating the effective properties of other wood species as well as bamboo is also discussed.