Computations for materials – discovery, design and the role of data

Resume : JARVIS (Joint Automated Repository for Various Integrated Simulations) is a unique integrated framework to accelerate material design using classical force-fields (FF), density functional theory (DFT) and machine learning (ML). The JARVIS-DFT hosts data for more than 30000 materials. We discovered more than 1500 2D materials using lattice parameter criteria and exfoliation energy calculations. We charted improved lattice parameters, formation energies and elastic tensors using van der Waals functional for more than 12000 materials and established relation between exfoliation energies and elastic constants. To alleviate bandgap underestimation in conventional DFT and improve frequency dependent dielectric function predictions, we evaluated meta-GGA based approaches for more than 15000 materials. We use spectroscopic limited maximum efficiency approach to identify potential photovoltaic materials. Using spin-orbit spillage criteria, we discovered more than 1500 potential topological materials including topological insulators, Weyl and Dirac semimetals, and topological crystalline insulators. The database is publicly available at https://jarvis.nist.gov/.

Resume : I will introduce The Computational 2D Materials Database (C2DB) -- a new platform for modeling and discovering atomically thin crystals and their heterostructures [Haastrup et al., 2D Materials 5, 042002 (2018)]. The C2DB organises a variety of structural, thermodynamic, elastic, electronic, magnetic, and optical properties of around 3500 different 2D materials distributed over more than 40 different crystal structures. Materials stability and properties are systematically calculated by state-of-the-art density functional theory and many-body perturbation theory (GW and the Bethe–Salpeter equation for the simpler materials) following a semi-automated workflow for maximal consistency and transparency. The C2DB is fully open and can be browsed online (http://c2db.fysik.dtu.dk) or downloaded in its entirety. I will show how the we utilised the database to design novel Janus structures for predictive tuning of band alignment at heterointerfaces, new 2D semiconductors with high mobilities and strong light-matter coupling, and materials with non-trivial topologies.

Resume : Exploratory synthesis in novel chemical spaces is the essence of solid-state chemistry. However, uncharted chemical spaces can be difficult to navigate, especially when materials synthesis is challenging. Nitrides represent one such space, where stringent synthesis constraints have limited the exploration of this important class of functional materials. Here, we employ a suite of computational materials discovery and informatics tools to construct a large stability map of the inorganic ternary metal nitrides. Our map clusters the ternary nitrides into chemical families with distinct stability and metastability, and highlights hundreds of promising new ternary nitride spaces for experimental investigation—from which we experimentally realized 7 new Zn- and Mg-based ternary nitrides. By extracting the mixed metallicity, ionicity, and covalency of solid-state bonding from the DFT-computed electron density, we reveal the complex interplay between chemistry, composition, and electronic structure in governing large-scale stability trends in ternary nitride materials. An interactive version of the map can be found at https://wenhaosun.github.io/TernaryNitridesMap.html

Resume : Two-dimensional (2D) materials have attracted tremendous amount of interest. Their unique properties are expected to lead to new technologies. In an effort of systematic 2D materials discovery, we have been using both the top-down and the bottom-up approaches to generate 2D structures. On one hand, monolayer structures are theoretically exfoliated from layered three-dimensional structures. On the other hand, chemical elements in known 2D materials are systematically substituted by similar elements to generate new 2D materials. High throughput first-principles calculations are carried out to study their physical properties. Our progress in this project will be reported and some preliminary applications, such as bilayer heterostructures for excitonic solar cell applications, of the 2D materials database will be discussed.

Resume : There has been an ongoing controversy on 1T-TiS2’s semimetallic1,2 or semiconducting nature3 for several decades due to the difficulty to synthetize stoichiometric TiS2 and the dispersion of calculation results. We tackle this issue by performing calculations on the CdI2 layered structure type with the ideal 1:2 Ti:S stoichiometry using some of the most refined models, e.g. hybrid functional including Grimme’s dispersion effects, the GW scheme and the Bethe-Salpether equation for the optical properties. All the calculations on the bulk (ideal) structure demonstrate that a (small) bandgap exists for the ideal material. Besides, van der Waals interactions must be taken into account during the calculation in order to reproduce accurately the TiS2 structure, more precisely a correct c/a ratio. Then, for the first time we study point defects in the material using the same level of accuracy. These calculations demonstrate unambiguously that the most stable defect is the titanium interstitial located in an octahedral site inside the van der Waals gap. This point defect may explain the distinguishable electronic behavior of TiS2 reported in the literature, i.e. non degenerated semiconductor (SC) or semimetal (SM) depending on defect concentration as a result of the synthesis conditions. (1) Liu, B.; Yang, J.; Liu, C.; Hu, T.; Han, Y.; Gao, C. The Ground Electronic State of TiS2: Experimental and Theoretical Studies. Phys. Status Solidi Curr. Top. Solid State Phys. 2011. https://doi.org/10.1002/pssc.201000612. (2) Thompson, A. H.; Pisharody, K. R.; Koehler, R. F. Experimental Study of the Solid SolutionsTixTa1-xS2. Phys. Rev. Lett. 1972, 29 (3), 163–166. https://doi.org/10.1103/physrevlett.29.163. (3) Logothetis, E. M.; Kaiser, W. J.; Kukkonen, C. A.; Faile, S. P.; Colella, R.; Gambold, J. Transport Properties and the Semiconducting Nature of TiS2. Phys. B+C 1980, 99 (1–4), 193–198. https://doi.org/10.1016/0378-4363(80)90231-4. (4) Stoliaroff, A.; Jobic S.; Latouche, C. Optoelectronic properties of TiS2: a never ended story tackled by hybrid functional and many body methods, Inorg. Chem. 2019 (5) Stoliaroff, A.; Latouche, C. ; Jobic, S. 2019

Resume : This talk will highlight both some of the successes and the stumbling blocks along the way to bring material science into the 21st century. Namely, the combination of (i) high-throughput experiment, (ii) massive computational resources, (iii) machine learning, and (iv) high energy beamline-based studies is more than summative. Given finite human resources (ie graduate students), these tools serve as force multipliers to enhance their capabilities. Over the last decade, we have applied these techniques in concert to the field of thermoelectric materials. This high-throughput search of known and hypothetical compounds has led to the discovery of new classes of materials for these two applications. Most notably, the search reveals entire classes of materials that had been ignored due to historical bias within the thermoelectric community. Through the interplay of materials physics and machine learning, general principles for designing the next generation materials have emerged. The talk will conclude with discussion of the prospects for sequential learning coupled with parallelization and automation to accelerate functional materials discovery.

Resume : Materials discovery is increasingly going beyond the prediction of thermodynamically stable materials at Daltonian compositions and their crystallographic primitive cells, including, e.g., metastable compounds, solid solutions, as well as defect- and disorder-enabled materials. Due to their relative instability against N2, nitrides are often grown as thin films via non-equilibrium deposition at relatively low temperatures, leading to atomic disorder. Following a broad search for ternary nitrides spaces, we identified novel wurtzite and rock-salt nitrides by first principles crystal structure prediction, including Zn-M (M = Sb, Mo, W) and Mg-M (M = Nb, Ti, Zr, Hf) [1, 2]. The effects of disorder on structure selection and electronic structure properties are studied by Monte-Carlo sampling. Disorder is also the inherent driving force for the formation of solid solutions. We recently studied the formation and the phase diagram of heterostructural alloys, i.e., the mixture of compounds with different crystal structures, such as MnO-ZnO and SnS-CaS [3]. We showed that, compared to the more conventional case of isostructural alloys (e.g. InGaN), heterostructural alloys can exhibit a markedly increased range of metastable compositions between the binodal and spinodal lines, thereby enabling the design of novel homogeneous single-phase alloys. [1] E. Arca et al, J. Am. Chem. Soc. 140, 4293 (2018) [2] W. Sun et al, arxiv/1809.09202 (2018) [3] Holder et al, Sci. Adv. 3, e1700270 (2018)

Resume : The recent observation of ferromagnetic order in two-dimensional (2D) materials highlights the importance of understanding the fundamental aspects of magnetism in 2D materials. In contrast to bulk materials in three dimensions, magnetic anisotropy plays a crucial role for magnetic order in 2D and is therefore intimately linked to spin-orbit coupling. In the present work we have used the Computational 2D Materials Database (C2DB) to search for new ferromagnetic 2D materials. In addition to the well-known ferromagnetic 2D materials CrI$_3$ and VSe$_2$, we find 12 novel insulating materials that exhibit magnetic order and evaluate their critical temperatures from Monte Carlo simulations using our calculated values for exchange and anisotropy parameters. A few of these exhibit critical temperatures exceeding 100 K and could comprise promising candidates for future investigations of magnetism in 2D. We also investigate the effect of Hubbard correction in the frame work of DFT+U and find that the chosen value of U can have a crucial influence on the magnetic properties.

Resume : Two-dimensional materials are promising candidates for use as tunnel barrier in atomically thin magnetic tunnel junctions (MTJs) [1]. High magneto resistance ratios have been predicted theoretically and recent progress in large scale manufacturing of these materials has paved the way towards their integrations in functional devices. Yet, the experimental results available so far vary greatly depending on the integration pathways. Seeking for increased performances, it has been shown lately that direct CVD growth of tunnel barriers improves significantly the quality of the ferromagnet-2D materials interfaces [2-4]. Following these recent developments, new phenomena such as the bias induced reversal of the magneto resistance were reported [5]. Here, we show that first-principles calculations can provide direct insights into the close relation that links the interface morphology to its magneto resistive behavior. In particular, we report on the interaction between ferromagnet electrodes and graphene/h-BN 2D layers as well as on the origin of TMR reversal in h-BN based MTJs [5]. References [1] M. Piquemal-Banci et al., J. Phys.D. Appl. Phys., 50 (2017), 203002 [2] S. Caneva et al., Nano Lett.,16 (2016), 1250. [3] M.-B. Martin et al., Appl. Phys. Lett., 107 (2015), 012408. [4] S. Caneva et al., ACS Appl. Mater. Interfaces, 9 (2017), 29973. [5] M. Piquemal-Banci et al., ACS Nano, 12 (2018), 4712.

Resume : We use numerical calculations to study effects of structural imperfections (point and line defects) and external mechanical or magnetic field on electronic and transport properties of graphene sheet comprising millions of atoms, i.e. layer of several hundred nanometres in size. Point defects are modelled as resonant (neutral) adsorbed atoms or molecules, vacancies, charged impurities, and local distortions. Line defects are attributed to atomic steps and terraces in epitaxial graphene, and grain boundaries, nanoripples or wrinkles in polycrystalline graphene. Results are obtained numerically using the quantum-mechanical Kubo-Greenwood formalism and tight-binding approach. Calculated behaviours of electronic density of states and conductivity indicate that deviations from perfection can be useful: they make possible tailoring graphene electrotransport properties for achievement of new functionalities. Particularly, the ordering of point defects can open a band gap in the energy spectrum of graphene, enhance its conductivity up to dozens (10-30) of times; and orientation correlation of linear defects can increase the conductivity up to 4-5 times. If there are both types defects, their ordering and correlation may improve the conductivity up to hundreds of times as compared with their random distribution. The presence of both point and extended defects reduces Landau levels, broadens, smears, and can even suppress them. The utilizing of the combination of the shear strain and uniaxial tensile strain is found to be the easy and stable way for the band-gap opening and tuning.

Resume : Synthetic polymers have been widely experimented in its applications for endocytosis drug delivery. To advance the molecular understanding of polymer interaction with charged lipid bilayers, Molecular Dynamics (MD) simulations have been performed using poly (L-lysine iso-phthalamide) grafted with L-phenylalanine in 75 mol% substitution degree (PP-75) and charged binary lipid membranes (DOPC/DOPS). The impact of chain lengths, degree of protonation and monomer sequence have been investigated systematically using MD. Area per lipid and thickness of the bilayer was calculated to determine the membrane geometry change after penetration. These results are helpful to interpret AFM experiments of the polymer adsorbed at biological membranes. Comparison of results with charged and uncharged membranes should provide an understanding of the role of lipid charge on polymer adsorption and penetration in the biological membranes.

Resume : The performance of polymer electrolyte membrane fuel cells (PEMFCs) is highly dependent on perfluorosulfonic acid (PFSA) polymer electrolyte membranes. In this investigation, we used full atomistic molecular dynamics (MD) simulations to reveal the effect of side chain pendants in PFSA polymers. Through these MD simulations, the structural and transport properties of PFSA polymers are analyzed with different contents of water at room temperature (298.15K) and operating temperature (353.15K) of PEMFCs. We performed the pair correlation function (PCF) analysis and free volume analysis to check morphologies of water inter-connections in PFSA polymer electrolyte membranes which is crucial to generate the water channel for proton conduct and oxygen transport. We also checked the diffusion of water, hydronium and oxygen by mean square displacement (MSD) analysis with different side chain pendants in PFSA polymers at various water contents. Thus, our primary objective in this simulation study is to elucidate the role of side chain pendants in determining the structural and transport properties in PFSA polymers electrolyte membranes, which will shed light on the design of PFSA polymers at a molecular level. Acknowledgement This work was supported by the R&D Collaboration Programs of Hyundai Motor Company.

Resume : Small-molecule organic semiconductors are used in a wide spectrum of applications, ranging from organic light emitting diodes1 to organic photovoltaics. A number of factors determine mobility, such as molecular packing, electronic structure, dipole moment and polarizability. Presently, quantitative ab-initio models to assess the influence of these molecule-dependent properties, including the influence of dopants, are lacking. Here, we present a multi-scale model, which provides an accurate prediction of experimental data over ten orders of magnitude in mobility,2 and allows for the decomposition of the carrier mobility into molecule-specific quantities. The model consists of a multi-step procedure, incorporating single molecule parameterization, generation of atomistic morphologies,3 DFT based electronic structure calculations yielding site energies, energy disorder, electronic couplings and reorganization energies.4 These parameters are used in an analytic model5 to compute the charge carrier mobility of the amorphous materials. We also provide molecule-specific quantitative measures how two single molecule properties, the dependence of the orbital energy on conformation and the dipole induced polarization determine mobility for hole-transport materials. On the basis of this methodology we are able to computationally predict novel pure ETL materials with three orders of magnitude higher mobility than their precursors and elucidate the molecular mechanism of doping these materials with kinetic Monte-Carlo simulations. The availability of first-principles based models to compute key performance characteristics of organic semiconductors may enable in-silico screening of numerous chemical compounds for the development of highly efficient opto-electronic devices. 1 Groves, C. Organic light-emitting diodes: Bright design. Nat. Mater. 12, 597-598 (2013). 2 Friederich, P. et al. Molecular origin of the charge carrier mobility in small molecule organic semiconductors. Accepted in Adv. Functional Mater.31, 5757, (2016). 3 Neumann, T., Danilov, D., Lennartz, C. & Wenzel, W. Modeling disordered morphologies in organic semiconductors. J. Comput. Chem. 34, 2716-2725, doi:10.1002/jcc.23445 (2013). 4 Friederich, P., Symalla, F., Meded, V., Neumann, T. & Wenzel, W. Ab Initio Treatment of Disorder Effects in Amorphous Organic Materials: Toward Parameter Free Materials Simulation. J. Chem. Theory Comput. 10, 3720-3725, doi:10.1021/ct500418f (2014). 5 Rodin, V. et al. A generalized effective medium model for the carrier mobility in amorphous organic semiconducturs. Phys. Rev. B 91, doi:10.1103/PhysRevB.91.155203 (2015).

Resume : Molecular π-conjugated semiconductors are used as active materials enabling charge transport for a range of applications in organic electronics. However, the lack of long-range order and elusiveness of structure-property rules hinders the systematic improvement of device performances for these amorphous systems. Molecular dynamics (MD) simulations, combined with the estimation of electronic couplings between molecular neighbors, are often used to study charge transport pathways and their dimensionality. Recent efforts have sought to adapt concepts and algorithms from graph theory and classical resistor theory to define molecular networks made of “electronically connected” units within a bulk amorphous molecular solid. In this representation, every molecule is a node of the graph and electronic couplings are the edges connecting the nodes. We present here an open-source molecular network code that implements and expands these concepts, offers powerful visualization tools and shows a >100x speedup compared to previously published benchmarks. The dynamical properties of the charge percolation networks, such as their lifetime and robustness, can eventually be used to build a macroscopic transport model (ie Kinetic Monte Carlo) including dynamic disorder, similarly to what is being done for crystalline systems. As a case study, we consider the impact of alkyl side chain length on the electronic connectivity, film morphology and photovoltaic efficiency of a series of ITIC derivatives.

Resume : Electro-active polymer β-polyvinylidene fluoride (β-PVDF) based ferroelectric composites have gained significant technological importance over conventional ceramic ferroelectrics. But synthesis of β-PVDF has been a challenge owing to its structural instability. Hydrated metal salt systems are found as one of the additive materials which effectively induce polar β-PVDF from non-polar PVDF blend (which contains majorly α phase). This article provides quantum chemical description of PVDF – hydrated aluminium nitrate salt composite system in the light of density functional theory. Four monomer units of pristine α and β-PVDF, pure Al(NO3)3.9H2O, and PVDF/ Al(NO3)3.9H2O structures are optimized using dispersion corrected exchange correlation functional B3LYP-D3 and 6-311 G(d,p) basis set. Similar to the experimental findings, the current theoretical investigation also suggests that hydrogen bond interaction between PVDF and the hydrated salt molecule plays the major role for the enhancement of ferroelectric properties in this composite system. Non-covalent interaction phenomenon is elucidated in detail on the basis of natural bond orbital analysis, Bader’s quantum theory of atoms in molecules and reduced density gradient analysis. Chemical reactivity and charge transfer mechanisms are explained using molecular electrostatic potential plot, frontier molecular orbital analysis and atomic-dipole corrected Hirshfeld population analysis, respectively.

Resume : Polymer electrolyte membrane fuel cells (PEMFCs) are expected to be useful as energy convertors for transportation and stationary applications, due to its high energy efficiency, low emissions and low working temperature. However, its high production cost and poor long-term durability impede the widespread commercialization process. A significant portion of the total cost is the cost of the membrane electrolyte assembly (MEA). The performance of the cathode catalyst layer in the MEA is a major controlling factor of the ultimate cell performance. Therefore, numerous studies on the understanding of the morphology and characteristics on catalyst layer have been carried out. However, the morphology and accompanying properties about the distribution of ionomer and dispersion solvent on the catalyst layer still involve many unknown aspects. In the present study, a molecular-level morphological insight is gained by means of molecular dynamics (MD) simulation. For controlling the distribution of Nafion ionomer and dispersion solvent on catalyst layer, we alter two key factors: (i) the type of dispersion solvent (dipropylene glycol (DPG), water and DPG/water mixture) and (ii) catalyst surface properties (Pt surface and alkylthiol-modified Pt surface). Depending on the type of dispersion solvent, the dispersion state of Nafion ionomer was changed. As increasing the ratio of dipropylene glycol (DPG) in dispersion solvent, the Nafion ionomer was well distributed in the dispersion solvent. Moreover, we find that the alteration of interfacial components in contact with the surface of catalyst layer take place by changing the surface property of catalyst. These findings can provide new insights into designing the optimized interface structure of catalyst layer, with excellent performance of PEMFC. Acknowledgements This research was supported by the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (Nos. NRF-2015M1A2A2057129 and NRF-2016M1A2A2937151).

Resume : The current work presents a methodology of simulation based design of Magnetic Resonance Imaging (MRI) magnets, modified with a symmetric ferromagnetic shell (pure Fe). This design employs the gradient-based optimization solver of COMSOL Multiphysics to obtain an optimum geometrical arrangement of symmetrical electromagnets in MRI. The aim of this work is, to design an MRI magnet system with minimum ampere-turns, high field homogeneity and having magnetic coils with shorter radii. The optimized design solution provides the dimensions of coil blocks, the total length of conductor in each coil block, positions of the coil blocks, current and the resultant magnetic field. The present work reveals that the modification of an MRI system with a ferromagnetic shell significantly improves the field uniformity within a central spherical zone having a diameter of 40-50 cm and greatly reduces the current requirement. The technique used for the optimization can be extended to other kinds of geometries.

Resume : Propane dehydrogenation (PDH) has been used for producing propylene, which is a crucial precursor for synthesizing a variety of industrial products. Besides the well-known Pt-based catalysts, other catalysts with high reactivity, high selectivity and lower cost have been sought for this application. Among metal-based catalysts, Ni was reported as an active catalyst for hydrogenation and dehydrogenation reactions of small alkanes. However, poor selectivity of propylene production and fast deactivation of Ni were observed in experiment. In this work, a plane-wave based density functional theory (DFT) method had been used to investigate PDH reaction mechanism on bare Ni surface and to study a role of dopant on its catalytic efficiency. Our calculated results revealed that the 1-propyl intermediate is the preferable pathway on bare Ni (111). Our finding also suggested that the strong propylene adsorption on Ni results in poor selectivity and leads to further deep hydrogenation or C-C bond cracking. The calculations clarified the effect of Au or Sn on reactivity of Ni, its selectivity towards propylene production and coke deposition on the surface. Electronic charge properties were elucidated to explain the nature of active sites of those catalysts. This study provides a good guidance for developing Ni-based catalysts used in dehydrogenation of light alkanes.

Resume : Numerous of applications require water wettability of material surface to be precisely controlled. Superhydrophobic surfaces, defined by the water contact angle being over 150 degree, have received rapidly increasing research interest due to their potential application mostly in areas such as self-cleaning and anti-icing. Following our previous studies regarding cerium, neodymium and europium doped Al2O3, in this work we applied spin-polarized density functional theory (DFT) method to analyse the possibility of tuning the wettability of aluminium surface by graphene coating. The computational model which includes long-range van der Waals interactions were used in order to get proper calculations of graphene geometries and properties. The calculations were performed in order to analyse the interaction between graphene layer, Al(111) surface, and ice-Ih bilayer. Moreover, the obtained results were used to characterise influence of substratum as well as functionalization (fluorination) of graphene to the atomic contact angle. Acknowledgements This work was supported by the European Commission under Horizon 2020 Programme and CARIC/NSERC from Canada in the framework of project PHOBIC2ICE (Grant No. 690819). Computing resources were provided by High Performance Computing facilities of the Interdisciplinary Centre for Mathematical and Computational Modeling (ICM) of the University of Warsaw under Grant No. G64-8.

Resume : Sr2FeNbO6 is a double perovskite compound which exhibits antiferromagnetic and semiconductor behaviors, an energy gap of ~2 eV and Néel temperature of ~35 K; that make it optimum to be used as SOFCs’ electrode or photocatalyst for H2 generation [1, 2]. However, there are very few research about this compound, possibly due to the high cost associated with their synthesis process. Nevertheless, recently we obtained the SFNO compound by using the molten salts method, in a six hours process, at 900°C, diminishing drastically the associated synthesis cost. The experimental details has been accepted for publication but, in this work, we report supplementary results on the structural and electronic properties of the SFNO compound, under high pressure, studied in the framework of the of density functional theory. Acknowledgments: This work was partially supported by the projects PAPIIT IN109718 and IPN-SIP-2019-6659. Calculations were done using resources the Supercomputing Center DGTIC-UNAM. J. L. Rosas, J. M. Cervantes and J. A. León F. want to acknowledge the scholarship from CONACYT. References: [1] E. D. Jeong, S.M. Yu, J.Y. Yoon, et al., J. Ceram. Process. Res. 3, 305 (2012). [2] K. K. Pandey, et al., AIP Conf. Proc. 1730, 030029-1 (2015).

Resume : 3D carbon allotropes formed by a carbon nanotubes network can present interesting mechanical and electronic properties. It has been previously shown in the literature that it is possible to grow carbon nanostructures inside the channels of the zeolites. As there is a large number of zeolites, a candidate structure to support the synthesis of 3D carbon nanotubes network could be the beta zeolite, that has its channels interconnected and with same diameter (~5.6Å). Then, we study theoretically from molecular dynamics simulation, the possibility of the existence of one (or more) 3D nanostructures of carbon nanotubes formed inside beta zeolite template, and its resulting mechanical properties. We verified that the carbon nanotube (6,0) can be embedded inside the channel of the beta zeolites keeping the system with the lowest possible energy. From this carbon nanotube, it is possible to built 3D structures with high density (all zeolite channels filled) and lower densities (zeolite channels partially filled). The mechanical properties of the resulting 3D carbon nanotube network with low density indicates mechanical anisotropy and enormous elasticity in the x direction when stretched, supporting more than 100% of strain without fracture. These results suggest that the 3D carbon nanotube network synthesized inside the beta zeolite can be an interesting structure for technological applications.

Resume : Ab initio study of thermoelectric properties of model materials made from the d-metallic nanoclusters is presented. Semiclassical thermoelectric transport coefficients had been obtained from the DFT-calculated band structure. Upto 50 times increasing of the Seebeck coefficient was obtained numerically for nanoclustered materials with respect to the bulk one. It stays in agreement with the experimentally observed trends of nanoclusters properties. Relationship of structural, electronic and thermoelectric properties of nanoclusters and structures formed from it, is discussed in order to developing of highly efficient thermoelectric materials.

Resume : The growth of the human population and the related increase of energy consumption force humanity to find new sources of energies. One of the most promising approach consists of using renewable solar energy sources, e.g. based on photovoltaic panels, which can supplement the limited fossil resources (coal, oil and gas). A semiconductor layer is the heart of every solar panel, which exhibits the photovoltaic effects responsible for the conversion of solar light energy into electricity. One of the newest and most promising photovoltaic solar cells (PVSCs) are based on the perovskite CH3NH3PbI3 material, which have reached high efficiencies of 23% [1, 2]. The main reasons of high conversion efficiency in the perovskite cells are a large absorption coefficient [3], high charge carrier mobility [4], a long electron–hole diffusion length [5] and a direct band-gap of about 1.5 eV being close to the optimal band-gap [6]. Furthermore, perovskite materials have numerous advantages from the social and economical point of views, like low cost of preparation, flexibility, small weight and high transparency [7]. All these points make perovskite PVSCs one of the most promising materials for photovoltaics. However, despite a large amount of investigations on this class of materials, an in-depth understanding of the physical mechanisms lying at the basis of the PVSCs high performance is still to be achieved. In particular, the impact of the orientational disorder of the methylammonium (MA) molecules on the electronic and optical properties still requires clarification. In this poster, we use theoretical calculations coupled with experimental methods to obtain precise information about structural, electronic and optical properties of the CH3NH3PbI3 material. For this purpose, Density Functional Theory (DFT) and Time Dependent DFT (TD-DFT) methods were used to study the perovskite material. A thorough numerical investigation of the possible orientations of the MA molecules within the crystal provided the most stable structures. Then, electronic structure calculations at the TD-DFT level of theory applied to chosen relaxed structures allowed the simulation of the absorption spectrum, which was compared to experimental data.

Resume : Organic light emitting diodes (OLEDs) have seen great improvements since the first prototype in 1987 by Tang and Van Slyke. Today they are widely used in consumer electronics such as smartphones, TVs and general lighting applications. OLEDs in these applications are based on thin films of amorphous organic semiconductors consisting of small organic molecules or polymers exhibiting hopping transport due to the localized nature of the charge carriers. Electrons and holes recombine in emissive layers embedded within multiple charge transport layers. Charge carrier mobility in individual layers can vary by orders of magnitudes depending on the materials, their composition and processing. The main challenge of device optimization is balancing the intricate interdependencies between the individual layers requiring great experimental effort. Here we developed an ab initio multiscale workflow to determine the structure property relations for OLED devices and apply it to single-carrier devices commonly used as charge transport layers. This method, extended to include emissive and injection layers, will allow inexpensive virtual prototyping of OLED devices, complimenting established experimental techniques.

Resume : A number of Metal Chalcogenides are promising candidates for cutting-edge applications[1]. IrTe2 displays peculiar T-induced phase transitions[2] that can be tuned by doping and triggered by photoexcitation[3]. These transitions display hysteresis and they are accompanied by a change in the physical properties of the material, thus making IrTe2 attractive for technological applications. The various phases differ by the length of certain Ir-Ir and Te-Te bonds, that change up to 30% [4]. While a wealth of experimental results exist, the atomic-level understanding of IrTe2 behaviour is still poor. In this contribution, we present an in-depth computational investigation of the electronic structure and lattice dynamics of IrTe2. We demonstrate -through the combined use of several chemical bonding analysis approaches- that the electronic stability of low-T phases is determined by the formation of Ir2Te2 4-centre bonds. The high-T phases, instead, is stabilized by the higher entropic contribution of its phonons. We rationalize the observed hysteresis by analyzing the energetics of phase transitions trough the Nudged Elastic Bands method. Finally, we present our preliminary results on the effect of doping on both the band structures and the phonon spectra of the IrTe2 phases. [1] C. Tan et al. , Adv. Mater., 29, 1701392 (2017). [2] Q. Li et al. , Nat. Comm., 5, 5358 (2014). [3] S. I. Ideta et al., Sci. Adv., 4, eaar3867 (2018). [4] J. Dai et al., Phys. Rev. B, 90, 235121 (2014)

Resume : With the aim of finding an interpretation for the isomerization reaction of icosadeca-ene by quantum methods, we have studied a series of three molecules giving the following results:  The studied segments (C20H22, C20H11F11, C20H11Cl11, C20H11Br11 and C20H11I11) are very stable. This stability is justified by the HOMO-LUMO found energy gap. However, examination of the stability of several conformations shows that the trans conformer is more stable than the cis conformer in the general assembly.  According to the study of different reaction profiles, we noticed that the size and nature of the dopant plays a very important role on the evolution of the activation energy.  From the obtained values of the activation energy, we find that the speed constants of the isomerization reaction are in the order: kC20H22 >> k C20H11F11 >> k C20H11Cl11 >> k C20H11Br11 >> k C20H11I11  The search for intermediate products during the transition Cis-Trans shows that the geometric parameters (angles and dihedral angles) are the most varied settings, this remark has been observed in the case of substituted and non-substituted icosadeca-ene.  The methods of calculations performed in this work are the Ab-initio and DFT methods, with the bases (6-31G, 3-21G **). All these calculations are performed with the Hyperchem software, where parameters obtained are in a closer order to those obtained with the Gaussian 03W software  Examination of different molecules obtained during the Cis-Trans isomerization reaction shows that the total energy of the resulting intermediate product is of the order of -10487.05 eV, corresponding to a 0.87eV activation energy (23.67 kcal / mol).  With the same HF method (6-31G and 3-21 G**), a close geometry was obtained for the intermediate product in the isomerization reaction with a total energy of 0.93 eV (25.30 kcal/mole), which shows that the different values of the activation energy obtained by the HF and DFT methods at the 6-31G level can be compared to those obtained by Ito, Montaner and Bernier. Keywords: Ab-initio; DFT; kinetics; isomerisation; substituted icosadeca-ene

Resume : Negative thermal expansion (NTE) is an interesting property of some materials leading to their lattice contraction upon heating. Recently metal fluorides like scandium fluorine (ScF3) have attracted attention as new class of NTE materials. In the present study the NTE effect in ScF3 was studied in the temperature range from 300 K to 1600 K using ab initio molecular dynamics (AIMD) as implemented in the CP2K code. The simulations were performed in the isothermal-isobaric ensemble for several different supercell sizes (from 2a×2a×2a to 5a×5a×5a) to investigate the stability of AIMD simulation results depending of the supercell size. The information on the temperature dependence of the lattice constant, inter-atomic bond angle distributions and radial distribution functions was obtained. The temperature dependence of the experimental Sc K-edge EXAFS spectra was also simulated based on the multiple-scattering formalism to additionally validate the accuracy of the AIMD method. Our results suggest that AIMD calculations are able to reproduce qualitatively the NTE effect in ScF3 which is attributed to the tilting motion of ScF6 octahedra. Financial support provided by project No. 1.1.1.2/VIAA/l/16/147 (1.1.1.2/16/I/001) under the activity ”Post-doctoral research aid” realized at the Institute of Solid State Physics, University of Latvia is greatly acknowledged.

Resume : A density functional theory (DFT) study of the structural and electronic properties of isolated neutral clusters of the type Au12M, with M = Cu, Ag, or Ir has been performed. Since there is experimental evidence that the gold-iridium [1], gold-copper [2] and gold-silver [3] nanoparticles have an enhanced catalytic activity for the carbon monoxide oxidation reaction, we also performed DFT calculations of the adsorption and dissociation of molecular oxygen on these nanoparticles. Moreover, to understand the effects of Cu, Ag, and Ir impurity atoms on the dissociation of molecular oxygen, we also analyze this reaction in the corresponding pure gold cluster. The results indicate that the substitution of one gold atom in a Au13 cluster by Ag, Cu or Ir diminishes the activation energy barrier for the oxygen molecule dissociation by nearly 1eV. This energy barrier is similar for Au12Ag and Au12Cu, whereas for Au12Ir is even lower [4]. These results suggest that the addition of other transition metal atoms to gold nanoclusters can enhance their catalytic activity towards the CO oxidation reaction, independently of the effect that the substrate could have on supported nanoclusters. This work was supported by UNAM-PAPIIT IN10717. Computations were performed at the supercomputer Miztli of DGTIC-UNAM (Project LANCAD-UNAM-DGTIC-180). A.L.M.V. would like to thank CONACYT and BEIFI-IPN for her scholarship. References [1] A. Gómez-Cortés, G. Díaz, R. Zanella, H. Ramírez, P. Santiago, J.M. Saniger, J. Phys. Chem. C 113, 9710 (2009). [2] A. Sandoval, C. Louis, R. Zanella, Appl. Catal. B 140-141, 363 (2013). [3] A. Sandoval, A. Aguilar, C. Louis, A. Traverse, R. Zanella, J. Catal. 281, 40 (2011). [4] L.M. Jiménez-Díaz, L.A. Pérez, accepted for publication in ‎Eur. Phys. J. D (2018).

Resume : In the austenitic lightweight steels, the nano-size precipitates within the austenitic matrix significantly contribute to the improvement of the yield strength. In the present work, the yield strength was predicted by computational approach. Thermodynamic, diffusion mobility and volume databases for lightweight steels was constructed by using both computational and experimental techniques, including CALPHAD, ab initio calculations and diffusion-couple method. Based on the present databases, the precipitation of the ?-carbide during aging was simulated to obtain the volume fraction and mean particle radius. Combining with the models for the solid solution strengthening, grain boundary strengthening, coherency strengthening, anti-phase boundary strengthening and modulus strengthening, the yield strength is predicted for the austenitic lightweight steel with finely dispersed nanometre-sized ?-carbide. The quantitative relationship between the alloy composition, aging parameters, precipitation behavior and the yield strength was established for lightweight steels in the present work.

Resume : The first-principle study with the use of density functional theory (DFT) was performed to evaluate the electrical properties of the boron-doped diamond (BDD) sensor. In calculations, general gradient approximation exchange-correlation was used for the description of energy functional. In this work optimized norm-conserving, Vanderbilt pseudopotential basis set was used [1]. All methods were implemented in software QuantumATK 2018.06 from Synopsys [2]. Periodic slab models were created with six carbon layers passivated with hydrogen atoms on (111) surface. Two boron atoms were placed within the second carbon layer from the top surface. The results show that ligand-functionalized BDD electrodes expose lower electron density on the plane passing through the top amino acids of the protein complex. The BDD structure donates electrons to ligand and protein complexes, resulting in cross-sectional tunelling. The corresponding pathways revealed a high probability of intermolecular charge transfers, which has demonstrated high diffusion to the electrode, due to the high in-plane surface charge transfer. Analysis of structure with bias potential in the range from 0.0 to 2.0 V exposed that the current transport is higher as the length of the molecule is shorter. Bias potential applied to the electrode showed that both systems have exhibited a minor increase in the current up to a certain threshold, after which a step increase was seen due to resonant transmission. The benzoic group forms a near perfect resonant transmission couple [3], which results in excellent performance of the biosensing electrode. References 1. Garrity, K. F., Bennett, J. W., Rabe, K. M. & Vanderbilt, D. Computational Materials Science 81, (2014), 446–452. 2. QuantumATK Atomic-Scale Modeling for Semiconductor & Materials, (2018) 3. Stokbro, K., Taylor, J., Brandbyge, M., Mozos, J.-L. & Ordejón, P., Computational Materials Science 27, (2003), 151–160.

Resume : Double perovskite compounds could be used in fuel cells, spintronics or information storage devices or Li-ion batteries. This variety of applications is linked to physical and chemical properties which is possible to change if, in the perovskite type compounds, the transition metal atoms are replaced total or partially, generating specific solid solutions. This is the case for the compounds Sr2FeMo1-xNbxO6, for which it has been reported experimental results: the ferromagnetic and the half-metallic behaviors shown by Sr2FeMoO6 change, when Nb is incorporated instead of Mo, for x≈0.7. There are enough theoretical and experimental information about the electronic, magnetic and structural properties of Sr2FeMoO6; however, theoretical studies for Sr2FeMo1-xNbxO6 are scarce. In this work we present the results from a first principles study, in the framework of the density functional theory, through the exchange-correlation functional CA-PZ within LDA+U as it is implemented in the CASTEP code. The semiconductor and antiferromagnetic behaviors of Sr2FeNbO6 were reproduced, as well as the corresponding half-metallic and ferromagnetic for Sr2FeMoO6, then the solid solution Sr2FeMo1-xNbxO6 is studied by means of a supercell model for x = 0, 0.25, 0.5, 0.75 and 1. Acknowledgments: This work was partially supported by the projects DGAPA-UNAM IN109718, IA106117 and IPN-SIP-2019-6659. J. L. Rosas, J. M. Cervantes and J. A. León and acknowledge the scholarship from CONACYT.

Resume : Reaction mechanism of the selective catalytic reduction of nitric oxide (NO) by ammonia (SCR-NH3 of NO) on the Ru doped CeO2 surface was investigated using density functional theory calculation corrected by on-site Coulomb interactions (DFT+U). The SCR-NH3 of NO mechanisms on Ru-CeO2, consisted of two consecutive reduction pathways, were systematically examined. The calculated results reveals that the Ru dopant affects on the electronic charge property and enhances the Lewis acidity of the CeO2 (111) surface. NH3 adsorption/dissociation and NO reduction via the NHNO intermediate are facile when the Ru site presents on the catalyst surface. The second NO reduction, which occurs over the presence of the Brønsted acid site, is slightly difficult in terms of NH3 adsorption and dissociation. For the overall reactions, the water formation is the rate determining step. Our results revealed that the Ru dopant greatly enhances the efficiency of the SCR-NH3 of NO reaction of the CeO2 catalyst. The performance of this catalyst can be further enhanced by improving the water formation aspect. The obtained results deepen the fundamental understanding the role of the different active sites on the crucial steps during the reaction.

Resume : In the last few decades, many types of new materials have been proposed and sometimes their synthesis was experimentally achieved. Relevant examples are new allotropic forms of carbon like Graphenylene, which was recently synthesized, or other interesting low dimensional materials, such as Porous Graphene. Both are of great technological interest due to their special mechanical and electronic properties. In our investigation, we used Density Functional Theory and Molecular Dynamics methods to investigate the possibility of tuning the mechanical and electronic properties of Porous Graphene and Graphenylene using different configurations of adsorbed molecules on these materials. Our calculations show that mechanical properties such as the Young modulus, Poisson ratio and the bulk modulus can be modified as well as some electronic features, like the band gap openings, the local density of states and electronic transport properties. For this regard, in the calculations presented in this work we considered the effect of small molecules (e.g. CO2 and O2) or individual atoms, as H and F, adsorbed on graphenylene and Porous Graphene membranes as well as the effect of metallic atoms, such as Co and Fe on the electronic and mechanical properties of those two materials.

Resume : Macrocyclic dyes such as phthalocyanine and porphyrin molecules are well-known harvesting compounds for dye-sensitized solar cells (DSSC). A theoretical DFT approach was used to study the effects of orientation of the dyes on (10-10) wurtzite surface as well as the role of adsorbate-induced surface dipoles on the dye-semiconductor electron transfer (ET). Tilting of the dyes has a dual effect on the photoinduced ET. Shortening of the distance between moieties enhances the tunneling probability for the electrons due to the stronger coupling between the conduction band minimum and LUMO of the dye. By contrast, strengthening of the electronic interactions upon the tilting may significantly affect the relative position of electronic levels and eventually hinder ET. Furthermore, linker of the molecule provides an intermediate state which may affect the hole transfer from the semiconductor to the dye. The peripheral substituents both reduce aggregation effects and shift the energy position of dye’s frontier orbitals that provides alternative ways for excited state relaxation. In addition, various adsorbates on wurtzite surface induce dipole moments, which may facilitate or hinder the ET. By combining spectroscopy studies with DFT methods we deduce that motion of the dye linked to the surface can be in control of the ET at few tens of ps or longer time. We conclude that proper choice of quantum dot ligands and the peripheral substituents can facilitate the photoinduced ET in DSSC.

Resume : Quantum wells (QW), quantum steps (QS), and quantum dots are usually assumed to have a rectangular potential profiles. We examine the potential profiles of a narrow QW and determine its effective width. We also study the charge transport via molecular dynamics (MD) simulations. We take the single and double dipole layers as electrostatic models for QS and QW. In addition the molecular dynamics (MD) is performed to study transport properties. As a practical example, we study the dipole layers formed by the ions in the GaN semiconductor. We show that the potential profiles are not rectangular. A quantum step is a continuous change of potential over a distance larger than one monolayer. Our approach provides insight into the relation between the potential profile and the positions of the ions in the crystal planes. Remarkably, the minimal thickness of the QS is not determined by the distance between the charged planes, but by the lateral spacing between ions. Due to the smooth potential profiles of the narrow QWs, the energy levels of localized states are shifted in comparison to a rectangular QWs. In the example of GaN, QS cannot be thinner than about 3 Å.

Resume : Mathematical modeling makes it possible to reduce time and cost for the development of complex technological processes. The paper presents a complex study of the metallurgical-grade silicon purification technology by vacuum refining and plasma-chemical purification under the conditions of electromagnetic stirring of silicon melt by mathematical modeling. The paper presents the numerical models of the processes, using COMSOL Multiphysics software. The initial and boundary conditions of these processes are formed with HSC Chemistry software. To study vacuum refining under the conditions of electromagnetic stirring, mathematical models including Maxwell equations, hydrodynamics equations for incompressible liquid, mass transfer equations, including directional mass transfer and diffusion for the impurity transportation calculation inside silicon melt are used. For mathematical modeling of plasma-chemical silicon purification, HSC Chemistry software is used. For the thermodynamic estimation of the chemical processes, the change of the Gibbs free energy has been calculated. The research has shown the efficiency of using mathematical modeling for the study of complex technological processes, especially, the technology for purification of metallurgical-grade silicon to solar-grade silicon level using vacuum refining and plasma-chemical purification of silicon melt under the conditions of electromagnetic stirring.

Resume : Mathematical modeling makes it possible to reduce time and cost for the development of complex technological processes. The paper presents a complex study of the metallurgical-grade silicon purification technology by vacuum refining and plasma-chemical purification under the conditions of electromagnetic stirring of silicon melt by mathematical modeling. The paper presents the numerical models of the processes, using COMSOL Multiphysics software. The initial and boundary conditions of these processes are formed with HSC Chemistry software. To study vacuum refining under the conditions of electromagnetic stirring, mathematical models including Maxwell equations, hydrodynamics equations for incompressible liquid, mass transfer equations, including directional mass transfer and diffusion for the impurity transportation calculation inside silicon melt are used. For mathematical modeling of plasma-chemical silicon purification, HSC Chemistry software is used. For the thermodynamic estimation of the chemical processes, the change of the Gibbs free energy has been calculated. The research has shown the efficiency of using mathematical modeling for the study of complex technological processes, especially, the technology for purification of metallurgical-grade silicon to solar-grade silicon level using vacuum refining and plasma-chemical purification of silicon melt under the conditions of electromagnetic stirring.

Resume : Currently, multicrystalline silicon (mc-Si) is the main material for solar cell production. Constant process to the improvement of solar cell specifications requires better crystals quality. Multicrystalline silicon ingots manufactured by directional solidification (DS) contain peripheral areas with contaminations that diffuse during crystallization from the crucible to the ingot. The availability of areas with contamination in silicon wafers results in production of solar cells with lower efficiency. For the registration of these areas, the silicon wafer photoluminescence (PL) method is used, since these areas give a signal with reduced average PL intensity. To control diffusion from the crucible, silicon nitride (Si3N4) protective diffusion coatings are used. The protective coating efficiency depends on its composition, thickness and the layer stability. In the paper, the data science as the most demonstrative and information intensive approach for studying contaminated areas in the ingot volume is used. The method of 3D visualization of recombination active defects with a large number of PL images of mc-Si wafers is applied. The analysis of the effect of using protective coating of different composition and thickness on the area of contaminated zones has been investigated.

Resume : FeSe is unique among other iron-based superconductors, which exhibits no magnetic ordering and becomes superconducting below 8 K in the undoped bulk system while the superconducting transition temperature soars by an order of magnitude for a monolayer FeSe on SrTiO3 substrate. Using a DFT+DMFT framework, we perform a comparative study on the magnetic properties of FeSe systems and LaFeAsO, another representative iron-based superconducting material. Calculated magnetic moment in the typical stripe-type magnetic ordering pattern is finite for LaFeAsO while negligible for bulk FeSe, consistently with experiments, and is also finite for monolayer FeSe suggesting the magnetic order is restored in monolayer FeSe. We suggest a mechanism explaining why the systems with the similar magnitudes of fluctuating spin (S2) have very different magnitude of ordered moment (Sz) focusing on the different aspects of local charge fluctuation. Our work provides a comprehensive understanding of magnetism in iron-based superconducting materials, and also emphasizes on the potential importance of magnetism in the high superconducting transition temperature of monolayer FeSe on SrTiO3 substrate.

Resume : Nowadays the great demand for ecologically appreciable ways of creating electricity is observed worldwide. One of the possible solution options is closely linked with creation and commercialization of intermediate temperature solid oxide fuel cells (IT SOFCs) – electrochemical devices capable for efficient energy production without sufficient damage to the environment. The development of the respective materials for various IT SOFC components is a complex, nontrivial task requiring combined theoretical and experimental approach. In particular, design of good SOFC cathode acquires of great importance. Cobaltites containing Ba and Pr are among the best possible candidates, however, their tolerance to carbon dioxide is not high enough. It was already shown that doping with highly charged cations, i.e. Ta5+ or Nb5+, can sufficiently improve cathodes’ working stability limit. This work is aimed at finding out how Nb and Ta doping can influence on PrBaCo2O6–δ oxide’s crystal and electronic structure, defect formation peculiarities and physical properties. The main focus of this study was shifted towards ab initio calculations, however, experimental methods including X-Ray and electron diffraction were also utilized. One of the main findings in this study is the discovery of the different cationic orderings for Nb and Ta doped samples observed experimentally and validated theoretically.

Resume : We show that a stacking defect or a shift has a striking effect on carrier transport of bilayer graphene. Charge transport through a ballistic n–neutral–n junction of shifted bilayer graphene may result in minimal conductivity and shot noise anomalies which are found to be sensitive to the shift defect. Minimum conductivity and shot noise in shifted bilayer graphene exhibit an anisotropic beahavior as a function of the orientation of the electrodes. The minimum conductivity could be suppressed for somme specific value of twist defect while the shot noise takes the unit value. Our results provide a way to control transport properties in an undoped graphene bilayer structure by adjusting the layer stacking.

Resume : Surface modification of nanostructures with functional moieties can substantially alter the surface charge, functionality, and reactivity of the surface to extend their usage in practical applications. In the present study, the numerical modeling is employed to examine the field emission properties of graphene-CNT (g-CNT) hybrids via hydrofluoric (HF) acid functionalization. The reaction kinetics of HF acid with acetylene as a precursor gas in an argon plasma medium is considered for the growth of functionalized g-CNT hybrids. The model includes the charging of hybrid surface, the kinetics of the plasma species (neutrals, ions, and electrons), and various processes exclusive to a plasma-exposed surface such as adsorption, thermal dissociation, ion-induced dissociation, and interaction between neutral species among others. The modeling results show that HF acid functionalization remarkably enhances the field enhancement factor of g-CNT hybrids. The enhancement in the electron emission characteristics is attributed to the chemical modification of the carbon network on the hybrid surface. The HF acid acts as an etchant to remove the oxygen atom and replacing the carbon-oxygen (C–O) molecules with carbon-fluorine(C-F) bond. The results of the work can be extended to improve the field emission from g-CNT hybrids. References [1]A. Tewari, S. Ghosh, and P. Srivastava, Phys. Plasmas 25, 043503(2018). [2] A. Tewari, P. Srivastava and S. Ghosh, Phys. Plasmas 25, 083520(2018).

Resume : Developing novel energetic materials with high detonation performance and low sensitivity is one of the hottest concerns in the field of explosive research. By employing quantum mechanical tools, a series of energetic ionic salts (EIS) based on 5,5'-bitetrazole-1,1'diolate (BTO) such as ABTOX, DMA-BTO, DU-BTO, HA-BTO, M2-BTO, TKX-50 were thoroughly investigated to obtain more comprehensive insight into structure-property-performance interrelationship. The physicochemical and detonation characteristics of these EIS including structural, electronic, vibrational and performance parameters (heats of formation, detonation pressures, and detonation velocities) were discussed in detail. The D-H stretching vibrations of hydrazine cations are located at much lower frequency followed by hydroxylammonium and their corresponding shorter D...A contacts represent the presence of strong hydrogen bonding environments in the crystal structures of HA-BTO and TKX-50. Careful inspection of various EIS revealed that the hydroxylammonium and hydrazine cations produce the highest density relative to other cations when combined with the BTO anion. The calculated detonation characteristics demonstrate these BTO salts as potential explosives, especially, HA-BTO and TKX-50 provide superior detonation pressure (38.85 and 40.23 GPa) and detonation velocities (9.94 and 9.91 km/s), superior to those of traditional nitrogen-rich energetic materials with moderate sensitivities. Therefore, in this study, we propose that enhancing hydrogen bonding interactions is the main strategy in designing high energy density materials for next-generation explosives, propellants, and pyrotechnics.

Resume : Models of thermodynamic properties and phase transformations of materials are necessary for simulations of different processes at high energy densities. In this work, an equation-of-state model for metals is presented with taking into account melting and evaporation effects. Equation of state for molybdenum is developed. As distinct from the previously obtained equation of state for the metal [1], a new form of functional expression of thermodynamic potential terms is proposed. These terms provide for more correct thermal contributions of atoms in the solid and liquid phases as well as the phase changes effects. A critical analysis of obtained results is made in comparison with available experimental data at high temperatures and pressures. [1] V. E. Fortov, K. V. Khishchenko, P. R. Levashov and I. V. Lomonosov, Nucl. Instr. Meth. Phys. Res. A 415, 604 (1998).

Resume : PrBa0.5Sr0.5Co1.5Fe0.5O5+δ (PBSCF), a mixed ionic-electronic conductor (MIEC), is a potential cathode of proton-conducting ceramic fuel cells because it shows high surface reaction with oxygen and its diffusivity into bulk [1]. Pr, Ba and Sr go to A site, and Co and Fe to B site in the ABO3 perovskite structure. Pr is known to be separated from Ba and Sr, and they form two different A site layers, A' and A'' respectively [2]. Oxygen vacancies are also known to form in the A' layer and Co adjacent to the oxygen vacancies. We used genetic algorithm (GA) to study energetically favorable distribution of atoms in PBSCF [3]. Separation of Pr from Ba and Sr was first confirmed by GA at a 2x2x2 supercell, and the rule was employed for 4x2x2 and 4x4x2 supercells; this substitution region restriction method reduces the number of configurations for atom distribution, accelerating the GA search further. We will discuss the distribution of Co and Fe atoms in B site. References [1] S. Choi, et al., Sci. Rep., 2013, 3, 2426. [2] A. A. Taskin, et al., Prog. Solid State Chem., 2007, 35, 481. [3] J. Kim, et al., Comput. Mat. Sci., 2017, 138, 219.

Resume : The production of solar fuels consists of harvesting sunlight as an energy source and transforming it in storable fuel form. One feasible way of doing this is by splitting water. After first example of photoelectrical system , the idea of using dye-sensitized semiconductor electrodes, linked to the water oxidation catalyst was embraced by the community. The interface between electrode and photosensitive component is notoriously complex system as it involves processes of light absorption, charge separation and charge transport. The advantage of this strategy of separating the semiconductor from the photo-sensitive part mimics the process of natural photosynthesis , separating different functions into different parts of the device. Efficiency of these type of cells heavily depends on dye’s characteristics and affinity towards fast electron injection into a semiconductor electrode with low charge recombination. For classes of molecules like naphthalene diimides it is known that properties like HOMO-LUMO energies and their relative gaps are easily tuned by choosing appropriate functional groups . Screening large number of molecules through automated procedure and usage of fast theoretical methods, will provide the molecule with most promising optical and redox properties. These properties are explored with the Amsterdam Density Functional (ADF) code at DFT and DFTB level and also time dependent DFT (TD-DFT and TD-DFTB).

Resume : DFT calculations, carried out with different basis sets, for the static longitudinal linear polarizability, αL, and second order hyperpolarizability, γL, of small doubly charged hexadecaocta-ene chains, are presented. To demonstrate the stability of several conformations studies of different reaction profiles, determination of rate constants of isomerization reactions and activation energies were made. A theoretical study of a new range based molecules substituted hexadecaocta-ene having the structure D-Π-A by the methods DFT, AM1, PM3 and PM6 finally we can highlight: • The effectiveness of the methods used to calculate the polarizabilities αL, hyperpolarizabilities γL and their components. • The correlations found between the various physical parameters and electronic substituted hexadecaocta-ene and its value exceeds 0.88 average polarizability allowing us and precise quantitative determination of the polarizability for systems with more than 80 atoms by simple extrapolation. • The PM6 method is an effective alternative methods CPHF / TDHF, DFT for the calculation of polarizabilities. The results obtained for positive and negative chains show that the ionization state effect decreases more rapidly, as the chain length is increased, for αL than for γL. Or both types of charged chains, the incorporation of the electron correlation increase the αL, and γL values, as compared to the DFT values. A comparison between the results obtained using the standard 6-31G basis set and augmented versions of this set, obtained by the addition of diffuse and polarization functions, shows that 6-31G basis set does not provide a good description of the negative chains studied here and that the addition of extra diffuse functions on the basis set is needed in order to obtain reliable estimates for polarizabilities, specially for γL. Key words : polarizabilities; hyperpolarisability; hexadecaocta-ene; DFT, AM1, PM3 , PM6

Resume : One promising method to achieve higher solar cell efficiencies is to replace fullerenes with strongly-absorbing dye molecules, namely, nonfullerene acceptors (NFAs). However, the systematic design of acceptor molecules with tailored properties has yet to be demonstrated. The main difficulty here is that, while fullerenes are electrostatically inert, new acceptor molecules typically have strong static quadrupole moments. By exploring the long-range electrostatic interaction at the interface, we demonstrate that, for a set of recently developed NFAs, the electrostatic bias potential can be directly related to the stabilization (or destabilization) of charge-transfer (CT) states as well as changes of the photovoltaic gap. We find that the correlation between quadrupole moments, charge separation efficiency, and CT-state energy predicted by our model, is experimentally reproduced for several different donor /acceptor combinations. This allows us to predict new NFA structures using combinations of readily available molecular building blocks that can potentially reach even higher performances than those currently achieved in stateof-the-art NFA devices.

Resume : Properties of ABO3 perovskite surfaces are strongly dependent not only on their perfectness, but also on reconstruction, defects, as well on the polarity, termination (AO or BO2), and the temperature (because of the both bulk and surface-related phase transitions). In this study the surface phase transitions on SrTiO3 (001) and related surfaces are simulated for the first time at the ab initio level using the state of the art DFT hybrid approach (HSE06 Hamiltonian), with accuracy much higher than in any previous theoretical studies. Raman frequencies for surface defects and their liminescence properties have been calculated. Predicted tiny surface reconstruction due to phase transitions and vibrational defect properties can be checked with RHEED, Raman and CL measurements. Understanding and control of the surface-related properties of advanced SrTiO3 perovskites is important for applications in microelectronics and catalysis. Funding from Latvian-Ukraine biletaral project is greatly acknowledged.

Resume : The new solar cell architecture is relatively complex and requires a thorough understanding of the materials behavior, involving accurate knowledge of the structural, morphological, optical and electrical properties of these materials. In this work, we have used the Silvaco-Atlas software to optimize a specified conception of the FTO/Co-ZnO/CuO hetero-junction structure by taking account of the materials parameters. Firstly, the CuO solar cell and its layer structure were prepared by spray pyrolysis method on glass substrates in order to characterize physical material parameters using Raman spectroscopy, Transmission electron microscopy, Scanning electron microscopy, and Hall Effect measurements. Secondly, theobtained physical parameters were used in simulation. Thickness and carrier concentration effects of the cell structure were studied to optimize the solar cell performances.The combination of the optimum parameters enables us to optimize and to obtain the efficiency of 24.55 % which is improved with 8.88 % compared to the reference one which is the high record conversion efficiency found experimentally in the CuO solar cell[1].

Resume : Vertically aligned carbon nanotube (VACNT) arrays or “forests” behave mechanically as foams when compressed, exhibiting a characteristic nonlinear stress e strain response. However, the fiber structure of VACNT forests is unlike that of cellular foams, and the microscopic mechanisms of the deformation are quite different. While numerous studies have addressed the mechanical response of VACNT forests undergoing uniaxial compression, the underlying deformation mechanisms are not yet fully established. In this presentation, we report the results of large-scale mesoscopic simulations of several structurally distinct VACNT forests subjected to compressive deformation. The simulations reveal that the deformation proceeds as a phase transformation from an original low-density phase composed of vertically aligned CNT bundles to a densified phase with horizontal alignment of CNTs. The two phases are separated by a well-defined interfacial layer, which advances during the compressive deformation through localized bending and folding of nanotubes. The characteristic three-stage stress-strain dependence (an "elastic" peak followed by an extended plateau region and a sharp rise of stress in the densification regime), commonly observed in experimental probing of the mechanical properties of VACNT forests, is reproduced in all of the simulations, suggesting that the heterogeneous propagation of densification front is the general mechanism of mechanical deformation of VACNT forests.

Resume : The search for new photovoltaic (PV) materials by high throughput screening among crystalline materials databases - experimental or computed - is generally based on selection criteria such as optical absorption bandgap energy, Eg and majority carrier transport properties (n- or p- semiconductor type and mobility or effective mass). Unfortunately, a selection criterion is missing regarding minority carrier transport properties, especially their lifetime, that is not easily accessible from computation. Starting from the fact that the best performing photovoltaic materials (with >15% light to electricity conversion efficiency) are all crystals with perfect cubic structures, we first discuss on the possible links between photovoltaic performance, minority carrier lifetime and deviation from perfect cubic structure. Depending on the crystalline structure, cubicity can be defined as e.g. the Goldschmidt tolerance factor for perovskites or as the tetragonal distortion for Stannites, etc. Such cubicity criterion would thus be easy to derive from available computed crystalline structures, and very selective to possibly discover new absorbers for PV, LEDs or photocatalytic devices.

Resume : In the semiconductor industries, it is necessary to develop a probe card to quickly test as many chips as possible at on the entire wafer on the time. Since a large amount of switches are not able to be mounted on the limited PCB area of probe card, the logic design of STA(Samsung Test Application) which contains 64 switches has been developed. As the automotive semiconductor products and the next generation STA require more and more switches in probe cards to shorten test times, a new format of program coding and sequences are needed. The existing coding methods require a lot of effort when detecting errors with hard coding. With over 70000 coding lines and FPGA (Field Programmable Gate Array)LUT(Look Up Table) usage, the configuration time and errors increase. This paper reports on the efficient FPGA logic to solve the problems caused by increased number of probe card switches due to multi-parallel test. We developed the verilog coding and program sequence firmware using standardization of ROM data by utilizing block memory from xilinx. We found FPGA logic utilization & configuration time reduced as the data structure becomes simple due to the use of ROM data, and it can be expected to reduce human error by coding automation. in addition, the switch configuration becomes free and scalable when designing a probe card

Resume : Density functional theory (DFT) has become an important tool for atomic scale understanding of materials. Recently, PrBa0.5Sr0.5Co1.5Fe0.5O5-δ is receiving high attention because it shows mixed ionic and electronic conduction [1, 2]. Studying this material by DFT, however, is a big challenge because the supercell should be a reasonable size to consider the five metal elements, and there are many different atomic configurations to compute. We employed a deep neural network to obtain a neural network potential (NNP) and, therefore, to reduce the computation time by several orders of magnitude. Energy data from 2,000 atomic configurations was first obtained by DFT, and the data was enlarged to 300,000 by considering crystal symmetry, such as translation and rotation. The data was used to train and test the NNP. We will further discuss the accuracy of the NNP and ways to improve it. References [1] S. Choi, et al., Sci. Rep., 2013, 3, 2426. [2] S. Choi, et al., Nature Energy, 2018.

Resume : In this work, we used spin-polarized, hybrid density functional theory method to investigate the electronic structure and magneto-optical properties of various hydrogen-vacancy clusters in a diamond. Our theoretical results indicate very strong tendency toward the formation of HnV complexes up to 4 hydrogen atoms that are mostly electrically and optically active centres. One of the investigated defects introduce highly correlated electronic states that pose a challenge for density functional theory and therefore, require the special treatment when charge- and spin-density related properties are determined. We introduced an of extended Hubbard-like model of Hamiltonian with fully ab initio provided parameters to analyse the complex electronic structure of highly correlated H2V(0) defect. The role of quantum tunnelling of hydrogen in HV centre and its impact on the hyperfine structure was discussed. We demonstrate that experimentally observed HV(1-) centre is similar to well-known NV(1-), i.e. I) it possesses triplet 3A ground state and 3E excited state in C3v symmetry; II) the calculated zero-phonon line is 1.71 eV (1.945 eV for NV1-). A detailed experimental reinvestigation based on optically detected electron paramagnetic resonance spectroscopy is suggested to verify whether the HV1- centre has metastable singlet shelving states between the ground and excited state triplets and as a result, whether it may exhibit a spin-selective decay to the ground state. Acknowledgments This research was financially supported by the Polish National Science Centre under contract no. 2012/05/E/ST8/03104. Computing resources were provided by High Performance Computing facilities of the Interdisciplinary Centre for Mathematical and Computational Modeling (ICM) of the University of Warsaw under Grant No. GB69-32.

Resume : In this study, we used quantum chemistry calculations in order to determine some kinetic parameters of the isomerization reaction of the substituted tetradecahepta-ene. The studied molecules are: (C14H16, C14H8F8, C14H8Cl8, C14H8Br8 et C14H8I8) Cis and Trans. One of the adopted ways to access these parameters (activation energy, rate constant, etc ...) is looking for the transition state that is based on the exploration of intermediaries during the passage of Cis-Trans isomerization process. The study of a ten molecules series gives the following results:  The Trans conformer is more stable than the Cis. The activation energy changes very greatly depending on the size and nature of the substituent according to the reaction profile.  The constants of the isomerization reaction rates are in the following order: kC14H16 >> k C14H8F8 >> k C14H8Cl8 >>k C14H8Br8 >> k C14H8I8. The geometrical parameters vary considerably according to intermediate products The calculation methods are DFT and Ab-initio methods at STO-3G (d,p) levels Keywords: substituted tetradecahepta-ene, kinetics; isomerisation, Ab-initio, DFT

Resume : Localized corrosion is one of the complex forms of corrosion which makes it difficult to detect and design-against. During metal corrosion in the corrosive environment, corrosion products are also formed as a result of electrochemical reactions inside the electrolyte. These products can precipitate on the corroding surface and stop/ slowdown the overall corrosion process. In order to understand this complex phenomenon, a multi-phase field model is proposed to simulate metal corrosion with corrosion products formation. The free energy of the system is described in terms of its metal ion concentration and the order parameters. Rather than considering linear kinetics (Allen-Cahn equation), inspired from classical rate theory, non-linear kinetics (Butler-Volmer) is considered to describe the temporal evolution of order parameters. The time dependent evolution of ionic species is governed by Nernst-Plank equations while electrostatic potential is governed by Poisson equation. The proposed model results are compared with the experimental findings and several examples are presented to show the practical applications of this model. These examples also suggest that the proposed model is, in general, not limited to corrosion process but can be a good tool for other complex phase change processes. KEYWORDS: pitting corrosion, multi-phase field model and Butler-Volmer.

Resume : For powder library obtained from high-throughput synthesis systems, we have developed high-speed evaluation tools for database creation. We are thinking that crystallographic data such as not only crystalline phases but also atomic coordinates and so on are necessary for prediction of new materials using machine learning. We developed a tool to efficiently measure synchrotron radiation XRD and XAFS measurements of powder libraries. Generally, in synchrotron radiation XRD powder must be filled into a thin capillary. However, our developed tool only adheres to the tape. The powder library fixed to the tape can be continuously moved to the X-ray irradiation position by the tool. By swinging the tool around the irradiation position, diffraction data close to the conventional method could be obtained. Acknowledgements: These XRD and XAFS experiments were conducted at the BL5S1 and BL5S2 of Aichi Synchrotron Radiation Center, Aichi Science & Technology Foundation, Aichi, Japan (Approval No.2017P0202).

Resume : Resistive switching devices are promising candidates for fast computer memories. Despite the simple structure of memristive elements the physical mechanisms behind the switching process are complex [1]. The kinetic Monte Carlo (kMC) method is an established model to simulate filament formation in memristive devices [2]. We present a kMC simulation to investigate the formation of silver (Ag) filaments in titanium dioxide (TiO2). We analyze under which conditions filaments form, how long the formation takes and compare the results with experimental data. The oxidation of Ag atoms at the anode, their migration through the TiO2 layer and the reduction of Ag+ ions at the cathode or in proximity to Ag atoms are modeled using an Arrhenius law. The activation energy ions have to overcome when hopping from one site to another is assumed to have a Gaussian disorder. Introducing a correlated disorder model leads to preferred pathways for Ag+ ions through the TiO2 layer and results in reduced filament formation times and allows filament formation at lower fields. [1] S. Dirkmann, et. al., J. Appl. Phys., 2015, 118, 214501 [2] S. Aldana, et. al., J. Phys. D: Appl. Phys., 2017, 50, 335103

Resume : While the quasi-Fermi level (or imref) is a standard concept in semiconductor physics employed to describe the finite-bias non-equilibrium operations of electronic and optoelectronic devices, its first-principles determination has not been achieved previously. Herein, based on the multi-space constrained-search density functional theory (MS-DFT) formalism we have recently developed [1], we calculate the non-equilibrium electronic structure of molecular junctions under a finite bias voltage and extract the quasi-Fermi level profiles across molecular device systems. Comparing with the conventional non-equilibrium Green’s function (NEGF) calculations based on DFT, we first confirm the practical equivalence between MS-DFT and DFT-NEGF. An important feature of MS-DFT that differentiates it from DFT-NEGF is that it relies on the determination of the quasi-Fermi level or electrochemical potentials across the channel, which are not explicitly provided within DFT-NEGF unlike their electrostatic potential counterparts. Analyzing the spatial profiles of electrochemical potentials of molecular junctions based on different electrode-molecule contact geometries [2] at varying bias voltages, we extract important insights into the nature of nonequilibrium quantum transport at the nanoscale. References [1] H. S. Kim and Y.-H. Kim, arXiv:1808.03608 [cond-mat.mes-hall]. [2] Y.-H. Kim, H. S. Kim, J. Lee, M. Tsutsui, T. Kawai, J. Am. Chem. Soc. 139, 24, 8286-8294 (2017).

Resume : Mathematical modeling makes it possible to reduce time and cost for the development of complex technological processes. The paper presents a complex study of the metallurgical-grade silicon purification technology by vacuum refining and plasma-chemical purification under the conditions of electromagnetic stirring of silicon melt by mathematical modeling. The paper presents the numerical models of the processes, using COMSOL Multiphysics software. The initial and boundary conditions of these processes are formed with HSC Chemistry software. To study vacuum refining under the conditions of electromagnetic stirring, mathematical models including Maxwell equations, hydrodynamics equations for incompressible liquid, mass transfer equations, including directional mass transfer and diffusion for the impurity transportation calculation inside silicon melt are used. For mathematical modeling of plasma-chemical silicon purification, HSC Chemistry software is used. For the thermodynamic estimation of the chemical processes, the change of the Gibbs free energy has been calculated. The research has shown the efficiency of using mathematical modeling for the study of complex technological processes, especially, the technology for purification of metallurgical-grade silicon to solar-grade silicon level using vacuum refining and plasma-chemical purification of silicon melt under the conditions of electromagnetic stirring.

Resume : Nano- and ultrafinegrained structure in metals after severe plastic deformation is formed as a result of movement, accumulation, and interaction of lattice defects. After such kind of metal processing the defects concentration can be estimated by analysis of stored energy which is necessary for metal grain breaking. A method showing good results is a thermal gravimetric analysis (TG), applied together with differential thermal analysis and differential scanning calorimetry (DSC). DSC has been used to fix a temperature difference, which is in proportion to a difference in a heat flow between a reference and a sample in another crucible from the same material. TG has been used to measure changes in the specimen weight, depending on temperature at specific controlled conditions. Preliminary results showed that after combined deformational processing by drawing with bending and torsion the ultrafine grained structure was formed. On the DSC curve obtained after different kinds of plastic deformational processing the single isothermal peak in the temperature range 400-500 C was observed. With increasing of deformational processing degree the temperature of the peak formation decreased whereas the peak square, as a characteristic of stored energy, increased approximately in 1.3 times. It was connected with change of internal stresses level which increased with dislocation density growth.

Resume : Singlet Fission is one of the possible outcomes occurring upon photoexcitation of a material in its aggregated states, where the generated singlet excited state can split in two entangled triplet states showing an overall singlet multiplicity [1]. However, only a few molecules so far have shown significant singlet fission activity [2], thus the mechanism of the process is still unclear and extracting design rules is still a challenge. Additionally, the interest in this process is also due to the possibility to generate two excitons from a photon, which is clearly beneficial for photovoltaic applications. The requirements for a material to give singlet fission are energetic in primis [3], due to the fact that, to be able to generate two triplet states from a singlet, the energy of the triplet state should be about half of the energy of the singlet state. In this contribution we present the results of a (TD)DFT protocol to screen for singlet fission materials [4]. The protocol consists in evaluating the energies of the first singlet/triplet excited states through standard computational procedures. We search publicly available databases (~1 Million materials) for potential Organic Semiconductors (OSCs) based on their computed HOMO-LUMO gap. We do this by estimating the gap from low level calculations. Once potential OSCs have been identified, we apply the protocol to evaluate whether they are promising singlet fission materials. We show the robustness of our results by re-discovering materials that are known to be singlet fission active. Additionally, we propose a high number of new singlet fission candidates that can be experimentally studied to verify the results of our predictions, and to obtain a more general view on the process to study the mechanism and extract design rules. References [1] M. B. Smith, J. Michl, Chem. Rev., 110, 6891 (2010). [2] I. Paci, J. C. Johnson, X. Chen, G. Rana, D. Popovic, D. E. David, A. J. Nozik, M. A. Ratner, J. Michl, J. Am. Chem. Soc., 128, 16546 (2006). [3] D. Casanova, Chem. Rev., 118, 7164 (2018). [4] R. Grotjahn, T. M. Maier, J. Michl, M. Kaupp, J. Chem. Theory Comput., 13, 4984 (2017).

Resume : Although the phosphorescence and, to a lesser extent, delayed fluorescence of polycyclic aromatic hydrocarbons in host matrices has been studied in the middle of the last century at low as well as at room temperature, this class of materials is notably absent from the recent surge of interest in room-temperature phosphorescence (RTP) and delayed fluorescence (DF) from single-component crystalline organic materials. Recently investigated metal-free RTP and DF materials either contain heavy atoms such as bromine to facilitate singlet-triplet intersystem crossing (ISC), or polarizing heteroatoms such as oxygen or nitrogen that induce charge-transfer states, but pure hydrocarbons are absent and are even commonly assumed to not present any of these properties. Electronic transitions between non-coplanar (i.e. orthogonal or tilted) orbitals are necessary to allow for spin-orbit coupling (SOC) and thus present the aforementioned properties, however the vast majority of polycyclic aromatic compounds do not present this kind of structure. Here we describe the successful synthesis of a hydrocarbon named homotruexene and make preliminary studies on its persistent room-temperature phosphorescence and concomitant delayed fluorescence in both powder and film forms, presenting unusually high quantum yields. We study this material with both experiments and theory, using our newly developed method for computing SOC matrix elements and excited states dynamics implemented on the free software ORCA[1-2]. Here we show that it is indeed possible to explain and, to some extent, even predict these unusual properties based on theoretical grounds and detail the emission mechanisms. [1] On the theoretical prediction of fluorescence rates from first principles using the path integral approach. J. Chem. Phys. 148, 034104 (2018). [2] Submitted to J. Chem. Theory. Comp.

Resume : Phosphorus-doped n-type diamond is currently one of the most fundamental wide bandgap materials for next generation high-power electronics and optoelectronics. Artificial diamond growth methods such as CVD involve hydrogen-containing precursors. Therefore, the hydrogen atoms can be simultaneously introduced into the diamond lattice as a contamination and form complexes with other defects. In this work, we used the spin-polarized hybrid density functional theory method to investigate the electronic structure, stability, and magnetic and optical properties of phosphorus, vacancy, and hydrogen clusters in diamond. Our results indicate a thermodynamic driving force for the formation of previously unidentified phosphorus-vacancy-hydrogen complexes that can be electrically, magnetically, or optically active centres. We found an unusual extremely large hyperfine coupling with the 31P nuclei (A > 2 GHz) for some of the investigated defects that requires further experimental verification. Finally, we demonstrate that the PV2H0 complex has two metastable triplets between the ground- and excited-state singlets, and it may exhibit a highly selective spin decay channel to a ground state, which makes the defect a promising candidate for realizing long-living solid-state quantum memory. These results provide deep insight into the donor compensation effect associated with vacancy-related clusters, and they may be useful in future identification of P-related defects suitable for quantum information processing applications.

Resume : The production difficulty of the DRAM devices is rapidly increasing due to integration process complexity and growth of the low power devices such as mobile application. These difficulties also affect product testing and yield analysis. In this study, we propose a precise way of predicting the package level yield using machine learning of imaged wafer data. We found the relationship between the wafer test data and package yield. It covered data pre-processing, feature selection and we verified the relevant data to derive an optimized relationship and the actual production process applied. There are over 200 measurement data in one chip of the wafer. The conventional way takes a long time since engineers have to investigate and judge the data themselves in order to find the causes of the failure. This leads to significant chip loss as well as man power loss in mass production. It is suggested to image a lot of data through a pre-processing process in this study. It is planning to predict package yields through machine learning of variety of data that is imaged. It is the first attempt to predict package yields through machine learning by imaging wafer data. CNN, which is easy to analyze images, was used as a machine learning tool.

Resume : In the last years, the development of computational methods permitted to obtain a super-probe of the atomic scale. Furthermore, it is now well understood that defects in structures (few %) can either lead to better or worse performances on the targeted properties of functional semiconductors. Therefore, combining modelling and experimental procedures become crucial if one wants to fully understand a targeted material in order to improve its properties for an application. In this talk, the attention will be drawn to two buffer layers to replace CdS in CIGS devices, i. e. CdIn2S4 and In2S3 (β-phase).1–4The native point defects of both materials were investigated using DFT (GGA) together with the VASP and PyDEF codes.5,6 All the computed results were confronted to experimental data. The theoretical results fully support the observed properties and explain which defects are responsible for them, e. g. CdIn2S4 and In2S3 give n-type conductivity. Finally, computations explain why it is impossible to get p-type conductivity for In2S3. 1 E. V. Péan, N. Barreau, J. Vidal, C. Latouche, and S. Jobic, Phys. Rev. Mater. 1, 064605 (2017). 2 N. Barreau et al. Sol. RRL 1, 1700140 (2017). 3 C. Guillot-Deudon, M.T. Caldes, A. Stoliaroff, L. Choubrac, M. Paris, C. Latouche, N. Barreau, A. Lafond, and S. Jobic, Inorg. Chem. acs. inorgchem.8b01771 (2018). 4 A. Stoliaroff, N. Barreau, S. Jobic, and C. Latouche, Theor. Chem. Acc. 137, 102 (2018). 5 E. Péan, J. Vidal, S. Jobic, and C. Latouche, Chem. Phys. Lett. 671, 124 (2017). 6 A. Stoliaroff, S. Jobic, and C. Latouche, J. Comput. Chem. 39, 2251 (2018).

Resume : Small-molecule organic semiconductors are used in a wide spectrum of applications, ranging from organic light emitting diodes (OLEDs) to organic photovoltaics. A number of factors determine their charge carrier mobility, such as molecular packing, electronic structure, dipole moment and polarizability. Here, we present a multi-scale model, which provides an accurate prediction of experimental data over ten orders of magnitude in charge carrier mobility and demonstrate the de novo design of a novel organic semiconductor with improved mobility [1,2]. Molecular orientation anisotropy of molecules used in OLEDs can give rise to an enhanced light-outcoupling efficiency or to the spontaneous buildup of an electrostatic potential perpendicular to the substrate. We will present our work on the orientation anisotropy of widely used organic semiconductors using a simulation approach which mimics the physical vapor deposition process of amorphous thin films. Our simulations reveal for all studied systems significant orientation anisotropy which is in agreement with experimental results [3,4]. The availability of first-principles based models to compute key performance characteristics of organic semiconductors may enable in-silico screening of numerous chemical compounds for the development of highly efficient opto-electronic devices. [1] Friederich et al., Adv. Funct. Mater., 2016, 26 (31), pp 5757-5763 [2] Friederich et al., Adv. Mater., 2017, 29 (43), pp 1703505 [3] Friederich et al., ACS Appl. Mater. Interfaces, 2018, 10 (2), pp 1881–1887 [4] Friederich et al., Chem. Mater., 2017, 29 (21), pp 9528–9535

Resume : The success of the Materials Genome Initiative [1] has led to opportunities for data-driven approaches for materials discovery. The recent development of Polymer Genome (PG), which is a machine learning (ML) based data-driven informatics platform for polymer property prediction, has significantly increased the efficiency of polymer design. Nevertheless, continuous expansion of the ‘training data’ is necessary to improve the robustness, versatility and accuracy of the ML predictions. In order to test the performance and transferability of the predictive models presently available in PG (which were previously trained on a dataset of 451 polymers), we have carefully collected additional experimental glass transition temperature (Tg) data for 871 polymers from multiple data sources. [2,3] The Tg values predicted by the present PG models for the polymers in the newly collected dataset were compared directly with the experimental Tg to estimate the accuracy of the present model. Using the full dataset of 1322 polymers, a new ML model for Tg was built following past work [4]. The RMSE of prediction for the extended dataset, when compared to the earlier one, decreased to 21.0 K from 48.7 K.To further improve the model’s performance, we are continuing to accumulate new data and exploring new ML approaches.

Resume : High titania slag is one of the raw material for the production of titanium. Titania slag is produced by the carbothermic reduction of the ilmenite ore in an electrical arc furnace. In this paper, effect of the FeO, TiO2, Ti2O3 on the liquidus temperatures and viscosity of the titania slag has studied. Cooling characteristics of titania slag was performed by using the equilibrium cooling and Scheil-Gulliver cooling equations. CALPHAD approach is used for the analysis of the equilibrium calculations and the phase diagram study of different systems. Analysis shows that temperature, FeO and Ti2O3 content have the inverse relation on viscosity of slags. Cooling curve shows that, in the equilibrium cooling 1100oC is the final solidification temperature, while in Scheil-Gulliver cooling liquid phases are still existing in the room temperature. Cooling rate or step has the major influence on the final solidification temperature. Key Words: titania slag; liquidus temperatures; viscosity; CALPHAD, Scheil-Gulliver cooling curve.

Resume : In this study, we analyze the wetting dynamics of two immiscible liquid systems with the aid of molecular dynamics simulations. An underwater oil droplet was placed on the smooth and a rough silicon surface and its wetting states were observed. The coupling parameter which defines the relative affinities between the atoms has found to have significant effect on the wetting characteristics of the oil droplet. The simulation results revealed that the solid surface became oleophobic when the coupling parameter of two liquids is kept low and the surface showed oleophilic characteristics when its value is increased. In addition, the contact angle of underwater oil droplet with the solid surface is also measured by experiments. Furthermore, the simulation and experimental contact angle data is compared with the theoretical models defined by Wenzel and Cassie. The results exhibit that for molecular dynamics simulation of two-liquid system, defining the coupling parameter is of great importance.

Resume : We investigate the existence of a body centered tetragonal (BCT) phase of Cu when single crystal FCC Cu is subjected to both high pressure and high temperature. The results have been demonstrated through DFT calculations (which are typically done at 0 K) followed by Helmholtz free energy calculations (for high temperature). The new metastable BCT phase of Cu demonstrates higher thermal conductivity compared to that of the FCC phase and thereby may be beneficial for high temperature engineering applications.

Resume : The structural, electronic and thermodynamic properties of XZrS3 (X=Ba,Sr,Ca) compounds with orthorhombic Pbnm and cubic Pm-3m phases have been investigated and reported. The calculations have been performed using various density functionals within the generalized gradient approximation (GGA). The obtained lattice parameters for Pnma phase reveal very good accord with experiment. The computed electronic band structures show that in the cubic phase the material of interest is an indirect band-gap (R-Г) semiconductor, whereas it is a direct band-gap (Г-Г) in the orthorhombic phase. The semiconducting XZrS3 (X=Ba,Sr,Ca) compounds are found to verify the stability criteria against volume change. Based on the quasi-harmonic Debye model, the thermodynamic properties of the material in question have been predicted taking into account of the lattice vibrations. The variation of the lattice constant, bulk modulus, heat capacity, Debye temperature and thermal expansion coefficient as a function of pressure in the range 0-30 GPa and temperatures of 0-1500 K is computed. Our findings show that external effects such as temperature and pressure are highly effective in tuning some of the macroscopic properties of the compounds under study. Keywords: Chalcogenide perovskite; Phase stability; Electronic Properties; Thermodynamic properties; Ab–initio calculations.

Resume : In this study, molecular dynamics (MD) was used to simulate a single layer of germanene quantum dots. The mechanical properties of buckled honeycomb germanene quantum dots under tension were investigated at different temperatures. The results show that Young's modulus of germanene quantum dots in armchair and zigzag directions at a temperature of 300 K are about 169 and 154 GPa, respectively. The ultimate strength of the single layer germanene in armchair and zigzag directions are 8.9 and 15.4 GPa, respectively. As the temperature was increased from 100 to 500 K, the ultimate strength and Young's modulus were significantly decreased. The ultimate strength of the single layer germanene in zigzag and armchair directions decreased about 65.8% and 66.6%, respectively. The results of this investigation are helpful for the application of two-dimensional germanene as optoelectronics and semiconductor-topological insulators.

Resume : Electrocatalytic reduction of nitrogen to ammonia, as an alternative to the energy-consuming Habor–Bosch nitrogen fixation, is a fascinating yet challenging topic. Here, by means of density functional theory and ab initio molecular dynamics calculations, we design a novel group of materials—two-dimensional (2D) pentagonal transition metal phosphides (penta-MP, M=Ti, Zr, Hf) and study their potential applications in the nitrogen fixation. Penta-MP are predicted to have dynamical, thermal, and mechanical stabilities. Their quasi-planar crystal structures and metallic properties facilitate the strong N2 adsorption on their surface. Gibbs free energy diagram suggests that NRR on penta-MP prefers the distal reaction mechanism, with the low overpotential of 0.56 eV for penta-TiP, which can benefit efficient electrocatalytic nitrogen fixation. Our findings open up a new pathway for designing novel 2D materials as well as electrocatalysts.

Resume : In this paper, we study the solar selective absorbing properties of metal nanowire array (NWA)/ anodic aluminum oxide (AAO) composites at 1000K via numerical simulation. The materials and structural parameters which influence the wavelength absorption between 0.28 and 10 μm are simulated and optimized. The results reveal that W NWA/AAO composites with nanowire length of 7.3 μm, fill factor of 0.03, and AAO template thickness of 0.1μm exhibits the best selective absorption. This composite has solar absorptivity α = 0.90 and thermal emissivity ε = 0.045 in 1000K, and shows a photothermal conversion efficiency of 71.5%. The effect of materials and structural parameters on the photothermal conversion efficiency was detailed discussed in this paper.

Resume : For the clean (001) surface of silicon, owing to a lack of upper bonding partners, each atom has two single occupied dangling bonds pointing out of the surface. Pairs of Si atoms dimerize, using up one dangling bond per atom to form the dimer bond. By depositing SrTiO3, the structural complexity of both the ferroelectric SrTiO3 itself and the buffer layers offers the possibility to break the "dimer-row" reconstruction of Si atoms. In this work, we systematically investigate SrTiO3/Si heterostructures from first-principles modeling in order to contribute to a thorough understanding of the driving force for this issue. Using first-principles density functional theory calculations, we firstly analyze the influence of the buffer layer on structural and electronic properties of the Si layers. It is proved that Si layers with full monolayer Sr coverage is the key to break Si dimers. Based on this point, we test all the possible SrTiO3/Si heterostructures with one monolayer Sr buffer layer to see how defective surfaces affact on Si dimers. The polarization of SrTiO3 layers and the distance between interface Sr and Si are sensitive to the surface configurations, which are the two main factors for whether the surface Si atoms prefer to dimerize or not. Using a parallel plate capacitor model, the effects mentioned above are well understood. Furthermore, our calculations show that for the cases with silicon dimers, the heterostructures possess semiconducting behavior. However, when the dimer is broken, the systems become metallic. Such an interesting phenomenon has been explained through detailed analysis of electronic states. This work helps to a fundamental understanding of the important role played by interfacial and surface structures in the SrTiO3/Si heterostructures and indicates the possibility to realize metal-insulator transition in such systems. It may suggest guidelines for the future design of coupled functional oxide-semiconductor devices in nanaotechonology.