Computer modelling in nanoscience and nanotechnology: an atomic-scale perspective IV

Resume : First-principles calculations using density-functional theory (DFT) are a very useful tool to study materials properties in depth. However, the big computational cost of such calculations imposes a limitation in the size of the simulated systems; thus, studying complex situations involving large simulation boxes becomes infeasible for such methods. Many solutions have been proposed for this problem, at the expense of losing accuracy and introducing approximations; for example, by using coarse-graining methods, Landau-like potentials, Lennard-Jones potentials, and other approaches, the community has developed a large collection of energy functionals able to reproduce key features of the materials of interest. Recently, Wojdeł and collaborators [1] presented an atomistic approach aiming at retaining the accuracy of the DFT calculations while being computationally affordable. In order to do so, an adapted Taylor expansion explicitly compliant with the acoustic sum rule is performed in terms of atomic displacements and cell strains around a certain reference structure. The method thus constitutes a systematically improvable atomistic (electrons are not treated explicitly) approach that can in principle reproduce exactly the potential energy surface, and the main technical difficulty is to find a way to compute the model parameters in a manner that is as automatic as possible. Here we present an efficient and flexible methodology to build such models via an optimization problem. Our approach has the peculiarity of employing a suitably chosen goal function that permits an analytic solution to the fitting problem, which makes the calculations very fast and allows us to explore the parameter space in an essentially complete way. We apply the method to construct a model for feroelastic perovskite strontium titanate as a test subject. We show that our model reproduces the DFT-computed ground state properties and even high temperature energetics very accurately, and allow us to simulate the structural phase transition in this compound. Time allowing, we will describe more recent developments, e.g. to describe chemically inhomogeneous situations. [1] Wojdeł et al., JPCM 25, 305401 (2013)

Resume : Transition metal­fullerenes complexes with metal atoms bound on the external surface of C60 are promising building blocks for next-generation fuel cells and catalysts. Yet, at variance with endohedral M@C60, they have received so far a limited attention. In the work presneted here, by resorting to first principles simulations, we elucidate structural and electronic properties for the M­C60 and M2-C60, M = Pt and / or Pd complexes. These specific transition metals have been selected as representative of the best candidates for their intrinsic catalytic properties and their ability to bind to fullerenes. In this context, we show that the most stable structures feature the metal atom located above a high electron density site, namely, the bond between two adjacent hexagons (-66 bond). If two metal atoms are simultaneously added, the most stable configuration turns out to be the one in which metal atoms still stand on -66 bonds but the two metals tends to clusterize. The electronic structure, rationalized in terms of localized Wannier functions, provides a clear picture of the underlying interactions responsible for the stability or instability of the complexes, showing a strict relationship between structure and electronic gap.

Resume : Recent computational and theoretical work [1-5] on understanding of thermal impedance of a crystal lattice with a monatomic unit cell due to phonon-phonon scattering processes is reviewed. It is demonstrated that equilibrium molecular dynamics simulation in conjunction with the Green-Kubo formalism provides an effective basis to explore the peculiarities of phonon dynamics. The key role of equilibrium molecular dynamics simulation is the ability to give unique and direct access to the heat current autocorrelation function which is shown, for a monatomic lattice, to reveal consistently two-stage relaxation. It is found that the two-stage relaxation can be universally modelled by an analytical expression which provides an exceptional basis for the development of a general analytical treatment of the lattice thermal impedance. A detailed analysis of the calculation data on the basis of this expression suggests that the phonon gas in a monatomic lattice decomposes into a mixture consisting of two 'thermal fluids' associated with the acoustic short and long range phonon modes. The decomposition of the phonon gas into the acoustic short and long range phonon modes is due to the decomposition of each of vibrational modes of the energy spectrum of a monatomic lattice into two fundamental contributions. These two contributions are related to some basic (elemental) energy spectra, a kind of two basic 'vectors' on which each mode of the energy spectrum can be decomposed with appropriate parametrization. One of them, associated with the acoustic short range phonon modes, is available only at frequencies ω≳ω_c and leads to thermal motion which is able to relax quickly to equilibrium directly via resistive Umklapp (or U) processes. In contrast, the other elemental energy spectrum, associated with the acoustic long range phonon modes, starts from the origin ω=0 and leads to thermal motion which: (i) initially relaxes to the so-called displaced equilibrium via non-resistive normal (or N) processes (the displaced equilibrium state moves as a whole with a drift velocity relative to the lattice), and (ii) only then can gradually relax to the true equilibrium via resistive Umklapp processes. The first elemental energy spectrum and associated with it the acoustic short range phonon modes might be related to the vibrational degrees of freedom which can be viewed as an extension of the Einstein picture of the coupled harmonic oscillators vibrating at the same frequency with the assumption of a random phase between them. This extension requires including into the consideration of: (i) both larger localized oscillating entities and varying vibrational frequencies (ω_c≲ω≲ω_D, ω_D is the Debye frequency - the highest allowed phonon frequency in the crystal lattice) compare to the single atom and the same frequency considered by Einstein, and (ii) the anharmonic effects. Meanwhile, the second elemental energy spectrum and associated with it the acoustic long range phonon modes are supposed to be related to the vibrational degrees of freedom which result in collective thermal motion produced by lattice waves of different frequencies (0≤ω≲ω_D) as originally hypothesized by Debye. Also, the two-fluid microscopic picture logically incorporates the momentum conservation criterion of the Callaway treatment as well as naturally dividing the phonon modes into two groups ('fluids') as hypothesized by Klemens but with explicitly defined characteristic frequency ω_c which turns out to have a fundamental physical importance, determining the energy gap ~ℏω_c between the lowest energy levels of the acoustic short and long range phonon modes. On the basis of the insight, a frequency 'window' is predicted for an external periodic temperature perturbation to generate thermal waves (so called second sound) in a monatomic lattice. The exposed two-fluid microscopic picture of phonon dynamics naturally captures and unites the essences of treatments due to Einstein, Debye, Klemens and Callaway, among others, within the framework of the fundamental concept due to Peierls based on the Boltzmann transport equation for the phonon distribution function and the two anharmonic interactions, non-resistive normal and resistive Umklapp processes. References: [1] A.V. Evteev, L. Momenzadeh, E.V. Levchenko, I.V. Belova, G.E. Murch, Phil. Mag., 94, 731 (2014). [2] A.V. Evteev, L. Momenzadeh, E.V. Levchenko, I.V. Belova, G.E. Murch, Phil. Mag., 94, 3992 (2014). [3] E.V. Levchenko, A.V. Evteev, L. Momenzadeh, I.V. Belova, G.E. Murch, Phil. Mag., 95, 3640 (2015). [4] A.V. Evteev, E.V. Levchenko, I.V. Belova, G.E. Murch, Phil. Mag., 95, 2571 (2015). [5] A.V. Evteev, E.V. Levchenko, L. Momenzadeh, I.V. Belova, G.E. Murch, Phil. Mag., 96, 596 (2016).

Resume : Self-consistent simulations with open boundaries for electrons are typically expensive to perform, and labour intensive to set up. Hairy Probes [1] is a formalism that allows these simulations to be made both simple to perform, and computationally efficient. We present a reformulation of the Hairy Probe method [2] for steady state calculations involving non-orthogonal atomic basis sets. Results will be presented for a perfect atomic wire of Cu atoms, and a perfect non-orthogonal chain of H atoms to showing the correctness of the method. Then further results relevant to molecular electronics with graphene contacts will be presented. [1] McEniry, E. J.; Bowler, D.; Dundas, D.; Horsfield, A. P.; Sánchez, C. G. & Todorov, T. N. Dynamical simulation of inelastic quantum transport Journal of Physics: Condensed Matter, 2007, 19, 196201 [2] Horsfield, A. P.; Boleininger, M.; D'Agosta, R.; Iyer, V.; Thong, A.; Todorov, T. N. & White, C. Efficient simulations with electronic open boundaries. In preparation

Resume : Atomic Scale Materials Design via Simple Electro-Structural Descriptors Controlling the macroscopic properties at the atomic scale and designing new materials is challenging and requires the identification of simple electronic and structural features that can be accessed and tuned experimentally. Using non-standard methods, we developed a systematic approach that enables us to decouple the electro-structural contributions to the macroscopic response of a generic material, irrespective of its composition and geometric structure. We characterize the electronic features by inspecting the atomic orbital polarization and exploiting our new formulation of bond covalency analysis; we then combine such outcomes with structural and dynamic information from group theoretical analysis and the recently developed lattice dynamic metric named “cophonicity”. This protocol has been successfully applied to the study of diverse phenomena and class of materials like charge-ordering perovskites, non-linear optically active compounds, hybrid-improper ferroelectrics and Transition Metal Dichalcogenides. References: Chem. Mater. 28, 1965 (2016) (Review) DOI: 10.1021/acs.chemmater.6b00430 Inorg. Chem. 54, 5739 (2015) DOI: 10.1021/acs.inorgchem.5b00431 Chem. Mater. 26, 5773 (2014) DOI: 10.1021/cm502895h ACS Photonics 1, 96 (2014) DOI: 10.1021/ph400049h Phys. Rev. B 92, 014102 (2015) DOI: 10.1103/PhysRevB.92.014102 Phys. Rev. B 87, 155135 (2013) DOI: 10.1103/PhysRevB.87.155135 Phys. Rev. B 86, 195144 (2012) DOI: 10.1103/PhysRevB.86.195144

Resume : The high-throughput computation of materials properties by ab initio methods has become the foundation of an effective approach to materials design, discovery and characterization. This data driven approach to materials science currently presents the most promising path to the development of advanced technological materials that could solve or mitigate important social and economic challenges of the 21st century. Enhanced repositories such as AFLOWLIB open novel opportunities for structure discovery and optimization, including uncovering of unsuspected compounds, metastable structures and correlations between various properties. However, the practical realization of these opportunities depends on the the design of efficient algorithms for electronic structure simulations of realistic material systems beyond the limitations of the current standard theories. In this talk, I will review recent progress in theoretical and computational tools, and in particular, discuss the development and validation of novel functionals within Density Functional Theory and of local basis representations for effective ab-initio tight-binding schemes.

Resume : We have developed a software package KLMC (Knowledge Led Master Code) 1 that utilises massively parallel computer platforms and third-party computational chemistry software to perform stochastic sampling and systematic searches for local and global minima on energy landscapes. Our software enables us to investigate diverse physical and chemical phenomena including prediction of atomic structure of nanoparticles, surface reconstructions of polar surfaces, and defect complexes. Automation within KLMC 1 alleviates repetitive or mundane computational tasks or processes typically required in simple task farming, global optimisation routines such as basin hopping and genetic algorithms, and statistical sampling. KLMC has recently been successfully exploited to predict plausible structures of nanoparticles of binary compounds 1,2 (e.g., ZnO, CdSe, MgO, KF, LaF3, BaO); mechanisms of polar surfaces reconstruction3 (including KTaO3 and ZnO); nucleation and growth of nanoparticles on metal supports 4 (e.g., ZnO on Ag, and Cu on ZnO). Complementary to KLMC, we are now developing a novel nanocluster database WASP@N (Web Assisted Structure Prediction at the Nanoscale) 5, in which we intend to link the web-interfaced database and compute nodes dedicated to help scientific community to search, discover and disseminate nanoclusters. The web-interfaced database is designed to calculate physical properties of the uploaded nanoclusters (e.g. symmetry, moments of inertia, and electrical dipoles), to find similar structures which were previously uploaded and to perform on-demand calculations. Combining KLMC and WASP@N provides a powerful solution to the problem of structure prediction at the nanoscale for materials scientists from data mining to visualisation. 1. Farrow et al. (2014) Phys. Chem. Chem. Phys. 16, 21119 2. Woodley (2013) J. Phys. Chem. C 117, 24003 3. Deacon-Smith et al. (2014) Adv. Mater. 26, 7252 4. Demiroglu et al. (2014) Nanoscale 6, 14754 5. http://hive.chem.ucl.ac.uk

Resume : Recent developments in the field of Ultrafast Dynamics have made it possible to access electron dynamics with sub-femtosecond resolution, giving insight into fundamental physical processes. Computational modelling is able to assist the interpretation of data from Ultrafast experiments, although these systems are particularly complex due to the prevalence of strong, time-dependent electric fields, and the effects of electron correlation. Time-dependent fields require computational methods with a reliable description of electronic excited states, as well as being sufficiently efficient to enable demanding time-dependent simulations of molecules in condensed phases. The challenge is hence to develop a theory with predictive capabilities comparable to density-functional theory (DFT), but with significantly less computational effort. We present a new method based on density-functional tight-binding (DFTB)[1]. DFTB is an approximate density functional theory (DFT), which is made orders of magnitude faster by the generation of integral look-up tables. Current formulations provide a poor description of electronic polarisability, necessitating further improvements. We formulate a new DFTB approach including self-consistent polarised charges [2] and polarisation orbitals. The electrostatic integrals are evaluated analytically for the chosen basis set by using Gaussian expansions, resulting in an internally consistent theory without fitting. The resulting Gaussian Tight Binding (GTB) model gives molecular polarisabilities in excellent agreement with experimental data, having errors of the same order as DFT using the PBE exchange-correlation functional. Finally we propagate the electronic-nuclear motion in time using the Ehrenfest approximation, which adds a basic description of non-adiabaticity to our model. We have applied the GTB method to investigate molecular polarisabilities, as well as charge dynamics following an excitation event of various hydrocarbons and polymer chains and report on these.. [1] Elstner, Marcus, et al. "Self-consistent-charge density-functional tight-binding method for simulations of complex materials properties." Physical Review B58.11 (1998): 7260. [2] Finnis, M. W., et al. "Self-consistent tight-binding approximation including polarisable ions." MRS Proceedings. Vol. 491. Cambridge University Press, 1997.

Resume : We investigate redox levels in aqueous solution using a combination of ab initio molecular dynamics (MD) simulations and thermodynamic integration methods. The molecular dynamics are performed with both the semilocal Perdew-Burke-Ernzerhof functional and a nonlocal functional (rVV10) accounting for van der Waals (vdW) interactions. The band edges are determined through three different schemes, namely, from the energy of the highest occupied and of the lowest unoccupied Kohn-Sham states, from total-energy differences, and from a linear extrapolation of the density of states. It is shown that the latter does not depend on the system size while the former two are subject to significant finite-size effects. For the redox levels, we provide a formulation in analogy to the definition of charge transition levels for defects in crystalline materials. We consider the H+/H2 level defining the standard hydrogen electrode, the OH?/OH? level corresponding to the oxidation of the hydroxyl ion, and the H2O/OH? level for the dehydrogenation of water. In spite of the large structural modifications induced in liquid water, vdW interactions do not lead to any significant structural effect on the calculated band gap and band edges. The effect on the redox levels is also small since the solvation properties of ionic species are little affected by vdW interactions. Since the electronic properties are not significantly affected by the underlying structural properties, it is justified to perform hybrid functional calculations on the configurations of our MD simulations. The redox levels calculated as a function of the fraction ? of Fock exchange are found to remain constant, reproducing a general behavior previously observed for charge transition levels of defects. Comparison with experimental values shows very good agreement. At variance, the band edges and the band gap evolve linearly with ?. For ? ? 0.40, we achieve a band gap, band-edge positions, and redox levels in overall good agreement with experiment.

Resume : The majority of high-throughput materials exploration efforts employ a DFT dataset of perfect crystals. However, construction of another DFT database by systematic surface calculations is essential to achieve better performance on screening of surface related properties, such as ionization potentials and catalytic properties. There are two major issues in this task. 1) Automatic slab model generation. Nonpolar slabs are generally preferred over polar ones in high-throughput calculations as the latter require case-by-case treatment to avoid the "polar catastrophe”. In addition, case-by-case surface reconstruction becomes more relevant in the latter. Furthermore, a scheme is needed to automatically generate unique surface termination models given the crystal, orientation, and minimum slab and vacuum thicknesses [1]. 2) Automatic nonpolar orientation determination. Checking all orientations to see whether a nonpolar slab can be obtained or not is inefficient. Moreover, logic must be employed to automatically choose specific orientations over others. In addition, the number of symmetry search and supercell generation should be reduced as much as possible for the sake of efficiency. We will show how these issues could be addressed and actually demonstrate generation of a high-throughput database on binary oxide surfaces. [1] Y. Hinuma et al., Comp. Mater. Sci. 113 (2016) 221.

Resume : We introduce a new model for calculating the change in time of three-dimensional atomic configurations. The method is based on the kinetic mean field (KMF) approach [1], however we have transformed that model into a stochastic approach by introducing dynamic Langevin noise. The result is a stochastic kinetic mean field model (SKMF) which produces results similar to lattice kinetic Monte Carlo (KMC). SKMF is, however, more cost-effective and the algorithm is easier to implement. [2] The group made the software and the program code (together with tutorials) freely available to the scientific community at the http://skmf.eu webpage. We plan to keep this open source approach with the model's further developments, too. [3] [1] Martin G., Atomic mobility in Cahn's diffusion model, Phys. Rev. B 41, 2279-2283 (1990) [2] Erdélyi Z., Pasichnyy M., Bezpalchuk V., Tomán J.J.,Gajdics B., Gusak A.M., Stochastic Kinetic Mean Field Model, Computer Physics Communications 204: pp. 31-37. (2016), http://dx.doi.org/10.1016/j.cpc.2016.03.003. [3] http://skmf.eu

Resume : Recently, non-linear self-consistent version of atomistic kinetic mean-field model suggested by George Martin [1], later applied to the systems with strong diffusion asymmetry [2] and generalized to 3D case [3], was essentially developed [4,5]. New development, SKMF (Stochastic Kinetic Mean-Field), contains Langevin noise of jump frequencies. Noise enables overcoming the barriers in first-order transformations. Also, changing noise amplitude enables “fine-tuning” from stochastic Kinetic Monte Carlo (KMC) description to deterministic Kinetic Mean-Field (KMF) description. Nucleation is a natural topic for application of SKMF since noise helps to overcome the nucleation barrier. Application of SKMF to following problems is discussed: 1. Nucleation rate in decomposition of solid solution as a function of noise amplitude and composition. 2. Ordering kinetics in FCC structures. 3. Decomposition of FCC solid solution with negative mixing energy with precipitation of ordered phase. 4. Ordering in the sharp concentration gradients of binary systems with diffusion asymmetry. 5. Phase competition in A-B interaction with formation of A3B, AB, AB3 phases. 6. SKMF model of phase formation and competition in exothermic SHS reactions. [1] Martin, G. Phys. Rev. B 41, 2279-2283 (1990). [2] Erdélyi, Z., Sladecek, M., Stadler, L., Zizak, I., Langer, G., Kis-Varga, M., Beke D., Sepiol, B. Science,306(5703), 1913-1915. (2004). [3] N. Storozhuk, K. Sopiga, A. Gusak. Philosophical Magazine 93, 1999-2012 (2013). [4] Erdélyi, Z., Pasichnyy, M., Bezpalchuk, V., Tomán, J. J., Gajdics, B., Gusak, A.M. Computer Physics Communications. (2016), http://dx.doi.org/10.1016/j.cpc.2016.03.003 [5] http://skmf.eu

Resume : Micro- and nano-structured materials are crucial for future energy technologies. Key processes during production and life-time are governed by self-organization in phase separation processes at the micro and nano scale. Examples include nano-structured Silicon absorber layers in solar cells providing tailored band-gaps [Apl. Phys. Lett. 103, 133106 (2013)] as well as cheap production and micro-patterned electrolyte-matrices] enhancing life-time and efficiency in a range of fuel cell technologies. Simulations of these out-of-equilibrium, inhomogeneous real world systems provide important insights, finding potential for optimization of structures and process parameters. To this end, kinetic lattice Monte Carlo simulations can be used to model physical systems at experimental scales in an atomistic way, thereby side-stepping many caveats connected with the alternative phase-field simulations. In this contribution, we present two massively parallel implementations for large-scale simulations on GPUs: One is optimized to offer fast time-to-solution on experimental-scale simulations [Eur. J. Phys.: Spec. Top. 210, 175 (2012)], the other provides highly efficient parameter studies or large sample sizes for large-scale simulations [Phys. Rev. E (2016) submitted]. Harnessing the compute power of modern (multi-)GPU installations leads to increased energy efficiency as well as reduced time-to-solution.

Resume : Free-surface-induced selective destabilization of L10 superstructure variants in nanolayers, nanowires and nanoparticles of FePt were studied by means of Monte Carlo simulations implemented with vacancy mechanism for atomic migration. The generated samples initially either perfectly ordered in the c-variant L10 superstructure, or completely disordered in the fcc structure were modeled with nn and nnn interatomic pair interactions deduced from “ab-initio” studies of Fe-Pt. The heterogeneous nucleation of a- and b-L10 variant domains was induced by all the (100)-type surfaces limiting the nanostructures. It was observed that in the case of nanocubes the competition between the a- and b-variant L10 domains nucleating at the (100), (010) and (001) surfaces resulted in suppression of their growth. As a consequence, most of the cube volume remained untransformed and showed the c-variant L10 chemical long-range order (LRO). The initially disordered samples were transformed by the creation of a specific L10 domain structure with a mosaic of particular L10-variant domains at the surfaces and almost homogeneous long-range order in the inner volume. The analysis of correlation effects revealed that chemical ordering in the initially disordered nanosystems was initiated at the free surfaces. The obtained results are important for the development of magnetic storage media technologies requiring a stability L10 superstructure variants determining easy magnetization directions.

Resume : The impact of reduced dimensionality on the magnetic properties of the tetragonal L10 CoPt alloy is investigated from ab-initio considering several kinds of surface defects. By exploring the dependence of the magnetic anisotropy energy (MAE) on the thickness of CoPt thin films, we demonstrate the crucial role of the chemical nature of the surface. For instance, Pt-terminated thin films exhibit huge MAEs which can be 500% larger than those of Co-terminated films. Besides the perfect thin films, we scrutinize the effect of defective surfaces such as stacking faults or anti-sites on the surface layers. Both types of defects reduce considerably the MAE with respect to the one obtained for Pt-terminated thin films. Keywords : magnetic anisotropy energy, thin films, L10 CoPt alloy.

Resume : Vacancy formation energy is an important parameter that determines the structural and phase transitions in solids and can be calculated using the ab initio simulation methods. But some issues require clarification. Firstly straight forward answer to the question of the preciseness of the calculation is absent due to the shielding effect of the electrostatic interactions of short-range order and as a result – impossibility of the accurate consideration of the electron density gradient near the vacant site. Secondly the assumption is used, that the energy and entropy of vacancy formation are independent of the temperature and are usually calculated for the temperature of 0 K followed by the comparison of the obtained results with experimental data for higher temperatures. The aim of this work is to investigate the influence of the temperature factor on the free energy of vacancy formation and its components (free energy of lattice vibrations and the energy of thermal electronic excitation) on the example of metals with different type of crystal lattice – FCC (Al) and BCC (Mo) – within the quasiharmonic approximation with using of Density Functional Theory. It was established that the vacancy formation energy sufficiently increases with the temperature in all cases related to the thermal expansion of the crystal lattice. This effect is underestimated in the calculations due to certain restrictions of quasiharmonic approach. Its consideration allows to increase the accuracy of calculations and to obtain the correct results for real objects.

Resume : The Curie temperature of magnetic nanoparticles has a dependence upon both the size and shape of the particle and has been observed to both decrease and increase with reducing particle size in agreement with finite scaling theory. However, due to the complex interplay between finite-size and surface effects, predictions as to the exact behaviour of any given material are difficult. A quantitative understanding of the change in magnetic ordering temperature for nano-sized materials of interest is important for understanding their potential applications. The ordering temperature for magnetite nanoparticles was investigated using an atomistic mean field model applied to a realistic crystal structure, with experimentally determined values for the exchange energies. The arising system of coupled non-linear equations was numerically solved using the SNES library of the Portable, Extensible Toolkit for Scientific Computation (PETSc). Some preliminary calculations have shown a reduction in both Curie temperature and surface magnetisation for magnetic nanoparticles, in qualitative agreement with both experimental and computational results.

Resume : Modification of CNTs by non-isovalent dopants can change their physical and chemical properties, and therefore improve performance of the CNT-based materials is specific applications. Oxyanions of heavy metals are toxic industrial pollutants and their removal from industrial wastewaters is an urgent technological problem [1]. In this report, adsorption of XO42- (X = Cr, Mo, W) molecular oxyanions on the surfaces of pristine and B(N)-doped carbon nanotubes is analyzed in computational studies. A DFT-based geometry-optimized calculations of the electronic structure of undoped, B- or N-doped CNTs of (3,3) and (5,5) chiralities and graphene sheets with adsorbed XO42- oxyanions were carried out using Gaussian 03 program package [2]. Relaxed geometries, binding energies, charge states of the adsorbates and the electronic wave-function profiles were calculated and analyzed. Calculation results are discussed in view of potential application of the CNT-based materials as efficient adsorbents of toxic oxides of hexavalent metals. The publication is based on the research provided by the grant support of the State Fund For Fundamental Research (project F64/42-2016). 1.Tarutani N., Tokudome Ya., Fukui M., et al. // RSC Adv. -2015.-5.-P.57187–57192. 2. Frisch M.J., Trucks G.W., Schlegel H.B., et al. Gaussian 03 (Gaussian, Inc., Wallingford, CT, 2003).

Resume : Pentameric ligand-gated ion channels (pLGICs), embedded in the membrane of nerve cells, are important neuroreceptors that mediate fast synaptic transmission, are involved in several neurological disorders and are target sites for drugs and, in invertebrates, insecticides. Here we investigate the first crucial step of their activation mechanism, which consists in the binding of a neurotransmitter to their extracellular domain, focussing of the prototypical case of the neurotransmitter GABA binding to the insect RDL receptor, linked to the resistance to the insecticide dieldrin. Using the innovative ``funnel-metadynamics'' computational technique, which efficiently enhance the sampling of bound and unbound states using a funnel-shaped restraining potential to limit the exploration in the solvent, we describe the binding free energy landscape, identify the chain of events leading GABA from the solvent into the binding-pocket and estimate the binding affinity. Moreover, we show how this landscape is disrupted by mutations which prevent the receptor to function. The RDL receptor shares structural and functional features with other pLGICs, hence our work provides a valuable protocol to study the binding of ligands to pLGICs beyond conventional docking and molecular dynamics techniques.

Resume : Hexagonal ice (Ih), the common form of crystalline ice, plays a crucial role in controlling and maintaining the natural environment. At low vapour pressures, Ih forms hexagonal prisms that can be thin plates or elongated needles depending on the temperature. The macroscopic shape is determined by the relative rate of growth of Ih basal and prism surfaces. To understand the mechanisms of growth at the molecular level and ultimately predict the shape of Ih crystals, we use a combination of molecular dynamics and metadynamics calculations in a range of temperatures. Our simulations show the formation of a quasi-liquid layer (QLL) at the surfaces, which mediates crystal growth and has a thickness which varies with temperature. The QLL gives rise to a vapour/liquid and a liquid/solid interfaces and we investigate the interplay between these two interfaces in the growth processes.

Resume : Graphene on Ni(100) forms a variety of Moire’ patterns which can be well explained in atomistic models by the mismatch with the substrate, with periodicity depending on the relative angle between the hexagonal graphene and the square surface lattices. Evidence of the different Moire’ structures is given by high resolution scanning tunneling microscopy images, that are well reproduced by ab-initio simulations. Beyond providing the detailed atomic-scale structures, the numerical simulations allow a deep local characterization of the chemical bonding between the graphene layer and the support. We also discuss the possible formation beneath the graphene of a surface-confined nickel-carbide in specific regions of the Moire’, whose presence is suggested by experimental STM images.

Resume : Graphene is a material, which is exhibiting a rising research interest over the last few years due to its unique properties and various possible applications. Possible applications range from optical sensing and drug delivery up to electric devices. As graphene is a zero band gap semiconductor, various approaches to open a band gap for electrical applications have been made over the last decade. In this work, the opening of a band gap by covalent functionalisation of graphene sheets with halogenated carbenes is examined. Furthermore, the effect of activating and deactivating substituents on the charge distribution and therefore on the self-assembly of molecules on the graphene surface is studied.

Resume : This project is to understand thermodynamics of bulk metals such as Zr and NiTi in order to understand martensitic transformation in metals. This requires the calculations of free energy of the possible phases of bulk metals. From a theoretical point of view, studying the temperature dependence of a particular phase is challenging because the thermal expansion of the system has to be taken into account. This may not preserve the space group of a crystal and can be studied using molecular dynamics simulations. Ab initio molecular dynamics (AIMD) simulations have been used to simulate the possible phases of bulk metals. The crystal structures at a number of temperatures and pressures have been established. The AIMD simulations are then run in a canonical ensemble and the vibrational frequencies of the crystals are derived from the AIMD data. The free energies are calculated within the quasi-harmonic approximation.

Resume : Fullerenes and Carbon nanotubes are among the most important discoveries in material science of the last century and motivated a large number of studies since then. One interesting possibility in this area of investigation is to obtain the so called inorganic fullerenes, like the already synthesized B40 and the WS2 and MoS2 closed cage clusters. With the goal of finding new inorganic fullerenes, the present investigation includes molecular structures with architectures inspired by the usual carbon fullerenes but formed with different combinations of atoms, including atoms from 3 columns of the periodic table of the elements. Our results show several structures, minimized considering atomic elements like Ge, Si, In, Ga, Sn, and others. Some of these need to be combined with other atom, like H in order to stabilize the molecule.

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. ScF3 exhibits NTE effect over a wide range of temperatures from 10 K to 1100 K. The NTE range can be also controlled by changing chemical composition as in solid solutions Sc1-xYxF3 and Sc1-xTixF3. Here 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 2x2x2 to 5x5x5) to investigate the stability of AIMD simulation results. 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.

Resume : Modeling the atomic structure of materials is crucial to achieve a precise understanding of the effects of scale, dimensionality, surface, interfaces, disorder... on their thermal response. Morever, using molecular dynamics (MD) ensures a complete description of all the scattering mechanisms experienced by phonons via the anharmonicity of the energy landscape. We have recently proposed a method to study and quantify thermal transport at the atomic scale, called the approach-to-equilibrium molecular dynamics (AEMD). This method is faster than standard ones and allows to extend the study of good bulk conductor over um lengths. In this presentation, we will first show the capability of the method to study heat transport in nanostructures of real-size, in particular nanowires and membranes. Then we shall highlight the insight on the thermal conductivity and phonon mean free paths that can be extracted from the length dependence of the thermal conductivity obtained by AEMD. In a second part, we will show that the lower computational cost of the methodology allows the calculation of thermal properties via first principles MD (FPMD). A temperature difference can be created within the time-scale of FPMD and the heat transient can be monitored to determine the thermal conductivity of a-GeTe4. This is the first calculation of a thermal conductivity via FPMD, to be considered as a step to overcome severe limitations of empirical potentials for complex materials and interfaces.

Resume : By means of proof-of-concept computer experiments, we provide evidence that thermal rectification in bulk Si by a population of defects, gradually distributed along the direction of the applied thermal gradient. We consider a graded population of both Ge substitutional defects and nanovoids, distributed along the direction of an applied thermal bias, and predict a rectification factor comparable to what observed in other low-dimensional Si-based nanostructures. By considering several defect distribution profiles, thermal bias conditions, and sample sizes our results suggest that a possible way for tuning thermal rectification is by defect engineering. This result is obtained by state-of-the-art atomistic simulations. The applied physics presented and discussed could help improving both the understanding and the managing of nanoscale thermal transport, a topic of potentially large impact in some emerging nanotechologies like, e.g., phononics. Interestingly enough, the practical fulfilment of the concept we have elaborated does not require sophisticated nanofabrication technique and, therefore, it could be useful for large-scale applications.

Resume : In this contribution, we review recent advances in understanding of transport coefficients in binary melts [1-4]. Equilibrium molecular dynamics simulation in conjunction with the Green-Kubo formalism is employed to study the transport properties of the model of Ni50Al50 melt with two different embedded-atom method potentials developed in [5,6]. The main objective of the work is to quantitatively characterize and analyze thermotransport in the system, i.e. diffusion driven by a temperature gradient. In addition, direct phenomenological coefficients for mass and thermal transport are also evaluated and analyzed in the process. The obtained results for these two different potentials are compared between each other as well as with experiment where it is possible. It is found that both potentials are able to consistently predict both direct transport coefficients over a wide temperature range. Meanwhile, these two potentials are found to be inconsistent in characterizing of the cross-coupled heat and mass transport, predicting even different direction (sign) of thermotransport. The origin of this difference will be discussed in the contribution in detail. References: [1] A.V. Evteev, E.V. Levchenko, I.V. Belova, R. Kozubski, Z.-K. Liu, G.E. Murch, Phil. Mag., 2014, 94, 3574. [2] A.V. Evteev, E.V. Levchenko, I.V. Belova, R. Kozubski, Z.-K. Liu, G.E. Murch, Diffusion Foundations, 2014, 2, 159. [3] A.V. Evteev, E.V. Levchenko, L. Momenzadeh, Y. Sohn, I.V. Belova, G.E. Murch, Phil. Mag., 2015, 95, 90. [4] E.V. Levchenko et al , Phil. Mag., 2016, in press. [5] Y. Mishin, M.J. Mehl and D.A. Papaconstantopoulos, Phys. Rev., 2002, B 65, 224114. [6] G.P. Purja Pun, Y. Mishin, Philosophical Magazine, 2009, 89, 3245.

Resume : From the point of view of the mainstream microelectronics technology, growing graphene on Ge substrates by chemical vapor deposition (CVD) is an interesting alternative to the usual growth done by CVD on Cu substrates. This is because by using Ge instead of Cu one avoids the hazard of contaminating the production tools with Cu [1]. In addition, in some cases graphene grown directly on Ge(001) may be directly used in a transistor [2]. For these reasons, CVD growth of graphene on Ge has recently been the subject of numerous studies [3]. We discuss the results of ab initio density functional theory (DFT) calculations for the interaction between C, H, and Ge during deposition of C2H4 and CH4 on Ge(001). Decomposition of the precursors, surface diffusion and surface reactions of the components, and nucleation and of growth of graphene seeds are discussed. The role of surface defects (missing dimers and atoms, surface steps) in the process of crystallographic orientation transfer from Ge(001) to graphene is addressed. The similarities and differences between the mechanisms governing the growth during CVD with various precursors and at various pressures are highlighted. [1] G. Lupina et al., ACS Nano, 9, 4776 (2015). [2] W. Mehr et al., IEEE EDL. 33, 691 (2012). [3] E.g., J. H. Lee et al., Science 344, 286–289 (2014), R. M. Jacobberger et al., Nature Comm. 6, 8006 (2015), J. Dabrowski et al., arXiv 1604.02315 (2016), I. Pasternak et al., Nanoscale (2016), DOI: 10.1039/C6NR01329E.

Resume : The Metal/Molecule/Metal nanojunctions are powerful devices for investigating phenomena involved in the domain of molecular electronics. Nowadays, thanks to the scanning tunneling microscope, it is possible to assemble atoms and molecules for building molecular junctions, to measure and manipulate their electric properties at the nanometric scale[1], and to excite these devices to induce optical transitions from one state to another[2]. In addition, molecular junctions are suitable systems for theoretical investigations with ab initio methods, since both size (few thousands atoms) and characteristic timescale of optical phenomena occurring (picoseconds) are accessible to the current high performance computing systems, allowing for a direct comparison to experiment. In this talk, I will present density functional theory (DFT) and time-dependent DFT results concerning a single-molecule light-emitting diode, composed of a thiophene- and porphyrine-based molecule grafted between two gold electrodes, a device recently developed in our laboratory [2,3]. The electric current crossing the nanodiode acts as a local source for molecular excitations. We will focus on the emission spectra of such a device, paying particular attention to the vibronic transitions occurring in the nanojunction. [1] G. Schull, Y. J. Dappe, C. Gonzalez, H. Bulou, and R. Berndt, Nano Letters 11, 3142-3146 (2011). [2] G. Reecht, F. Scheurer, V. Speisser, Y. J. Dappe, F. Mathevet, and G. Schull, Phys. Rev. Lett. 112, 047403 (2014). [3] M. C. Chong, G. Reecht, H. Bulou, A. Boeglin, F. Scheurer, F. Mathevet, and G. Schull, Phys. Rev. Lett. 116, 036802 (2016).

Resume : Porous organic molecules are a relatively new class of porous materials that could find application in sensing, storage, catalysis and separation. Unlike other porous materials, such as metal organic frameworks (MOFs) or zeolites, porous organic molecules are composed of discrete molecules with no direct covalent bonding between them. These can be studied in the solid state, for the properties arising from crystal packing and pores, or at the individual molecular level. The latter approach is especially interesting allowing for relatively fast and cheap computational analysis. Here we present a computational study focused on a single molecule analysis of previously reported porous organic molecules. This includes so called organic “cages”, but also cucurbiturils, cyclodextrins and cryptophanes. We present a new insight into the Xe/Kr selectivity performance of these molecules followed by experimental results for the most promising candidates. Separation of noble gases, krypton and xenon in particular, is a complex and costly industrial process with prices for xenon reaching $5000 per kg. Driven by the small computational demand of the single molecule analysis approach we employ our in-house software for hypothetical cage assembly, a new genetic algorithm and a structural analysis algorithm allowing for a “hands free” on-the-fly assessment of structural parameters such as the void and window diameters. This, combined with binding energies calculations, results in a cage structure optimization tool for a rational design of organic cage molecules for molecular separations and guest encapsulation (e.g. C60 binding).

Resume : The localised deformation of molecular monolayers constrained between the spherical surfaces of Au nanoparticles is studied by means of molecular dynamics simulations. Long-chain alkyl or polyethilene glycol molecules were densely distributed over the curved Au surface, and pushed against each other by repeated cycles of force relaxation and constant-volume equilibration, at temperatures increasing from 50 to 300K, before being slowly quenched down to near-zero temperature. Curves of minimum configurational energy can be obtained as a function of the nanoparticle distance, according to different directions of approach; therefore such simulations describe a range of deformations, from perfectly uniaxial compression, to a combination of com- pression and shear. The deformation is always found to be localised at the interface between the opposing molecular monolayers. We can also deduce the apparent Young and shear moduli of a dense nanostructure, composed by a triangular arrangement of identical MUDA- decorated Au nanoparticles, which is found to be smaller than estimates indirectly deduced by atomic- force experiments, but very close to previous computer simulations of molecular monolayers on flat surfaces, or bulk nanoparticle assemblies.

Resume : Understanding how molecular structure and temperature controls the molecular self-assembly process will enable the fabrication of desired nanomaterials. This presentation will introduce an original simulation technique for molecular self-assembly on metal surfaces called GAMMA modeling. GAMMA modeling deals with the long time-scales of the self-assembly process by replacing the model configuration space with a reduced space that only contains information on the shapes of the molecular assemblies that can form. As well as allowing for fast convergence rates in Monte Carlo simulation, this approach leads to a data visualization method that identifies how molecular structure and temperature determine the molecular assemblies that emerge from the self-assembly process. These points will be illustrated by considering self-assembly of substituted bianthracene molecules on Cu(111) surfaces, a system for which high-quality scanning tunneling microscopy data is available to verify model predictions.

Resume : Development of efficient and economic methods for fresh water production by desalination of salt water becomes a significant scientific and industrial challenge because of global water shortage. Separation of salt ions from seawater by stacked graphene membrane is an emerging desalination method. However, the structure and dynamics of water molecules intercalated between graphene sheets is not know, and this has proven to be a hindrance in understanding how this system functions. Therefore, we developed a model system composed of two parallel graphene sheets (25 Å × 20 Å) in a water box (44 Å × 44 Å × 44 Å), and performed molecular dynamics simulations of the system with SPC/E model for water, and the parameters obtained by MP2 method for the carbon-water nonbonded interaction. We varied the distance between graphene sheets for examining the diffusion of water molecules from the bulk phase to the confined space, and we found that the translational and rotational diffusion coefficients of confined water molecules depend on the distance between graphene sheets and the concentration of water. At different distances (8 – 12 Å) between graphene sheets, we also obtained the free energy profiles of water permeation through the slit of stacked graphene using steered molecular dynamics simulations and analysis with Jarzynski equation. We also obtained the free energy profiles of cations and anions of permeation, and we found that sodium ions would penetrate the stacked graphene membrane when the graphene-graphene distance is longer than 10 Å. We believe that our simulation results would be a significant contribution for designing a new stacked graphene- based membrane for desalination.

Resume : ß-amyloid (Aß) aggregate is an assembly built of 36 to 43 amino acids long peptides formed into tightly bound ß-sheets. Such folded Aß is the major component of pathological aggregates in the central nervous system, known as amyloid fibers, associated with Alzheimer’s disease. Numerous studies in the context of finding new therapeutics for that neurodegenerative disease mostly focus on conditions that favor the formation of amyloidogenic fibrils and aim at hindering them. A range of inhibitors (e.g. small organic molecules, modified peptides and peptidomimetics) have been extensively studied for modulating the peptide assembly structures. Fine structures of amyloid proteins provide a platform for rational drug design aiming at stabilizing or disrupting the assembly of Aß structures, and understanding interactions between drugs and targets. Hence, at the moment scientific attention and synthetic creativity have mostly focused on finding molecules that prevent aggregation of neurotoxic Aß assemblies to form. The aim of this work is to prop up the design of anti-amyloid drugs, applicable at the later stages of Aßopathy process, when aggregation-inhibiting approach is not efficient anymore. We have been working on a design of new azobenzene (azo) derivatives, which can act as molecular levers for the photo-controlled mechanical dissociation of neurotoxic Aß assemblies, and optimizing their structure-property characteristics using state-of-the-art computational techniques. The choice of azo-derivatives as the photo-controlled molecular levers has not been random, as they are known to undergo a well-de?ned isomerization from the trans to the cis conformation around the N=N double bond, in response to an external stimulus, such as UV or X-ray light, and vice versa if exposed to visible light or heating. Accordingly, azobenzenes have been used for a large variety of applications, including modulation of the interactions between azo-protected nanoparticles, liquid-crystal alignment for displays, optical switching and data storage devices. In particular, azobenzene derivatives hold potential to be incorporated in electronic switchable devices, working as photo-mechanical responsive systems. Azobenzene derivatives have been also used for biotechnological applications, like reversible photo-control of a bacterial enzyme activity. Moreover, recent study on zebrafish embryos provide a direct evidence of an azobenzene photo-switching in vivo. Noteworthy, the embryos with the azo-switches developed normally and no toxicity was evident monitoring growth alongside uninjected embryos. That results indicate that it is be possible, in general, to photo-control peptide and protein function in living systems, and constitute a big encouragement for designing new azo-derivatives for biomedical applications. Our idea was to check in silico whether it would be possible for a planar trans azobenzene derivative to intercalate between Aß layers.. If this is the case, the azo-molecule should disrupt some of the main-chain hydrogen bonds between the two ß-strands, as other small molecules usually do. The novelty of our idea is based on the ability of an azobenzene molecule to change conformation from trans to cis upon photo-excitation. We believe that nonplanar cis azo-molecule shall put a mechanical stress on the protein strands, eventually leading to the mechanical dissociation of the assembly. We have started our study by simulating the dynamic behavior of Aß complexes. For this purpose Molecular Mechanics (MM) and Dynamics (MD) methodology in all-atom representation and explicit solvent has been applied in conditions as close to native ones as possible. As a basis for simulations, structures of 42-aminoacids long Aß complexes available in Protein Data Bank were used. After modelling residues that were not resolved in experiments, two sets of simulations were performed per each Aß form, one showing the dynamics of Aß at the endings of the complex and the other – taking advantage of periodic representations of Aß macromolecules – in the midst of the complexes. The next step was to check the behavior of a ß-sheet structure upon twisting the azobenzene molecule, placed inside the Aß assembly. Our analysis mostly focus on the redistribution of the non-covalent interactions (i.e. hydrophobic interactions, hydrogen bonds, and electrostatic interactions) between azo-molecules and amyloid fibrils and their relative importance upon the photo-induced conformational changes. We have been introducing and checking several chemical structure modifications of the basic azobenzene molecule, adjusting it to the molecular character of the targeted protein. As first choices we functionalized azobenzene with groups present in other small molecules known for interacting with amyloid proteins and disturbing their aggregations (e.g. hydroxyl groups present in Morin flavonol).

Resume : Molecular junctions are perceived as the ultimate step toward the miniaturization of electronic components based on organic materials. Although the realization of integrated molecular devices remains a long-term goal, understanding the parameters influencing the charge transport through a single molecular bridge is a key step toward complex functionnal architectures. In this work, a scanning tunnelling microscope is used to pull a polythiophene wire from a Au(111) surface while measuring the current traversing the junction. Abrupt current increases measured during the lifting procedure are associated with the detachment of molecular subunits, in apparent contradiction with the expected exponential decrease of the conductance with wire length. In this presentation, I will show extensive density functional theory (DFT) simulations that reproduce the overall lifting procedure which have been performed to interpret the transport data. On the basis of the agreement between experiment and theory, the sudden increases of conductance are associated with releases of the stress applied on the suspended wire when thiophene units detach from the surface. This stress relaxation produces a gain in conjugation along the wire and a better electronic coupling with the electrodes, yielding a substantial increase of the transport efficiency of the wire junction, which overcomes the expected loss of conductance due to the wire elongation. This work opens the way to electronic devices made of single molecular wires whose transport properties can be tuned mechanically. [1] Pulling and Stretching a Molecular Wire to Tune its Conductance. G. Reecht, H. Bulou, F. Scheurer, V. Speisser, F. Mathevet, C. González, Y. J. Dappe, and G; Schull, J. Phys. Chem. Lett. 6, 2987 (2015). [2] Electroluminescence of a Polythiophene Molecular Wire Suspended between a Metallic Surface and the Tip of a Scanning Tunneling Microscope. G. Reecht, F. Scheurer, V. Speisser, Y. J. Dappe, F. Mathevet, and G. Schull, Phys. Rev. Lett. 112, 047403 (2014).

Resume : The primary purpose of this project is to study functionalized carbon nanotube (CNT)- polymer composites in order to propose a mechanism capable of measuring strain and/or monitoring failure of the composite element via changes in electronic conductivity. It has been shown that changes in bonding at the functionalized site of the CNT can cause dramatic changes in conductivity [1]. It is not clear whether this could provide a practical mechanism for monitoring the deformation of real polymeric-CNT composites. This is because the mechanism of load transfer from the polymer matrix to the functionalized site of the CNT is not known and whether the forces that are transmitted through the interface are sufficient to cause a change in conductivity is still an open question. We investigate this question using a range of computational methodologies including a combination of first-principles quantum-mechanical simulations and classical Molecular Dynamics (MD). Large scale MD is used to simulate a model CNT-polymer composite element consisting of a functionalized CNT embedded in plolyethylene while Density-Functional Theory is employed to focus on the functionalized CNT in isolation. Applicability of hybrid methods for combining these two approaches will also be investigated. [1] Li, E. Y., Poilvert, N., & Marzari, N. (2011). ACS Nano, 5(6), 4455–4465.

Resume : Metal oxide materials exhibit a wide range of electronic, optical, chemical and magnetic properties and find diverse technological applications in areas such as electronics, energy generation, catalysis and medicine. Whether in the form of thin crystalline films, nanoparticles or bulk polycrystals extended defects such as grain boundaries and dislocations are ubiquitous in metal oxides. These defects often significantly perturb structural and electronic properties affecting functionality and performance for many applications. In recent work we have shown how first principles materials modelling can provide invaluable insights into the role of extended defects which are often challenging to unravel by experiment alone. In this talk, I will present a number of examples including grain boundaries and dislocations in the oxides MgO, TiO2, Fe3O4 and HfO2 with relevance to applications in electronics, photovoltaics, magnetism and spintronics [1-7]. [1] K. P. McKenna et al, Nature Communications 5 (2014) 5740. [2] S. Wallace and K. P. McKenna, Advanced Materials Interfaces 1 (2014) 1400078. [3] Z-C. Wang, M. Saito, K. P. McKenna, Y. Ikuhara, Nature Communications 5 (2014) 3239. [4] K. P. McKenna, Journal of the American Chemical Society 135 (2013) 18859. [5] Z. Wang et al, Nature 479 (2011) 380. [6] K. P. McKenna and A. L. Shluger, Applied Physics Letters 95 (2009) 222111. [7] K. P. McKenna and A. L. Shluger, Nature Materials 7 (2008) 859.

Resume : Hydrogen is a promising candidate for the clean energy carrier that may replace fossil fuels. For the production of hydrogen, water splitting with efficient catalysts has been intensively studied over the past decades. Platinum, which is known to be the best catalyst for water splitting, is too expensive to be used in large-scale applications. Therefore, numerous earth-abundant materials have been investigated as a replacement of Pt. Recently, transition metal dicalcogenides (TMDs), most notably MoS2, are receiving a great deal of attention as a novel catalyst for water splitting. Although the basal plane of TMDs are efficient as catalysts, it was found recently that the sulfur vacancy in MoS2 can increase the catalytic activity for hydrogen evolution. In this presentation, motivated by the previous work, we explore the detailed mechanism for hydrogen production from the sulfur vacancy in MoS2 and calculate the activation energies along the reaction path. Furthermore, we evaluate the catalytic efficiency of vacancy sites in various TMDs and suggest TMDs that may show high catalytic effects in hydrogen evolution reaction.

Resume : Porous solids have been widely studied for their applications in heterogeneous catalysis, gas storage, and molecular separations. These materials include extended networks such as zeolites, metal-organic frameworks (MOFs), covalent organic frameworks (COFs), and porous aromatic frameworks (PAFs). In addition, porous solids self-assembled from discrete molecules have sparked recent interest because of the intrinsically higher solubility of small molecules compared with extended networks, which makes these materials easier to process. This presentation will discuss our work on the computational design of new porous organic cages for enhanced gas storage in amorphous solids [1], high-throughput screening of crystalline porous molecular solids for gas storage [2], and coarse-grained modelling of the effects of chemical functionality on the mesoscale self-assembly and porosity of porous aromatic frameworks. References: [1] J.D. Evans, D.M. Huang, M.R. Hill, C.J. Sumby, D.S. Sholl, A.W. Thornton, C.J. Doonan, J. Phys. Chem. C 119, 7746–7754 (2015) [2] J.D. Evans, D.M. Huang, M. Haranczyk, A.W. Thornton, C.J. Sumby, C.J. Doonan, CrystEngComm (2016), in press, doi: 10.1039/C6CE00064A

Resume : Chemical warfare agents (CWAs) are chemicals that are very toxic and lethal to humans in milligram quantities. Despite global conventions that prohibit their use in combat situations,1 there are reports of their ongoing use in Syria.2 The interactions of these compounds with porous materials found in the field, such as sand, brick, soil, etc. are not well understood and so molecular simulations are an ideal way to investigate the chemistry of these very toxic materials at an atomistic level. The current study is designed to model the sorption and diffusion of CWAs within a variety of porous materials through the application of molecular dynamics and grand canonical Monte Carlo (GCMC) simulations. The modelling studies being conducted are being correlated with experimental measurements, in particular solid state NMR, on the same materials. One material upon which this study has thus far focused is an amorphous conjugated microporous organic polymer, known as CMP-8.3 CMP-8 has been selected as a model substrate to mimic the chemical functionalities found in organic rich soils, mainly aromatic phenols. Presented here is the process of generating a realistic model of CMP-8 using the Polymatic software for generating amorphous polymers released by the Colina Group.4 Various attempts have been made to generate realistic structure models, validated through surface area and pore size measurements, as well as using GCMC simulated adsorption of nitrogen. Copyright Imperial College London 2016 1 http://www.opcw.org/ 2 http://www.reuters.com/article/2013/09/16/us-syria-crisis-un-idUSBRE98F0ED20130916# 3 R. Dawson, A. Laybourn, R. Clowes, Y. Z. Khimyak, D. J. Adams and A. I. Cooper, Macromolecules, 2009, 42, 8809–8816. 4 L. J. Abbott, K. E. Hart and C. M. Colina, Theor. Chem. Acc., 2013, 132, 1334.