main-content

This book provides a unique and comprehensive overview of the latest advances, challenges and accomplishments in the rapidly growing field of theoretical and computational materials science. Today, an increasing number of industrial communities rely more and more on advanced atomic-scale methods to obtain reliable predictions of materials properties, complement qualitative experimental analyses and circumvent experimental difficulties. The book examines some of the latest and most advanced simulation techniques currently available, as well as up-to-date theoretical approaches adopted by a selected panel of twelve international research teams. It covers a wide range of novel and advanced materials, exploring their structural, elastic, optical, mass and electronic transport properties. The cutting-edge techniques presented appeal to physicists, applied mathematicians and engineers interested in advanced simulation methods in materials science. The book can also be used as additional literature for undergraduate and postgraduate students with majors in physics, chemistry, applied mathematics and engineering.

### Chapter 1. Making Computer Materials Real: The Predictive Power of First-Principles Molecular Dynamics

Abstract
First-principles molecular dynamics (FPMD) is a well-established method to study materials at the atomic scale by taking advantage of three ingredients: the laws of statistical mechanics, the theoretical foundations of density functional theory and powerful computers. FPMD does its best when the atomic structures are unknown or poorly known and when their time trajectories are required to extract, via statistical averages, a thermodynamical evolution as a function of temperature. In this paper, key concepts of molecular dynamics are recalled and made simple, by insisting on the proper use of some definitions and by showing, via a prototypical example, the sensitivity to a crucial part (the exchange-correlation energy) of the total energy functional.
Carlo Massobrio, Mauro Boero, Sébastien Le Roux, Guido Ori, Assil Bouzid, Evelyne Martin

### Chapter 2. Assessing the Versatility of Molecular Modelling as a Strategy for Predicting Gas Adsorption Properties of Chalcogels

Abstract
Modelling gas adsorption of porous materials is nowadays an undeniable necessary in order to complement experiment findings with the purpose to enrich our fundamental understanding of adsorption mechanisms as well as develop better performing materials for gas mixture separation. In this contribution, we explore the possibility to use first-principles molecular dynamics (FPMD) and grand canonical Monte Carlo (GCMC) simulations to target the gas adsorption of disordered nanoporous chalcogenides (i.e. chalcogels). This computational scheme allows us to take advantage of the ability of FPMD to accurately describe the structure and bonding of the disordered nature of chalcogels as well as the potential of GCMC to model the adsorption mechanisms of porous networks. We assess the versatility of such scheme by evaluating the role of pore size, chemical stoichiometry and composition for multiple chalcogenide-based systems on nitrogen adsorption isotherms.
Iréné Berenger Amiehe Essomba, Carlo Massobrio, Mauro Boero, Guido Ori

### Chapter 3. Exploring Defects in Semiconductor Materials Through Constant Fermi Level Ab-Initio Molecular Dynamics

Abstract
We focus on the determination of point defects in semiconductor materials through constant-Fermi-level ab initio molecular dynamics and demonstrate that this technique can be used as a computer-based tool to reveal and control relevant defects in semiconductor materials. In this scheme, the Fermi level can be set at any position within the band gap during the defect generation process, in analogy to experimental growth conditions in the presence of extra electrons or holes. First, the scheme is illustrated in the case of GaAs, for which we generate melt-quenched amorphous structures through molecular dynamics at various Fermi levels. By a combined analysis that involves both the atomic structure and a Wannier-function decomposition of the electronic structure, we achieve a detailed description of the generated defects as a function of Fermi level. This leads to the identification of As–As homopolar bonds and Ga dangling bonds for Fermi levels set in the vicinity of the valence band. These defects convert into As dangling bonds and Ga–Ga homopolar bonds, as the Fermi level moves toward the conduction band. Second, we investigate defects at the InGaAs/oxide interface upon inversion. We adopt a substoichiometric amorphous model for modelling the structure at the interface and investigate the formation of defect structures upon setting the Fermi-level above the conduction band minimum. Our scheme reveals the occurrence of In and Ga lone-pair defects and As–As dimer/dangling bond defects, in agreement with previous studies based on physical intuition. In addition, the present simulation reveals hitherto unidentified defect structures consisting of metallic In–In, In–Ga, and Ga–Ga bonds. The defect charge transition levels of such metallic bonds in Al$$_2$$O$$_3$$ are then determined through a hybrid functional scheme and found to be consistent with the defect density measured at InGaAs/Al$$_2$$O$$_3$$ interfaces. Hence, we conclude that both In and Ga lone pairs and metallic In–In bonds are valid candidate defects for charge trapping at InGaAs/oxide interfaces upon charge carrier inversion. These two studies demonstrate the effectiveness of constant-Fermi-level ab initio molecular dynamics in revealing and identifying semiconductor defects in an unbiased way.
Assil Bouzid, Alfredo Pasquarello

### Chapter 4. Enhancing the Flexibility of First Principles Simulations of Materials via Wavelets

Abstract
We illustrate how the properties of a Daubechies wavelet basis set can be exploited to build an effective computational method that enables one to perform electronic structure calculations of systems containing up to many thousands of atoms. This is achieved by implementing a ladder of approaches of different scaling behaviours and decreasing computational complexity. We will explain that such an approach is suitable both for extended systems and for systems with molecular character. We define quantitative indicators that provide guidelines to the end-user about the pertinence of the employed methodology, thereby guaranteeing limited impact on the precision of the result. We provide a quantitative illustration of these concepts to defective systems with an extended character, by presenting the differences in computational walltime and in precision among the various methodological steps of the ladders.
Laura E. Ratcliff, Luigi Genovese

### Chapter 5. Self-consistent Hybrid Functionals: What We’ve Learned So Far

Abstract
These are exciting times for computational materials science. We are witnessing a wide-spread availability of high-performance computing facilities, a huge increase in accessible computational resources, and an accompanying development of new exchange-correlation functionals within density functional theory. All this contributes to the establishment of density functional theory as an indispensable tool for materials science investigations in general. Here, we want to highlight some examples utilising a recently developed self-consistent hybrid functional, proposed by Shimazaki and Asai (J Chem Phys 130:164702, 2009 [3]) and Skone et al. (Phys Rev B 89:195112, 2014 [4]), allowing for the calculation of accurate material properties using a fully ab initio procedure. The obtained structural and electronic properties of a range of oxide semiconductors will be critically discussed with respect to experimental results, and pave the way towards open questions in the field.
Daniel Fritsch

### Chapter 7. Diffusion Kinetics in Binary Liquid Alloys with Ordering and Demixing Tendencies

Abstract
Theoretical relationship between collective and tracer diffusion coefficients has been derived and tested for different types of binary melts: (i) with an ordering tendency (case study on Ni–Al and Ni–Zr melts) and (ii) with a demixing tendency (case study on Cu–Ag melts). The obtained relationship explicitly demonstrates microscopic cross-correlation effects in the kinetics of collective diffusion. Our approach incorporates molecular dynamics calculations, modelling and statistical mechanical analysis based on fundamental concepts of the fluctuation-dissipation theorem, generalized Langevin equation and Mori-Zwanzig formalism. We also applied the developed theory to interpret recent available experimental data as well as our molecular dynamics data of diffusion kinetics in different types of binary melts: with chemical ordering and contrarily with demixing tendency.
Andreas Kromik, E. V. Levchenko, Alexander V. Evteev

### Chapter 8. Advanced Monte Carlo Simulations for Ion-Channeling Studies of Complex Defects in Crystals

Abstract
This chapter describes the most important features of computational software called ‘McChasy’, which is a Monte Carlo (MC) simulation code developed for the evaluation of the Rutherford Backscattering Spectrometry data, in particular, recorded in the channeling mode (RBS/C). RBS/C is an experimental technique used in the analysis of defects in single crystals. Lattice distortions affect materials modified by ion beams or exposed to irradiation. Therefore, the analysis of damage in crystals is of high importance in materials science. Various types of defects can be created due to the interaction of ions with targets. However, RBS/C has different sensitivity to each of them so the analytical analysis of experimental data is hardly possible. MC simulations are a powerful tool used to overcome this limitation. The McChasy code simulates the movement of light ions in crystals. The software provides a fitting procedure of RBS/C spectra based on independent depth profiles of different defect types: interstitials, edge dislocations, substitutions, stacking faults or grain boundaries. The code works well not only with materials containing complex defects but also with heterostructures and superlattices. Recent improvements of the code include a unique approach of 3D-interaction between ions and target atoms. Application of the McChasy code in the analysis of crystal defects is described and possible ways of its further development are pointed out.
Przemyslaw Jozwik, Lech Nowicki, Renata Ratajczak, Cyprian Mieszczynski, Anna Stonert, Andrzej Turos, Katharina Lorenz, Eduardo Alves

### Chapter 9. Electronic and Optical Properties of Polypyrrole as a Toxic Carbonyl Gas Sensor

Abstract
Chronic formaldehyde and acetaldehyde exposure is known to cause various health problems and the detection of these toxic carbonyl gases is very important and a subject of interest both experimentally and theoretically. In this study, the interaction of various oligomers (n = 1, 3, 5, 7, and 9) of polypyrrole towards toxic carbonyl species: acetaldehyde and formaldehyde, and less toxic carbonyl species: acetone and butanone were studied using density functional theory (DFT). The interactions of the carbonyl species with oligopyrrole lead to differences in interaction energies and changes in structural features: H-bond distances, bond angles, and dihedral angles. The changes resulted in variations in the electronic properties of the pyrrole-gas complexes: HOMO/LUMO energies, ionization potentials (IP), electron affinity (EA), and energy gap (EGap). The pyrrole-carbonyl complexes resulted in higher HOMO energies due to electron charge donation from the carbonyl gases and lower LUMO energies resulting in smaller EGap values compared to pyrrole. It was observed that the smallest carbonyl molecule, formaldehyde (For), had the lowest LUMO energy and lowest EGap value, while the largest carbonyl molecule, butanone (MEK), had the highest LUMO energy and highest EGap value. Furthermore, simulated UV–Vis absorption studies showed red-shifted first singlet excited state, λ1st, for the pyrrole-gas complexes. The results do not only demonstrate the potential of polypyrrole as a toxic carbonyl gas sensor but also its selectivity towards different carbonyl species.
Francisco C. Franco

### Chapter 10. Thermoelectric Power Factor Under Strain-Induced Band-Alignment in the Half-Heuslers NbCoSn and TiCoSb

Abstract
Band convergence is an effective strategy to improve the thermoelectric performance of complex bandstructure thermoelectric materials. Half-Heuslers are good candidates for band convergence studies because they have multiple bands near the valence bad edge that can be converged through various band engineering approaches providing power factor improvement opportunities. Theoretical calculations to identify the outcome of band convergence employ various approximations for the carrier scattering relaxation times (the most common being the constant relaxation time approximation) due to the high computational complexity involved in extracting them accurately. Here, we compare the outcome of strain-induced band convergence under two such scattering scenarios: (i) the most commonly used constant relaxation time approximation and (ii) energy dependent inter- and intra-valley scattering considerations for the half-Heuslers NbCoSn and TiCoSb. We show that the outcome of band convergence on the power factor depends on the carrier scattering assumptions, as well as the temperature. For both materials examined, band convergence improves the power factor. For NbCoSn, however, band convergence becomes more beneficial as temperature increases, under both scattering relaxation time assumptions. In the case of TiCoSb, on the other hand, constant relaxation time considerations also indicate that the relative power factor improvement increases with temperature, but under the energy dependent scattering time considerations, the relative improvement weakens with temperature. This indicates that the scattering details need to be accurately considered in band convergence studies to predict more accurate trends.
Chathurangi Kumarasinghe, Neophytos Neophytou

### Chapter 11. Prediction of Energy Gaps in Graphene—Hexagonal Boron Nitride Nanoflakes Using Artificial Neural Networks

Abstract
Machine learning methods are currently applied in conjunction with ab initio density functional theory (DFT) simulations in order to establish computationally efficient alternatives for high-throughput processing in atomistic computations. The proposed method, based on artificial neural networks (ANNs), was used to predict the HOMO-LUMO energy gap in quasi-0D graphene nanoflake systems with randomly generated boron nitride embedded regions. Several artificial neural network (ANN) algorithms were tested in order to optimize the network parameters for the problem at hand. The trained ANNs prove to be computationally efficient at determining the energy gap with good accuracy and show a significant speedup over the classical DFT approach.
Tudor Luca Mitran, George Alexandru Nemnes

### Chapter 12. Hydrogen in Silicon: Evidence of Independent Monomeric States

Abstract
The data on hydrogen in saturated/quenched samples and in samples exposed to plasma have been revisited. It is concluded that the monomeric hydrogen in intrinsic silicon is represented mostly by two neutral species: Hb (presumably a ground state of tetrahedral hydrogen) and Hs (a slow monomer in a different interstitial position). At high T these species are in equilibrium, with a concentration ratio close to 1. At lower T (at least at T ≤ 500 °C) they become independent one of the other. This conclusion differs from a conventional notion that considers bond-centred H+(BC) ions to be dominant in intrinsic Si. In p-Si, boron is passivated not only by H+(BC) ions (denoted H+(1)) but also by another kind of H+ denoted H+(2). A presence of several independent species (Hb, Hs, H+(1) and H+(2)) gives rise to a rich variety of hydrogen depth profiles in plasma-exposed silicon; these profiles are well reproduced by simulations.
V. V. Voronkov

### Chapter 13. Architecture and Function of Biohybrid Solar Cell and Solar-to-Fuel Nanodevices

Abstract
In recent years, an immense research effort has been devoted to the generation of hybrid materials which change the electronic properties of one constituent by changing the optoelectronic properties of the other one. The most appealing and commonly used approach to design such materials relies on combining organic materials or metals with biological systems like redox-active proteins. Such hybrid systems can be used e.g. as bio-sensors, bio-fuel cells, biohybrid photoelectrochemical cells and nanosctuctured photoelectronic devices. Although experimental efforts have already resulted in the generation of a number of biohybrid materials, the main bottleneck of this technology is the formation of a stable and efficient (in terms of electronic communication) interface between the biological and the organic/metal counterparts. In particular, the efficiency of the final devices is usually very low due to two main problems related with the interfacing of such different materials: charge recombination at the interface and the high possibility of losing the function of the biological component, which leads to the inactivation of the entire device. In this chapter, we explore the power of computation to answer pressing questions for a rational design of the different components of the biohybrid interface.
Silvio Osella, Joanna Kargul, Miriam Izzo, Bartosz Trzaskowski

### Chapter 14. Mathematical Modeling of the Kinetics of Counter Diffusion During the Formation of Boron-Containing Coatings on Steels

Abstract
For the first time, a mathematical model of the kinetics of the formation of diffusion boride coatings on iron has been created, allowing to calculate the increase in the dimensions of parts due to the displacement of the outer surface of the saturated sample. Models of diffusant distribution in boride phases of different chemical composition are obtained. The experimental results confirmed the validity of the developed models.
A. S. Borsyakov, V. A. Yuryev, V. V. Ozyerelyev, E. V. Levchenko