Abstract. It is essential to know the arrangement of the atoms in a material in order to compute and understand its properties. Searching for stable structures of materials using first-principles electronic structure methods, such as density functional theory (DFT), is a rapidly growing field. Here we describe our simple, elegant and powerful approach to searching for structures with DFT which we call ab initio random structure searching (AIRSS). Applications to discovering structures of solids, point defects, surfaces, and clusters are reviewed. New results for iron clusters on graphene, silicon clusters, polymeric nitrogen, hydrogen-rich lithium hydrides, and boron are presented.

Hydrogen, being the first element in the periodic table, has the simplest electronic structure of any atom, and the hydrogen molecule contains the simplest covalent chemical bond. Nevertheless, the phase diagram of hydrogen is poorly understood. Determining the stable structures of solid hydrogen is a tremendous experimental challenge 1-3 , because hydrogen atoms scatter X-rays only weakly, leading to low-resolution diffraction patterns. Theoretical studies encounter major difficulties owing to the small energy differences between structures and the importance of the zero-point motion of the protons. We have systematically investigated the zerotemperature phase diagram of solid hydrogen using firstprinciples density functional theory (DFT) electronic-structure methods 4 , including the proton zero-point motion at the harmonic level. Our study leads to a radical revision of the DFT phase diagram of hydrogen up to nearly 400 GPa. That the most stable phases remain insulating to very high pressures eliminates a major discrepancy between theory 5 and experiment 6 . One of our new phases is calculated to be stable over a wide range of pressures, and its vibrational properties agree with the available experimental data for phase III.The low-pressure phase I of solid hydrogen, which consists of freely rotating molecules on a hexagonal close-packed lattice 2 , transforms at pressures of about 110 GPa to the broken-symmetry phase II, in which the mean molecular orientations are ordered, and then to phase III at about 150 GPa (ref. 1). However, even the combination of X-ray and neutron scattering data and Raman and infrared vibrational data has not so far yielded the structures of phases II and III of hydrogen.The theoretical prediction of stable crystal structures is very difficult because of the need to search the very large space of possible structures, and the necessity of obtaining accurate energies for each of these structures. First-principles DFT methods have proved an efficient means of calculating quite accurate energies, and they have provided many insights into the properties of materials, including solid hydrogen 5,7 . At present, DFT offers the highest level of theoretical description at which we can carry out searches over many possible candidate structures.Our approach is to relax many random structures to minima in the enthalpy at fixed pressure 8 . This method does not rely on previous theoretical or experimental results, and it allows for the possibility of finding radically new structures. In some cases we used the intuition gained from the random searches to build other candidate structures. We then calculated the enthalpies of the most stable phases at a larger number of pressures, generating the data shown in Fig. 1. We refer to each structure by its short HermannMauguin space-group symbol, giving additional information where an ambiguity might occur.The lowest-enthalpy structures found around 100 GPa were those of space groups Pca2 1 and P2 1 /c, which were considered in previous studies 5,7 , and ...

Room-temperature superconductivity has been a long-held dream and an area of intensive research. Recent experimental findings of superconductivity at 200 K in highly compressed hydrogen (H) sulfides have demonstrated the potential for achieving room-temperature superconductivity in compressed H-rich materials. We report first-principles structure searches for stable H-rich clathrate structures in rare earth hydrides at high pressures. The peculiarity of these structures lies in the emergence of unusual H cages with stoichiometries H_{24}, H_{29}, and H_{32}, in which H atoms are weakly covalently bonded to one another, with rare earth atoms occupying the centers of the cages. We have found that high-temperature superconductivity is closely associated with H clathrate structures, with large H-derived electronic densities of states at the Fermi level and strong electron-phonon coupling related to the stretching and rocking motions of H atoms within the cages. Strikingly, a yttrium (Y) H_{32} clathrate structure of stoichiometry YH_{10} is predicted to be a potential room-temperature superconductor with an estimated T_{c} of up to 303 K at 400 GPa, as derived by direct solution of the Eliashberg equation.

High-pressure phases of silane SiH4 are predicted using first-principles electronic structure methods. We search for low-enthalpy structures by relaxing from randomly chosen initial configurations, a strategy which is demonstrated to work well for unit cells containing up to at least ten atoms. We predict that silane will metallize at higher pressures than previously anticipated but might show high-temperature superconductivity at experimentally accessible pressures.

We use first principles calculations to study structural, vibrational and superconducting properties of H2S at pressures P ≥ 200 GPa. The inclusion of zero point energy leads to two different possible dissociations of H2S, namely 3H2S → 2H3S + S and 5H2S → 3H3S + HS2, where both H3S and HS2 are metallic. For H3S, we perform non-perturbative calculations of anharmonic effects within the self-consistent harmonic approximation and show that the harmonic approximation strongly overestimates the electron-phonon interaction (λ ≈ 2.64 at 200 GPa) and Tc. Anharmonicity hardens H-S bond-stretching modes and softens H-S bond-bending modes. As a result, the electronphonon coupling is suppressed by 30% (λ ≈ 1.84 at 200 GPa). Moreover, while at the harmonic level Tc decreases with increasing pressure, the inclusion of anharmonicity leads to a Tc that is almost independent of pressure. High pressure hydrogen sulfide is a strongly anharmonic superconductor.Cuprates [1] have for many years held the world record for the highest superconducting critical temperature (T c = 133 K) [2]. However, despite almost 30 years of intensive research, the physical mechanism responsible for such a high T c is still elusive, although the general consensus is that it is highly non-conventional. The discovery by Drozdov et al.[3] of T c = 190 K in a diamond anvil cell loaded with hydrogen sulfide (H 2 S) and compressed to about 200 GPa breaks the cuprates record and overturns the conventional wisdom that such a high T c cannot be obtained via phonon-mediated pairing.The claim that hydrogen at high pressure could be superconducting is not new [4] and it was recently supported by first principles calculations based on the harmonic approximation applied to dense hydrogen [5][6][7][8] and several hydrides [9][10][11][12][13][14][15]. More recently, two theoretical papers predicted the occurrence of high T c superconductivity in high-pressure sulfur-hydrides [16,17]. However, as shown in Refs. [18,19], anharmonicity can be crucial in these systems. For example, in PdH, the electron-phonon coupling λ parameter is found to be 1.55 at the harmonic level, while a proper inclusion of anharmonic effects leads to λ = 0.40 [18], in better agreement with experiments. Thus, in hydrogen-based compounds, the phonon spectra are strongly affected by anharmonic effects.Several first principles calculations [16,17,20,26] suggested that decomposition of the H 2 S sample occurs within the diamond-anvil cell at high pressures. The high-T c superconducting material is therefore very unlikely to be H 2 S, while H 3 S is the obvious candidate for the H-rich decomposition product.Here we study the structural, vibrational and superconducting properties of H 2 S above 200 GPa, where the highest T c occurs. We show that the inclusion of zero point motion in the convex hull at 200 and 250 GPa stabilizes two metallic structures, H 3 S and HS 2 . Finally, we show that, contrary to suggestions in previous work [16,20], the harmonic approximation does not explain the measured T c ...

The results of the sixth blind test of organic crystal structure prediction methods are presented and discussed, highlighting progress for salts, hydrates and bulky flexible molecules, as well as on-going challenges.

In this paper we investigate the construction of nomanserving 'soft-core' pseudopotentials with improved convergence properties of lhe plane-wave and penurbation expansions. The key factor is found to be the kinelic energy of the valence pseudo-orbitals. l h e total kinetic energy controls the convergence of lhe perturbation expansion 01 thc total energy, the kinelic energy contained in the Fourier mmponents beyond a cerlain cut-off limits the convergence or the plane-wave expansion. The simultaneous optimization of both expansions allows us lo use the Same pseudopotential in a rapidly convergent lotal-energy calculation for the nyslalline phases, and in the calculation of interatomic forces to be used in atomistic simulations of the disordered phases.

Progress in the discovery of new materials has been accelerated by the development of reliable quantum-mechanical approaches to crystal structure prediction. The properties of a material depend very sensitively on its structure, therefore structure prediction is the key to computational materials discovery. Structure prediction was considered to be a formidable problem, but the development of new computational tools has allowed the structures of many new and increasingly complex materials to be anticipated. These widely applicable methods, based on global optimisation and relying on little or no empirical knowledge, have been used to study crystalline structures, point defects, surfaces and interfaces. In this Review we discuss structure prediction methods, examining their potential for the study of different materials systems, and present examples of computationally-driven discoveries of new materials -including superhard materials, superconductors and organic materials -that will enable new technologies. Advances in firstprinciple structure predictions also lead to a better understanding of physical and chemical phenomena in materials.