Abstracts of individual topics to be covered at the workshop can be found below.

**Stefano de Gironcoli **^{1,2}

^{1} SISSA, Via Beirut 2/4, 34151 Trieste, Italy

^{2} IOM-DEMOCRITOS, Via Beirut 2/4, 34151 Trieste, Italy

QUANTUM ESPRESSO [1] is an integrated suite of computer codes for electronic-structure calculations and materials modeling, based on density-functional theory, plane waves, and pseudopotentials (norm-conserving, ultrasoft, and projector-augmented wave). The acronym ESPRESSO stands for opEn Source Package for Research in Electronic Structure, Simulation, and Optimization. It is freely available [2] to researchers around the world under the terms of the GNU General Public License.

Innovation and efficiency are still its main focus, with special attention paid to massively parallel architectures, and a great effort being devoted to user friendliness. QUANTUM ESPRESSO is evolving towards a distribution of independent and interoperable codes in the spirit of an open-source project, where researchers active in the field of electronic-structure calculations are encouraged to participate in the project by contributing their own codes or by implementing their own ideas into existing codes.

A QE-forge web portal [3] offering an integrated software development environment has been created and is maintained to promote scientific software development projects, not necessarily connected to Quantum ESPRESSO. Researchers participating in the QE-forge initiative retain complete control over their project (including the right not to release it), while enjoying the advanced development tools that QE-forge provides: CVS or SVN repository, mailing lists, public forums, download space, wiki pages and much more.

A general overview of the Quantum ESPRESSO suite of codes will be given, describing the functionalities of its main components, illustrating them by a few representative applications. The installation procedure will be described, as well as some details of the parallelization strategies used to exploit modern HPC architectures.

__References__

[1] P. Giannozzi, *et al.*, J. Phys.: Condens. Matter 21, 395502 (2009). [ link to article ]

[2] www.quantum-espresso.org

[3] www.qe-forge.org

**Arrigo Calzolari **^{1}

^{1} Democritos National Simulation Center, CNR-IOM Istituto Officina dei Materiali, Trieste, Italy.

Recent combinatorial biotechnologies have shown that the molecular recognition capability of proteins can be specifically oriented toward inorganic surfaces. However, at present the principles regulating protein-surface interactions are poorly understood. In addition, the exact role of the water molecules belonging to the protein solvation shell in the process of protein-surface interaction is unknown. The comprehension of the microscopic mechanisms that regulate protein-surface interaction, and the description of the protein hydration layer structure close to the surface would foster several technological applications, ranging from biomaterials to nanobioelectronics.

In order to address some of the afore mentioned problems, we present a first-principles description of the interaction between a prototype protein molecule (serin-based beta-sheet) and a gold surface, by explicitly taking into account the hydration layer in a liquid environment surrounding the molecule and the surface. In particular, we performed extensive (several ps) ab initio molecular dynamics (Car-Parrinello) simulations at finite temperature, using plane-wave Density Functional Theory. Our results [1] on the microscopic structure of the hydration layer, and its modifications induced by confinement effects close to the gold surface unravel the role of the interfacial water [2] in facilitating peptide adsorption. In particular, the hydrated beta-sheet dynamically adheres to gold surface, recognizing selectively the atomic corrugation of the substrate.

Such calculations represent a huge computational effort, due to the size of the studied systems (hundreds of atoms, thousands of electrons) and the need for meaningful statistics (hundreds of thousands of molecular dynamics steps). This computational grand-challenge requires extreme supercomputing resources, (such as those made available by DEISA-DECI) and high performing parallel codes, as quantum espresso.

__References__

[1] A. Calzolari, G. Cicero, C. Cavazzoni, R. Di Felice, A. Catellani and S. Corni, "Hydroxyl-rich beta-sheet adhesion to the gold surface in water by first-principle simulations", J. Am. Chem. Soc. **132**, 4790 (2010). [ link to article ]

[2] G. Cicero, S. Corni, A. Calzolari and A. Catellani, "Anomalous wetting layer at liquid water/gold interface", preprint (2010).

**Giovanni la Penna **^{1}

^{1} National research council, Institute for Chemistry of Organo-Metallic Compounds, via Madonna del Piano 10, I-50019 Sesto fiorentino (Firenze), Italy.

Weak interactions between transition metal ions and proteins play en essential role in biological processes influenced by structural disorder and bond fluxionality.

Because of the efficient implementation of ultra-soft pseudopotential technique, the Quantum-Espresso (QE) package provides many tools for calculation, elaboration and analysis of accurate electron structure and atomic forces in samples of condensed matter [1,2]. The high level of density-functional theory, the robust numerical algorithms and the efficient parallelization of the QE code, allow first-principle molecular dynamics simulations of large complexes in the presence of explicit water solvent, thus including essential interactions and chemical exchange into the models.

In this contribution I will describe the set-up and analysis of several models for Zn and Cu binding to fragments of the amyloid peptide responsible of the Alzheimer's disease [3]. The systems are in the range of 900-1400 atoms. The calculations have been performed with 1024 parallel tasks of the IBM-BlueGene/P architecture in Juelich (Jugene) and 128 tasks the Cray XT4/XT5 architecture in Helsinki (Louhi) [4]. Both architectures are accessible within the Deisa european infrastructure.

__References__

[1] P. Giannozzi *et al.*, J. Chem. Phys., **120**, 5903 (2004). [ link to article ]

[2] P. Giannozzi *et al.*, J. Phys.: Condens. Matt., **21**, 395502 (2009). [ link to article ]

[3] S. Furlan *et al.*, Phys. Chem. Chem. Phys., **11**, 6468-6481 (2009). [ link to article ]

[4] http://www.deisa.eu/science/deci/projects2008-2009/BICaPS

**Stefano de Gironcoli **^{1,2}

^{1} SISSA, Via Beirut 2/4, 34151 Trieste, Italy

^{2} IOM-DEMOCRITOS, Via Beirut 2/4, 34151 Trieste, Italy

After a general description of the input structure of the main codes in the QUANTUM ESPRESSO suite [1], a step-by-step description of how to perform basic calculations is given, from simple fixed atomic-positions self-consistent total energy calculations, to the determination of the electronic band structure across the BZ, to structural optimization, both at fixed cell geometry or including variable cell shape optimization. In each case a minimal introduction to the relevant theoretical concepts will be given. The relevant input variables to be specified will be introduced and illustrated by some practical hands-on examples.

__References__

[1] P. Giannozzi, *et al.*, J. Phys.: Condens. Matter 21, 395502 (2009). [ link to article ]

**Andrea Dal Corso **^{1,2}

^{1} SISSA, Via Beirut 2/4, 34151 Trieste, Italy

^{2} IOM-DEMOCRITOS, Via Beirut 2/4, 34151 Trieste, Italy

In the harmonic approximation, the dynamical matrices of a molecule or a solid are given, within the Born-Oppenheimer approximation, by the second derivatives of the total electron energy with respect to atomic displacements. These second derivatives are calculated very efficiently within density functional perturbation theory (DFPT).[1,2] No supercell is required to get the phonon frequencies at finite wavelength. In this lecture, I will present the fundamental ideas of DFPT applied to the calculation of the dynamical matrices of molecules and solids.[3] Within the Quantum ESPRESSO package, DFPT is implemented in the ph.x code which is illustrated in the practical section and used in a few examples. Its input and output are discussed in detail. To get the phonon dispersions, a mesh of wave vectors is introduced and, for each wave vector, the response to 3 × Nat different modes (here Nat is the number of atoms in one unit cell) is calculated. The response calculations can be carried out almost independently and I will show how to exploit this intrinsic parallelism to divide the total effort into simpler independent tasks.[4]

__References__

[1] S. Baroni, S. de Gironcoli, A. Dal Corso, and P. Giannozzi, Rev. Mod. Phys. **73**, 515 (2001). [ link to article ]

[2] A. Dal Corso, Phys. Rev. B 64, 235118 (2001); A. Dal Corso, Phys. Rev. B **81**, 075123 (2010). [ link to article ]

[3] A. Dal Corso, Introduction to density functional perturbation theory. [ http://people.sissa.it/∼dalcorso/lectures.html ]

[4] R. di Meo, A. Dal Corso, P. Giannozzi, and S. Cozzini, ICTP lecture notes **24**, 163 (2009). [ link to article ]

**Stefano de Gironcoli **^{1,2}

^{1} SISSA, Via Beirut 2/4, 34151 Trieste, Italy

^{2} IOM-DEMOCRITOS, Via Beirut 2/4, 34151 Trieste, Italy

A number of post-processing tools present in the QUANTUM ESPRESSO suite of codes [1] will be described and demonstrated by hands-on examples. In particular how to calculate and visualize electronic charge density, electronic density of states, atomic-orbital projected density of states and STM images will be explained.

__References__

[1] P. Giannozzi, *et al.*, J. Phys.: Condens. Matter 21, 395502 (2009). [ link to article ]

**Prof. Nicola Marzari **^{1}** & Dr. Jonathan Yates **^{1}

^{1} Department of Materials, University of Oxford, Parks Road, Oxford OX1 3PH, United Kingdom

The electronic structure of materials is usually described in terms of extended Bloch orbitals. Whilst this is a natural choice for efficient calculations it obscures the simple chemistry picture of bonding in solids.

We will discuss a scheme for transforming calculations in a post-processing step to real-space through the use of so-called Maximally Localised Wannier functions. This retains the accuracy of first principles approaches whilst taking advantage of a real-space view point e.g. a chemically intuitive picture of bonding, fast computation of Fermi surface/spectral properties, or contruction of model Hamiltonians as a basis for calculation of transport properties or LDA+U / LDA+DMFT.

The interface between QE and the open source Wannier90 code [1] will be demonstrated.

__References__

[1] http://www.wannier.org (link to the Wannier90 code, as well as to key theory papers and recent applications)

**Stefano de Gironcoli **^{1,2}

^{1} SISSA, Via Beirut 2/4, 34151 Trieste, Italy

^{2} IOM-DEMOCRITOS, Via Beirut 2/4, 34151 Trieste, Italy

Reaction rate prediction for thermally activated processes represents a challenge for computational physics. Being characterized by multiple time scales, often separated by several order of magnitude, these processes cannot be straightforwardly tackled via molecular dynamics simulations and require some alternative theoretical approach.

In the (harmonic) Transition State Theory the rate can be estimated from the knowledge of the saddle-point energy along the minimum energy path connecting stable states of the potential energy surface.

The nudged elastic band (NEB) method [1,2] is one of the main methods currently used to identify the minimum energy path and the corresponding saddle-point. NEB implementation [3] in the QUANTUM ESPRESSO [4] suite of codes will be described and some practical examples given.

__References__

[1] G. Henkelman, B.P. Uberuaga, and H. Jonsson, J. Chem. Phys., 113, 9901, (2000) [ link to article ]

[2] G. Henkelman, and H. Jonsson, J. Chem. Phys., 113, 9978, (2000) [ link to article ]

[3] C. Sbraccia, "Computer Simulation of Thermally Activated Processes", SISSA Ph.D. Thesis 2005, [ http://www.sissa.it/cm/thesis/2005/sbraccia.pdf ]

[4] P. Giannozzi, *et al.*, J. Phys.: Condens. Matter 21, 395502 (2009). [ link to article ]

**Andrea Dal Corso **^{1,2}

^{1} SISSA, Via Beirut 2/4, 34151 Trieste, Italy

^{2} IOM-DEMOCRITOS, Via Beirut 2/4, 34151 Trieste, Italy

The transport properties of many atomic scale nanostructures, such as nanocontacts and nanowires, are oftentimes well described by the Landauer-Büttiker theory. Within this theory, the ballistic conductance of an open quantum system is proportional to the transmission of the electronic wavefunctions at the Fermi level. After briefly introducing this approach to transport, I will discuss how to calculate the transmission as a function of energy for real materials described, within density functional theory, by pseudopotentials and plane waves. The method, based on the calculation of the scattering states, was introduced by Choi and Ihm [1] for norm conserving pseudopotentials and has been extended to both scalar [2] and fully relativistic [3] ultrasoft pseudopotentials. It is now implemented in the pwcond.x code which is part of the Quantum ESPRESSO package.[4] In the practical section, I will illustrate both the input and the output of this code as well as its main output files. I will present a few examples of transmission calculation for an open quantum system consisting of a scattering region and left and right leads. Key ingredients of the method are the complex band structures along the transport direction which are used to expand the scattering states in the leads, so a part of the examples is devoted to their calculation.

__References__

[1] H. J. Choi and J. Ihm, Phys. Rev. B **59**, 2267 (1999). [ link to article ]

[2] A. Smogunov, A. Dal Corso, and E. Tosatti, Phys. Rev. B **70**, 045417 (2004). [ link to article ]

[3] A. Dal Corso, A. Smogunov, and E. Tosatti, Phys. Rev. B **74**, 045429 (2006). [ link to article ]

[4] P. Giannozzi, *et al.*, J. Phys.: Condens. Matter **21**, 395502 (2009). [ link to article ]

**Prof. Nicola Marzari **^{1}** & Dr. Jonathan Yates **^{1}

^{1} Department of Materials, University of Oxford, Parks Road, Oxford OX1 3PH, United Kingdom

Solid-State Nuclear Magnetic Resonance (SSNMR) spectroscopy provides extremely detailed insights into ordered and disordered materials at the sub-nanometre scale. Electron Paramagnetic Resonance (EPR) provides information on the structure of radicals and defects. In both cases first-principles calculations can play an important role in providing the link between the observed spectrum and the underlying atomic level structure.

The Gauge-Including Projector Augmented Wave Method (GIPAW) [1] has enabled the prediction of magnetic resonance parameters within the planewave pseudoptoential approach. It is now possible to compute the main NMR (chemical shift, electric field gradient, J-coupling) and EPR (hyperfine, g-tensor) parameters for solids within Density Functional Theory. The implementation of the GIPAW approach within QE will be described.

__References__

[1] http://www.gipaw.net/ (links to key theory papers, and recent applications)

**Carlo Cavazzoni **^{1,2}

^{1} CNR-INFM S3, Dip. Fisica, Univ. Modena e Reggio E., Via Campi 213/A, 41100 Modena, Italy

^{2} CINECA, Via Magnanelli 6/3, 40033 casalecchio di Reno, Bologna, Italy

In this lecture the basic elements of the Car-Parrinello ab-initio molecular dynamics and their connection with the CP code input parameters and output will be introduced. It will be shown how to perform the main types of CP simulations: ground state energy, constant energy, constant temperature and constant energy molecular duynamics. Finally some important issues about phase space sampling and how to check the energy drift of the CP trajectory will be discussed.

**Carlo Cavazzoni **^{1,2}

^{1} CNR-INFM S3, Dip. Fisica, Univ. Modena e Reggio E., Via Campi 213/A, 41100 Modena, Italy

^{2} CINECA, Via Magnanelli 6/3, 40033 casalecchio di Reno, Bologna, Italy

The main goal of this lecture is to make attendees aware of the number of possibilities there are to make Quantum Espresso simulations computationally efficient and scalable on parallel machines. The different levels of parallelism implemented in the code will be shown, and it will be explained how and when to activate them to exploit the computing power of different parallel architecture, in particular massively parallel machines like IBM BluGene and Linux clusters. Some insight on the parallel programming techniques used in Quantum Espresso will also be discussed, like the implementation of parallel linear algebra, parallel FFT and hybrid MPI+OpenMP programming.

We would like to pay special thanks to Inspire, NVIDIA and CAPS entreprise for their sponsorship of this event.