[en] We constructed regular fractal SiOH atomic clusters which pending bonds are saturated with OH molecules. We calculated the binding energies of these clusters as well as for sp² hybridization as for sp³ hybridizations. The result are the following: for the two hybridizations, the total binding energies have a linear dependence on the size of the fractal cluster, which comes directly from the scaling law of the fractal characteristic of the building of the cluster. We related by a scaling law, the number of electronic bonds and the total bonding energy

[en] We built graphene nanoflakes doped or not with C atoms in the sp³ hybridization or with Si atoms. These nanoflakes are isolated, i.e. are not connected to any object (substrate or junction). We used a modified tight binding method to compute the π and σ density of states. The nanoflakes are semiconducting (due to the armchair geometry of their boundaries) when they are pure but they become conducting when doped because doping removes the degeneracy of the density of states levels. Moreover, we showed that the π Fermi level and the Fermi level of both π and σ electrons are not superimposed for small isolated nanoflakes.

[en] We computed by a Monte Carlo method, derived from the solid on solid model, the evolution during thermal treatment of a polycrystalline thin film deposited on a substrate with no further deposition. Two types of substrates have been studied: a single crystalline substrate with no defects and a single crystalline substrate with defects. We obtain islands which are either flat (i.e., with a height which does not overcome a given value) or grow in height like narrow towers. The numerical results have been qualitatively compared with experimental data: the fragmentation after thermal treatment of an yttria stabilized zirconia thin film deposited on an Al₂O₃ substrate. A good agreement was found regarding the morphology of simulated and experimental nanoislands

[en] We performed a numerical study of multiple scattering on regular fractal aggregates. We used a recursive method to compute the backscattered intensity resulting from a multiple scattered coherent wave. The wavelength of the incident field was supposed to be in the optical domain. The resulting backscattered intensity is analyzed in the light of diffraction on deterministic fractals. Due to multiple scattering, the scaling of the signals is lost, on the contrary of single scattering signals

[en] We performed two-dimensional molecular dynamics simulations of cohesive discs under shear. The cohesion between the discs is added by the action of springs between very next neighbouring discs, modelling capillary forces. The geometry of the cell allows disc-disc shearing and not disc-cell wall shearing as is commonly found in the literature. Does a stick-slip phenomenon happen though the upper cover moves at a constant velocity, i.e. with an infinite shearing force? We measured the forces with which the discs acted on the upper cover for different shearing rates, as well as the disc velocities as a function of the distance to the bottom of the cell. It appears that the forces measured versus time present a periodic behaviour, very close to a stick-slip phenomenon, for shearing rates larger than a given threshold. The discs' collective displacements in the shearing cell (back and ahead) are the counterpart of the constant velocity of the upper cover, leading to a periodic behaviour of the shear stress

[en] We computed a Lennard-Jones frozen liquid with a free surface using classical molecular dynamics. The structure factor curves on the free surface of this sample were calculated for different depths knowing that we have periodic boundary conditions on the other parts of the sample. The resulting structure factor curves show an horizontal shift of their first peak depending on how deep in the sample the curves are computed. We analyze our resulting curves in the light of spatial correlation functions during melting. The conclusion is that the differences between bulk and surface are quite small during melting and that at the end of melting, only the very surface happens to be less dense than the bulk. This result is intrinsic to the shape of the Lennard-Jones potential and does not depend on any other parameter.

[en] We computed by a Monte Carlo method the thermal relaxation of a polycrystalline thin film deposited on a square lattice. The thin film was modelled by a two-dimensional array of elementary domains, which have each a given height and a given crystallographic orientation. During the Monte Carlo process, the height of each of these elementary domain is allowed to change as well as their crystallographic orientation. After equilibrium is reached at a given numerical temperature, all elementary domains have changed their orientation into the same one and small islands appear. This method is a numerical approach analogous to the solid on solid model, which includes the evolution of a thin film in the vertical direction as well as parallel to the substrate. Moreover, the effects of defects, if added to the substrate, on the thin film evolution, can be easily studied

[en] Sol-gel obtained yttria stabilized zirconia (YZS) nanocrystals have different morphologies when they grow on an α-alumina substrate after thermal treatment. When the substrate has planar defects, the nanocrystals grow in height and are narrow with a crystallographic orientation [111] in the vertical direction, while when the substrate is a perfect plane at the nanometric scale, the nanocrystals are rather extended over the substrate and do not grow in height, with a crystallographic orientation [100] in the vertical direction. We present here a Monte Carlo approach which computes the actions of the substrate on the nanocrystals during thermal treatment: one action is the change in crystallographic orientation depending on the presence of defects and the other is the action on the morphology of the nanocrystals. The equivalent of thermal treatment is obtained after applying the Metropolis algorithm with adequate expressions of the energy depending on the inter-plane spacings and surface diffusion. Our numerical approach is in good agreement with the experimental results on the orientations and morphologies of the YSZ nanocrystals growing on a α-alumina substrate with planar defects [R. Bachelet, A. Boulle, B. Soulestin, F. Rossignol, A. Dauger, R. Guinebretiere, Thin Solid Films 515 (2007) 7080].

[en] We computed by a Monte Carlo method derived from the solid on solid model, the thermal relaxation of a polycrystalline thin film deposited on a Penrose lattice. The thin film was modeled by a 2-dimensional array of elementary domains, which have each a given height. During the Monte Carlo process, the height of each of these elementary domains is allowed to change as well as their crystallographic orientation. After equilibrium is reached at a given numerical temperature, all elementary domains have changed their orientation into the same one and small islands appear, preferentially on the domains of the Penrose lattice located in the center of heptagons. This method is a numerical approach to study the influence of the substrate and its defects on the islanding process of polycrystalline films