Invierno 2000

Polycarbonates are important engineering thermoplastic materials with applications is fields as diverse as compact disks, mobile telephones and artificial kidneys.
In my talk I present a computational study, based on the density functional theory, of a hierarchical sequence of molecular analogs of polycarbonates, going from fragments of the monomeric unit to small polymeric rings. The equilibrium geometry is determined by DFT-molecular dynamics, and the vibrational properties are analyzed by diagonalization of the dynamical matrix, computed by a finite difference differentiation of the potential energy surface.
The resulting structural and vibrational properties agree well with available experimental data, and provide an extended data base to gauge empirical force field models. Moreover, I discuss the reactivity of the largest molecular analogs in the presence of smaller species (phenol, Li- and Na-phenoxides), that are used as reagents in crucial steps of the polymerization reaction.
Our computations show that the carbonate group provides the docking site for the positive end of polar species (i.e., Li+ and Na+ in Li- and Na-phenoxides), giving rise to moderately stable complexes (DE 10 Kcal/mol). In turn, the native complexes are connected by low energy paths to a variety of different chemical species, that are known to be present in the polymerization reaction.
The description of reactivity in polycarbonates offered by DFT simulations provides important parameters (like approximate reaction barriers, characteristic vibrational frequencies for the different chemical species, etc.) useful in the interpretation of experimental results, and represents a challenge for more empirical modellization of these systems.

Spontaneus polarization has long been known to take place in ferroelectrics, but its existence in compound semiconductors has been long disregarded becuase technologically important semiconductors cristallize in the zincblende structure, whose high symmetry does not allow the existence of spontaneous polarization fields.
III-V nitrides naturally crystallize in the low symmetry wurtzite structure. Recent calculations have shown that this materials own extremely high spontaneous polarization fields. As a consequence of these anomalous polarization properties huge (about 1 to 3 MV/cm) electric fields will show up in nanostructures like superlattices and MQW's.
We show, for a large variety of nanostructure geometries and compositions, how the values of the resulting electrostatic fields can be predicted using simple rules, from the knowledge of the total polarization and dielectric properties.

The Supercomputing Center of Catalonia (CESCA) has been an IBM user since its creation back in 1991, initially with an IBM 3090 and after 1995 with an IBM SP2. The machine has grown from 12 to 44 processors of different classes (wide, thin, thin2 ...) and it has made it to 7 out of 8 TOP500 lists during the period 1995-98. We will comment on our experience on the successive processor generations and their usage and performance.

The combined quantum mechanics/molecular mechanics (QM/MM) method is emerging as a viable computational molecular modeling tool. This method allows for the incorporation of effects that are often ignored in high level calculations, but may be critical to the real chemistry of the simulated system. In the combined QM/MM method, part of the system is treated quantum mechanically, while the remainder of the system is treated with a faster molecular mechanics force field.
This way, large molecules, for which accurate calculations are too expensive or out of reach, can be investigated in a reasonable amount of time.

Metals subjected to irradiation by energetic atomic particles, such as those generated in nuclear power systems, undergo changes in their properties due to 'radiation damage'. This damage arises from collisions of the irradiating particles with the atoms in the host lattice. Each collision can create a cascade of secondary collisions, with the result that vacant lattice sites and interstitial atoms are formed. Atomic-scale computer simulation by molecular dynamics has now shed considerable light on the mechanisms associated with defect production by collision cascades in metals and on the number and disposition of the defects themselves. For example, simulations have shown that the thermal spike phase results in the efficiency for production of interstitial and vacancy defects being much less than estimates given by earlier binary-collision models and that self-interstitial atoms can form clusters during the cascade process itself. In addition, it has been demonstrated that such clusters can have high mobility. The present talk will describe recent progress in this area. It will include research on: defect production in cascades as a function of energy; the effects of varying the irradiation temperature; the influence of crystal structure on defect number and clustering fraction; and the mobility of the interstitial clusters. The consequences for evolution of radiation damage will be discussed.

The knowledge of the structure of the Ziegler-Natta's catalyst in heterogeneous phase is of primer importance to modelise the reaction of polymerisation of the a-olefins. Starting from the study of small clusters, we investigated the acido-basic properties of titanium chlorides in its three oxidation states, and their reactivity toward ethylene. The whole calculations performed on CESCA with the VASP package, concern the structure and the reactivity of the main active surfaces of magnesium chloride that is commonly used as a support for the titanium chloride's active sites in the Ziegle-Natta catalyst. We present here the mains results of the former clusters study and the preliminary results of the investigation on the periodic systems.

CCR5 is a G-protein coupled receptor (GPCR) that has the primary function to activate chemotaxism in response to its natural ligand, the CC-chemokines RANTES, MIP-1a, MIP-1b and MCP-2. It is also the main co-receptor allowing infection by the HIV virus on M-tropic cells. Although no cristallographic data of its structure is currently available, we are trying to understand how CCR5 interacts with its ligand using computational modeling techniques. Combining the low resolution structural data of the rhodopsin (a related receptor), data from molecular biology experiments and evolutionary informations, we use molecular dynamics simulations as the guiding tool in the modeling process. Indeed, we have deeply studied some structural particularities of the chemokine receptors family that were potentially important for the receptor activation mechanisms. Our computer simulations show that the studied structural motif should be of great importance for the function of the receptor. First results from ongoing experiments are confirming this theoretical study. This leads to very interesting insights into the molecular activation mechanism of the chemokine receptor sub-family of GPCR's.