The effect of the nanostructure on the photochemistry of TiO2 is an active field of research owing to its applications in photocatalysis and photovoltaics. Despite this interest, little is known of the structure of small particles of this oxide with sizes at the nanometer length scale. Here we present a computational study that locates the global minima in the potential energy surface of Ti(n)O2n clusters with n = 1-15. The search procedure does not refer to any of the known TiO2 polymorphs, and is based on a novel combination of simulated annealing and Monte Carlo basin hopping simulations, together with genetic algorithm techniques, with the energy calculated by means of an interatomic potential. The application of several different methods increases our confidence of having located the global minimum. The stable structures are then refined by means of density functional theory calculations. The results from the two techniques are similar, although the methods based on interatomic potentials are unable to describe some subtle effects. The agreement is especially good for the larger particles, with n = 9-15. For these sizes the structures are compact, with a preference for a central octahedron and a surrounding layer of 4- and 5-fold coordinated Ti atoms, although there seems to be some energy penalty for particles containing the 5-fold coordinated metal atoms with square base pyramid geometry and dangling Ti=O bonds. The novel structures reported provide the basis for further computational studies of the effect of nanostructure on adsorption, photochemistry, and nucleation of this material.

The hydration enthalpy and Gibbs free energy of proton and hydroxide are calculated by means of a combination of ab initio density functional theory and a polarizable continuum model within the self-consistent reaction field method. The ion-water cluster models here used include up to 13 water molecules solvating the ions. This allows the first and second solvation shells to be described explicitly from first principles. Vibrational contributions to the enthalpy and entropy have been taken into account. Our best model of the hydrated proton includes three molecules in the first hydration shell and nine molecules in the second shell. The calculated proton hydration enthalpy is ϷϪ1150 kJ/mol, which is in rather good agreement with the most recent results from cluster-ion solvation data. The hydration free energy of the proton has a larger error of Ϸ50-80 kJ/mol as compared to recently reported values. The calculated hydroxide hydration enthalpy, ϷϪ520 kJ/mol, and hydration free energy, ϷϪ400 kJ/mol, are consistent with well-established values taken from experiment. Two different sources of error in our calculations, namely, the nature of the hydrated complex and the outlying charge correction, are discussed. Moreover, we compare the results from three slightly different methods for the calculation of hydration energies.

In the southern Iberian Peninsula, Rhododendron ponticum occurs in restricted and vulnerable populations as a Tertiary relict. Population structure and the main phases of the reproductive process were examined in order to shed light on recruitment patterns and limitations. Rhododendron ponticum flowers are self‐compatible and attract a diverse array of insects, which are responsible for a considerable number of seeds set in the populations. Nevertheless, only adults form populations, whilst seedlings are scarce and saplings virtually absent (only two juveniles out of 2489 adults sampled). Non‐specialized vegetative multiplication by layering was observed. Recruitment failure seems to depend on the scarcity of safe microsites, which are free from drought, for seedling establishment. The observations contrast with R. ponticum's reputation as an aggressive invader in temperate Atlantic areas. It is proposed that the species shows a variable balance between sexual reproduction and vegetative multiplication depending on environmental conditions. At present, only the latter seems to be prevailing in relict populations in the Iberian Peninsula. This flexible reproductive strategy is also discussed as a mechanism allowing persistence during geological climatic oscillations. © The Linnean Society of London, Botanical Journal of the Linnean Society, 2002, 140, 297–311.

The OH radical is an important species in natural and man made aqueous environments, influencing diverse processes such as the oxidation of atmospheric pollutants or the development of some diseases. Yet, little is known about the solvation thermodynamics and structure of the hydration shell of OH. Here, we present a computational study of the hydration of OH in small H2 n +1O n +1 (n = 1−5) clusters. We begin by comparing three different quantum chemical methods, UMP2, BLYP, and BHLYP. We find that BLYP does not describe correctly the OH−H2O interaction as compared to the current MP2 or other high ab initio calculations found in the literature. BLYP favors the formation of hemibonded H2O−OH structures, whereas MP2 predicts that hydrogen-bonded complexes are more stable. Mixing Becke's exchange functional with 50% Hartree−Fock exchange improves the DFT description, yielding results that are similar to those from MP2. We find that the H2 n +1O n +1 clusters form structures in which all species are donors and acceptors in hydrogen bonded rings similar to those of pure water clusters. OH participates in two or three hydrogen bonds. Structures in which OH forms more than three hydrogen bonds are not favored energetically. We report values of energy, enthalpy, and Gibbs free energy of complexation in the gas phase, OH(g) + H2 n O n (g) → H2 n +1O n +1(g), as a function of cluster size. We also estimate values of thermodynamic parameters of hydration in the liquid phase from OH(g) + H2 n O n (aq) → H2 n +1O n +1(aq), where the energies of the aqueous species, H2 n O n (aq) and H2 n +1O n +1(aq), are calculated by means of a hybrid solvation model in which part of the solvent is treated explicitly and the long-range interactions are added into the Hamiltonian by means of the PCM version of the self-consistent reaction field. The implications of our work as well as the accuracy of the results are also discussed.

No abstract