Boron nitride (BN) nanotubes have the same nanostructure as carbon nanotubes but are found to exhibit significant resistance to oxidation at high temperatures. Our systematic study has revealed that BN nanotubes are stable at 700 °C in air and that some thin nanotubes (diameter less than 20 nm) with perfect multiwalled cylindrical structure can survive up to 900 °C. Thermogravimetric analysis reveals an onset temperature for oxidation of BN nanotubes of 800 °C compared with only 400 °C for carbon nanotubes under the same conditions. This more pronounced resistance of BN nanotubes to oxidation is inherited from the hexagonal BN and also depends on the nanocrystalline structure. This high level of resistance to oxidation allows promising BN nanotube applications at high temperatures.

No abstract

Hyperfine studies of disordered materials often yield the distribution of the electric field gradient (EFG) or related quadrupole splitting (QS). The question of the structural information that may be extracted from such distributions has been considered for more than fifteen years. Experimentally most studies have been performed using Mössbauer spectroscopy, especially on . However, NMR, NQR, EPR and PAC methods have also received some attention. The EFG distribution for a random distribution of electric charges was for instance first investigated by Czjzek et al [1] and a general functional form was derived for the joint (bivariate) distribution of the principal EFG tensor component and the asymmetry parameter . The importance of the Gauss distribution for such rotationally invariant structural models was thus evidenced. Extensions of that model which are based on degenerate multivariate Gauss distributions for the elements of the EFG tensor were proposed by Czjzek. The latter extensions have been used since that time, more particularly in Mössbauer spectroscopy, under the name `shell models'. The mathematical foundations of all the previous models are presented and critically discussed as they are evidenced by simple calculations in the case of the EFG tensor. The present article only focuses on those aspects of the EFG distribution in disordered solids which can be discussed without explicitly looking at particular physical mechanisms. We present studies of three different model systems. A reference model directly related to the first model of Czjzek, called the Gaussian isotropic model (GIM), is shown to be the limiting case for many different models with a large number of independent contributions to the EFG tensor and not restricted to a point-charge model. The extended validity of the marginal distribution of in the GIM model is discussed. It is also shown that the second model based on degenerate multivariate normal distributions for the EFG components yields questionable results and has been exaggeratedly used in experimental studies. The latter models are further discussed in the light of new results. The problems raised by these extensions are due to the fact that the consequences of the statistical invariance by rotation of the EFG tensor have not been sufficiently taken into account. Further difficulties arise because the structural degrees of freedom of the disordered solid under consideration have been confused with the degrees of freedom of QS distributions. The relations which are derived and discussed are further illustrated by the case of the EFG tensor distribution created at the centre of a sphere by m charges randomly distributed on its surface. The third model, a simple extension of the GIM, considers the case of an EFG tensor which is the sum of a fixed part and of a random part with variable weights. The bivariate distribution is calculated exactly in the most symmetric case and the effect of the random part is investigated as a function of its weight. The various models are more part...

This paper, by no means exhaustive, focuses on high-energy ball-milling of oxides, on their mechanically induced changes and on the consequences of such changes on their physical and chemical properties. High-energy ball-milling offers a fortunate combination of technical simplicity and of complexity both of physical mechanisms which act during milling and of mechanosynthesized materials. Its basic interest, which stems from the large diversity of routes it offers to prepare oxides either directly or indirectly, is illustrated with various families of oxides. The direct path is to be favoured when as-milled oxides are of interest per se because of their nanocrystalline characteristics, their defects or their modified structures which result from mechanically driven phase transformations. The indirect path consists of a sequence of steps starting with mechanically activated oxides which may be subsequently just annealed or submitted to a combination of thermal treatments, with the possible occurrence of various chemical reactions, to prepare the sought-after materials with potential gains in processing temperatures and times. High energy ball-milling of oxides is more and more currently used to activate powders and to prepare nano-oxides at moderate temperatures. The interest of an activation step is well illustrated by the broad development of doped titania powders, synthesized by heat treatment of pre-ground reactants, for photocatalytic applications or to develop antibacterial materials. Another important class of applications of high-energy ball-milling is the formation of composites. It is exemplified here with the case of oxide-dispersed strengthened alloys whose properties are considerably improved by a dispersion of ultra-stable nanosized oxides whose formation mechanisms were recently described. The basic understanding of the mechanisms by which oxides or oxide mixtures evolve by high-energy ball-milling appears to be less advanced than it is for metallic materials essentially because of the overall complexity of the oxide structures, of their surfaces, of their defects and of their mechanical behavior.