No abstract

Magnetite nanoparticles (Fe3O4) of three different sizes below the limit for single domain magnetic behaviour have been obtained by thermal decomposition of an iron precursor in an organic medium in the presence of a surfactant. Good agreement between mean particle size obtained by TEM, crystal size calculated from x-ray diffraction and magnetic diameter calculated from magnetization curves measured at room temperature shows that the samples consist of uniform, crystalline and isolated magnetite nanoparticles with sizes between 5 and 11 nm. High saturation magnetization and high initial susceptibility values have been found, the latter decreasing as the particle size decreases. The main contribution to the anisotropy is magnetocrystalline and shape anisotropy, since surface anisotropy is suppressed by the oleic acid molecules which are covalently bonded to the nanoparticle surface.

Isolated nanometric particles (D < 30 nm) of γ-Fe2O3 in a silica matrix have been prepared by heating at 400 °C the gel formed in the hydrolysis of an ethanol solution of Fe(NO3)3‚9H2O and tetraethylorthosilicate (TEOS). However, when FeCl3‚6H2O was used as precursor, well-developed hematite particles were obtained in the final composite. This different behavior was already manifest in the initial gels. Thus, the gel obtained from iron nitrate salt shows a compact appearance as a result of its higher degree of network connectivity (polymeric gel) whereas the one from the iron chloride appears more loose and highly hygroscopic (colloidal gel). In addition, small superparamagnetic nuclei are formed during the hydrolysis and condensation of the gel obtained from the iron nitrate salt. The γ-Fe 2O3 nanoparticle formation takes place through a reduction-oxidation reaction which occurs during the burning of the organic species trapped inside the gel pore. The growth mechanism of the γ-Fe2O3 nanoparticles in the silica network has been studied as well as the optimum conditions for their preparation. Thus, γ-Fe2O3 nanocomposites with different particle sizes and distributions can be prepared by adequate modification of the initial gel microstructure through different gelation times, salt concentrations, and mechanical treatment. Superparamagnetic behavior has been found in all nanocomposites at room temperature, meanwhile at 70 K, a transition from superparamagnetic to ferrimagnetic behavior is observed as the particle size increases. In all cases, the variation in particle size observed by X-ray diffraction corresponds well with changes in the saturation magnetization for the γ-Fe 2O3 nanocomposites. Similar size effects are also found via the coercivity values at 70 and 5 K.

No abstract

The effects of interactions (dipolar and exchange) on the magnetic behavior of granular solid systems are examined using a Monte Carlo model capable of predicting the temperature and time dependence of the magnetic properties. Using this model the interaction effects on the magnetization and the magnetoresistance are studied. The results show that these properties depend critically on the strength and nature of the interactions. Magnetostatic interactions are found to decrease both remanence and coercivity and Hc is predicted to decrease linearly with concentration. It is shown that spatial disorder may be responsible for an increase of coercivity with exchange coupling which has been observed in some experimental studies. In systems with no hysteresis, magnetostatic interaction effects are found to increase the superparamagnetic transition temperature, in agreement with experimental data and previous analytical treatments. Calculations of the giant magnetoresistance (GMR) show that magnetostatic interaction effects give rise to a finite positive resistivity at zero field which increases with concentration. This causes the value of the maximum change in resistivity, which occurs near the coercivity, to be greater than the value at zero field. These calculations are in agreement with experimental observations of GMR in granular solids. It is predicted that the GMR is strongly dependent on the spin diffusion length via the local spin–spin correlation function.