Ensembles consisting of Co nanoparticles fixed in space by freezing ferrofluids are investigated. In frozen ferrofluids a well defined low temperature magnetic state of the nanoparticle system can be achieved. We distinguish between textured samples (preferred orientation of easy axes) and non-textured with randomly oriented easy axes. Considering the magnetization curves of such systems bears in principle the possibility to distinguish between influences arising from particle anisotropy and particle-particle interaction. Our intention is to present new experimental strategies aiming to separate these influences from magnetization measurements versus temperature. In another approach we analysed the susceptibility derived from magnetization measurements and determined the mean magnetic moment of the particles and an additional field which may be caused by the anisotropy of the particles or the interaction between them. Transmission electron microscopy was used for the direct measurement of the particles' sizes and their size distribution.

No abstract

No abstract

PACS 75.50.Mm, 75.75.+a The best way to achieve a virgin state of a magnetized sample is to heat it up to a temperature high enough to destroy every magnetic order and thereafter cooling it down in a zero magnetic field. This procedure is often not possible because high temperatures may cause irreversible changes in the sample's constitution e. g. by crystallization in amorphous systems. Because of magnetic history and memory effects it is difficult to measure the same initial magnetization curves after different demagnetizing procedures for materials consisting of nanoparticles with high Curie temperatur TC . Very often the demagnetization by alternating decreasing magnetic fields (AC demagnetization) is used instead. Ferrofluids offer the possibility to demagnetize the sample by heating it above the melting point of the liquid in zero external field. The brownian motion destroys any "frozen in" magnetic order and relieves single domain particles in the fluid. After zero field cooling the virgin state may be reproducibly achieved. We present an experimental study showing the influence of different demagnetization sequences on the initial magnetization curves in frozen ferrofluids. In this contribution we show the deviation between thermally demagnetized and AC demagnetized samples by substracting both initial curves from each other. We detect a "frozen in" pattern which shows the "steps" of the AC demagnetization sequences. In a further step we focus on the experimental determination of the switching field distribution (SFD) of the particles. We compare the SFD with a plot proposed by Thamm and Hesse (1996). ExperimentsMagnetization experiments were performed in a Quantum Design SQUID Magnetometer (MPMSR2) localized at the Physikalisch-Technische Bundesanstalt (PTB) Braunschweig, Germany. We present investigations on one ferrofluid consisting of Co-particles with o-dichlorobenzene (ODCB) as solvent. The ferrofluid was prepared using high-temperature solution-phase synthesis ([1] and references herein). The particle radii derived from room temperature magnetization measurements can be fitted very well by a lognormal distribution. The expectation value of the fitted log-normal distribution is approximately 1.6 nm and the standard deviation 0.5 nm. These values are in very good agreement with results from electron microscopy. For high resolution electron microscopy (HRTEM) the samples were prepared by dipping the electron transparent TEM carbon grids into the highly diluted ferrofluid. During this procedure the particles tend to self-organize and form a hexagonal superlattice of up to 1µm×1µm in size. The ODCB has a melting point of around 220 K. The further discussed measurements are all performed at 5 K which implies that the solution is frozen and the particles are fixed in space.

No abstract