To test their theory, the scientists embedded 4 nm diameter cobalt nanoparticles in carbon, alumina and cobalt-oxide matrices. The first two materials are paramagnetic while cobalt-oxide is antiferromagnetic. The team deposited a 15-20 nm thick layer of matrix material followed by a layer of nanoparticles, repeating the process up to ten times. They made the cobalt nanoparticles by gas condensation of sputtered atoms in a cluster gun, incorporating some oxygen inside the sputtering chamber so that the surface of the particles was oxidized. This produced a CocoreCoOshell core-shell structure: on average the cobalt oxide shell was about 1 nm thick.
"Perhaps the main difficulty was to ensure a good contact between the nanoparticle and the matrix, to maximize the effects," Josep Nogués of the Universitat Autònoma de Barcelona told nanotechweb.org. "In our case this was solved by growing core-shell nanoparticles rather than purely metallic nanoparticles."
Cobalt nanoparticles in an alumina or carbon matrix lost their magnetic moment at a temperature of roughly 10 K. In contrast, when embedded in an antiferromagnetic cobalt-oxide matrix, the nanoparticles remained ferromagnetic up to about 290 K.
"The system we studied - Co/CoO - is not suitable for applications due to its low transition temperature," added Nogués. "One of our first research goals is to find new systems that will work at room temperature and above."
The researchers, who reported their work in Nature, believe that choosing the right combination of nanoparticles and matrix could result in magnetically stable dots only a few nanometres in size, an achievement that would surpass the magnetic-storage density goal of 1 Tbit/sq. inch.