In the last issue of Materials Today we reported on the ultra-fast melting of charge order, where the charge disproportionation in TiSe2 disappeared after just 20 femtoseconds [Gone in 20 femtoseconds, May, 2011]. Now, a collaboration between researchers from The Netherlands, Germany, the USA, the UK, and Japan has found evidence for an ultra-fast magnetic reversal with a surprising twist [Radu et al., Nature (2011) 472, 205].
The team studied a ferrimagnetic alloy containing gadolinium, iron, and small amounts of cobalt. The iron and cobalt form one ferromagnetic sublattice, while the gadolinium forms another. These two sublattices are antiferromagnetic with respect to each other, but as the magnitudes of the magnetic moments between the two are not equal, the result is a material with a net moment.
The two sublattices possess different temperature dependencies, and so upon warming the material reaches a point at which the magnetic moments from the two sublattices cancel. Several years ago, several of the researchers involved in this new study discovered that by quickly heating the alloy above this critical temperature, the orientation of the FeCo sublattice could be flipped in under 700 fs.
In this new study the researchers again used a short burst of laser light to heat the sample above the critical temperature, but also used 100 fs pulses of x-rays to rapidly probe the individual sublattices. X-ray magnetic circular dichroism (XMCD) involves using circularly polarized beams of x-rays, tuned close to the absorption edges of the elements present in the material. By comparing the amount of right and left circularly polarized light absorbed, at particular energies, it is possible to study the magnetic state of individual elements.
Following the initial laser pulse, the magnetic structures in both lattices began to collapse. After just 300 fs the moments of the Fe lattice had been reduced to zero, but the Gd lattice took much longer, around 1500 fs, to collapse. The magnetic structures quickly reorganized to form new sublattices, although each sublattice now pointed in the opposite direction to how it had before the laser pulse. The startling result reported in this new study is thus that between 300 and 1500 fs the system is actually in a ferromagnetic state, and therefore possesses a large magnetic moment.
The team has not only managed to demonstrate that it is possible for two lattices to have “totally different dynamics” [Radu et al.], they have also shown that it is possible to control the magnetization of a material in the femtosecond regime. As such, we could one day look forward to an increase in data storage speeds of several orders of magnitude.
Stewart Bland