From cassette tapes to disk drives, the most popular way to store data is with magnetic materials. Yet exactly how individual magnetic bits reverse direction during writing and re-writing is still poorly understood. In the 22 January PRL a Canadian group reports the fastest and most spatially detailed movies of the flipping of micrometer-sized magnets. For the first time, the experiments reveal how the slow and complex flipping patterns of magnetic domains within a micromagnet can be coordinated and sped up with magnetic fields. The results help researchers better understand magnetic switching and may help designers optimize magnetic storage devices.
Mark Freeman, of the University of Alberta in Edmonton, says that magnetization reversal is tough to understand. “It is strongly dependent on the material and sample shape, defects, temperature, and so on.” But past experiments haven’t had the combined temporal and spatial resolution to follow the switching in detail.
To catch the quick change act, the Alberta group used a kind of high tech strobe light called a scanning Kerr microscope, in which femtosecond laser pulses bounce off the surface of a magnetic sample. Each light pulse has its polarization altered slightly by the surface magnetism. This polarization shift is measured as the pulsing beam scans across the sample, building up an image of the state of the magnet. With this setup they’ve combined high spatial resolution with picosecond timing to get the best magnetic images ever recorded.
The samples were 10 by 2 µm rectangles of 15 nm thick nickel-iron () grown on a small strip of gold. The researchers applied a magnetic field parallel to the long direction of a sample (“longitudinal bias field”) to align its atomic spins and then sent a 10-ns-long jolt of current through the gold strip. The current pulse briefly created a magnetic field in the direction opposite to the static field, causing the nickel-iron’s magnetization to flip twice in rapid succession. Using the fast scanning Kerr microscope, the team imaged the time course and domain structure of the sample’s magnetic state.
Freeman and his co-workers found evidence for two entirely different modes of reversal. With a longitudinal bias field, there was a lag of about 3.5 ns as the magnetization responded to the switching pulse. But with an additional field applied along the short dimension of the sample (“transverse field”), the magnetization responded much faster to the switching pulse, settling down within a nanosecond.
The Kerr imaging explains this difference. Without the transverse field, the magnet reverses in the usual way–by nucleating small regions which then grow together, until the whole magnet has flipped. But a transverse field causes the region to flip almost uniformly. Coordination within the sample seems to be the key to increased speed, just as a team of synchronized swimmers turns most efficiently if they all go the same way at once. In the magnetic structure, a transverse field tells the individual magnetic spins which way to flip. “It’s a remarkable result,” says David Awschalom of the University of California at Santa Barbara, “and counter to the way many people thought small magnetic particles would switch.”
Such improvements in magnetic switching speed may provide a way for disk drive makers to optimize read/write heads for faster data flow and higher storage density. Freeman says that the technique should also permit more stringent tests of the models used in designing virtually all micro- and nanomagnetic devices. “These pictures tell a thousand words,” says Awschalom.