|
ScienceWeek
CONDENSED MATTER: ON MAGNETIC DOMAIN WALLS
The following points are made by E. Saitoh et al (Nature 2004 432:203):
1) A magnetic domain wall (DW) is a spatially localized change of magnetization configuration in a magnet. This topological object has been predicted to behave at low energy as a composite particle with finite mass(1). This particle will couple directly with electric currents as well as magnetic fields, and its manipulation using electric currents(2-5) is of particular interest with regard to the development of high-density magnetic memories. The DW mass sets the ultimate operation speed of these devices, but has yet to be determined experimentally.
2) The DW mass is, in principle, detectable directly by resonance absorption measurements; magnetic-resonance effects have long been investigated for DWs trapped in pinning centres. However, owing to their low sensitivity, these resonance measurements are limited only to samples containing huge numbers of DWs with inevitable distribution. As a result, single DW properties have never been accessible.
3) Recent studies have revealed that a DW can be driven by an electric current by way of several different mechanisms(2-5). Theories predict two dominant mechanisms in submicrometer wires, both arising from the exchange coupling between the local magnetization and the spin of conduction electrons. The first is due to the reflection of conduction electrons by the DW, called momentum transfer; this effect is proportional to the charge current and the DW resistance. The second is the transfer of the spin angular momentum from conduction electrons to the DW as the electrons pass through the DW. This effect, called the spin-transfer effect, is proportional to the spin polarization of the current (spin current). In the case of narrow metallic wires under a steady current, spin transfer is expected to be dominant and recent experiments using DC currents(3-5) are consistent with the above mechanism.
4) The authors report the direct observation of the dynamics of a single DW in a ferromagnetic nanowire, which demonstrates that such a topological particle has a very small but finite mass of 6.6 10^(-23) kg. This measurement was realized by preparing a tunable DW potential in the nanowire, and detecting the resonance motion of the DW induced by an oscillating current. The resonance also allows low-current operation, which is crucial in device applications; a DW displacement of 10 microns was induced by a current density of 10^(10) A m^(-2).
References (abridged):
1. D÷ring, V. W. ber die Troegheit der Woende zwischen Weisschen Bezirken. Z. Naturforsch. 3a, 373-379 (1948)
2. Berger, L. Exchange interaction between ferromagnetic domain wall and electric current in very thin metallic films. J. Appl. Phys. 55, 1954-1956 (1984)
3. Freitas, P. P. & Berger, L. Observation of s-d exchange force between domain walls and electric current in very thin Permalloy films. J. Appl. Phys. 57, 1266-1269 (1988)
4. Hung, C.-Y. & Berger, L. Exchange forces between domain wall and electric current in permalloy films of variable thickness. J. Appl. Phys. 63, 4276-4278 (1988)
5. Gan, L. et al. Pulsed-current-induced domain wall propagation in Permalloy patterns observed using magnetic force microscope. IEEE Trans. Magn. 36, 3047-3049 (2000)
Nature http://www.nature.com/nature
--------------------------------
Related Material:
MATERIALS SCIENCE: ON NONVOLATILE RANDOM ACCESS MEMORY
The following points are made by J. Campbell Scott (Science 2004 304:62):
1) Various inadequacies of current data storage technologies are motivating a widespread search (1-5) for new fast, nonvolatile, and inexpensive ways of storing information. The short data retention time of the most common form of memory, DRAM (dynamic random access memory), places severe constraints on the design of computers. Nonvolatile RAM would permit instant restart, encourage more frequent standby operation, save energy, and extend battery life.
2) Although a hard disk drive (HDD) can store copious amounts of data essentially indefinitely, its data access time of several milliseconds is more than 1 million times the duration of the CPU clock cycle of today's computers, and the mechanical reliability of its moving parts is an increasing concern. Flash memory is currently the most successful nonvolatile solid-state data storage technology, but its fastest writing speed is still slower than the clock cycle rate by a factor of 1000, and it is more expensive, in terms of dollars per gigabyte, than DRAM and, especially, HDD.
3) Although the demand and opportunity for new types of memory clearly exist, any successful newcomer must ultimately exceed the existing speed and cost constraints of today's entrenched technologies. DRAM is fast but relatively expensive. A price premium is paid for the nonvolatility of flash memory but at a substantial reduction in speed. HDD is much less expensive but is too slow to compete with solid-state memories.
4) Other performance criteria must also be satisfied. Nonvolatile data should be retained for at least 10 years, and archiving may require as long as 100 years. Power consumption must be low --ideally zero, unless data are being accessed. Flash memory is the clear winner here, consuming just a few milliwatts to operate an average-size memory card (corresponding to an energy of about 1 nanojoule per bit to write and about 1 picojoule per bit to read), whereas typical DRAM and HDD continuously use several watts for refresh and spindle rotation, respectively. The number of rewrite cycles is also of great importance. Today's flash memories are typically rated at 10^(6) cycles, whereas HDD achieves 10^(12) cycles and DRAM greater than 10^(15) cycles.
References (abridged):
1. D. Ma, M. Aguiar, J. A. Freire, I. A. Hummelgen, Adv. Mater. 12, 1063 (2000)
2. M. A. Reed et al., Appl. Phys. Lett. 78, 3735 (2001)
3. C. Rossel et al., J. Appl. Phys. 90, 2892 (2001)
4. D. M. Taylor et al., J. Appl. Phys. 90, 306 (2001)
5. J. H. Krieger et al., Synth. Metals 122, 199 (2001)
Science http://www.sciencemag.org
ScienceWeek http://scienceweek.com
|