Total Pageviews

Sunday, March 25, 2012

Recording the data at the “ultrasfast” rate for your digital device.

The principle of magnetic storage used by most hard drives is an important pillar in the evolution of modern digital world. Before the advent of flash memory, it dominated the way we saved our data. Simply speaking, binary information (0 or 1) is presented by small magnets pointing forward (0) or backward (1); let’s say the north is the head. Writing the data is by changing the pointing directions of these magnets, usually fulfilled by an electric coiled wrapped head (by applying the current into the head, you create a strong external magnetic field that realign the directions of the small magnets in the hard drive, one at a time). In addition, packing in as many magnets as possible in a limited volume will define the capacity of a hard drive, and this is improving ever since the first device available. Imaging the first computer I had came with a hard drive of 400 MB, and now a decent one has a few TB storage capacities. By comparing the number, you can realize how much effort and advances in the business of data storage. For a very nice introduction, you can find at hard drive 101: magnetic storage.

Figure 1.  A generic ferrimagnet, composed of Fe and Gd, shows the alignment of magnetic moment. Courtesy of I.Radu et al., Nature 472 205 (2011).
A nice paper where ultrafast laser pulses (sub 100 fs) instead of external magnetic field are used to write data intrigued my curiosity. I know immediately that it is the heating effect that causes the change of the magnetization of the small magnets in the hard drive. But for me, the heat has no directionality, how it can tell the magnet to point forward or backward. It should just erase the information since an ultrafast laser pulse can easily create a hot environment above Curie temperature where the magnetization is destroyed. So in my mind, an ultrafast laser is a hard drive terminator, not a hard drive writer. Driven by this curiosity, I dug in to find out, and this is how:

Strong magnets are either ferromagnetic of ferrimagnetic (that is right, only one letter difference). Iron is ferromagnetic, since when it gains magnetization, all the molecules have the magnetic moments (or the moment generator – the atoms’ spins) aligned in one direction. On the other hand, ferrimagnetic materials contain different atoms or same atoms in different chemical forms. For example, the alloy of GdFeCo or magnetite is ferrimagnetic.  When these materials gain magnetic power, the magnetic moments of different atoms (or same atoms with different chemical forms) are pointing in the opposite direction. However the magnitudes are different. As a result, the cancellation is not complete, and some magnetic power retains. This lengthy discussion has a purpose, since in order to use ultrafast laser to write on the device, we need ferrimagnetic materials as our small magnets. Figure 1 is an example of a ferrimagnetic material made of the alloy of GdFeCo. The small arrows are the direction of the spins or magnetic moments if you prefer.

The working principle is like this: When an ultrafast laser pulse hits the ferrimagnetic material (in this paper, a alloy where the active components are Gadolinium (Gd) and Iron (Fe)), the temperature shoots up and all spins are free to flip (or randomize) due to the energy shot. As a result, magnetic moments are decreasing toward zero. In other words, they start to demagnetize (I got this part right). Here is the catch: Fe demagnetize faster than Gd. Like figure 2 shows, when the overall magnet moment of Fe reaches 0 before 0.4 ps, Gd is still decreasing toward zero. At this instant, another principle plays a pivotal role – exchange interaction between Gd and Fe. This exchange interaction says that the flips of the spins of the Gd and Fe can undergo collaboratively, with the spin of a Gd atom flips from up to down, that of a Fe atom goes from down to up. Mother Nature likes the exchange interaction, since it is more energy efficient. Now incorporate this to our discussion. Right after the 0.4 ps, the overall magnetic moment of Gd is still decreasing; it means the spins are flipping down. This gives the spins of Fe atoms a direction, which favors to flip up against Gd’s to please Mother Nature. So the magnetic moment of Fe starts to build up, instead of staying randomized. After the overall magnetic moment of Gd hits zero, the overall magnetic moment of Fe has built up quite a bit. This built-up will guide the entire alloy to relax into an overall ferrimagnetic form. If you are still with me, you know the secret of this technology. With this principle at hands, you can use an ultrafast laser to write the data in an unprecedented speed -- for a few hundreds of picoseconds you can write a bit. And maybe scientists can push this limit even further. 

Figure 2. The evolution of the overall magnetic moment after the sample is hit by an ultrafast laser pulse. As can be seen from the diagram (inset of (b)), Fe demagnetizes faster than Gd. The difference in demagnetization rate makes the reversal of the magnetization possible by an ultrafast laser pulse. Courtesy of I. Radu, et al., Nature 472 205 (2011).
Researchers have demonstrated this phenomenon theoretically and experimentally (this article). As can be seen from figure 3, the magnetic moment flips back and forth after the ultrafast pulses hit the sample.

Figure 3. The magneto-optical image of GdFeCo alloy obtained after the action of a sequence of ultrafast laser pulses. (a) and (b) shows the images of the film with magnetic moment pointing down and up, respectively. (c) and (d) show the magnetization reversal after interacting with ultrafast pulses. The boundary of the circles shows the spot size of the light, and the scale bar is 20 um. Courtesy of T.A Ostler and et al., Nature Communication DOI:10.1038/ncomms1666.
Amazingly, ultrafast lasers are finding their applications at the frontiers in so many fields. They shine and rise in the fundamental chemical physics, advanced spectroscopy, astronomy, machining (Jim has a nice article about it), and now, computer science. Who knows what their next stop will be?

The opinions expressed herein are those of the author and do not represent the Optical Society of America (OSA) or any OSA affiliate.

No comments:

Post a Comment