One of my primary research interest is focused on the field of:

1. Laser-induced ultrafast magnetization dynamics in (metallic) thin films;
2. Laser Induced magnetization reversal.

Introduction

Magnetism is certainly one of the cornerstones of what we today call the information technology era. Following the huge development of electronic devices such as the personal computer, this era has been strongly driven by the growing demand to increase the density and the speed of writing and retrieving data in memory devices. Today, the demand for information storage is enormous and expected to increas
even further as new technologies such as high-definition video and on-demand TV are established in the market. In this quest, driven by the two words "smaller and faster", magnetic recording remains the dominant data storage technology. The reason for this is the unmatched combination offered by magnetic recording: large storage capacity, small physical size, random and fast access to data, non-volatility, radiation hardness and low cost. Consequently, the science and technology of magnetic materials is strongly fuelled by the global market for magnetic storage devices, which was estimated at $20 billion in 2005, and is expected to grow to $40 billion in 2010. Some examples of storage devices that are based on magnetism such as hard disk drive (HDD), magneto-optical disk and magnetic random access memory (MRAM),
are shown in Figure below.

Figure: Illustration of various memory devices based on magnetism. In a Hard Disk Drive the binary state of the storage element is represented by magnetic domains with magnetization vector oriented either "up" or "down" corresponding to the values "1" or "0". On the other hand, in MRAM (magnetic random access memory) the values "1" or "0" are recorded on the two opposite orientation of the magnetization in one of the magnetic layers of magnetic tunnel junctions, which are connected to the crossing points of the two perpendicular arrays of parallel conducting lines. Here, the information is written by current pulses sent through one line of each array. Only at the crossing point of these lines the resulting magnetic field is high enough to reorient the magnetization in one of the two magnetic layers - called the free layer.

Magnetic Data Storage and its Speed

In magnetic memory devices, logical bits ("ones" and "zeros") are stored by setting the magnetization vector of individual magnetic domains either 'up' or 'down'. The size of these domains is what defines the density of information in a memory device. In other words, the smaller the bit size is, the higher is the storage capacity of data storage media. The evolution of the bit size has been impressive since 1955, when IBM built the first hard disk drive [this first HDD called RAMAC (Random Access Method of Accounting and Control) consisted
out of 50 disks, each with a diameter of 24 inch (61 cm)], featuring a storage capacity of about 5MB (5 million 8-bit characters) with areal recording density of 2 kbit/in^2 [the capacity of a HDD is in general referred to as the number of bits (or bytes - series of 8 bits) that can be squeezed on a square inch of a storage medium]. This corresponded to a bit size of ~0.55mm x 0.55mm. The barrier of 100 Gbit/in^2 has already been passed in 2002. Today there are new technologies such as Seagate's Heat Assisted Magnetic Recording (HAMR) that promises recording densities beyond 1 Tbit/in^2, i.e. bit sizes of ~10nm x 65nm. As the bit size becomes smaller, the information recording speed must also became faster. The current disk drives are operating at an internal data transfer rate of approximately 200 MB/s, which corresponds to a channel data rate of about 1.6 Gbit/s [2, 3]. Therefore, the writing time for a single bit, or in other words the magnetization reversal time in a bit, is not far below 1 nanosecond. Although today a recording speed in the nanosecond range might be considered high enough with respect to the available densities, there is no doubt that the future Tbit/in2 asks for a much higher recording speed of the information, that is a faster magnetization reversal time.

Besides to disk storage, a tremendous research eort has been devoted in recent years to MRAM. This magnetic memory has the potential to store data at a relatively high density, high speed, and to have a low power consumption. Although currently Flash memory still offers higher areal density, MRAM has a potentially innite endurance compared with ~ 105 cycles for a Flash. Such characteristics together with its non-volatility makes MRAM an "ideal memory". Generally speaking, the most common design for MRAM uses a magnetic tunnel junction: two ferromagnetic thin lms as electrodes and a thin tunneling barrier separating them. The resistance of the tunneling junction is modifed as the magnetic moments of the two ferromagnetic
layers change their relative orientation. The diference in junction resistance corresponding to the stable parallel and anti-parallel orientations, respectively, allow the defnition of binary memory states. In July 2006, Freescale started selling the first commercial MRAM chip, with 4Mbit of memory while recently Toshiba and NEC have announced a 16 Mbit MRAM chip, with 34 ns read and write cycles [4]. Obviously, there is a large space here for improving the speed of manipulating data if appropriate new technological concepts are introduced.

Traditionally, in order to reverse the magnetization, and thus write or rewrite the information, an external magnetic field pulse is applied. The operation time for this magnetization reversal mechanism lies in the nanosecond regime. Increasing the strength of the magnetic field, the magnetization reversal time can be pushed into the picosecond range. However, by trying to do this new challenges appear. In particular, the write poles approach their limits in achieving strong and short field pulses for HDD. In addition, it must be noted that the increase in the density of the recorded information is achieved by using materials with very high magnetic anisotropy. In these conditions, the strength of the writing field must increase even more. There are more challenges regarding the actual writing process in which a coil is used. For example, in order to not affect the neighboring data the writing field distribution must be scaled down as the density of information increases. On top of all these challenges, it has been recently predicted that no matter how short and strong the magnetic field pulse, magnetic recording cannot be made ever faster than about 2 picoseconds (2 x 10^-12 s) [5]. Having all this in mind, it becomes clear that the actual technology is fast approaching its speed limitations.

Ultrafast laser pulses [as it was recently defined [1], ultrafast is everything what is happen on a time-scale shorter than 100 picoseconds (0.000 000 000 1 seconds)] could represent the key to access a new ground for the magnetic data storage technology. In fact, the laser has already proven to be able to solve some of the above mentioned problems. In particular, the need to use materials with high magnetic anisotropy, converted the conventional magnetic recording into a hybrid magnetic recording - HAMR - in which a laser beam is used to locally heat the storage medium and thus decrease its anisotropy, while the data is simultaneously written magnetically with the traditional scheme. However in this scheme the laser is only used to heat the material while as was shown in my PhD thesis, its potential for data processing goes beyond simple heating.

With the recent developments of ultrafast femtosecond lasers (currently, ultrafast lasers are generating pulses with a typical duration of 10^-13 to 10^-14 s, that represent some of the shortest man-made events), the study of ultrafast magnetization dynamics has become one of the most active fields of magnetism fuelled by both scientifc and technological interest. In this quest, the ultrafast optical manipulation of the magnetization promises to become a real alternative to the magnetic field pulses. Note that the time-scale offered by femtosecond laser pulses for manipulating the magnetization is orders of magnitude shorter than the magnetization reversal time in actual memory devices. Indeed, the first experimental studies on magnetization dynamics using femtosecond lasers have uncovered a sub-picosecond demagnetization of magnetic metals [6]. Following this experiment, a wave of exciting result appeared recently: excitation of coherent spin waves via optically changing the magnetic anisotropy fields [7, 8]; optical excitation of high frequency spin oscillations (~400 GHz) in antiferromagnets [9, 10]; small-angle ultrafast switching of magnetization in garnets via the opto-magnetic inverse Faraday effect [11] and laser-induced coherent spin dynamics at a frequency of several THz [12, 13]. In all these experiments, though fast, the laser excitation only bring the spin out of equilibrium (about several degrees) for a certain amount of time but does not accomplish a complete ultrafast magnetization reversal, as required for the data storage. Consequently, one of the biggest challenge in the field of ultrafast magnetization dynamics is to find ways to ultrafast induce (180^0) magnetization reversal.

If you would like to know how this challenge can be solved, please read my PhD thesis.

References

[1] J. Stöhr, H. C. Siegmann, Magnetism. From fundamentals to Nanoscale Dynamics (Springer-Verlag, Berlin, 2006).
[2] A. Taratrin, et al., IEEE Trans. Magn. 38, 1873, 2002.
[4] http://www.mram-info.com/history
[5] C. H. Back and D. Pescia, Nature 428, 808 (2004).
[6] E. Beaurepaire, J.-C. Merle, A. Daunois, and J.-Y. Bigot, Phys. Rev. Lett. 76, 4250 (1996).
[7] G. Ju, A. V. Nurmikko, R. F. C. Farrow, R. F. Marks, M. J. Carey, and B. A. Gurney, Phys. Rev. Lett. 82, 3705 (1999).
[8] M. van Kampen, C. Jozsa, J. T. Kohlhepp, P. LeClair, L. Lagae, W. J. M. de Jonge, and B. Koopmans, Phys. Rev. Lett. 88, 227201 (2002).
[9] A. V. Kimel, A. Kirilyuk, A. Tsvetkov, R. V. Pisarev, and Th. Rasing, Nature 429, 850 (2004).
[10] A. V. Kimel, C. D. Stanciu, P. A. Usachev, R. V. Pisarev, V. N. Gridnev, A. Kirilyuk, and Th. Rasing, Phys. Rev. B 74, 060403(R) (2006).
[11] F. Hansteen, A. Kimel, A. Kirilyuk, and Th. Rasing, Phys. Rev. Lett. 95, 047402 (2005).
[12] A. Melnikov, I. Radu, U. Bovensiepen, O. Krupin, K. Starke, E. Matthias, and M.Wolf, Phys. Rev. Lett. 91, 227403 (2003).
[13] I. Radu, Ph.D. thesis, Berlin, Germany, 2006.