Static information storage and retrieval – Systems using particular element – Magnetic thin film
Reexamination Certificate
2000-10-26
2002-11-12
Nelms, David (Department: 2818)
Static information storage and retrieval
Systems using particular element
Magnetic thin film
C365S171000, C365S145000
Reexamination Certificate
active
06480412
ABSTRACT:
RELATED APPLICATION DATA
The present application claims priority to Japanese Application No. P11-305520 filed Oct. 27, 1999, and to Japanese application No. P2000-119589 filed Apr. 20, 2000, which applications are incorporated herein by reference to the extent permitted by law.
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a magnetization control method, information storage method, magnetic functional device and information storage device, which are suitable for application to a solid magnetic memory, for example.
2. Description of the Related Arts
Along with epoch-making diffusion of information communication apparatuses, especially, personal-use information apparatuses such as portable terminals, their constituent devices such as memory and logic are requested to have higher performances, such as higher integration, higher speed, lower power consumption, and so on. Especially, progression of nonvolatile memory toward higher densities and larger capacities is getting more and more important as complementary techniques with magnetic hard disks that are essentially difficult to progress reduction in size, increase the speed and decrease the power consumption because of the existence of movable members and other reasons.
Semiconductor flash memory and FeRAM (ferroelectric nonvolatile memory) are currently used in practice as nonvolatile memory, and various efforts are still continued toward higher performances. However, flash memory and FeRAM involve the following essential drawbacks derived from their fundamentals of operation, structures, and materials employed.
<Flash Memory>
1. For writing, injection of hot electrons into floating gates is employed. However, injection efficiency is usually as low as 10
−6
, and it needs a time in the order of ps for accumulating an electric charge sufficient for storage.
2. If an effort is made to increase the injection efficiency in (1) above, (for example, the use of Fowler-Nordheim tunneling injection), the device structure is inevitably complicated. This makes it difficult to realize a higher integration, and such a complicated structure causes a high cost.
3. Either hot electron injection or tunneling injection needs a high voltage (typically more than 10 V), and its power consumption is large. Additionally, an inverter is required for portable use, and this is disadvantageous for miniaturization.
4. A tunneling oxide film around floating gates deteriorates progressively with occurrences of writing. It results in inducing the leak current, promoting outflow of charges for storage, and degrading the reliability. Typical endurance is about 100,000 times.
<FeRAM>
1. Ferroelectric element, which is a metal oxide, is easily affected by a reducing atmosphere that is indispensable for a silicon process, and it does not match such process.
2. Oxides, in general, need a process under a high temperature, which is a disadvantageous condition in microminiaturization of design rules. That is, this disturbs high-density integration. Although it has a device structure enabling integration as high as DRAM, or the like, it is generally considered that those factors limit the degree of integration to 10 Mb/inch
2
, maximum.
As nonvolatile memory not involving those drawbacks, magnetic memory called MRAM (magnetic random access memory) as disclosed by Wang et al., IEEE Trans. Magn. 33(1997), 4498, for example, has been recently brought into attention.
MRAM is a magnetic information storage device in which fine storage media of a magnetic body are regularly arranged, and wirings are provided to allow each of the storage media to be accessed to. Writing is performed by using a current-magnetic field generated by flowing a current in both a selection line (word line) and a read line (bit line) provided above the storage media and controlling magnetization of individual magnetic elements forming the storage media. Reading is performed by using a current-magnetic effect. Since MRAM has a simple structure, high integration thereof is easy, and because of its way of storage by rotation of a magnetic moment, its maximum endurance is large. Additionally, it does not need a high voltage, and needs almost no oxide difficult to make. Just after its proposal, a long access (read) time was a problem. Today, however, where a high output can be obtained by using GMR (giant magnetoresistance effect and TMR (tunneling magnetoresistance) effect, the problem of its access time has been improved significantly.
This MRAM, however, involves an essential problem in its writing method from the standpoint of more advanced integration. More specifically, as the wirings become thinner along with progressively higher integration, the critical value of the current that can be supplied to the write line decreases, and therefore, the magnetic field obtained becomes so small that the coercive force of the storage region must be lowered. This means that the reliability of the information storage device degrades. Further, since magnetic field, in general, cannot be converged unlike light or electron beams, its high integration may cause cross-talk. Thus, writing relying upon a current-magnetic field essentially involves those problems, which may become large drawbacks of MRAM.
Similarly to flash memory and FeRAM, those problems will be removed if magnetization can be controlled by mere electric stimulation, i.e., without using a current-magnetic field. A method according to this concept was already proposed, and there is a technique that uses a structure in which two ferromagnetic layers are separated by a semiconductor layer, as disclosed by Mattson et al., Phys. Rev. Lett. 71(1993) 185, for example.
This technique uses the fact that magnetic coupling between ferromagnetic layers depends on carrier concentration of a semiconductor layer interposed between them. In a multi-layered structure stacking such ferromagnetic/semiconductor/ferromagnetic layers, magnetic coupling between ferromagnetic layers can be changed from parallel to anti-parallel, for example, by controlling carrier concentration of the semiconductor layer as the intermediate layer. Therefore, if the coercive force in one of the ferromagnetic layers (fixed layer) is set large, then magnetization of the other ferromagnetic layer (movable layer) can be rotated relative to the fixed layer. Such method of rotating magnetization through an electric input is hopeful as a technique for realizing a compact all-solid-state component.
In the above-mentioned multi-layered structure stacking ferromagnetic/semi-conductor/ferromagnetic layers, magnetic interaction occurs indirectly between the ferromagnetic layers via the semiconductor layer. In this case, the semiconductor layer as the intermediate layer must be sufficiently thin because intensity of interaction between the ferromagnetic layers via the semiconductor layer attenuates exponentially with the thickness of the semiconductor layer.
Assume here that, for the purpose of obtaining an actual intensity of interaction, a coercive force of 100 Oe (Oersted) is given by a method like exchange biasing in a nickel-iron alloy having the thickness of 2 nm and the saturated magnetization of 12500 G (Gauss), for example. In order to give an energy equivalent to an energy necessary for reversing magnetization of this nickel-iron alloy by indirect interaction via the semiconductor layer, it is estimated by simple calculation that the exchange coupling constant J must be at least 0.02 mJ(Joule)/m
2
. Apparently, to ensure interaction not smaller than that value, thickness of the semiconductor layer as the intermediate layer must be 2.5 nm or less, as indicated by J. J. de Vries, Physical Review Letters 78(1997) 3023, for example. This is a term imposed to the semiconductor layer as the intermediate layer to provide a practical device.
On the other hand, in order to control magnetic coupling between the ferromagnetic layers by controlling carrier concentration of the semiconductor layer as the intermediate layer, an electrode has to be attache
Bessho Kazuhiro
Iwasaki Yoh
Le Thong
Nelms David
Sonnenschein Nath & Rosenthal
Sony Corporation
LandOfFree
Magnetization control method, information storage method,... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Magnetization control method, information storage method,..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Magnetization control method, information storage method,... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2946282