JP2002368195A - Memory material and method for manufacturing the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a memory material which is nonvolatile, high-speed, randomly accessible, static and capable of being updated. SOLUTION: A composition material with ferromagnetic, piezoelectric and electro-optical characteristics is disclosed. In a preferred embodiment, this composition material is composed of a first layer of M(1-x-y) Cdx Siy (M being Te or Zn), a second layer of Se(1-z) Sz and a third layer of Fe(1-w) Crw , where x, y, z and w have values within the range of 0.09<=x<=0.11, 0.09<=y<=0.11, 0.09<=z<=0.11 and 0.18<=w<=0.30. Further, each layer contains at least one element from among Ag, Bi, O, or N. Also, a nonvolatile memory, made of this composition material which is randomly accessible, is disclosed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a material having a novel composition which has ferromagnetic, piezoelectric and electro-optical properties and is used as a storage medium.

[0002]

2. Description of the Related Art Computer technology requires large-capacity and high-speed memories. Generally, in modern computers, a semiconductor memory is used as a high-speed main memory, and a magnetic disk is used as a large-capacity secondary memory.

Prior to the development of semiconductor memories, high-speed main memories using magnetic core memories were used. The magnetic core memory includes a matrix-shaped annular ferromagnetic core. Each memory cell of a magnetic core memory includes a ferromagnetic core having two or more wires passing through the center of the core, and a sensing coil disposed around the core.

When a current I is applied to a wire passing through a core,
A magnetic field is formed, the magnetic field strength H of which is a function of the current I. The magnetic field generated by the current causes a permanent magnetization of the core, which is determined by the induced magnetism B. The relationship between B and H is substantially hysteretic,
As a result, the BH diagram, known as the magnetization curve or BH loop, is substantially square.

The induced magnetic field B in the core has two states, Br and -Br, and is opposite to the direction of the magnetic field. As a result, each core combines one state of "1" with another state of "0" to form one bit of 2 bits.
Stores hexadecimal data. In the example, + Br is associated with a binary "1" and -Br is associated with a binary "0".

[0006] Binary data is written into one core memory cell by passing an appropriate current through the wire. If the total current through the core is greater than the critical current Ic, the core induction changes from -Br to + Br. Similarly, if the current is less than -Ic, the induced magnetism is converted from + Br to -Br. Preferably, 1
In the magnetic cores of a row, the conversion is performed by the simultaneous application of signals on two or more wires. Thus, initially, if the induced magnetism has a value of -Br corresponding to "0", a binary "1" means that a current of I> Ic / 2 is applied to each of the two wires. So that the total current through the core is greater than + Ic which causes the induced magnetism to change to + Br.

[0007] The data stored in the core is read by sensing the coil voltage induced by the conversion between the two magnetic states described above. The polarity of the induced voltage indicates the magnetic state of the core before conversion.

Although the above-mentioned magnetic core memory is randomly accessible and non-volatile, such a memory is large, has high power consumption, operates slowly, and cannot be manufactured with a high storage density. . To overcome these problems, magnetic thin film memory devices have been developed. The magnetic thin film memory is composed of a strip-like ferromagnetic thin film, and two or more wires for writing data are formed on the thin film, and a coil for reading data is formed around the thin film.

In a thin film memory, the magnetic moment M of the film represents stored information. The magnetic moment M is initially arranged in the direction of the film surface and has two discrete arrangements or states representing binary "1" and "0", namely M and -M. A current is passed through a wire formed on the thin film to store one bit of binary data. These currents induce a magnetic field sufficient to change the direction of the magnetic moment M. The stored information is
It is read by passing a current through the wire and measuring the voltage induced in the coil. As in a magnetic core memory, the current is usually selected to be large enough that a single current cannot reverse the magnetic moment of the film, so that at least two simultaneously applied currents are used to store data. Needed for However, there are significant weaknesses in magnetic thin film memory technology. First, the thin film device has an open flux structure, so that the BH loop is compromised by the self-demagnetizing effect. To reduce this effect, the membrane is usually manufactured in a rectangle whose length is much larger than its width. Since the induced voltage of the coil around the film is proportional to the cross-sectional area of the film, reducing the width of the film also reduces the induced voltage. As a result, the read signal is easily affected by noise.

Second, in existing magnetic films, the magnetic moment is usually in-plane. Thus, the device is complicated by the need to apply different magnitudes of current to store and read data in the selected orientation. Moreover, this thin film device is too large for high density.

[0011] Compared to magnetic core memories and thin film memories, semiconductor memories are faster, consume less power, and have higher storage densities. Typical semiconductor memories include dynamic random access memory (DRAM), static random access memory (SRAM), and read-only memory (ROM).

A DRAM is high-speed, high-density, low-power consuming, and is readable and writable. However, both DRAM and SRAM are volatile, and stored information is lost when power is cut off. In addition, DRAM
It requires constant refresh of stored data that requires complicated circuits. Although the SRAM does not require refreshing, it consumes a large amount of power and does not have a high storage density.

The ROM is non-volatile, but cannot update the information stored in the ROM. That is, data cannot be easily written to the ROM.

In a typical disk storage system, a ferromagnetic material having a substantially square BH loop is coated on the disk; and a magnetic head reads and writes information on the disk rotating across the head. The disk is divided into circumferential tracks. Each track is further divided into sub-regions, where the magnetic moment has two states representing binary values. The external magnetic field generated by the read / write head converts the magnetic moment of each sub-region to store a binary value in that region. Thus, to write data, the magnetic head magnetizes a small area adjacent to the rotating disk material. The stored data is read out in the form of a voltage induced on the head by the magnetic moment of a small area moving across the head.

[0015] Magnetic disk storage systems are capable of storing large amounts of data, for example, 500 megabytes or more. However, magnetic disk storage systems are not randomly accessible, operate slowly because they require mechanical drive, and require complex mechanical and electrical equipment.

[0016]

Obviously, none of the above memory technologies provide all the required characteristics in a memory storage system. Thus, there is a current need to develop storage systems and memory materials that are non-volatile, fast, randomly accessible, static, and updatable.

[0017]

The material composition according to the present invention is preferably composed of Pb _(1-xy) Cd _x Si _y , S
made from a layer of e _{(1-z) S} z, and _{Fe (1-w) Cr w} , where x, y, z and w represents the composition ratio in each layer. These values are usually preferably in the following range: 0.09 ≦ x ≦ 0.11, 0.09
≦ y ≦ 0.11, 0.09 ≦ z ≦ 0.11, and 0.
22 ≦ w ≦ 0.36. In a preferred embodiment, the material layer of this composition also comprises the following elements: Bi, Ag,
Contains O and N. These elements are Bi ₂ O ₃ and A
It is mixed in by electrolysis of the gNO ₃ containing solution.

In a memory device according to the present invention, two sets of parallel address lines are arranged at right angles on opposite sides of a flat substrate. As described above, this layer of the novel composition material, together with the outermost FeCr layer, is provided on both sides of the base on the address line, and the electrodes are connected to the outermost FeCr layers on each side of the base. I have. Each memory cell is located at each intersection of two sets of address lines.

By applying appropriate current pulses to the two address lines, two independent information bits are
Magnetically stored in a single memory cell. This information is read out as a piezoelectric voltage between the electrodes generated in response to the appropriate current pulses applied to the two address lines.

More specifically, two synchronous current pulses of the same strength and polarity are applied to two orthogonal address lines to store and read a first bit of information in a memory cell. The second bit is stored and read into the memory cell by applying two synchronous current pulses of the same strength but of opposite polarity to the two address lines. The current pulse used to store the binary information has an intensity which is not sufficient to modify the stored information in a single pulse, but sufficient for the stored information in two synchronization pulses. Have. The current pulse used to read the stored binary information is of an intensity that does not cause a change in the stored information.

Such a memory cell is non-volatile, randomly accessible, static, operates at high speed,
It requires low power, is readable and writable, and can be manufactured in a high-density array.

[0022]

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a memory composition material having ferromagnetic and electro-optical piezoelectric properties. Also,
A random accessible non-volatile memory device using the composition material of the present invention is disclosed. Preferably, the memory device is capable of storing two independent information bits.

The composition material is preferably Pb
_(1-xy)Cd_xSi_y, Se_{(1- z)}S_z,
And Fe_(1-w)Cr_wLayers. x, y, z and
The value of w is 0.09 ≦ x ≦ 0.11, 0.09 ≦ y ≦
0.11, 0.09 ≦ z ≦ 0.11, and 0.22 ≦
It is preferable that w is within the range of 0.36. Also good
Preferably, each layer is one of Bi, Ag, O and N or
Contains more elements.

Alternatively, Pb _(1-xy) Cd _x Si
_{For the y-} layer, Ge may be used instead of Si and / or Zn or Te may be used instead of Pb.
Also, other conductors such as Au, Pt or Cu
It can be added to the layer structure instead of g. Also, the present invention
In the _{Fe (1-w) Cr w} , w is from 0.18 0.
It is also achieved using a Cr concentration such as in the range of 30.

In particular, as shown in FIG.
One preferred embodiment of the composition material of_0.80Cd
_0.10Si_0.10Layer 110, Se_0.90S
_0.10Layer 120, and Fe_0.76Cr_0.24Layer 1
30. This Fe_{0. 76}Cr_0.24layer
Plays a major role in the ferromagnetic properties of this material composition.
Yes, Pb_0.80Cd_0.10Si_0.10And Se
_0.90S_0.10Is mainly due to its electro-optical properties.
Take charge. All three layers have piezoelectric properties.

In the apparatus described below, these layers are
0, Pb_{0. 80}Cd_0.10S
i_0.10, Se_0.90S_0.10, And Fe
_0.76Cr_0.24Each layer is 0.5 μm thick
is there.

The physical properties of the composition material of the present invention will be described later. By knowing these properties, the operation of a memory device using this composition material can be understood.

According to the background, a ferromagnetic material forms a permanent magnetic field in the absence of an external magnetic field. Such materials are described in the form of a number of tiny magnets known as magnetic dipoles. An external magnetic field applied to the ferromagnet aligns the magnetic dipoles in the material in the direction of the applied magnetic field,
The total magnetic field in the material is then the sum of the magnetic fields generated by the magnetic dipoles aligned with the external magnetic field. When the external magnetic field is interrupted, the arrangement of the magnetic dipoles does not change, thereby creating a constant magnetic field in the material. Magnetic information storage is based on this property of ferromagnetic materials.

FIG. 2 is an example of a magnetization curve of a typical ferromagnetic material. Also, the magnetization curve is referred to as a BH loop. In this figure, the y-axis represents the induction magnetism B, and the x-axis represents the magnetic field strength H of the external magnetic field. Thus, BH
The loop shows the change in the induction magnetism B with respect to the magnetic field strength H.

The BH loop of FIG. 2 will be analyzed in further detail. Initially, assuming that the directions of the magnetic dipoles of the ferromagnetic material are uniformly distributed in all directions, the total value of B will be zero (point "a" on the curve) in the absence of an external magnetic field. When an external magnetic field is applied to the ferromagnetic material, the B value gradually increases as H increases, and reaches a point (point “b” on the curve) where the induced magnetic field B starts to saturate. In other words, when H reaches a certain value, B remains substantially at B ₀ even if H increases. After saturation, the induced magnetic field B does not return to point "a" (B = 0) even if the external magnetic field decreases to H = 0. Instead, the B value remains approximately at B = B ₀ (point “c” on the curve).

At point "c", the direction of the external magnetic field H is reversed. The external magnetic field changes the polarity of the magnetic field B to approximately H = -Hc, and at point "e" the magnetic field has the opposite polarity B = -B
Saturates at _zero .

As the magnetic field strength H increases, B changes from point "e" to point "b" as shown in FIG.

FIG. 3 shows a BH loop of the composition material according to the present invention. 2, the x-axis represents the magnetic field strength H of the external magnetic field, and the y-axis represents the induced magnetism B. It is important for the composition material according to the present invention that the shape of the BH loop is that the angle α between the y-axis and the BH loop at B = 0 is 1 ° or less and is substantially square. Due to the substantially square shape of the magnetization curve, the induced magnetism B is stably stable in two discrete and stable states + B ₀ and -B
One of ₀ . As a result, the new composition material is suitable for storing binary information.

The composition material according to the present invention also has piezoelectric characteristics. Generally, when the mechanical pressure on the piezoelectric material decreases, a piezoelectric voltage is generated. In the present invention, when the mechanical pressure on the composition material decreases substantially perpendicular to the plane of the layer in the composition material, a piezoelectric voltage is developed across said layer. In the present invention, the change in mechanical pressure is caused by a change in the magnetic state of the composition material.

The structure shown in FIG. 4A shows the piezoelectric characteristics of the present invention. The operation is explained in FIG.
-(J).

Referring to FIG. 4A, the structure 190 is composed of two layers of the composition material of the present invention. For more information,
This structure includes a first FeCr layer 200, a first SeS layer 210, a first PbCdSi layer 220, a second PbCd
Si layer 240, second SeS layer 240, second FeCr
It is composed of a layer 250. In addition, a wire 260 passes through the center of the structure, parallel to the layer.

As shown in FIG. 4 (b), when a current is applied to the wire in a direction to enter the drawing page, the wire is moved in a clockwise direction indicated by an arrow as indicated by a circle Br. , A substantially annular magnetic field is generated. Arrow 27
0 indicates that the FeCr layers 200, 250
Shows the direction of the magnetic dipoles inside. As shown in FIG. 4 (b), this structure is divided into two parts 275, 280 symmetric about a vertical axis 265 perpendicular to the wire 260.
275, 280, the dipole arrangement of FIG.
Equivalent to two magnets of the same strength with north and south poles indicated by arrows 282, 284 in (c). The length of each arrow represents the magnitude of the induced magnetism B of the corresponding magnet. By the attractive force between the S pole S and the N pole N of each magnet,
The storage medium is mechanically pressed in the vertical direction of the layers of the structure.

The BH loop of the induced magnetic Br is shown in FIG. As described above, the BH loop is substantially square and has two discrete and stable magnetic states + B
It represents ₀ and -B _0.

Furthermore, the magnetic field has a critical magnetic field strength Hc, defined as the magnitude of the magnetic field strength, which causes a conversion between + B ₀ and −B ₀ . As a result, if Ikere size than H is Hc, induced magnetism Br becomes + _{B 0} value. H is -H
is smaller than c is, Br becomes -B ₀ value.

Initially, under an applied external magnetic field, the magnetic state becomes
Assume that the point is “a” on the curve of FIG. 4D where the induced magnetism is + B ₀ . Set the magnetic state of the storage medium to + B ₀
To change from to −B ₀ , the current through wire 260 is reduced to reduce magnetic field strength H. When the current is zero, the magnetic intensity H is also zero (point “b” on the BH loop). As described above, due to the ferromagnetic properties, the magnetic state of the storage medium is B even when the external magnetic field disappears.
_It remains at ₀ . That is, the information represented by the induced magnetic B ₀ is held.

When the direction of the current is reversed, the magnetic field strength continues to decrease. Induced magnetism B at point "c" reaches the smaller Bc value than _{B 0.} At this point, the dipole moment continues to decrease as shown in FIG. 4 (e), as the dipole begins to rearrange in the opposite direction. As a result, the mechanical pressure applied to the FeCr layers 200, 250 by the attraction force of the layers continues to decrease. A change in the pressure of the layer causes a piezoelectric voltage to develop vertically across the layer. H = -H
At point "c" where c and the induced magnetic field B are zero, the pressure on the layer is minimized because the dipoles are arranged in opposite directions. At this point, the induced piezoelectric voltage reaches a maximum because the pressure change in the layer is at a maximum.

As H continues to decrease below -Hc,
Magnetic states is converted into point "e" from the point "d", then converted to a point "f" to reach the second stable state B = -B _0. FIG. 4F illustrates that the magnetic pole is reversed at the point “f”. Thus, at point "f", the mechanical pressure on the layer returns to its initial value,
Decrease the piezoelectric voltage. Increasing the current in the opposite direction (from point "f" to "g") does not increase the magnitude of the dipole moment and therefore does not increase the mechanical pressure of the layer.

[0043] FIG. 4 (g) is between _{the B 0} changes to -B _0, shows a piezoelectric voltage corresponding to some point on the BH loop of Fig. 4 (d). In FIG. 4H, the piezoelectric voltage generated in response to the current pulse is shown in the time domain. The piezoelectric voltage is one piezoelectric voltage pulse,
It is later than the application point of the current pulse. Current pulse is applied to the wire has a sufficient strength -I to convert from B ₀ to -B _0.

[0044] Similarly, the conversion from the magnetic state -B ₀ to _{B 0} is occurring a negative piezoelectric voltage pulse.

As shown in FIG. 4D, Hc
A current that produces a magnetic field with a greater intensity is required for conversion between magnetic states. However, FIG.
If a current smaller than the intensity indicated by the point “c” in (e) is applied, the magnetic state becomes unstable. In this case, the induction magnetic B will oscillate between _{B 0} value (point "b") Bc value (point "c"). FIG. 4 (i) shows a piezoelectric voltage pulse generated in response to such vibration. Strength of the piezoelectric voltage pulse V2 is from + _{B 0} -B ₀ smaller than the pulse intensity caused by the conversion into (FIG. 4 (g)).

FIG. 4 (j) shows a current pulse in which B is a Bc value. The piezoelectric voltage pulse generated in response to this current is shown at the bottom of the figure. The shaded area reflects the oscillation between the two states (Bc and + B ₀ ) and is visible on an oscilloscope. As will be explained subsequently, a piezoelectric voltage generated in response to the current that perturbs but does not change the magnetic state is used to read magnetically stored information.

FIGS. 5A and 5B are partial cross-sectional views (not to scale) of a memory device 290 according to a preferred embodiment of the present invention.
And a top view. This memory device comprises a silicon planar substrate 330, a first address line 320 formed on the surface of the substrate, and a second address line 340 formed on the opposite surface of the substrate and orthogonal to the first line. A first set 310 and a second set 350 of material layers according to the present invention are disposed on address lines on opposite sides of the substrate. The electrodes 300, 360 are connected to layers 310, 350 of this composition material, respectively.

The first and second address lines have a width of about 2 μm.
m, a silver strip having a thickness of about 1 μm. As shown,
The spacing between adjacent address lines is about 9-20 μm, depending on the required density of the memory device. For example, in one embodiment, the spacing is 9.5 μm;
9 μm. Each set of material layers 310, 35
0 is a Pb _0.80 Cd _0.10 Si _0.10 layer sequentially formed on one of the address lines 320 and 340 on the Si substrate on which the two FeCr layers are outermost, and Se
_0.90 S _0.10 layer and Fe _0.76 Cr _0.24
Consists of layers. Each layer preferably has a thickness of 0.5 μm.
And therefore each set is 1.5 μm. Each layer is homogeneously filled with Bi, Ag, O and N. Preferably, the substrate is 40 μm thick and the electrodes are 1 μm
It is a thick silver layer.

The manufacture of this device was performed first with a 40 μm thick silicon
A 1 μm thick metal (preferably,
Silver) layer. Alternatively, BaF2
Substrate made of another material such as
Can be used for Coating is performed by thermal evaporation, electron beam evaporation,
This is done using conventional techniques such as taring. Next
The coated silver layer is patterned by photolithography.
And a set of metal strips, each about 2 μm wide.
Etched to form a loop. One of Si base
The set of strips on one side is orthogonal to the strips on the other side.
You. The strips on both sides of the base have a crossbar structure.
ing. Then, Pb_0.80Cd_{0.1 0}Si
_0.10, Se_0.90S_0.10, And Fe_0.76
Cr_0.24Are sequentially coated. Prior to coating, the layer
Elements are mixed with the appropriate amount of powder for each required element
It is made by doing. The amount of powder for each element corresponds
The required ratio of the elements in the layer to be formed. example
If Pb_0.80Cd_0.10Si_{0 . 10}About the coating
The layers of Pb, Cd and Si powder are 80: 10: 1
Mix at 0 ratio. Pb, Cd and Si are well mixed
After being mixed, the mixture is
Pressed and calcined to form a material source. S
e_0.90S_0.10And Fe_0.76Cr_0.24of
Coated source materials are similarly made.

Pb_0.80Cd_0.10Si_0.10,
Se_0.90S_0.10And Fe_{0 . 76}Cr_0.24
The layers are then sequentially coated on both sides of the substrate. This coating
Is performed by a known method. For example, the preferred implementation
In the example, plasma sputtering technology creates a multilayer structure
Used to Each layer is coated by sputtering
After that, the temperature of the layer rapidly rises to about 500 ° C.
(Eg, about 1.5 seconds) and then the next layer
To about room temperature. As usual
Sputtering is performed in a vacuum using Ar gas.
You. As described above, Pb_0.80Cd_0.10Si
_0.10, Se_0.90S_0.10And Fe_0.76C
r_0.24Each of the layers is 0.5 μm thick and 2
One FeCr layer has a thickness of 1. on the opposite side of the outermost substrate.
Two 5 μm structures are formed.

Next, Bi, Ag, 0 and N are Bi₂
O₃And AgNO₃Using High Temperature Electrolyte Containing Oxygen
Added to the layer by the process. Electrolysis is at the bottom of the vessel
Pure water in a stainless steel container with stirring means
This is done by heating to 97 ° C. Weight of added powder
The ratio is preferably Bi₂O₃Is 40% and AgNO ₃
Is 60%. The amount of these powders in the electrolyte
It is adjusted to obtain the expected current. Add the powder
After that, the electrolyte is maintained at 97 ° C. to form a homogeneous solution.
Stirring for at least one hour.

Prior to the electrolytic process, all metal strips on both sides of the substrate are connected to form a single electrode. Next, the substrate is immersed in an electrolyte maintained at a temperature of 97 ° C. and continuously stirred. Preferably, multiple substrates are simultaneously immersed in the electrolyte. For example, one hundred
A cm × 1 cm substrate is processed simultaneously. In this case, the metal strips of all the substrates should be connected to a single electrode.

The complete electrolysis process takes 45 days. Every day, the same process cycle is repeated. During the first 10 hours of the cycle, a voltage of +60 V is applied to the substrate, and for the next 14 hours, a voltage of -60 V is applied to the substrate. Stainless steel containers are always kept at ground voltage. Also, every 12 hours during the electrolysis process, the position of the substrates in the vessel is rearranged to make the process uniform. During the process, the electrolyte is continuously stirred.

Throughout the process, the current intensity in the electrolyte is monitored. FIG. 12 (a) illustrates the current amperes I for the first 40 days of the process. The number of days is shown on the horizontal axis "t". FIG. 12B illustrates the value of the current I for the last five days. The current I is in mA.

At the end of the 45th day, the electrodes are detached from the substrate and the substrate is withdrawn from the electrolyte. At this point, the ions of the above-described elements sufficiently penetrate the multilayer structure. It should be noted that in different embodiments, ion implantation techniques are used to introduce these elements into the multilayer structure. The composition material on both sides of the substrate is then polished until its surface is sufficiently smooth. Subsequently, a silver layer having a thickness of about 1 μm is coated on the surface of each substrate to form the electrodes 300 and 360.

When this procedure is completed, a new two-dimensional memory array is manufactured. A region near each intersection of the orthogonal address lines 320 and 340 becomes a memory cell.

More specifically, as shown in FIGS. 5 (a) and 5 (b), one set of metal strips on the lower surface of the Si substrate is provided with a first set of address lines (X lines). ), And one set of metal strips on the top surface of the substrate form a second set of address lines (Y lines). When two currents Ii and Ij are applied simultaneously to a line Xi with an X line and to a Yj with a Y line, respectively, the memory device (i, i,
j) will be selected. By properly choosing the intensity and polarity of the currents Ii, Ij, information is stored and read out in the memory device (i, j). As described above, the memory array including the memory device according to the present invention can be randomly accessed.

The process of storing information in one cell is shown in FIGS. 6 (a), 6 (b), 7 (c), 7 (b), 8
(A), 8 (b), 9 (a) and 9 (b), which are evident from FIGS. 10 and 11. 6A and 6B are top views of a single cell of the present memory device.

The orthogonal address lines 325, 345
As shown in FIG. 6A, the cell is divided into four sections 37.
0, 375, 380 and 385. As described below, the first bit of information is magnetically stored in sections 370 and 380, and the second bit is magnetically stored in sections 375 and 385. For simplicity, the sections 370 and 380 where the first information bits are stored are collectively referred to as carrier "a" and the sections 375 and 385 where the second bits are stored are collectively referred to as carrier "b". You.

To store one bit of information on one carrier of the memory device, two currents having a particular strength and polarity are applied to the first and second address lines. Information is extracted by applying two currents to the address line and detecting a piezoelectric voltage generated between the upper and lower electrodes. The current applied to the first address line is represented as Ii, and the current applied to the second address line is represented as Ij. The directions of Ii and Ij are indicated by arrows written on the address lines. In the preferred embodiment, current Ii and Ij have the same intensity _{I 0.} Each current is
An induced annular magnetic field is created around the address line, as indicated by arrows 390 and 395.

The directions of the magnetic fields Bi and Bj generated by Ii and Ij in each section are shown in FIGS. 6 (a) and 6 (b). A dot (•) indicates that the magnetic field is in the upward direction, and a cross (x) indicates that the magnetic field is in the opposite direction and is downward.

As shown in FIG. 6 (a), in compartments 485 and 475 (carrier "b"), Bi and Bj are in opposite directions, thereby canceling each other. For this reason,
The current shown in FIG. 6 (a) does not affect the information stored in the carrier "b". Conversely, compartments 470 and 480
In (carrier "a"), the magnetic fields Bi and Bj are induced in the same direction. As a result, these fields strengthen each other,
As a result, the stored information can be changed.

Thus, a current of the same polarity and strength applied to the address line affects only the magnetic state of carrier "a", thereby selecting this carrier. Similarly, select the two negative pulses and store the data in carrier "a". It should also be noted that the magnitude of the current selecting carrier "a" may not be equal, as long as the combined effect does not change the magnetic state in carrier "b".

FIG. 6 (b) illustrates the process of selecting carrier "b". Opposite polarity Ii = + _{I 0} and Ij =
Current having a -I ₀ is applied to the first and second address lines. As shown using the dots and crosses described above, in carrier "a", the magnetic fields generated by this current cancel each other out without affecting its magnetic state. However, in carrier "b", the magnetic fields generated by the current reinforce each other, so that carrier "b" is selected.

Similarly, the signals applied to the two address lines
Current Ii = -I₀And Ij = + I ₀Also select carrier “b”
Select. Thus, two of the same strength but of opposite polarity
Currents carry the carrier "b" when storing and reading information.
select.

In order to store information, the sum of the two current strengths should be large enough to magnetize one carrier between the magnetic states B ₀ and -B ₀ . Furthermore, the strength of the two currents combined should be small enough that a single current cannot change the magnetic state of one carrier. In a memory array, it is important to ensure that only one carrier is selected by the signal on one address line.

For reading out the information, the magnitude of the two currents summed should be small enough that the induced magnetic field is not strong enough to change the magnetization state of the carrier. However,
The total strength is such that it disrupts the magnetic state of the carrier, which produces a piezoelectric voltage across the storage medium. As mentioned above, this direction of the piezoelectric voltage represents the binary data stored in the carrier.

FIG. 7 (a) shows the process of writing a binary "1" into carrier "a" using synchronous current pulses on two address lines as shown. First, all cells of the array are assumed to be in the corresponding "0" state to the induced magnetism of -B _0. Two synchronous current pulses Ii = + 20 μm for writing a binary “1”
A and Ij = + 20 μA are applied to the two address lines, respectively. As a result, a magnetic field H for magnetizing the FeCr layer of the composition material is generated. Induction magnetic B of these layers
a is shown in FIG. 7A as a closed loop with an arrow. For the structure of FIG. 5 having the dimensions described above, the critical magnetic field strength H required for conversion between two discrete states
The critical current Ic of the intensity required to produce c is approximately 3
5 μA. In a cell where two pulses are synchronized, two +20 μA currents create a magnetic field H that will be generated by applying 40 μA. This current is larger than Ic, induced magnetism becomes B _0, As a result, binary "1" is stored. As described above, even after the pulse is closed, the induction of cell Ba magnetic remains equal to B _0, As a result, binary "1" is retained in carrier "a".

As shown in FIG. 7B, in order to store the binary number "0" in the carrier "a", two synchronous current pulses Ii = -20 .mu.A and Ij = + 20 .mu.A correspond to two address lines. Respectively. Since the sum of these currents is −40 μA, which is smaller than −Ic,
The current pulses will change the magnetic state from + _{B 0} to -B _0.

The conversion between + B ₀ and −B ₀ produces a piezoelectric voltage pulse between the first and second electrodes with a delay of Δt from the application of this current pulse. The piezoelectric pulse is + B
₀ to -B ₀ is positive and -B ₀ to + B ₀
Conversion to is negative.

If the magnetic state does not change, no piezoelectric pulse occurs. Therefore, the generated piezoelectric voltage pulse is used to prove that 1-bit binary data is stored.

To read the information stored in the carrier "a" of the memory, two synchronous current pulses Ii = -15 μm
A and Ij = −15 μA are applied to the address line. Since the critical current is Ic = −35 μA, the magnetic state cannot be changed from + B ₀ to −B ₀ with a total of −30 μA of these currents. However, this current
It does not, but is sufficient to disrupt the magnetic state. As shown in FIG. 8 (a), assuming that a binary "1" is stored in the carrier, the applied current pulse is Ba
From the value corresponding to a point such as “a” on the magnetization curve to the value corresponding to the point “b” on the curve. Due to the piezoelectric properties of the cell described above, this change in induced magnetism produces a positive piezoelectric voltage of about +15 μV, indicating that a “1” is stored in the carrier.

If a "0" is stored in the carrier "a", the applied current pulse moves Ba to the point "c" on the curve.
From the value corresponding to the point “d” to the value corresponding to the point “d”. However, in this case, the induced magnetism of carrier "a" remains Ba = -B _0, the result, the piezoelectric voltage is not at all occur, "0" indicates that it is stored. The information dumped in this device does not change during the read process because I is less than Ic.

This data reading process is shown in FIG.
It is illustrated in the time domain of (b). The delay between the piezoelectric voltage pulse and the synchronization current pulse is approximately 0.75 ns.

The storage and readout of data for carrier "b" are similar. As shown in FIG. 9A, two synchronous current pulses Ii = −20 μA and Ij = + 20 μA
A is applied to the address line, storing a "1" in the carrier "b". As mentioned above, this current pulse does not affect the carrier "a". A magnetic field is induced at the point where the pulse is synchronized, which is equivalent to the magnetic field induced by a current of 40 μA. Since this is greater than Ic = 35 μA, “1” is stored on the carrier “b”. FIG.
As shown in (b), the current pulse Ii = + 20
μA and Ij = −20 μA are applied to store “0” in carrier “b”.

The conversion from “1” to “0” on the carrier “b” generates a negative piezoelectric voltage pulse between the electrodes with a time delay of Δt, and the conversion from “0” to “1” causes a positive piezoelectric voltage pulse. Of the piezoelectric voltage pulse occurs. If the state does not change, no piezoelectric voltage is generated.

The data stored in carrier "b" is read out in a similar manner as described in connection with carrier "a".
As shown in FIG. 10, to read the data stored in the carrier "b", two synchronous current pulses Ii = + 1
5 μA and Ij = −15 μA are applied. The sum of these pulses is not large enough to change the magnetic state of carrier "b". If "0" is stored, the current pulse does not change the induced magnetism Bb of the carrier "b" and no piezoelectric voltage is generated between the electrodes. If “1” is stored,
The applied current pulse disturbs the magnetic state Bb = B ₀ , but does not change it, producing a positive piezoelectric voltage pulse with a delay of Δt. Thus, on the carrier "b", no piezoelectric voltage pulse indicates that a "0" is stored, but a positive piezoelectric voltage pulse indicates that a "1" is stored.

FIG. 11 summarizes the above-described method for storing and reading data on the carriers "a" and "b" of the memory device.

Other methods for storing and reading information from the memory device of the present invention may be utilized. For example, two synchronization currents Ii = + 15 μA and Ij = + 15 μA can be used to read information from carrier “a”. Similarly, two synchronous currents Ii = −15 μA and Ij = + 15
μA can be used to read information from carrier “b”. For example, two synchronous current pulses Ii = + 20 μA and Ij = + 20 μA for writing “1” to the carrier “a” are applied, and the piezoelectric voltage generated in response to these pulses recognizes the immediately preceding stored data, and Data is read from “a” in an erasable manner.

One of the advantages of the memory device of the present invention is that it consumes less power than a conventional nonvolatile magnetic memory device. That is, since the storage medium used in this device is highly sensitive to the magnetic field generated by the drive current, conversion between "0" and "1" can be performed with a relatively small drive current of about 20 μA in each line. Done quickly. As a result, power consumption for storing and reading data is small. In one embodiment, the read is approximately 3.
4 × 10 ⁻¹⁰ w, almost 6 × for storage in 1-bit device
It consumes 10 ⁻¹⁰ w.

The reading of information as a piezoelectric voltage generated between the sensing electrodes is essentially faster than the occurrence induced piezoelectric voltage as in a conventional magnetic memory device.

Typically, the delay between the current pulse and the corresponding piezoelectric voltage is on the order of nanoseconds or less. The conversion between "1" and "0" is typically a few nanoseconds.

Thus, a randomly accessible, non-volatile, statically operating memory device has been described. This memory device provides a device that operates at high speed and consumes low power, and can store information at high density.

It is understood that the appended claims cover all equivalent structures and methods. The present invention is not limited to the above disclosed examples, but is defined only by the appended claims.

[0085]

As described above, according to the present invention, it is possible to provide a nonvolatile, high-speed, randomly accessible, static, and updatable storage system and memory material.

[Brief description of the drawings]

FIG. 1 is a cross-sectional view of one preferred embodiment of a composition material according to the present invention.

FIG. 2 is a diagram showing a magnetization curve (BH loop) of a normal ferromagnetic material.

FIG. 3 shows a substantially square BH of the composition material according to the invention.
It is a figure showing a loop.

FIGS. 4A to 4J are views showing a process in which a piezoelectric voltage is generated in the composition material.

5 (a) and 5 (b) are a cross-sectional view and a top view of a preferred embodiment of the memory device according to the present invention.

6 (a) and 6 (b) show a process for selecting a carrier in a memory device.

FIGS. 7A and 7B are diagrams showing storage of a first information bit in a memory device. FIG.

FIGS. 8 (a) and (b) are diagrams illustrating a process of reading stored first information bits.

FIGS. 9A and 9B are diagrams showing storage of a second information bit in a memory device.

FIG. 10 shows a current pulse used to read a second information bit stored in a memory device and a corresponding output.

FIG. 11 is a list of preferred methods for storing and reading information from memory.

FIGS. 12 (a) and 12 (b) are diagrams showing a current with respect to a process time in an electrolytic process used in a process for manufacturing the composition material.

[Explanation of symbols]

Reference _Signs List 100 base 110 Pb _0.80 Cd _0.10 Si _0.10 layer 120 Se _0.90 S _0.10 layer 130 Fe _0.76 Cr _0.24 layer 190 Piezoelectric structure 200 First FeCr layer 210 First SeS Layer 220 First PbCdSi layer 230 Second PbCdSi layer 240 Second SeS layer 250 Second FeCr layer 260 Wire 265 Vertical axis 270 Arrow indicating direction of magnetic dipole 275,280 Dipole array 290 Memory device 300 electrode 310 first set of material layer 320 first address line 330 silicon planar substrate 340 second address line 350 second set of material layer 360 electrode

────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] June 25, 2001 (2001.6.2)
5)

[Procedure amendment 2]

[Document name to be amended] Drawing

[Correction target item name] All figures

[Correction method] Change

[Correction contents]

FIG.

FIG. 2

FIG. 3

FIG. 5

FIG. 4

FIG. 6

FIG. 7

FIG. 9

FIG. 10

FIG. 8

FIG. 11

FIG.

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5E049 AA00 BA12 CB01 GC01 5F083 FZ10 JA60 LA12 LA16

Claims (2)

[Claims] The composition is M _(1-xy) Cd _x R _y , where M is one element selected from the group consisting of Zn and Te, and R consists of Si and Ge. A first material layer selected from the group, wherein x and y each take a value in the range of 0.09 ≦ x ≦ 0.11, 0.09 ≦ y ≦ 0.11; A second layer formed on the layer and having a value of z in a range of 0.09 ≦ z ≦ 0.11.
Of _{Se (1-z) S} z layer; formed on the second layer,
The third F in which w takes a value in the range of 0.18 ≦ w ≦ 0.36
e _(1-w) Cr _w layer; a composition material for memory. 2. The composition having a composition of M_(1-xy)Cd_xR_ySo
Where M is selected from the group consisting of Zn and Te.
One selected element, where R is composed of Si and Ge
X and y are one element selected from the group
0.09 ≦ x ≦ 0.11, 0.09 ≦ y ≦ 0.11
A first material layer taking the value of the enclosure; on said first layer, z is
Second Se having a value in the range of 0.09 ≦ z ≦ 0.11
_(1-z)S_zForming a layer; on said second layer, w is
Third Fe having a value in the range of 0.18 ≦ w ≦ 0.36
_(1-w)Cr_wFerromagnetism consisting of the steps of forming a layer
, Electro-optic and piezoelectric composition material for memory
How to make the ingredients.