Great research starts with great data.

Learn More
More >
Patent Analysis of

Memory cell structure of magnetic memory with spin device elements

Updated Time 12 June 2019

Patent Registration Data

Publication Number

US10002655

Application Number

US15/377111

Application Date

13 December 2016

Publication Date

19 June 2018

Current Assignee

BLUESPIN, INC.

Original Assignee (Applicant)

BLUESPIN, INC.

International Classification

G11C11/00,H01L43/08,H01L43/02,G11C11/16,H01L27/22

Cooperative Classification

G11C11/1673,G11C11/161,G11C11/1675,H01L27/222,H01L43/02

Inventor

FUKUZAWA, HIDEAKI

Patent Images

This patent contains figures and images illustrating the invention and its embodiment.

US10002655 Memory cell structure magnetic 1 US10002655 Memory cell structure magnetic 2 US10002655 Memory cell structure magnetic 3
See all images <>

Abstract

A magnetic memory includes a plurality of memory cells and a data identification circuit. Each of the memory cells includes: a first bias node to which a first voltage is applied in data reading, the first voltage being a positive voltage; a second bias node to which a second voltage is applied in the data reading, the second voltage being a negative voltage having substantially the same absolute value as the first voltage; a connection node; a first spin device element connected between the first bias node and the connection node; and a second spin device element connected between the connection node and the second bias node. The first and second spin device elements operate differentially. The data identification circuit identifies data stored in each of the memory cells based on a polarity of a voltage generated on the connection node.

Read more

Claims

1. A magnetic memory, comprising: a memory cell configured to store a one-bit data; and a data identification circuit, wherein the memory cell includes: a first bias node; a second bias node; a connection node; a first spin device element connected between the first bias node and the connection node; and a second spin device element connected between the connection node and the second bias node,wherein the first spin device element includes: a first recording layer having a first reversible magnetization; a first reference layer having a first fixed magnetization; and a first spacer layer provided between the first recording layer and the first reference layer,wherein the second spin device element includes: a second recording layer having a second reversible magnetization; a second reference layer having a second fixed magnetization; and a second spacer layer provided between the second recording layer and the second reference layer, wherein the first recording layer and the second recording layer are connected to each other via the connection node, wherein the first reference layer is connected to the first bias node, wherein the second reference layer is connected to the second bias node, wherein the first fixed magnetization of the first reference layer and the second fixed magnetization of the second reference layer are directed in a same direction, wherein the magnetizations of the first recording layer and the second recording layer are directed in opposite directions, wherein, in data reading, a bias voltage is applied between the first bias node and the second bias node across the first and second spin device elements to allow a read current to flow between the first bias node and the second bias node through the first and second spin device elements, and wherein the data identification circuit is configured to identify the one-bit data stored in the memory cell based on a voltage generated on the connection node.

2. The magnetic memory according to claim 1, further comprising: a write circuit configured to supply a write current between the first bias node and the second bias node in writing a one-bit write data into the memory cell, wherein the write current flows through the first and second spin device elements.

3. The magnetic memory according to claim 2, wherein a direction of the write current is selected depending on the one-bit write data.

4. The magnetic memory according to claim 1, wherein the connection node includes a connection electrode, and wherein the first recording layer and the second recording layer are connected to a same surface of the connection electrode.

5. The magnetic memory according to claim 1, wherein the first bias node includes a first electrode, wherein the second bias node includes a second electrode, wherein the connection node includes a connection electrode, wherein the first reference layer is formed on an upper surface of the first electrode, wherein the second reference layer is formed on an upper surface of the second electrode, wherein the first recording layer and the second recording layer are connected to a lower surface of the connection electrode.

6. A magnetic memory, comprising: a memory cell configured to store a one-bit data; and a data identification circuit, wherein the memory cell includes: a first bias node; a second bias node; a connection node; a first spin device element connected between the first bias node and the connection node; and a second spin device element connected between the connection node and the second bias node,wherein the first spin device element includes: a first recording layer having a first reversible magnetization; a first reference layer having a first fixed magnetization; and a first spacer layer provided between the first recording layer and the first reference layer,wherein the second spin device element includes: a second recording layer having a second reversible magnetization; a second reference layer having a second fixed magnetization; and a second spacer layer provided between the second recording layer and the second reference layer, wherein the first recording layer is connected to the first bias node, wherein the second recording layer is connected to the second bias node, wherein the first reference layer and the second reference layer are connected to each other via the connection node, wherein the first fixed magnetization of the first reference layer and the second fixed magnetization of the second reference layer are directed in a same direction, wherein the magnetizations of the first recording layer and the second recording layer are directed in opposite directions, wherein, in data reading, a bias voltage is applied between the first bias node and the second bias node across the first and second spin device elements to allow a read current to flow between the first bias node and the second bias node through the first and second spin device elements, and wherein the data identification circuit is configured to identify the one-bit data stored in the memory cell based on a voltage generated on the connection node.

7. The magnetic memory according to claim 6, further comprising: a write circuit configured to supply a write current between the first bias node and the second bias node in writing a one-bit write data into the memory cell, wherein the write current flows through the first and second spin device elements.

8. The magnetic memory according to claim 7, wherein a direction of the write current is selected depending on the one-bit write data.

9. The magnetic memory according to claim 7, wherein the connection node includes a connection electrode, and wherein the first reference layer and the second reference layer are connected to a same surface of the connection electrode.

10. The magnetic memory according to claim 6, wherein the first bias node includes a first electrode, wherein the second bias node includes a second electrode, wherein the connection node includes a connection electrode, wherein the first and second reference layers are formed on an upper surface of the connection electrode, wherein the first electrode is formed on the upper surface of the first recording layer, and wherein the second electrode is formed on the upper surface of the second recording layer.

Read more

Claim Tree

  • 1
    1. A magnetic memory, comprising:
    • a memory cell configured to store a one-bit data
    • and a data identification circuit, wherein the memory cell includes: a first bias node
    • a second bias node
    • a connection node
    • a first spin device element connected between the first bias node and the connection node
    • and a second spin device element connected between the connection node and the second bias node,wherein the first spin device element includes: a first recording layer having a first reversible magnetization
    • a first reference layer having a first fixed magnetization
    • and a first spacer layer provided between the first recording layer and the first reference layer,wherein the second spin device element includes: a second recording layer having a second reversible magnetization
    • a second reference layer having a second fixed magnetization
    • and a second spacer layer provided between the second recording layer and the second reference layer, wherein the first recording layer and the second recording layer are connected to each other via the connection node, wherein the first reference layer is connected to the first bias node, wherein the second reference layer is connected to the second bias node, wherein the first fixed magnetization of the first reference layer and the second fixed magnetization of the second reference layer are directed in a same direction, wherein the magnetizations of the first recording layer and the second recording layer are directed in opposite directions, wherein, in data reading, a bias voltage is applied between the first bias node and the second bias node across the first and second spin device elements to allow a read current to flow between the first bias node and the second bias node through the first and second spin device elements, and wherein the data identification circuit is configured to identify the one-bit data stored in the memory cell based on a voltage generated on the connection node.
    • 2. The magnetic memory according to claim 1, further comprising:
      • a write circuit configured to supply a write current between the first bias node and the second bias node in writing a one-bit write data into the memory cell, wherein the write current flows through the first and second spin device elements.
    • 4. The magnetic memory according to claim 1, wherein
      • the connection node includes a connection electrode, and wherein
    • 5. The magnetic memory according to claim 1, wherein
      • the first bias node includes a first electrode, wherein
  • 6
    6. A magnetic memory, comprising:
    • a memory cell configured to store a one-bit data
    • and a data identification circuit, wherein the memory cell includes: a first bias node
    • a second bias node
    • a connection node
    • a first spin device element connected between the first bias node and the connection node
    • and a second spin device element connected between the connection node and the second bias node,wherein the first spin device element includes: a first recording layer having a first reversible magnetization
    • a first reference layer having a first fixed magnetization
    • and a first spacer layer provided between the first recording layer and the first reference layer,wherein the second spin device element includes: a second recording layer having a second reversible magnetization
    • a second reference layer having a second fixed magnetization
    • and a second spacer layer provided between the second recording layer and the second reference layer, wherein the first recording layer is connected to the first bias node, wherein the second recording layer is connected to the second bias node, wherein the first reference layer and the second reference layer are connected to each other via the connection node, wherein the first fixed magnetization of the first reference layer and the second fixed magnetization of the second reference layer are directed in a same direction, wherein the magnetizations of the first recording layer and the second recording layer are directed in opposite directions, wherein, in data reading, a bias voltage is applied between the first bias node and the second bias node across the first and second spin device elements to allow a read current to flow between the first bias node and the second bias node through the first and second spin device elements, and wherein the data identification circuit is configured to identify the one-bit data stored in the memory cell based on a voltage generated on the connection node.
    • 7. The magnetic memory according to claim 6, further comprising:
      • a write circuit configured to supply a write current between the first bias node and the second bias node in writing a one-bit write data into the memory cell, wherein the write current flows through the first and second spin device elements.
    • 10. The magnetic memory according to claim 6, wherein
      • the first bias node includes a first electrode, wherein
See all independent claims <>

Description

CROSS REFERENCE

This application claims priority of Japanese Patent Application No. 2015-075407, filed on Apr. 1, 2015, the disclosure which is incorporated herein by reference.

TECHNICAL FIELD

The present invention relates to a magnetic memory and a method of operating the same.

BACKGROUND ART

Magnetic memories are devices storing a data as the direction of a spontaneous magnetization (hereinafter, simply referred to as “magnetization”) of a magnetic layer. Extensive research and development activities have been conducted for magnetic memories, since they are expected as non-volatile memories with high speed operation, large capacity and reduced power consumption. Most typically, a magnetic memory is configured to achieve data reading by using a magnetoresistance effect, such as a tunnel magnetoresistance effect (TMR effect) and a giant magnetoresistance effect (GMR). An element including two magnetic layers and a spacer layer disposed therebetween (hereinafter, referred to as “spin device element”) exhibits a resistance depending on the relative direction of the magnetizations of the two magnetic layers due to the magnetoresistance effect. Most typically, a spin device element is placed into the “low resistance” state when the magnetizations of the two magnetic layers are directed in “parallel” and placed into the “high resistance” state when the magnetizations of the two magnetic layers are directed in “antiparallel”. In a magnetic memory which incorporates a spin device element in each memory cell, data stored in each memory cell can be identified from the signal level of a voltage or current signal generated so as to depend on the resistance value of the spin device element.

FIG. 1 is a conceptual diagram illustrating an exemplary read operation of a magnetic memory. Discussed below is the case in which a spin device element is used as a memory cell 101 and the memory cell 101 can take two states: the “high resistance” state of a resistance value RHigh and the “low resistance” state of a resistance value RLow, where RHigh>RLow.

Most typically, a read current Isense is fed to the memory cell 101 when data reading from the memory cell 101 is performed. When the read current Isense flows through the memory cell 101, a read voltage Vmemory is generated across the memory cell 101. The read voltage Vmemory, which depends on the state of the memory cell 101, that is, the resistance value of the memory cell 101, can be used as a read signal obtained from the memory cell 101. It is possible to identify the data stored in the memory cell 101 by comparing the read voltage Vmemory generated across the memory cell 101 with a predetermined reference voltage Vref, for example, with a sense amplifier 102 and providing a value as the “OUTPUT” shown in FIG. 1. In detail, the data stored in the memory cell 101 can be identified by comparing the read voltage Vmemory with a reference voltage Vref adjusted between voltages VHigh and VLow where VHigh (=RHigh·Isense) is an expected value of the read voltage Vmemory for the high resistance state and VLow (=RLow·Isense) is an expected value of the read voltage Vmemory for the low resistance state. In the case when the “high resistance” state is associated with data “0” and the “low resistance” state is associated with data “1”, for example, the data stored in the memory cell 101 can be identified as data “0” if the read voltage Vmemory is higher than the reference voltage Vref and as data “1” if the read voltage Vmemory is lower than the reference voltage Vref.

In the above-described read operation, the effective signal window for data identification based on the read signal obtained from the memory cell 101 is the difference ΔV between the voltages VHigh and VLow. The data stored in the memory cell 101 can be identified more surely, as the signal window ΔV is increased.

Although the read voltage Vmemory dependent on the data stored in the memory cell 101 is obtained by feeding a given read current Isense through the memory cell 101 in the above-described operation, a read current dependent on the data stored in the memory cell 101 may be obtained in a read operation by applying a given read voltage across the memory cell 101. In this case, the effective signal window of the read signal obtained from a memory cell 101 is the difference between a current flowing through the memory cell 101 placed into the “low resistance” state and the current flowing through the memory cell 101 placed into the “high resistance” state.

One current issue of the magnetic memory is the insufficiency in the effective signal window of the read signal obtained from the memory cell. Read signals obtained from memory cells have a distribution due to variations in the resistance value of the spin device elements (that is, the memory cells) and variations in the interconnection resistance of long interconnections. In the meantime, the signal level of a reference signal (the reference voltages Vref obtained in the operation illustrated in FIG. 1) also have a distribution due to variations in the circuitry which generates the reference signal. FIG. 2 illustrates one example of the distributions in the signal levels of the read signals and the reference signal. FIG. 2 illustrates the distributions of the reference voltage Vref, the read voltage VHigh obtained as the read voltage Vmemory when the memory cell 101 is placed in the high-resistance state, and the read voltage VLow obtained as the read voltage Vmemory when the memory cell 101 is placed in the low-resistance state. The horizontal axis of the graph illustrated in FIG. 2 corresponds to the voltage V and the horizontal axis corresponds to the frequency N. As is understood from FIG. 2, an insufficient effective signal window of the read signal undesirably causes read errors due to the overlap of the distributions of the read voltage VLow and the reference voltage Vref and the overlap of the distributions of the read voltage VHigh and the reference voltage Vref.

It is preferable to sufficiently increase the MR (magnetoresistance) ratio of a spin device element incorporated in a memory cell for achieving a sufficient signal window, because the effective signal window of the read signal obtained from the memory cell depends on the MR ratio of the spin device element. The current technology, however, does not offer an MR ratio for obtaining a sufficient signal window. It is technologically difficult to achieve an increase in the MR ratio, because it requires a remarkable breakthrough in materials used in spin device elements.

Related art documents are listed in the following. U.S. Patent Application Publication No. 2013/0121066 A1 discloses the structure of a memory cell which includes four spin device elements. U.S. Pat. No. 6,424,562 B1 discloses read/write architecture for a magnetoresistive random access memory (MRAM). U.S. Patent Application Publication No. 2013/0272059 A1 discloses a differential MRAM structure.

SUMMARY OF INVENTION

Therefore, an objective of the present invention is to increase an effective signal window of a read signal obtained from a memory cell of a magnetic memory. Other objectives and new features of the present invention would be understood by a person skilled in the art from the attached drawings and the following disclosure.

In an aspect of the present invention, a magnetic memory includes a plurality of memory cells and a data identification circuit. Each of the memory cells includes: a first bias node to which a first voltage is applied in data reading, the first voltage being a positive voltage; a second bias node to which a second voltage is applied in the data reading, the second voltage being a negative voltage having substantially the same absolute value as the first voltage; a connection node; a first spin device element connected between the first bias node and the connection node; and a second spin device element connected between the connection node and the second bias node. Each of the first and second spin device elements is configured to have a first magnetization which is reversible and to take a first or second state depending on a direction of the first magnetization. The resistance of each of the first and second spin device elements in a case when each of the first and second spin device elements is placed in the first state is higher than the resistance of each of the first and second spin device elements in a case when each of the first and second spin device elements is placed in the second state. The second spin device element is placed in the second state when the first spin device element is placed in the first state, and placed in the first state when the first spin device element is placed in the second state. The data identification circuit identifies data stored in each of the memory cells based on a polarity of a voltage generated on the connection node.

In another aspect of the present invention, a magnetic memory includes a memory cell and a data identification circuit. The memory cell includes: a first bias node to which a first voltage is applied in data reading from the memory cell; a second bias node to which a second voltage lower than the first voltage is applied in the data reading; a first connection node; a second connection node; a first spin device element connected between the first bias node and the first connection node; a second spin device element connected between the first connection node and the second bias node; a third spin device element connected between the first bias node and the second connection node; and a fourth spin device element connected between the second connection node and the second bias node. Each of the first to fourth spin device elements is configured to have a first magnetization which is reversible and to take selected one of first and second states depending on a direction of the first magnetization. The resistance of each of the first to fourth spin device elements in a case when each of the first to fourth spin device elements is placed in the first state is higher than the resistance of each of the first to fourth spin device elements in a case when each of the first to fourth spin device elements is placed in the second state. The first and fourth spin device elements have the same state selected from the first and second states and the second and third spin device elements have the same state selected from the first and second states. The second and third spin device elements are placed in the second state when the first and fourth spin device elements are placed in the first state, and placed in the first state when the first and fourth spin device elements are placed in the second state. The data identification circuit identifies data stored in the memory cell based on a third voltage generated on the first connection node and a fourth voltage generated on the second connection node.

In still another aspect of the present invention, a magnetic memory includes a memory cell and a data identification circuit. The memory cell includes: a first bias node to which a first voltage is applied in data reading from the memory cell; a second bias node to which a second voltage lower than the first voltage is applied in the data reading; a first connection node; a second connection node; a first spin device element connected between the first bias node and the first connection node; a first resistor element connected between the first connection node and the second bias node; a second resistor element connected between the first bias node and the second connection node; and a second spin device element connected between the second connection node and the second bias node. Each of the first and second spin device elements is configured to have a first magnetization which is reversible and to take selected one of first and second states depending on a direction of the first magnetization. The resistance of each of the first and second spin device elements in a case when each of the first and second spin device elements is placed in the first state is higher than the resistance of each of the first and second spin device elements in a case when each of the first and second spin device elements is placed in the second state. The first and fourth spin device elements have the same state selected from the first and second states. The data identification circuit identifies data stored in each of the memory cells based on a third voltage generated on the first connection node and a fourth voltage generated on the second connection node.

In still another aspect of the present invention, a method of operating a magnetic memory is provided, which includes a plurality of memory cells each comprising first and second bias nodes, a connection node, a first spin device element connected between the first bias node and the connection node, and a second spin device element connected between the connection node and the second bias node. Each of the first and second spin device elements is configured to have a first magnetization which is reversible and to take selected one of first and second states depending on a direction of the first magnetization. The resistance of each of the first and second spin device elements in a case when each of the first and second spin device elements is placed in the first state is higher than the resistance of each of the first and second spin device elements in a case when each of the first and second spin device elements is placed in the second state. The second spin device element is placed in the second state when the first spin device element is placed in the first state, and placed in the first state when the first spin device element is placed in the second state. The method includes:

applying a first voltage to the first bias node, the first voltage being a positive voltage;

applying a second voltage to the second bias node, the second voltage being a negative voltage having substantially the same absolute value as the first voltage; and

identifying data stored in each of the memory cells based on a polarity of a voltage generated on the connection node.

In still another aspect of the present invention, a method of operating a magnetic memory is provided, which includes a memory cell comprising: first and second bias nodes; first and second connection nodes; a first spin device element connected between the first bias node and the first connection node; a second spin device element connected between the first connection node and the second bias node; a third spin device element connected between the first bias node and the second connection node; and a fourth spin device element connected between the second connection node and the second bias node. Each of the first to fourth spin device elements is configured to have a first magnetization which is reversible and to take selected one of first and second states depending on a direction of the first magnetization. The resistance of each of the first to fourth spin device elements in a case when each of the first to fourth spin device elements is placed in the first state is higher than the resistance of each of the first to fourth spin device elements in a case when each of the first to fourth spin device elements is placed in the second state. The first and fourth spin device elements have the same state selected from the first and second states and the second and third spin device elements have the same state selected from the first and second states. The second and third spin device elements are placed in the second state when the first and fourth spin device elements are placed in the first state, and placed in the first state when the first and fourth spin device elements are placed in the second state. The method includes:

applying a first voltage to the first bias node;

applying a second voltage lower than the first voltage to the second bias node; and

identifying data stored in the memory cell based on a third voltage generated on the first connection node and a fourth voltage generated on the second connection node.

The present invention effectively increases the effective signal window of a read signal obtained from a memory cell of a magnetic memory.

BRIEF DESCRIPTION OF THE DRAWINGS

The above and other advantages and features of the present invention will be more apparent from the following description taken in conjunction with the accompanied drawings, in which:

FIG. 1 is a conceptual diagram illustrating an exemplary read operation of a magnetic memory;

FIG. 2 is a graph illustrating one example of the distributions in the signal levels of read signals and a reference signal;

FIG. 3 illustrates an exemplary configuration of a magnetic memory in one embodiment of the present invention;

FIG. 4A is a circuit diagram illustrating a schematic configuration of each memory cell in the present embodiment;

FIG. 4B is a sectional view illustrating an exemplary structure of each spin device element in the present embodiment;

FIG. 5A is a conceptual diagram illustrating two allowed states of the memory cell configured as illustrated in FIG. 4A;

FIG. 5B is a graph illustrating one example of the distribution of read voltages obtained from memory cells configured as illustrated in FIG. 4A and the effective signal window of the read voltages;

FIG. 6 is a circuit diagram illustrating the schematic structure of each memory cell of a magnetic memory in another embodiment of the present invention;

FIG. 7 a conceptual diagram illustrating two allowed states of the memory cell illustrated in FIG. 6;

FIG. 8 is a graph illustrating one example of the distribution of read voltages obtained from memory cells illustrated in FIG. 6 and the effective signal window of the read voltages;

FIG. 9A a circuit diagram illustrates a modification of the memory cell in the present embodiment;

FIG. 9B a circuit diagram illustrating another modification of the memory cell in the present embodiment;

FIG. 10 is a conceptual diagram illustrating two allowed states of the memory cell illustrated in FIG. 9A;

FIG. 11A is a conceptual diagram illustrating one example of the configuration of each memory cell when the magnetic memory of the present embodiment is configured as an STT-MRAM (Spin Transfer Torque Magnetoresistive Random Access Memory);

FIG. 11B is a conceptual diagram illustrating another example of the configuration of each memory cell when the magnetic memory of the present embodiment is configured as an STT-MRAM;

FIGS. 12A and 12B illustrate data writing into the memory cell illustrated in FIG. 11A;

FIG. 13 is a diagram illustrating data reading from the memory cell illustrated in FIG. 11A;

FIGS. 14A to 14F are conceptual diagrams illustrating other examples of the configuration of each memory cell when the magnetic memory of the present embodiment is configured as an STT-MRAM;

FIGS. 15A and 15B are diagrams illustrating data writing into the memory cell illustrated in FIG. 14A;

FIGS. 16A and 16B are diagrams illustrating data reading from the memory cell illustrated in FIG. 14A;

FIG. 17A is a plan view illustrating one example of the configuration of a memory cell in the case when a magnetic memory is configured to achieve data writing with a current-induced magnetic field;

FIG. 17B is a side view illustrating the configuration of the memory cell illustrated in FIG. 17A;

FIG. 17C is a cross-sectional view illustrating the structure of the memory cell on section A-A indicated in FIG. 17A;

FIG. 17D is a cross-sectional view illustrating the structure of the memory cell on section B-B indicated in FIG. 17A;

FIGS. 18A and 18B illustrate data writing into the memory cell illustrated in FIG. 17A;

FIG. 19 is a conceptual diagram illustrating electrical connections among spin device elements in a memory cell and data reading from the memory cell, in the case when the magnetic memory is configured to achieve data writing with a current-induced magnetic field;

FIG. 20 is a plan view illustrating an example of the configuration of a memory cell array of a magnetic memory in which spin device elements are integrated on a deformable base plate in one embodiment;

FIG. 21 is a plan view illustrating the structure of each block of the memory cell array illustrated in FIG. 20;

FIG. 22A is a cross-sectional view illustrating the structure of each block on section C-C indicated in FIG. 21;

FIG. 22B is a cross-sectional view illustrating the structure of each block on section D-D indicated in FIG. 21;

FIG. 23 is a perspective view illustrating the structure of each block;

FIG. 24 is a cross-sectional view illustrating the structure of the memory cell array on section E-E indicated in FIG. 20;

FIGS. 25A to 25D are cross-sectional views illustrating an exemplary data write procedure into a memory cell of the memory cell array illustrated in FIGS. 20 to 24;

FIGS. 26A and 26B are cross-sectional views illustrating another example of the structure of the memory cell array of the magnetic memory in which spin device elements are integrated on a deformable base plate;

FIG. 27 is a plan view illustrating still another example of the configuration of the memory cell array of the magnetic memory in which spin device elements are integrated on a deformable base plate;

FIG. 28 is a cross-sectional view illustrating the structure of the memory cell array on section F-F indicated in FIG. 27;

FIG. 29 is a perspective view illustrating an exemplary configuration of each block of the memory cell array in the case when spin device elements are provided on a deformable base plate and data writing into a memory cell is achieved with a current-induced magnetic field; and

FIG. 30 is a plan view illustrating the structure of the block illustrated in FIG. 29.

DESCRIPTION OF PREFERRED EMBODIMENTS

The invention will be now described herein with reference to illustrative embodiments. Those skilled in the art would appreciate that many alternative embodiments can be accomplished using the teachings of the present invention and that the invention is not limited to the embodiments illustrated for explanatory purposed. It would be appreciated that the same or similar components may be denoted by the same or corresponding reference numerals in the description of the preferred embodiments. It would be also appreciated that for simplicity and clarity of illustration, elements in the Figures have not necessary drawn to scale. For example, the dimensions of some of the elements are exaggerated relative to other elements.

(Configuration and Operation of Magnetic Memory)

FIG. 3 illustrates an exemplary configuration of a magnetic memory 10 in one embodiment of the present invention. The magnetic memory 10 includes a plurality of memory cells 1 arrayed in row and columns, a row selection circuit 3, a column selection circuit 4, a sense amplifier circuit 5 and a write circuit 6. The row selection circuit 3 selects a desired row of the memory cells 1 and the column selection circuit 4 selects a desired column of the memory cells 1. In read and write operations, a memory cell 1 to be accessed is selected by the row selection circuit 3 and the column selection circuit 4. The sense amplifier circuit 5 includes sense amplifiers 12. In read operation, data stored in a selected memory cell 1 is identified by the corresponding sense amplifier 12. The write circuit 6 feeds a write current to a selected memory cell 1 in a write operation.

FIG. 4A is a circuit diagram illustrating a schematic configuration of each memory cell 1 of the magnetic memory 10. In one embodiment, each memory cell 1 includes two spin device elements 11A and 11B. It should be noted that the spin device elements 11A and 11B may be collectively referred to as the spin device elements 11 if they are not distinguished from each other.

FIG. 4B is a sectional view illustrating an exemplary structure of each spin device element 11. In the present embodiment, each spin device element 11 includes a reference layer 21, a spacer layer 22 and a recording layer 23. The reference layer 21 and the recording layer 23 are opposed to each other across the spacer layer 22. In the structure illustrated in FIG. 4B, the spacer layer 22 is coupled with the upper surface of the reference layer 21 and the recording layer 23 is coupled with the upper surface of the spacer layer 22. It should be noted that the positions of the reference layer 21 and the recording layer 23 may be exchanged.

The reference layer 21 and the recording layer 23, which are both configured to exhibit a magnetization, include at least one magnetic film. The spacer layer 22 is formed of non-magnetic material. The magnetization direction of the referent layer 21 is fixed, while the magnetization direction of the recording layer 23 is reversible. In one embodiment, the reference layer 21 and the recording layer 23 both have perpendicular magnetic anisotropy. In this case, the reference layer 21 is formed so that the magnetization thereof is fixed in a film thickness direction, and the recording layer 23 is formed so that the magnetization thereof is reversible between the film thickness directions. Illustrated in FIG. 4B is the structure in which the magnetization of the reference layer 21 is fixed in the upward direction (+Z direction) and the magnetization of the recording layer 23 is reversible between the upward and downward directions (+Z and −Z directions). It should be noted however that the reference layer 21 and the recording layer 23 may have in-plane magnetic anisotropy instead.

The reference layer 21 and the recording layer 23 may be formed of an elementary substance of magnetic metal, such as iron (Fe), cobalt (Co) and nickel (Ni), or a ferromagnetic alloy including at least one of these magnetic metals, for example. The reference layer 21 and the recording layer 23 may be formed of magnetic metal or alloy doped with one or more non-magnetic elements. Non-magnetic elements which may be contained in the reference layer 21 and the recording layer 23 include boron, carbon, nitrogen, oxygen, aluminum, silicon, titanium, vanadium, chromium, manganese, copper, zinc, zirconium, niobium, molybdenum, ruthenium, rhodium, palladium, silver, hafnium, tantalum, tungsten, iridium, platinum and gold.

In one embodiment, the spacer layer 22 may be formed of a dielectric film having such a thin thickness that a tunnel current can flow through the spacer layer 22. In this case, the spin device element 11 operates as a TMR (tunnel magnetoresistance) element that exhibits a TMR effect. To obtain a sufficiently large TMR effect, it is preferable that the spacer layer 22 is formed of, for example, magnesium oxide (MgO), aluminum oxide (AlOx) or the like. Alternatively, the spacer layer 22 may be formed of oxide, nitride or oxynitride of magnesium (Mg), aluminum (Al), silicon (Si), zirconium (Zr), hafnium (Hf), tantalum (Ta) or the like.

The spacer layer 22 may be formed of a metal conductor to reduce the resistance of the spin device element 11. In this case, the spin device element 11 is configured as a spin valve element that exhibits a giant magnetoresistance (GMR) effect. To achieve a sufficiently large GMR effect, the spacer layer 22 may be formed of non-magnetic metal, such as copper (Cu), gold (Au), silver (Ag) and aluminum (Al) or an alloy of these non-magnetic metals. A composite spacer layer that includes an oxide matrix and metal columns penetrating through the oxide matrix in the thickness direction may be used as the spacer layer 22. In this case, the oxide matrix of the composite spacer layer may be formed of aluminum oxide and the metal columns, which penetrates through the oxide matrix in the thickness direction, may be formed of copper (Cu). The spacer layer 22 preferably has a thickness of 1 to 3 nm.

Each spin device element 11 thus structured is allowed to have two states: the “low resistance” state and the “high resistance” state. In detail, a spin device element 11 is placed into the “low resistance” state when the magnetizations of the reference layer 21 and the recording layer 23 are directed in “parallel” and the spin device element 11 is placed into the “high resistance” state when the magnetizations of the reference layer 21 and the recording layer 23 are directed in “antiparallel”. This implies that the spin device element 11 functions as a variable resistor element exhibiting a resistance value depending on the relative direction of the reference layer 21 and the recording layer 23.

Although the simplest structure of the spin device element 11 is illustrated in FIG. 4B, the structure of the spin device element 11 may be variously modified. For example, the reference layer 21 may be formed of a film stack including a magnetic film and an antiferromagnetic film that fixes the magnetization of the magnetic film. The recording layer 23 may be formed of a film stack including multiple magnetic layers and one or more non-magnetic films each providing ferromagnetic coupling between adjacent two of the magnetic layers.

Referring back to FIG. 4A, the two spin devices elements 11A and 11B are connected in series between a node Nbias1 and a node Nbias2 in the memory cell 1, where the nodes Nbias1 and Nbias2 are first and second bias nodes to which bias voltages are applied in a read operation. In detail, the spin device element 11A is connected between the node Nbias1 and a node N1 and the spin device element 11B is connected between the node N1 and the node Nbias2, where the node N1 is a connection node which provides electrical connection between the spin devices elements 11A and 11B.

The memory cell 1 illustrated in FIG. 4A is configured to offer two allowed states with two spin device elements 11 and to thereby store one bit data. In the configuration illustrated in FIG. 4A, as illustrated in FIG. 5A, each memory cell 1 stores one-bit data, correlating one of the first and second states defined below with data “0” and the other with data “1”:

First state: a state in which the spin device element 11A is placed in the “low resistance” state and the spin device element 11B is placed in the “high resistance” state, and

Second state: a state in which the spin device element 11A is placed in the “high resistance” state and the spin device element 11B is placed in the “low resistance” state.

FIG. 5A illustrates two allowed states of the memory cell 1 in the case when the first state is correlated with data “0” and the second state is correlated with data “1”.

Data reading from a memory cell 1 configured as illustrated in FIG. 4A is achieved by determining the polarity of the voltage V1 generated on the node N1 with a data identification circuit, e.g. a sense amplified 12, in the state in which a positive bias voltage +Vbias is applied to the node Nbias1 and a negative bias voltage −Vbias is applied to the node Nbias2. It should be noted that the bias voltages applied to the nodes Nbias1 and Nbias2 have substantially the same absolute value (magnitude) but have opposite polarities. In an actual implementation, the absolute values of the bias voltages applied to the nodes Nbias1 and Nbias2 may slightly differ from each other due to manufacturing process variations or other reasons; however, the bias voltages applied to the nodes Nbias1 and Nbias2 are adjusted to have at least substantially the same absolute value. In the example illustrated in FIG. 5A, in which the first state is correlated with data “0” and the second state is correlated with data “1”, the data stored in the memory cell 1 is identified as data “0” when the polarity of the voltage V1 is positive and as data “1” when the polarity of the voltage V1 is negative.

The memory cell configuration illustrated in FIG. 4A effectively offers a large signal window even when the MR ratios of the spin device elements 11 are not so large, due to the differential operation using the two spin device elements 11. Accordingly, as illustrated in FIG. 5B, the reliability of data reading from the memory cell 1 can be effectively improved even when the voltage V1 on the node N1 varies on some level. Note that the signal window is denoted by symbol ΔV1 in FIG. 5B. Additionally, the memory cell configuration illustrated in FIG. 4A allows identifying the data stored in the memory cell 1 with a simple method; the data identification can be achieved by determining the polarity of the voltage V1.

FIG. 6 is a circuit diagram illustrating the schematic structure of each memory cell 1 of the magnetic memory 10 in another embodiment of the present invention. In the present embodiment, each memory cell 1 includes four spin device elements 11. It should be noted that, suffixes “A1”, “A2”, “B1” and “B2” may be used in the following, if the spin device elements 11 are distinguished from one another.

The four spin device elements 11 forms a bridge circuit in each memory cell 1 of the magnetic memory 10 in the present embodiment. More specifically, the spin device element 11A1 is connected between a node Nbias1 and a node N1 and the spin device element 11B1 is connected between the node N1 and a node Nbias2, where the node Nbias1 is a first bias node to which a first voltage is applied in a read operation and the node Nbias2 is a second bias node to which a second voltage lower than the first voltage is applied in the read operation. In one embodiment, in the read operation, a positive bias voltage +Vbias is applied to the node Nbias1 while the node Nbias2 is grounded. The node N1 is a first connection node which provides an electrical connection between the spin device elements 11A1 and 11B1. Furthermore, the spin device element 11B2 is connected between the node Nbias1 and a node N2 and the spin device element 11A2 is connected between the node N2 and the node Nbias2, where the node N2 is a second connection node which provides an electrical connection between the spin device elements 11B2 and 11A2.

The memory cell 1 illustrated in FIG. 6 is configured to offer two allowed states with four spin device elements 11 and to thereby store one bit data. FIG. 7 illustrates the two allowed states of the memory cell 1 in the present embodiment. In the present embodiment, each memory cell 1 stores one-bit data, correlating one of the first and second states defined below with data “0” and the other with data “1”:

First state: a state in which the spin device elements 11A1 and 11A2 are placed in the “low resistance” state and the spin device elements 11B1 and 11B2 are placed in the “high resistance” state, and

Second state: a state in which the spin device elements 11A1 and 11A2 are placed in the “high resistance” state and the spin device elements 11B1 and 11B2 are placed in the “low resistance” state. FIG. 7 illustrates two allowed states of the memory cell 1 in the case when the first state is correlated with data “0” and the second state is correlated with data “1”. It should be noted that the spin device elements 11A1 and 11A2 are a pair of spin device elements which are always placed in the same state and the spin device elements 11B1 and 11B2 are another pair of spin device elements which are always placed in the same state. It should be also noted that the state of the spin device elements 11A1 and 11A2 is always different from the state of the spin device elements 11B1 and 11B2.

Data reading from a memory cell 1 configured as illustrated in FIG. 6 is achieved by comparing the voltages V1 and V2 generated on the nodes N1 and N2 by using a data identification circuit, e.g. a sense amplifier 12, in the state in which a first voltage is applied to the node Nbias1 and a second voltage lower than the first voltage is applied to the node Nbias2. In one example, the voltages V1 and V2 generated on the nodes N1 and N2 are compared by the sense amplifier 12 in the state in which a positive bias voltage +Vbias is applied to the node Nbias1 and the node Nbias2 is grounded. When a first voltage is applied to the node Nbias1 and a second voltage lower than the first voltage is applied to the node Nbias2 (for example, when a positive bias voltage +Vbias is applied to the node Nbias1 with the node Nbias2 grounded), a read current Isense1 flowing through the spin device elements 11A1 and 11B1 and a read current Isense2 flowing through the spin device elements 11A2 and 11B2 are generated and the voltages V1 and V2 are generated by the read currents Isense1 and Isense2 on the nodes N1 and N2, respectively. In the example illustrated in FIG. 7, in which the first state is correlated with data “0” and the second state is correlated with data “1”, the data stored in the memory cell 1 is identified as data “0” when the voltage V1 is higher than voltage V2 and as data “1” when the voltage V1 is lower than voltage V2.

The above-described memory cell 1, in which the four spin device elements 11 form a bridge circuit generating a pair of differential signals, effectively offers a large signal window ΔV even when the MR ratios of the spin device elements 11 are not so large. Accordingly, as illustrated in FIG. 8, the reliability of data reading from the memory cell 1 can be effectively improved even when the voltages V1 and V2 on the nodes N1 and N2 vary on some level.

Additionally, when the four spin device elements 11 are arranged closed to each other, this effectively suppresses the influences of variations in the properties of the spin device elements 11. Even when a manufacturing process of the magnetic memory suffers from manufacturing process variations, the influences of the manufacturing process variations on the voltages V1 and V2 on the nodes N1 and N2 can be effectively suppressed, because spin device elements placed close to each other exhibit reduced property variations.

FIG. 9A is a circuit diagram illustrating a modification of the memory cell 1 in the present embodiment. In the configuration of the memory cell 1 illustrated in FIG. 9A, two spin device elements 11 and two resistor elements 13 with a fixed resistance form a half bridge circuit. It should be noted that, suffixes “A1”, “A2”, “B1” and “B2” may be used in the following, if the resistor elements 13 are distinguished from one another.

The configuration of the memory cell 1 illustrated in FIG. 9A is derived from that of the memory cell 1 illustrated in FIG. 6 by replacing the spin device elements 11B1 and 11B2 with resistor elements with a fixed resistance value. More specifically, the memory cell 1 illustrated in FIG. 9A includes two spin device elements 11A1, 11A2 and two resistor elements 13B1 and 13B2 having a fixed resistance value. The spin device element 11A1 is connected between the node Nbias1 and the node N1 and the resistor element 13B1 is connected between the node N1 and the node Nbias2. Furthermore, the resistor element 13B2 is connected between the node Nbias1 and the node N2 and the spin device element 11A2 is connected between the node N2 and the node Nbias2. As described above, the node Nbias1 is a first bias node to which a first voltage is applied in a read operation and the node Nbias2 is a second bias node to which a second voltage lower than the first voltage is applied in the read operation.

Preferably, the resistance value Rx of the resistor elements 13B1 and 13B2 is in the range between the resistance values RHigh and RLow, where RHigh is the resistance value of the spin device element 11A1 and 11A2 in the case when the spin device element 11A1 and 11A2 are placed in the “high resistance” state, and RLow the resistance value of the spin device element 11A1 and 11A2 in the case when the spin device element 11A1 and 11A2 are placed in the “low resistance” state. More preferably, the resistance value Rx of the resistor elements 13B1 and 13B2 is adjusted to the average value of the resistance values RHigh and RLow. In an actual implementation, it is preferable that the design value of the resistance value Rx of the resistor elements 13B1 and 13B2 is determined as the average of the design values of the resistance values RHigh and RLow of the spin device elements 11A1 and 11B2, since the resistance values RHigh and RLow actually vary due to manufacturing process variations.

The memory cell 1 illustrated in FIG. 9A is configured to offer two allowed states with two spin device elements 11 and to thereby store one bit data. FIG. 10 illustrates the two allowed states of the memory cell 1 configured as illustrated in FIG. 9A. The memory cell 1 configured as illustrated in FIG. 9A stores one-bit data, correlating one of the first and second states defined below with data “0” and the other with data “1”:

First state: a state in which the spin device elements 11A1 and 11A2 are placed in the “low resistance” state, and

Second state: a state in which the spin device elements 11A1 and 11A2 are placed in the “high resistance” state.

FIG. 10 illustrates two allowed states of the memory cell 1 in the case when the first state is correlated with data “0” and the second state is correlated with data “1”. It should be noted that the spin device elements 11A1 and 11A2 are a pair of spin device elements which are always placed in the same state.

Data reading from a memory cell 1 configured as illustrated in FIG. 9A is achieved by comparing the voltages V1 and V2 generated on the nodes N1 and N2 by using a data identification circuit, e.g. a sense amplifier 12, in the state in which a first voltage is applied to the node Nbias1 and a second voltage lower than the first voltage is applied to the node Nbias2. In one embodiment, the voltages V1 and V2 generated on the nodes N1 and N2 are compared by the sense amplifier 12 in the state in which a positive bias voltage +Vbias is applied to the node Nbias1 and the node Nbias2 is grounded. In the example illustrated in FIG. 10, in which the first state is correlated with data “0” and the second state is correlated with data “1”, the data stored in the memory cell 1 is identified as data “0” when the voltage V1 is higher than voltage V2 and as data “1” when the voltage V1 is lower than voltage V2.

The memory cell 1 illustrated in FIG. 9A, in which the two spin device elements 11 and the two resistor elements 13 form a bridge circuit generating a pair of differential signals, effectively offers a large signal window ΔV even when the MR ratios of the spin device elements 11 are not so large. Accordingly, the reliability of data reading from the memory cell 1 can be effectively improved even when the voltages V1 and V2 on the nodes N1 and N2 vary on some level.

FIG. 9B is a circuit diagram illustrating another modification of the memory cell 1 in the present embodiment. Also in the configuration of the memory cell 1 illustrated in FIG. 9B, two spin device elements 11 and two resistor elements 13 with a fixed resistance form a half bridge circuit.

The configuration of the memory cell 1 illustrated in FIG. 9B is derived from that of the memory cell 1 illustrated in FIG. 6 by replacing the spin device elements 11A1 and 11A2 with resistor elements with a fixed resistance value. More specifically, the memory cell 1 illustrated in FIG. 9B includes two spin device elements 11B1, 11B2 and two resistor elements 13A1 and 13A2 having a fixed resistance value. The resistor element 13A1 is connected between the node Nbias1 and the node N1 and the spin device element 11B1 is connected between the node N1 and the node Nbias2. Furthermore, the spin device element 11B2 is connected between the node Nbias1 and the node N2 and the resistor element 13A2 is connected between the node N2 and the node Nbias2.

The memory cell 1 illustrated in FIG. 9B is also configured to offer two allowed states with four spin device elements 11 and to thereby store one bit data, as is the case with that illustrated in FIG. 9A. The memory cell 1 configured as illustrated in FIG. 9B stores one-bit data, correlating one of the first and second states defined below with data “0” and the other with data “1”:

First state: a state in which the spin device elements 11B1 and 11B2 are placed in the “high resistance” state, and

Second state: a state in which the spin device elements 11B1 and 11B2 are placed in the “low resistance” state.

It should be noted that the spin device elements 11B1 and 11B2 are a pair of spin device elements which are always placed in the same state.

Data reading from a memory cell 1 configured as illustrated in FIG. 9B is achieved by comparing the voltages V1 and V2 generated on the nodes N1 and N2 by using a data identification circuit, e.g. a sense amplifier 12, in the state in which a first voltage is applied to the node Nbias1 and a second voltage lower than the first voltage is applied to the node Nbias2, similarly to that from a memory cell 1 configured as illustrated in FIG. 9A. In one embodiment, the voltages V1 and V2 generated on the nodes N1 and N2 are compared by the sense amplifier 12 in the state in which a positive bias voltage +Vbias is applied to the node Nbias1 and the node Nbias2 is grounded. When the first state is correlated with data “0” and the second state is correlated with data “1”, for example, the data stored in the memory cell 1 is identified as data “0” when the voltage V1 is higher than voltage V2 and as data “1” when the voltage V1 is lower than voltage V2.

The memory cell 1 illustrated in FIG. 9A, in which the two spin device elements 11 and two resistor elements 13 form a bridge circuit generating a pair of differential signals, also offers a large signal window ΔV even when the MR ratios of the spin device elements 11 are not so large. Accordingly, the reliability of data reading from the memory cell 1 can be effectively improved even when the voltages V1 and V2 on the nodes N1 and N2 vary on some level.

In the following, a description is given of more specific configurations of the memory cells of the magnetic memory in the present embodiment. It should be noted that the memory cell configuration described below is designed to facilitate data writing into the memory cells.

FIG. 11A is a diagram conceptually illustrating one example of the configuration of each memory cell 1A, when the magnetic memory 10 of the present embodiment is configured as an STT-MRAM (Spin Transfer Torque Magnetoresistive Random Access Memory). In an STT-MRAM, a spin transfer torque is used in data writing into a memory cell. The configuration of the memory cell 1A illustrated in FIG. 11A corresponds to that illustrated in FIG. 4A. In the following, a description is given of the configuration of the memory cell 1A illustrated in FIG. 11A.

The memory cell 1A includes two spin device elements 11A and 11B. Each of the spin device elements 11 includes a reference layer 21, a spacer layer 22 and a recording layer 23. The structure of the spin device elements 11 is as described above with reference to FIG. 4B.

The spin device elements 11A and 11B are formed on the upper surfaces of lower electrodes 24A and 24B, respectively. In detail, the reference layers 21 of the spin device elements 11A and 11B are formed on the upper surfaces of the lower electrodes 24A and 24B, respectively. In each of the spin device elements 11A and 11B, the spacer layer 22 is formed on the upper surface of the reference layer 21 and the recording layer 23 is formed on the upper surface of the spacer layer 22.

The lower electrode 24A is connected to a node Nbias1 and the lower electrode 24B is connected to a node Nbias2, where the nodes Nbias1 and Nbias2 are bias nodes to which bias voltages are applied in read and write operations. The lower electrode 24A functions as an interconnection which electrically connects the spin device element 11A to the node Nbias1 and the lower electrode 24B functions as an interconnection which electrically connects the spin device element 11B to the node Nbias2.

An upper electrode 25 is coupled with the upper surfaces of the spin device elements 11A and 11B. The upper electrode 25, which functions as an interconnection that provides an electrical connection between the spin device elements 11A and 11B, is a circuit component which corresponds to the node N1 of the memory cell 1 illustrated in FIG. 4A.

As illustrated in FIG. 11B, the spin device elements 11A and 11B may be electrically connected to each other via a lower electrode 24. In this case, an upper electrode 25A formed on the upper surface of the spin device element 11A is connected to the node Nbias1, and an upper electrode 25B formed on the upper surface of the spin device element 11B is connected to the node Nbias2.

It should be noted that the positions of the reference layer 21 and the recording layer 23 may be exchanged in each spin device element 11 in both of the configurations illustrated in FIGS. 11A and 11B.

The memory cells 1A illustrated in FIGS. 11A and 11B, similarly to the memory cell 1 illustrated in FIG. 4A, are configured to store one-bit data, correlating one of the first and second states defined below with data “0” and the other with data “1”:

First state: a state in which the spin device element 11A is placed in the “low resistance” state and the spin device element 11B is placed in the “high resistance” state, and

Second state: a state in which the spin device element 11A is placed in the “high resistance” state and the spin device element 11B is placed in the “low resistance” state.

It should be noted that, in the configurations illustrated in FIGS. 11A and 11B, the reference layers 21 of the two spin device elements 11A and 11B are connected to each other (see FIG. 11B) or the recording layers 23 of the two spin device elements 11A and 11B are connected to each other (see FIG. 11A). Such a connection aims at placing the memory cell 1A into the first or second state by feeding a write current which flows between the node Nbias1 and Nbias2. When a write current is caused to flow through the memory cell 1A, one of the spin device elements 11A and 11B, in which the directions of the write currents are different, is placed into the “high resistance” state and the other is placed into the “low resistance” state, since the direction into which the magnetization is directed in the magnetization reversal based on the spin transfer torque depends on the direction of the write current. In the configuration illustrated in FIG. 11A, for example, when a write current which flows from the node Nbias1 to the node Nbias2 is fed, the write current flows from the reference layer 21 to the recording layer 23 in the spin device element 11A and from the recording layer 23 to the reference layer 21 in the spin device element 11B. This allows placing the memory cell 1A into the first state defined above. When a write current is fed from the node Nbias2 to the node Nbias1, the write current flows from the recording layer 23 to the reference layer 21 in the spin device element 11A and the write current flows from the reference layer 21 to the recording layer 23 in the spin device element 11B. This allows placing the memory cell 1A into the second state defined above. A person skilled in the art would appreciate it that this discussion also applies to the configuration illustrated in FIG. 11B.

Next, a description is given of data writing and reading into and from the memory cell 1A illustrated in FIG. 11A. FIGS. 12A and 12B illustrate data writing into the memory cell 1A illustrated in FIG. 11A. It should be noted that FIGS. 12A and 12B illustrates a write operation in the case when the first state is correlated with data “0” and the second state is correlated with data “1”. As described above, the first state is the state in which the spin device element 11A is placed in the “low resistance state” and the spin device element 11B is placed in the “high resistance” state” and the second state is the state in which the spin device element 11A is placed in the “high resistance” state and the spin device element 11B is placed in the “low resistance” state.

In data writing, bias voltages are applied to the nodes Nbias1 and Nbias2 depending on data to be written into the memory cell 1A to generate a write current IW between the nodes Nbias1 and Nbias2.

In detail, to write data “0”, a voltage lower than the voltage on the node Nbias2 is applied to the node Nbias1 as illustrated in FIG. 12A. In the operation illustrated in FIG. 12A, the node Nbias2 is grounded and a negative bias voltage −Vbias is applied to the node Nbias1. This causes a write current to flow from the node Nbias2 to the node Nbias1, directing the magnetizations of the reference layer 21 and the recording layer 23 in “parallel” in the spin device element 11A, and in “antiparallel” in the spin device element 11B. As a result, the memory cell 1A is placed into the first state, in which the spin device element 11A is in the “low resistance” state and the spin device element 11B is in the “high resistance” state.

To write data “1”, on the other hand, a voltage higher than the voltage on the node Nbias2 is applied to the node Nbias1 as illustrated in FIG. 12B. In the operation illustrated in FIG. 12B, the node Nbias2 is grounded and a positive bias voltage Vbias is applied to the node Nbias1. This causes a write current to flow from the node Nbias1 to the node Nbias2, directing the magnetizations of the reference layer 21 and the recording layer 23 in “antiparallel” in the spin device element 11A and in “parallel” in the spin device element 11B. As a result, the memory cell 1A is placed into the second state, in which the spin device element 11A is in the “high resistance” state and the spin device element 11B is in the “low resistance” state.

FIG. 13 illustrate data reading from the memory cell 1A illustrated in FIG. 11A. In data reading, a positive bias voltage +Vbias is applied to the node Nbias1 and a negative bias voltage −Vbias is applied to the node Nbias2. It should be noted that the bias voltages applied to the nodes Nbias1 and Nbias2 have substantially the same absolute value (magnitude) but have opposite polarities. When the positive bias voltage +Vbias is applied to the node Nbias1 and the negative bias voltage −Vbias is applied to the node Nbias2, a read current IR flowing through the spin device elements 11A and 11B is generated and a voltage V1 is generated by the read current IR on the upper electrode 25, which corresponds to the node N1.

The data stored in the memory cell 1A is identified in the same way as the memory cell 1 illustrated in FIG. 4A. When the first state is correlated with data “0” and the second state is correlated with data “1”, for example, the data stored in the memory cell 1A is identified as data “0” if the polarity of the voltage V1 is positive and as data “1” if the polarity of the voltage V1 is negative.

FIG. 14A schematically illustrates another example of the configuration of each memory cell 1B, when the magnetic memory 10 of the present embodiment is configured as an STT-MRAM. The configuration of the memory cell 1B illustrated in FIG. 14A corresponds to that illustrated in FIG. 6. In the following, a description is given of the configuration of the memory cell 1B illustrated in FIG. 14A.

The memory cell 1B includes four spin device elements 11A1, 11A2, 11B1, 11B2 and switching elements SW1 and SW2. Each of the four spin device elements 11 includes a reference layer 21, a spacer layer 22 and a recording layer 23. The configuration of each spin device element 11 is as described above with reference to FIG. 4B.

The spin device elements 11A1, 11A2, 11B2 and 11B2 are formed on the upper surfaces of lower electrodes 24A1, 24A2, 2421 and 2422, respectively. In detail, the reference layers 21 of the spin device elements 11A1, 11A2, 11B2 and 11B2 are formed on the upper surfaces of the lower electrodes 24A1, 24A2, 24B1, and 24B2, respectively. In each of the spin device elements 11A1, 11A2, 11B2 and 11B2, the spacer layer 22 is formed on the reference layer 21 and the recording layer 23 is formed on the spacer layer 22.

The lower electrode 24A1 is connected to a node N11 and the lower electrode 24B2 is connected to a node N12, where the nodes N11 and N12 are connection nodes used to establish electrical connections with the spin device elements 11A1 and 11B2. In other words, the lower electrode 24A1 functions as an interconnection which electrically connects the spin device element 11A1 to the node N11 and the lower electrode 24B2 functions as an interconnection which electrically connects the spin device element 11B2 to the node N12. The node N11 is connected to the node Nbias1 and the node N12 is connected to the switching element SW1. As described layer, the node Nbias1 is a bias node to which a bias voltage is applied in data writing and data reading.

The lower electrodes 24B1 and 24A2 are commonly connected to a node N13. In other words, the lower electrodes 24B1 and 24A2 function as interconnections which electrically connect the spin device elements 11B1 and 11A2 to the node N13. The node N13 is a connection node which provides an electrical connection between the spin device elements 11B1 and 11A2.

An upper electrode 251 is coupled with the upper surfaces of the spin device elements 11A1 and 11B1, and an upper electrode 252 is coupled with the upper surfaces of the spin device elements 11A2 and 11B2. The upper electrode 251, which functions as an interconnection which provides an electrical connection between the spin device elements 11A1 and is a component corresponding to the node N1 illustrated in FIG. 6. The upper electrode 252, which functions as an interconnection which provides an electrical connection between the spin device elements 11A2 and 11B2, is a component corresponding to the node N2 illustrated in FIG. 6.

The switching element SW1 connects the node N12 to selected one of the nodes Nbias1 and Nbias3, where the node Nbias3 is a bias node which is kept at a predetermined voltage; in the present embodiment, the node Nbias3 is grounded. As described later, the node Nbias3 is used in write operations. The switching element SW2 electrically connects or disconnects the node N13 to or from the node Nbias2, where the node Nbias2 is a bias node which kept at a predetermined voltage level; in the present embodiment, the node Nbias2 is grounded. As described later, the node Nbias2 is used in read operations. The switching elements SW1 and SW2 are used to switch the memory cell 1B between data writing and data reading. In data writing, the switching element SW1 is placed into a state in which the node N12 is connected to the node Nbias3 (that is, the circuit ground), and the switching element SW2 is placed into the off-state. In data reading, on the other hand, the switching element SW1 is placed into a state in which the node N12 is connected to the node Nbias1, and the switching element SW2 is placed into the on-state.

The memory cell 1B illustrated in FIG. 14A, similarly to the memory cell 1 illustrated in FIG. 6, is configured to store one-bit data, correlating one of the first and second states defined below with data “0” and the other with data “1”:

First state: a state in which the spin device elements 11A1 and 11A2 are placed in the “low resistance” state and the spin device elements 11B1 and 11B2 are placed in the “high resistance” state, and

Second state: a state in which the spin device elements 11A1 and 11A2 are placed in the “high resistance” state and the spin device elements 11B1 and 11B2 are placed in the “low resistance” state. It should be noted that the spin device elements 11A1 and 11A2 are always placed in the same state, and the spin device elements 11B1 and 11B2 are always placed in the same state. It should be also noted that the state of the spin device elements 11A1 and 11A2 is always different from the state of the spin device elements 11B1 and 11B2.

In the configuration illustrated in FIG. 14A, the spin device elements 11A1, 11B1, 11A2 and 11B2 are serially connected in this order between the nodes N11 and N12. It should be noted that a connection between adjacent two spin device elements 11 is achieved by connecting the reference layers 21 of the adjacent two spin device elements 11 each other or by connecting the reference layers 23 of the adjacent two spin device elements 11 each other. In detail, the recording layers 23 of the spin device elements 11A1 and 11B1 are electrically connected to each other via the upper electrode 251 and the recording layers 23 of the spin device elements 11B2 and 11A2 are electrically connected to each other via the upper electrode 252. Furthermore, the reference layers 21 of the spin device elements 11B1 and 11A2 are electrically connected via the lower electrode 24B1, the node N13 and the lower electrode 24A2.

As thus described, when a write current is caused to flow through the memory cell 1B, in which the reference layers 21 or recording layers 23 of adjacent two spin device elements 21 are connected to each other, the directions of the write currents flowing between the reference layers 21 and the recording layers 23 via the space layers 22 are different between the spin device elements 11A1 and 11B1. One of the spin device elements 11A1 and 11B1, in which the directions of the write currents are different, is placed into the “high resistance” state and the other is placed into the “low resistance” state, since the direction into which the magnetization is directed in the magnetization reversal based on the spin transfer torque depends on the direction of the write current. This discussion also applies to the spin device elements 11A2 and 11B2. The above-described electrical connections allow placing one of adjacent two spin device elements 11 into the “high resistance” state and the other into the “low resistance state”.

Accordingly, the memory cell 1B, configured as described above, can be placed into the first or second state by causing a write current to flow between the node N11 and N12. When a write current which flows from the node N11 to the node N12 is generated, for example, the write current flows from the reference layer 21 to the recording layer 23 in the spin device elements 11A1 and 11A2, and the write current flows from the recording layer 23 to the reference layer 21 in the spin device elements 11B1 and 11B2. This allows placing the memory cell 1B into the first state. When a write current which flows from the node N12 to the node N11 is generated, on the other hand, the write current flows from the recording layer 23 to the reference layer 21 in the spin device elements 11A1 and 11A2, and the write current flows from the reference layer 21 to the recording layer 23 in the spin device elements 11B1 and 11B2. This allows placing the memory cell 1B into the second state described above.

It should be noted that the positions of the reference layer 21 and the recording layer 23 may be exchanged in each of the spin device elements 11, as illustrated in FIG. 14B. Also in this case, with respect to every adjacent two spin device elements 11, the reference layers 21 of the adjacent two spin device elements 11 are connected to each other, or the recording layers 23 of the adjacent two spin device elements 11 are connected to each other. In detail, also in the configuration illustrated in FIG. 14B, the reference layers 21 of the spin device elements 11A1 and 11B1 are electrically connected to each other via the upper electrode 251 and the reference layers 21 of the spin device elements 11A2 and 11B2 are electrically connected to each other via the upper electrode 252. Furthermore, the recording layers 23 of the spin device elements 11B1 and 11A2 are electrically connected to each other via the lower electrode 24B1, the node N13 and the lower electrode 24A2. A person skilled in the art would appreciate that the memory cell 1B illustrated in FIG. 14B can be also placed into the first or second state by feeding a write current flowing between the nodes N11 and N12.

Although the memory cell 1B illustrated in FIG. 14A is configured so that the node N11 is connected to the node Nbias1 and the switch element SW1 connects the node N12 to selected one of the node Nbias1 and Nbias3, the memory cell 1B may be configured so that the node N12 is connected to the node Nbias1 and the switch element SW1 connects the node N11 to selected one of the node Nbias1 and Nbias3, as illustrated in FIG. 14C. Also in this case, the positions of the reference layer 21 and the recording layer 23 may be exchanged in each of the spin device elements 11 as illustrated in FIG. 14D.

In an alternative embodiment, as illustrated in FIG. 14E, the memory cell 1B may be configured so that the spin device elements 11A1 and 11B1 are electrically connected to each other via a lower electrode 241 and the spin device elements 11B2 and 11A2 are electrically connected to each other via a lower electrode 242. In this case, an upper electrode 25A1 formed on the upper surface of the spin device element 11A1 is connected to the node N11 is connected, and an upper electrode 25B2 formed on the upper surface of the spin device element 11B2 is connected to the node N12. Also, upper electrodes 25B1 and 25A2 respectively formed on the upper surfaces of the spin device elements 11B1 and 11A2 are commonly connected to the node N13. It should be noted that, also in this configuration, with respect to every adjacent two spin device elements 11, the reference layers 21 of the adjacent two spin device elements 11 are connected to each other, or the recording layers 23 of the adjacent two spin device elements 11 are connected with each other.

As illustrated in FIG. 14F, the memory cell 1B may be configured so that the node N12 is connected to the node Nbias1 and the switching element SW1 connects the node N11 to selected one of the nodes Nbias1 and Nbias3, also in the case when the spin device elements 11A1 and 11B1 are electrically connected to each other via the lower electrode 241 and the spin device elements 11B2 and 11A2 are electrically connected to each other via the lower electrode 242. It should be noted that the positions of the reference layer 21 and the recording layer 23 may be exchanged in each spin device element 11, also in the configurations illustrated in FIGS. 14E and 14F.

Next, a description is given of data writing and reading into and from the memory cell 1B illustrated in FIG. 14A. FIGS. 15A and 15B illustrate data writing into the memory cell 1B illustrated in FIG. 14A. It should be noted that FIGS. 15A and 15B illustrates a write operation in the case when the first state is correlated with data “0” and the second state is correlated with data “1”. As described above, the first state is the state in which the spin device elements 11A1 and 11A2 are placed in the “low resistance state” and the spin device elements 11B1 and 11B2 are placed in the “high resistance” state” and the second state is the state in which the spin device elements 11A1 and 11A2 are placed in the “high resistance” state and the spin device elements 11B1 and 11B2 are placed in the “low resistance” state.

In data writing, the switching element SW1 is placed into the state in which the node N12 is connected to the node Nbias3 and the switching element SW2 is placed into the off-state. Furthermore, a bias voltage depending on write data is applied to the node Nbias1 to generate a write current IW which flows between the nodes N11 and N12.

In detail, to write data “0”, as illustrated in FIG. 15A, a bias voltage lower than the voltage on the node Nbias3 is applied to the node Nbias1. In the present embodiment, in which the node Nbias3 is grounded, a negative bias voltage −Vbias is applied to the node Nbias1. This allows the write current to flow from the node N12 to the node N11, resulting in that the magnetizations of the reference layer 21 and the reference layer 23 are directed in “parallel” in each of the spin device elements 11A1 and 11A2 and directed in “antiparallel” in each of the spin device elements 11B1 and 11B2. This operation places the memory cell 1B into the first state, in which the spin device elements 11A1 and 11A2 are placed in the “low resistance” state and the spin device elements 11B1 and 11B2 are placed in the “high resistance” state. It should be noted that a pair of the “low resistance” state and the “high resistance” state is generated in this operation, because the write current flows in different directions in the adjacent spin device elements 11A1 and 11B1. The same applies to the spin device elements 11A2 and 11B2.

To write data “1”, on the other hand, as illustrated in FIG. 15B, a bias voltage higher than the voltage on the node Nbias3 is applied to the node Nbias1. In the present embodiment, in which the node Nbias3 is grounded, a positive bias voltage +Vbias is applied to the node Nbias1. This allows the write current to flow from the node N11 to the node N12, resulting in that the magnetizations of the reference layer 21 and the reference layer 23 are directed in “antiparallel” in each of the spin device elements 11A1 and 11A2 and directed in “parallel” in each of the spin device elements 11B1 and 11B2. This operation places the memory cell 1B into the second state, in which the spin device elements 11A1 and 11A2 are placed in the “high resistance” state and the spin device elements 11B1 and 11B2 are placed in the “low resistance” state.

FIGS. 16A and 16B illustrates data reading from the memory cell 1B illustrated in FIG. 14A. In data reading, the switching element SW1 is placed into a state in which the node N12 is connected to the node Nbias1, and the switching element SW2 is placed into the on-state. Furthermore, a bias voltage higher than the voltage on the node Nbias2 is applied to the node Nbias1. In the present embodiment, in which the node Nbias2 is grounded, a positive bias voltage Vbias is applied to the node Nbias1. This operation generates a read current Isense1 flowing through the spin device elements 11A1 and 11B1 and a read current Isense2 flowing through the spin device elements 11B2 and 11A2. The read currents Isense1 and Isense2 generate a voltage V1 on the upper electrode 251, which corresponds to the node N1, and a voltage V2 on the upper electrode 252, which corresponds to the node N2.

The data stored in the memory cell 1B is identified in the same way as the memory cell 1 illustrated in FIG. 6. When the voltage V1 generated on the upper electrode 251 is higher than the voltage V2 generated on the upper electrode 252, as illustrated in FIG. 16A, the data stored in the memory cell 1B is identified as data “0”, because the spin device elements 11A1 and 11A2 are placed in the “low resistance” state and the spin device elements 11B1 and 11B2 are placed in the “high resistance” state. When the voltage V2 generated on the upper electrode 252 is higher than the voltage V1 generated on the upper electrode 251, as illustrated in FIG. 16B, the data stored in the memory cell 1B is identified as data “1”, because the spin device elements 11A1 and 11A2 are placed in the “high resistance” state and the spin device elements 11B1 and 11B2 are placed in the “low resistance” state.

Data reading and data writing can be similarly achieved when the configurations illustrated in FIGS. 14B to 14F are used instead. In data writing, the switching element SW1 is placed into the state in which the node N11 or N12 is connected to the node Nbias3 (that is, the circuit ground) and the switching element SW2 is placed into the off-state. Furthermore, a bias voltage depending on write data is applied to the node Nbias1 to generate a write current IW which flows between the nodes N11 and N12. In data reading, the switching element SW1 is placed into a state in which the node N11 or N12 is connected to the node Nbias1, and the switching element SW2 is placed into the on-state.

FIG. 17A is a plan view illustrating one example of the configuration of a memory cell 1C in the case when the magnetic memory 10 in the present embodiment is configured to achieve data writing with a current-induced magnetic field, and FIG. 17B is a side view illustrating the configuration of the memory cell 1C. Also in the configuration illustrated in FIG. 17A, the memory cell 1C includes four spin device elements 11A1, 11A2, 11B1 and 11B2. When data writing is achieved with a current-induced magnetic field, a write current line 26 is additionally provided close to each memory cell 1C. The write current line 26 is disposed to extend in the X-axis direction. The spin device elements 11A1 and 11A2 are positioned opposed to the spin device elements 11B1 and 11B2 across the write current line 26. More specifically, the spin device elements 11A1 and 11A2 are positioned shifted in the +Y direction (a first direction) from the write current line 26 and the spin device elements 11B1 and 11B2 are positioned shifted in the −Y direction (a second direction opposite to the first direction) from the write current line 26. This arrangement allows generating such a magnetic field that the magnetizations of the recoding layers 23 of the spin device elements 11A1 and 11A2 and those of the recoding layers 23 of the spin device elements 11B1 and 11B2 are directed in opposite directions, by using only one write current line 26. Note that the spin device element 11A2 is positioned shifted from the spin device element 11A1 in the +X direction, and the spin device element 11B1 is positioned shifted from the spin device element 11B2 in the +X direction.

FIGS. 17C and 17D are cross-sectional views illustrating the structure of the memory cell 1C on sections A-A and B-B indicated in FIG. 17A, respectively. As illustrated in FIGS. 17C and 17D, the spin device elements 11A1 and 11A2 are formed on the upper surfaces of the lower electrodes 24A1 and 24A2, respectively, and the spin device elements 11B1 and 11B2 are formed on the upper surfaces of the lower electrodes 24B1 and 24B2, respectively. In detail, the reference layers 21 of the spin device elements 11A1, 11A2, 11B1 and 11B2 are formed on the upper surfaces of the lower electrodes 24A1, 24A2, 24B1 and 24B2, respectively. In each of the spin device elements 11A1, 11A2, 11B1 and 11B2, the spacer layer 22 is formed on the upper surface of the reference layer 21, and the recording layer 23 is formed on the upper surface of the spacer layer 22. The upper electrodes 25A1, 25A2, 25B1 and 25B2 are formed on the upper surfaces of the recording layers 23 of the spin device elements 11A1, 11A2, 11B1 and 11B2, respectively.

Data writing into the memory cell 1C illustrated in FIGS. 17A to 17D can be achieved by generating a current-induced magnetic field with a write current Iw flowing through the write current line 26. The direction of the write current Iw is determined depending on data to be written into the memory cell 1C.

Discussed below is the case when a first state is correlated with data “0” and a second state is correlated with data “1”, where the first state is a state in which the spin device elements 11A1 and 11A2 are placed in the “low resistance” state and the spin device elements 11B1 and 11B2 are placed in the “high resistance” state and the second state is a state in which the spin device elements 11A1 and 11A2 are placed in the “high resistance” state and the spin device elements 11B1 and 11B2 are placed in the “low resistance” state.

To write data “0”, as illustrated in FIG. 18A, a write current Iw is generated which flows through the write current line 26 in the +X direction. When the write current Iw flows in the +X direction, a current-induced magnetic field having a +Z component is applied to the recording layers 23 of the spin device elements 11A1 and 11A2 and a current-induced magnetic field having a −Z component is applied to the recording layers 23 of the spin device elements 11B1 and 11B2. This allows directing the magnetizations of the recording layers 23 of the spin device elements 11A1 and 11A2 in the +Z direction to place the spin device elements 11A1 and 11A2 into the “low resistance” state and directing the magnetizations of the recording layers 23 of the spin device elements 11B1 and 11B2 in the −Z direction to place the spin device elements 11B1 and 11B2 into the “high resistance” state. As thus described, data “0” can be written into the memory cell 1C by feeding a write current Iw flowing in the +X direction.

To write data “1”, on the other hand, as illustrated in FIG. 18B, a write current Iw is generated which flows through the write current line 26 in the −X direction. When the write current Iw flows in the −X direction a current-induced magnetic field having a −Z component is applied to the recording layers 23 of the spin device elements 11A1 and 11A2 and a current-induced magnetic field having a +Z component is applied to the recording layers 23 of the spin device elements 11B1 and 11B2. This allows directing the magnetizations of the recording layers 23 of the spin device elements 11A1 and 11A2 in the −Z direction to place the spin device elements 11A1 and 11A2 into the “high resistance” state and directing the magnetizations of the recording layers 23 of the spin device elements 11B1 and 11B2 in the +Z direction to place the spin device elements 11B1 and 11B2 into the “low resistance” state. As thus described, data “1” can be written into the memory cell 1C by generating a write current Iw flowing in the −X direction.

FIG. 19 is a perspective view conceptually illustrating electrical connections among the spin device elements 11A1, 11A2, 11B1 and 11B2 in the memory cell 1C for achieving data reading. The electrical connections among the spin device elements 11A1, 11A2, 11B1 and 11B2 in the memory cell 1C are similar to those in the memory cell 1 illustrated in FIG. 6.

In detail, the lower electrode 24A1 and 24B2, which are coupled with the spin device elements 11A1 and 11B2, respectively, are connected to the node Nbias1 and the lower electrode 24A2 and 24B1, which are coupled with the spin device elements 11A2 and 11B1, respectively, are connected to the node Nbias2. The upper electrode 24A1 and 24B1, which are connected coupled with the spin device elements 11A1 and 11B1, respectively, are commonly connected to the node N1, and the upper electrode 24B2 and 24A2, which are coupled with the spin device elements 11A2 and 11B2, respectively, are commonly connected to the node N2.

Data reading from the memory cell 1C is achieved by comparing the voltages V1 and V2 generated on the nodes N1 and N2 by using a sense amplifier 12, in the state in which a voltage higher that the voltage on the node Nbias2 is applied to the node Nbias1. In the read operation illustrated in FIG. 19, a positive bias voltage +Vbias is applied to the node Nbias1 with the node Nbias2 grounded. When the positive bias voltage +Vbias is applied to the node Nbias1, a read current Isense1 flowing through the spin device elements 11A1 and 11B1 and a read current Isense2 flowing through the spin device elements 11A2 and 11B2 are generated and the voltages V1 and V2 are generated by the read currents Isense1 and Isense2 on the nodes N1 and N2, respectively. When the first state is correlated with data “0” and the second state is correlated with data “1”, for example, the data stored in the memory cell 1 is identified as data “0” if the voltage V1 is higher than voltage V2 and as data “1” if the voltage V1 is lower than voltage V2.

(Integration of Spin Device Element on Deformable Base Plate)

One known issue in commercialization of magnetic memories is a difficulty in concurrently satisfying these three requirements: data read sensitivity, data stability and data write power (electric power consumed in a data writing operation). This issue involves three types of conflicts. When the MR (magnetoresistance) ratio is increased to improve the data read sensitivity and to thereby reduce the error rate in read operations, for instance, this makes it difficult to reduce the data write power required for generating a spin transfer torque. The opposite also applies. As for the relation between the data write power and data stability, the data stability increases proportionally with the magnetic anisotropy energy KuV (where Ku is the magnetic anisotropy constant and V is the volume) while the data write power also increases proportionally with the magnetic anisotropy energy KuV. Accordingly, the improvement in the data stability inevitably increases the data write power. As thus discussed, although it is possible to individually satisfy each of the above-described three requirements with the current technologies, it is difficult to satisfy the three requirements at the same time with the current technologies, due to the trade-off relationship. If a conflict between two of the three above-described requirements is resolved, this would effectively contribute commercialization of magnetic memories.

To address the above-described trade-off relationship, as one approach to relieve a conflict between the data stability and data write power in a magnetic memory, the inventor has been studying a technique in which spin device elements of a memory cell are integrated on a deformable base plate and a bending mechanism which bends the deformable base plate is also integrated. Further details are disclosed in U.S. patent application Ser. No. 14/728,651, filed on Jun. 2, 2015, the disclosure of which is incorporated herein by reference. A mechanism which uses a piezoelectric effect or a mechanism which uses a force exerted between a pair of capacitor electrodes may be used as the bending mechanism to bend the deformable base plate.

In a magnetic memory thus configured, a mechanical stress is applied to each spin device element when the deformable base plate is bent and this stress generates a strain in each spin device element. When the strain is generated in each spin device element, the magnetization direction of the recording layer of each spin device element is tilted from the original magnetization direction, which is determined depending on the magnetic anisotropy of the recording layer, due to a magnetostrictive effect. The magnetostrictive effect is known as a phenomenon in which the strain of a magnetic body varies depending on the magnetization status and the magnetization status of a magnetic body varies depending on the strain applied to the magnetic body. Strictly, the latter effect should be referred to as an inverse magnetostrictive effect; however, these effects are collectively referred to as the magnetostrictive effect in the present application, since the magnetostrictive effect in the broad sense usually means to encompass both of the magnetostrictive effect and inverse magnetostrictive effect in the narrow sense. The magnetization of a recording layer is easily reversed in the state in which the magnetization direction of the recording layer is tilted from the original magnetization direction of the recording layer. By making use of this phenomenon, it is possible to achieve data writing with a reduced data write power, even if the recording layer of each spin device element is formed of material superior in the data stability. In the following, a description is given of embodiments in which spin device elements forming a memory cell are integrated on a deformable base plate.

FIG. 20 is a plan view illustrating an example of the configuration of a memory cell array of a magnetic memory in one embodiment, and FIG. 21 is a plan view illustrating the configuration of each block of the memory cell array illustrated in FIG. 20. Note that a block is one unit structure of the memory cell array and each block includes a plurality of memory cells as described later. FIGS. 20 and 21 conceptually illustrate the configuration of the memory cell array and that of each block in the case when the magnetic memory of the present embodiment is configured as an STT-MRAM.

As illustrated in FIG. 20, the memory cell array includes blocks 2 arranged in rows and columns in the present embodiment. As illustrated in FIG. 21, each block 2 includes a plurality of memory cells 1B. In the structure illustrated in FIGS. 20 and 21, each block 2 includes four memory cells 1B. The configuration of each memory cell 1B is basically similar to that illustrated in any of FIGS. 14A to 14F, and therefore the memory cells included in each block 2 are denoted by the same reference numeral “1B”.

FIG. 22A is a cross-sectional view illustrating the structure of each block 2 on section C-C indicated in FIG. 21, FIG. 22B is a cross-sectional view illustrating the structure of each block 2 on section D-D indicated in FIG. 21, and FIG. 23 is a perspective view illustrating the configuration of each block 2. As illustrated in FIGS. 22A and 22B, each block 2 includes a deformable base plate 33. The deformable base plate 33 is configured to be deformable so that the deformable base plate 33 can be bent. The deformable base plate 33 may be formed of silicon, silicon oxide, silicon nitride, silicon oxynitride or the like. Materials and processes used in a SON (Si-on-nothing) technology may be used for the formation of the deformable base plate 33. The deformable base plate 33 preferably has a thickness of 200 nm to 5 μm, for example.

The lower electrodes 24A2, 24A2, 24B1 and 24B2 are formed on the upper surface of the deformable base plate 33 and the spin device elements 11A1, 11A2, 11B2 and 11B2 are formed on the upper surfaces of the lower electrodes 24A1, 24A2, 24B1 and 24B2, respectively. The upper electrodes 25A1, 25A2, 25B1 and 25B2 are formed on the upper surfaces of the spin device elements 11A2, 11B2 and 11B2, respectively. The upper electrodes 25A1 and 25B1 are electrically connected to each other via an interconnection and, the upper electrodes 25B2 and 25A2 are electrically connected to each other via an interconnection. It would be understood that such electrical connections correspond to the configuration in which the upper electrode 251 is coupled with the upper surfaces of the spin device elements 11A1 and 11B1 and the upper electrode 252 is coupled with the upper surfaces of the spin device elements 11B2 and 11A2, as illustrated in FIGS. 14A to 14F.

Piezoelectric layers 34 and 35 are coupled with both of the side surfaces (the surfaces facing the X-axis directions) of the deformable base plate 33. The piezoelectric layers 34 and 35 are formed of piezoelectric material, such as AlN, lead zirconate titanate (PZT) and zirconium oxide (ZrOx). As described later, the deformable base plate 33 is bent by using a piezoelectric effect occurring in the piezoelectric layers 34 and 35. The piezoelectric layers 34 and 35 are coupled and fixed on the upper surface of a fixture base 32. The fixture base 32 is coupled with the upper surface of a semiconductor substrate 31 in which a transistor circuit are integrated.

Electrode layers 36 and 37 are coupled with the surfaces of the piezoelectric layers 34 and 35 opposing to the surfaces coupled to the side surfaces of the deformable base plate 33. An additional electrode layer (not illustrated) on which a potential difference is generated with respect to the electrode layer 36 may be coupled with the piezoelectric layer 34. Similarly, an additional electrode layer (not illustrated) on which a potential difference is generated with respect to the electrode layer 37 may be coupled with the piezoelectric layer 35. Although the additional electrode layers on which potential differences are respectively generated with respect to the electrode layers 36 and 37 are not illustrated in FIGS. 22A and 22B, the additional electrode layers may be also used as the lower electrodes 24 connected to the spin device elements 11 or other electrodes.

In the present embodiment, the deformable base plate 33 is formed so that at least one of the lower surface (that is, the surface opposed to the surface on which the spin device elements 11 are formed) and the upper surface (that is, the surface on which the spin device elements 11 are formed) of the deformable base plate 33 faces a “space which is not filled with solid substance”. The “space which is not filled with solid substance” referred to herein may be filled with fluid, including gas (such as air, nitrogen) and liquid. Alternatively, the “space which is not filled with solid substance” may be vacuumed. The structure in which at least one of the lower and upper surfaces of the deformable base plate 33 faces the “space which is not filled with solid substance” allows the deformable base plate 33 to be deformed with a sufficiently large displacement. This is effective for generating a sufficiently-large strain, that is, a magnetostrictive effect in the recording layer 23 of each spin device element 11.

Referring back to FIG. 20, every two deformable base plates 33 of the blocks 2 adjacent to each other in the Y-axis direction are separated by a gap 39 in the present embodiment. The gap 39 is a space which is not filled with solid substance, as is the case with the space 38. FIG. 24 is a cross-sectional view schematically illustrating the structure on section E-E indicated in FIG. 20. Two deformable base plates 33 adjacent to each other in the Y-axis direction are opposed across the gap 39. The gap 39 is formed to communicate with the space 38. In this structure, two deformable base plates 33 adjacent to each other in the Y-axis direction are mechanically separated and this allows individually bending two deformable base plates 33 adjacent to each other in the Y-axis direction.

In the magnetic memory thus structured, the deformable base plate 33 can be bent by applying electric fields to the piezoelectric layers 34 and 35 with electrode layers coupled with the piezoelectric layers 34 and 35 (the electrode layers 36, 37 and the additional electrode layers (not illustrated)). More specifically, when electric fields are applied to the piezoelectric layers 34 and 35, strains are generated in the piezoelectric layers 34 and 35 due to the piezoelectric effect. A force is applied to the deformable base plate 33 due to the strains of the piezoelectric layers 34 and 35, and thereby the deformable base plate 33 is bent. The conflict between the data stability and data write power is effectively relieved by feeding to the spin device elements 11 a write current Iw in the state in which the deformable base plate 33 is bent.

It should be also noted that substantially no current is generated (other than a leakage current and a temporary charging current) to bend the deformable base plate 33 in the structure of the magnetic memory illustrated in FIGS. 20 to 24 in the present embodiment. In the structure of the magnetic memory illustrated in FIGS. 20 to 24, in which a piezoelectric effect is used to bend the deformable base plate 33, it is not necessary to feed a current to bend the deformable base plate 33. This is advantageous for reducing the data write power (electric power consumed in a data writing operation).

FIG. 25A to FIG. 25D are cross-sectional views illustrating an exemplary data write procedure into a memory cell 1B of the memory cell array illustrated in FIGS. 20 to 24. FIG. 25A illustrates the initial state of a selected memory sell 1B into which a data is to be written. In FIG. 25A, the states of only the spin device elements 11A1 and 11B1 of the selected memory cell 1B is illustrated. Discussed below is the case when the memory cell 1B stores data “1” in the initial state. In this state, the spin device element 11A1 is placed in the “high resistance” state and the spin device element 11B1 is placed in the “low resistance” state. Although not illustrated in FIG. 25A, a person skilled in the art would understand that the spin device element 11A2 is placed in the “high resistance” state and the spin device element 11B2 is placed in the “low resistance” state.

As illustrated in FIG. 25B, the deformable base plate 33 is bent in data writing into the memory cell 1B. As described above, the deformable base plate 33 can be bent by applying electric fields to the piezoelectric layers 34 and 35 with the electrode layers coupled with the piezoelectric layers 34 and 35 (that is, the electrode layers 36, 37 and the additional electrode layers (not illustrated)). When the deformable base plate 33 is bent, a strain is generated in each of the spin device electrode 11 and the magnetization direction of the recording layer 23 of each spin device element 11 is tilted from the original magnetization direction of the recording layer 23, which is determined by the magnetic anisotropy thereof, due to a magnetostrictive effect.

By tilting the magnetization direction of the recording layer 23 from the original magnetization direction, the recoding layer 23 is placed into a state in which the magnetization of the recoding layer 23 is easily reversible. This means that applying a strain to each spin device element 11 by bending the deformable base plate 33 allows temporarily placing each spin device element 11 into a state in which data stability is low. To apply a sufficiently-large strain to each spin device element 11, it is effective to increase the displacement (bending amount) of the deformable base plate 33. It is also preferable to increase the magnetostriction of the recording layer 23 to enhance the magnetostrictive effect.

It should be noted that the structure in which at least one of the lower and upper surfaces of the deformable base plate 33 faces a “space which is not filled with solid substance” in the present embodiment allows deforming the deformable base plate 33 with a sufficiently-large displacement. The “space which is not filled with solid substance” referred to herein may be filled with fluid, including gas (such as air, nitrogen) and liquid. Alternatively, the space which is not filled with solid substance may be vacuumed.

More specifically, as illustrated in FIGS. 22A and 22B, the lower surface of the deformable base plate 33 (the surface of the deformable base plate 33 opposite to the surface on which the spin device elements 11 are formed) partially faces the space 38 which is not filled with solid substance, and the spin device elements 11 are disposed opposed to the space 38 across the deformable base plate 33. This structure is especially advantageous for increasing the displacement of the deformable base plate 33 and thereby increasing the strain generated in each spin device element 11. It should be noted that the portion of the upper surface of the deformable base plate 33 on which the spin device elements 11 are not formed may be covered with a dielectric film for protection.

As illustrated in FIG. 25C, a write current Iw for writing a desired data is generated which flows through the spin device elements 11 of the selected memory cell 1B in the state in which the deformable base plate 33 bent. Since the magnetostrictive effect is a uniaxial effect rather than a unidirectional effect, it is impossible to direct the magnetization direction of the recording layer 23 of each spin device element 11 (which corresponds to data “0” or “1) into the desired direction only with the magnetostrictive effect; the magnetostrictive effect exerted in the recording layer 23 only achieves an effect of tilting the magnetization of the recording layer 23 by about 90 degree from the direction of the magnetic anisotropy of the recording layer 23 at the maximum. The write current Iw corresponding to the desired data is fed to each spin device element 11 to limit the magnetization direction of the recording layer 23 to only one direction.

Illustrated in FIG. 25C is the operation for writing data “0”. The write current Iw is generated in such a direction that the magnetizations of the reference layers 21 and the recording layers 23 are directed in “parallel” in the spin device elements 11A1 and 11A2 and those of the reference layers 21 and the recording layers 23 are directed in “antiparallel” in the spin device elements 11B1 and 11B2. This allows placing the selected memory cell 1B into the first state in which the spin device elements 11A1 and 11A2 are placed in the “low resistance” state and the spin device elements 11B1 and 11B2 are placed in the “high resistance” state.

This is followed by stopping bending the deformable base plate 33. This completes the data writing. FIG. 25D illustrates the state of the selected memory cell 1B after the data writing is completed. In FIG. 25D, the magnetizations of the recording layers 23 of the spin device elements 11A1 and 11B1 are illustrated as being directed in the opposite directions to those in the original state (initial state) illustrated in FIG. 25A.

Since the data writing is assisted by the magnetostrictive effect caused by the strain applied to each spin device element 11, the above-described operation effectively achieves data writing with a reduced data write power, even if magnetic material exhibiting superior data stability is used as the recording layer 23 of each spin device element 11 in the memory cell 1B. In other words, the magnetic memory and data writing method in the present embodiment advantageously relieves a conflict between the data stability and data write power.

Although the mechanism illustrated in FIGS. 20 to 24 uses a piezoelectric effect to bend the deformable base plate 33, various other mechanisms may be used to bend the deformable base plate 33. For example, the deformable base plate 33 may be bent by using a force exerted between a pair of capacitor electrodes. FIGS. 26A and 26B are cross-sectional views illustrating the configuration of each block 2 in this case.

In the configuration illustrated in FIGS. 26A and 26B, a capacitor electrode 41 is partially embedded in the fixture base 32. The capacitor electrode 41 includes a flat plate section 41a and a contact section 41b. The lower surface of the flat plate 41a faces a space 38A which is not filled with solid substance.

The deformable base plate 33 is coupled on the upper surface of the fixture base 32. The deformable base plate 33 is opposed to the flat plate 41a of the capacitor electrode 41 across a space 38B which is not filled with solid substance. In other words, the lower surface of the deformable base plate 33 faces the space 38B which is not filled with solid substance.

The deformable base plate 33 includes a dielectric layer 42, a capacitor electrode layer 43 and a main body 44. The dielectric layer 42 is coupled with the upper surface of the fixture base 32 and the capacitor electrode layer 43 is coupled with the upper surface of the dielectric layer 42. The main body 44 is coupled with the upper surface of the capacitor electrode layer 43. The capacitor electrode layer 43 is opposed to the flat plate section 41a of the capacitor electrode 41 across the dielectric layer 42 and the space 38B which is not filled with solid substance, and a capacitor is formed with the capacitor electrode 41 and the capacitor electrode layer 43.

The lower electrodes 24A1, 24B1, 24A2 and 24B2 are formed on the upper surface of the deformable base plate 33 (that is, the upper surface of the main body 44) and the spin device elements 11A1, 11B1, 11A2 and 11B2 are formed on the upper surfaces of the lower electrodes 24A1, 24B1, 24A2 and 24B2, respectively. The structure of each spin device element 11 is as described above with reference to FIG. 4B. The upper electrodes 25A2, 25B2, 25A2 and 25B2 are formed on the upper surfaces of the recording layers 23 of the spin device elements 11A1, 11B2, 11A2 and 11B2, respectively.

This structure allows bending the deformable base plate 33 by applying a voltage between the capacitor electrode 41 and the capacitor electrode layer 43. In detail, when a voltage is applied between the capacitor electrode 41 and the capacitor electrode layer 43, an electric field is generated between the capacitor electrode 41 and the capacitor electrode layer 43, and this electric field works on charges generated on the surface of the capacitor electrode layer 43 to generate an attracting force which pulls the capacitor electrode layer 41 towards the capacitor electrode 43, that is, an attracting force which pulls the deformable base plate 33 towards the capacitor electrode 41. Since the lower surface of the deformable base plate 33 is only partially coupled with the fixture base 32 and faces the space 38B which is not filled with solid substance, the deformable base plate 33 is bent by the force attracting the deformable base plate 33 towards the capacitor electrode 41.

It should be noted here that, in the structure illustrated in FIGS. 26A and 26B, a part of the lower surface of the deformable base plate 33 faces the space 38B which is not filled with solid substance. The structure in which the surface of the deformable base plate 33 opposed to the surface on which the spin device elements 11 is formed faces the space 38B which is not filled with solid substance effectively enlarges the displacement of the deformable base plate 33.

It should be noted that substantially no current is generated (other than a leakage current and a temporary charging current) to bend the deformable base plate 33 also in the structure illustrated in FIGS. 26A and 26B. The structure illustrated in FIGS. 26A and 26B eliminates the need for feeding a current to bend the deformable base plate 33, since the deformable base plate 33 is bent with a force exerted between capacitor electrodes. This effectively reduces the data write power (that is, the power consumed in a write operation).

In an alternative embodiment, the memory cells 1B illustrated in FIGS. 11A and 11B, which each includes two spin device elements 11, may be integrated on the deformable base plate 33. FIG. 27 is a plan view illustrating the configuration of each block 2 in a magnetic memory thus structured and FIG. 28 is a cross-sectional view illustrating the structure of each block 2 on section F-F indicated in FIG. 27.

The lower electrodes 24A and 24B are formed on the upper surface of the deformable base plate 33 and the spin device elements 11A and 11B are formed on the upper surfaces of the lower electrodes 24A and 24B. The upper electrode 25 is coupled with the upper surfaces of the spin device elements 11A and 11B. The upper electrode 25 functions as a connection node which provides an electrical connection between the spin device elements 11A and 11B.

Also in the structure illustrated in FIGS. 27 and 28, data writing is achieved by feeding a write current through each of the spin device elements 11A and 11B of a selected memory cell 1A in the state in which the deformable base plate 33 is bent. More specifically, the deformable base plate 33 is bent when a desired data is written into the selected memory cell 1A. As described above, the deformable base plate 33 can be bent by applying electric fields to the piezoelectric layers 34 and 35 with the electrode layers coupled with the piezoelectric layers 34 and 35 (that is, the electrode layers 36, 37 and the additional electrode layers (not illustrated)). When the deformable base plate 33 is bent, a strain is generated in each of the spin device electrode 11 and the magnetization direction of the recording layer 23 of each spin device element 11 is tilted from the original magnetization direction of the recording layer 23, which is determined by the magnetic anisotropy thereof, due to a magnetostrictive effect. By tilting the magnetization direction of the recording layer 23 from the original magnetization direction, the recoding layer 23 is placed into a state in which the magnetization of the recoding layer 23 is easily reversible.

A write current is fed to the selected memory cell 1A in the state in which the deformable base plate 33 is bent. The direction of the write current is determined on the basis of the value of the data to be written into the selected memory cell 1A. This operation effectively allows writing the desired data into the selected memory cell 1A. This is followed by stopping bending the deformable base plate 33 to complete the data writing.

Also in the structure illustrated in FIGS. 27 and 28, the above-described operation effectively achieves data writing with a reduced data write power, since the data writing is assisted by the magnetostrictive effect caused by the strain generated in each spin device element 11. It should be noted that the mechanism which uses a force exerted between a pair of capacitor electrodes as illustrated in FIGS. 26A and 26B may be used as the bending mechanism to bend the deformable base plate 33.

Also in the structure in which the spin device elements 11 of the memory cells are formed on the deformable base plate 33, data writing into a selected memory may be achieved with a current-induced current. FIG. 29 is a perspective view illustrating the configuration of each block of the memory cell array in the case when data writing into a selected memory 1C is achieved with a current-induced magnetic field, and FIG. 30 is a plan view illustrating the structure of each block.

In the structure illustrated in FIGS. 29 and 30, a write current line 26 is provided close to each memory cell 1C. FIGS. 29 and 30 illustrate a structure in which the write current lines 26 are provided under the deformable base plate 33. The write current lines 26 are formed to extend in the X-axis direction. The spin device elements 11A1 and 11A2 of each memory cell 1C are positioned opposed to the spin device elements 11B2 and 11B2 across the write current line 26. More specifically, as illustrated in FIG. 30, the spin device elements 11A1 and 11A2 of each memory cell 1C are positioned shifted from the write current line 26 corresponding to each memory cell 1C in the +Y direction (a first direction), and the spin device elements 11B1 and 11B2 of each memory cell 1C are positioned shifted from the write current line 26 corresponding to each memory cell 1C in the −Y direction (a second direction opposite to the first direction). This arrangement preferably allows generating such a magnetic field that the magnetizations of the recoding layers 23 of the spin device elements 11A1 and 11A2 and those of the recoding layers 23 of the spin device elements 11B1 and 11B2 are directed in opposite directions with only one write current line 26. Note that the spin device element 11A2 is positioned shifted from the spin device element 11A1 in the +X direction, and the spin device element 11B1 is positioned shifted from the spin device element 11B2 in the +X direction.

Although electrical connections among the spin device elements 11A1, 11A2, 11B1 and 11B2 in each memory cell 1C are not illustrated in FIGS. 29 and 30, a person skilled in the art would appreciate that the spin device elements 11A1, 11A2, 11B1 and 11B2 may be electrically connected in a similar way to the configuration illustrated in FIG. 19.

Also in the configuration illustrated in FIGS. 29 and 30, data writing into a selected memory cell 1C can be achieved by applying a current-induced magnetic field to the recording layer 23 of each spin device element 11 of the selected memory cell 1C in the state in which the deformable base plate 33 is bent. In detail, the deformable base plate 33 is bent in data writing into the selected memory cell 1C. As described above, the deformable base plate 33 can be bent by applying electric fields to the piezoelectric layers 34 and 35 with the electrode layers coupled with the piezoelectric layers 34 and 35 (that is, the electrode layers 36, 37 and the additional electrode layers (not illustrated)). When the deformable base plate 33 is bent, a strain is generated in each of the spin device electrode 11 and the magnetization direction of the recording layer 23 of each spin device element 11 is tilted from the original magnetization direction of the recording layer 23, which is determined by the magnetic anisotropy thereof, due to a magnetostrictive effect. By tilting the magnetization direction of the recording layer 23 from the original magnetization direction, the recoding layer 23 is placed into a state in which the magnetization of the recoding layer 23 is easily reversible.

Furthermore, a write current Iw flowing through the write current line 26 corresponding to the selected memory cell 1C is generated in the state in which the deformable base plate 33 is bent. The direction of the write current Iw is determined depending on data to be written into the memory cell 1C. By the write current Iw flowing through the corresponding write current line 26, a current-induced magnetic field is applied to the recording layer 23 of each spin device element 11 of the selected memory cell 1C, and a desired data is thereby written into the selected memory cell 1C. This is followed by stopping bending the deformable base plate 33 to complete the data writing.

Also in the case when data writing is achieved by using a current-induced magnetic field, the above-described operation effectively achieves data writing with a reduced data write power, since the data writing is assisted by the magnetostrictive effect caused by the strain generated in each spin device element 11. It should be noted that the mechanism which uses a force exerted between a pair of capacitor electrodes as illustrated in FIGS. 26A and 26B may be used as the bending mechanism to bend the deformable base plate 33.

Although various specific embodiments of the present invention have been described in the above, the present invention must not be construed as being limited to the above-described embodiments. It would be apparent to a person skilled in the art that the present invention may be implemented with various modifications without departing from the scope of the invention.

Read more
PatSnap Solutions

Great research starts with great data.

Use the most comprehensive innovation intelligence platform to maximise ROI on research.

Learn More

Patent Valuation

$

Reveal the value <>

15.0/100 Score

Market Attractiveness

It shows from an IP point of view how many competitors are active and innovations are made in the different technical fields of the company. On a company level, the market attractiveness is often also an indicator of how diversified a company is. Here we look into the commercial relevance of the market.

70.0/100 Score

Market Coverage

It shows the sizes of the market that is covered with the IP and in how many countries the IP guarantees protection. It reflects a market size that is potentially addressable with the invented technology/formulation with a legal protection which also includes a freedom to operate. Here we look into the size of the impacted market.

73.95/100 Score

Technology Quality

It shows the degree of innovation that can be derived from a company’s IP. Here we look into ease of detection, ability to design around and significance of the patented feature to the product/service.

34.0/100 Score

Assignee Score

It takes the R&D behavior of the company itself into account that results in IP. During the invention phase, larger companies are considered to assign a higher R&D budget on a certain technology field, these companies have a better influence on their market, on what is marketable and what might lead to a standard.

22.08/100 Score

Legal Score

It shows the legal strength of IP in terms of its degree of protecting effect. Here we look into claim scope, claim breadth, claim quality, stability and priority.

Citation

Patents Cited in This Cited by
Title Current Assignee Application Date Publication Date
磁気メモリ用途のための応力支援による電流駆動式スイッチング グランディス インコーポレイテッド 12 November 2004 24 May 2007
Magnetic random access memory KABUSHIKI KAISHA TOSHIBA 19 January 2005 08 June 2006
Fully Integrated Tuneable Spin Torque Device For Generating An Oscillating Signal And Method For Tuning Such Apparatus IMEC 24 December 2004 13 December 2007
Planar Multiferroic/Magnetostrictive Nanostructures as Memory Elements, Two-Stage Logic Gates and Four-State Logic Elements for Information Processing VIRGINIA COMMONWEALTH UNIVERSITY 16 April 2012 25 October 2012
磁気メモリ素子、磁気メモリ、及び磁気記録再生装置 シャープ株式会社 23 January 2006 09 August 2007
See full citation <>

More like this

Title Current Assignee Application Date Publication Date
Magnetic memory TDK CORPORATION 06 March 2017 21 September 2017
Spin transfer torque structure for MRAM devices having a spin current injection capping layer SPIN TRANSFER TECHNOLOGIES, INC. 08 March 2016 27 October 2016
Bit cell with shared bit line QUALCOMM INCORPORATED 23 May 2016 22 December 2016
High speed, low power spin-orbit torque (SOT) assisted spin-transfer torque magnetic random access memory (STT-MRAM) bit cell array QUALCOMM INCORPORATED 25 July 2017 01 March 2018
磁気抵抗効果素子、および磁気ランダムアクセスメモリ 日本電気株式会社 18 March 2013 13 June 2013
Memory circuit provided with variable-resistance element TOHOKU UNIVERSITY 16 May 2016 24 November 2016
Magnetoresistance effect element and magnetic memory TOHOKU UNIVERSITY 21 January 2016 28 July 2016
增强的磁性存储器元及其实施方法 HGST荷兰公司 25 August 2016 15 March 2017
Perpendicular magnetic memory with reduced switching current INTEL CORPORATION 26 June 2015 29 December 2016
Switching skyrmions with VCMA/electric field for memory, computing, and information processing VIRGINIA COMMONWEALTH UNIVERSITY 01 March 2017 08 September 2017
Complementary magnetic tunnel junctions MTJ bit cell employing shared source line, and related methods QUALCOMM INCORPORATED 10 August 2016 02 March 2017
Low stray field magnetic memory INTEL CORPORATION 26 June 2015 29 December 2016
Magnetic tunnel junction element and magnetic memory TOHOKU UNIVERSITY 18 November 2016 26 May 2017
Magnetic memory device KOREA UNIVERSITY RESEARCH AND BUSINESS FOUNDATION 12 May 2016 17 November 2016
Perpendicular magnetic memory with filament conduction path INTEL CORPORATION 26 June 2015 29 December 2016
Magnetoresistive element and magnetic memory KABUSHIKI KAISHA TOSHIBA 12 November 2015 26 May 2016
Magnetic memory, method for writing data into magnetic memory, and semiconductor device BLUESPIN, INC. 30 July 2015 15 September 2016
Spin hall effect magnetic random access memory bitcell INTEL CORPORATION 24 September 2015 30 March 2017
Memory device and method for operating thereof AGENCY FOR SCIENCE,TECHNOLOGY AND RESEARCH 31 December 2015 21 July 2016
Magnetoresistive element, and magnetic random access memory NEC CORPORATION 28 October 2008 08 May 2012
See all similar patents <>

More Patents & Intellectual Property

PatSnap Solutions

PatSnap solutions are used by R&D teams, legal and IP professionals, those in business intelligence and strategic planning roles and by research staff at academic institutions globally.

PatSnap Solutions
Search & Analyze
The widest range of IP search tools makes getting the right answers and asking the right questions easier than ever. One click analysis extracts meaningful information on competitors and technology trends from IP data.
Business Intelligence
Gain powerful insights into future technology changes, market shifts and competitor strategies.
Workflow
Manage IP-related processes across multiple teams and departments with integrated collaboration and workflow tools.
Contact Sales
Clsoe
US10002655 Memory cell structure magnetic 1 US10002655 Memory cell structure magnetic 2 US10002655 Memory cell structure magnetic 3