Magnetic storage device having magnetoresistance element
By using the combination of rectangular recording layer shape and short-side magnetic field in the voltage writing system, the problem of high write error rate caused by thermal fluctuations is solved, and a high speed and low power consumption is realized.
Patent Information
- Application Number
- JP2022016669
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2022-02-04
- Publication Date
- 2025-04-18
- Estimated Expiration
- 2042-02-04
AI Technical Summary
In the voltage writing system, thermal fluctuations lead to unstable magnetization direction of the recording layer, resulting in high write error rates. The prior art methods such as increasing perpendicular magnetic facilitation and reverse bias can suppress thermal fluctuations, but increase power consumption.
The rectangular shape of the recording layer is adopted, and the effective magnetic field is applied to the short side direction. The effective magnetic field in the short side direction is applied when the voltage pulse is written is controlled to reduce the impact of thermal fluctuations on the magnetization direction, reduce the write error rate, and reduce power consumption.
Through the combination of the rectangular recording layer shape and the magnetic field in the short side direction, a high-speed and low-power write operation is realized, and the write error rate is effectively reduced.
Smart Images

Figure 0007664631000008 
Figure 0007664631000009 
Figure 0007664631000010
Abstract
Description
[Technical field]
[0001] The present invention relates to a magnetic storage device having a magnetoresistive element of a thin film stack including a magnetic thin film. [Background technology]
[0002] A magnetoresistive element is an element consisting of a laminate of a magnetization fixed layer, an intermediate layer, and a magnetization free layer. A magnetoresistive element has a characteristic that its resistance value changes depending on the relative angle of magnetization between the magnetization fixed layer and the magnetization free layer. By providing anisotropy to the magnetization, the magnetoresistive element can provide maximum and minimum energy values for the magnetization angle in the magnetization free layer. For example, when the magnetization of the magnetization fixed layer and the magnetization free layer has positive anisotropy on one axis, the magnetization energy becomes minimum when the relative angle of magnetization is parallel or antiparallel, and two stable states can be taken. The resistance value of the magnetoresistive element is lowest when the magnetization direction is parallel, and highest when it is antiparallel. When the magnetoresistive element is applied as a memory, the magnetization free layer is used as the recording layer.
[0003] In the voltage writing method, the magnetization direction of the magnetization free layer in the recording layer can be reversed by applying a voltage pulse to the magnetoresistance element. This phenomenon is called voltage-controlled magnetization reversal. Voltage-controlled magnetoresistance memory (VC-MRAM) is a non-volatile memory that uses this reversal as a data writing method. Voltage-controlled magnetization reversal can occur at a rate of several tens of kΩ. B By applying a voltage pulse with a duration of subnanoseconds to a few nanoseconds, the energy barrier between two magnetization states separated by an energy barrier of T can be lowered, making it easier to move between them, resulting in high-speed switching with extremely low power consumption. [Prior art documents] [Patent documents]
[0004] [Patent Document 1] US Patent Application Publication No. 2018 / 0158525 [Non-patent literature]
[0005] [Non-Patent Document 1] Y. Shiota, et al., Appl. Phys. Lett. 111, 022408 (2017) Summary of the Invention [Problem to be solved by the invention]
[0006] In the voltage writing method, one of the causes of writing errors is thermal fluctuations that the magnetization of the recording layer undergoes between the initial state before writing and the final state after writing, which causes the magnetization to deviate from the direction in the equilibrium state at absolute zero and vary in direction. In order to suppress the variation in the magnetization direction, it is considered to increase the perpendicular magnetic anisotropy of the recording layer, and a reverse bias method has been proposed (see, for example, Patent Document 1). In the reverse bias method, a bias voltage of the opposite polarity to the write voltage is applied to the magnetoresistance element before and after applying a write voltage for reversing the magnetization of the recording layer. When the bias voltage is applied, the perpendicular magnetic anisotropy energy increases and acts in the perpendicular direction to the film surface, making the thermal energy relatively small, and suppressing the variation in the magnetization direction. However, in the reverse bias method, a voltage is applied to the magnetoresistance element from before writing to after writing, which causes an undesirable phenomenon of increased power consumption.
[0007] Another cause of write errors is thermal disturbance of the magnetization of the recording layer, which undergoes precessional reversal during writing, which may result in a final state in which the direction of magnetization does not reverse. It has been reported that the write error rate can be reduced by increasing the external magnetic field applied parallel to the film surface of the recording layer in order to suppress the effects of thermal disturbance (see, for example, Non-Patent Document 1). However, with this method, the stronger the external magnetic field, the shorter the voltage pulse width is required, allowing for high-speed, low-power writing; however, if the external magnetic field is excessively large compared to the perpendicular magnetic anisotropy energy, an undesirable phenomenon occurs in which the write error rate cannot be sufficiently reduced.
[0008] An object of the present invention is to provide a magnetic storage device that is capable of writing data at high speed with low power consumption and that can reduce the write error rate. [Means for solving the problem]
[0009] According to one aspect of the present invention, there is provided a magnetic storage device comprising: a magnetoresistive element having a layered structure of a recording layer including a first magnetic layer, a tunnel barrier layer, and a reference layer including a second magnetic layer; and a control unit; wherein the in-plane shape of the recording layer is a rectangle that circumscribes the in-plane shape and has a smallest area, the rectangle having short and long sides of different lengths; and when writing to the recording layer, the control unit applies a voltage pulse to the magnetoresistive element while applying an effective magnetic field in the direction of the short side in the plane of the recording layer.
[0010] According to the above aspect, the recording layer of the magnetoresistive element has a shape in which a rectangle circumscribing its in-plane shape and having the smallest area has short and long sides of different lengths, and by applying an effective magnetic field in the direction of the short side in the plane of the recording layer while the control unit applies a voltage pulse to the magnetoresistive element, high speed writing with low power consumption is possible, and the shape magnetic anisotropy of the recording layer suppresses thermal fluctuations of the magnetization of the recording layer before and after writing, and suppresses the effects of thermal disturbance during writing, thereby reducing the write error rate. [Brief description of the drawings]
[0011] [Figure 1] 1 is a schematic configuration diagram of a main part of a magnetic storage device according to a first embodiment. [Diagram 2] FIG. 2 is a plan view of a recording layer of a magnetoresistive element, illustrating an in-plane shape. [Diagram 3] 4A and 4B are diagrams illustrating the behavior of magnetization in a recording layer of a magnetoresistive element. [Figure 4] FIG. 11 is an explanatory diagram of a sequence of a write model for applying a voltage pulse. [Diagram 5] 11 is a diagram showing an example of calculation of the distribution of the initial state of magnetization of the recording layer of a magnetoresistive element; FIG. [Figure 6]FIG. 4 is a diagram showing the energy density in the initial state of magnetization of the recording layer of the magnetoresistive element. [Figure 7] FIG. 11 is a diagram (part 1) showing a calculation example of a write error rate of the magnetic memory device according to the first embodiment. [Figure 8] FIG. 4 is a diagram showing the relationship between the time width of a write voltage pulse and a write error rate in the first embodiment. [Figure 9] FIG. 2 is a diagram (part 2) showing a calculation example of a write error rate of the magnetic memory device according to the first embodiment. [Figure 10] FIG. 11 is a diagram showing the relationship between the time width of a write voltage pulse and a write error rate in the second embodiment. [Figure 11] 1 is a schematic diagram of a magnetic storage device according to an embodiment of the present invention; [Figure 12] FIG. 1 is an explanatory diagram (part 1) of an operation example of a magnetic storage device according to an embodiment of the present invention. [Figure 13] FIG. 2 is an explanatory diagram (part 2) of an operation example of the magnetic storage device according to one embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0012] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Note that elements common to multiple drawings are given the same reference numerals, and detailed descriptions of those elements will be omitted.
[0013] [First embodiment] Fig. 1 is a schematic diagram of a main part of a magnetic storage device according to a first embodiment. Fig. 2 is a plan view of a recording layer of a magnetoresistive element, and is a diagram for explaining an in-plane shape. The XY plane in Figs. 1 and 2 is a plane parallel to the in-plane shape of the recording layer. The X axis is the long axis direction of the elliptical in-plane shape of the recording layer, the Y axis is the short axis direction of the elliptical in-plane shape of the recording layer, and the Z axis is a direction perpendicular to the recording layer.
[0014] 1 and 2, a magnetic memory device 10 has a magnetoresistive element 11 that stores information by utilizing two stable magnetization states, and a control unit 12 that writes information and controls reading by applying a voltage pulse to the magnetoresistive element 11. The magnetoresistive element 11 has a recording layer 13, a reference layer 15, and a tunnel barrier layer 14 sandwiched between the recording layer 13 and the reference layer 15.
[0015] The magnetic storage device 10 of this embodiment is different from conventional magnetic field-based writing method so-called toggle magnetoresistance memory (toggle MRAM) and current-based writing method so-called spin torque writing magnetic memory (STT-RAM), and is a voltage-controlled magnetic memory (VC-MRAM) that applies a voltage pulse to the magnetoresistance element 11 to reverse the direction of magnetization of the recording layer 13. The magnetic storage device 10 controls the height of the energy barrier existing between two stable magnetization states of the recording layer 13 by a short voltage pulse on the order of subnanoseconds to nanoseconds to perform magnetization reversal with little Joule loss, thereby enabling high-speed writing with low power consumption.
[0016] The magnetic storage device 10 applies an effective magnetic field in a predetermined direction in the plane of the recording layer 13, for example, during writing in order to reverse the magnetization due to voltage. The effective magnetic field may be applied not only during writing but also during non-writing. The predetermined direction of the effective magnetic field is the short side direction (Y-axis direction) of the in-plane shape. The short side direction will be described below. The in-plane shape of the recording layer 13 shown in FIG. 2 is uniquely determined as a rectangle RTG that circumscribes the in-plane shape (contour) and has the smallest area. The in-plane shape of the recording layer 13 is a shape having short sides SS and long sides LS of different lengths, and does not include the case where the short sides SS and long sides LS are the same length, that is, the case where the rectangle RTG is a square. In this way, the effective magnetic field is applied in the short side direction (SS direction) of the rectangle RTG specified as described above for the in-plane shape of the recording layer 13.
[0017] If the magnetoresistance element 11 shown in Fig. 1 has a three-dimensional shape of an elliptical cylinder, the in-plane shape of the recording layer 13 will be an ellipse, with the short axis of the ellipse corresponding to the short side SS of the rectangle RTG and the long axis of the ellipse corresponding to the long side LS, as shown in Fig. 2. Even if the in-plane shape of the recording layer 13 is deformed due to influences during the manufacturing process, etc., the effective magnetic field is applied in the direction of the short side of the rectangle specified as above.
[0018] The longer the length of the short side SS, the larger the volume of the recording layer 13 and the greater the stability of memory, which is preferable, but the greater the element area, the fewer the number of elements that can be arranged on one substrate, which is undesirable because it reduces the memory capacity per unit area. For this reason, the length of the short side SS is preferably 5 nm or more and 500 nm or less, and more preferably 10 nm or more and 300 nm or less.
[0019] The effective magnetic field is the magnetization m of the recording layer 13. rec This is part of the magnetic field acting on the magnetoresistive element 11, and is a composite magnetic field including an external magnetic field applied from outside the magnetoresistive element 11, a magnetic field due to induced magnetic anisotropy of the recording layer 13, a magnetic field due to interlayer exchange coupling, an exchange bias magnetic field, and a leakage magnetic field from magnetic layers other than the recording layer 13.
[0020] The external magnetic field applied from outside the magnetoresistance element 11 may be a static magnetic field of a permanent magnet placed near the magnetoresistance element 11. Alternatively, a magnetic field generated by passing a direct current through a wire placed near the magnetoresistance element 11 may be used.
[0021] The induced magnetic anisotropy of the recording layer 13 is a magnetic anisotropy formed when the recording layer 13 is formed in a magnetic field or heat-treated in a magnetic field. When the recording layer 13 is made of a material in which at least one element selected from the group consisting of B, C, N, Si, P, and Cr is added to Fe, Co, Ni, or an alloy of these elements, induced magnetic anisotropy occurs when the recording layer 13 is formed in a magnetic field or heat-treated in a magnetic field. An internal magnetic field is generated in the recording layer 13 due to the induced magnetic anisotropy.
[0022] Interlayer exchange coupling is a phenomenon in which a magnetic field is generated between the magnetization of the magnetic fixed layer and the magnetization of the recording layer 13, which tend to become antiparallel or parallel to each other, by providing a magnetic fixed layer via a nonmagnetic layer on the opposite side of the recording layer 13 from the reference layer 15. The magnetic field due to interlayer exchange coupling is a magnetic field generated in the recording layer 13 by this action.
[0023] The exchange bias magnetic field is a magnetic field that the recording layer 13 receives from a bias layer provided in contact with the opposite side of the recording layer 13 from the reference layer 15. The bias layer is made of an alloy containing at least one element selected from the group consisting of Cr, Mn, Fe, Co, and Ni, and at least one element selected from the group consisting of Ru, Rh, Pd, Ag, Os, Ir, Pt, and Au. This alloy is preferably antiferromagnetic. Examples of alloys that constitute the bias layer include a Pt-Mn alloy, an Ir-Mn alloy, and an Fe-Mn alloy.
[0024] The leakage magnetic field from a magnetic layer other than the recording layer 13 is a magnetic field that leaks from a magnetic layer other than the recording layer 13. As a part of the laminate of the magnetoresistive element 11, a magnetic layer for generating a leakage magnetic field may be provided.
[0025] Magnetization m of the recording layer 13 rec In addition to the effective magnetic field, a perpendicular anisotropy magnetic field and a magnetic field due to shape magnetic anisotropy also act on the recording layer 13. The magnetic field due to shape magnetic anisotropy is a magnetic field caused by the in-plane shape of the recording layer 13, that is, a demagnetizing field. The in-plane shape of the recording layer 13 is a rectangle RTG that circumscribes the in-plane shape and has the smallest area, and has short sides SS and long sides LS of different lengths. In the case of such a shape, when the recording layer 13 is magnetized, a demagnetizing field, which is an internal magnetic field, is generated in the direction opposite to the magnetization.
[0026] The recording layer 13 and the reference layer 15 of the magnetoresistance element 11 are magnetic layers. The magnetic layer contains at least one element selected from the group consisting of Fe, Co, Ni, Cr, Mn, Gd, Nd, Sm, and Tb. The thickness of the magnetic layer is preferably set in the range of, for example, 0.5 nm to 10 nm.
[0027] The recording layer 13 has a magnetization m ref Compared to magnetization m rec The reference layer 15 includes a magnetic layer (referred to as a magnetization free layer) whose magnetization direction is easily changed. The reference layer 15 includes a magnetization fixed layer whose magnetization direction is difficult to change.
[0028] The recording layer 13 preferably has perpendicular magnetic anisotropy. When no voltage pulse is applied, the magnetization of the recording layer 13 is oriented in a direction inclined perpendicular to the film surface due to competition between the perpendicular magnetic anisotropy and the effective magnetic field.
[0029] The recording layer 13 is preferably a magnetic layer containing Co, for example, Co--Fe--B, since the magnetoresistance effect is relatively larger than that of other magnetic materials.
[0030] A laminate in which a magnetic layer containing Co and a magnetic layer containing at least one of Pt and Pd are laminated can be used for the recording layer 13. A laminate in which a magnetic layer containing Co and a magnetic layer containing at least one of Pt and Pd are alternately laminated can be used for the recording layer 13.
[0031] The reference layer 15 may have a ferromagnetic layer / intermediate layer / ferromagnetic layer of a synthetic ferrimagnetic (SAF) structure. The SAF structure is a structure in which two ferromagnetic layers are antiferromagnetically coupled and the magnetizations of each layer are antiparallel. The intermediate layer is a thin conductive layer. The magnetization direction of the reference layer 15 is fixed in the same direction with almost no time change. When the magnetization direction coincides with the stacking direction, it is called a perpendicular magnetization film, and when the magnetization direction is perpendicular to the stacking direction, it is called an in-plane magnetization film. The reference layer 15 is, for example, a perpendicular magnetization stacked film containing Co-Fe-B / Ru layer / CoPt in that order from the tunnel barrier layer 14.
[0032] The tunnel barrier layer 14 is disposed between the recording layer 13 and the reference layer 15. The tunnel barrier layer 14 can be any of a non-magnetic material known as a tunnel barrier layer, such as an insulator, a semiconductor, or a dielectric. The tunnel barrier layer includes, for example, an oxide, a nitride, or a fluoride containing at least one element selected from the group consisting of Mg, Si, Al, Ti, Zr, Hf, Ta, Zn, Sr, Bi, and Ba. Specifically, for example, Al 2 O 3 , SiO 2 , MgO, AlN, Ta-O, Al-Zr-O, Bi 2 O 3 , MgF 2 , CaF 2 , SrTiO 3 , AlLaO 3 , Al—NO, Si—NO, etc. The tunnel barrier layer 14 may be made of, for example, a nonmagnetic semiconductor (ZnOx, InMn, GaN, GaAs, TiOx, Zn, Te, or any of these doped with a transition metal).
[0033] The tunnel barrier layer 14 is preferably an oxide containing Mg, since the magnetoresistance effect is relatively larger than that of other magnetic materials, and is preferably an oxide containing Mg and Al, since the magnetoresistance effect is even larger.
[0034] The thickness of the tunnel barrier layer 14 is preferably 1.2 nm or more in terms of suppressing Joule loss, more preferably 1.3 nm or more, even more preferably 1.4 nm or more, even more preferably 1.5 nm or more, and particularly preferably 1.6 nm or more. The thickness of the tunnel barrier layer 14 is preferably 4 nm or less in terms of high-speed readout.
[0035] A simulation was performed on the magnetic storage device 10 of this embodiment. The simulation was performed on the magnetic storage device 10 of this embodiment. x ,m y ,m zThe Langevin equation (1) below, which expresses the time evolution of
number
number
[0036] Composite magnetic field H rec is defined by the following equation (3).
number
number
[0037] FIG. 3 is an explanatory diagram of the behavior of magnetization of the recording layer of the magnetoresistive element. The x direction is the long axis (long side) direction of the recording layer 13, and the y direction is the short axis (short side) direction of the recording layer 13. Magnetization m is a unit vector indicating the direction of magnetization. Referring to FIG. 3 together with FIG. 1, the effective magnetic field H eff is applied in the y-axis direction, and the magnetization m of the recording layer has a positive z-component (m z >0). As shown in FIG. 3, when the magnetization direction of the recording layer 13 is expressed in a spherical coordinate system, it is tilted in the latitudinal direction at an angle θ from the z-axis direction. The torque acting during the application of a voltage pulse is indicated by an arrow starting from the tip of the arrow of the magnetization m. The magnetization m has a damping torque due to the composite magnetic field H rec , and the precession torque is generated by the resultant magnetic field H rec This occurs in a direction that rotates the magnetization around the axis.
[0038] The sequence of a write model for applying a voltage pulse to the magnetoresistance element 11 was set as follows: The write error rate of this write model was calculated by simulation using the Langevin equation of the above formula (1).
[0039] 4 is an explanatory diagram of a write model sequence for applying a voltage pulse. The horizontal axis of (a) and (b) represents time. The vertical axis of (a) represents the voltage applied to the magnetoresistive element 11. The vertical axis of (b) represents the effective perpendicular magnetic anisotropy constant K acting on the recording layer 13. eff represents K eff =K u -(1 / 2)μ 0 M s 2 (N z -Nx ) was decided.
[0040] First, when the temperature is 0 K (Kelvin), the voltage applied to the magnetoresistance element 11 is set to zero, m z 4(a), the magnetoresistive element 11 is relaxed for 10 nanoseconds (ns) in state 1 where no voltage is applied to the magnetoresistive element 11 at room temperature (300K), and then a voltage pulse (time width t p , applied voltage V 1 ) is applied (state 2), and then the material is allowed to relax for 10 nanoseconds (ns) in state 3 where no applied voltage is applied. As shown in (b), the effective perpendicular magnetic anisotropy constant K eff is K eff (0) , the effective perpendicular magnetic anisotropy constant K in state 2 eff is K eff (+V 1 ) The magnetization m at the end of the relaxation period of 10 ns in state 3 is m z If m is less than 0, it is determined that magnetization reversal has occurred. z If the error rate is >0, the magnetization reversal does not occur, so it is determined to be a write error. The write error rate is defined as (error frequency) / number of trials. In this calculation example, the number of trials is 10 5 ~10 6 The number of times was 100.
[0041] In the simulation of the magnetic memory device according to the first embodiment, the following various parameters were applied to the magnetoresistance element 11 and the write conditions. -Recording layer thickness: 1.1 nm -Magnetic resistance element junction area: 289πnm 2 (=17 2 π nm 2 ) -In-plane shape of recording layer, aspect ratio (Aspect Ratio (AR) = major axis / minor axis): Example 1: Circular, AR=1 (as a comparison example), Example 2: Elliptical, AR=2, Example 3: Ellipse, AR=3, Example 4: Ellipse, AR=4, Example 5: Ellipse, AR=5 - Demagnetization factor (N x ,N y ,N z ): Example 1 (0.04447, 0.04447, 0.91106), Example 2 (0.02608, 0.06836, 0.90556), Example 3 (0.01817, 0.08445, 0.89738), Example 4 (0.013778, 0.096766, 0.889456), Example 5 (0.0110, 0.1069, 0.8821) - Saturation magnetization M of the recording layer s :0.955×10 6 A / m -Magnetic friction constant (Gilbert damping constant) α=0.1 - Effective magnetic field H eff (Recording layer in-plane minor axis direction): 200 Oe to 1600 Oe - Effective perpendicular magnetic anisotropy constant K eff : When the applied voltage is zero, K eff (0) =110kJ / m 3 , Voltage pulse (applied voltage V 1 ) applied, K eff (+V 1 ) =-200~50kJ / m 3 -Time width of the voltage pulse t p : 0.01ns~1.00ns
[0042] FIG. 5 is a diagram showing a calculation example of the distribution of the initial state of magnetization of the recording layer of the magnetoresistive element. FIG. 5 shows the distribution of magnetization of the recording layer 13 after passing through state 1 shown in FIG. 4. The distribution of magnetization directions was obtained by performing a simulation up to relaxation in state 1 100,000 times and calculating the magnetization direction after passing through state 1 each time. K eff (0) =110kJ / m 3 , H eff= 800 Oe. (a) of FIG. 5 shows the effective magnetic field H eff When the in-plane shape is the example 3, (b) shows the effective magnetic field H eff When the magnetic field is applied in the long axis direction in the plane, (c) shows the effective magnetic field H eff The upper spherical diagrams in (a) to (c) show the position of the tip of magnetization m on a sphere. In (a) and (b), the depth direction is the x-component of m (m x ), the horizontal axis is the y component of m (m y ), and the vertical axis is the z component of m (m z In (c), the y component of the depth direction is m (m y ), the horizontal axis is the x-component of m (m x ), and the vertical axis is the z component of m (m z The lower figures in (a) to (c) represent m z The horizontal axis is the z-component of m (m z ) and the vertical axis represents the count value.
[0043] Referring to FIG. 5(a), when the in-plane shape of the recording layer 13 is a circle (Example 1), the spherical diagram and histogram show m z It can be seen that there exists a state where the z component m of magnetization m is negative and there are trials in which the z component m is reversed within the relaxation time before the voltage pulse is applied. z The standard deviation of is 0.0800. As shown in FIG. 5(b), even if the recording layer has an in-plane shape of AR=3 (Example 3), the effective magnetic field H eff If is applied in the major axis direction, m z It can be seen that there exists a negative state, and there are attempts to reverse it within the relaxation time before the voltage pulse is applied. z In contrast, when the recording layer has an AR=3 (the in-plane shape of Example 3 and the effective magnetic field H eff If is applied in the minor axis direction, m z There is no negative state. The z component m of magnetization m zThe standard deviation of is 0.0389, which is much smaller than that of Example 1. Furthermore, it is found that the variation is much smaller than that of Example 3, in which the effective magnetic field is applied in the in-plane major axis direction. This shows that the in-plane shape of the recording layer 13 is elliptical and the effective magnetic field H eff It was found that the spread of the distribution of the initial state of magnetization of the recording layer 13 can be suppressed by applying a magnetic field in the in-plane minor axis direction.
[0044] 6 is a diagram showing the energy density in the initial state of magnetization of the recording layer of the magnetoresistive element. (a) shows the effective magnetic field H eff (b) shows the effective magnetic field H eff The horizontal axis shows the z-component m of the magnetization of the recording layer. z The vertical axis indicates the energy density of the recording layer 13. In (a), the effective magnetic field H eff is in the x direction, and in the initial state, magnetization m is unlikely to be oriented in the y direction, so m y = 0, and the energy density E in equation (4) is m z In (b), the effective magnetic field H eff is in the y direction, and in the initial state, magnetization m is unlikely to be oriented in the x direction, so m x = 0, and the energy density E in equation (4) is m z It is plotted as a dependency.
[0045] Referring to FIG. 6(a), m z The energy density has a minimum value near -0.9 and 0.9, and m z The energy density reaches its maximum value near 0. The difference between the maximum and minimum values is 47 kJ / m 3 Referring to FIG. 6(b), m z The energy density has a minimum value near -0.95 and 0.95, and m z The energy density reaches its maximum value near 0. The difference between the maximum and minimum values is 81 kJ / m 3 From these facts, it is assumed that the in-plane shape of the recording layer 13 is elliptical and the effective magnetic field Heff When the magnetic field is applied in the in-plane short axis direction, the magnetic field is larger than that when the in-plane shape of the recording layer 13 is circular and the magnetic field is applied in the in-plane direction due to the shape magnetic anisotropy. z Since the energy barrier at which the energy becomes maximum near 0 is high, the spread of the distribution of the initial state of magnetization of the recording layer 13 can be suppressed.
[0046] 7 is a diagram (part 1) showing a calculation example of the write error rate of the magnetic storage device according to one embodiment of the present invention. The horizontal axis represents the effective magnetic field H eff The vertical axis is the minimum write error rate (WER) min In each of the combinations of the in-plane shape and the effective magnetic field in Examples 1 to 5, the voltage pulse duration t p Dependency K eff (+V 1 ) =-200~50kJ / m 3 The minimum write error rate (WER) is calculated in the range of min For Examples 2 to 5, the index (M s (N y -N x ) are 510, 800, 1000, and 1150 Oe, respectively. The effective perpendicular magnetic field due to the perpendicular magnetic anisotropy is H k eff (=2K eff / (μ 0 M s )) then 0.2 × H k eff The value of 0.2×H is 460 Oe. k eff <M s (N y -N x ) is the case.
[0047] 7, the minimum value of the write error rate is smaller in Examples 2 to 5 than in Example 1. That is, when the in-plane shape of the recording layer 13 is elliptical, the effective magnetic field H effBy applying L2 / L1, the write error rate can be reduced more than in the case of a circular shape. Furthermore, it can be seen that the minimum value of the write error rate decreases as the aspect ratio increases from Example 1 to Example 5. For this reason, a larger aspect ratio is preferable. On the other hand, if the aspect ratio is too large, the recording layer is likely to enter a multi-domain state, which impairs recording stability. Therefore, the aspect ratio, i.e., L2 / L1 in FIG. 2, is preferably in the range greater than 1 and equal to or less than 5, more preferably in the range of 1.5 to 4, and even more preferably in the range of 2 to 3.5. Furthermore, from the calculation results of Examples 2 to 5 shown in FIG. 7, it can be seen that the effective magnetic field H eff is 0.2×H k eff <|H eff | <M s (N y -N x ) is preferably satisfied.
[0048] FIG. 8 is a diagram showing the relationship between the time width of the write voltage pulse and the write error rate. eff (+V 1 ) is 0 kJ / m 3 8(b) shows the case of Example 3. eff (+V 1 ) -60kJ / m 3 The horizontal axis represents the time width of the voltage pulse, and the vertical axis represents the write error rate.
[0049] Referring to FIG. 8(a), in the case of the circular recording layer of Example 1, the write error rate was minimized (2.6×10 -4 8(b), in the case of the elliptical recording layer 13 of Example 3, when an external magnetic field is applied in the minor axis direction in the plane, the write error rate is minimized (3.2×10 -5) It can be seen that the write error rate in Example 3 is one order of magnitude lower than in Example 1. Also, the time width of the write voltage pulse at which the write error rate is minimized in Example 3 is about 50% of that in Example 1, demonstrating that high-speed writing is possible.
[0050] 9 is a diagram (part 2) showing a calculation example of the write error rate of the magnetic storage device according to the embodiment of the present invention. The horizontal axis represents the effective magnetic field H eff The vertical axis is the minimum write error rate (WER) min FIG. 9 shows the index (M s (N y -N x )) is the effective perpendicular magnetic field H due to perpendicular magnetic anisotropy. k eff About 0.2×H k eff is smaller than M s (N y -N x )<0.2×H k eff The simulation was performed with the thickness of the recording layer set to 0.5 nm and the junction area of the magnetoresistive element set to 900πnm. 2 The in-plane shape of the recording layer was an ellipse with an aspect ratio of 2, which was similar to that of Example 2. s (N y -N x ) is 180 Oe, and 0.2 × H k eff is 460 Oe.
[0051] Referring to FIG. 9, the write error rate is proportional to the effective magnetic field H eff It can be seen that the minimum value is around 450 Oe to 520 Oe. s (N y -N x )<0.2×H k eff In the case of , the effective magnetic field H eff is 0.2×H k eff That is, M s (N y -N x )<|Heff |≒0.2×H k eff It is preferable to satisfy the following relationship.
[0052] [Second embodiment] The magnetic memory device according to the second embodiment includes a magnetoresistive element that can obtain a low write error rate even if the precision of the time width of the write voltage pulse is lower than that of the first embodiment. The schematic configuration of the main part of the magnetic memory device according to the second embodiment and a plan view of the recording layer of the magnetoresistive element are the same as those of the first embodiment in Figs. 1 and 2, respectively, so the same reference numerals are used for the components, and the illustration and description thereof are omitted.
[0053] The magnetoresistance element 11 of the magnetic storage device according to the second embodiment is a case where the magnetic friction constant (Gilbert damping constant) α of the recording layer 13 is relatively large. In the magnetoresistance element 11 including the recording layer 13 with a large Gilbert damping constant α, the write error rate can be reduced by suppressing thermal fluctuation in the initial state. In addition, since the write error rate is low over a long range of the time width of the write voltage pulse, there is an effect that a highly accurate time width is not required for the write voltage pulse.
[0054] In the simulation of the magnetic memory device according to the second embodiment, the following various parameters were applied to the magnetoresistance element 11 and the write conditions. -Recording layer thickness: 2 nm -Magnetic resistance element junction area: 19600πnm 2 (=140 2 π nm 2 ) -In-plane shape of recording layer, aspect ratio (Aspect Ratio (AR) = major axis / minor axis): Example 6: Circular, AR=1 (as a comparative example), Example 7: Elliptical, AR=3 (second embodiment) - Demagnetization factor (N x ,N y ,N z ): Example 6 (0.01325,0.01325,0.97350), Example 7 (0.00535, 0.02574, 0.96891) - Saturation magnetization M of the recording layer s :1.400×10 6 A / m -Magnetic friction constant (Gilbert damping constant) α: Example 6: α=0.17, Example 7: α=0.20 - Effective magnetic field H eff (Recording layer in-plane minor axis direction): 400 Oe - Effective perpendicular magnetic anisotropy constant K eff : When the applied voltage is zero, K eff (0) =70kJ / m 3 , Voltage pulse (applied voltage V 1 ) applied, K eff (+V 1 ) : Example 6:K eff (+V 1 ) =33kJ / m 3 , Example 7:K eff (+V 1 ) =10kJ / m 3 -Time width of the voltage pulse t p : 0.5ns~10.0ns
[0055] 10 is a diagram showing the relationship between the time width of the write voltage pulse and the write error rate in the second embodiment. Referring to FIG. 10, in both Example 6 and Example 7, the time width t p The write error rate is low over a long range of 1.0 ns to 10.0 ns. This is different from the cases of Examples 1 and 3 shown in FIG. pThis writing method does not require high-precision control of . This type of writing technique is called damped writing. In damped writing, in order to relax to the state after writing in a short time by using magnetic damping (friction of magnetization motion), it is preferable that α is 0.05 or more and 0.5 or less. In order to obtain this relatively large α, it is preferable that the recording layer 13 contains one or more of the 5d transition metals Hf, Ta, W, Os, Ir, Pt, and Au.
[0056] Comparing Example 6 and Example 7, the in-plane shape of the recording layer in Example 6 is circular and the effective magnetic field H eff The in-plane shape of the recording layer 13 in Example 7 is elliptical, and the effective magnetic field H eff The write error rate is about two orders of magnitude lower when the magnetic field is applied in the in-plane minor axis direction of the recording layer. In Example 7, the write error rate is reduced more than in Example 6 by using the same effect as in the first embodiment. Therefore, even in the braking write, 0.2×H k eff <M s (N y -N x ), the effective magnetic field H eff is 0.2×H k eff <|H eff | <M s (N y -N x ) is preferably satisfied. s (N y -N x )<0.2×H k eff In the case of , the effective magnetic field H eff is 0.2×H k eff That is, M s (N y -N x )<|H eff |≒0.2×H k eff It is preferable that the following relationship is satisfied.
[0057] In Example 7, K eff (0) =70kJ / m 3 , K eff(+V 1 ) =33kJ / m 3 However, when the in-plane shape of the recording layer 13 is an ellipse, the following relationship is satisfied. eff (0) , K eff (+V 1 ) is preferred for causing braking writing.
[0058] (1)h eff ≦N y -N x in the case of,
number
number
number
[0059] [Third embodiment] In the magnetic memory device according to the third embodiment, the effective magnetic field applied to the recording layer 13 of the magnetoresistive element 11 has a component perpendicular to the film surface, i.e., a component perpendicular to the surface, in addition to the short side direction in the plane of the recording layer 13 (the minor axis direction when the recording layer 13 is elliptical).The schematic configuration of the main parts of the magnetic memory device according to the third embodiment and a plan view of the recording layer of the magnetoresistive element are respectively similar to Figures 1 and 2 of the first embodiment, so the same reference numerals are used for the components and the illustration and description thereof are omitted.
[0060] In the simulation of the magnetic memory device according to the third embodiment, the following various parameters were applied to the magnetoresistance element 11 and the write conditions. -Recording layer thickness: 1.1 nm -Magnetic resistance element junction area: 289πnm 2 (=17 2 π nm 2 ) -In-plane shape of recording layer, aspect ratio (Aspect Ratio (AR) = major axis / minor axis): Example 8: Circular, AR=1 (as a comparison example), Example 9: Ellipse, AR=3 - Demagnetization factor (N x ,N y ,N z ): Example 8 (0.04447, 0.04447, 0.91106), Example 9 (0.01817, 0.08445, 0.89738) - Saturation magnetization M of the recording layer s :0.955×10 6 A / m -Magnetic friction constant (Gilbert damping constant) α=0.1 - Effective magnetic field H eff : Component perpendicular to the recording layer (-z): 100 Oe, Component along the minor axis of the recording layer: Example 8: 400Oe, Example 9: 800Oe - Effective perpendicular magnetic anisotropy constant K eff : When the applied voltage is zero, K eff (0) =110kJ / m 3 , Voltage pulse (applied voltage V 1 ) applied, K eff (+V 1 ) : Example 8, magnetization reversal from +z to -z direction: K eff (+V 1 ) =38kJ / m 3 Example 9, magnetization reversal from +z to -z: Keff (+V 1 ) =26kJ / m 3 Example 8, magnetization reversal from -z direction to +z direction: K eff (+V 1 ) =-10kJ / m 3 Example 9, magnetization reversal from -z direction to +z direction: K eff (+V 1 ) =-80kJ / m 3 -Time width of the voltage pulse t p : Example 8 and Example 9, magnetization reversal from +z to -z direction: t p =10ns Example 8, magnetization reversal from -z direction to +z direction: t p =0.37ns Example 9, magnetization reversal from -z direction to +z direction: t p =0.18ns Hereinafter, the magnetization reversal direction from the +z direction to the −z direction will be referred to as the first magnetization reversal direction, and the magnetization reversal direction from the −z direction to the +z direction will be referred to as the second magnetization reversal direction. In this embodiment, the effective magnetic field H eff A form in which the perpendicular component of is set in the -z direction will be described.
[0061] In the writing where the first magnetization reversal direction occurs, the effective magnetic field H eff The magnetization direction in the final state is stabilized by the −z-direction component, and damped writing occurs. The WER is 0.0378 in Example 8 and 0.0542 in Example 9, and the WERs are comparable.
[0062] On the other hand, in the writing where the second magnetization reversal direction occurs, the effective magnetic field H eff The -z component of the magnetization direction of the final state is destabilized. In Example 8, K eff (+V 1 ) =-10kJ / m 3 , t p =0.37ns (WER) min= 0.00633. The effective magnetic field H eff (WER) in Example 1, which does not include the -z component min =0.00026, which is about 24 times higher (WER) min In contrast, in Example 9, K eff (+V 1 ) =-80kJ / m 3 , t p =0.18ns (WER) min = 0.00014. The effective magnetic field H eff (WER) in Example 3, which does not include a component in the -z direction min =0.000032, (WER) min Although the number of cases will increase, the increase will be only about 4.3 times.
[0063] In this embodiment, the effective magnetic field H eff The perpendicular component of may be set in the +z direction, in which case the magnetization reversal direction from the -z direction to the +z direction becomes the first magnetization reversal direction, and the magnetization reversal direction from the +z direction to the -z direction becomes the second magnetization reversal direction, thereby achieving the action and effect of the present embodiment described above.
[0064] FIG. 11 is a schematic diagram of a magnetic storage device according to an embodiment of the present invention. Referring to FIG. 11 together with FIG. 1, the magnetic storage device 100 includes a plurality of magnetoresistance elements 110 having the configuration of the magnetoresistance element 11 shown in FIG. 1, an insulating section 120 for insulating and isolating each of the magnetoresistance elements 110, a control section 12, a first wiring 121, a second wiring 122, and a switch 123. The control section 12 applies a voltage pulse for writing to each of the magnetoresistance elements 110 via the first wiring 121, the second wiring 122, and the switch 123 to perform writing, and applies a voltage pulse for reading to perform reading. The control section 12 preferably determines that the voltage pulse for writing is a unipolar pulse, thereby enabling bidirectional writing of the magnetoresistance elements 110.
[0065] The magnetoresistance element 110 includes a multilayer film 111 including a recording layer, a reference layer 112, a tunnel barrier layer 113 sandwiched between the multilayer film 111 and the reference layer 112, a first electrode 114 in contact with an upper side of the multilayer film 111, and a second electrode 115 in contact with a lower side of the reference layer 112. A first wiring 121 is electrically connected to the first electrode 114. One terminal 123a of a switch 123 is electrically connected to the second electrode 115. The other terminal 123b of the switch 123 is electrically connected to a second wiring 122.
[0066] The multilayer film 111 may have, for example, a stacked structure (not shown) of a magnetization free layer / a non-magnetic layer / a magnetization fixed layer, or may be a single layer of a magnetization free layer. The magnetoresistance element 110 may be appropriately provided with a cap layer, a protective film, a seed layer, a buffer layer, etc. (not shown).
[0067] The magnetic memory device 100 may be formed by laminating a first electrode 114 or a second electrode 115, a seed layer, a buffer layer, etc., on a substrate by a sputtering method, a vacuum deposition method, a chemical vapor deposition (CVD) method, etc., and then laminating a multilayer film 111, a tunnel barrier layer 113, a reference layer 112, etc. thereon. In addition, in FIG. 11, the multilayer film 111 including the recording layer is above the reference layer 112, but the multilayer film 111 may be laminated below the reference layer 112.
[0068] The magnetic memory device 100 may be formed using a process technique, known as a three-dimensional stacking process technique, in which a magnetoresistance element 110 is formed on a substrate, and then the magnetoresistance element 11 is pressure-bonded to a circuit including a first wiring 121, a second wiring 122, a switch 123, etc., formed on another substrate, and finally the substrate is removed from the magnetoresistance element 110.
[0069] The first electrode 114 and the second electrode 115 are made of a non-magnetic conductive material, and contain at least one element selected from the group consisting of Ta, Ru, W, Ir, Au, Ag, Cu, Al, Cr, Pt, and Pd. The thickness of the first electrode 114 and the second electrode 115 is preferably 1 nm or more and 200 nm or less. The thickness (th el) is greater than the length L1 of the short side SS or the length L2 of the long side LS in the in-plane shape of the recording layer 13 shown in FIG. 2 and is 200 nm or less, i.e., L1 or L2 <th el It is more preferable that the thickness is ≦200 nm. This allows the first electrode 114 and the second electrode 115 to have good flatness and a sufficiently low resistance value.
[0070] The insulating section 120 includes a non-magnetic insulating compound. The insulating compound is, for example, an oxide, a nitride, or a fluoride of at least one element selected from the group consisting of Si, Al, Ti, Mg, and Ta.
[0071] Fig. 12 is an explanatory diagram (part 1) of an operation example of a magnetic memory device according to an embodiment of the present invention. The following description will be given with reference to Fig. 11 as well. The horizontal axis of Fig. 12(a) and (b) represents time, and the vertical axis represents the voltage V of a signal applied between the first wiring 121 and the second wiring 122. The signal substantially corresponds to a signal applied between the multilayer film 111 including the recording layer of the magnetoresistance element 110 and the reference layer 112.
[0072] [Write and read operation example] Referring to FIG. 12(a), as an example of an information write operation, the control unit 12 of the magnetic storage device 100 performs a first operation OP1 in which a switch is turned on (ON) to apply a first pulse P1 (pulse height H1, pulse width T1) between the first wiring 121 and the second wiring 122. In the first operation OP1, the first pulse P1 is supplied between the multilayer film 111 including the recording layer and the reference layer 112. The effective magnetic field described in FIG. 1 and FIG. 2 is applied to the recording layer 13. The effective magnetic field may be kept applied regardless of the operation time of the first operation OP1, may be applied before the first operation OP1 is performed, or may be applied simultaneously with the first pulse P1. The first pulse P1 reverses the magnetization of the recording layer included in the multilayer film 111, changing the electric resistance value between the multilayer film 111 and the reference layer 112, and the information stored in the magnetoresistance element 110 is rewritten. Furthermore, the change in electrical resistance value between the multilayer film 111 and the reference layer 112 corresponds to a change in electrical resistance value between the first electrode 114 and the second electrode 115 (i.e., a change in the electrical resistance value of the magnetoresistance element 110), and further corresponds to a change in electrical resistance value between the first wiring 121 and the second wiring 122.
[0073] If the electrical resistance value R before the first pulse P1 is supplied is the first electrical resistance value R1, and the electrical resistance value R after the first pulse P1 is supplied is the second electrical resistance value R2, the second electrical resistance value R2 is different from the first electrical resistance value R1. These electrical resistance values R1 and R2 are based on the relative directions of the magnetization of the recording layer and the magnetization of the reference layer 112 included in the multilayer film 111. For example, the electrical resistance value R is higher when the relative directions of the magnetization are antiparallel than when they are parallel. The multiple states in which the electrical resistance values R differ from each other correspond to the information to be stored. The first operation OP1 changes the relative directions of the magnetization, changing the electrical resistance value R between the multilayer film 111 and the reference layer 112.
[0074] Referring to FIG. 12(b), the control unit 12 may perform a second operation OP2, which is a read operation of the stored information, by turning on the switch 123 and applying a second pulse P2 (pulse height H2) between the first wiring 121 and the second wiring 122 before performing the first operation OP1. The polarity of the second pulse P2 may be the same as or opposite to the polarity of the first pulse P1. From the viewpoint of preventing erroneous writing, it is preferable that the polarity of the second pulse P2 is opposite to the polarity of the first pulse P1. In addition, when the polarity of the second pulse P2 is the same as the polarity of the first pulse P1, the pulse height H2 is set so that the relationship of |H2|<|H1| is satisfied. The electric resistance value (third electric resistance value R3) between the multilayer film 111 and the reference layer 112 after the second operation OP2 is different from the second electric resistance value R2 after the first operation OP1. The third electric resistance value R3 is, for example, the same as the first electric resistance value R1.
[0075] When using the second pulse P2 having the opposite polarity in this way, the absolute value of the height of the second pulse P2 (|H2|) may be smaller than, larger than, or the same as the absolute value of the height of the first pulse P1 (write pulse) (|H1|). As a result, in this embodiment, the magnetic anisotropy of the recording layer is controlled by voltage, so that by using the second pulse having the opposite polarity to the first pulse P1, it is possible to suppress the magnetization direction of the recording layer from changing during reading.
[0076] For example, even if the absolute value of the height of the second pulse P2 is set to be the same as the absolute value of the height of the first pulse, the absolute value of the difference in the electrical resistance R before and after application of the second pulse P2 in the second operation OP2 (i.e., the absolute value of the difference between the first electrical resistance R1 and the third electrical resistance R3) is smaller than the absolute value of the difference in the electrical resistance before and after application of the first pulse in the first operation OP1 (i.e., the absolute value of the difference between the first electrical resistance R1 and the second electrical resistance R2). In other words, even if a second pulse P2 of opposite polarity to the first pulse P1 is applied as a voltage pulse of the same height as the first pulse P1, no information is rewritten.
[0077] If the information is not to be rewritten, the control unit 12 may perform the second operation OP2 and not perform the first operation OP1 thereafter.
[0078] The control unit 12 applies a first pulse P1 having a pulse height H1 and a time width T1 to the magnetoresistance element 110, thereby enabling the information of the magnetoresistance element 110 to be rewritten, and the magnetoresistance element 110 changes from a high resistance state to a low resistance state, or from a low resistance state to a high resistance state. An appropriate pulse height H1 can be determined by performing the first operation OP1 multiple times or more times, measuring the electrical resistance values R1 and R2 before and after the first operation OP1, and determining the probability that a desired change has been obtained. The lower the pulse height H1, the smaller the power consumption. The maximum value of the pulse height H1 that the control unit 12 can output is preferably 1.8V or less, more preferably 1.5V or less, even more preferably 1.2V or less, even more preferably 0.9V or less, and particularly preferably 0.7V or less.
[0079] In the writing of the first embodiment and the writing in which the magnetization reversal in the second magnetization reversal direction occurs in the third embodiment, applying the first pulse P1 with a time width twice the appropriate time width T1, that is, 2×T1, is not appropriate because it reduces the probability of the electric resistance value changing before and after application. On the other hand, in the writing of the second embodiment and the writing in which the magnetization reversal in the first magnetization reversal direction occurs in the third embodiment, the time width of the first pulse P1 may be longer than the appropriate time width T1. However, the longer the time width, the greater the power consumption, so it is not preferable to unnecessarily increase the pulse time width. Also, in the first to third embodiments, a time width shorter than the appropriate time width T1, for example, a time width of 30% or less of T1, is not preferable.
[0080] 13 is an explanatory diagram (part 2) of an operation example of the magnetic storage device according to one embodiment of the present invention. The following description will be given with reference to FIG. 11. The horizontal and vertical axes in FIG. 13 are the same as those in FIG.
[0081] Referring to FIG. 13, the write pulse of the first operation OP1 is a pulse with a time width T1, and the time it takes for the pulse to rise to a predetermined height H1 is τ R and the time it takes for the voltage to fall to 0, τ FFor example, in the second magnetization reversal direction 2 of the first, second, and third embodiments, the time τ R is the time τ F It is preferable that the time τ is equal to or smaller than the time τ R From the viewpoint of stabilizing writing, it is preferable that the time τ is shorter, and is preferably closer to zero. F is preferably longer, but from the viewpoint of power consumption, it is preferable that it is not excessively long, and is preferably substantially greater than 0 ns and less than 10 ns. Note that, although not shown, the time taken for the second pulse P2 of the second operation OP2 shown in Fig. 12(b) to rise and fall are similar to the time taken for the first pulse P1 to fall.
[0082] Although the preferred embodiment of the present invention has been described above in detail, the present invention is not limited to the specific embodiment, and various modifications and changes are possible within the scope of the present invention described in the claims. For example, the above-mentioned embodiments may be combined with each other. The above macro spin simulation was performed using a simulation program written in C language. [Explanation of symbols]
[0083] 10,100 Magnetic storage devices 11,110 Magnetoresistance element 12 Control section 13 Recording layer 14,113 Tunnel Barrier Layer 15,112 Reference layer 111 Multilayer film
Claims
1. a magnetoresistive element having a laminated structure including a recording layer including a first magnetic layer, a tunnel barrier layer, and a reference layer including a second magnetic layer; A control unit, the in-plane shape of the recording layer is a shape in which a rectangle that circumscribes the in-plane shape and has a minimum area has short and long sides that are different in length; When writing to the recording layer, the control unit applies a voltage pulse to the magnetoresistive element while applying an effective magnetic field in the direction of the short side of the plane of the recording layer; A magnetic storage device in which the effective magnetic field H eff satisfies the relationship 0.2×H k eff <|H eff |<M s (N y -N x ), where H k eff is the effective perpendicular magnetic field due to perpendicular magnetic anisotropy, M s is the saturation magnetization of the recording layer, N x is the demagnetizing factor in the long side direction, and N y is the demagnetizing factor in the short side direction.
2. A magnetoresistive element having a laminated structure of a recording layer including a first magnetic layer, a tunnel barrier layer, and a reference layer including a second magnetic layer; A control unit, the in-plane shape of the recording layer is a shape in which a rectangle circumscribing the in-plane shape and having a smallest area has short and long sides of different lengths, and when the length of the short side of the rectangle is L1 and the length of the long side is L2, the ratio L2 / L1 is 1.5 or more and 5 or less, The magnetic storage device, wherein, when writing to the recording layer, the control unit applies a voltage pulse to the magnetoresistance element while applying an effective magnetic field in the direction of the short side of the plane of the recording layer.
3. A magnetic storage device as claimed in claim 1, wherein the voltage pulse applied when writing to the recording layer is a unipolar pulse voltage, and application of the unipolar voltage pulse causes bidirectional magnetization reversal in the recording layer.
4. The magnetic storage device according to claim 1, wherein the recording layer contains one or more of Hf, Ta, W, Os, Ir, Pt and Au.
5. The magnetic storage device according to claim 1, wherein the effective magnetic field further has a component perpendicular to the film surface of the recording layer.
6. The magnetic storage device according to claim 1, wherein the recording layer has perpendicular magnetic anisotropy.
7. The magnetic storage device according to claim 1, wherein the recording layer contains Co--Fe--B.
8. The magnetic memory device according to claim 1, wherein the tunnel barrier layer is an oxide containing Mg.
9. The magnetic memory device according to claim 1, wherein the tunnel barrier layer is an oxide containing Mg and Al.
10. The magnetic memory device according to claim 1, wherein the tunnel barrier layer has a thickness of 1.2 nm or more.
11. The magnetic storage device according to claim 1, wherein the in-plane shape of the recording layer is such that, when the length of the short side of the rectangle is L1 and the length of the long side is L2, the ratio L2 / L1 is greater than 1 and not greater than 5.
12. The magnetic storage device according to claim 1, wherein the in-plane shape of the recording layer is such that, when the length of the short side of the rectangle is L1 and the length of the long side is L2, the ratio L2 / L1 is 1.5 or more and 5 or less.
Citation Information
Patent Citations
Non-volatile storage device
JP2014067929A
Magnetoresistive element and electronic device
JP2017157662A
Bipolar voltage writing type magnetic memory element
JP2018014376A
Magnetic storage device
JP2020021782A
Nonvolatile memory device
US20140085969A1