JP6548415B2 - Method of providing a perpendicular magnetic anisotropic magnetic junction usable in spin transfer torque magnetic devices using a sacrificial insertion layer - Google Patents

Description

  The present invention relates to magnetic junctions and methods of providing magnetic junctions and magnetic memory usable in magnetic devices on a substrate.

  Magnetic memories, in particular Magnetic Random Access Memories (MRAMs), are gaining attention because of their potential for high read / write speeds, excellent durability, non-volatility, and low power consumption during operation. . The MRAM uses a magnetic material as an information storage medium to store information. One type of MRAM is an STT-RAM (Spin Transfer Torque Random Access Memory). STT-RAM utilizes magnetic junctions (elements) at least partially recorded by the current passing through the magnetic junctions. The spin-polarized current passing through the magnetic junction adds spin torque to the magnetic moment in the magnetic junction. Thus, a layer having a magnetic moment responsive to spin torque is switched in a desired manner.

  As an example, FIG. 1 shows a conventional Magnetic Tunneling Junction (MTJ) 10 that can be used in a conventional STT-RAM. A conventional MTJ 10 is typically disposed on a substrate 12. Bottom contact 14 and top contact 22 are used to drive current through a conventional MTJ 10. A conventional MTJ utilizes a conventional seed layer (not shown) and includes a capping layer and a conventional antiferromagnetic layer (AFM) (not shown). The conventional MTJ 10 includes a conventional pinned layer 16, a conventional tunneling barrier layer 18, and a conventional free layer 20. Also, the upper contact 22 is shown. Conventional contacts (14, 22) are used to drive current in the current-perpendicular-to-plane (CPP) direction or in the z-axis shown in FIG. Typically, the conventional pinned layer 16 is adjacent to the substrate 12 of the layer (16, 18, 20).

  The conventional pinned layer 16 and the conventional free layer 20 have magnetism. The magnetization 17 of the conventional pinned layer 16 is fixed or pinned in a specific direction. Although shown in the drawings as a single layer, the conventional pinned layer 16 may include multiple layers. For example, the conventional pinned layer 16 is a synthetic antiferromagnetic (SAF) layer that includes a magnetic layer that is antiferromagnetically coupled through a thin conductive layer such as Ru. In such an SAF, a plurality of magnetic layers in which thin Ru layers are inserted are used. As another example, the coupling across the Ru layer can be ferromagnetic.

  The conventional free layer 20 has a variable magnetization 21. Although shown as a single layer, the conventional free layer 20 can include multiple layers. For example, the conventional free layer 20 is a synthetic layer that includes an antiferromagnetically or ferromagnetically coupled magnetic layer through a thin conductive layer such as Ru. Although shown perpendicular to the plane, the magnetization 21 of the conventional free layer 20 is in plane. Therefore, the pinned layer 16 and the free layer 20 each have magnetization (17, 21) in the direction perpendicular to the plane of the layer.

  In order to switch the magnetization 21 of the conventional free layer 20, a current is driven in the direction perpendicular to the plane (z-direction). The magnetization 21 of the conventional free layer 20 is switched parallel to the magnetization 17 of the conventional pinned layer 16 when sufficient current is driven from the top contact 22 to the bottom contact 11. When sufficient current is driven from the lower contact 11 to the upper contact 24, the magnetization 21 of the conventional free layer 20 is switched antiparallel to the magnetization 17 of the pinned layer 16. The differences in magnetic configurations correspond to different magnetoresistances. That is, they correspond to different logic states (eg, logic "0" and logic "1") of the conventional MTJ 10.

  Research on magnetic memory is in progress as it may be used in a variety of applications. For example, a mechanism for improving the performance of STT-MRAM is required. That is, there is a need for a method and system for improving the performance of spin transfer torque based memories. The methods and systems described herein address such needs.

  The present invention has been made in view of the above-mentioned prior art, and an object of the present invention is to provide a magnetic junction and a method of providing a magnetic junction and a magnetic memory usable on a magnetic device on a substrate. .

  A method of providing on a substrate a magnetic junction usable in a magnetic device according to an aspect of the present invention made to achieve the above object comprises the steps of: writing current being passed through the magnetic junction between a plurality of stable magnetic states. Providing a switched free layer, providing a nonmagnetic spacer layer, and providing a pinned layer, said nonmagnetic A spacer layer is disposed between the free layer and the pinned layer, and at least one of providing the free layer and providing the pinned layer includes a plurality of steps. At least one step of the step of providing a free layer includes a first plurality of steps, and At least one step of providing a layer includes a second plurality of steps, wherein the first plurality of steps comprises: depositing a first region of the free layer; and depositing a first sacrificial layer Annealing at least a first region of the free layer and the first sacrificial layer at a first temperature higher than 25 ° C., removing the first sacrificial layer, and a second region of the free layer Depositing, the second plurality of steps including depositing a first region of the pinned layer, depositing a second sacrificial layer, and at least a first region of the pinned layer Annealing the second sacrificial layer at a second temperature above 25 ° C., defining a region of a magnetic junction including the free layer, the nonmagnetic spacer layer, and the first region of the pinned layer Removing the second sacrificial layer, and depositing a second region of the pinned layer.

The step of providing the free layer includes the first plurality of steps, wherein the free layer has perpendicular magnetic anisotropy energy greater than out-of-plane demagnetization energy. It can have
The free layer may have a thickness greater than 15 Å.
The thickness of the free layer may be 25 Å or less.
The first sacrificial layer is made of Bi, W, I, Zn, Nb, Ag, Cd, Hf, Os, Mo, Ca, Hg, Sc, Y, Sr, Mg, Ti, Ba, K, Na, Rb, Pb And at least one of Zr.
The method may further include depositing a MgO seed layer before providing the free layer.
The annealing may include performing a rapid thermal anneal (RTA).
The first region of the free layer has a first thickness, the second region of the free layer has a second thickness, the first thickness is less than 15 Å, and the second thickness is 15 Å. It may be less than.
Providing a pinned layer comprising the second plurality of steps may further include depositing at least one refill material prior to removing the second sacrificial layer.
The method may further include performing planarization after depositing the at least one refill material.
The pinned layer is a synthetic antiferromagnetic material that includes a first ferromagnetic layer, a second ferromagnetic layer, and a coupling layer between the first ferromagnetic layer and the second ferromagnetic layer. The depositing of the second region of the pinned layer may include depositing at least one nonmagnetic layer spacer and depositing the second ferromagnetic layer.
The depositing the second region of the pinned layer may further include depositing the region of the first ferromagnetic layer.
At least one of the first ferromagnetic layer and the second ferromagnetic layer may be a multilayer.
The method of providing a magnetic junction may further include the step of defining a remaining area of the pinned layer.
Providing a pinned layer comprising the second plurality of steps further comprising defining a region of a magnetic junction prior to depositing the at least one refill material, defining the region of the magnetic junction Providing a photoresist mask on the second sacrificial layer covering a region of the second sacrificial layer corresponding to the magnetic junction, an exposed region of the second sacrificial layer, a first region of the fixed layer And removing the nonmagnetic spacer layer and the free layer exposed by the photoresist mask.
The method further includes the steps of providing an additional nonmagnetic spacer layer and providing an additional pinned layer, wherein the free layer comprises the additional nonmagnetic spacer layer and the additional nonmagnetic spacer layer. Between the nonmagnetic spacer layer and the additional nonmagnetic spacer layer may be between the additional pinned layer and the free layer.

  A method of providing on a substrate a magnetic memory usable in a magnetic device according to one aspect of the present invention made to achieve the above object is a method of providing a first ferromagnetic layer of a free layer including a CoFeB layer having a thickness of 15 Å or less Depositing a layer on the substrate, Bi, W, I, Zn, Nb, Ag, Cd, Hf, Os, Mo, Ca, Hg, Sc, Y, Sr, Mg on the first ferromagnetic layer Depositing a first sacrificial layer containing at least one of Ti, Ba, K, Na, Rb, Pb, and Zr and having a thickness of 4 Å or less, at least the first ferromagnetic layer, and Annealing the first sacrificial layer by a first rapid thermal annealing (RTA) at a first temperature above 25 ° C., removing at least the first sacrificial layer, and over the remaining area of the first ferromagnetic layer The first ferromagnetism Depositing a second ferromagnetic layer of said free layer comprising a CoFeB layer less than 15 Å thick so as to have a thickness less than 25 Å combined with the remaining area of the second layer, between the pinned layer and the free layer Providing a nonmagnetic spacer layer formed on the substrate, depositing the first region of the pinned layer, Bi, W, I, Zn, Nb, Ag, Cd, Hf, Os, Mo, Ca, Depositing a second sacrificial layer having a thickness of 4 Å or less including at least one of Hg, Sc, Y, Sr, Mg, Ti, Ba, K, Na, Rb, Pb, and Zr; The first region of the pinned layer, the remaining region of the first ferromagnetic layer, the second ferromagnetic layer, and the second sacrificial layer are subjected to a second rapid thermal annealing (RTA) to a second temperature higher than 25 ° C. Annealing at the second height, and Providing a photoresist mask on the sacrificial layer covering a region of the sacrificial layer corresponding to the at least one magnetic junction after the annealing step by thermal annealing; and using the photoresist mask, the free layer, the Defining a region of at least one magnetic junction including a nonmagnetic spacer layer and a first region of the pinned layer, depositing at least one refill material, and depositing the at least one refill material Subsequent to the step of planarizing, the step of removing the second sacrificial layer after the step of planarizing, the step of depositing at least a second region of the layer to be fixed, the layer to be fixed Of the at least one magnetic junction after depositing at least a second region of Defining a region, wherein the free layer has a perpendicular magnetic anisotropy energy greater than out-of-plane demagnetization energy, and a plurality of stable magnetic states as a write current flows through the magnetic junction. Switched between.

  A magnetic junction usable in a magnetic device according to one aspect of the present invention made to achieve the above object is switched between multiple stable magnetic states as write current flows through the magnetic junction and out-of-plane demagnetization energy And a nonmagnetic spacer layer having a larger perpendicular magnetic anisotropy energy and a ferromagnetic layer thicker than 15 Å, a nonmagnetic spacer layer, and a pinned layer, wherein the nonmagnetic spacer layer is the pinned layer. It is formed between a layer and the free layer.

The ferromagnetic layer may include a CoFeB layer having perpendicular magnetic anisotropy energy greater than the out-of-plane demagnetization energy.
The nonmagnetic spacer layer is a crystalline MgO tunnel barrier layer, the magnetic junction further includes a MgO seed layer, and the ferromagnetic layer is between the MgO seed layer and the crystalline MgO tunnel barrier layer. It can be arranged.

It is a figure which shows the conventional magnetic junction. FIG. 5 illustrates a method of providing a useable magnetic junction in programmable magnetic memory using spin transfer torque according to one embodiment. FIG. 5 illustrates a magnetic junction usable in magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates a magnetic junction usable in magnetic memory programmable with spin transfer torque according to another embodiment. FIG. 5 illustrates a method of providing a portion of a magnetic junction usable in programmable magnetic memory using spin transfer torque according to another embodiment. FIG. 5 illustrates a magnetic junction usable in magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates a method of providing a useable magnetic junction in programmable magnetic memory using spin transfer torque according to another embodiment. FIG. 5 illustrates a magnetic junction usable in magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates a method of providing a useable magnetic junction in programmable magnetic memory using spin transfer torque according to another embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to one embodiment. FIG. 7 illustrates a manufacturing process of an in-process magnetic junction usable with a magnetic memory programmable using spin transfer torque according to one embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to another embodiment. FIG. 5 illustrates an in-process magnetic junction usable in a magnetic memory programmable with spin transfer torque according to another embodiment. FIG. 5 illustrates a memory utilizing magnetic junctions in a memory element of a storage cell according to one embodiment.

  The present embodiments relate to magnetic junctions that can be used in magnetic devices such as magnetic memories and devices that use such magnetic junctions. Magnetic memories include spin transfer torque magnetic random access memories (STT-MRAMs) and are used in electronic devices that employ non-volatile memory. Such electronic devices include, but are not limited to, cell phones, smart phones, tablets, laptops, and other portable and non-portable computing devices. The following description is provided to enable any person skilled in the art to which the present invention pertains to practice the present invention and is provided as part of the patent application and its requirements. Various modifications of the exemplary embodiments described herein, and the principles and forms thereto, will be apparent to those skilled in the art to which the present invention pertains. Although this embodiment primarily describes particular methods and systems provided in particular embodiments, the above methods and systems work well with other embodiments. Words such as "this embodiment", "one embodiment", and "other embodiments" may relate to the same or other embodiments as well as the plurality of embodiments. Although the embodiments describe systems and / or devices having a certain configuration, the systems and / or devices may include more or less than the illustrated configurations, and the arrangement and configuration form is modified within the scope of the present invention It can. Also, while the present embodiment describes a particular method having predetermined steps, such method and system may have other sequences of steps not inconsistent with the exemplary embodiment having other and / or additional steps. It works effectively in other methods that it has. Accordingly, the present invention is not intended to be limited to the illustrated embodiments. It has the broadest scope consistent with the principles and features described herein.

  The method and system provide magnetic junctions as well as magnetic memory utilizing magnetic junctions. The present embodiments provide magnetic junctions and methods of providing magnetic junctions that can be used in magnetic devices. The magnetic junction providing method provides a free layer that is switched between multiple stable magnetic states as the write current flows through the magnetic junction, and provides a nonmagnetic spacer layer. Providing a pinned layer, the nonmagnetic spacer layer is disposed between the free layer and the pinned layer, and the step of providing the free layer includes a first plurality of steps; The step of providing the layer comprises a second plurality of steps, the first plurality depositing the first region of the free layer, depositing the first sacrificial layer, at least the first region of the free layer and the first sacrificial Annealing the layer at a first temperature greater than 25 ° C., removing the first sacrificial layer, and depositing a second region of the free layer; Depositing a first region of the pinned layer, depositing a second sacrificial layer, annealing at least the first region of the pinned layer and the second sacrificial layer at a second temperature higher than 25 ° C .; Defining the region of the magnetic junction including the free layer, the nonmagnetic spacer layer, and the first region of the pinned layer, removing the second sacrificial layer, and depositing the second region of the pinned layer.

  The present embodiment describes a specific method, a magnetic junction, and a magnetic memory having a configuration. Those of ordinary skill in the art to which the present invention pertains will appreciate that the present invention is consistent with the use of magnetic junctions and magnetic memories having other and / or additional configurations and / or other features not inconsistent with the present invention. . The methods and systems also describe the understanding of spin transfer phenomena, magnetic anisotropy, and other physical aspects. As a result, those of ordinary skill in the art to which the present invention pertains will appreciate that theoretical explanations for the operation of the present methods and systems are such as spin transfer, magnetic anisotropy, and other physical conditions. It is easy to see what is done based on the current understanding. However, the methods and systems described herein do not rely on specific physical explanations. Also, those of ordinary skill in the art to which the present invention pertains will readily appreciate that the methods and systems describe structures having a specific relationship to the substrate. However, those of ordinary skill in the art to which the present invention pertains will readily appreciate that the method and system are consistent with other structures. Also, the method and system describe a combined and / or single layer. However, those of ordinary skill in the art to which the present invention pertains readily appreciate that the layers may have other structures. Furthermore, the method and system describe magnetic junctions and / or substructures with special layers. However, those of ordinary skill in the art to which the present invention pertains can easily use magnetic junctions and / or substructures with additional and / or other layers not inconsistent with the present method and system. I understand. In addition, some configurations are described as magnetic, ferromagnetic, and ferrimagnetic. The term magnetic as used herein may include ferromagnetic, ferrimagnetic, or similar structures. Thus, the terms "magnetic" or "ferromagnetic" as used herein include, but are not limited to ferromagnets and ferrimagnets. The methods and systems also describe single magnetic junctions and substructures. However, those of ordinary skill in the art to which the present invention pertains will readily appreciate that the method and system relate to the use of magnetic memory with multiple magnetic junctions and multiple infrastructures. Furthermore, as used herein, "in-plane" is substantially in or parallel to the plane of one or more layers of the magnetic junction. Conversely, "perpendicular" applies in a direction substantially perpendicular to one or more layers of the magnetic junction.

  FIG. 2 is a diagram illustrating one embodiment of a method 100 of fabricating a magnetic junction usable in magnetic memory, such as STT-RAM. Thus, the method 100 may be used in various types of electronic devices. For the sake of simplicity, some steps may be omitted or performed separately or in combination. Additionally, method 100 may begin after other steps of magnetic memory formation have been performed.

  A free layer comprising material deposition for the free layer is provided by step 102. The free layer is deposited on the seed layer. The seed layer is selected for various purposes including, but not limited to, the crystal structure of the free layer, the magnetic anisotropy, and / or the magnetic damping of the free layer. For example, the free layer is provided on a seed layer such as a crystalline MgO layer that improves the perpendicular magnetic anisotropy of the free layer. When dual magnetic junctions are fabricated, free layers are formed on other nonmagnetic spacer layers. Such nonmagnetic spacer layer is the MgO seed layer described above. The layer to be fixed is formed under such a spacer layer.

  The free layer provided in step 102 is required to have a perpendicular magnetic anisotropy that exceeds the demagnetization energy. Thus, the magnetic moment of the free layer is stable out-of-plane, including being perpendicular-to-plane. In addition, a polarization enhancement layer (PEL) is provided as part of or additionally to the free layer. PEL contains a high spin polarization material. Also, the free layer provided in step 102 is formed such that the write current is switched between the plurality of stable magnetic states as it flows through the magnetic junction. Thus, the free layer is switched using spin transfer torque. The free layer provided in step 102 is magnetic and is thermally stable at the operating temperature. However, although providing a free layer is described in step 102, the edges of the free layer are defined in the stack provided later.

  In one embodiment, step 102 includes additional steps. In such embodiments, the first region of the free layer is deposited first. The first region of the free layer comprises a magnetic layer comprising Co, Fe and / or B. For example, a CoFeB layer having a B of 20 atomic percent or less is deposited. Also, in such an embodiment, step 102 includes depositing a sacrificial insertion layer on the first ferromagnetic layer to share the interface between the layers. The sacrificial insertion layer comprises a material that has an affinity for boron, has low diffusion and has a relatively good lattice match with the underlying layer. For example, the difference in lattice parameter between the lower ferromagnetic layer and the sacrificial insertion layer is 10% or less. The grid insertion layer is thin. In one embodiment, the sacrificial insertion layer is 10 Å or less thick. In such embodiments, the sacrificial insertion layer is no greater than 4 Å and greater than 1 Å. In other embodiments, different thicknesses are used. The sacrificial insertion layer and the lower layer are then annealed at room temperature (e.g. 25 [deg.] C) or higher. For example, rapid thermal annealing (RTA) is used in a temperature range of 300 to 400 degrees Celsius. In other embodiments, the annealing may be performed in other manners, including but not limited to block heating. Also, annealing is performed at other temperatures. After annealing, the sacrificial insertion layer is removed by plasma etching, for example. In other embodiments, the sacrificial insertion layer is removed by yet other methods including, but not limited to, ion milling or CMP (chemical mechanical planarization). In the removing step, a partial region of the lower ferromagnetic layer is removed. Then, optionally, the remaining area of the free layer is deposited. For example, the second ferromagnetic layer is deposited on the exposed first ferromagnetic layer. Also, such a second ferromagnetic layer is another CoFeB layer. In one embodiment, the total amount of magnetic material provided will cause the free layer to have a perpendicular magnetic anisotropy that exceeds the demagnetization energy. For example, at the end of step 102, the first and second ferromagnetic layers have an overall thickness greater than 15 Å, without exceeding 30 Å together. In one embodiment, the overall thickness does not exceed 25 Å. For example, the total thickness is at least 16 Å and less than or equal to 25 Å. In other embodiments, the free layer is formed in another manner. In the present invention, the sacrificial insertion layer is also referred to as a sacrificial layer.

  Step 104 provides a nonmagnetic spacer layer. In one embodiment, a crystalline MgO tunnel barrier layer is required for the magnetic junction to be formed. Step 104 includes MgO deposition to form a tunnel barrier layer. In one embodiment, step 104 includes MgO deposition using, for example, high frequency (RF) sputtering. After the Mg metal is deposited, it is oxidized in step 104 to provide a native oxide of Mg. Also, the MgO barrier layer / nonmagnetic spacer layer is formed by another method. For step 102, as described above, the edge of the nonmagnetic spacer layer is later defined, for example, after deposition of the remaining area of the magnetic junction. Step 104 includes annealing the remaining region of the magnetic junction already formed to provide a crystalline MgO tunnel barrier having a "100" orientation for enhanced tunneling magnetoresistance (TMR). .

  Step 106 provides the anchored layer. Thus, the nonmagnetic spacer layer is between the pinned layer and the free layer. In one embodiment, in step 106, the pinned layer is formed after forming the free layer in step 102. In another embodiment, the free layer is formed first. The pinned layer is magnetic and the magnetization of the pinned layer is pinned or fixed in a particular direction during operation of at least some of the magnetic junctions. Thus, the pinned layer is thermally stable at the operating temperature. The pinned layer formed in step 106 is a single (single) layer or includes multiple layers. For example, the pinned layer formed in step 106 is an SAF that includes a thin nonmagnetic layer such as ruthenium (Ru) and an antiferromagnetic or ferromagnetic coupled magnetic layer. Also, within such SAFs, each magnetic layer comprises multiple layers. Moreover, a to-be-fixed layer is another multilayer. The pinned layer formed in step 106 has perpendicular anisotropy energy exceeding out-of-plane demagnetization energy. Thus, the pinned layer has a magnetic moment oriented perpendicular to the plane. Other orientations of the magnetization of the pinned layer are possible. Also note that other layers such as PEL or coupling layers are inserted between the pinned layer and the nonmagnetic spacer layer.

  In one embodiment, step 106 includes multiple steps similar to those described above for step 102. For example, the first region of the layer to be fixed is deposited first. The first region of the pinned layer includes a magnetic layer containing Co, Fe, and / or B. For example, a CoFeB layer containing B not exceeding 20 atomic percent is deposited. Also, a PEL or other structure is deposited between the pinned layer and the nonmagnetic spacer layer. Also, in such an embodiment, step 106 includes depositing another sacrificial insertion layer on the area of the formed pinned layer. In one embodiment, the sacrificial insertion layer is deposited directly on the ferromagnetic layer. In another embodiment, another layer is deposited between the ferromagnetic layer and the sacrificial insertion layer. The sacrificial insertion layer comprises a material that has a relatively good lattice match with the lower layer, has low diffusion, and has an affinity for boron. For example, the difference in lattice parameter between the lower ferromagnetic layer and the sacrificial insertion layer is 10% or less. The sacrificial insertion layer is thin. In one embodiment, the sacrificial insertion layer has the same thickness as the free layer described above. In other embodiments, other thicknesses are used. However, the sacrificial insertion layer is required to be continuously formed for patterning as described later. The sacrificial insertion layer and the lower layer are later annealed at a temperature higher than the room temperature. For example, RAT at a temperature in the range of 300-400 degrees Celsius is used. In other embodiments, the annealing is performed in other ways. After annealing, the area of the magnetic junction under the sacrificial insertion layer is defined. For example, the edges of the magnetic junctions are defined using photolithographic masks and ion milling or other mechanisms for layers. A nonmagnetic insulating layer such as alumina is deposited to refill the peripheral region of the magnetic junction. Also, planarization is performed. The sacrificial layer is later removed by, for example, plasma etching. Also, other removal methods are used. In the removing step, a partial region of the lower ferromagnetic layer is removed. Later, the remaining area of the anchored layer is optionally deposited. For example, an additional ferromagnetic layer is deposited directly on the exposed first ferromagnetic layer. In embodiments where the pinned layer is an SAF, a nonmagnetic layer such as Ru is deposited. Other magnetic layers are also provided on the nonmagnetic layer. In another embodiment, the anchored layer is formed in another manner.

  FIG. 3 is a diagram illustrating one embodiment of a magnetic junction 200 and peripheral structure fabricated using method 100. Clearly, FIG. 3 is not a ratio of actual sizes, but is intended to aid understanding. The magnetic junction 200 is used in magnetic devices such as spin transfer torque magnetic random access memory (STT-MRAM) and in various types of electronic devices. The magnetic junction 200 includes a free layer 210 having a magnetic moment 211, a nonmagnetic spacer layer 220, and a pinned layer 230 having a magnetic moment 231. In addition, although a transistor is formed on the lower substrate 201, the device is not limited thereto. Also shown are bottom contact 202 and top contact 208, optional seed layer 204, and optional capping layer 206. As shown in FIG. 3, the pinned layer 230 is adjacent to the top of the magnetic junction 200 (far from the substrate 201). A selective pinned layer (not shown) is used to pin the magnetization (not shown) of pinned layer 230. In one embodiment, the selective pinned layer is an AFM layer or a multiayer that pins the magnetization (not shown) of pinned layer 230 by exchange-bias interactions. However, in other embodiments, the optional pinned layer is omitted or other structures are used. Furthermore, in one embodiment, the orientation of the pinned layer 230 and the free layer 210 with respect to the substrate 201 is reversed. That is, the pinned layer 230 is closer to the substrate than the free layer 210 in another embodiment.

  In the embodiment shown in FIG. 3, the perpendicular magnetic anisotropy energy of each of the pinned layer 230 and the free layer 210 exceeds the out-of-plane demagnetization energy of the pinned layer 230 and the free layer 210. Thus, each of the magnetic moment 211 of the free layer 210 and the magnetic moment 231 of the pinned layer 230 is perpendicular to the plane. States are different, and the stable magnetic state for the free layer 210 is a moment oriented in the + z or −z direction. Each of the free layer 210 and the pinned layer 230 includes dotted lines that point to areas of the layer 210 and / or layer 230 that are separately formed with the use of a sacrificial insertion layer that is removed prior to completing the magnetic junction 200.

  The magnetic junction 200 is configured to switch the free layer 210 between a plurality of stable magnetic states as a write current flows through the magnetic junction 200. That is, the free layer 210 is switched using a spin transfer torque when the write current is driven by the magnetic junction 200 in a current perpendicular-to-plane (CPP) direction. Therefore, the data stored in the magnetic junction 210 and the magnetization direction of the free layer 210 can be read by the read current flowing through the magnetic junction 200. Also, the read current is driven by the magnetic junction 200 in the CPP direction. That is, the magnetoresistance of the magnetic junction 200 provides a read signal.

  The magnetic junction 200 and free layer 210 have enhanced performance by manufacturing using step 102 and / or step 106. Such effects are discussed below in connection with specific physical mechanisms. However, one of ordinary skill in the relevant art can readily recognize that the methods and systems described do not rely on specific physical descriptions. If the free layer 210 is formed utilizing the sacrificial insertion layer in step 102, the free layer 210 may be thick and perpendicular to the plane for magnetic moment 211, enhanced reluctance, and / or low damping ( Perpendicular-to-plane) stable state. When formed without a sacrificial insertion layer, the free layer typically does not exceed 12 Å in thickness to maintain the magnetic moment perpendicular to the plane. For example, a ferromagnetic CoFeB layer that is approximately 15 Å thick has an in-plane magnetic moment. However, although thinner free layers have a perpendicular-to-plane magnetic moment, the reluctance is reduced. Such a reduction is particularly noticeable when the free layer is between two MgO layers, such as a MgO seed layer and a MgO nonmagnetic spacer layer. The decrease in tunnel magnetoresistance is believed to be due to a conflict in the crystallinity of the free layer and the MgO layer. Also, the free layer forms a permanent insertion layer between the two magnetic layers. Such free layers have an overall thickness greater than 12 Å. The magnetic layers are separated by a permanent insertion layer. Each of the magnetic layers is in an order not thicker than 12 Å thick to maintain the magnetic moment perpendicular to the plane. Such thin magnetic / free layers have a magnetic moment perpendicular to the plane. Furthermore, the magnetoresistance can be improved. For example, a permanent insertion layer such as W reduces the collisions between the crystallinity of peripheral layers such as the MgO layer and the free layer. This results in a higher magnetic resistance. However, the attenuation may be higher than required. Such high attenuation increases the switch current (the write current required to change the state of the free layer magnetic moment). High switch currents are generally undesirable. That is, problems occur in the performance of such a magnetic junction.

  In contrast to such magnetic junctions, the magnetic junction 200 has higher magnetic resistance by using a sacrificial insertion layer (not shown in FIG. 3) in the manufacturing process. Use of the sacrificial insertion layer and subsequent annealing of the lower region of the free layer 210 causes crystallization of the free layer 210 prior to formation of the nonmagnetic spacer layer 220. This seems to be at least partially due to the affinity of the sacrificial insertion layer for B and O, unlike the free layer 210. Thus, the free layer 210 is fabricated to a greater thickness while maintaining the crystal structure and vertical anisotropy. For example, free layer 210 is thicker than 15 Å and has a magnetic moment 231 perpendicular to the plane. In one embodiment, free layer 210 does not exceed a thickness of 25 Å. For example, the free layer 210 is at least 16 Å thick and does not exceed 20 Å thick. Thus, the magnetic junction 200 has higher magnetic resistance. Also, removal of the sacrificial insertion layer reduces the damping of the free layer 210. That is, the free layer 210 exhibits lower switch current. Smaller write currents are used for magnetic junction programming. Thus, the performance is improved.

  Also, fabrication of pinned layer 230 in step 106 improves the performance of magnetic junction 200 in a magnetic device. Thus, the bottom layer (part of 204, 210, 220, 230) is defined before the entire pinned layer 230 is deposited, and thin areas of the magnetic junction 200 are removed during such a definition step. During such a definition step, shadowing by the closest magnetic junction in the magnetic device is mitigated. A similar effect is achieved in defining the remaining area of pinned layer 230 and the remaining area of magnetic junction 200 such as capping layer 206. Accordingly, the magnetic junction 200 is disposed adjacent to another magnetic junction (not shown in FIG. 3) without negatively affecting the manufacturing process. That is, the manufacturing process may be improved to realize a higher density memory device. When a sacrificial insertion layer is used in step 102 and step 106, the effect of performance improvement is achieved in all of the later described magnetic junction and magnetic device packaging / manufacturing processes.

  FIG. 4 is a diagram illustrating one embodiment of a magnetic junction 200 'and peripheral structures fabricated using method 100. As shown in FIG. Clearly, FIG. 4 is not a ratio of actual size, but is intended to aid understanding. The magnetic junction 200 'is used in magnetic devices such as spin-transfer magnetic random-access memory (STT) and various types of electronic devices. The magnetic junction 200 ′ is similar to the magnetic junction 200. Thus, similar components are shown as well. The free layer 210 having the magnetic moment 211 of the magnetic junction 200 ', the nonmagnetic spacer layer 220, and the pinned layer 230 having the magnetic moment 231 are the free layer 210 having the magnetic moment 211 shown in the magnetic junction 200, the nonmagnetic spacer layer 220 and similar to the pinned layer 230 with a magnetic moment 231. Also, the substrate 201 of the magnetic junction 200, the lower contact 202, the upper contact 208, the selective seed layer 204, and the lower substrate 201 of the magnetic junction 200 'similar to the selective capping layer 206, the lower contact 202, the upper contact 208, select Seed layer 204 and selective capping layer 206 are shown.

  The magnetic junction 200 'shown in FIG. 4 is a dual magnetic junction. That is, the magnetic junction 200 ′ includes an additional nonmagnetic spacer layer 240 and an additional pinned layer 250. The layer to be fixed 250 is similar to the layer to be fixed 230. Thus, the pinned layer 250 has a magnetic moment 251 perpendicular to the surface. In the illustrated embodiment, the magnetic junction 200 'is in a dual state. That is, the magnetic moments (231, 251) are antiparallel. In another embodiment, the magnetic moment (231, 251) is in an antidual or parallel state. In yet another embodiment, the magnetic moments (231, 251) are switched between dual and antidual states during operation. Nonmagnetic spacer layer 240 is similar to nonmagnetic spacer layer 220. However, the nonmagnetic spacer layer 240 may be formed of a material different from that of the nonmagnetic spacer layer 220 or may have the same or a different thickness. For example, layers (220, 240) are both "100" MgO. However, one layer, such as nonmagnetic spacer layer 240, is thin. In one embodiment, nonmagnetic spacer layer 240 is 30% thinner than nonmagnetic spacer layer 220.

  The dual magnetic junction 200 ′ shares the advantages of the magnetic junction 200. Thus, the magnetic junction 200 'is more closely packed with magnetic devices, with improved reluctance, reduced damping, and switch current.

  FIG. 5 illustrates one embodiment of a method 110 of fabricating a region of a magnetic junction usable with a magnetic device such as STT-RAM. For the sake of simplicity, some steps may be omitted or performed separately or in combination. Furthermore, method 110 may begin after other steps of magnetic memory formation have been performed. Method 110 is used in performing step 102 of method 100. However, in other embodiments, the method 110 is used in fabricating other regions of the magnetic junction 200, such as pinned layers, in conjunction with or separately from this, in conjunction with other fabrication steps.

  Method 110 may begin after another layer, such as a seed layer, is formed. For example, in one embodiment, method 110 may begin after a crystalline MgO seed layer having a "100" orientation is deposited. If a dual magnetic junction is fabricated, the MgO "seed" layer is another nonmagnetic spacer layer formed on the pinned layer. Furthermore, PEL is provided as part of or additionally to the free layer.

  A first region of the free layer is deposited by step 112. The first region of the free layer comprises a magnetic layer comprising Co, Fe and / or B. For example, a CoFeB layer with B not exceeding 20 atomic percent is deposited. In one embodiment, the thickness of such a ferromagnetic layer extends to 25 Å. In one embodiment, the ferromagnetic layer is at least 15 Å. However, in other embodiments, other thicknesses and / or other layers are possible.

  A sacrificial insertion layer is deposited by step 114 on the first ferromagnetic layer, as the layers share an interface. Thus, the sacrificial insertion layer comprises a material that has an affinity for boron, low diffusion, and relatively good lattice match with the lower CoFeB layer. For example, the difference in lattice parameter between the lower ferromagnetic layer and the sacrificial insertion layer is less than 10%. The sacrificial insertion layer is made of Bi, W, I, Zn, Nb, Ag, Cd, Hf, Os, Mo, Ca, Hg, Sc, Y, Sr, Mg, Ti, Ba, K, Na, Rb, Pb, and At least one of Zr is included. In one embodiment, the sacrificial insertion layer comprises Bi, W, I, Zn, Nb, Ag, Cd, Hf, Os, Mo, Ca, Hg, Sc, Y, Sr, Mg, Ti, Ba, K, Na, It is composed of Rb, Pb and / or Zr. The sacrificial insertion layer is thin, for example less than 10 Å thick. In such embodiments, the thickness of the sacrificial insertion layer does not exceed 4 Å and is greater than 1 Å. In other embodiments, other thicknesses are used.

  The sacrificial insertion layer and the lower layer are then annealed in step 116 at a temperature above room temperature. For example, RTA is used in the temperature range of 300-400 degrees Celsius. In other embodiments, the annealing is performed at other modes and / or at other temperatures. The annealing of step 116 is performed such that the lower CoFeB layer is crystallized to the preferred structure and orientation. Also, excess B in the CoFeB layer and / or excess oxygen in the ferromagnetic layer is absorbed by the sacrificial insertion layer during annealing.

  After annealing, step 118 removes the sacrificial insertion layer. For example, plasma etching is used. In another embodiment, the sacrificial insertion layer is removed by ion milling or other method including chemical mechanical planarization (CMP), but is not limited thereto. At step 118, a portion of the lower CoFeB layer is removed. After step 118, the remaining thickness of the CoFeB layer is required to be greater than 0 and not exceed 15 Å. In one embodiment, the remaining area of the CoFeB layer formed in step 112 does not exceed 20 Å. In such embodiments, the CoFeB layer does not exceed 10 Å after step 118. However, complete removal of the CoFeB layer is not required.

  The remaining area of the free layer is optionally deposited by the following step 120. For example, a second ferromagnetic CoFeB layer is deposited on the exposed first ferromagnetic layer. Thus, the first and second magnetic layers (eg, CoFeB layers) share an interface. Also, other layers including multiple layers are formed. Regardless of the total amount that the magnetic material is present, the free layer has a perpendicular magnetic anisotropy that exceeds the demagnetization energy. The remaining area of the first ferromagnetic layer has an overall thickness greater than 15 Å after the second ferromagnetic layer is provided together in steps 118 and 120. The total thickness of such a bilayer does not exceed 30 Å. In such embodiments, the overall thickness does not exceed 25 Å. For example, the overall thickness is at least 16 Å and less than 20 Å. In one embodiment, the thickness of each of the first and second ferromagnetic layers does not exceed a thickness of 15 Å.

  FIG. 6 is a diagram illustrating one embodiment of a magnetic junction 200 ′ ′ fabricated using method 110. Clearly, FIG. 6 is not a ratio of actual size, but is intended to aid understanding. The magnetic junction 200 ′ ′ is used in magnetic devices such as STT-MRAM and in various types of electronic devices. The magnetic junction 200 ′ ′ is similar to the magnetic junction 200. Thus, similar components are shown as well. The magnetic junction 200 ′ ′ comprises a free layer 210 with a magnetic moment 211, similar to the free layer 210 with a magnetic moment 211 shown in the magnetic junction 200, the nonmagnetic spacer layer 220 and the pinned layer 230 with a magnetic moment 231, It includes a nonmagnetic spacer layer 220 and a pinned layer 230 'having a magnetic moment (231A / 231B). Also, a lower selective seed layer 204 similar to the selective seed layer 204 of the magnetic junction 200 is shown. The seed layer 204 found in this embodiment is a crystalline MgO seed layer. The seed layer (MgO) 204 improves the perpendicular magnetic anisotropy of the free layer 210.

  FIG. 6 also shows the selective Fe insertion layer 260 and the selective PEL 270. For example, the selective PEL 270 is a CoFeB alloy layer, a FeB alloy layer, an Fe / CoFeB bilayer, a half metallic layer, or a Heusler alloy layer. Other high spin polarization materials are also provided. In one embodiment, selective PEL 270 is configured to improve the perpendicular magnetic anisotropy of pinned layer 230. Furthermore, the pinned layer 230 ′ is an SAF that includes ferromagnetic layers (232, 236) separated by a nonmagnetic layer 234. The ferromagnetic layers (232, 236) are coupled antiferromagnetically via the nonmagnetic layer 234. In one embodiment, the ferromagnetic layer 232 is one or more multilayers. The pinned layer 230 ′ is manufactured utilizing step 106 of the method 100. Thus, the area of the magnetic junction 200 ′ ′ is defined prior to being formed as an area of the pinned layer 230 ′. In another embodiment, layers (232, 234, 236) are deposited before the edge of the magnetic junction 200 '' is defined.

  The magnetic junction 200 ′ ′ shown in FIG. 6 is formed utilizing the method 110 of step 102 of the method 100. Thus, free layer 210 includes two regions separated by dotted lines. The lower region of the free layer 210 below the dotted line is deposited at step 112. A portion of this layer is removed at step 118. The upper region of free layer 210 above the dotted line is deposited at step 120. Although the free layer 210 is generally divided in half with dotted lines, the upper and lower dotted lines of the free layer 210 may have different proportions. Thus, free layer 210 is considered to include a single ferromagnetic layer having a thickness greater than 15 Å. However, regions of such ferromagnetic layers are deposited in other steps of method 110. In the embodiment shown in FIG. 6, the free layer 210 is comprised of a single ferromagnetic layer. In one embodiment, such a ferromagnetic layer is a CoFeB layer in which B does not exceed 20 atomic percent.

  In method 110, since free layer 210 is formed using a sacrificial insertion layer, free layer 210 is thick, has a stable state perpendicular to the plane for magnetic moment 211, and has enhanced magnetoresistance and low damping. Have. The sacrificial insertion layer and the annealing used in steps 116-118 improve the crystallinity of free layer 210. This is a high reluctance. Removal of the sacrificial insertion layer in step 118 prior to deposition of the remaining area of the free layer 210 improves the damping of the free layer 210. Thus, free layer 210 is fabricated with a large thickness while maintaining the crystal structure and vertical anisotropy. For example, the free layer 210 has a magnetic moment 211 greater than 15 Å but perpendicular to the plane. In one embodiment, free layer 210 does not exceed a thickness of 25 Å. For example, the free layer 210 is at least 16 Å thick and does not exceed 20 Å thick. Thus, the magnetic junction 200 '' has a higher magnetic resistance. Also, removal of the sacrificial insertion layer reduces attenuation in the free layer 210. That is, the free layer 210 exhibits low switch current. A smaller write current is used to program the magnetic junction. Thus, the performance is improved.

  In addition, the pinned layer 230 'improves the performance of the magnetic junction 200' 'in the magnetic device. In particular, the area of the magnetic junction comprising the layers (parts of 210, 260, 220, 270, 230 ') is defined above. The remaining area of the pinned layer 230 'will be defined later. Shadowing is mitigated while such definition steps are performed. Thus, the manufacturing process can be improved to realize a more tightly packed memory device.

  FIG. 7 is an illustration of an embodiment of a method 130 of fabricating a region of a magnetic junction usable with magnetic devices such as STT-RAM and various types of electronic devices. For the sake of simplicity, some steps may be omitted or performed separately or in combination. Also, method 130 may begin after other steps of magnetic memory formation have been performed. Method 130 is similar to the embodiment of step 106 of method 100. Thus, method 130 may begin after the free layer and the nonmagnetic spacer layer are provided.

  A first region of the pinned layer is deposited in step 132. Such first region of the anchored layer is a single layer or multilayer. For example, the first region of the pinned layer includes a magnetic layer comprising Co, Fe, and / or B. For example, a CoFeB layer is deposited which contains B not to exceed 20 atomic percent. Also, a PEL or other structure is deposited between the pinned layer and the nonmagnetic spacer layer. Also, a multilayer is deposited that includes a ferromagnetic layer sandwiched by nonmagnetic layers such as Co / Pt multilayers. If the pinned layer formed by the method 130 is an SAF, step 132 may be a magnetic (multi) layer, a magnet (multi) layer, a part or all of a nonmagnetic layer, or a magnetic (multi) layer, a nonmagnetic Depositing a layer, and a region of the upper magnetic (multi) layer. However, generally smaller areas of the pinned layer are deposited at step 132. This will allow a thinner structure to be defined in step 138 below.

  A sacrificial insertion layer is deposited by step 134 on the area of the formed fixed layer. The sacrificial insertion layer comprises a material that is compatible with boron, has low diffusion, and has a relatively good lattice match with the underlying layer. For example, the difference in lattice parameter between the lower ferromagnetic layer and the sacrificial insertion layer is less than 10%. For example, the sacrificial insertion layer may be made of Bi, W, I, Zn, Nb, Ag, Cd, Hf, Os, Mo, Ca, Hg, Sc, Y, Sr, Mg, Ti, Ba, K, Na, Rb, Pb And at least one of Zr. In one embodiment, the sacrificial insertion layer comprises Bi, W, I, Zn, Nb, Ag, Cd, Hf, Os, Mo, Ca, Hg, Sc, Y, Sr, Mg, Ti, Ba, K, Na, It is composed of Rb, Pb and / or Zr. The sacrificial insertion layer is thin. However, the sacrificial insertion layer is required to be continuous to be patterned as described later.

  The sacrificial insertion layer and the lower layer are then annealed in step 136. For example, RTA is used at temperatures in the range of 300-400 <0> C. In another embodiment, the annealing is performed in another manner. That is, the nonmagnetic spacer layer and the free layer, as well as the area of the pinned layer deposited in step 132 and the sacrificial insertion layer annealed in step 136, are disposed below the sacrificial insertion layer. Thus, the annealing temperature and other properties are required to be low enough to eliminate the negative impact on nonmagnetic spacer layers such as crystalline MgO tunnel barrier layers.

  After annealing, the area of the magnetic junction under the sacrificial insertion layer is photolithographically defined by step 138. Thus, step 138 includes providing a photoresist layer and patterning the photoresist layer to provide a photoresist mask. Also, other substances are used for the mask. The mask covers the area of the deposited layer that forms the area of the magnetic junction. The peripheral region of the magnetic junction is exposed. The edges of the magnetic junctions are defined using ion milling or other mechanisms to etch exposed areas of the layer. Ion milling is performed on the top of the sacrificial insertion layer at a small angle to the vertical.

  Thereafter, in step 140, a refill step is performed. That is, a nonmagnetic insulating layer such as alumina is deposited. Also, planarization is performed to provide a flat surface for subsequent processing.

  The sacrificial insertion layer is then removed in step 142. Step 142 is performed by plasma etching. Also, other removal methods are used. In the removing step, a partial region of the lower region of the fixed layer is removed. Optionally, the remaining area of the anchored layer is later deposited by step 144. For example, an additional ferromagnetic layer is deposited directly on the exposed first ferromagnetic layer. In embodiments where the pinned layer is an SAF, the layer is deposited depending on the proportion of the pinned layer deposited in step 132. For example, at step 132 the entire lower ferromagnetic layer (or multilayer) is step deposited, then at step 144 a nonmagnetic layer such as Ru and other magnetic layers are deposited. In another embodiment, the anchored layer is formed by another method.

  The remaining area of the magnetic junction is defined by step 146. Step 146 is performed by photolithography in a manner similar to step 138. However, because the free layer has already been defined in step 138, a low density pattern is used in step 146. Thus, the top of the magnetic junction is not wider than the bottom. In other embodiments, the upper region of the magnetic junction is the same size or wider than the lower region of the magnetic junction. In one embodiment, the upper region of the pinned layer is extended by multiple magnetic junctions.

  FIG. 8 is a diagram illustrating one embodiment of a magnetic memory including a magnetic junction 200 ′ ′ ′ fabricated using method 130. Clearly, FIG. 8 is not a ratio of actual size, but is intended to aid understanding. The magnetic junction 200 ′ ′ ′ is used in magnetic devices such as STT-MRAM and in various types of electronic devices. The magnetic junction 200 ′ ′ ′ is similar to the magnetic junction 200, the magnetic junction 200 ′, and / or the magnetic junction 200 ′ ′. However, the individual layers of the magnetic junction 200 '' 'are not shown for simplicity.

As can be seen from FIG. 8, the lower region of the magnetic junction 200 ′ ′ ′ defined in step 138 is separated by a distance d1. The top region of the magnetic junction 200 '''defined in step 146 is separated by a distance d2. Furthermore, the distance d ₁ <the distance d ₂ . Thus, the photoresist masks used in steps 138 and 146 have other densities. In another embodiment, the density is the same: distance d ₁ = distance d ₂ . Also, in other embodiments, the density of the mask used in step 146 is greater than the density of the mask used in step 138. Thus, in such an embodiment, the distance d ₁ > the distance d ₂ . In another embodiment, the top regions of the magnetic junctions 200 ′ ′ ′ are linked. In addition, the top and bottom aspect ratios, footprints, and other geometric parameters of the magnetic junction 200 '''are different. Also, although only three magnetic junctions are shown, in general, other numbers are manufactured together. Furthermore, generally, a two-dimensional array of magnetic junctions are fabricated together on a substrate. Only three lines are shown for clarity.

  Method 130 is used to improve the performance and fabrication process of the magnetic junction 200 ′ ′. The lower region of the magnetic junction 200 '' 'is previously defined. The remaining area of the pinned layer 230 'will be defined later. The area of the stack defined in steps 138 and 146 is thin. As a result, shadowing is mitigated during such definition steps. Thus, the lower region of the magnetic junction 200 '' 'is packed to be more contiguous and is better defined. The upper region of the magnetic junction 200 '' 'does not include the free layer. The spacing between such regions of the magnetic junction 200 '"is less important. Such areas may be further separated. Thus, better process control and integration is achieved. Furthermore, the separated configuration of such sections of the magnetic junction 200 '' 'may be shaped to improve performance. As a result, manufacturing may be improved and more closely packed memory devices may be embodied. Performance is further improved if the free layer of the magnetic junction 200 ′ ′ ′ is fabricated using the method 110.

  FIG. 9 is an illustration of an embodiment of a method 150 of fabricating a magnetic junction usable in magnetic devices such as STT-RAM and various types of electronic devices. For the sake of simplicity, some steps may be omitted or performed separately or in combination. In addition, method 150 may begin after other steps in magnetic memory formation have been performed. 10-24 illustrate an embodiment of an in-process magnetic junction utilizing method 150. FIG. 10 to 24 are not the ratio of the actual size.

  Step 152 deposits a crystalline MgO seed layer. In one embodiment, step 152 forms one nonmagnetic spacer layer, such as a double magnetic junction. Thus, the pinned layer is disposed below the crystalline MgO layer. In another embodiment, the layer deposited in step 152 is a seed layer for the lower magnetic junction.

  Step 154 deposits a first ferromagnetic CoFeB layer of the free layer. Such layers are similar to steps 102 and 112 described above. In one embodiment, the ferromagnetic layer is at least 15 Å. However, in other embodiments, different thicknesses and / or different layers are also possible. FIG. 10 shows the magnetic junction 300 after performing step 154. That is, the seed layer (MgO) 302 of the free layer and the first ferromagnetic layer (CoFeB) 312 are shown.

  In step 156, a sacrificial insertion layer is deposited on the first ferromagnetic layer 312. That is, step 156 is similar to step 114. Accordingly, the material and thickness of the sacrificial insertion layer are as described above. FIG. 11 shows the magnetic junction 300 after step 156 has been performed. That is, a sacrificial insertion layer (W) 304 is shown. In one embodiment, the material and thickness of the sacrificial insertion layer 304 are similar to those of the methods 100 and 110 described above.

  The layers (302, 304, 312) are then annealed according to step 158. For example, RTA is used at temperatures in the range of 300-400 <0> C. That is, the annealing step 158 is similar to that of step 116. After annealing, the sacrificial insertion layer 304 is removed by step 160. Step 160 is similar to step 118. For example, plasma etching is used. FIG. 12 shows the magnetic junction 300 after step 160 has been performed. That is, the sacrificial insertion layer 304 is removed. In addition, a partial region of the first ferromagnetic layer 312 'is removed. This shows a slightly thinner ferromagnetic layer 312 '.

  In one embodiment, the remaining area of the free layer is deposited by step 162. For example, a second ferromagnetic CoFeB layer is deposited on the exposed first ferromagnetic layer 312 '. FIG. 13 shows the magnetic junction 300 after step 162. That is, the second ferromagnetic layer (CoFeB) 314 is deposited. The layers (312 ', 314) combine to form the free layer 310.

  A nonmagnetic spacer layer is provided by step 164. In one embodiment, a crystalline MgO barrier layer is provided at step 164. FIG. 14 shows the magnetic junction 300 after step 164 has been performed. That is, the nonmagnetic spacer layer (MgO) 320 is manufactured.

  Step 166 deposits a first region of the pinned layer. Step 166 is similar to step 132. Thus, a single layer or multilayer comprising a ferromagnetic layer and / or a nonmagnetic layer is deposited. FIG. 15 shows the magnetic junction 300 after step 166. That is, the ferromagnetic layer 332 is shown. In the embodiment shown in FIGS. 15-24, the entire lower layer / SAF pinned layer multilayer is provided at step 166. However, in other embodiments, more or less layers of ferromagnetic layer 332 are deposited at step 166.

  An additional sacrificial insertion layer is deposited on the ferromagnetic layer 332 by step 166 '. Step 166 'is similar to step 134, so the materials and thicknesses described above are used. FIG. 16 shows the magnetic junction 300 after step 166 'has been performed. That is, a sacrificial insertion layer (W) 306 is shown.

  Step 168 anneals the layers (302, 312 ', 314, 320, 306). Step 168 is similar to step 136. For example, RTA is performed at the above-mentioned temperature. The annealing temperature and other properties are required to be low enough so that nonmagnetic spacer layers such as crystalline MgO tunnel barrier layers are not adversely affected.

  After annealing, the region of the magnetic junction 300 under the sacrificial insertion layer is photolithographically defined by step 170. Step 170 is similar to step 138. FIG. 17 shows the magnetic junction in step 170. That is, a mask 360 is provided on the sacrificial insertion layer 306. FIG. 18 shows the magnetic junction after step 170. That is, the area of two magnetic junctions is defined. Specifically, the free layer 310, the nonmagnetic layer 320, and the ferromagnetic layer 332 are defined.

  Thereafter, in step 172, a refill step is performed. Thereby, a nonmagnetic insulating layer such as alumina is deposited and planarized. Step 172 is similar to step 140. 19 and 20 show the magnetic junction during and after step 172. FIG. That is, the refill material 308 is shown in FIG. FIG. 20 shows the magnetic junction 300 after step 172 is completed. That is, the top surface of the refill material 308 is planarized.

  Thereafter, the sacrificial layer is removed in step 174. Step 174 is similar to step 142. Optionally, the remaining area of the anchored layer is later deposited by step 176. Step 176 is similar to step 144. FIG. 21 illustrates one embodiment of the magnetic junction 300 after step 174 is completed. In the illustrated embodiment, the entire lower ferromagnetic layer (or multilayer) 332 is deposited at step 166. Thus, nonmagnetic layers such as Ru and other magnetic layers are deposited at step 176. That is, a nonmagnetic layer such as the Ru layer 334 and a ferromagnetic layer 336 are shown. It can be seen that layers (334, 336) extend across the two magnetic junctions 300. Layers (332, 334, 336) form a SAF pinned layer.

  Step 178 defines the remaining area of the magnetic junction. Step 178 is similar to step 146. Step 178 is performed by a photolithographic method similar to step 170. However, a different density pattern is used in step 178 because the free layer has already been defined in step 170. Thus, the top of the magnetic junction is not wider than, the same size as the bottom, or wider than the bottom. In one embodiment, the upper region of the pinned layer is extended by a multilayer magnetic bond. FIG. 22 illustrates one embodiment of the magnetic junction 300 after step 178 has been performed. That is, the layer to be fixed 330 is defined. In the illustrated embodiment, the top of the pinned layer 330 is the same size as the bottom.

  Figures 23 and 24 illustrate one embodiment of a magnetic junction 300 'that includes a portion of the layer 332 deposited in step 166. FIG. 23 shows an embodiment after step 176 has been performed. That is, the layers (333, 334, 336) are shown. The layers (333, 331) together form the lower ferromagnetic layer 332 'of the SAF pinned layer 330'. FIG. 24 shows the magnetic junction after performing step 178. That is, the upper region of the magnetic junction 300 'is defined.

  The magnetic junctions (300, 300 ′) share the effects of the magnetic junction 200, the magnetic junction 200 ′, the magnetic junction 200 ′ ′, and / or the magnetic junction 200 ′ ′ ′. Thus, the magnetic junction (300, 300 ') has enhanced reluctance, reduced damping, and switch current, and is packed more closely with the magnetic device, either together or separately.

  25 illustrates a magnetic memory 400 using at least one of magnetic junction 200, magnetic junction 200 ′, magnetic junction 200 ′ ′, magnetic junction 200 ′ ′ ′, magnetic junction 300, and / or magnetic junction 300 ′. 1 shows an embodiment. The magnetic memory 400 includes not only the read / write column select driver (402, 406) but also the word line select driver 404. It is noted that other and / or different configurations are provided. The storage area of the magnetic memory 400 includes magnetic storage cells 410. Each of the magnetic storage cells includes at least one magnetic junction 412 and at least one selection element 414. In one embodiment, select element 414 is a transistor. Magnetic junction 412 may be one of magnetic junction 200, magnetic junction 200 ′, magnetic junction 200 ′ ′, magnetic junction 200 ′ ′ ′, magnetic junction 300, and / or magnetic junction 300 ′ disclosed herein. is there. Although the illustrated embodiment shows one magnetic junction 412 per magnetic storage cell 410, in other embodiments different numbers of magnetic junctions 412 are provided per cell. Thereby, the magnetic memory 400 has the above-described effect.

  A method and system for providing magnetic junctions and memory fabricated using the above magnetic junctions has been described. Although the above method and system have been described in accordance with the illustrated exemplary embodiment, those of ordinary skill in the art to which the present invention belongs can modify the above embodiment, and It is easily understood that any changes can be made without departing from the spirit and scope. Accordingly, the present invention can be variously modified by those skilled in the art to which the present invention belongs without departing from the spirit and scope of the claims.

200, 200 ', 200'',200''', 200 '''', 300, 300 ', 412 magnetic junction 201 substrate 202 lower contact 204, 302 (selective) seed layer (MgO)
206 (Selective) capping layer 208 top contact 210, 310 free layer 220, 240, 320 nonmagnetic spacer layer (MgO)
230, 230 ', 250 pinned layers 211, 231, 251 magnetic moments 232, 236, 331, 332, 333 ferromagnetic layers 234 nonmagnetic layers 260 selective Fe insertion layers 270 selective PEL
304, 306 sacrificial insertion layer (W)
308 Refill (substance)
312, 312 'first ferromagnetic layer (CoFeB)
314 Second ferromagnetic layer (CoFeB)
332 'lower ferromagnetic layer 334 Ru layer 360 mask 400 magnetic memory 402, 406 read / write column select driver 404 word line select driver 410 magnetic storage cell 414 select element

Claims (16)

A method of providing a magnetic junction usable in a magnetic device on a substrate, the method comprising:
Providing a free layer that is switched between the plurality of stable magnetic states as the write current flows through the magnetic junction;
Providing a nonmagnetic spacer layer;
Providing a pinned layer.
The nonmagnetic spacer layer is disposed between the free layer and the pinned layer,
At least one of the steps of providing the free layer and providing the fixed layer includes a plurality of steps,
The at least one step of providing the free layer comprises a first plurality of steps,
At least one step of providing the fixed layer includes a second plurality of steps,
The first plurality of steps are:
Depositing a first region of the free layer;
Depositing a first sacrificial layer;
Annealing at least a first region of the free layer and the first sacrificial layer at a first temperature above 25 degrees Celsius;
Removing the first sacrificial layer;
Depositing the second region of the free layer;
The second plurality of steps are:
Depositing a first region of the layer to be fixed;
Depositing a second sacrificial layer;
Annealing at least a first region of the pinned layer and the second sacrificial layer at a second temperature higher than 25 ° C .;
Defining a region of a magnetic junction including the free layer, the nonmagnetic spacer layer, and a first region of the pinned layer;
Removing the second sacrificial layer;
Look including the steps of: depositing a second region of the object to be fixed layer,
The step of providing a pinned layer comprising the second plurality of steps further comprises the step of depositing at least one refill material prior to the step of removing the second sacrificial layer. How to provide bonding. Providing the free layer comprises the first plurality of steps,
The method according to claim 1, wherein the free layer has perpendicular magnetic anisotropy energy greater than out-of-plane demagnetization energy.   3. The method of claim 2, wherein the free layer has a thickness greater than 15 Å.   4. The method of claim 3, wherein the thickness of the free layer is 25 Å or less.   The first sacrificial layer is made of Bi, W, I, Zn, Nb, Ag, Cd, Hf, Os, Mo, Ca, Hg, Sc, Y, Sr, Mg, Ti, Ba, K, Na, Rb, Pb The magnetic junction providing method according to claim 2, further comprising at least one of Zr and Zr.   3. The method of claim 2, further comprising depositing a MgO seed layer prior to providing the free layer.   The method of claim 2, wherein the annealing comprises performing a rapid thermal anneal (RTA). The first region of the free layer has a first thickness, and the second region of the free layer has a second thickness,
3. The method of claim 2, wherein the first thickness is less than 15 angstroms and the second thickness is less than 15 angstroms. The method of claim 1 , further comprising performing planarization after depositing the at least one refill material. The pinned layer is a synthetic antiferromagnetic material that includes a first ferromagnetic layer, a second ferromagnetic layer, and a coupling layer between the first ferromagnetic layer and the second ferromagnetic layer. Yes,
Depositing the second region of the layer to be fixed;
Depositing at least one nonmagnetic spacer layer;
Magnetic junction method of claim 1, characterized in that it comprises a step of depositing the second ferromagnetic layer. The method of claim 10 , wherein depositing the second region of the pinned layer further comprises depositing a region of the first ferromagnetic layer. The method of claim 10 , wherein at least one of the first ferromagnetic layer and the second ferromagnetic layer is a multilayer. The method of claim 10 , further comprising defining a remaining area of the pinned layer. Providing the pinned layer comprising the second plurality of steps further comprises defining a region of the magnetic junction prior to depositing the at least one refill material,
Defining the area of the magnetic junction comprises
Providing a photoresist mask over the second sacrificial layer covering a region of the second sacrificial layer corresponding to the magnetic junction;
Removing the exposed region of the second sacrificial layer, the first region of the pinned layer, the nonmagnetic spacer layer, and the free layer exposed by the photoresist mask. magnetic junction providing method according to claim 1. Providing an additional nonmagnetic spacer layer;
Providing an additional anchored layer,
The free layer is between the additional nonmagnetic spacer layer and the nonmagnetic spacer layer,
The method of claim 1, wherein the additional nonmagnetic spacer layer is between the additional pinned layer and the free layer. A method of providing magnetic memory usable in a magnetic device on a substrate, the method comprising:
Depositing on the substrate a first ferromagnetic layer of a free layer comprising a CoFeB layer which is less than 15 Å thick;
On the first ferromagnetic layer, Bi, W, I, Zn, Nb, Ag, Cd, Hf, Os, Mo, Ca, Hg, Sc, Y, Sr, Mg, Ti, Ba, K, Na, Rb Depositing a first sacrificial layer containing at least one of: Pb, Pb, and Zr and having a thickness of 4 Å or less;
Annealing at least the first ferromagnetic layer and the first sacrificial layer by a first rapid thermal anneal (RTA) at a first temperature above 25 ° C .;
Removing at least the first sacrificial layer;
A second strength of the free layer comprising a CoFeB layer of 15 Å thickness or less on the remaining region of the first ferromagnetic layer so as to have a thickness of 25 Å or less combined with the remaining region of the first ferromagnetic layer Depositing a magnetic layer;
Providing a nonmagnetic spacer layer formed between the pinned layer and the free layer;
Depositing a first region of the layer to be fixed;
At least one of Bi, W, I, Zn, Nb, Ag, Cd, Hf, Os, Mo, Ca, Hg, Sc, Y, Sr, Mg, Ti, Ba, K, Na, Rb, Pb, and Zr Depositing a second sacrificial layer that is less than 4 Å thick including one;
At least a first region of the pinned layer, a remaining region of the first ferromagnetic layer, the second ferromagnetic layer, and the second sacrificial layer are subjected to a second rapid thermal annealing (RTA) to a temperature higher than 25 ° C. Annealing at a temperature,
Providing a photoresist mask on the sacrificial layer covering a region of the sacrificial layer corresponding to the at least one magnetic junction after the step of annealing by the second rapid thermal annealing;
Defining a region of at least one magnetic junction including the free layer, the nonmagnetic spacer layer, and a first region of the pinned layer using the photoresist mask;
Depositing at least one refill material;
Performing a planarization after the depositing of the at least one refill material;
Removing the second sacrificial layer after the planarizing step;
Depositing at least a second region of the layer to be fixed;
After depositing at least a second region of the pinned layer, defining a remaining region of the at least one magnetic junction.
The magnetic memory characterized in that the free layer has perpendicular magnetic anisotropy energy greater than out-of-plane demagnetization energy and is switched between a plurality of stable magnetic states as a write current flows through the magnetic junction. How to provide.

Images