Magnetic tunnel junction structure and method for manufacturing the same
By using the chromium seed crystal layer and cobalt multilayer structure combined with the SAF coupling layer in the magnetic tunnel junction, the problem of SDF coupling loss in MTJ at high temperature is solved, and good magnetic memory performance is maintained at high temperature.
Patent Information
- Application Number
- CN201980030970.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2019-04-05
- Publication Date
- 2024-11-29
- Estimated Expiration
- 2039-04-05
AI Technical Summary
Magnetic tunnel junction (MTJ) is prone to lose SAF coupling after high temperature treatment, resulting in a degradation of magnetic memory performance. It is difficult for the prior art to effectively maintain the magnetic and electrical characteristics of MTJ stacking.
A multi-layer structure made of chromium (Cr)-based seed crystal layer and cobalt (Co) is used as the first and second pinning layers, and a synthetic anti-ferromagnetic (SAF) coupling layer and structural barrier layer are combined to form a physical vapor deposition ( PVD) technology forms an MTJ stack to ensure that perpendicular magnetic anisotropy and electrical characteristics are maintained at high temperatures.
At high temperatures above 400°C, MTJ stacking can maintain 100% to 175% tunnel magnetoresistance (TMR) and maintain good performance after annealing of 0.5 hours to 3 hours, solving the loss problem of SAF coupling and improving the Memory stability and performance.
Smart Images

Figure CN112166509B_ABST
Abstract
Description
Technical Field
[0001] Embodiments of the present disclosure generally relate to fabricating magnetic tunnel junction structures for magnetic random access memory (MRAM) applications. Background Art
[0002] Spin transfer torque magnetic random access memory, or STT-MRAM, employs a magnetic tunnel junction structure in its memory cell in which two ferro-magnetic layers are separated from each other by a thin insulating or "dielectric" layer. One of the magnetic layers has a fixed magnetic polarity and the other has a magnetic polarity that can be selectively changed between two states. In the case where the magnetic layers have perpendicular magnetic anisotropy, the polarity of the variable polarity layer can be switched between having the same polarity as the fixed polarity layer or having a polarity opposite to the polarity of the fixed polarity layer in the depth direction of the film layer stack comprising the magnetic tunnel junction or "MTJ" structure. The resistance across the MTJ is a function of the polarity of the variable polarity layer relative to the fixed polarity layer. The resistance across the MTJ is low when the polarities of the two layers are the same in the depth direction of the MTJ, and high when they are opposite to each other in the depth direction of the MTJ. Thus, the resistance across a memory cell can be used to indicate a value of 1 or 0, and thus store a data value, such as by using a low resistance state to represent a data value of 1 and a high resistance state to represent a data value of 0.
[0003] In the case of an MTJ that is a synthetic anti-ferromagnet (SAF) comprising two or more ferromagnetic layers separated by a nonmagnetic layer, SAF coupling may be lost after high temperature processing thereof, for example, at or above about 400°C.
[0004] To form the MTJ stack, a film layer stack including a seed layer is used to ensure proper crystal orientation in the overlying pinned layer, here face-centered cubic <111> The seed layer contains platinum (Pt) or ruthenium (Ru) to establish a perpendicular magnetic isotropy or PMA therein. To properly form an MRAM cell, the film stack is annealed at about 400°C for about 0.5 to 3 hours. However, during this annealing, the crystallographic lattice and the resulting crystallographic and magnetic polarity orientation in the pinned layer are lost, or at least partially lost, resulting in a loss of the desired (SAF) coupling.
[0005] Therefore, there remains a need for improved MTJ stacks. Summary of the invention
[0006] The present disclosure generally relates to the design and manufacture of a magnetic tunnel junction (MTJ) stack for a memory cell. In one example, a device includes a magnetic tunnel junction stack. The magnetic tunnel junction stack includes a seed layer and a first pinned layer, the seed layer contacts a buffer layer, wherein the seed layer includes chromium (Cr), and the first pinned layer contacts the seed layer. The first pinned layer is formed by a plurality of bilayers, wherein each bilayer of the plurality of bilayers includes: a first interlayer formed of cobalt (Co), and a second interlayer formed of platinum (Pt), iridium (Ir), nickel (Ni), or palladium (Pd). The MTJ stack further includes: a coupling layer and a second pinned layer, the coupling layer contacts the first pinned layer, and the second pinned layer contacts the coupling layer.
[0007] In another example, a magnetic tunnel junction stack includes: a seed layer formed on a buffer layer, and a first pinned layer formed on the seed layer, wherein the seed layer includes chromium (Cr), wherein the first pinned layer includes a first plurality of bilayers. Each bilayer in the first plurality of bilayers includes: a first interlayer formed of cobalt (Co), and a second interlayer formed of at least one of platinum (Pt), iridium (Ir), nickel (Ni), or palladium (Pd). The MTJ stack may further include: a coupling layer formed on the first pinned layer; and a second pinned layer formed on the coupling layer, wherein the second pinned layer includes cobalt (Co).
[0008] In another example, a magnetic tunnel junction stack includes: a seed layer and a first pinning layer, wherein the seed layer is formed on the buffer layer, wherein the seed layer includes chromium (Cr) and is to and a first pinning layer is formed on the seed layer, the first pinning layer having a thickness from to The MTJ stack further comprises: a coupling layer formed between the first pinned layer and the first cobalt capping layer; a second pinned layer formed on the coupling layer; and a first cobalt capping layer having a thickness of to The second pinned layer comprises a second plurality of bilayers; and a structural barrier layer formed on the second pinned layer. The structural barrier layer comprises at least one of tantalum (Ta), molybdenum (Mo) or tungsten (W), and is formed from to The MTJ stack further includes: a magnetic reference layer formed on the structural barrier layer; a tunnel barrier layer formed on the magnetic reference layer; and a magnetic storage layer formed on the tunnel barrier layer. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] The above briefly summarizes the manner in which the above-described features of the present disclosure can be understood in detail, as well as a more specific description of the present disclosure, which can be obtained by referring to the embodiments, some of which are shown in the accompanying drawings. However, it should be noted that the drawings only depict exemplary embodiments and are not to be construed as limiting the scope of the present invention, which may allow for other equally effective embodiments.
[0010] Figure 1A Schematic diagram of an example magnetic tunnel junction (MTJ) stack.
[0011] Figure 1B Is manufactured including Figure 1A A flow chart of a method of manufacturing a magnetic tunnel junction (MTJ) stack and a memory device according to an embodiment of the present disclosure.
[0012] Figure 2A is a schematic diagram of an MTJ stack according to an embodiment of the present disclosure.
[0013] Figure 2B is an enlarged view of a buffer layer of an MTJ stack according to an embodiment of the present disclosure.
[0014] Figure 2C is an enlarged view of a first pinned layer of an MTJ stack according to an embodiment of the present disclosure.
[0015] Figure 2D is an enlarged view of the second pinned layer of the MTJ stack according to an embodiment of the present disclosure.
[0016] Figure 2E is an enlarged view of an example magnetic storage layer of an MTJ stack according to an embodiment of the present disclosure.
[0017] Figure 2F is an enlarged view of an example capping layer of an MTJ stack according to an embodiment of the present disclosure.
[0018] To facilitate understanding, identical reference numerals have been used, where possible, to designate identical elements that are common to the figures. It is contemplated that elements and features of one embodiment may be beneficially incorporated in other embodiments without further description. DETAILED DESCRIPTION
[0019] Embodiments of the present disclosure relate to magnetic tunnel junction (MTJ) stacks and STT MRAM memory cells and memories. Here, the MTJ stack is combined in a film stack including an upper electrode and a lower electrode, wherein the MTJ stack is sandwiched between the upper electrode and the lower electrode, and can be used to form a plurality of memory cells for magnetoresistive random-access memory (MRAM). In each MTJ stack of the MRAM, there are two magnetic layers, one of which has a fixed polarity and the other has a polarity that can be switched by applying a voltage to the layer or by applying a current to the magnetic layer. Based on the relative polarity between the first and second magnetic layers, the resistance varies across the MRAM. The first and second magnetic layers are referred to herein as magnetic reference layers and magnetic storage layers. The memory cell formed by the MTJ stack operates when a voltage is applied to the memory cell or a current passes through the memory cell. The polarity of the switchable magnetic layer can change in response to a voltage application of sufficient strength. In addition, the resistivity of the memory cell can be determined by measuring the current-to-voltage relationship on the memory cell at a relatively low voltage below the threshold required to switch the magnetic polarity of the magnetic storage layer.
[0020] The basic MTJ stack discussed herein is formed using multiple deposition chambers, which are used to deposit thin film layers on a substrate and ultimately pattern and etch the deposited film layers. The deposition chamber used to form the MTJ stack discussed herein includes a physical vapor deposition (PVD) chamber. Here, the PVD chamber is used to form multiple thin film layers of the MTJ stack. Therefore, the MTJ stack includes a buffer layer, a seed layer on the buffer layer, a first pinned layer on the seed layer, a synthetic anti-ferromagnetic (SAF) coupling layer on the first pinned layer, a second pinned layer on the SAF layer, and a barrier layer on the second pinned layer. In the PVD operation described herein, an inert gas or a rare gas such as argon (Ar), helium (He), krypton (Kr) and / or xenon (Xe) is ionized into a plasma in a sputtering chamber while maintaining the chamber in a vacuum state. The PVD processing chamber also contains at least one sputtering target, and the substrate is disposed facing a substantially flat surface of the sputtering target. The sputtering target is coupled to a power source so that the sputtering target is electrically driven or self-established as a cathodic state in a circuit of the power source, through the plasma to a ground, such as a grounded portion of the sputtering chamber. The substrate is disposed on a pedestal or other structure in the sputtering chamber, which may be at a floating potential, connected to ground, or may be biased to form an anode in the cathodic target to create a plasma to anode or ground circuit. The positively ionized portion of the noble gas atoms in the sputtering chamber are electrically attracted to the negatively biased target, and therefore, the ions of the plasma bombard the target, which causes atoms of the target material to be ejected and deposited on the substrate to form a thin film composed of the target material(s) on the substrate.
[0021] In order to form a thin film of the compound, a sputtering target including the compound can be used in a PVD chamber with Ar plasma. In another example, there are multiple sputtering targets in the PVD chamber, each sputtering target includes one or more elements of the compound to be formed as a film on the substrate, and is used with Ar plasma to form the desired compound. In addition, in the PVD operation used to form the layers of the MTJ stack herein, metal oxides and metal nitrides are formed using metal oxide or metal nitride sputtering targets. In an alternative embodiment, in the presence of Ar plasma and oxygen (O2) or nitrogen (N2) in the PVD chamber, a metal oxide or metal nitride layer is formed in the PVD chamber by sputtering one or more metal sputtering targets composed of the metal of the metal oxide or nitride. In one example, a PVD chamber has a plurality of sputtering targets disposed therein, each of the sputtering targets in the PVD chamber being biased by a power supply to establish a negative bias thereon, or by directly applying a negative DC bias to the sputtering target, or by using a waveform to electrically drive or self-establish a cathodic state on the target, or a combination thereof. In this example, a shield within the PVD chamber is configured to block one or more of the plurality of targets while allowing ions of a plasma formed in the PVD chamber to bombard at least one of the targets to eject or sputter target material atoms from the at least one target to form a thin film on a substrate while protecting other sputtering targets in the chamber from the plasma. In this example, one or more targets are exposed to Ar plasma, and the sputtering of the targets sequentially or simultaneously forms a desired film composition on a substrate, and separate PVD chambers are used to form metal oxide and metal nitride layers when O2 or N2 are used together with Ar plasma to form film layers on a substrate, respectively.
[0022] The PVD system may include one or more PVD sputtering chambers coupled to a central robotic substrate transfer chamber. The central robotic substrate transfer chamber is configured to move substrates between a loading station coupled thereto and the sputtering chambers connected thereto. The PVD system is maintained at, for example, 10E -9The process pressure in the one or more PVD chambers may be from about 2 mTorr to about 3 mTorr for the substrate to be formed thereon. The process pressure in the one or more PVD chambers may be from about 2 mTorr to about 3 mTorr for the substrate to be formed thereon. The process pressure in the one or more PVD chambers may be from about 2 mTorr to about 3 mTorr for the substrate to be formed thereon. The process pressure in the one or more PVD chambers may be from about 2 mTorr to about 3 mTorr for the substrate to be formed thereon. Depending on the implementation, the chamber of the PVD platform is maintained at -200° C. to 600° C. during fabrication of at least the seed layer, first and second pinned layers, SAF layer, and buffer layer of the MTJ stack.
[0023] Figure 1A is a schematic diagram of a magnetic tunnel junction (MTJ) stack. Figure 1A A conventional MTJ stack 100A is shown, and the MTJ stack 100A includes a substrate 102 having a conductive layer of tungsten (W), tantalum nitride (TaN), or titanium nitride (TiN) or having other metal layers formed of tungsten (W), tantalum nitride (TaN), or titanium nitride (TiN). In some examples, the substrate 102 includes one or more transistors, bit lines or source lines, and other storage lines previously manufactured therein or thereon, or other elements to be used in an MRAM memory and previously manufactured or formed thereon. The substrate on which the MTJ stack is formed may have dimensions that may include: a diameter of less than 200 mm, a diameter of 200 mm, a diameter of about 300 mm, a diameter of about 450 mm, or another diameter, and may have a circular or rectangular or square panel shape.
[0024] The buffer layer 104 in the conventional MTJ stack 100A is formed on the substrate 102 by sputtering one or more targets in a PVD chamber, and includes one or more of the following layers: Co x Fe y B z, TaN, Ta or a combination of the foregoing, and the substrate is located in the PVD chamber. The seed layer 106 is formed above the buffer layer 104 via sputtering in the PVD chamber. The buffer layer 104 is used in the conventional MTJ stack 100A to improve the adhesion of the seed layer 106 to the substrate. The seed layer 106 here includes platinum (Pt) or ruthenium (Ru), and is formed by sputtering a target of Pt or Ru or an alloy of the foregoing in a PVD chamber in which the substrate is located. By reducing or eliminating the lattice mismatch between the buffer layer 104 and the seed layer 106, the seed layer 106 can be used to improve the adhesion and seeding of subsequently deposited layers in the conventional MTJ stack 100A.
[0025] A first pinned layer 108 is formed on the seed layer 106 by sputtering. Here, the first pinned layer 108 includes a cobalt (Co) layer, one or more Co-containing bilayers, or a combination of a cobalt layer and one or more Co-containing bilayers. A synthetic antiferromagnetic (SAF) coupling layer 110 is formed above the first pinned layer 108 by sputtering. The SAF coupling layer 110 may be formed of ruthenium (Ru), rhodium (Rh), Cr, or iridium (Ir) sputtered from a target thereof. A second pinned layer 112 is formed above the SAF coupling layer 110 by sputtering. The second pinned layer 112 is formed here by a single cobalt (Co) layer, at least one cobalt-platinum bilayer, or a combination of a cobalt layer and one or more Co-containing bilayers, such as a Co-Pt bilayer, formed by sputtering targets composed of Co, Pt, or Co-Pt, respectively. The SAF coupling layer 110 is located between the first pinned layer 108 and the second pinned layer 112, and when exposed to a magnetic field, causes the surface atoms of the first pinned layer 108 and the second pinned layer 112 to align with the surface atoms of the SAF coupling layer 110, thereby pinning the orientation of the first pinned layer 108 and the second pinned layer 112. The first pinned layer 108 and the second pinned layer 112 each exhibit similar magnetic moments, and will therefore react similarly when an external magnetic field is applied to the conventional MTJ stack 100A. The SAF coupling layer 110 maintains the anti-parallel alignment of the magnetic moments of the first pinned layer 108 and the second pinned layer 112.
[0026] The structural barrier layer 114 is formed over the second pinned layer 112 and includes tantalum (Ta), molybdenum (Mo), tungsten (W), or a combination thereof. The structural barrier layer 114 is utilized because of its crystalline structure, which is different from the crystalline structure of the first pinned layer 108 and the second pinned layer 112. The structural barrier layer 114 prevents a short circuit from being formed between the conventional MTJ stack 100A and a metal contact that may be coupled to the conventional MTJ stack 100A to form an MRAM memory cell.
[0027] In addition, in the conventional MTJ stack 100A, a magnetic reference layer 116 is formed over the structural barrier layer 114 by sputtering in a PVD chamber. A tunnel barrier layer 118 is formed over the magnetic reference layer 116, and a magnetic storage layer 120 is formed over the tunnel barrier layer 118. Each of the tunnel barrier layer 118, the magnetic reference layer 116, and the magnetic storage layer 120 is formed by sputtering in one or more PVD chambers in the presence of Ar plasma. The magnetic reference layer 116 and the magnetic storage layer 120 each include Co with a variable composition. x Fe y B z alloy. In addition, the magnetic storage layer 120 may include one or more layers of Ta, Mo, W, or Hf, or a combination of the foregoing. The tunnel barrier layer 118 includes an insulating material and may be made of a dielectric material such as MgO. The composition and thickness of the tunnel barrier layer 118 may be selected to produce a large tunnel magnetoresistance ratio (TMR) in the tunnel barrier layer 118 of the conventional MTJ stack 100A. TMR is the ratio of the magnetic field in the conventional MTJ stack 100A from the anti-parallel state (R ap ) to parallel state (R p ) is a measure of the resistance change and can be expressed using the formula ((R ap -R p ) / R p When a bias voltage is applied to the conventional MTJ stack 100A, the tunnel barrier layer 118 is penetrated by spin-polarized electrons, and the transmission of such electrons through the tunnel barrier layer 118 results in conduction between the magnetic reference layer 116 and the magnetic storage layer 120 .
[0028] A capping layer 122 is formed on the magnetic storage layer 120 by sputtering in a PVD chamber, and the capping layer 122 includes a plurality of interlayers. The plurality of interlayers of the capping layer 122 include a first capping interlayer 122A made of a dielectric material such as MgO. A second capping interlayer 122B including a metal material such as Ru, Ir, Ta, or a combination thereof is formed over the first capping interlayer 122A. The first capping interlayer 122A acts as an etch stop layer for hard mask etching and protects the MTJ stack 100A from corrosion. The second capping interlayer 122B is configured to be electrically connected to a transistor or a contact when the conventional MTJ stack 100A is subsequently patterned, such as Figure 1B As shown and discussed below, a hard mask layer 124 is formed in a PVD chamber by sputtering and is formed over the second capping interlayer 122B to protect the conventional MTJ stack 100A and may be patterned during subsequent operations.
[0029] Figure 1B1 is a flow chart of a method 100B for manufacturing a memory device, the memory device including an MTJ stack 100A, and the MTJ stack is manufactured according to an embodiment of the present disclosure and shown in FIG. Figures 2A to 2F The method 100B is performed in part in a plurality of PVD chambers of a PVD system configured to deposit thin film layers by sputtering. The substrate 102 may be moved in and between sputtering chambers via a central robotic transfer chamber of the PVD system to form various thin film layers, including Figure 1A 100B and those thin film layers made according to embodiments of the present disclosure and shown and discussed below. In another example, as discussed above, multiple sputtering targets are disposed in a PVD chamber, and a shield within the PVD chamber is configured to selectively protect the multiple sputtering targets from exposure to plasma or to expose the targets to plasma. The shield is rotated under different operations of method 100B to expose two or more targets to plasma in the PVD chamber sequentially or simultaneously.
[0030] Therefore, this article refers to method 100B. Figure 1A The operations of method 100B are performed in one or more PVD chambers using one or more gases including argon (Ar), helium (He), krypton (Kr), xenon (Xe), oxygen (O2), or nitrogen (N2). During method 100B, the process pressure in the PVD chamber may be from about 2 mTorr to about 3 mTorr. Depending on the implementation, the PVD chamber may be maintained at -200° C. to 600° C. during the fabrication of the pinning layer and the seed layer of the MTJ stack.
[0031] Depending on the composition of the sputtering target(s) used for the various layers of the MTJ stack 100A, the substrate 102 may be moved in and between PVD chambers, or as discussed herein, multiple targets may be coupled to a power source and shields may be configured to selectively protect some targets such that one or more targets are exposed sequentially or simultaneously to form a desired film composition, or both approaches may be performed. During sputtering in the PVD chamber, Ar ions of the plasma bombard one or more exposed sputtering targets such that surface atoms of the sputtering targets are ejected and deposited as a thin film on the substrate. In method 100B, at operation 128A, a substrate such as a PVD chamber is subjected to a plurality of sputtering targets. Figure 1AA substrate, such as substrate 102, is subjected to operations including degassing and pre-cleaning in Ar gas plasma or in He / H plasma, and is moved between processing chambers by or via a central robotic substrate transfer chamber. In operation 128B, the substrate 102 is transferred from the central robotic substrate transfer chamber to a PVD chamber among a plurality of PVD chambers. Subsequently, in operation 130, a buffer layer 104 is deposited on the substrate 102 by sputtering in a target in the PVD chamber. A power of 1 kW to 100 kW is applied to one or more PVD chambers discussed herein to ionize a portion of the Ar and form the plasma used in operation 130. The ejected surface atoms of the target are deposited on the substrate 102 to form the buffer layer 104. During the formation of the buffer layer 104 in operation 130, a target including Co is used in conjunction with the Ar plasma in the PVD chamber. x Fe y B z , TaN and / or Ta one or more sputtering targets are used to form the buffer layer 104. In an embodiment where the buffer layer 104 is Ta or includes Ta, the buffer layer 104 is sputtered in a PVD chamber using a Ta target and Ar plasma. In an example where the buffer layer 104 is TaN or includes TaN, when nitrogen (N2) is present in the PVD chamber with the Ar plasma and the Ta sputtering target, operation 130 is performed to form the TaN buffer layer 104. In another example where the buffer layer 104 is TaN or includes TaN, operation 130 is performed in a PVD chamber using a TaN sputtering target and Ar plasma to form the TaN buffer layer 104. During the formation of the buffer layer 104 and subsequent layers, the one or more PVD chambers used are maintained at a vacuum pressure.
[0032] Subsequently, in operation 132, a seed layer 106 is deposited on the buffer layer 104 by sputtering a target in a PVD chamber. In an embodiment of operation 132, the seed layer 106 is formed in the same PVD chamber as the buffer layer 104 using a sputtering target different from the sputtering target used to form the buffer layer 104. In an example of method 100B according to an embodiment of the present disclosure, the following may be manufactured in operation 132 using Ar plasma and one or more sputtering targets in a PVD chamber: Figure 2A Seed layer 206 is shown. Figure 2AA coordinate system is shown with an x-axis 226 perpendicular to a y-axis 228. The thickness discussed herein may be measured in the direction of the y-axis 228. In one example of forming the seed layer 206, the sputtering target used to form the seed layer in the PVD chamber is comprised of Cr, or a combination of Cr and one or more of Ni, Fe, Ru, Ir, Co. In another example, the sputtering target(s) used in the PVD chamber at operation 132 is comprised of one or more alloy targets including NiCr, NiCrFe, RuCr, IrCr, CoCr, or a combination of the foregoing. The seed layer 206 may be formed to or less. In one example, the seed layer 206 includes NiCr and is manufactured in a PVD chamber using a NiCr target. In this example, Ar is introduced into the PVD chamber at a flow rate of about 5sccm to 60sccm, and in some examples, the flow rate of Ar entering the chamber is about 25sccm. In another example, Ar is introduced into the PVD chamber at a flow rate of 10sccm to 40sccm, and in some examples, at a flow rate of 25sccm. When Ar is present, a power of about 50W to 1200W is applied to the sputtering target at a negative voltage to form a plasma for sputtering. In some examples, the power applied to the sputtering target at a negative voltage is 100W to 800W, and in another example, the power applied to the sputtering target is about 500W.
[0033] At operation 134, a first pinned layer 108 is deposited on the seed layer 106 by sputtering a target in a PVD chamber. Taking the first pinned layer 108 as an example in a conventional MTJ stack 100A, the first pinned layer 108 may be formed at operation 134 by sputtering one or more targets in the presence of Ar plasma. In an example where the first pinned layer 108 is Co, a Co sputtering target is used in the presence of Ar plasma in the PVD chamber. In an example where the first pinned layer 108 includes one or more bilayers of cobalt and another element, operation 134 uses a Co sputtering target and another sputtering target composed of the other element of the bilayer. Depending on the implementation, the Co sputtering target and the sputtering target of the other metal may be sputtered in the same PVD chamber in the presence of Ar plasma, or each layer of the bilayer may be formed in a separate PVD chamber.
[0034] In an example according to an embodiment of the present disclosure, at operation 134, a PVD chamber is used to form Figure 2A and Figure 2C. In this example, in operation 134, the first pinned layer 208 is formed in the PVD chamber using xenon (Xe) or argon (Ar) introduced into the PVD chamber at a flow rate of about 2 sccm to 40 sccm while applying a power of 50 W to 10000 W to the target under a negative voltage to form a plasma. In another example, Xe or Ar is introduced into the PVD chamber at a flow rate of 5 sccm to 20 sccm, and in some examples, at a flow rate of 10 sccm. In another example, the power applied to the sputtering target is 100 W to 800 W, and in another example, the power applied to the sputtering target may be 400 W. Depending on the composition of the first pinned layer 208, in operation 134, Xe may be used in the sputtering operation in the PVD chamber because it is a heavier gas than Ar and therefore produces ions having a higher atomic weight than ions formed using Ar or other lighter gases and therefore bombards the target with more energy. In one example of the first pinned layer 208 of the present disclosure, Xe, Ar, or a mixture thereof is introduced into the PVD chamber at a flow rate of about 10 sccm, and a power of 400 W is applied to the target under a negative voltage to form Ar or Xe plasma. In one example, the first pinned layer 208 is made of cobalt (Co) by sputtering a Co target, and the first pinned layer 208 is formed from about to about In another example, Figure 2C As shown, the first pinned layer 208 is made of one or more bilayers of multiple materials. In multiple embodiments, the bilayer used to form the first pinned layer 208 includes a first interlayer of Co and a second interlayer of another element. The bilayer of the first pinned layer 208 can be formed in a PVD chamber including multiple targets, including a Co target and a target formed of the other element; or the bilayer of the first pinned layer 208 can be formed in separate PVD chambers, one of which contains a Co target and the other contains a target of the other element. In an example where multiple targets are set in a single PVD chamber, in the presence of Ar plasma and / or Xe plasma, the shield discussed herein can be used to selectively expose each of the Co target and the target of the other metal to form a bilayer of Co interlayer and form an interlayer of the other element to form the resulting bilayer. These depositions may be repeated at operation 134 for multiple iterations to form multiple bi-layers of the first pinned layer 208 .
[0035] At operation 136, a SAF coupling layer 110 is deposited on the first pinned layer 108 by sputtering a target of Ru, Cr, Rh, or Ir in a PVD chamber in the presence of Ar plasma. At operation 138, a second pinned layer 112 is deposited on the SAF coupling layer in a PVD chamber. In one example, the second pinned layer 112 is formed of Co in a PVD chamber using a Co target and Ar plasma. In another example, the second pinned layer 112 includes a bilayer and may or may not include a Co layer formed in contact with the bilayer. In this example, the second pinned layer 112 is formed in a PVD chamber using a Co sputtering target and a second metal sputtering target, and the shield is adjusted in at least one iteration to expose each of the Co and second metal sputtering targets, respectively, to form one or more bilayers of the second pinned layer 112. In other examples, each layer of the bilayer of the second pinned layer 112 can be formed in different PVD chambers, where one PVD chamber includes a Co sputtering target and the other PVD chamber includes a sputtering target of the second metal.
[0036] In one example of the second pinning layer 212 according to an embodiment of the present disclosure, Figure 2A and 2DAs shown, in operation 138, a second pinned layer 212 is deposited in a PVD chamber using Ar gas and / or Xe gas depending on the material to be sputtered. In an embodiment of the present disclosure, in operation 138, a Co sputtering target is used to manufacture the second pinned layer 212 as a single layer of cobalt (Co) in the presence of Ar plasma or a plasma consisting of Ar and Xe. In another example, the second pinned layer 212 includes a bilayer including a first interlayer formed of Co and a second interlayer of one or more second metals such as Pt, Ir, Ni, or Pd. In another example that may be combined with other examples herein, the second pinned layer 212 may be formed by a single Co layer alone, or by a combination of a single Co layer and at least one bilayer, wherein the bilayer has a first interlayer of Co and a second interlayer of at least one different metal. In this example, the second pinned layer 112 is formed in a PVD chamber using a Co sputtering target and a second metal sputtering target, and the shield is adjusted in at least one iteration to expose each of the Co and second metal sputtering targets, respectively, to form one or more double layers of the second pinned layer 112. When a metal such as Pt is used to form the second pinned layer 212, xenon can be used to deposit the second pinned layer 212 because Xe is a heavier gas than Ar and can therefore interact more effectively with heavier metals including Pt during the sputtering process in the PVD chamber. In an embodiment using Xe, the Xe gas is introduced into the PVD chamber at a flow rate of about 2 sccm to about 40 sccm, or 5 sccm to 20 sccm, and in some embodiments, the Xe gas is introduced into the PVD chamber at a flow rate of about 10 sccm. During the formation of the second pinned layer 212, a power of 50 W to about 1000 W is applied to the target at a negative voltage to form and maintain Ar and / or Xe plasma. In some examples, a power of 100 W to 600 W is applied to the target at a negative voltage to form and sustain the Ar and / or Xe plasma, and in some embodiments, a power of about 200 W is used.
[0037] At operation 140, the structural barrier layer 114 is formed in a PVD chamber that includes sputtering targets of Ta, Mo, and / or W, depending on the desired composition of the structural barrier layer 114. When two or more sputtering targets of Ta, Mo, and W are used, each target may be used in a separate PVD chamber, or the two or more sputtering targets may be sputtered sequentially or simultaneously in the PVD chamber using the shielding adjustments discussed above, depending on the desired composition of the structural barrier layer 114. A magnetic reference layer 116 is then deposited on the structural barrier layer 114 at operation 142, and the magnetic reference layer 116 may be formed in the PVD chamber in which other layers of the MTJ stack 100A may also be formed. If both the buffer layer 104 and the magnetic reference layer 116 are Cox Fe y B z If the composition of the buffer layer 104 is not sufficient, this may depend on, for example, the composition of other layers, such as the composition of the buffer layer 104. Co x Fe y B z The magnetic reference layer 116 is formed in a PVD chamber by using a sputtering target of an alloy, or by using independent sputtering targets of Co, Fe or B, or by a combination of an alloy sputtering target and a single element sputtering target, such as a CoFe target and a B target.
[0038] At operation 144, a tunnel barrier layer 118 is deposited on the magnetic reference layer 116. In one example of operation 144, the tunnel barrier layer 118 is formed in a PVD chamber using a metal oxide target such as MgO in the presence of Ar plasma. In an alternative embodiment, at operation 144, the tunnel barrier layer 118 is formed in a PVD chamber using a metal target such as Mg, Ti, Hf, Ta, or Al in the presence of Ar plasma and O2 to form a metal oxide of the tunnel barrier layer 118. At operation 146, a magnetic storage layer 120 is formed in a PVD chamber. The formation of the magnetic storage layer 120 can occur in a variety of ways depending on the desired composition. The magnetic storage layer 120 may include Co x Fe y B z In some examples that may be combined with other examples, the magnetic storage layer 120 may include one or more layers of Ta, Mo, W, or Hf. Accordingly, the deposition of the magnetic storage layer 120 in the PVD chamber may include: Ar plasma, Co x Fe y B z Alloy targets, independent targets of Co, Fe and B, or combinations of alloy targets and element targets such as CoFe targets and B targets. In the example where the magnetic storage layer 120 includes Ta, Mo, W or Hf, a sputtering target of Ta, Mo, W or Hf and a plasma formed of Ar are used in the chamber.
[0039] In one example, one or more targets (such as those discussed above for forming Co) may be exposed or protected by adjusting the shield in the presence of a plasma formed using Ar. x Fe y B z and Ta, Mo, W or Hf layers) in a single PVD chamber to form the magnetic storage layer 120. In another example, in the presence of Ar plasma, Co x Fe y B zAlloy target, Co of the magnetic storage layer 120 is sputtered in the PVD chamber x Fe y B z In another example, Co is formed in a PVD chamber using separate Co, Fe, and B targets in the presence of Ar plasma. x Fe y B z In another example, a Co layer is formed in a PVD chamber using an alloy target and an elemental target in the presence of Ar plasma. x Fe y B z Layers, the element targets are, for example, CoFe targets and B targets. Ta, Mo, W or Hf layers can be sputtered in a PVD chamber using Ta, Mo, W or Hf targets.
[0040] At operation 148, the capping layer 122 is deposited on the magnetic storage layer 120. In an embodiment, the first capping interlayer 122A of the capping layer 222 is formed in a PVD chamber, which may be different from the PVD chamber in which the non-oxide layer is formed because both Ar plasma and O2 are present in the PVD chamber during operation 148 when the oxide layer is formed. The first capping interlayer 122A is deposited in the PVD chamber by sputtering a Mg target and a plasma formed by ionized Ar, and O2 is also present in the PVD chamber. In another example, at operation 148, the first capping interlayer 122A is formed in the PVD chamber using a MgO sputtering target and Ar plasma. In an example where the first capping interlayer 122A is formed of the same material as the tunnel barrier layer 118 (e.g., Mg), the PVD chamber used for operation 144 may be the same PVD chamber as operation 148 used to form the first capping interlayer 122A. At operation 150, a second capping interlayer 122B is deposited on the first capping interlayer 122A. If O2 is used in operation 148, operation 150 may occur in a separate PVD chamber from the PVD chamber used to sputter the first capping interlayer 122A because O2 is not used in the PVD chamber to form the first capping interlayer 122A. The second capping interlayer 122B is formed in the PVD chamber using Ar plasma and one or more sputtering targets composed of Ru, Ir, and / or Ta. Depending on the composition of the second capping interlayer 122B, operation 150 may occur in the PVD chamber also used to form the SAF coupling layer 110, such as at operation 136.
[0041] Further, in the method 100B, at operation 152, a hard mask layer 124 is deposited over the second capping interlayer 122B in a PVD chamber. Depending on the type of hard mask layer 124 used in the MTJ stack 100A, operation 152 may or may not occur in the presence of O2. For example, if the hard mask layer 124 is a metal oxide hard mask, O2 may be used with one or more metal targets to form the metal oxide layer during operation 152, or a metal oxide sputtering target may be used to deposit the hard mask layer 124, in which case O2 is not used to form the hard mask layer 124 in operation 150. In some embodiments, when the hard mask layer 124 is amorphous carbon or spin-on carbon, operation 152 is performed in a CVD chamber.
[0042] Further, in the method 100B, the MTJ stack 100A (or hereinafter referred to as Figure 2A 100B) may undergo one or more processes collectively indicated by operation 154 in method 100B. These operations may include high temperature (approximately 400°C) operations. In one example, the process at operation 154 may include a pre-patterning annealing operation followed by an MTJ patterning operation. In alternative embodiments, the MTJ patterning at operation 154 may include multiple processes, such as patterning the hard mask layer 124, and may further include the following operations: after the hard mask layer 124 is patterned to form a plurality of independent pillars from the MTJ stack 100A, etching the MTJ stack 100A.
[0043] In an alternative embodiment, at operation 154, a thermal annealing operation is performed to repair, crystallize, and enhance the lattice structure of the film stack, which includes the (several) magnetic storage layer and (several) magnetic reference layer in the MTJ stack 100A. The thermal annealing performed at operation 154 may serve to further crystallize the material of at least the (several) magnetic reference layer 116 and the (several) magnetic storage layer 120. The crystallization of the magnetic reference layer and the magnetic storage layer as they are deposited establishes the perpendicular anisotropy of the MTJ stack 100A while maintaining the desired electrical and mechanical properties thereof. An embodiment of an MTJ stack fabricated following the operations of method 100B is shown and discussed below. The embodiment is configured to maintain the as-deposited fcc of the pinned layer after the thermal annealing operation performed at operation 154, and / or during additional or alternative back-end processing operations occurring at elevated temperatures on the order of 400°C. <111> Crystal structure.
[0044] Unlike conventional use of Ru or Ir for seed layers, the MTJ stack manufactured according to the embodiments of the present disclosure uses a chromium-based seed layer, including a seed layer of at least one of chromium (Cr), nickel chromium (NiCr), nickel iron chromium (NiFeCr), ruthenium chromium (RuCr), iridium chromium (IrCr) or cobalt chromium (CoCr). The seed layer is disposed in a film stack, and the memory cell is defined between the first pinned layer and the buffer layer from the film stack. Since the lattice matching between the seed layer and the first pinned layer helps to stabilize the first pinned layer and the layer formed on top of the first pinned layer, the Cr-containing seed layer discussed herein can improve the adhesion of the first pinned layer. In one example, the seed layer is in direct contact with both the first pinned layer and the buffer layer, and in other examples, there is a transition layer between the seed layer and the first pinned layer or between the seed layer and the buffer layer. The pinned layer discussed herein can be made of cobalt, or as one or more bilayers, each of which includes cobalt, or a combination of a bilayer structure and a capping layer of cobalt, as discussed below. As discussed herein, an "overlayer" is a layer formed over another structure, and the overlayer may include one or more layers, including a bilayer. Depending on the implementation, the first and second pinned layers may include the same layer structure, material, and / or thickness, or may vary in layer structure, material, and / or thickness. Using the MTJ stack discussed herein, the crystal structure of the SAF layer and the magnetic coupling of the magnetic reference layer to the magnetic storage layer remain substantially in a crystalline state, even after annealing the layers at about 400°C for a duration of 0.5 hours to at least 3 hours. Further, in an embodiment herein, the MTJ stack continues to exhibit a tunnel magnetoresistance (TMR) of 100% to 175%, even after annealing the MTJ stack at about 400°C for a duration of 0.5 hours to at least 3 hours.
[0045] Figure 2A 2 is a schematic diagram of an MTJ stack 200 according to an embodiment of the present disclosure. In the illustrated embodiment, a buffer layer 204 is formed on a conductive portion of a substrate 202, or on a conductive film layer on the substrate 202, via sputtering in a PVD chamber. The substrate 202 may include tungsten (W), tantalum nitride (TaN), titanium nitride (TiN), or other metal layers. The buffer layer 204 improves the adhesion of the seed layer 206 to the substrate 202, which helps the formation and performance of subsequently deposited layers of the MTJ stack 200. The buffer layer 204 includes Co x Fe y B z , Ta and / or TaN, and are formed in one or more PVD deposition operations in a PVD chamber in the presence of Ar plasma. In various examples, Ar plasma and Co x Fe y B zThe buffer layer 104 is formed in a PVD chamber by using a sputtering target of an alloy, or by using a separate sputtering target of Co, Fe or B, or by a combination of an alloy sputtering target and a single element sputtering target, such as a combination of a CoFe target and a B target. In an example where the buffer layer 204 includes a Ta layer, the Ta layer can be formed in a PVD chamber using a Ta target and Ar plasma.
[0046] In one example, the buffer layer 204 includes TaN and is sputtered onto the substrate 202 in a PVD chamber using a Ta target in the presence of both Ar plasma and N2, and the N2 reacts with the Ta material sputtered from the Ta target to form a TaN layer. In another example, the buffer layer 204 is formed in a PVD chamber using a TaN sputtering target and Ar plasma. In one example, the buffer layer 204 is sputtered directly onto and in contact with the conductive layer on the substrate 202. In other examples, there is a conductive transition layer between the conductive layer on the substrate 202 and the buffer layer 204 that does not affect the performance of the MTJ stack. The buffer layer 204 is used in the illustrated embodiment as appropriate and may not be used in some embodiments discussed herein. The total thickness of the buffer layer 204 is (without buffer layer) to about In one example, the buffer layer 204 is Ta, TaN or Co x Fe y B z The monolayer is sputtered directly on and in contact with the conductive layer on the substrate 202, and up to In another example, the buffer layer 204 is a combination of layers, and each layer of the buffer layer 204 is Ta, TaN or Co. x Fe y B z , and the thickness is from to In the case of using TaN instead of Ta or Co in the buffer layer 204 x Fe y B z In the example of When used alone, Co x Fe y B z In the example of forming the buffer layer 204, the layer thickness is In another example of the buffer layer 204, Ta or TaN and Co x Fe y B z Combined, and the thickness of the buffer layer 204 is
[0047] The seed layer 206 is deposited in a PVD chamber via sputtering of one or more targets. The seed layer 206 is deposited on the buffer layer 204. The seed layer 206 comprises Cr, NiCr, NiCrFe, RuCr, IrCr, CoCr, or a combination of the foregoing. The seed layer 206 may be formed as one or more layers of Cr, NiCr, NiCrFe, RuCr, IrCr, or CoCr, which may include a combination of those elements or a combination of alloys in a single layer. The formation of the seed layer 206 in the PVD chamber is discussed in detail above in operation 132. In an embodiment, the seed layer 206 has a thickness of 1000 Å. or less, and in one example, the thickness of the seed layer 206 is from about to about In one example, the seed layer 206 is directly formed on and contacts the buffer layer 204. In other examples, there is a transition layer between the seed layer 206 and the buffer layer 204 that does not affect the performance of the MTJ stack.
[0048] Further, in the MTJ stack 200, the first pinned layer 208 is formed on the seed layer 206 in a PVD chamber. Figure 1B The formation of the first pinned layer 208 is shown in detail at operation 134 of method 100B in FIG. 1 and occurs in a PVD chamber using a plasma of ionized Ar and one or more sputtering targets. In one example, the first pinned layer 208 is fabricated to have a thickness of approximately to about In another example, the first pinning layer 208 is made of one or more bilayers of various materials, each of which may include two interlayers. The first pinning layer 208 may include only one or more bilayers, or a combination of one or more bilayers and a Co layer, such as Figure 2CAs shown. In the example where one or more bilayers are included in the first pinned layer 208, each bilayer contains a first interlayer of Co and a second interlayer of another element. The first pinned layer 208 is formed by sputtering a Co target using Ar plasma, and then sputtering a second target of Pt, Ir, Ni, or Pd using Ar plasma. In the example where Pt is used together with Co to form a bilayer, Xe plasma can be used instead of Ar plasma or Xe plasma and Ar plasma can be used. In the embodiment where one or more bilayers are used to form the first pinned layer, repeated deposition cycles can be performed in a PVD chamber in the following manner: a first interlayer of a bilayer is formed by shielding a target that does not include Co, wherein the first interlayer contains Co. Subsequently, the Co target and other targets are shielded to expose a second target, which contains a second element for the second interlayer of the bilayer. This can be repeated in an iterative manner to form one or more bilayers of the first pinned layer 208. In one example, the first pinned layer 208 is formed directly on the seed layer 206 in the PVD chamber and in contact with the seed layer 206. In other examples, there is a transition layer between the seed layer 206 and the first pinned layer 208 that does not affect the performance of the MTJ stack.
[0049] A synthetic antiferromagnet (SAF) coupling layer 210 is sputter-deposited on the first pinned layer 208 in a PVD chamber, and a second pinned layer 212 is sputter-deposited on the SAF coupling layer 210. The SAF coupling layer 210 is formed in a PVD chamber using a Ru sputtering target, a Rh sputtering target, a Cr sputtering target, or an Ir sputtering target in the presence of a plasma formed from ionized Ar gas. The SAF coupling layer 210 includes about to about In an embodiment, the second pinned layer 212 is fabricated by sputtering a cobalt (Co) target in a PVD chamber in the presence of a plasma formed by ionized Ar gas. In one example, the SAF coupling layer 210 is formed directly on and contacts the first pinned layer 208 and the second pinned layer 212. In other examples, there is a transition layer between the SAF coupling layer 210 and either or both of the first pinned layer 208 or the second pinned layer 212, and the transition layer does not affect the performance of the MTJ stack.
[0050] In the above Figure 1BThe formation of the second pinned layer 212 is shown in operation 138 of method 100B in FIG. In an example where the second pinned layer 212 includes one or more bilayers of Co and another element, the second pinned layer 212 is formed by sputtering a Co target using an Ar plasma and then sputtering a second target of Pt, Ir, Ni, or Pd using an Ar plasma. Repeated deposition cycles using the Co target and the second target in a PVD chamber in the presence of an Ar plasma may be used to form the one or more bilayers of the second pinned layer 212. In one example, the second pinned layer 212 is sputtered to a thickness ranging from about 100 Å to about 100 Å. to about In another example, the thickness of the second pinned layer 212 is about Figure 2D An alternative configuration of the second pinned layer 212 is shown.
[0051] In addition, in the MTJ stack 200, a structural barrier layer 214 is formed on the second pinned layer 212, optionally by sputtering in a PVD chamber. The structural barrier layer 214 prevents a short circuit from forming between the MTJ stack 200 and a metal contact that can be coupled to the MTJ stack 200 to form an MRAM memory cell. During sputtering of the structural barrier layer 214, one or more independent Ta, Mo, or W sputtering targets can be used in the PVD chamber together with Ar plasma, depending on the desired composition of the layer. In other examples, during the formation of the structural barrier layer 214, one or more alloy targets including an alloy of Ta, Mo, and / or W can be used in the PVD chamber with Ar plasma. Each of the targets can be oriented in a face-centered cubic manner. <111> Compared with the seed layer 206, the first pinned layer 208 and the second pinned layer 212, the structural barrier layer 214 is oriented in the <100> The structure barrier layer 214 is from 0 (no layer) to to about thick, and in one example, sputter deposited In one example, the structural barrier layer 214 is formed directly on and contacts the second pinned layer 212 during sputtering deposition. In other examples, there is a transition layer between the structural barrier layer 214 and the second pinned layer 212 that does not affect the performance of the MTJ stack.
[0052] The magnetic reference layer 216 is formed on the structural barrier layer 214 by sputtering deposition in a PVD chamber in the presence of Ar plasma. The magnetic reference layer 216 can be formed in the PVD chamber using a single Co-Fe-B alloy sputtering target, or by using two or more of a Co sputtering target, a Fe sputtering target, or a B sputtering target. In another example, the magnetic reference layer 216 can be formed in the PVD chamber using an alloy target and an element target, such as a CoFe target and a B target, in the presence of Ar plasma. The magnetic reference layer 216 can be sputtered to a thickness of 1000 Å. to In one example, the magnetic reference layer 216 may be formed as The magnetic reference layer 216 includes Co x Fe y B z , in one example, z is from about 10 wt % to about 40 wt %, y is from about 20 wt % to about 60 wt %, and x is equal to or less than 70 wt %. In another example that may be combined with other examples herein, z is at least 20 wt %. In one example, the magnetic reference layer 216 is formed directly on and in contact with the structural barrier layer 214. In other examples, there is a transition layer between the magnetic reference layer 216 and the structural barrier layer 214 that does not affect the performance of the MTJ stack.
[0053] The tunnel barrier layer 218 is formed on the magnetic reference layer 216 using sputtering of a target in Ar plasma in a PVD chamber. The tunnel barrier layer 218 includes metal oxides such as magnesium oxide (MgO), hafnium oxide (HfO2), titanium oxide (TiO2), tantalum oxide (TaO x ), aluminum oxide (Al2O3), or other materials suitable for various applications. Thus, the tunnel barrier layer 218 can be formed in a PVD chamber using a sputtering target of a metal oxide in the presence of Ar. Alternatively, the tunnel barrier layer 218 can be formed in a PVD chamber using a sputtering target of a metal of a desired metal oxide in the presence of Ar and O2, wherein the metal oxide layer is exposed to O2 when the metal layer is sputtered from the metal sputtering target. The tunnel barrier layer 218 has to In some embodiments, the thickness is In one example, the tunnel barrier layer 218 is formed directly on and contacts the magnetic reference layer 216. In other examples, there is a transition layer between the tunnel barrier layer 218 and the magnetic reference layer 216 that does not affect the performance of the MTJ stack.
[0054] In an embodiment, the MTJ stack 200 further includes a magnetic storage layer 220 formed on the tunnel barrier layer 218 using a sputtering operation in a PVD chamber as discussed herein. The magnetic storage layer 220 may include Co x Fe y B z In some examples, the magnetic storage layer 220 may alternatively or additionally include one or more layers of Ta, Mo, W, or Hf. Accordingly, the deposition of the magnetic storage layer 220 in the PVD chamber may include: using Ar plasma, Co x Fe y B z Alloy target, or independent targets of Co, Fe and B, or a combination of alloy target and element target, such as CoFe target and B target. In the example where the magnetic storage layer 220 includes Ta, Mo, W or Hf, a sputtering target of Ta, Mo, W or Hf is used in the chamber together with a plasma formed of Ar.
[0055] The thickness of the magnetic storage layer 220 may range from about 1000 Å to about 5000 Å, depending on factors including the material or materials used to form the magnetic storage layer 220. to about In one embodiment, the magnetic storage layer is composed of Co x Fe y B z %, wherein z is from about 10 wt % to about 40 wt %, y is from about 20 wt % to about 60 wt %, and x is equal to or less than 70 wt %. In this example, the thickness of the magnetic storage layer 220 is from to And in some embodiments, the target layer thickness is In one example, the magnetic storage layer 220 is formed directly on and in contact with the tunnel barrier layer 218. In other examples, there is a transition layer between the magnetic storage layer 220 and the tunnel barrier layer 218 that does not affect the performance of the MTJ stack. The magnetic storage layer is shown below Figure 2E middle.
[0056] Further, in an embodiment of the MTJ stack 200, the capping layer 222 is formed on the magnetic storage layer 220 and includes a plurality of interlayers forming the capping layer 222, the interlayers including an oxide containing iron (Fe). In addition, in some embodiments, the hard mask layer 224 is directly formed on the capping layer 222 and in contact with the capping layer 222. In another example, the hard mask layer 224 is formed on the capping layer 222, and there is a transition layer between the capping layer 222 and the hard mask layer 224; such a transition layer does not affect the performance of the MTJ stack 200. The hard mask layer 224 can be formed of a metal oxide, amorphous carbon, ceramic, a metal material, or a combination of the foregoing. In one example, the magnetic storage layer 220 is directly formed on the capping layer 222 and in contact with the capping layer 222. In other examples, there is a transition layer between the magnetic storage layer 220 and the capping layer 222 that does not affect the performance of the MTJ stack 200. The capping layer 222 is shown below Figure 2F middle.
[0057] In one example of the MTJ stack 200, the MTJ stack 200 includes a seed layer 206 made of NiCr, a first pinned layer 208 made of a Co / Pt bilayer stack, and a second pinned layer 212 made of Co. In this example, for example, the MTJ stack 200 exhibits a TMR of 175% when annealed at about 400° C. for 0.5 hours. When the MTJ stack 200 is annealed at about 400° C. for 3 hours, the TMR of the MTJ stack 200 is maintained at about 175%.
[0058] Figure 2B FIG. 2 is an enlarged view of a buffer layer 204 according to an embodiment of the present disclosure. The buffer layer 204 includes tantalum (Ta) or TaN, or a layered stack of Ta and TaN, and in some examples, includes a single Co x Fe y B z , or Co x Fe y B z Combination of layered stacking with Ta, TaN or Ta / TaN. In one example of the buffer layer 204, the buffer layer 204 includes at least one bilayer 204D. The at least one bilayer 204D includes a first buffer interlayer 204A and a second buffer interlayer 204B formed on the substrate 202 in an alternating manner for at least one iteration of the at least one bilayer 204D. In this example, the first buffer interlayer 204A includes Ta, and the second buffer interlayer 204B includes TaN, and the first buffer interlayer 204A contacts the substrate 202. In another example, the first buffer interlayer 204A includes TaN, and the second buffer interlayer 204B includes Ta, and thus the TaN is in direct contact with the substrate 202.
[0059] In other examples of the buffer layer 204, Figure 2A As shown, the buffer layer 204 is made of Co alone. x Fe y B z , and therefore Co x Fe y B z may directly contact the substrate 202. In another example, Figure 2B As shown, the third buffer layer 204C is formed on at least one bilayer 204D. In this example, the third buffer layer 204C is composed of Co x Fe y B z Manufacturing and forming up to Therefore, depending on the configuration of the buffer layer 204, the thickness of the buffer layer 204 ranges from Thick to Thick. x Fe y B z In an example of the third buffer layer 204C, z is from about 10 wt % to about 40 wt %, y is from about 20 wt % to about 60 wt %, and x is equal to or less than 70 wt %.
[0060] Figure 2C 2 is an enlarged view of the first pinned layer 208 according to an embodiment of the present disclosure. In an embodiment, the first pinned layer 208 is made of at least one bilayer 230, and when two or more bilayers are used, the bilayers can be said to form a bilayer stack 234. Each bilayer 230 is made of a first interlayer 208A and a second interlayer 208B. The bilayers of the first pinned layer 208 are represented as (X / Y) n , (208A / 208B) n , where each bilayer is a combination of X and Y materials, and n is the number of bilayers in the first pinned layer 208. In an embodiment, X is Co and Y is one of Pt, Ir, Ni, or Pd. Figure 2C In the example of FIG. 1 , n=4, but in alternative embodiments, n is from 1 to 10. In an embodiment, at least one bilayer 230 includes from about to about In one example, the first interlayer 208A includes Co and has a thickness of about to about The second interlayer 208B is about to about In another example, the desired layer thickness of each layer in the first interlayer 208A and the second interlayer 208B is from about to about Furthermore, in another embodiment, the first pinned layer 208 includes at least one bilayer 230 formed directly on and in contact with the seed layer 206, and a capping layer 208C of Co formed on top of the at least one bilayer 230. In this example, the capping layer 208C is formed from about to about Depending on the implementation, the total thickness of the first pinned layer 208, which may include one or more layers including at least one bilayer 230 as discussed herein, is from 0.3 nm to about 18 nm. In other examples, one or more transition layers that do not negatively affect the characteristics of the MTJ stack may be formed between the first pinned layer 208 and the seed layer 206.
[0061] Figure 2D FIG. 2 is an enlarged view of the second pinned layer 212 according to an embodiment of the present disclosure. Figure 2A As shown, the second pinned layer 212 is made of a single cobalt layer. In an alternative embodiment, the second pinned layer 212 is made of at least one bilayer 232. The at least one bilayer 232 includes a first interlayer 212A formed of Co and a second interlayer 212B formed of one or more of Pt, Ir, Ni, or Pd. When two or more bilayers such as the bilayer 232 are used in the second pinned layer 212, the plurality of bilayers may be referred to as a bilayer stack 236. Figure 1B When one or more bilayers are used, separate sputtering targets may be used to form each layer of the bilayer. The at least one bilayer 232 of the second pinned layer 212 may be represented as (X / Y) n , (212A / 212B) n , where n is the number of double layers. Figure 2D In the example of n=4, but in alternative embodiments, n is from 1 to 10. In an embodiment, at least one bilayer 232 includes from about to about In one example, the first interlayer 212A is about to about The second interlayer 212B is about to about In another example, the thickness of each layer in the first interlayer 212A and the second interlayer 212B is from about to about
[0062] Furthermore, in another embodiment, the second pinned layer 212 includes at least one bilayer 232 formed directly on and in contact with the SAF coupling layer 210, and a capping layer 212C of Co located on top of the at least one bilayer 232. In some examples, a transition layer may be used between the at least one bilayer 232 and the second pinned layer 212, or between the at least one bilayer 232 and the SAF coupling layer 210, or between both, wherein such (several) transition layers do not affect the performance of the MTJ stack. In this example, the capping layer 212C is about to about Depending on the implementation, the second pinned layer 212, which may include one or more layers including at least one bilayer 232 discussed herein, has a total thickness of 0.3 nm to about 18 nm.
[0063] In an embodiment, the first pinned layer 208 and the second pinned layer 212 each include at least one of the same composition or the same thickness. In an alternative embodiment, the first pinned layer 208 and the second pinned layer 212 each include different compositions and / or thicknesses. In an embodiment, the first pinned layer 208 is Co, and the second pinned layer 212 is formed by one or more bilayers. Each bilayer of the second pinned layer 212 includes a first interlayer of Co and a second interlayer of Pt. In another embodiment, the first pinned layer 208 is Co, and the second pinned layer 212 is formed by one or more Co / Ni bilayers. In yet another embodiment, the first pinned layer 208 includes one or more bilayers, each bilayer containing a first interlayer of Co and a second interlayer of Ni, and the second pinned layer 212 includes one or more bilayers, each bilayer containing a first interlayer of Co and a second interlayer of Pt. In yet another embodiment, the first pinned layer 208 includes one or more bilayers, each bilayer containing a first interlayer of Co and a second interlayer of Pt, and the second pinned layer 212 includes one or more bilayers, each bilayer containing a first interlayer of Co and a second interlayer of Ni.
[0064] Figure 2E FIG. 2 is an enlarged view of an exemplary magnetic storage layer 220 according to an embodiment of the present disclosure. Figure 2E As shown, the first magnetic layer 220A of the magnetic storage layer 220 and the second magnetic layer 220B of the magnetic storage layer 220 are each made of Co x Fe y B zA third layer 220C made of Ta, Mo, W, Hf or a combination thereof is disposed therebetween, and the third layer 220C contains a dopant, such as boron, oxygen or other dopants. Therefore, the magnetic storage layer 220 is made of the following three layers: the first magnetic layer 220A and the second magnetic layer 220B, and the third layer 220C disposed between the first magnetic layer 220A and the second magnetic layer 220B. The third layer 220C strengthens the pinning moment perpendicular to the substrate plane (e.g., a plane perpendicular to the substrate 202), which enhances the magnetic anisotropy, the directional dependence of the magnetism of the structure.
[0065] Figure 2F FIG. 2 is an enlarged view of an exemplary cover layer 222 according to an embodiment of the present disclosure. The total thickness of the cover layer 222 is from to And in some embodiments, the total desired thickness of the covering layer (e.g., including Figure 2F All interlayers shown) are approximately In an embodiment, the capping layer 222 includes a plurality of interlayers. The first capping interlayer 222A is formed directly on the magnetic storage layer 220 to a thickness of about 100 mm. to about On top of the first cover interlayer 222A, a second cover interlayer 222B of Ru, Ir or a combination thereof is formed. to about In an embodiment, a third cover interlayer 222C to 222B is formed of Ta as appropriate. to about Therefore, some embodiments of the cover layer 222 do not include the third cover interlayer 222C. In an embodiment, a fourth cover interlayer 222D is formed on the third cover interlayer 222C as appropriate. The fourth cap interlayer 222D is formed of Ru, Ir, or a combination thereof. In various embodiments, the cap layer 222 includes only the first cap interlayer 222A, or the first cap interlayer 222A and the second cap interlayer 222B, or the first cap interlayer 222A, the second cap interlayer 222B, and the third cap interlayer 222C, or the first, second, and third cap layers 222A to 222C. In some embodiments, a transition layer may be used between some or all of the first cap interlayer 222A, the second cap interlayer 222B, and the third cap interlayer 222C, or a transition layer may be used between the cap layer 222 and the magnetic storage layer 220, so that the (several) transition layers do not negatively affect the performance of the MTJ stack.
[0066] The MTJ stack discussed herein has improved performance after being subjected to treatment at a temperature of 400°C or above. The use of a Cr seed layer or a Cr-containing seed layer in combination with a first pinning layer also comprising Cr results in a lattice match between the structure of the seed layer and the first pinning layer, thereby promoting the improved performance. The lattice match between the structure of the seed layer and the first pinning layer suppresses the formation of roughness at the interface of the seed layer and the first pinning layer, which may result in a lack of flatness in one or more layers. Therefore, the MTJ stack can maintain structural integrity and also maintain desired magnetic and electrical properties. Furthermore, in the embodiments discussed herein, the combination of the first and second pinning layer components and the seed layer can still enhance the desired magnetic properties even after high temperature treatment.
[0067] While the foregoing is directed to embodiments of the present disclosure, other and further embodiments of the disclosure may be devised without departing from the basic scope thereof and the scope of the disclosure is determined by the claims that follow.
Claims
1. A device comprising: A magnetic tunnel junction stack comprising: a seed layer contacting the buffer layer, wherein the seed layer comprises chromium (Cr); a first pinning layer contacting the seed layer, wherein the first pinning layer comprises a plurality of bilayers, wherein each bilayer of the plurality of bilayers comprises: a first interlayer and a second interlayer, the first interlayer being formed of cobalt (Co), and the second interlayer being formed of platinum (Pt), iridium (Ir), nickel (Ni), or palladium (Pd); a coupling layer, contacting the first pinning layer; a second pinning layer, contacting the coupling layer, a structural barrier layer contacting the second pinning layer; a magnetic reference layer contacting the structural barrier layer; a tunnel barrier layer contacting the magnetic reference layer; and The magnetic storage layer contacts the tunnel barrier layer.
2. The device of claim 1, wherein the first pinned layer and the second pinned layer each comprise 1 to 10 bilayers, and wherein the total thickness of the first pinned layer is 0.3 nm to 18 nm, and the total thickness of the second pinned layer is 0.3 nm to 18 nm.
3. The device of claim 1, wherein the first pinned layer further comprises a cobalt layer, the cobalt layer contacting the plurality of bilayers and contacting the coupling layer, wherein the cobalt layer is formed from to thick.
4. The device of claim 1, wherein the first pinned layer is to thick, and wherein the seed layer is from to thick.
5. The device of claim 1, wherein the second pinned layer is formed as a single Co layer and is to thick.
6. A device as described in claim 1, wherein the second pinned layer is formed of a plurality of bilayers, wherein each bilayer of the plurality of bilayers includes a first interlayer and a second interlayer, the first interlayer is formed of cobalt (Co), and the second interlayer is formed of at least one of platinum (Pt), iridium (Ir), nickel (Ni) or palladium (Pd), wherein the second pinned layer is from 0.3 nm to 18 nm thick.
7. The device of claim 1, wherein the seed layer is formed from to thick.
8. The device according to claim 7, further comprising a capping layer formed on the magnetic storage layer, wherein the capping layer is formed of an oxide containing iron (Fe) and is formed from to thick.
9. A magnetic tunnel junction stack, comprising: a seed layer formed on the buffer layer, wherein the seed layer comprises chromium (Cr); a first pinned layer formed on the seed layer, wherein the first pinned layer comprises a first plurality of bilayers, wherein each bilayer of the first plurality of bilayers comprises: a first interlayer and a second interlayer, the first interlayer being formed of cobalt (Co), and the second interlayer being formed of at least one of platinum (Pt), iridium (Ir), nickel (Ni), or palladium (Pd); a coupling layer formed on the first pinning layer; a second pinned layer formed on the coupling layer, wherein the second pinned layer comprises cobalt (Co); a structural barrier layer contacting the second pinning layer; a magnetic reference layer contacting the structural barrier layer; a tunnel barrier layer contacting the magnetic reference layer; and The magnetic storage layer contacts the tunnel barrier layer.
10. The stack of claim 9, wherein the first pinning layer further comprises to A cobalt layer having a thickness of from 0.04 to 0.13 W is formed on the first plurality of bilayers and in contact with the coupling layer, and wherein the first pinned layer has a thickness of from to thick.
11. The stack of claim 9, wherein the structural barrier layer comprises at least one of tantalum (Ta), molybdenum (Mo), or tungsten (W), and has a thickness ranging from to 12. The stack of claim 9, wherein the seed layer is formed from to Thick and composed of two or more layers.
13. The stack of claim 11, wherein the second pinning layer is formed from to thick.
14. A magnetic tunnel junction stack, comprising: A seed layer is formed on the buffer layer, wherein the seed layer comprises chromium (Cr) and is formed from to thick; A first pinning layer is formed on the seed crystal layer, wherein the first pinning layer has to a thickness of , wherein the first pinned layer comprises a first plurality of bilayers and a first cobalt capping layer, the first cobalt capping layer being formed over the first plurality of bilayers; a coupling layer formed between the first pinning layer and the first cobalt coating layer; A second pinning layer is formed on the coupling layer, wherein the second pinning layer has to a thickness of , wherein the second pinned layer comprises a second plurality of bilayers; A structural barrier layer is formed on the second pinned layer, wherein the structural barrier layer comprises at least one of tantalum (Ta), molybdenum (Mo) or tungsten (W), and is formed from to thick; a magnetic reference layer formed on the structural barrier layer; a tunnel barrier layer formed on the magnetic reference layer; and The magnetic storage layer is formed on the tunnel barrier layer.
15. A stack as described in claim 14, wherein each bilayer in the first plurality of bilayers comprises: a first interlayer and a second interlayer, the first interlayer is formed of cobalt (Co), and the second interlayer is formed of at least one of platinum (Pt), iridium (Ir), nickel (Ni) or palladium (Pd), and wherein the number of bilayers in the first plurality of bilayers is from 2 to 10.
16. The stack of claim 15, wherein the first cobalt coating is formed from to thick.
17. The stack of claim 14, wherein the seed layer is formed from to thick.
18. The stack of claim 14, wherein the second pinned layer comprises a second cobalt capping layer formed on top of the second plurality of bilayers, the second cobalt capping layer contacting the structural barrier layer.
Citation Information
Patent Citations
Magnetic memory devices having perpendicular magnetic tunnel structures therein
CN104658593A
Spin transfer torque tunneling magneto-resistive device having a laminated free layer with perpendicular magnetic anisotropy
CN105679930A
