Magnetoresistive stack and method of fabricating same

CROSS REFERENCE TO RELATED APPLICATIONS

0001This application is a reissue application of U.S. Pat. No. 10,622,554, which issued on Apr. 14, 2020, from U.S. patent application Ser. No. 16/419,165, filed May 22, 2019, the disclosure of which is expressly incorporated herein by reference. U.S. patent application Ser. No. 16/419,165, filed May 22, 2019, is a continuation of U.S. patent application Ser. No. 16/230,031, filed Dec. 21, 2018 (now U.S. Pat. No. 10,347,828), which is a continuation of U.S. patent application Ser. No. 15/941,153, filed Mar. 30, 2018 (now U.S. Pat. No. 10,199,574), which is a continuation of U.S. patent application Ser. No. 15/400,889, filed Jan. 6, 2017 (now U.S. Pat. No. 9,947,865), which is a divisional of U.S. patent application Ser. No. 14/860,657, filed Sep. 21, 2015 (now U.S. Pat. No. 9,553,258), which is a continuation of U.S. patent application Ser. No. 14/219,532, filed Mar. 19, 2014 (now U.S. Pat. No. 9,159,906), which is a divisional of U.S. patent application Ser. No. 13/158,171, filed Jun. 10, 2011 (now U.S. Pat. No. 8,686,484).

TECHNICAL FIELD

0002The exemplary embodiments described herein generally relate to magnetoresistive random access memory (MRAM) and more particularly to spin-torque MRAM elements.

BACKGROUND

0003Magnetoelectronic devices, spin electronic devices, and spintronic devices are synonymous terms for devices that make use of effects predominantly caused by electron spin. Magnetoelectronics are used in numerous information devices to provide non-volatile, reliable, radiation resistant, and high-density data storage and retrieval. The numerous magnetoelectronics information devices include, but are not limited to, Magnetoresistive Random Access Memory (MRAM), magnetic sensors, and read/write heads for disk drives.

0004Typically an MRAM includes an array of magnetoresistive memory elements. Each magnetoresistive memory element typically has a structure that includes multiple magnetic layers separated by various non-magnetic layers, such as a magnetic tunnel junction (MTJ), and exhibits an electrical resistance that depends on the magnetic state of the device. Information is stored as directions of magnetization vectors in the magnetic layers. Magnetization vectors in one magnetic layer are magnetically fixed or pinned, while the magnetization direction of another magnetic layer may be free to switch between the same and opposite directions that are called “parallel” and “antiparallel” states, respectively. Corresponding to the parallel and antiparallel magnetic states, the magnetic memory element has low and high electrical resistance states, respectively. Accordingly, a detection of the resistance allows a magnetoresistive memory element, such as an MTJ device, to provide information stored in the magnetic memory element. There are two completely different methods used to program the free layer: field switching and spin-torque switching. In field-switched MRAM, current carrying lines adjacent to the MTJ bit are used to generate magnetic fields that act on the free layer. In spin-torque MRAM, switching is accomplished with a current pulse through the MTJ itself. The spin angular momentum carried by the spin-polarized tunneling current causes reversal of the free layer, with the final state (parallel or antiparallel) determined by the polarity of the current pulse. Spin-torque transfer is known to occur in MTJ devices and giant magnetoresistance devices that are patterned or otherwise arranged so that the current flows substantially perpendicular to the interfaces, and in simple wire-like structures when the current flows substantially perpendicular to a domain wall. Any such structure that exhibits magnetoresistance has the potential to be a spin-torque magnetoresistive memory element. The mean current required to switch the magnetic state of the free layer is called the critical current (Ic). The critical current density (Jc) is the average critical current per area of the bit (Jc=Ic/A), and the current supplied by the circuit to switch spin-torque MRAM elements in a memory array is the write current (Iw). Reducing the write current Iw is desirable so that a smaller access transistor can be used for each bit cell and a higher density, lower cost memory can be produced. Lowering Jc is desirable to prevent tunnel barrier damage during programming.

0005In order to reduce write current, some spin-torque MRAM elements incorporate a dual-spin-filter structure, in which the MTJ stack includes two different spin-polarizing layers, one on each side of the free layer, to lower Jc by improving spin-torque transfer efficiency through increased spin torque on the free layer, resulting in a lower write current. Some dual-spin-filter devices have two tunnel barriers for providing a lower Jc, and a more symmetrical write current in the current up/down direction, than single tunnel barrier devices.

0006Dual-spin-filter devices require that the spin-polarizing fixed layers on either side of the free layer have opposite magnetization directions, so that the spin-torque effect from each of the two fixed layers will act together to switch the free layer magnetization into the desired direction when a current flows either up or down through the device. One way to provide such opposed fixed layers is to use a pinned synthetic antiferromagnetic (SAF) fixed region on one side and a single pinned layer on the opposite side of the free layer. Another known dual-spin-filter device includes a three-layer SAF and a two-layer SAF on opposed sides of the free layer. However, a device having such opposed fixed layers has reduced magnetoresistance ratio (MR) compared to a single-tunnel-barrier device since one tunnel junction is in the parallel state when the other is in the antiparallel state.

0007The structure will have a different resistance depending on the stable magnetic states in which the free magnetic layer has been written. In order to achieve a magnetic element which includes a better read signal, or an improved MR, a larger difference between the individual resistances, and thus a larger MR, is desirable.

0008Accordingly, it is desirable to provide a spin-torque magnetoresistive memory element having a low critical current density and a high MR. Furthermore, other desirable features and characteristics of the exemplary embodiments will become apparent from the subsequent detailed description and the appended claims, taken in conjunction with the accompanying drawings and the foregoing technical field and background.

BRIEF SUMMARY

0009A spin-torque magnetoresistive memory element is formed on a substrate having a surface defining a plane. The spin-torque magnetoresistive memory element comprises a first electrode comprising a ferromagnetic material formed over the substrate; a second electrode; a free magnetic layer; a first tunnel barrier positioned between the free magnetic layer and the first electrode to form a first tunnel junction having a first magnetoresistance ratio and a first resistance-area product; and a second tunnel barrier positioned between the free magnetic layer and the second electrode to form a second tunnel junction having a second magnetoresistance ratio and a second resistance-area product, wherein the first magnetoresistance ratio and the first resistance-area product are one of less than half or more than double the second magnetoresistance ratio and the second resistance-area product, respectively.

0010A method for forming the spin-torque magnetoresistive memory element on a substrate having a surface defining a plane, comprises forming a first electrode comprising a ferromagnetic material over the substrate, forming a first tunnel barrier over the first electrode, forming a free magnetic layer over the first tunnel barrier, thereby forming a first tunnel junction having a first magnetoresistance ratio and a first resistance-area product, forming a second tunnel barrier over the free magnetic layer, and forming a second electrode over the second tunnel barrier, thereby forming a second tunnel junction having a second magnetoresistance ratio and a second resistance-area, wherein the first magnetoresistance ratio and the first resistance-area produce are one of less than half or more than double the second magnetoresistance ratio and the second resistance-area product, respectively.

0011Another method for forming a spin-torque MRAM element comprising forming a first tunnel barrier; forming a second tunnel barrier; forming a free layer between the first and second tunnel barriers; forming a first electrode on a side of the first tunnel barrier opposed to the free layer, thereby forming a first tunnel junction having a first magnetoresistance ratio and a first resistance-area product; and forming a second electrode on a side of the second tunnel barrier opposed to the free layer. thereby forming a second tunnel junction having a second magnetoresistance ratio and a second resistance-area product, wherein the first magnetoresistance ratio is more than double the second magnetoresistance ratio and the first resistance-area produce is more than double the second resistance-area product.

BRIEF DESCRIPTION OF THE DRAWINGS

0012The present invention will hereinafter be described in conjunction with the following drawing figures, wherein like numerals denote like elements, and

0013<figref idref="DRAWINGS">FIG. <b>1</b></figref> is a cross section of a spin-torque magnetoresistive memory element in accordance with an exemplary embodiment;

0014<figref idref="DRAWINGS">FIG. <b>2</b></figref> is a cross section of a spin-torque magnetoresistive memory element in accordance with another exemplary embodiment;

0015<figref idref="DRAWINGS">FIG. <b>3</b></figref> is a cross section of a spin-torque magnetoresistive memory element in accordance with yet another exemplary embodiment;

0016<figref idref="DRAWINGS">FIG. <b>4</b></figref> is a cross section of a spin-torque magnetoresistive memory element in accordance with still another exemplary embodiment

0017<figref idref="DRAWINGS">FIG. <b>5</b></figref> is a graph of magnetoresistance versus the resistance/area of a top and a bottom tunnel barrier with the free layer comprising Ta, and a top and a bottom tunnel barrier without the Ta insertion in the free layer;

0018<figref idref="DRAWINGS">FIG. <b>6</b></figref> is a graph of magnetoresistance versus the resistance/area of a dual tunnel barrier device with a Ta insertion within the free layer;

0019<figref idref="DRAWINGS">FIG. <b>7</b></figref> is a graph of magnetoresistance versus the resistance/area of dual tunnel barrier device with a Ru layer within the free layer;

0020<figref idref="DRAWINGS">FIG. <b>8</b></figref> is a cross section of a spin-torque magnetoresistive memory element in accordance with a further exemplary embodiment;

0021<figref idref="DRAWINGS">FIG. <b>9</b></figref> is a cross section of a spin-torque magnetoresistive memory element in accordance with yet a further exemplary embodiment;

0022<figref idref="DRAWINGS">FIG. <b>10</b></figref> is a flow chart in accordance with an exemplary embodiment of a process for fabricating a spin-torque magnetoresistive memory element;

0023<figref idref="DRAWINGS">FIG. <b>11</b></figref> is a flow chart in accordance with another exemplary embodiment of a process for fabricating a spin-torque magnetoresistive memory element;

0024<figref idref="DRAWINGS">FIG. <b>12</b></figref> is a flow chart in accordance with yet another exemplary embodiment of a process for fabricating a spin-torque magnetoresistive memory element; and

0025<figref idref="DRAWINGS">FIG. <b>13</b></figref> is a flow chart in accordance with still another exemplary embodiment of a process for fabricating a spin-torque magnetoresistive memory element.

DETAILED DESCRIPTION

0026The following detailed description is merely illustrative in nature and is not intended to limit the embodiments of the subject matter or the application and uses of such embodiments. Any implementation described herein as exemplary is not necessarily to be construed as preferred or advantageous over other implementations. Furthermore, there is no intention to be bound by any expressed or implied theory presented in the preceding technical field, background, brief summary, or the following detailed description.

0027For simplicity and clarity of illustration, the drawing figures depict the general structure and/or manner of construction of the various embodiments. Descriptions and details of well-known features and techniques may be omitted to avoid unnecessarily obscuring other features. Elements in the drawings figures are not necessarily drawn to scale: the dimensions of some features may be exaggerated relative to other elements to assist improve understanding of the example embodiments.

0028Terms of enumeration such as “first,” “second,” “third,” and the like may be used for distinguishing between similar elements and not necessarily for describing a particular spatial or chronological order. These terms, so used, are interchangeable under appropriate circumstances. The embodiments of the invention described herein are, for example, capable of use in sequences other than those illustrated or otherwise described herein.

0029The terms “comprise,” “include,” “have” and any variations thereof are used synonymously to denote non-exclusive inclusion. The term “exemplary” is used in the sense of “example,” rather than “ideal.”

0030In the interest of conciseness, conventional techniques, structures, and principles known by those skilled in the art may not be described herein, including, for example, standard magnetic random access memory (MRAM) process techniques, fundamental principles of magnetism, and basic operational principles of memory devices.

0031In general, what is described herein is a spin-torque magnetoresistive memory device structure with a high magnetoresistance ratio and a low spin-torque critical current density. The structure includes a free layer positioned between first and second electrodes, a first tunnel barrier positioned between the first electrode and the free layer forming a first tunnel junction, and a second tunnel barrier positioned between the second electrode and the free layer forming a second tunnel junction. The tunnel barriers may comprise MgO, for example. One or both of the first and second electrodes may comprise fixed magnetic layers which act as spin polarizers that provide polarized tunneling currents. In a first exemplary embodiment in which the first and second electrodes comprise spin polarizers, the first electrode comprises a ferromagnetic alloy with low Fe content and a high B content compared to the second electrode, and the first tunnel junction has a lower resistance-area product (RA) compared to the second. In a second exemplary embodiment, the free layer includes a high-Fe interface region in contact with the second tunnel barrier. In a third exemplary embodiment, the first electrode is a spin polarizer, the second electrode is a non-ferromagnetic material, and the first tunnel junction has a higher RA compared to the second. The free layer may include an optional high-Fe interface region in contact with the first tunnel barrier and it may include a second high-Fe interface region in contact with the second tunnel barrier. In a fourth exemplary embodiment, the free layer is a compositionally-modulated structure comprising layers of ferromagnetic material, preferably a CoFeB alloy, separated by one or more thinner layers including a non-ferromagnetic transition metal, such as Ta, Nb, V, Zr, or Ru. The non-ferromagnetic transition metal lowers the magnetization of the free layer which thereby allows for thicker layers that typically have better magnetic switching characteristics, and it can be used to tune the exchange coupling for reduced spin-torque critical current. The free layer may comprise multiple alternating layers of CoFeB and thinner layers comprising one or more non-ferromagnetic transition metals.

0032During the course of this description, like numbers are used to identify like elements according to the different figures that illustrate the various exemplary embodiments.

0033The spin-torque effect is known to those skilled in the art. Briefly, a current becomes spin-polarized after the electrons pass through the first magnetic layer in a magnetic/non-magnetic/magnetic trilayer structure, where the first magnetic layer is substantially more stable than the second magnetic layer. The higher stability of the first layer compared to the second layer may be determined by one or more of several factors including: a larger magnetic moment due to thickness or magnetization, coupling to an adjacent antiferromagnetic layer, coupling to another ferromagnetic layer as in a SAF structure, or a high magnetic anisotropy. The spin-polarized electrons cross the nonmagnetic spacer and then, through conservation of spin angular momentum, exert a spin torque on the second magnetic layer that causes precession of the its magnetic moment and switching to a different stable magnetic state if the current is in the proper direction. When net current of spin-polarized electrons moving from the first layer to the second layer exceeds a first critical current value, the second layer will switch its magnetic orientation to be parallel to that of the first layer. If a bias of the opposite polarity is applied, the net flow of electrons from the second layer to the first layer will switch the magnetic orientation of the second layer to be antiparallel to that of the first layer, provided the magnitude of the current is above a second critical current value. Switching in this reverse direction involves a fraction of the electrons reflecting from the interface between the spacer and the first magnetic layer and traveling back across the nonmagnetic spacer to interacting with the second magnetic layer.

0034Magnetoresistance is the property of a material to change the value of its electrical resistance depending on its magnetic state. Typically, for a structure with two ferromagnetic layers separated by a conductive or tunneling spacer, the resistance is highest when the magnetization of the second magnetic layer is antiparallel to that of the first magnetic layer, and lowest when they are parallel. The MR is defined as (R_H−R_L)/R_L, where R_Land R_Hare the device resistance in the low and high resistance states, respectively. When the spacer layer is a dielectric tunnel barrier, the tunneling resistance is measured by the resistance-area product (RA), such that the tunneling resistance R of a device having an area α, for a tunneling current passing perpendicular to the film plane, is given by RA/α. As used herein, the term “film” is the equivalent of a thin layer, and the term “film plane” is a plane to the surface of a film or layer.

0035<figref idref="DRAWINGS">FIG. <b>1</b></figref> is a side sectional view of an MRAM device <b>100</b> configured in accordance with an exemplary embodiment. In practice, an MRAM architecture or device will include many MRAM devices <b>100</b>, typically organized in a matrix of columns and rows. The exemplary MRAM bit structure (or “stack”) <b>100</b> generally includes a free magnetic layer (or “free layer”) <b>102</b> separated from a top electrode <b>104</b> and a bottom electrode <b>106</b> by tunnel barriers <b>108</b> and <b>110</b>, respectively. Either or both of the tunnel barriers <b>108</b> and <b>110</b> may be dielectrics, typically oxides such as MgO or AlOx. For the purposes of clarity, some commonly-used layers have not been illustrated in the drawings, including various protective cap layers, seed layers, and the underlying substrate (which may be a conventional semiconductor substrate or any other suitable structure). For the exemplary embodiments described below, the bottom electrode <b>106</b> is a ferromagnetic polarizer, while the top electrode <b>104</b> may be either a non-ferromagnetic material or a ferromagnetic polarizer. Generally, a ferromagnetic polarizer would include a pinning layer, a pinned magnetic layer, a coupling spacer layer, and a fixed magnetic layer adjacent to the tunnel barrier (none of which are shown) as is well known in the industry.

0036The three layers including the free layer <b>102</b>, the tunnel barrier <b>110</b> and bottom electrode <b>106</b>, forms a first magnetic tunnel junction having a MR greater than zero and a first RA. The three layers including the free layer <b>102</b>, the tunnel barrier <b>108</b>, and top electrode <b>104</b>, forms a second magnetic tunnel junction having a MR equal to or greater than zero and a second RA. For the MRAM device <b>100</b> to have a good MR, it is desirable to have the second MR less than half the first MR, and the second RA is less than half of the first RA. Most preferably, the second MR is less than one-fourth the first MR, and the second RA is less than one-fourth of the first RA.

0037The difference in the RA of the two tunnel barriers can be adjusted by either changing the thickness of the tunnel barrier layers <b>108</b>, <b>110</b> or by using different doses of oxidation when forming the dielectrics. The MR for each junction can be adjusted by using thin layers at the tunnel barrier interfaces <b>107</b>, <b>109</b> that are either low polarization or high polarization interfacial layers, as well as through the choice of alloys for the bottom and top electrodes.

0038<figref idref="DRAWINGS">FIG. <b>2</b></figref> is a side sectional view of an MRAM device <b>200</b> configured in accordance with an exemplary embodiment in which both electrodes are spin polarizers resulting in what may be referred to as a dual spin-filter MTJ. The exemplary dual spin-filter MTJ bit structure (or “stack”) <b>200</b> generally includes a free magnetic layer (or “free layer”) <b>202</b> separated from a top electrode <b>204</b> and a bottom electrode <b>206</b> via tunnel barriers <b>208</b> and <b>210</b>, respectively. Both of the layers <b>208</b> and <b>210</b> are dielectrics.

0039Bottom electrode <b>206</b> has a fixed magnetization state that does not change when the free layer <b>202</b> is switched between its two or more stable states. In the practical embodiment, bottom electrode <b>206</b> may include a template or seed layer <b>212</b> formed on a conductor <b>222</b> for facilitating the formation thereon of a pinning layer <b>214</b> made from an antiferromagnetic material, for example, IrMn, PtMn, or FeMn. The template/seed layer <b>212</b> is preferably a non magnetic material, for example Ta, TaN, Al, Ru, but can also be a magnetic material, for example NiFe or CoFe. The template/seed layer <b>212</b> may include two layers or may be omitted in cases where conductor <b>222</b> provides the desired growth characteristics for the subsequent layers. The bottom electrode <b>206</b> of device <b>200</b> includes three ferromagnetic layers <b>216</b>, <b>226</b>, and <b>220</b>, antiferromagnetically coupled through coupling layers <b>228</b> and <b>213</b>. The pinning layer <b>214</b> determines the orientation of a magnetic moment of the pinned ferromagnetic layer <b>216</b> formed thereon. Ferromagnetic layer <b>226</b> is antiferromagnetically coupled to pinned layer <b>216</b>, through coupling layer <b>228</b>, so that their magnetic moments orient antiparallel in the absence of an external field and fixed layer <b>220</b> is antiferromagnerically coupled to ferromagnetic layer <b>226</b>, through coupling layer <b>213</b>, so their magnetic moments orient antiparallel in the absence of an external field. The ferromagnetic layers <b>216</b>, <b>226</b>, and <b>220</b> may be formed from any suitable magnetic material, such as at least one of the elements Ni, Fe, Co, or their alloys including alloys incorporating additional elements such as B, C, Ta, V, Zr, and others, as well as so-called half-metallic ferromagnets such as NiMnSb, PtMnSb, Fe₃O₄, or CrO₂. In one embodiment, for example, pinned magnetic layer <b>216</b> and ferromagnetic layer <b>220</b> comprises 20-50 Å of CoFe, fixed ferromagnetic layer <b>220</b> comprise about 20-30 Å of CoFeB, and free magnetic layer <b>202</b> comprises about 20-35 Å of CoFeB. Coupling layers <b>228</b> and <b>213</b> are formed from any suitable nonmagnetic material, including at least one of the elements Ru, Os, Re, Cr, Rh, Cu, Cr, or their combinations. Such synthetic antiferromagnet structures are known to those skilled in the art and, therefore, their operation will not be described in detail herein. Bottom electrode is chosen to be a SAF with three ferromagnetic layers in device <b>200</b> while top electrode <b>204</b> is chosen to be a SAF with two ferromagnetic layers so that the magnetization direction of fixed layer <b>220</b> and the magnetization direction of top fixed layer <b>232</b> will be substantially antiparallel when processed under typical conditions for an MTJ stack as described below.

0040In this illustration, arrows are used to indicate the direction of the magnetic moment, or magnetization, for individual layers. The magnetization directions of the top and bottom fixed layers <b>220</b> and <b>232</b> are typically set using a high-temperature anneal in a strong applied magnetic field. During the anneal, the ferromagnetic layers align with the strong magnetic field. When the field anneal is complete, the antiferromagnetic pinning material, such as that used in pinning layer <b>214</b>, provides an exchange bias to the adjacent ferromagnetic pinned layer in the direction of the applied field.

0041Top electrode <b>204</b> includes a non-magnetic layer (“spacer layer,” or “coupling layer”) <b>230</b> between two ferromagnetic layers <b>232</b> and <b>234</b>. The magnetic moments of ferromagnetic layers <b>232</b> and <b>234</b> are antiferromagnetically coupled through coupling layer <b>230</b>, so that their magnetic moments orient antiparallel in the absence of an external field. A top pinning layer <b>244</b> can be used to orient the magnetic moment of ferromagnetic layer <b>234</b>, in the same way that pinning layer <b>214</b> orients pinned layer <b>216</b>. After the field anneal, the top pinned layer <b>234</b> and the bottom pinned layer <b>216</b> will be biased in the same direction by the pinning material. Since the SAF that forms the bottom electrode has one more ferromagnetic layer than does the top electrode, the magnetization of the bottom fixed layer <b>220</b> will be set in a direction antiparallel to the magnetization of the top fixed layer <b>232</b>, providing the necessary magnetic configuration for additive contributions from both fixed layers to the spin torque transferred to the free layer <b>202</b>.

0042Top electrode <b>204</b> is a synthetic antiferromagnet (SAF) in that it comprises two ferromagnetic layers separated by a non-magnetic coupling layer, the thickness of the coupling layer chosen to provide strong antiferromagnetic coupling between the two ferromagnetic layers. The useful materials for the layers in top electrode <b>204</b> are the same as for bottom electrode <b>206</b>. In one embodiment, for example, top pinned magnetic layer <b>234</b> comprises 20-30 Å of CoFe and ferromagnetic fixed layer <b>232</b> comprises about 20-30 Å of CoFeB. It is known in the art, for example, U.S. Pat. 7,605,437, that there can be advantages to eliminating the top pinning layer <b>244</b> and instead designing electrode <b>204</b> to work as an “unpinned” SAF. The magnetic orientation of the unpinned SAF can be set by designing a magnetic asymmetry into the structure, such as a moment imbalance between the ferromagnetic layers of the SAF <b>232</b> and <b>234</b>.

0043It is desirable for the magnetic moments of fixed layers <b>220</b> and <b>232</b> to be relatively unaffected by spin-transfer torque from free layer <b>202</b>, so that only the direction of the free layer <b>202</b> changes when a write current is applied. The fixed layers are made stable by the strong coupling between the layers in each SAF structure and the large magnetic volume of the SAFs compared to the free layer <b>202</b>. The strong exchange coupling to the pinning material contributes additional stability in addition to defining a reference direction.

0044First and second conductor <b>222</b>, <b>224</b> are formed from any suitable material capable of conducting electricity. For example, conductors <b>222</b>, <b>224</b> may be formed from at least one of the elements Al, Cu, Ta, TaNx, Ti or their combinations. The various ferromagnetic layers may comprise any suitable material having the desired ferromagnetic properties as described above. It is advantageous to have the net magnetic coupling experienced by the free layer to be near zero so that the switching characteristics of the free layer are symmetric. This can be achieved by adjusting the thickness of each ferromagnetic layer in the top and bottom electrodes. There is typically a ferromagnetic coupling between a fixed layer and the free layer, due to various mechanisms known in the art. When both top and bottom fixed layers are present, and oriented in opposite directions as shown in device <b>200</b>, the ferromagnetic interlayer coupling of one fixed layer opposes that of the other, reducing the net coupling. There is typically antiferromagnetic coupling between the layers in a patterned magnetic structure due to the poles that form at the patterned edges of the layers. Since the magnetization of each layer in a SAF structure is opposite to the nearest other ferromagnetic layer in SAF, they have a cancelling effect on each other. In a bottom electrode comprising a three-layer SAF as shown in device <b>200</b>, the middle ferromagnetic layer <b>226</b> is typically designed to have a higher magnetic moment than ferromagnetic layers <b>216</b> and <b>220</b> so that the dipolar field created by layer <b>226</b> substantially cancels the dipolar fields created by layers <b>216</b> and <b>220</b>. In an optimized structure, the layer thicknesses are adjusted so that all the sources of coupling experienced by the free layer sum to near zero.

0045In one embodiment, coupling layers <b>228</b>, <b>213</b>, <b>230</b> comprise Ru having a thickness of approximately 8 Å. In an alternate embodiment, some or all of the coupling layers may comprise a material, such as Ti or Ta, that does not produce any antiparallel coupling between continuous magnetic films, but merely causes exchange decoupling between the magnetic films. In this embodiment, the ferromagnetic SAF layers will be antiferromagnetically coupled due to the magnetostatic dipolar fields generated at the patterned edges of each layer. These alternate coupling layers will be useful for devices patterned to dimensions less than approximately 30 nm because this type of magnetostatic coupling is stronger for smaller patterned shapes.

0046During fabrication of MRAM structure <b>200</b>, each succeeding layer (i.e., layers <b>222</b>, <b>212</b>, <b>214</b>, <b>216</b>, <b>228</b>, <b>226</b>, <b>213</b>, <b>220</b>, <b>210</b>, <b>202</b>, <b>208</b>, <b>232</b>, <b>230</b>, <b>234</b>, <b>244</b>, <b>224</b>) is deposited or otherwise formed in sequence and each MRAM bit may be defined by selective deposition, photolithography processing, and etching in accordance with any of the various conventional techniques known in the semiconductor industry. During deposition of the various fixed and free magnet layers, a magnetic field may be provided to set a preferred easy magnetic axis of the layer (i.e., via induced anisotropy). Similarly, a strong magnetic field applied during the post-deposition high-temperature anneal step may be used to induce a preferred easy axis and a preferred pinning direction for any antiferromagnetically pinned materials.

0047Free magnetic layer <b>202</b> is formed from a ferromagnetic material having two or more stable magnetic states. For example, free magnetic element <b>202</b> may be formed of various ferromagnetic alloys comprising at least one of the elements Ni, Fe, and Co. Additional elements are added to the alloys to provide improved magnetic, electrical, or microstructural properties. As with conventional MRAM devices, the direction of the magnetization of free magnetic element <b>202</b> determines the resistance of the element. In practice, for a two-state device, the direction of the magnetization of free magnetic element <b>202</b> is either parallel or anti-parallel to the magnetization of a fixed magnetic layer, resulting in a low or high resistance representing a “0” bit state or a “1” bit state. Furthermore, the free magnetic element <b>202</b> may have an in-plane magnetization while the ferromagnetic spin polarizer has out-of-plane magnetization.

0048Free magnetic layer <b>202</b> has a magnetic easy axis that defines a natural or “default” axis of its magnetization. When MRAM device <b>200</b> is in a steady state condition with no current applied from conductor <b>222</b> to conductor <b>224</b>, the magnetization of free magnetic element <b>202</b> will naturally point along its easy axis. MRAM device <b>200</b> is suitably configured to establish a particular easy axis direction for free magnetic element <b>202</b>. From the perspective of <figref idref="DRAWINGS">FIG. <b>2</b></figref>, the easy axis of free magnetic element <b>202</b> points either to the right or to the left. In practice, MRAM device <b>200</b> utilizes anisotropy, such as shape, crystalline, or interface anisotropy, in the free magnetic layer <b>202</b> to achieve the orientation of the respective easy axes. It is understood by those skilled in the art that some materials have a strong perpendicular anisotropy which can be used to make free layers with the two magnetic states lying along a perpendicular easy axis so the two magnetic states are up and down in <figref idref="DRAWINGS">FIG. <b>2</b></figref>. For such devices, one or more perpendicular fixed layer is also used.

0049In addition to carrying the write current, conductors <b>222</b> and <b>224</b> also serve as the data read conductors for MRAM device <b>200</b>. In this regard, data in MRAM device <b>200</b> can be read in accordance with conventional techniques: a small current flows through MRAM device <b>200</b> and electrode <b>224</b>, and that current is measured to determine whether the resistance of MRAM device <b>200</b> is relatively high or relatively low. The read current is much smaller than the current required to switch the free layer by spin-torque in order to avoid disturbs caused by reading the cell.

0050In practice, MRAM device <b>200</b> may employ alternative and/or additional elements, and one or more of the elements depicted in <figref idref="DRAWINGS">FIG. <b>2</b></figref> may be realized as a composite structure or combination of sub-elements. The specific arrangement of layers shown in <figref idref="DRAWINGS">FIG. <b>2</b></figref> merely represents one suitable embodiment of the invention.

0051In order to determine a change in state of a magnetic element, a magnetoresistance must be sufficiently high. Three exemplary embodiments are described herein for providing this high magnetoresistance along with a low critical current density (Jc). For a structure with two tunnel junctions, the MR is maximized when one junction dominates the resistance change by having a much larger resistance change than the other junction when the free layer changes state. This is best accomplished by having a dominant junction with both a larger MR and a larger RA than the other junction.

0052In the present invention, high MR of the dominant tunnel junction is accomplished by using higher Fe content at the tunnel-barrier interfaces of the dominant junction as compared to the other junction. To make the tunnel junction formed by layers <b>208</b>, <b>202</b>, and <b>232</b> (<figref idref="DRAWINGS">FIG. <b>2</b></figref>) the high-MR junction, the surfaces in contact with tunnel barrier <b>208</b>, which are the surfaces of the adjacent ferromagnetic layers <b>202</b> and <b>232</b>, should comprise a higher Fe content than the surfaces in contact with tunnel barrier <b>210</b>. In accordance with one exemplary embodiment, it is preferred that fixed layer <b>220</b>, which is the fixed ferromagnetic layer adjacent to tunnel barrier <b>210</b>, have a low Fe content of less than 20% by atomic composition and a B content greater than 20% by atomic composition, and more preferably a low Fe content of approximately 5% and a B content of approximately 25% by atomic composition. The fixed layer <b>232</b>, adjacent to tunnel barrier <b>208</b> may have an Fe content greater than 20% by atomic composition and a B content of between 14% and 20% by atomic composition.

0053Referring to <figref idref="DRAWINGS">FIG. <b>3</b></figref> and in accordance with another exemplary embodiment, the device <b>300</b> includes a deposition of a small amount of iron (Fe) between the tunnel barrier <b>208</b> and the free magnetic layer <b>202</b>. The thin Fe interface deposition may form a continuous atomic layer of Fe or may mix with the underlying free ferromagnetic alloy in the final annealed structure, resulting in a high-Fe interface region <b>302</b> adjacent to tunnel barrier <b>208</b>. It should be noted that all components of this exemplary embodiment as illustrated in <figref idref="DRAWINGS">FIG. <b>3</b></figref> that are similar to components of the exemplary embodiment of <figref idref="DRAWINGS">FIG. <b>2</b></figref> are designated with like numbers. The top electrode <b>204</b> shown in <figref idref="DRAWINGS">FIG. <b>3</b></figref> as an unpinned SAF as described previously, and the bottom electrode <b>206</b> is a pinned SAF with two ferromagnetic layers. The amount of the Fe deposition may be in the range of 0.5 Å to 5 Å, but preferably is in the range of 1.5 Å-3 Å, expressed in equivalent continuous film thickness (see U.S. Pat. 7,098,495 assigned to the assignee of the present application regarding high polarization insertion layers). By adding a small amount of pure Fe at the interface between tunnel barrier <b>208</b> and the free layer <b>202</b>, to form high-Fe interface region <b>302</b>, one can obtain higher MR values even when free layer <b>202</b> is predominantly comprised of low-Fe, high-B CoFeB alloy. Whether the high-Fe interface region includes a continuous atomic layer of Fe, a discontinuous layer of Fe, or an interfacial layer of high-Fe alloy, it does result in at least an atomic layer of material at the surface of the free layer composed mainly of Fe atoms. That is, the interface region <b>302</b> will be at least 50% Fe by atomic percentage. Adding Fe under tunnel barrier <b>208</b> also improves the growth of <b>208</b> and increases RA for a given tunnel barrier process. The deposited Fe also improves the growth of the (001) crystallographically-oriented MgO layer upon it. The preferred device design has a low RA and low MR for the tunnel junction formed by tunnel barrier <b>210</b> through use of a low-RA tunnel barrier process and low-Fe, high-B alloys for fixed layer <b>220</b> and free layer <b>202</b>, combined with a high RA and high MR for the tunnel junction formed by tunnel barrier <b>208</b> by forming a high-Fe interface region <b>302</b> below tunnel barrier <b>208</b> combined with a high-RA tunnel barrier process. The Fe-rich surface may provide higher perpendicular interface anisotropy energy than for the case where the top tunnel barrier <b>208</b> is grown on a free layer, for example, of a CoFeB alloy, without the Fe deposition. The perpendicular interface anisotropy is desirable since it lowers the spin-torque switching critical current Ic by offsetting some of the thin-film demagnetization anisotropy that results in a strong in-plane anisotropy. The interfacial perpendicular anisotropy lowers Ic by making it easier for the free layer moment to precess out of plane as needed in the switching process.

0054In another aspect of this invention, it has been found that inserting certain materials into the free layer can increase the MR of the top tunnel junction formed by tunnel barrier <b>208</b> and sometimes decrease the MR of the bottom tunnel junction formed by tunnel barrier <b>210</b>.

0055In still another exemplary embodiment is device <b>400</b> as shown in <figref idref="DRAWINGS">FIG. <b>4</b></figref>, the free layer <b>202</b> comprises a thin layer <b>402</b> comprising either Ta or Ru positioned between a first portion <b>404</b> and a second portion <b>406</b>. The Ta insertion layer deposition is chosen to be of a thickness that does not form a continuous Ta layer, which would break the exchange between the adjacent layers, but rather mixes with the other free layer materials or forms a layer that is not continuous so that the adjacent ferromagnetic layers <b>404</b> and <b>406</b> are directly exchange coupled to each other and the entire structure <b>202</b> acts as a single ferromagnetic free layer. The typical thickness of Ta deposited to achieve this effect is less than 3.5 Å, and preferably in the range between 1 Å and 3 Å. Other similar materials that form alloys with Co, Fe, or Ni may yield similar results, for example: V, Zr, Ti, Nb, Mo, W, Hf, Mn, or Cr. The Ru insertion layer thickness is chosen to be of a thickness that results in a continuous layer for antiferromagnetic coupling but may have gaps when ferromagnetic coupling is desired, thus being as thin as 2 Angstrom, with little or no alloying with the adjacent ferromagnetic layers <b>404</b> and <b>406</b>. For Ru and similar materials, the ferromagnetic layers <b>404</b> and <b>406</b> are coupled through this nonmagnetic layer by the well-known oscillatory exchange coupling effect and are considered non-ferromagnetic coupling layers. The coupling strength associated with such non-ferromagnetic coupling layers is controlled by the layer thickness, preferably between 2 Angstrom and 30 Angstrom, and most often between 5 Angstrom and 15 Angstrom. Other similar materials that may yield similar results include: Rh, Os, Cu, Cr, Pd, Pt, or Ir. It should be noted that all components of this exemplary embodiment as illustrated in <figref idref="DRAWINGS">FIG. <b>4</b></figref> that are similar to components of the exemplary embodiment of <figref idref="DRAWINGS">FIG. <b>3</b></figref> are designated with like numbers.

0056The graph <b>500</b> of <figref idref="DRAWINGS">FIG. <b>5</b></figref> shows experimental data of magnetoresistance (MR) versus resistance-area product (RA) for single junctions of a top tunnel barrier <b>502</b> with the Ta insertion <b>402</b> in the free layer <b>202</b> (data <b>502</b>) compared with a top tunnel barrier without the Ta insertion <b>402</b> (data <b>504</b>), and a bottom tunnel barrier with the Ta insertion <b>402</b> in the free layer <b>202</b> (data <b>508</b>) compared with a bottom tunnel barrier without the Ta insertion <b>402</b> (data <b>506</b>). In the case where the top tunnel barrier forms the dominant magnetic tunnel junction of a dual tunnel barrier device, adding the Ta insertion <b>402</b> can be expected to increase the MR of the device since it would enhance the MR of the dominant junction and reduce the MR of the other junction. The symbols are measured data points for MTJ stacks made with MgO tunnel barriers and CoFeB ferromagnetic layers having a high-Fe interface region <b>302</b> for the top tunnel-barrier stacks. The various RA values were obtained by varying the oxidation dose of the tunnel barrier. Note that, to improve MR in a dual-spin-filter structure, the MR of the bottom junction can be further lowered by using a low-Fe alloy for the bottom fixed layer <b>220</b>.

0057The graph <b>600</b> of <figref idref="DRAWINGS">FIG. <b>6</b></figref> shows magnetoresistance (MR) versus RA data for the structure of <figref idref="DRAWINGS">FIG. <b>4</b></figref> with the Ta insertion <b>402</b> in the free layer <b>202</b> (data <b>602</b>) compared with the structure of <figref idref="DRAWINGS">FIG. <b>3</b></figref> without the Ta insertion <b>402</b> (data <b>604</b>). It is seen that the Ta insertion <b>402</b> provides an improvement in MR averaging approximately 10 percentage points over the resistance-area product range shown.

0058The graph <b>700</b> of <figref idref="DRAWINGS">FIG. <b>7</b></figref> shows magnetoresistance (MR) versus RA for the structure of <figref idref="DRAWINGS">FIG. <b>4</b></figref> with a Ru layer <b>402</b> in the free layer <b>202</b> (data <b>702</b>) compared with the same type of data for the structure of <figref idref="DRAWINGS">FIG. <b>3</b></figref> without a Ru layer <b>402</b> (data <b>704</b>), after a field anneal at 300 degrees C. It is seen that the Ru layer <b>702</b> provides an improvement in MR of 30 to 50 percentage points over the resistance-area product range shown.

0059<figref idref="DRAWINGS">FIG. <b>8</b></figref> is a side sectional view of a free layer <b>802</b> configured in accordance with another exemplary embodiment which may be used in lieu of the free layer <b>202</b> of <figref idref="DRAWINGS">FIGS. <b>2</b>, <b>3</b></figref>, and <b>4</b>. Two insertion layers <b>812</b> are deposited between ferromagnetic material layers <b>814</b>, <b>816</b>, <b>820</b>, wherein the material used for the insertions and the amount of the material are chosen as described for insertion layer <b>402</b> in <figref idref="DRAWINGS">FIG. <b>4</b></figref>. Fe is deposited on the top ferromagnetic layer <b>820</b> to form high-Fe interface region <b>302</b>, and an optional high-Fe interface region <b>818</b> is formed on the bottom tunnel barrier <b>210</b> by depositing Fe on tunnel barrier <b>210</b> before depositing ferromagnetic layer <b>814</b>. Though only two insertion layers are shown, additional such layers could be formed within the free layer <b>802</b>. In one preferred embodiment, both insertion layers <b>812</b> comprise Ta depositions including between 0.5 Å and 3.5 Å of continuous-film-equivalent material, although the layers are not continuous and do not break the exchange coupling between the ferromagnetic layers <b>814</b>, <b>816</b>, <b>820</b>. In the final structure, the regions of Ta deposition <b>812</b> are thin films of Ta-rich ferromagnetic alloy or discontinuous regions of Ta lying in the plane between the ferromagnetic materials. The additional Ta reduces the magnetization of the composite free layer material, enabling thicker free layers for a given desired magnetic moment, resulting in better magnetic properties than the thinner layers of the same CoFeB alloy without the Ta insertions. In a second preferred embodiment, one of the insertion layers <b>812</b> comprises Ta as described above and another comprises Ru with a thickness chosen as described above for layer <b>402</b> of <figref idref="DRAWINGS">FIG. <b>4</b></figref>. A Ru insertion layer with an optimized coupling strength can provide reduced switching current (as described in US Patent Publication 2009/0096042) while combining with the desirable reduced magnetization and perpendicular anisotropy provided by the other insertion layer and Fe-rich surface layer described in this invention. Note that ferromagnetic layers <b>814</b>, <b>816</b>, and <b>820</b> do not need to be the same thickness or material, but can contain different materials and compositions and thicknesses as desired for optimum performance. Ferromagnetic layers <b>814</b>, <b>816</b>, and <b>820</b> are preferably thin-film depositions of CoFeB alloy with less than 10% Fe and more than 14% B, and most preferably about 5% Fe and 25% B by atomic concentration, each deposition in the thickness range of 5 Å to 20 Å as needed to obtain the desired total magnetic moment for the free layer <b>802</b>.

0060As described previously for high-Fe interface region <b>302</b> in <figref idref="DRAWINGS">FIG. <b>3</b></figref>, the high-Fe interface regions <b>302</b> and <b>818</b> in <figref idref="DRAWINGS">FIG. <b>8</b></figref> may include a continuous atomic layer of Fe, a discontinuous layer of Fe, or an interfacial layer of high-Fe alloy, resulting in at least an atomic layer of material at the surface of the free layer composed mainly of Fe atoms. That is, the high-Fe interface regions <b>302</b> and <b>818</b> will be at least 50% Fe by atomic percentage. The Fe-rich interface regions may provide a higher perpendicular interface anisotropy energy than for the case where the top tunnel barrier <b>208</b> is grown on a surface without the Fe deposition.

0061<figref idref="DRAWINGS">FIG. <b>9</b></figref> is a side sectional view of an MRAM device <b>900</b> configured in accordance with an exemplary embodiment in which the bottom electrode <b>206</b> (<figref idref="DRAWINGS">FIG. <b>3</b></figref>) includes a ferromagnetic fixed layer <b>220</b> in contact with the bottom tunnel barrier <b>210</b> and the top electrode <b>204</b> of <figref idref="DRAWINGS">FIG. <b>3</b></figref> is replaced by a non-ferromagnetic material <b>930</b>. This double-tunnel-barrier structure is not a dual spin filter since it has a ferromagnetic layer polarizing the tunneling electrons on one side only. However, it is found that the top tunnel junction formed by tunnel barrier <b>208</b> enables a significant reduction in the critical current Ic required to switch the free layer, even though top electrode <b>930</b> is not ferromagnetic. The improvement may arise from magnetic heating of the free layer by the electrons tunneling through tunnel barrier <b>208</b> and from perpendicular interface anisotropy resulting from the interface between the surface of the free layer <b>202</b> and the top tunnel barrier <b>208</b> as described above for device <b>802</b> in <figref idref="DRAWINGS">FIG. <b>8</b></figref>. Since top electrode <b>930</b> is not ferromagnetic, MR for the junction formed by tunnel barrier <b>208</b> is zero and this junction will be the non-dominant junction in the double-tunnel-barrier device. For maximum MR of device <b>900</b>, the resistance of the tunnel junction formed by tunnel barrier <b>208</b> should be much less than that of the dominant junction formed by tunnel barrier <b>210</b>. The RA of barrier <b>210</b> should be at least two times higher than RA of barrier <b>208</b>, and most preferably it should be more than four times higher. The double-barrier structure <b>900</b> is useful because it provides many of the benefits of the dual spin filter structures <b>200</b>, <b>300</b>, and <b>400</b> but with a simpler and thinner top electrode, making the material stack much easier to pattern into devices. However, the choice of materials for the non-ferromagnetic top electrode <b>930</b> and the process for making the top tunnel barrier <b>208</b> are critical to proper functionality of the devices as described below. A further benefit of double-barrier structure <b>900</b> is that the top tunnel junction, having low RA and no magnetoresistance as described above, does not produce a resistance change opposite to the dominant bottom tunnel junction, resulting in a higher MR for the device <b>900</b> compared to dual spin filter devices <b>200</b>, <b>300</b>, and <b>400</b>.

0062The free layer <b>202</b> includes a ferromagnetic layer <b>201</b> and surface layer <b>302</b> in device <b>900</b> is most preferably the free layer <b>802</b> as shown in <figref idref="DRAWINGS">FIG. <b>8</b></figref> and described above. In the preferred embodiment, the ferromagnetic layers are a CoFeB alloy. Optionally and most preferably, a high-Fe interface region <b>818</b> is formed on tunnel barrier <b>210</b> by depositing Fe on tunnel barrier <b>210</b>. Preferably a thin Fe layer is deposited on top of the final ferromagnetic layer to form the high-Fe interface region <b>302</b> below the top tunnel barrier <b>208</b>. Since the tunnel junction formed by tunnel barrier <b>208</b> has little or no MR due to the non-ferromagnetic top electrode <b>930</b>, this Fe deposition does not affect MR as it does for the dual spin filter structures, but has been found to promotes the growth of high quality MgO for tunnel barrier <b>208</b> as well as promoting perpendicular magnetic anisotropy at the interface between the free layer <b>202</b> and the tunnel barrier <b>208</b>. The MTJ stack of <b>900</b> using the free layer <b>802</b> provides all benefits of the free layer <b>802</b> and the double tunnel barrier device <b>900</b>.

0063It is desirable that the interface between tunnel barrier <b>208</b> and the top electrode <b>930</b> is of a very high quality so that the tunnel junction formed by tunnel barrier <b>208</b> will be free from defects, shorting, and excessive spatial variation of the tunneling current. To form a high quality interface, the choice material for top electrode <b>930</b> is important as is the material under tunnel barrier <b>208</b>. The material in contact with tunnel barrier <b>208</b> must have properties that allow for a sharp interface with the tunnel barrier dielectric, typically MgO. Since Fe and Co form such sharp interfaces, those materials and alloys based on those materials can be used for non-ferromagnetic top electrode <b>930</b> if they are very thin, preferably less than or equal to 15 Angstrom of deposited ferromagnetic alloy, and a layer of Ta or similar material is deposited on them to suppress their ferromagnetism to the point where the resulting layer is not ferromagnetic within the operating temperature range of the device. Examples of such Fe and Co alloys include Fe, Co, CoFe, and alloys incorporating B, C, Ta, Ti, V, Nb, Zr, W, Hf, Cr, Mo, and Mn. Examples of layers to deposit on these materials to suppress their ferromagnetism include Ta, Ti, V, Nb, Zr, W, Hf, Cr, Ru, Mo, and Mn. Alternatively a non-ferromagnetic material that forms a sharp interface with the dielectric may be deposited on tunnel barrier <b>208</b>, and may optionally be followed by one of the material combinations described above. The benefit of using a non-ferromagnetic layer first is to eliminate MR and any magnetic coupling to the free layer that would be associated with any residual ferromagnetic material at the interface. The benefit of also including one of the Fe or Co alloys over the non-ferromagnetic layer is to provide an amorphous layer that is very resistant to interdiffusion between the MTJ stack and materials from the top contact <b>224</b>. Examples of such top electrodes <b>930</b> include: Ru, Ru/Ta, CoFeB (&lt;15 Å)/Ta, and Ru/CoFeB(&lt;15 Å)/Ta.

0064As described above with regards to device <b>300</b> in <figref idref="DRAWINGS">FIG. <b>3</b></figref>, free layer <b>802</b> in <figref idref="DRAWINGS">FIG. <b>8</b></figref>, and device <b>900</b> in <figref idref="DRAWINGS">FIG. <b>9</b></figref>, an Fe-rich surface at the interface with MgO may provide a higher perpendicular interface anisotropy energy than for the case where the MgO is in contact with a typical CoFeB alloy used for the free layer, without the Fe deposition <b>302</b> applied. For MTJ devices having in-plane magnetic easy axes, the perpendicular interface anisotropy lowers the spin-torque switching critical current Ic by offsetting some of the thin-film demagnetization anisotropy that results in a strong in-plane anisotropy. The interfacial perpendicular anisotropy lowers Ic by making it easier for the free layer moment to precess out of plane as needed in the switching process. However, if the interface anisotropy is strong enough, and the moment of a ferromagnetic layer low enough, it is possible for the perpendicular interface anisotropy to overcome the in-plane thin-film demagnetization anisotropy, resulting in a film with a perpendicular easy axis. An additional embodiment of the present invention employs the free layer <b>802</b> of <figref idref="DRAWINGS">FIG. <b>8</b></figref> or the free layer <b>202</b> of <figref idref="DRAWINGS">FIG. <b>4</b></figref>, in a double tunnel barrier structure or dual spin filter structure, designed with strong perpendicular anisotropy and a magnetic moment low enough to have a perpendicular easy axis. In this case, the two stable states of the free layer will be with the magnetization vector pointed perpendicular to the plan up toward tunnel barrier <b>208</b> or down toward tunnel barrier <b>210</b>. Similar layers, with or without an antiferromagnetic pinning layer, can be used to form all or part of the bottom and top electrodes with a perpendicular magnetization.

0065<figref idref="DRAWINGS">FIG. <b>10</b></figref> is a flow chart that illustrates an exemplary embodiment of a process <b>1000</b> for fabricating an MRAM device having a high magnetoresistance and a low critical current density. It should be appreciated that process <b>1000</b> may include any number of additional or alternative tasks, the tasks shown in <figref idref="DRAWINGS">FIG. <b>10</b></figref> need not be performed in the illustrated order, and process <b>1000</b> may be incorporated into a more comprehensive procedure or process having additional functionality not described in detail herein. Moreover, one or more of the tasks shown in <figref idref="DRAWINGS">FIG. <b>10</b></figref> could be omitted from an embodiment of the process <b>1000</b> as long as the intended overall functionality remains intact.

0066The method <b>1000</b> for forming a spin-torque magnetoresistive element comprises: forming <b>1002</b> a first electrode; forming <b>1004</b> a first tunnel barrier over the first electrode; forming <b>1006</b> a free magnetic layer over the first tunnel barrier to form first magnetic tunnel junction, wherein the first magnetic tunnel junction has a MR greater than zero and a first RA, forming <b>1008</b> a second tunnel barrier over the free layer, and forming <b>1010</b> a second electrode over the second tunnel barrier to form a second magnetic tunnel junction, wherein the second magnetic tunnel junction has a second MR equal to or greater than zero and a second RA, and wherein the second MR is less than half the first MR, and the second RA is less than half the first RA. Forming the layers typically involves thin-film deposition processes known in the art, including but not limited to physical vapor deposition techniques such as ion beam sputtering and magnetron sputtering. Forming thin insulating layers, such as the tunnel barrier layers, may involve physical vapor deposition from an oxide target, such as by radio-frequency (RF) sputtering, or by deposition of a thin metallic film followed by an oxidation step, such as oxygen plasma oxidation, oxygen radical oxidation, or natural oxidation by exposure to a low-pressure oxygen environment. Devices are typically defined by photolithography and etching steps known in the fields of integrated circuit manufacturing and magnetoresistive sensor manufacturing.

0067Referring to <figref idref="DRAWINGS">FIG. <b>11</b></figref>, a method for forming a dual tunnel barrier structure including an insertion layer and an Fe deposition includes forming <b>1102</b> a first electrode over a substrate, the first electrode including a first fixed magnetic layer, forming <b>1104</b> a first tunnel barrier on the first fixed magnetic layer, forming <b>1106</b> a free layer on the first tunnel barrier to create a first magnetic tunnel junction having a first magnetoresistance and a first resistance-area product, wherein forming the free layer includes the steps of depositing a first ferromagnetic portion, depositing an amount of a non-ferromagnetic material corresponding to less than 4 angstroms in thickness, depositing a second ferromagnetic portion, and depositing an amount of iron corresponding to less than or equal to 5 Angstroms in thickness, the non-ferromagnetic material comprising at least one of Ta, Nb, Hf, Zr, Ti, W, Cr, and Mn. A second tunnel barrier is formed <b>1108</b> on the free layer, and a second electrode is formed <b>1110</b> on the second tunnel barrier, the second electrode including a second fixed magnetic layer in contact with the second tunnel barrier to create a second magnetic tunnel junction having a second magnetoresistance and a second resistance-area product, wherein the magnitude of the second magnetoresistance is at least twice that of the first magnetoresistance and the magnitude of the second resistance-area product is at least twice that of the first resistance-area product.

0068Referring to <figref idref="DRAWINGS">FIG. <b>12</b></figref>, a method for forming a dual tunnel barrier structure including a coupling insertion layer and an Fe deposition includes forming <b>1202</b> a first electrode over a substrate, the first electrode including a first fixed magnetic layer, forming <b>1204</b> a first tunnel barrier on the first fixed magnetic layer, forming <b>1206</b> a free layer on the first tunnel barrier to create a first magnetic tunnel junction having a first magnetoresistance and a first resistance-area product, wherein forming the free layer includes the steps of depositing a first ferromagnetic portion, depositing an amount of a non-ferromagnetic coupling material corresponding to a thickness of between 2 Angstrom and 30 Angstrom, depositing a second ferromagnetic portion, and depositing an amount of iron corresponding to a thickness of less than or equal to 5 Angstroms, the non-ferromagnetic coupling material comprising at least one of Ru, Rb, Ir, Pt, Pd, Cu, Cr, and Os. A second tunnel barrier is formed <b>1208</b> on the free layer, and a second electrode is formed <b>1210</b> on the second tunnel barrier, the second electrode including a second fixed magnetic layer in contact with the second tunnel barrier to create a second magnetic tunnel junction having a second magnetoresistance and a second resistance-area product, wherein the magnitude of the second magnetoresistance is at least twice that of the first magnetoresistance and the magnitude of the second resistance-area product is at least twice that of the first resistance-area product.

0069Referring to <figref idref="DRAWINGS">FIG. <b>13</b></figref>, a method for forming a double tunnel barrier structure including a non-ferromagnetic layer in contact with the second tunnel barrier and an optional Fe deposition includes forming <b>1302</b> a first electrode over a substrate, the first electrode including a first fixed magnetic layer, forming <b>1304</b> a first tunnel barrier on the first fixed magnetic layer, forming <b>1306</b> a free layer on the first tunnel barrier to create a first magnetic tunnel junction having a first magnetoresistance and a first resistance-area product, wherein forming the free layer includes the steps of depositing a first ferromagnetic portion, depositing an amount of a non-ferromagnetic material, depositing a second ferromagnetic portion, and optionally depositing an amount of iron corresponding to a thickness of less than or equal to 5 angstroms, the non-ferromagnetic material comprising an amount resulting in magnetic exchange coupling between the first and second ferromagnetic portions. A second tunnel barrier is formed <b>1308</b> on the free layer, and a second electrode is formed <b>1310</b> on the second tunnel barrier, the second electrode including a non-ferromagnetic layer in contact with the second tunnel barrier to create a second magnetic tunnel junction having no magnetoresistance and a second resistance-area product, wherein the magnitude of the second resistance-area product is less than half that of the first resistance-area product.

0070In summary, a magnetic element and fabricating method thereof is disclosed in which the MR is improved based on the inclusion of Fe at the high-magnetoresistance tunnel barrier, an Fe layer between the free layer and the top tunnel barrier, and a portion comprising a non-ferromagnetic transition metal within the free layer, such as an interlayer including Ta or an Ru coupling interlayer between first and second portions of the free layer.

0071While at least one exemplary embodiment has been presented in the foregoing detailed description, it should be appreciated that a vast number of variations exist. It should also be appreciated that the exemplary embodiment or exemplary embodiments are only examples, and are not intended to limit the scope, applicability, or configuration of the invention in any way. Rather, the foregoing detailed description will provide those skilled in the art with a convenient road map for implementing an exemplary embodiment of the invention, it being understood that various changes may be made in the function and arrangement of elements described in an exemplary embodiment without departing from the scope of the invention as set forth in the appended claims.