JP5095765B2 - Method for manufacturing magnetoresistive element - Google Patents

Description

  The present invention relates to a method of manufacturing a magnetoresistive element that senses magnetism by flowing a sense current in a direction perpendicular to the film surface of the magnetoresistive film.

  By using a giant magneto-resistive effect (GMR), the performance of a magnetic device, particularly a magnetic head, has been dramatically improved. In particular, the application of a spin-valve (SV film) to a magnetic head, MRAM (Magnetic Random Access Memory), etc. has brought great technical progress in the field of magnetic devices.

  The “spin valve film” is a laminated film having a structure in which a nonmagnetic spacer layer is sandwiched between two ferromagnetic layers, and is also called a spin-dependent scattering unit. The magnetization of one of these two ferromagnetic layers (referred to as “pinned layer” or “magnetization pinned layer”) is pinned by an antiferromagnetic layer, and the other (“free layer”, “magnetization free layer”, etc.) Magnetization) can be rotated in response to an external magnetic field. In the spin valve film, a huge change in magnetoresistance can be obtained by changing the relative angle between the magnetization directions of the pinned layer and the free layer.

  The magnetoresistive effect element using the spin valve film includes a CIP (Current In Plane) -GMR element, a CPP (Current Perpendicular to Plane) -GMR element, and a TMR (Tunneling Magneto Resistance) element. In the CIP-GMR element, a sense current is passed in parallel to the surface of the spin valve film, and in the CPP-GMR and TMR elements, a sense current is passed in a direction substantially perpendicular to the surface of the spin valve film. The method of passing a sense current perpendicular to the film surface is attracting attention as a technology for future high recording density heads.

  Here, in a metal CPP-GMR element in which a spin valve film is formed of a metal layer, a resistance change amount due to magnetization is small, and it is difficult to detect a weak magnetic field (for example, a magnetic field in a magnetic disk having a high recording density). is there.

  A CPP element using an oxide layer [NOL (nano-oxide layer)] including a current path in the thickness direction as a spacer layer has been proposed (see Patent Document 1). In this element, both the element resistance and the MR change rate can be increased by a current confinement [CCP (Current-confined-path)] effect. Hereinafter, this element is referred to as a CCP-CPP element.

JP 2002-208744 A

  Currently, magnetic storage devices such as HDDs (Hard Disk Drives) are used for applications such as personal computers and portable music players. However, as the usage of magnetic storage devices further expands in the future, and the higher density storage progresses, the requirement for reliability becomes more severe. For example, it is necessary to improve the reliability under a higher temperature condition or under a higher speed operating environment. For this purpose, it is desirable to improve the reliability of the magnetic head as compared with the conventional case.

  In particular, since the CCP-CPP element has a lower resistance than the conventional TMR element, it can be applied to a high-end magnetic storage device for server enterprises that requires a higher transfer rate. Such high-end applications are required to satisfy high density and high reliability at the same time. In these applications, it is desirable to improve the reliability at higher temperatures. In other words, it is necessary to use the CCP-CPP element under a harsher environment (such as a high temperature environment) and a more severe usage condition (such as reading information on a magnetic disk rotating at high speed).

  The CCP-CPP element has low resistance and high frequency response and high density recording compatibility and great merit, but because it is a very complicated form with a three-dimensional nanostructure, it is actually an ideal structure as designed Is difficult to realize. In order to realize servers and enterprise applications with strict specifications, it is necessary to realize a more ideal CCP structure.

  It is an object of the present invention to provide a magnetoresistive effect element that can be applied to a magnetic storage device for high-density storage and is improved in reliability.

In order to achieve the above object, one embodiment of the present invention includes a magnetization pinned layer in which the magnetization direction is substantially pinned in one direction, a magnetization free layer in which the magnetization direction changes in response to an external magnetic field, and the magnetization pinned In the manufacturing method of the magnetoresistive effect element having the insulating layer and the spacer layer including the metal layer penetrating the insulating layer, which is provided between the layer and the magnetization free layer, A second metal layer is formed on the first metal layer, the second metal layer is converted into a first insulating layer, and the first metal layer is A first conversion process for forming a first conductive layer penetrating the first insulating layer by being sucked into the first insulating layer is performed, and the first insulating layer and the first conductive layer are formed on the first insulating layer. A third metal layer is formed, and the third metal layer is converted into a second insulating layer. Second conversion processing for forming a second conductive layer that penetrates the second insulating layer only on the first conductive layer by said first conductive layer is sucked into the second insulating layer To form the metal layer having the first conductive layer and the second conductive layer, and the insulating layer having the first insulating layer and the second insulating layer, At least one of the first conversion process and the second conversion process is performed by irradiating the film surface by ionizing or plasmaizing a gas containing at least one element of argon, xenon, helium, neon, and krypton. Oxygen gas and nitrogen in an atmosphere obtained by ionizing or plasmaizing a gas containing at least one of argon, xenon, helium, neon, and krypton A gas containing at least one of the gases is supplied, and at least one of oxidation treatment and nitridation treatment is performed on the second metal layer or the third metal layer in an atmosphere of the ionized gas or the plasma gas. And performing a second step.

  The inventors of the present invention have intensively studied to achieve the above object. As a result, a magnetization fixed layer whose magnetization direction is fixed substantially in one direction, a magnetization free layer whose magnetization direction changes corresponding to an external magnetic field, and the magnetization fixed layer and the magnetization free layer are provided. In a magnetoresistive effect element having a so-called CCP structure, which has a formed insulating layer and a spacer layer including a metal layer penetrating the insulating layer, when the spacer layer is formed, the spacer layer is converted into a conventional one-stage conversion. Instead of performing by processing, the spacer layer can be formed relatively thick by performing by two-stage conversion processing according to the present invention.

  Therefore, the thickness of the insulating layer constituting the spacer layer is increased, and the leakage current is completely prevented from flowing in a region that should function as an insulating layer other than the current path layer formed in the spacer layer. Will be able to. As a result, the CCP effect can be exerted more strongly and stronger reliability can be realized.

  Further, since the insulation breakdown voltage of the insulating layer is improved by increasing the thickness of the insulating layer, it is possible to improve resistance to electrostatic breakdown such as ESD (Electric Static Discharge). A high ESD withstand voltage leads to an improvement in yield when the head is assembled to the HDD.

  Such a significant improvement in reliability performance can improve not only the HDD production but also the destruction and thermal resistance in every situation, so servers and high reliability are required. It will be suitable as a head for enterprise use. High-reliability heads that do not simply improve recording density are becoming increasingly important in the recent situation where HDD applications are expanding. Increasing the life of the head is very important in expanding the usage environment of the HDD. Such a highly reliable head is very effective as an HDD for car navigation where the thermal environment is severe.

  Of course, such a highly reliable head is used not only as a high value-added HDD as described above, but also as an HDD for ordinary personal computers, portable music players, mobile phones and the like used for consumer use. it can.

  In addition, although explained in full detail below, in the manufacturing method of this invention, the 1st conversion process and 2nd conversion process which were mentioned above are implemented with respect to the 2nd metal layer and the 3rd metal layer. In this case, the atoms constituting the first metal layer formed under the second metal layer during the conversion process move upward by obtaining movement energy, As a result, a metal layer (current path) is formed in the insulating layer.

  Therefore, in the above aspect, since the spacer layer is implemented by the two-stage conversion process, in the second-stage conversion process, a metal layer (fourth metal layer) having the same function as the first metal layer is formed. Separately, the insulating layer and the metal layer formed through the first conversion treatment may be formed between the third metal layer. In this case, the atoms constituting the fourth metal layer move into the third metal layer in the process of being converted into the insulating layer in the process of the second conversion process, thereby forming a current path. . As a result, the metal layer (current path) penetrating through the spacer layer can be reliably performed.

By passing through the manufacturing method described above, for example, a magnetization fixed layer whose magnetization direction is fixed substantially in one direction, a magnetization free layer whose magnetization direction changes in response to an external magnetic field, and the magnetization fixed layer And a magnetoresistive effect element having a spacer layer including a metal layer penetrating the insulating layer, the insulating layer being provided between the magnetic free layer and the magnetization free layer.
A magnetoresistive effect element can be obtained in which the metal layer forming the spacer layer has a difference between the oxygen concentration in the lower part on the substrate side and the oxygen concentration in the upper part on the opposite side of the substrate within 10 atomic%.

A magnetization pinned layer in which the magnetization direction is substantially pinned in one direction; a magnetization free layer whose magnetization direction changes corresponding to an external magnetic field; and the magnetization pinned layer and the magnetization free layer. In a magnetoresistive effect element having an insulating layer and a spacer layer including a metal layer penetrating the insulating layer,
A magnetoresistive effect element can be obtained in which the metal layer forming the spacer layer has a difference between the opening area of the lower part on the substrate side and the opening area of the upper part on the opposite side of the substrate within 20%.

  As described above, according to the present invention, it is possible to provide a magnetoresistive effect element which can be applied to a magnetic storage device of high density storage and has improved reliability.

It is a perspective view showing an example of the magnetoresistive effect element (CCP-CPP element) of this invention. It is a flowchart showing the manufacturing process for forming the spacer layer of the magnetoresistive effect element which concerns on embodiment of this invention. It is a schematic block diagram of the apparatus used in order to manufacture the magnetoresistive effect element (CCP-CPP element) which concerns on embodiment of this invention. It is an example of a structure of the oxidation chamber shown in FIG. It is a flowchart showing the manufacturing process for forming the spacer layer of the magnetoresistive effect element which concerns on embodiment of this invention. It is a figure which shows the state which incorporated the magnetoresistive effect element which concerns on embodiment of this invention in the magnetic head. It is a figure which shows the state which incorporated the magnetoresistive effect element which concerns on embodiment of this invention in the magnetic head. It is a principal part perspective view which illustrates schematic structure of a magnetic recording / reproducing apparatus. It is the expansion perspective view which looked at the head gimbal assembly ahead from an actuator arm from the disk side. It is a figure which shows an example of the matrix structure of the magnetic memory which concerns on embodiment of this invention. It is a figure which shows the other example of the matrix structure of the magnetic memory which concerns on embodiment of this invention. It is sectional drawing which shows the principal part of the magnetic memory which concerns on embodiment of this invention. It is sectional drawing which follows the A-A 'line of FIG.

  Hereinafter, other features and advantages of the present invention will be described based on the best mode for carrying out the invention with reference to the drawings.

(Magnetoresistive element)
FIG. 1 is a perspective view showing an example of a magnetoresistive effect element (CCP-CPP element) of the present invention. In the present specification, all drawings are schematically drawn, and the size of each component (film thickness, etc.) and the ratio between components are drawn differently from the actual ones. . As shown in FIG. 1, the magnetoresistive effect element according to the present embodiment includes a magnetoresistive effect film 10, and a lower electrode 11 and an upper electrode 20 sandwiching the magnetoresistive effect film 10 from above and below, and is configured on a substrate (not shown).

  The magnetoresistive film 10 includes an underlayer 12, a pinning layer 13, a pin layer 14, a lower metal layer 15, a spacer layer (CCP-NOL) 16 (insulating layer 161, current path 162), an upper metal layer 17, and a free layer 18. The cap layer 19 is sequentially laminated. Among them, the pinned layer 14, the lower metal layer 15, the spacer layer 16, the upper metal layer 17, and the free layer 18 correspond to a spin valve film in which a nonmagnetic spacer layer is sandwiched between two ferromagnetic layers. To do. Further, the lower metal layer 15, the spacer layer (CCP-NOL) 16, and the upper metal layer 17 are defined as a broad spacer layer. For ease of viewing, the spacer layer 16 is shown separated from the upper and lower layers (the lower metal layer 15 and the upper metal layer 17).

  Hereinafter, components of the magnetoresistive element will be described.

  The lower electrode 11 is an electrode for energizing in the direction perpendicular to the spin valve film. When a voltage is applied between the lower electrode 11 and the upper electrode 20, a current flows through the inside of the spin valve film along the direction perpendicular to the film. By detecting a change in resistance caused by the magnetoresistive effect with this current, magnetism can be detected. For the lower electrode 11, a metal layer having a relatively small electric resistance is used in order to pass a current through the magnetoresistive element. NiFe, Cu or the like is used.

  The underlayer 12 can be divided into, for example, a buffer layer 12a and a seed layer 12b. The buffer layer 12a is a layer for reducing the roughness of the surface of the lower electrode 11. The seed layer 12b is a layer for controlling the crystal orientation and crystal grain size of the spin valve film formed thereon.

  As the buffer layer 12a, Ta, Ti, W, Zr, Hf, Cr, or an alloy thereof can be used. The thickness of the buffer layer 12a is preferably about 2 to 10 nm, and more preferably about 3 to 5 nm. If the buffer layer 12a is too thin, the buffer effect is lost. On the other hand, if the buffer layer 12a is too thick, the series resistance that does not contribute to the MR change rate is increased. When the seed layer 12b formed on the buffer layer 12a has a buffer effect, the buffer layer 12a is not necessarily provided. As a preferred example of the above, Ta [3 nm] can be used.

  The seed layer 12b may be any material that can control the crystal orientation of the layer formed thereon. As the seed layer 12b, an fcc structure (face-centered cubic structure), an hcp structure (hexagonal close-packed structure), or a bcc structure (body-centered cubic structure) A metal layer having a structure is preferable. For example, by using Ru having an hcp structure or NiFe having an fcc structure as the seed layer 12b, the crystal orientation of the spin valve film thereon can be changed to the fcc (111) orientation. Further, the fct structure (face-centered tetragonal structure) in which the crystal orientation of the pinning layer 13 (for example, PtMn) is regularized, or bcc (body-centered cubic structure) (110) orientation It can be. Besides these, Cr, Zr, Ti, Mo, Nb, W, alloy layers thereof, and the like can also be used.

  In order to sufficiently exhibit the function as the seed layer 12b for improving the crystal orientation, the film thickness of the seed layer 12b is preferably 1 to 5 nm, more preferably 1.5 to 3 nm. As a preferred example of the above, Ru [2 nm] can be used.

  The crystal orientation of the spin valve film and the pinning layer 13 can be measured by X-ray diffraction. Good orientation is obtained by setting the full width at half maximum of the rocking curve at the fcc (111) peak of the spin valve film, the fct (111) peak of the pinning layer 13 (PtMn), or the bcc (110) peak to 3.5 to 6 degrees. be able to. The orientation dispersion angle can also be determined from a diffraction spot using a cross-sectional TEM.

As the seed layer 12b, a NiFe-based alloy (for example, Ni _x Fe _100-x (x = 90 to 50%, preferably 75 to 85%) or a third element X is added to NiFe instead of Ru. and the non-magnetic _{(Ni x Fe 100-x)} 100-y X y (X = Cr, V, Nb, Hf, Zr, Mo)) can also be used. In the NiFe-based seed layer 12b, it is relatively easy to obtain good crystal orientation, and the half-value width of the rocking curve measured in the same manner as described above can be 3 to 5 degrees.

  The seed layer 12b has not only a function of improving the crystal orientation but also a function of controlling the crystal grain size of the spin valve film. Specifically, the crystal grain size of the spin valve film can be controlled to 5 to 40 nm, and even when the size of the magnetoresistive element is reduced, a high MR change rate can be realized without causing variations in characteristics.

  The crystal grain size here can be determined by the grain size of the crystal grains formed on the seed layer 12b, and can be determined by a cross-sectional TEM or the like. When the pinned layer 14 is a bottom type spin valve film positioned below the spacer layer 16, the pinning layer 13 (antiferromagnetic layer) or the pinned layer 14 (magnetization pinned) formed on the seed layer 12b. It can be determined by the crystal grain size of the layer.

  In a reproducing head compatible with high-density recording, the element size is, for example, 100 nm or less. A large ratio of the crystal grain size to the element size causes variations in element characteristics. It is not preferable that the crystal grain size of the spin valve film is larger than 40 nm. Specifically, the crystal grain size is preferably in the range of 5 to 40 nm, and more preferably in the range of 5 to 20 nm.

  When the number of crystal grains per element area decreases, it may cause variation in characteristics due to the small number of crystals, so it is not preferable to increase the crystal grain size. In particular, it is not preferable to increase the crystal grain size in a CCP-CPP element in which a current path is formed. On the other hand, it is generally difficult to maintain good crystal orientation even if the crystal grain size becomes too small. A preferable range of the crystal grain size in consideration of the upper limit and the lower limit of the crystal grain size is 5 to 20 nm.

  However, in MRAM applications and the like, the element size may be 100 nm or more, and even if the crystal grain size is as large as about 40 nm, there may be no problem. That is, by using the seed layer 12b, the crystal grain size may be increased in some cases.

In order to obtain the crystal grain size of 5~20nm described above, as the seed layer 12b, Ru2nm _{and, (Ni x Fe 100-x} ) 100-y X y (X = Cr, V, Nb, Hf, Zr, Mo )) In the case of a layer, the composition y of the third element X is preferably about 0 to 30% (including the case where y is 0%).

  On the other hand, in order to use the crystal grain size coarser than 40 nm, it is preferable to use a larger amount of additive element. When the material of the seed layer 12b is, for example, NiFeCr, it is preferable to use a NiFeCr layer having a bcc structure with a Cr amount of about 35 to 45% and a composition showing a boundary phase between fcc and bcc.

  As described above, the thickness of the seed layer 12b is preferably about 1 nm to 5 nm, and more preferably 1.5 to 3 nm. If the thickness of the seed layer 12b is too thin, effects such as crystal orientation control are lost. On the other hand, if the thickness of the seed layer 12b is too thick, the series resistance may be increased, and unevenness at the interface of the spin valve film may be caused.

  The pinning layer 13 has a function of fixing magnetization by imparting unidirectional anisotropy to a ferromagnetic layer to be a pinned layer 14 formed thereon. As a material of the pinning layer 13, an antiferromagnetic material such as PtMn, PdPtMn, IrMn, RuRhMn can be used. Of these, IrMn is advantageous as a material for a head corresponding to a high recording density. IrMn can apply unidirectional anisotropy with a film thickness smaller than that of PtMn, and is suitable for narrowing the gap necessary for high-density recording.

  In order to impart sufficient strength of unidirectional anisotropy, the film thickness of the pinning layer 13 is appropriately set. When the material of the pinning layer 13 is PtMn or PdPtMn, the thickness is preferably about 8 to 20 nm, and more preferably 10 to 15 nm. When the material of the pinning layer 13 is IrMn, unidirectional anisotropy can be imparted even with a film thickness thinner than PtMn, and is preferably 3 to 12 nm, more preferably 4 to 10 nm. As a preferred example of the above, IrMn [7 nm] can be used.

As the pinning layer 13, a hard magnetic layer can be used instead of the antiferromagnetic layer. As a hard magnetic layer, for _{example, CoPt (Co = 50~85%)} , (Co x Pt 100-x) 100-y Cr y (x = 50~85%, y = 0~40%), FePt (Pt = 40-60%) can be used. Since the hard magnetic layer (especially CoPt) has a relatively small specific resistance, an increase in the series resistance and the area resistance RA can be suppressed.

The pinned layer 14 includes a lower pinned layer 141 (for example, Co ₉₀ Fe ₁₀ 3.5 nm), a magnetic coupling layer 142 (for example, Ru), and an upper pinned layer 143 (for example, Fe ₅₀ Co ₅₀ [1 nm] / Cu [0] .25 nm]) × 2 / Fe ₅₀ Co ₅₀ [1 nm]) is a preferred example. The pinning layer 13 (for example, IrMn) and the lower pinned layer 141 immediately above the pinning layer 13 are exchange-magnetically coupled so as to have unidirectional anisotropy. The lower pinned layer 141 and the upper pinned layer 143 above and below the magnetic coupling layer 142 are strongly magnetically coupled so that the magnetization directions are antiparallel to each other.

As a material of the lower pinned layer 141, for example, a Co _x Fe _100-x alloy (x = 0 to 100%), a Ni _x Fe _100-x alloy (x = 0 to 100%), or a nonmagnetic element is added thereto Can be used. Further, as the material of the lower pinned layer 141, a single element of Co, Fe, Ni or an alloy thereof may be used.

The magnetic film thickness (saturation magnetization Bs × film thickness t (Bs · t product)) of the lower pinned layer 141 is preferably substantially equal to the magnetic film thickness of the upper pinned layer 143. That is, it is preferable that the magnetic film thickness of the upper pinned layer 143 and the magnetic film thickness of the lower pinned layer 141 correspond. As an example, when the upper pinned layer 143 is (Fe ₅₀ Co ₅₀ [1 nm] / Cu [0.25 nm]) × 2 / Fe ₅₀ Co ₅₀ [1 nm], the saturation magnetization of FeCo in the thin film is about 2.2 T. Therefore, the magnetic film thickness is 2.2T × 3 nm = 6.6 Tnm. Since the saturation magnetization of Co ₉₀ Fe ₁₀ is about 1.8 T, the thickness t of the lower pinned layer 141 that gives the same magnetic thickness as above is 6.6 Tnm / 1.8T = 3.66 nm. Therefore, it is desirable to use Co ₉₀ Fe ₁₀ having a film thickness of about 3.6 nm. Further, when using IrMn as the pinning layer 13, it is preferable that the composition of the lower pinned layer 141 slightly increases the Fe composition as compared with Co ₉₀ Fe ₁₀ . Specifically, Co ₇₅ Fe ₂₅ or the like is an example of a preferred embodiment.

The thickness of the magnetic layer used for the lower pinned layer 141 is preferably about 1.5 to 4 nm. Based on the unidirectional anisotropic magnetic field strength due to the pinning layer 13 (eg, IrMn) and the antiferromagnetic coupling magnetic field strength between the lower pinned layer 141 and the upper pinned layer 143 via the magnetic coupling layer 142 (eg, Ru) . If the lower pinned layer 141 is too thin, the MR change rate becomes small. On the other hand, if the lower pinned layer 141 is too thick, it is difficult to obtain a sufficient unidirectional anisotropic magnetic field necessary for device operation. A preferable example is Co ₇₅ Fe ₂₅ having a film thickness of 3.6 nm.

  The magnetic coupling layer 142 (for example, Ru) has a function of forming a synthetic pin structure by causing antiferromagnetic coupling in the upper and lower magnetic layers (lower pinned layer 141 and upper pinned layer 143). The film thickness of the Ru layer as the magnetic coupling layer 142 is preferably 0.8 to 1 nm. Note that any material other than Ru may be used as long as the material causes sufficient antiferromagnetic coupling in the upper and lower magnetic layers. Instead of the film thickness of 0.8 to 1 nm corresponding to the 2nd peak of RKKY (Ruderman-Kittel-Kasuya-Yosida) bond, the film thickness of 0.3 to 0.6 nm corresponding to the 1st peak of RKKY bond can also be used. . Here, as an example, 0.9 nm of Ru, which can stably obtain a highly reliable coupling, is obtained.

As an example of the upper pinned layer 143, a magnetic layer such as (Fe ₅₀ Co ₅₀ [1 nm] / Cu [0.25 nm]) × 2 / Fe ₅₀ Co ₅₀ [1 nm] can be used. The upper pinned layer 143 forms part of the spin dependent scattering unit. The upper pinned layer 143 is a magnetic layer that directly contributes to the MR effect, and both the constituent material and the film thickness are important in order to obtain a large MR change rate. In particular, the magnetic material located at the interface with the spacer layer 16 is particularly important in that it contributes to spin-dependent interface scattering.

The effect of using the Fe ₅₀ Co ₅₀ having the bcc structure used here as the upper pinned layer 143 will be described. When a magnetic material having a bcc structure is used as the upper pinned layer 143, a large MR change rate can be realized because the spin-dependent interface scattering effect is large. Examples of the FeCo-based alloy having a bcc structure include Fe _x Co _100-x (x = 30 to 100%) and those obtained by adding an additive element to Fe _x Co _100-x . Among these, Fe ₄₀ Co ₆₀ to Fe ₆₀ Co ₄₀ satisfying various characteristics are examples of easy-to-use materials.

  When the upper pinned layer 143 is formed of a magnetic layer having a bcc structure that easily realizes a high MR ratio, the total thickness of the magnetic layer is preferably 1.5 nm or more. This is to keep the bcc structure stable. Since the metal material used for the spin valve film often has an fcc structure or an fct structure, only the upper pinned layer 143 may have a bcc structure. For this reason, if the film thickness of the upper pinned layer 143 is too thin, it becomes difficult to keep the bcc structure stable, and a high MR ratio cannot be obtained.

Here, as the upper pinned layer 143, Fe ₅₀ Co ₅₀ including an ultrathin Cu stack is used. Here, the upper pinned layer 143 is made of FeCo having a total film thickness of 3 nm and Cu having a thickness of 0.25 nm stacked for every 1 nm of FeCo, and has a total film thickness of 3.5 nm.

  The film thickness of the upper pinned layer 143 is preferably 5 nm or less. This is to obtain a large pin fixing magnetic field. In order to achieve both the large pinned magnetic field and the stability of the bcc structure, the film thickness of the upper pinned layer 143 having the bcc structure is preferably about 2.0 nm to 4 nm.

For the upper pinned layer 143, instead of a magnetic material having a bcc structure, a Co ₉₀ Fe ₁₀ alloy having an fcc structure widely used in a conventional magnetoresistive effect element or a cobalt alloy having an hcp structure may be used. it can. As the upper pinned layer 143, any single metal such as Co, Fe, Ni, or an alloy material containing any one of these elements can be used. When the magnetic material of the upper pinned layer 143 is arranged from those advantageous for obtaining a large MR ratio, an FeCo alloy material having a bcc structure, a cobalt alloy having a cobalt composition of 50% or more, and an Ni composition of 50% or more are obtained. It becomes the order of nickel alloy.

  As an example here, as the upper pinned layer 143, a magnetic layer (FeCo layer) and a nonmagnetic layer (ultra-thin Cu layer) stacked alternately can be used. In the upper pinned layer 143 having such a structure, a spin-dependent scattering effect called a spin-dependent bulk scattering effect can be improved by an ultrathin Cu layer.

  The “spin-dependent bulk scattering effect” is used as a paired term with the spin-dependent interface scattering effect. The spin-dependent bulk scattering effect is a phenomenon that manifests the MR effect inside the magnetic layer. The spin-dependent interface scattering effect is a phenomenon in which the MR effect is exhibited at the interface between the spacer layer and the magnetic layer.

  Hereinafter, the improvement of the bulk scattering effect by the laminated structure of the magnetic layer and the nonmagnetic layer will be described.

  In the CCP-CPP element, since the current is confined in the vicinity of the spacer layer 16, the contribution of resistance in the vicinity of the interface of the spacer layer 16 is very large. That is, the ratio of the resistance at the interface between the spacer layer 16 and the magnetic layer (the pinned layer 14 and the free layer 18) to the resistance of the entire magnetoresistive element is large. This indicates that the contribution of the spin-dependent interface scattering effect is very large and important in the CCP-CPP element. That is, the selection of the magnetic material positioned at the interface of the spacer layer 16 has an important meaning as compared with the case of the conventional CPP element. This is the reason why the FeCo alloy layer having the bcc structure having a large spin-dependent interface scattering effect is used as the pinned layer 143, as described above.

  However, the use of a material having a large bulk scattering effect cannot be ignored, and it is still important to obtain a higher MR change rate. The film thickness of the ultrathin Cu layer for obtaining the bulk scattering effect is preferably 0.1 to 1 nm, and more preferably 0.2 to 0.5 nm. If the thickness of the Cu layer is too thin, the effect of improving the bulk scattering effect is weakened. If the Cu layer is too thick, the bulk scattering effect may be reduced, and the magnetic coupling of the upper and lower magnetic layers via the nonmagnetic Cu layer will be weak, and the characteristics of the pinned layer 14 will be insufficient. . Therefore, in a preferable example, 0.25 nm of Cu was used.

  As a material for the nonmagnetic layer between the magnetic layers, Hf, Zr, Ti, Al, or the like may be used instead of Cu. On the other hand, when these ultrathin nonmagnetic layers are inserted, the film thickness of one magnetic layer such as FeCo is preferably 0.5 to 2 nm, more preferably about 1 to 1.5 nm.

  As the upper pinned layer 143, a layer obtained by alloying FeCo and Cu may be used instead of the alternately laminated structure of the FeCo layer and the Cu layer. Examples of such an FeCoCu alloy include (FexCo100-x) 100-yCuy (x = 30 to 100%, y = about 3 to 15%), but other composition ranges may be used. Here, as an element added to FeCo, other elements such as Hf, Zr, Ti, and Al may be used instead of Cu.

The upper pinned layer 143 may be a single layer film made of Co, Fe, Ni, or an alloy material thereof. For example, as the upper pinned layer 143 having the simplest structure, a 2 to 4 nm Co ₉₀ Fe ₁₀ single layer that has been widely used in the past may be used. Other elements may be added to this material.

  Next, a film configuration for forming the spacer layer in a broad sense will be described. The lower metal layer 15 is a remaining layer after being used as a supply source of the current path 162 material in a process described later, and may not necessarily remain as a final form.

  The spacer layer (CCP-NOL) 16 includes an insulating layer 161 and a current path 162. As described above, the spacer layer 16, the lower metal layer 15, and the upper metal layer 17 are handled as a spacer layer in a broad sense.

The insulating layer 161 is made of oxide, nitride, oxynitride, or the like. The insulating layer 161 can have both an amorphous structure such as Al ₂ O ₃ and a crystal structure such as MgO. In order to exhibit the function as the spacer layer, the thickness of the insulating layer 161 is preferably 1 to 3.5 nm, and more preferably 1.5 to 3 nm.

  As shown in FIG. 1, it is difficult to form a spacer layer having such a complicated structure having a three-dimensional structure on the nano order. The present inventors have found that the manufacturing method according to the present invention is effective for realizing a CCP structure close to an ideal form. Since this manufacturing method is the most important part of the present invention, details will be described later.

As a typical insulating material used for the insulating layer 161, there are a material using Al ₂ O ₃ as a base material and a material obtained by adding an additive element thereto. Examples of additive elements include Ti, Hf, Mg, Zr, V, Mo, Si, Cr, Nb, Ta, W, B, C, and V. The addition amount of these additive elements can be appropriately changed within a range of about 0% to 50%. As an example, about 2 nm of Al ₂ O ₃ can be used as the insulating layer 161.

For the insulating layer 161, instead of Al oxide such as Al ₂ O ₃ , Ti oxide, Hf oxide, Mg oxide, Zr oxide, Cr oxide, Ta oxide, Nb oxide, Mo oxidation Materials, Si oxides, V oxides, and the like can also be used. Even in the case of these oxides, the above-described materials can be used as additive elements. Further, the amount of the additive element can be appropriately changed within a range of about 0% to 50%.

  Instead of these oxides, oxynitrides and nitrides based on Al, Si, Hf, Ti, Mg, Zr, V, Mo, Nb, Ta, W, B, and C as described above are used. However, any material having a function of insulating current may be used.

  The current path 162 is a path (path) through which current flows perpendicularly to the film surface of the spacer layer 16 and is used to confine the current. It functions as a conductor that allows current to pass in the direction perpendicular to the film surface of the insulating layer 161, and can be composed of, for example, a metal layer such as Cu. That is, the spacer layer 16 has a current confinement structure (CCP structure), and the MR change rate can be increased by the current confinement effect. Examples of the material forming the current path 162 (CCP) include Au, Ag, Al, Ni, Co, Fe, or an alloy layer including at least one of these elements, in addition to Cu. As an example, the current path 162 can be formed of an alloy layer containing Cu. An alloy layer such as CuNi, CuCo, or CuFe can also be used. Here, a composition having 50% or more of Cu is preferable in order to reduce the high MR ratio and the interlayer coupling field (Hin) between the pinned layer 14 and the free layer 18.

  The current path 162 is a region where the oxygen and nitrogen contents are significantly lower than that of the insulating layer 161 (there is at least a two-fold difference in oxygen and nitrogen content), and is generally a crystalline phase. Since the crystalline phase has a smaller resistance than the amorphous phase, it easily functions as the current path 162.

The upper metal layer 17 forms part of a broad spacer layer. It has a function as a barrier layer that protects the free layer 18 formed thereon from being oxidized in contact with the oxide of the spacer layer 16 and a function of improving the crystallinity of the free layer 18. For example, when the material of the insulating layer 161 is amorphous (for example, Al ₂ O ₃ ), the crystallinity of the metal layer formed thereon is deteriorated, but the layer that improves the fcc crystallinity (for example, By arranging the (Cu layer) (the film thickness may be about 1 nm or less), the crystallinity of the free layer 18 can be remarkably improved.

  Depending on the material of the spacer layer 16 and the material of the free layer 18, the upper metal layer 17 is not necessarily provided. By optimizing the annealing conditions, selection of the insulating layer 161 material of the spacer layer 16, the material of the free layer 18, etc., the crystallinity can be avoided and the metal layer 17 on the spacer layer 16 can be eliminated.

  However, considering the manufacturing margin, it is preferable to form the upper metal layer 17 on the spacer layer 16. As a preferred example, Cu [0.5 nm] can be used as the upper metal layer 17.

  In addition to Cu, Au, Ag, Ru, or the like can be used as a constituent material of the upper metal layer 17. The material of the upper metal layer 17 is preferably the same as the material of the current path 162 of the spacer layer 16. When the material of the upper metal layer 17 is different from the material of the current path 162, the interface resistance is increased. However, if both are the same material, the interface resistance is not increased.

  The thickness of the upper metal layer 17 is preferably 0 to 1 nm, and more preferably 0.1 to 0.5 nm. If the upper metal layer 17 is too thick, the current confined by the spacer layer 16 spreads in the upper metal layer 17 and the current confinement effect becomes insufficient, leading to a reduction in MR ratio.

The free layer 18 is a layer having a ferromagnetic material whose magnetization direction is changed by an external magnetic field. For example, a two-layer configuration of Co ₉₀ Fe ₁₀ [1 nm] / Ni ₈₃ Fe ₁₇ [3.5 nm] using NiFe with CoFe inserted at the interface is an example of the free layer 18. In this case, it is preferable to provide a CoFe alloy at the interface with the spacer layer 16 rather than a NiFe alloy. In order to obtain a high MR ratio, selection of the magnetic material of the free layer 18 located at the interface of the spacer layer 16 is important. In the case where the NiFe layer is not used, a Co ₉₀ Fe ₁₀ [4 nm] single layer can be used. A free layer having a three-layer structure such as CoFe / NiFe / CoFe may be used.

Among the CoFe alloys, Co ₉₀ Fe ₁₀ is preferable because the soft magnetic characteristics are stable. When a CoFe alloy near Co ₉₀ Fe ₁₀ is used, the film thickness is preferably 0.5 to 4 nm. In addition, Co _x Fe _100-x (x = 70 to 90) is preferable.

  Further, as the free layer 18, a CoFe layer or Fe layer having a thickness of 1 to 2 nm and an ultrathin Cu layer having a thickness of about 0.1 to 0.8 nm may be alternately stacked.

When the spacer layer 16 is formed of a Cu layer, similarly to the pinned layer 14, even in the free layer 18, when the bcc FeCo layer is used as an interface material with the spacer layer 16, the MR ratio is increased. As an interface material with the spacer layer 16, a bcc FeCo alloy may be used instead of the fcc CoFe alloy. In this case, Fe _x Co _100-x (x = 30 to 100) in which a bcc layer is easily formed, or a material obtained by adding an additive element thereto can be used. Of these configurations, Co ₉₀ Fe ₁₀ [1 nm] / Ni ₈₃ Fe ₁₇ [3.5 nm] can be used as an example of a preferred embodiment.

  The cap layer 19 has a function of protecting the spin valve film. The cap layer 19 can have, for example, a plurality of metal layers, for example, a two-layer structure of a Cu layer and a Ru layer (Cu [1 nm] / Ru [10 nm]). As the cap layer 19, a Ru / Cu layer in which Ru is disposed on the free layer 18 side can also be used. In this case, the film thickness of Ru is preferably about 0.5 to 2 nm. The cap layer 19 having this configuration is particularly desirable when the free layer 18 is made of NiFe. This is because Ru has a non-solid relationship with Ni, so that magnetostriction of the interface mixing layer formed between the free layer 18 and the cap layer 19 can be reduced.

  When the cap layer 19 is Cu / Ru or Ru / Cu, the thickness of the Cu layer is preferably about 0.5 to 10 nm, and the thickness of the Ru layer can be about 0.5 to 5 nm. . Since Ru has a high specific resistance value, it is not preferable to use a very thick Ru layer.

  As the cap layer 19, another metal layer may be provided instead of the Cu layer or the Ru layer. The configuration of the cap layer 19 is not particularly limited, and other materials may be used as long as the cap can protect the spin valve film. However, care should be taken because the MR ratio and long-term reliability may change depending on the selection of the cap layer. Cu and Ru are examples of desirable cap layer materials from these viewpoints.

  The upper electrode 20 is an electrode for energizing in the direction perpendicular to the spin valve film. When a voltage is applied between the lower electrode 11 and the upper electrode 20, a current in the vertical direction of the film flows in the spin valve film. The upper electrode layer 20 is made of an electrically low resistance material (for example, Cu, Au, NiFe, etc.).

(Method for manufacturing magnetoresistive element)
Hereinafter, a method for manufacturing the magnetoresistive effect element according to the present embodiment will be described.

  FIG. 2 is a flowchart showing a manufacturing process for forming the spacer layers 15, 16, and 17 of the magnetoresistive effect element according to the embodiment of the present invention. With reference to FIGS. 1A to 1G, a method of manufacturing a magnetoresistive element (CCP-CPP element) according to the present invention will be schematically described.

  As shown in FIG. 2A, an underlayer and a pinning layer (these members are not shown) are formed on a substrate on which a lower electrode has been formed in advance, and a pinned layer 14 is formed on the pinning layer. Film. On the pinned layer 14, a first metal layer m1 (for example, Cu) that forms a current path is formed. A second metal layer m2 (for example, AlCu or Al) to be converted into an insulating layer is formed on the first metal layer m1.

  Next, as shown in FIGS. 2B and 2C, a surface oxidation treatment or a surface nitridation treatment for forming a part of the CCP structure having an insulating layer and a current path on the surface of the second metal layer m2. I do. Specifically, the surface oxidation treatment and the surface nitridation treatment can be performed through the steps described in detail below. However, these treatments are different from simple oxidation treatment and the like, as shown in FIGS. ) To form an insulating layer of the second metal layer m2 and to form a CPP structure for forming a metal layer (current path) by applying transfer energy to the first metal layer m1. Structure conversion processing (first conversion processing).

  The CCP structure formed in FIG. 2C is formed with a relatively thin film thickness for both m1 and m2 so that a CCP structure close to the ideal form can be easily realized. When this film thickness is thick, it becomes difficult to realize a CCP structure having an ideal shape as shown in FIG. However, if the film thickness remains as it is, the function of the NOL as an insulating layer becomes insufficient, and a leak current flows or the electrostatic breakdown voltage becomes a low value. Therefore, the formation process after FIG. 2D is necessary.

Next, as shown in FIG. 2D, as in FIG. 2A, a fourth metal layer m4 (for example, Cu) that forms a current path is formed. A third metal layer m3 (for example, AlCu or Al) to be converted into an insulating layer is formed on the fourth metal layer m4.

  Next, as shown in FIG. 2E, as in FIG. 2B, surface oxidation for forming a part of the CCP structure having an insulating layer and a current path on the surface of the third metal layer m3. Treatment or surface nitriding treatment is performed. As a result, a structure as shown in FIG. Specifically, the surface oxidation treatment and the surface nitridation treatment can be performed through the steps described in detail below, but these treatments are different from simple oxidation treatments and the like, as shown in FIGS. ) To form an insulating layer of the third metal layer m3 and to form a CPP structure for forming a metal layer (current path) by imparting transfer energy to the fourth metal layer m4. Structure conversion processing (second conversion processing).

  In order to easily realize a CCP structure close to the ideal form, both m3 and m4 are formed with a relatively thin film thickness. When this film thickness is large, it becomes difficult to realize a CCP structure having an ideal shape as shown in FIG. In such a structure, the insulation function of the NOL is not sufficient with only one layer, but since one layer of the CCP-NOL structure has already been prepared in FIG. 2C, the CCP-NOL formed in FIG. A thin film thickness is sufficient. In this way, by dividing the CCP-NOL formation process into two stages, a CCP-NOL structure close to the ideal form as shown in FIG. 2F can be realized.

  Next, as shown in FIG. 2G, a metal layer 17 such as Cu is formed on the spacer layer 16 as necessary, and a free layer 18 is formed thereon. Note that the metal layer 17 on the spacer layer 16 has a function of preventing the free layer formed thereon from being affected by oxidation, but is not necessarily provided.

  By using the method according to the embodiment of the present invention, the purity of the current path 21 that penetrates the insulating layer 22 of the spacer layer 16 can be improved regardless of the position of the upper and lower layers, and the vertical symmetry of the CCP shape can be improved. A good magnetoresistive element having a high MR ratio and good reliability can be manufactured.

  In the above example, the fourth metal layer m4 for forming the metal layer (current path) in the CCP-NOL structure is provided, but this layer is not an essential element in the present invention. Even when the metal layer m4 is not present, by performing the second conversion treatment, the atoms constituting the first metal layer m1 obtain further transfer energy and move upward to form an insulating layer. A current path is formed in the metal layer m3.

  However, by forming the fourth metal layer m4, a current path can be easily formed in a good state in the insulating layer made of the third metal layer m3.

  Next, the above process will be described in more detail.

  In FIG. 2A, m1 is a material for forming a current path, and m2 is a metal layer that is converted into an insulating layer by oxidation, nitridation, or oxynitridation treatment. As m1, a metal layer of Cu, Au, Ag, Al or the like is preferable. m2 is formed of a material containing at least one element of Al, Si, Mg, Ti, Hf, Zr, Cr, Mo, Nb, and W having a good insulating function when oxidized and nitrided. Is preferred. These single metals or alloy materials may be used. The film thickness of m1 is preferably about 0.1 to 1.5 nm, and the film thickness of m2 is preferably about 0.3 to 1 nm.

  In FIG. 2B, the surface oxidation treatment or the surface nitridation treatment performed after the first metal layer m1 and the second metal layer m2 are formed on the pinned layer 14 is performed in the first metal layer m2. This is a step performed to allow a part of the metal layer m1 to penetrate and to convert the metal layer m2 into the insulating layer 22. As described above, in order to produce a current path in nano order together with the conversion process from the metal to the insulating layer, the surface oxidation treatment or the surface nitridation treatment can be performed as follows, for example.

  First, in order to allow m2 to enter m1, it is necessary to provide atom transfer energy. Therefore, when performing oxidation treatment or nitriding treatment, use energy by irradiating ions or plasma state gas, not natural oxidation process or natural nitridation process that simply flows oxygen gas or nitrogen gas into the chamber. Is preferred. Further, in order to make m1 have good insulation resistance as oxidation, nitridation, or oxynitride, the oxidation treatment and nitridation treatment are preferably formed in a state where energy is applied.

  From such a viewpoint, the oxidation treatment and nitridation treatment are performed by ionizing or plasmaizing a gas such as argon, xenon, helium, neon, or krypton, and supplying oxygen gas, nitrogen gas, or the like into the ionized atmosphere or plasma atmosphere. However, it is preferable to carry out the process under the assistance of ions or plasma in the atmosphere (first method).

  In order to effectively perform the above-described assist in the oxidation treatment or nitridation treatment, it is preferable to perform a plurality of steps as shown below.

(I) After the oxidation treatment or nitridation treatment in the first method, the film surface is irradiated with the rare gas ions or plasma (second method).

  In this method, after the oxidation treatment and nitridation treatment in the first method described above are performed, the film surface is irradiated with ions containing at least one element of argon, xenon, helium, neon, and krypton, or plasma. To do. Further, instead of these rare gases, oxygen or nitrogen gas ions or plasma can be irradiated.

  According to this method, the oxidation treatment and the nitridation treatment are assisted afterwards by ion irradiation or plasma irradiation. For example, further movement energy is added to the first metal layer m1, etc. A characteristic current path can be easily formed.

(II) Before the oxidation treatment or nitridation treatment in the first method, the film surface is irradiated with the above-described rare gas ions or plasma (third method).

  In this case, the film surface is irradiated with rare gas ions or plasma containing at least one element of argon, xenon, helium, neon, and krypton. Also in this case, for example, the transfer energy can be added in advance to the first metal layer m1 or the like, and a current path having a uniform size and characteristics can be easily formed through a subsequent oxidation process or the like. Can do.

(III) Before or after the oxidation treatment or nitridation treatment in the first method described above, the film surface is irradiated with the aforementioned rare gas ions or plasma (fourth method).

  This is a combination of the above methods (I) and (II). Therefore, the ion or plasma irradiation after the oxidation treatment or nitridation treatment in the first method can be performed in the same manner as in the case of the above (I), and therefore before the oxidation treatment or nitridation treatment in the first method. The ion or plasma irradiation can be performed in the same manner as in the above (II).

  Next, in the process as described above, the details of the ion beam or plasma irradiation conditions of the process performed after the oxidation process of (I) or the process performed before the oxidation process of (II) or the like will be described (second) Method to fourth method).

  When the above rare gas is ionized or converted into plasma, it is preferable to set the acceleration voltage V to +30 to 130 V and the beam current Ib to 20 to 200 mA. These conditions are significantly weaker than the conditions for ion beam etching. Although an ion beam is used here, it can be created in the same manner by using plasma such as RF plasma instead of the ion beam.

  The incident angle of the ion beam is defined as 0 degree when incident perpendicular to the film surface and 90 degrees when incident parallel to the film surface, and is appropriately changed within a range of 0 to 80 degrees. The treatment time in this step is preferably about 15 seconds to 180 seconds, and more preferably 30 seconds or more from the viewpoint of controllability. If the treatment time is too long, the productivity of the CCP-CPP element is inferior, which is not preferable. From these viewpoints, the treatment time is most preferably about 30 seconds to 180 seconds.

  By irradiating the ion beam having the energy as described above, the constituent atoms of the first metal layer m1 are sucked up and penetrated into the second metal layer m2 with high movement energy, and a current path is generated. Will be.

  As described above, AlCu or Al can be used as the second metal layer m2. When Al not containing Cu is used as the second metal layer m2, only Cu suction from the first metal layer m1 occurs. For the second metal layer m2, a metal material other than Al can be used. For example, as the second metal layer m2, Si, Hf, Zr, Ti, Mg, Cr, Mo, Nb, W, or the like that is converted into a stable oxide may be used.

  In addition, detailed conditions in the case where ions or plasma is used for the oxidation treatment will be described (one step in the first method and the second method to the fourth method). The ion or plasma irradiation conditions at this time are preferably set such that the acceleration voltage V is +40 to 200 V and the beam current Ib is about 30 to 300 mA. The oxidation treatment time is preferably about 15 seconds to 300 seconds, and more preferably about 20 seconds to 180 seconds. When a strong ion beam is used, the processing time is shortened, and when a weak ion beam is used, the processing time is lengthened.

The preferable range of the oxygen exposure amount during oxidation is 1000 to 5000 L (1 L = 1 × 10 ⁻⁶ Torr × sec) in the case of oxidation treatment using ions or plasma, and 3000 to 30000 L in the case of natural oxidation. is there.

  By using the appropriate conditions as described above in the respective steps shown in FIG. 2, it is possible to realize a CCP structure close to the ideal shape.

  In FIG. 2D, m3 and m4 may be the same material as m1 and m2, or may be different. In general, it is a preferred embodiment to use the same material. Specifically, m4 is formed of a metal layer containing at least one element of Cu, Au, Ag, and Al, and m3 is Al, Si having a good insulating function when oxidized or nitrided. , Mg, Ti, Hf, Zr, Cr, Mo, Nb, and W are preferably formed from a material containing at least one element. These single metals or alloy materials may be used.

  The film thickness of m3 is preferably about 0.1 to 1.5 nm, and the film thickness of m4 is preferably about 0.3 to 1 nm.

  Note that, in the above, detailed conditions for the case of performing the oxidation treatment in particular are described, but similar conditions can also be used for the nitriding treatment.

  In FIG. 2G, the upper metal layer 17 and the free layer 18 are formed, but the upper metal layer 17 may be the same as the material constituting the CCP or may be a different material. As desirable forms, Cu, Au, Ag, Al and the like are desirable embodiments. The thickness of the upper metal layer 17 is preferably 0 to 1 nm.

  FIG. 3 shows a schematic configuration of an apparatus used for manufacturing a magnetoresistive effect element (CCP-CPP element) according to an embodiment of the present invention. As shown in FIG. 3, a load lock chamber 51, a pre-cleaning chamber 52, a first metal film forming chamber (MC1) 53, and a second metal film forming chamber (MC2) 54 are centered on a transfer chamber (TC) 50. Each of the oxidation chambers (OC) 60 is provided via a vacuum valve. In this apparatus, since the substrate can be transported in a vacuum between the chambers connected via the vacuum valve, the surface of the substrate is kept clean.

  The metal film forming chambers 53 and 54 have multi-targets (5 to 10 yuan). Examples of the film forming method include sputtering methods such as DC magnetron sputtering and RF magnetron sputtering, ion beam sputtering methods, vapor deposition methods, CVD (Chemical Vapor Deposition) methods, and MBE (Molecular Beam Epitaxy) methods.

  For the surface oxidation treatment, a chamber having an ion beam mechanism, an RF plasma mechanism, or a heating mechanism can be used, and it is necessary to separate it from the metal film forming chamber.

A typical value of the vacuum degree of the vacuum chamber is on the order of 10 ⁻⁹ Torr, and the first half value of 10 ⁻⁸ Torr is acceptable.

  The film formation of m 1, m 2, m 3 and m 4, which is characteristic in the present invention, is performed in one of the metal film formation chambers 53 and 54, and the surface oxidation treatment is performed in the oxidation chamber 60. After the deposition of m1 and m2, the wafer is transferred to the oxidation chamber 60 through the transfer chamber 50 and an oxidation process is performed. Thereafter, the m3 and m4 films are formed by being transferred to one of the metal film forming chambers 53 and 54, and then the wafer is transferred again to the oxidation processing chamber 60 through the transfer chamber 50, and the oxidation process is performed. Thereafter, the upper metal layer 17 and the free layer 18 are formed by being transferred to one of the metal film formation chambers 53 and 54.

  FIG. 4 shows an example of the configuration of the oxidation chamber 60 shown in FIG. Here, the oxidation chamber 60 is a chamber having an ion beam. As shown in the figure, the oxidation chamber 60 is evacuated by a vacuum pump 61, and oxygen gas whose flow rate is controlled by a mass flow controller (MFC) 63 is introduced into the oxidation chamber 60 from an oxygen supply pipe 62. An ion source 70 is provided in the oxidation chamber 60. Examples of the ion source include ICP (Inductive coupled plasma) type, Capacitive coupled plasma type, ECR (Electron-cyclotron resonance) type, and Kaufmann type. The substrate holder 80 and the substrate 1 are arranged so as to face the ion source 70.

  Three grids 71, 72, and 73 that adjust ion acceleration are provided at the ion emission port from the ion source 70. A neutralizer 74 for neutralizing ions is provided outside the ion source 70. The substrate holder 80 is supported to be tiltable. Although the incident angle of ions on the substrate 1 can be varied in a wide range, a typical incident angle value is 15 ° to 60 °.

  In this oxidation chamber 60, by irradiating the substrate 1 with an ion beam of Ar or the like, energy assist associated with the surface oxidation treatment using ions can be performed, and Ar or the like while supplying oxygen gas from an oxygen supply pipe 62. By irradiating the substrate 1 with this ion beam, conversion from the metal layer to the oxide layer can be performed.

  Here, the chamber has an ion beam as an oxidation chamber, but an RF plasma chamber or the like has exactly the same effect. In any case, as described above, in the surface oxidation process, it is necessary to perform a process for applying energy. Therefore, it is necessary to perform the surface oxidation process in a chamber capable of generating ions or plasma.

  Further, heat treatment can be performed as a means for applying energy. In this case, the treatment is performed for several tens of seconds to several minutes at a temperature of 100 ° C. to 300 ° C. This treatment can also be performed as part of the surface oxidation treatment.

(Overall description of manufacturing method of magnetoresistive effect element)
Hereinafter, the whole manufacturing method of a magnetoresistive effect element is demonstrated in detail using FIG.

  First, on a substrate (not shown), a lower electrode 11, an underlayer 12, a pinning layer 13, a pin layer 14, a lower metal layer 15, a spacer layer 16, an upper metal layer 17, a free layer 18, a cap layer 19, The upper electrode 20 is formed in order.

Next, using the apparatus configuration as shown in FIG. 3, the substrate is set in the load lock chamber 51, the metal film formation is performed in the metal film formation chamber 53 or 54, and the oxidation is performed in the oxide layer / nitride layer formation chamber 60. Do each. The ultimate vacuum of the metal film forming chamber is preferably 1 × 10 ⁻⁸ Torr or less, and generally about 5 × 10 ⁻¹⁰ Torr to 5 × 10 ⁻⁹ Torr. The ultimate vacuum of the transfer chamber 50 is on the order of 10 ⁻⁹ Torr. The ultimate vacuum of the oxide layer / nitride layer forming chamber 60 is 8 × 10 ⁻⁸ Torr or less.

(1) Formation of the underlayer 12 The lower electrode 11 is formed in advance on a substrate (not shown) by a microfabrication process. On the lower electrode 11, for example, Ta [5 nm] / Ru [2 nm] is formed as the base layer 12. As described above, Ta is the buffer layer 12a for reducing the roughness of the lower electrode. Ru is a seed layer 12b for controlling the crystal orientation and crystal grain size of the spin valve film formed thereon.

(2) Formation of Pinning Layer 13 The pinning layer 13 is formed on the underlayer 12. As a material of the pinning layer 13, an antiferromagnetic material such as PtMn, PdPtMn, IrMn, RuRhMn can be used.

(3) Formation of Pinned Layer 14 The pinned layer 14 is formed on the pinning layer 13. The pinned layer 14 can be, for example, a synthetic pinned layer including a lower pinned layer 141 (Co ₉₀ Fe ₁₀ ), a magnetic coupling layer 142 (Ru), and an upper pinned layer 143 (Co ₉₀ Fe ₁₀ [4 nm]). .

(4) Formation of Spacer Layer 16 Next, a spacer layer (CCP-NOL) 16 having a current confinement structure (CCP structure) is formed. Since this is the most characteristic process in the present invention, it will be described in detail below.

  First, in order to form the spacer layer 16, the layers up to the pins 14 are formed in a metal film forming chamber, then transferred to the oxide layer / nitride layer forming chamber 60, and processed at 60.

  Here, a desirable shape of the spacer layer 16 will be described. It is desirable to realize a relatively thick spacer layer 16 as the spacer layer 16. The reason for this is that the increase in the merit and the thickness of the insulating layer can completely prevent the leakage current from flowing in a region that should otherwise function as an insulating layer other than the current path layer, and the CCP effect. It is possible to achieve stronger reliability and realize stronger reliability. Further, since the insulation breakdown voltage of the insulating layer is improved by increasing the thickness of the insulating layer, it is possible to improve resistance to electrostatic breakdown such as ESD (Electric Static Discharge). A high ESD withstand voltage leads to an improvement in yield when the head is assembled to the HDD.

  Such a significant improvement in reliability performance can improve not only the HDD production but also the destruction and thermal resistance in every situation, so servers and high reliability are required. It will be suitable as a head for enterprise use. High-reliability heads that do not simply improve recording density are becoming increasingly important in the recent situation where HDD applications are expanding. Increasing the life of the head is very important in expanding the usage environment of the HDD. Such a highly reliable head is very effective as an HDD for car navigation where the thermal environment is severe.

  Of course, such a highly reliable head is used not only as a high value-added HDD as described above, but also as an HDD for ordinary personal computers, portable music players, mobile phones and the like used for consumer use. it can.

  As described above, it is desirable to form a relatively thick spacer layer, but the realization process is not easy. As a conventional method for forming the spacer layer 16, there is, for example, Japanese Patent Application No. 2004-233641. However, it is difficult to form a thick spacer layer in the conventional process. Considering the process of sucking up m1 into m2, it is easily imagined that energy treatment from the film surface cannot be reached at a thick film thickness.

  On the other hand, if the energy of the ion beam or plasma is increased too much in order to give enough energy, this time, the etching will be etched away, and m1 will not be sucked up to m2, but it will be etched by m2 instead. In this case, both m2 and m1 are not etched. In this case, the energy assist effect is completely lost.

  From the above, it can be seen that it is difficult to form the thick spacer layer 16 by a conventional process.

  On the other hand, although it is relatively easy to form the thin spacer layer 16, it does not exhibit a sufficient CCP function. For example, a tunnel current also flows through a portion that should originally function as an insulating layer, which becomes a leak current. In addition, when the insulating layer is thin, the withstand voltage is weakened, so that it is weak against ESD, and various measures are required in the manufacturing process, resulting in a decrease in yield.

  As described above, although the conventional process exhibits a sufficient function in practical use, it has higher reliability in terms of applications that require higher reliability and in terms of further improving the manufacturing yield. A head is needed.

  In order to solve such a problem, the thickness of the insulating layer constituting the spacer layer is increased by using the manufacturing method of the present invention as described above, and other than the current path layer formed in the spacer layer. Thus, it is possible to completely suppress the leakage current from flowing through the region that should function as the insulating layer. As a result, the CCP effect can be exerted more strongly and stronger reliability can be realized. Further, since the insulation breakdown voltage of the insulating layer is improved by increasing the thickness of the insulating layer, it is possible to improve resistance to electrostatic breakdown such as ESD (Electric Static Discharge).

To form the spacer layer 16 according to the present invention, steps as shown in FIG. 2 are used. Here, the case where the spacer layer 16 including the current path 162 made of Cu having a metal crystal structure is formed in the insulating layer 161 made of Al ₂ O ₃ having an amorphous structure will be specifically described.

  First, after a metal layer m1 (for example, Cu) serving as a current path supply source is formed on the upper pinned layer 143, an oxidized metal layer that is converted as a part of the insulating layer 161 on the lower metal layer m1. m2 (for example, AlCu or Al) is deposited.

  Next, a conversion process is performed on the metal layer to be oxidized. This conversion process can be performed through the oxidation process or the nitridation process as described above. Further, such conversion processing can be performed through a plurality of steps. For example, as a first step of the conversion process, an ion beam of a rare gas (eg, Ar) is irradiated. This pretreatment is called PIT (Pre-ion treatment). As a result of this PIT, a part of the lower metal layer is sucked and invaded into the metal layer to be oxidized. PIT is effective as a means for performing energy processing.

  As another means for providing the energy, a means such as heating the film surface is also effective. At this time, there are means for heating at about 100 to 300 ° C. Further, as described above, a method of performing energy treatment using a rare gas such as Ar after conversion to an oxide film using oxygen gas may be used. This post-treatment is called AIT (After-ion treatment).

  In addition to the above-described PIT and the like, in this example, oxygen gas or nitrogen gas is supplied in the presence of the ion beam or the like, and oxidation or nitridation is performed with energy assistance. These processes are the main stream of the present invention, and are called, for example, IAO (Ion-beam assisted oxidation).

  Through the above-described treatment, Cu in the first metal layer m1 is sucked into the AlCu layer, and the second metal layer m2 is converted from the metal state to the insulating layer state.

  In PIT and AIT, Ar ions are irradiated under conditions of an acceleration voltage of 30 to 150 V, a beam current of 20 to 200 mA, and a processing time of 30 to 180 seconds. Among the acceleration voltages, a voltage range of 40 to 60 V is preferable. In the case of a voltage range higher than this, the MR change rate may decrease due to the influence of surface roughness after PIT or AIT. Further, the current value can be in the range of 30 to 80 mA, and the irradiation time can be in the range of 60 to 150 seconds.

  There is also a method of forming a metal layer before being converted as a part of the insulating layer 161 such as AlCu or Al by bias sputtering instead of the PIT or AIT processing. In this case, the energy of bias sputtering can be 30 to 200 V in the case of DC bias, and 30 to 200 W in the case of RF bias.

  In IAO, Ar ions are irradiated under conditions of an acceleration voltage of 40 to 200 V, a beam current of 30 to 200 mA, and a processing time of 15 to 300 seconds. Among the acceleration voltages, a voltage range of 50 to 100 V is preferable. If the acceleration voltage is higher than this, there is a possibility that the MR change rate is lowered due to the influence of surface roughness after PIT. In addition, a beam current of 40 to 100 mA and an irradiation time of 30 to 180 seconds can be adopted.

  The oxygen supply amount during oxidation with IAO is preferably in the range of 1000 to 3000 L. Oxidation not only to Al but also to the lower magnetic layer (pinned layer 14) during IAO is not preferable because the heat resistance and reliability of the CCP-CPP element deteriorate. In order to improve reliability, it is important that the magnetic layer (pinned layer 14) located below the spacer layer 16 is not oxidized and is in a metal state. In order to realize this, it is necessary to set the oxygen supply amount within the above range.

  In addition, in order to form a stable oxide with the supplied oxygen, it is desirable that oxygen gas is flowed only while the surface of the substrate is irradiated with an ion beam. That is, it is desirable not to flow oxygen gas when the substrate surface is not irradiated with an ion beam.

After the processing as described above, for example, a part of the spacer layer 16 having the insulating layer 161 made of Al ₂ O ₃ and the current path 162 made of Cu is formed. This is a process that utilizes the difference in oxidation energy that Al is easily oxidized and Cu is not easily oxidized.

  The processes up to the above are the same as in Japanese Patent Application No. 2004-233641, but in the present invention, in order to realize a CCP structure closer to the ideal form, both the thicknesses m1 and m2 are thin. This is because it is possible to realize a CCP structure in a good form more easily by reducing the thickness.

  The specific film thickness of m1 is preferably about 0.1 to 1.5 nm, and the film thickness of m2 is preferably about 0.3 to 1 nm.

  As a material of the first metal layer m1 (lower metal layer 15) forming the current path, Au, Ag, Al, or a metal containing at least one of these elements may be used instead of Cu. However, Cu is more preferable than Au and Ag because of its higher stability to heat treatment. As a material of the first metal layer m1, a magnetic material may be used instead of these nonmagnetic materials. Examples of the magnetic material include Co, Fe, Ni, and alloys thereof.

When Al ₉₀ Cu ₁₀ is used for the second metal layer m2, not only Cu in the first metal layer m1 is sucked up during the PIT process, but also Cu in the AlCu is separated from Al. That is, the current path 162 is formed from both the first and second metal layers m2. When ion beam assisted oxidation is performed after the PIT process, the oxidation proceeds while the separation of Al and Cu is promoted by the assist effect of the ion beam.

As the second metal layer m2, an Al single metal that does not contain Cu, which is a constituent material of the current path 162, may be used instead of Al ₉₀ Cu ₁₀ . In this case, Cu as a constituent material of the current path 162 is supplied only from the first metal layer m1 as a base. When AlCu is used as the second metal layer m2, Cu that is the material of the current path 162 is also supplied from the second metal layer m2 during the PIT process. For this reason, even when the thick insulating layer 161 is formed, the current path 162 can be formed relatively easily. When Al is used as the second metal layer m2, Cu is less likely to be mixed into Al ₂ O ₃ formed by oxidation, so that Al ₂ O ₃ with a high breakdown voltage can be easily formed. Since there are merits of Al and AlCu, they can be used properly according to the situation.

  The film thickness of the second metal layer m2 is preferably about 0.3 to 1 nm in the case of AlCu or Al. This film thickness range includes a film thickness range that can be said to be too thin when the spacer layer 16 is formed only by m1 and m2. For example, in the case of using 0.3 nm of Al, if the formation of the spacer layer 16 is completed only by this, the film thickness is insufficient. However, as will be described later, since the spacer layer 16 is formed again by another process and only a part of the spacer layer 16 is formed up to this step, the thin film thickness as described above forms CCP. This is a desirable range from the viewpoint.

The AlCu as the second metal layer m2 preferably has a composition represented by AlxCu100-x (x = 100 to 70%). Elements such as Ti, Hf, Zr, Nb, Mg, Mo, and Si may be added to AlCu. In this case, the composition of the additive element is preferably about 2 to 30%. When these elements are added, the formation of the CCP structure may be facilitated. Further, when these additive elements are distributed more richly in the boundary region between the Al ₂ O ₃ insulating layer 161 and the Cu current path 162 than in other regions, the adhesion between the insulating layer 161 and the current path 162 is improved. Electro-migration resistance may be improved.

In the CCP-CPP element, the density of the current flowing through the metal current path of the spacer layer 16 is a huge value of 10 ⁷ to 10 ¹⁰ A / cm ² . For this reason, it is important that the electromigration resistance is high and the stability of the Cu current path 162 during current application can be ensured. However, if an appropriate CCP structure is formed, sufficiently good electromigration resistance can be realized without adding an element to the second metal layer m2.

The material of the second metal layer m2 is not limited to an Al alloy for forming Al ₂ O _3, but may be an alloy containing Hf, Mg, Zr, Ti, Ta, Mo, W, Nb, Si, or the like as a main component. Good. The insulating layer 161 converted from the second metal layer m2 is not limited to an oxide, and may be a nitride or an oxynitride.

  Whatever material is used for the second metal layer m2, the film thickness at the time of film formation is preferably 0.5 to 2 nm, and the film thickness when converted to oxide, nitride or oxynitride Is preferably about 0.8 to 3.5 nm.

The insulating layer 161 may be not only an oxide containing a single element but also an oxide, nitride, or oxynitride of an alloy material. For example, using Al ₂ O ₃ as a base material, oxidation of any one element such as Ti, Mg, Zr, Ta, Mo, W, Nb, Si, or a material containing 0 to 50% of a plurality of elements in Al Things can also be used.

  The third metal layer m3 is preferably 0.3 to 1.0 nm, and the fourth metal layer m4 is preferably 0.1 to 1.5 nm. The same conversion process as described above is performed on the third metal layer m3 and the fourth metal layer m4. The third metal layer m3 can be composed of AlCu or Al, and the fourth metal layer m4 can be composed of Cu or the like.

  As described above, in this example, the conversion process is performed on the first metal layer m1 and the second metal layer m2, and the third metal layer m3 and the fourth metal layer m4, respectively. The spacer layer is made up of. Therefore, the spacer layer can be configured to exhibit a relatively thick CCP structure.

  In the above specific example, the conversion process is performed in two stages, but naturally, the conversion process of three or more stages may be performed to form a CCP type spacer layer. However, in the CPP structure type magnetoresistive effect element that is currently required, a CPP structure type magnetoresistive effect element having the desired characteristics can be obtained only through the two-step conversion process described above.

(5) Formation of Upper Metal Layer 17 and Free Layer 18 Next, for example, Cu [0.25 nm] is deposited on the spacer layer 16 as the upper metal layer 17. A preferable film thickness range is about 0.2 to 1.0 nm. When about 0.25 nm is used, there is an advantage that the crystallinity of the free layer 18 is easily improved. However, this upper metal layer may not necessarily be formed in some cases.

On the upper metal layer 17, a free layer 18, for example, Co ₉₀ Fe ₁₀ [1 nm] / Ni ₈₃ Fe ₁₇ [3.5 nm] is formed. In order to obtain a high MR ratio, selection of the magnetic material of the free layer 18 located at the interface with the spacer layer 16 is important. In this case, it is preferable to provide a CoFe alloy at the interface with the spacer layer 16 rather than a NiFe alloy. Among CoFe alloys, Co ₉₀ Fe ₁₀ [1 nm] having particularly stable soft magnetic properties can be used. CoFe alloys can also be used with other compositions.

When a CoFe alloy near Co ₉₀ Fe ₁₀ is used, the film thickness is preferably 0.5 to 4 nm. When using a CoFe alloy having another composition (for example, Co ₅₀ Fe ₅₀ ), the film thickness is preferably 0.5 to 2 nm. When, for example, Fe ₅₀ Co ₅₀ (or FexCo100-x (x = 45 to 85)) is used for the free layer 18 in order to increase the spin-dependent interface scattering effect, soft magnetism as the free layer 18 is used. Therefore, it is difficult to use a thick film such as the pinned layer 14. For this reason, 0.5-1 nm is a preferable film thickness range. When Fe containing no Co is used, since the soft magnetic characteristics are relatively good, the film thickness can be set to about 0.5 to 4 nm.

  The NiFe layer provided on the CoFe layer is made of a material having a stable soft magnetic property. The soft magnetic property of the CoFe alloy is not so stable, but the soft magnetic property can be supplemented by providing the NiFe alloy thereon. The use of NiFe as the free layer 18 makes it possible to use a material that can realize a high MR ratio at the interface with the spacer layer 16, and is preferable in terms of the total characteristics of the spin valve film.

The composition of the NiFe alloy is preferably NixFe100-x (x = about 78 to 85%). Here, it is preferable to use a Ni-rich composition (for example, Ni ₈₃ Fe ₁₇ ) rather than the commonly used NiFe composition Ni ₈₁ Fe ₁₉ . This is to realize zero magnetostriction. In the NiFe film formed on the spacer layer 16 having the CCP structure, the magnetostriction shifts to the positive side as compared with the NiFe film formed on the metal Cu spacer layer. In order to cancel the shift of magnetostriction to the plus side, a negative NiFe composition having a larger Ni composition than usual is used.

  The total film thickness of the NiFe layer is preferably about 2 to 5 nm (for example, 3.5 nm). When the NiFe layer is not used, a free layer 18 in which a plurality of CoFe layers or Fe layers of 1 to 2 nm and ultrathin Cu layers of about 0.1 to 0.8 nm are alternately stacked may be used.

(6) Formation of Cap Layer 19 and Upper Electrode 20 For example, Cu [1 nm] / Ru [10 nm] is laminated on the free layer 18 as the cap layer 19. An upper electrode 20 is formed on the cap layer 19 to vertically apply current to the spin valve film.

Example 1
Examples of the present invention will be described below. Below, the structure of the magnetoresistive effect film | membrane 10 based on the Example of this invention is represented.

・ Lower electrode 11
・ Underlayer 12: Ta [3 nm] / Ru [2 nm]
Pinning layer 13: Ir ₂₂ Mn ₇₈ [7 nm]
Pinned layer 14: Co ₉₀ Fe ₁₀ [3.6 nm] / Ru [0.9 nm] / (Fe ₅₀ Co
₅₀ [1 nm] / Cu [0.25 nm]) × 2 / Fe ₅₀ Co ₅₀ [1 nm]
Metal layer 15: Cu [0.1 nm]
Spacer layer (CCP-NOL) 16: Al ₂ O ₃ insulating layer 161 and Cu current path 162
Metal layer 17: Cu [0.25 nm]
Free layer 18: Co ₉₀ Fe ₁₀ [1 nm] / Ni ₈₃ Fe ₁₇ [3.5 nm]
Cap layer 19: Cu [1 nm] / Ru [10 nm]
-Upper electrode 20.

CCP-NOL Creation Method Only the CCP-NOL creation method characteristic of the present invention will be described more specifically below. The film formation up to the pinned layer 14 has been completed, and the substrate will be described from the state where it is transferred to the oxidation chamber 60.

  First, as a step in FIG. 2A, Cu was formed to 0.3 nm as m1 and AlCu was formed to 0.6 nm as m2.

  As the step in FIG. 2B, the first stage conversion process is performed. The specific conversion process was performed in a plurality of steps as follows. First, oxygen gas is flowed into the oxidation chamber in a state in which the sample surface is irradiated with an Ar ion beam. The energy of the Ar ion beam was 60 V (IAO). Thereafter, the oxygen gas flow was stopped and only Ar ion beam irradiation was performed. The conditions of the Ar ion beam at this time are exactly the same as those used for the oxidation. Ion beam irradiation was performed for 60 seconds (AIT). As a result, a form as shown in FIG. 2C is formed.

  Next, as a step in FIG. 2D, Cu was formed to 0.3 nm as m3 and AlCu was formed to 0.6 nm as m4.

  Next, a second-stage conversion process was performed as the process of FIG. The specific conversion process was performed in a plurality of steps as follows. First, oxygen gas is flowed into the oxidation chamber in a state in which the sample surface is irradiated with an Ar ion beam. The energy of the Ar ion beam was 60 V (IAO). Thereafter, the oxygen gas flow was stopped and only Ar ion beam irradiation was performed. The conditions of the Ar ion beam at this time are exactly the same as those used for the oxidation. Ion beam irradiation was performed for 60 seconds (AIT). As a result, a form as shown in FIG. 2F can be formed.

  In the process of FIG. 2G, the upper metal layer 17 is formed to 0.25 nm, and the spacer layers 15, 16, and 17 are completed.

  Here, the configuration shown in the figure is the final form by the heat treatment performed after all the CCP-CPP elements are formed. Therefore, the final configuration shown in the figure is in the middle of the film formation of the CCP-CPP elements. There may be no. Actually, the heat treatment performed after film formation up to the CCP-CPP film cap layer also has an energy assist effect, so that the final shape is obtained after the heat treatment is completed. The heat treatment is performed at 290 degrees for 4 hours.

  When the formation of the spacer layer is completed, the spacer layer is unloaded from the oxidation chamber 60 and transferred to the metal film forming chamber. A free layer is formed in a metal film forming chamber.

(Evaluation of Examples)
Examples were evaluated with comparative examples.

  As a comparative example, a case where the conventional process as shown in FIG. In the comparative example, the film thickness of m1 is the sum of the film thickness of m1 and m3 of the example, and 0.6 nm of Cu is formed, and the film thickness of m2 of the comparative example is m2 and m4 of the example. An AlCu film having a thickness of 1.2 nm was formed.

  As the conversion process, the same process as the conversion process performed in the example was performed only once. That is, the spacer layer was produced in the IAO / AIT process.

  Here, the current flowing direction from the pinned layer 14 to the free layer 18 was set as the energizing direction for which the characteristics were evaluated. In other words, the flow of electrons is reversed, and therefore flows from the free layer 18 to the pinned layer 14. Such an energizing direction is a desirable direction in order to reduce spin transfer noise. When a current is passed from the free layer 18 to the pinned layer 14 (in terms of electron flow, the pinned layer to the free layer) is said to have a larger spin transfer torque effect, and noise is generated in the head. From this point of view, the energization direction is preferably a direction in which the pin layer flows to the free layer.

When the characteristics of the CCP-CPP element according to the example were evaluated, RA = 500 mΩ / μm ² and MR change rate = 9%. On the other hand, when the characteristics of the CCP-CPP element according to the comparative example were evaluated, RA = 900 mΩ / μm ² and MR change rate = 7%. In both cases, although the NOL film thickness after the formation was almost equal, the RA and MR change rates were found to be significantly different between the example and the comparative example.

  Here, in order to investigate the cause of the difference as described above, a comparative study between the example and the comparative example was performed using a three-dimensional atom probe. A three-dimensional atom probe is a three-dimensional atom probe that processes a sample into a needle shape, sets the sample in a vacuum chamber, and applies a high voltage to the sample to generate one atom at the tip and one field evaporation. It is a destructive testing method that analyzes nanostructures in atomic order.

  As a result, in the case of the comparative example, the current path opening area is small in the upper part of the current path 21, and the oxygen concentration is larger than that in the lower part of the current path 21. Specifically, some current paths cause an area difference of 20% or more between the upper and lower sides. Here, as the definition of the lower part and the upper part of the current path, the lower part indicates the film thickness direction in the current path film thickness direction, where 0 is the substrate side and 100 is the film surface side. Called the lower part and more than 50 places called the upper part of the current path.

  Also, a difference is seen in the Cu purity of the current path 21 between the upper part and the lower part of the current path 21. Specifically, the oxygen concentration contained in Cu may be higher by 10 atomic% or more in the upper part of the current path than in the lower part of the current path. When the Cu purity of the current path 21 is deteriorated, the CCP effect is reduced and the performance is deteriorated. In addition, the vertical asymmetry of such a shape is not preferable because it causes a difference in reliability depending on the energization direction.

  On the other hand, in the case of the embodiment, there is no difference of 20% in the current path opening area in the upper part and the lower part of the current path 21, and they are uniformly formed. Also, there is no significant difference of 10 atomic% in the oxygen concentration in the current path between the upper part and the lower part.

  As a cause of such a difference, in the case of the comparative example, in the case of surface oxidation of 1.2 nm of AlCu as the film thickness of m2, the uptake of Cu of m1 is not perfect, This is thought to be due to an incomplete state. Regarding the oxygen concentration, when the film thickness of m2 is too thick, the oxygen concentration in the upper part is higher than that in the lower part when surface oxidation is performed. The diameter of the current path 162 that penetrates the insulating layer 161 is about 1 to 10 nm, and is about 2 to 6 nm. The current path 162 having a diameter larger than 10 nm is not preferable because it causes variation in characteristics of each element when the element size is reduced, and it is preferable that the current path 162 having a diameter larger than 6 nm does not exist.

(Example 2)
In the first embodiment, the magnetoresistive effect element having the CPP structure having the bottom type spin valve film has been described. In this example, the magnetoresistive effect element having the CPP structure having the top type spin valve film will be described. In the top type spin valve film, the pinned layer 14 is disposed above the free layer 18. That is, the SCT can be applied not only to the bottom type CCP-CPP element in which the pinned layer 14 is located below the free layer 18 but also to the top type CCP-CPP element. Even in this case, the process for forming the spacer layer 16 characteristic of the present invention is exactly the same. In FIG. 1, the free layer becomes the base of the spacer layer instead of the upper pinned layer, and instead of the free layer 18 in FIG. 1, a layer corresponding to the pinned layer 14 is disposed above the spacer layer.

  Specifically, the following configuration is an example.

・ Lower electrode 11
・ Underlayer 12: Ta [3 nm] / Ru [2 nm]
Free layer 18: Ni ₈₃ Fe ₁₇ [3.5 nm] / Co ₉₀ Fe ₁₀ [1 nm]
Metal layer 15: Cu [0.5 nm]
Spacer layer (CCP-NOL) 16: Al ₂ O ₃ insulating layer 161 and Cu current path 162
Metal layer 17: Cu [0.25 nm]
Pinned layer 14: (Fe ₅₀ Co ₅₀ [1 nm] / Cu [0.25 nm]) × 2 / Fe ₅₀ Co ₅₀ [1 nm] / Ru [ _0.9 nm] / Co ₉₀ Fe ₁₀ [3.6 nm]
Pinning layer 13: Ir ₂₂ Mn ₇₈ [7 nm]
Cap layer 19: Cu [1 nm] / Ru [10 nm]
-Upper electrode 20.

  As described above, when a top-type CCP-CPP element is manufactured, the layers between the base layer 12 and the cap layer 19 are formed in the reverse order of FIG. However, the order of the lower metal layer 15 and the upper metal layer 17 is not reversed due to the formation of the spacer layer 16 and the like. That is, the roles of m1, m2, m3, and m4 are exactly the same for the bottom type and the top type.

(Application of magnetoresistive effect element)
Hereinafter, application of the magnetoresistive effect element (CCP-CPP element) according to the embodiment of the present invention will be described.

In an embodiment of the present invention, the element resistance RA of the CPP element, from the viewpoint of high density corresponding, preferably 500mΩ / μm ² or less, 300mΩ / μm ² or less is more preferable. In calculating the element resistance RA, the resistance R of the CPP element is multiplied by the effective area A of the energized portion of the spin valve film. Here, the element resistance R can be directly measured. On the other hand, since the effective area A of the energized portion of the spin valve film is a value depending on the element structure, it needs to be carefully determined.

For example, when the entire spin valve film is patterned as an effective sensing region, the effective area A is the area of the entire spin valve film. In this case, from the viewpoint of appropriately setting the element resistance, the area of the spin valve film is set to at least 0.04 μm ^2, and at a recording density of 300 Gbpsi or less, it is set to 0.02 μm ² or less.

However, when the lower electrode 11 or the upper electrode 20 having a smaller area than the spin valve film is formed in contact with the spin valve film, the area of the lower electrode 11 or the upper electrode 20 becomes the effective area A of the spin valve film. When the area of the lower electrode 11 or the upper electrode 20 is different, the area of the smaller electrode is the effective area A of the spin valve film. In this case, from the viewpoint of appropriately setting the element resistance, the area of the smaller electrode is set to at least 0.04 μm ² or less.

  In the embodiment shown in FIGS. 6 and 7 to be described in detail later, the portion where the area of the spin valve film 10 is the smallest in FIG. 6 is the portion in contact with the upper electrode 20, so the width is considered as the track width Tw. Regarding the height direction, the portion in contact with the upper electrode 20 in FIG. 7 is the smallest, so the width is considered as the height length D. The effective area A of the spin valve film is considered as A = Tw × D.

  In the magnetoresistive effect element according to the embodiment of the present invention, the resistance R between the electrodes can be 100Ω or less. This resistance R is, for example, a resistance value measured between two electrode pads of a reproducing head unit attached to the tip of a head gimbal assembly (HGA).

  In the magnetoresistive effect element according to the embodiment of the present invention, when the pinned layer 14 or the free layer 18 has an fcc structure, it is desirable to have fcc (111) orientation. When the pinned layer 14 or the free layer 18 has a bcc structure, it is desirable to have bcc (110) orientation. When the pinned layer 14 or the free layer 18 has an hcp structure, it is desirable to have hcp (001) orientation or hcp (110) orientation.

  The crystal orientation of the magnetoresistive effect element according to the embodiment of the present invention is preferably 4.0 degrees or less, more preferably 3.5 degrees or less, and further preferably 3.0 degrees or less in terms of the orientation variation angle. This is calculated | required as a half value width of the rocking curve in the peak position obtained by (theta) -2 (theta) measurement of X-ray diffraction. Further, it can be detected as a dispersion angle of the spot at the nano-diffraction spot from the element cross section.

  Although depending on the material of the antiferromagnetic film, since the lattice spacing is generally different between the antiferromagnetic film and the pinned layer 14 / spacer layer 16 / free layer 18, the orientation variation angle in each layer is different. Can be calculated. For example, platinum manganese (PtMn) and pinned layer 14 / spacer layer 16 / free layer 18 often have different lattice spacings. Since platinum manganese (PtMn) is a relatively thick film, it is a suitable material for measuring variation in crystal orientation. Regarding the pinned layer 14 / spacer layer 16 / free layer 18, the pinned layer 14 and the free layer 18 may have different crystal structures such as a bcc structure and an fcc structure. In this case, the pinned layer 14 and the free layer 18 have different crystal orientation dispersion angles.

(Magnetic head)
6 and 7 show a state in which the magnetoresistive effect element according to the embodiment of the present invention is incorporated in a magnetic head. FIG. 6 is a cross-sectional view of the magnetoresistive element cut in a direction substantially parallel to a medium facing surface facing a magnetic recording medium (not shown). FIG. 7 is a cross-sectional view of the magnetoresistive element cut in a direction perpendicular to the medium facing surface ABS.

  The magnetic head illustrated in FIGS. 6 and 7 has a so-called hard abutted structure. The magnetoresistive film 10 is the CCP-CPP film described above. A lower electrode 11 and an upper electrode 20 are provided above and below the magnetoresistive effect film 10, respectively. In FIG. 6, a bias magnetic field application film 41 and an insulating film 42 are laminated on both sides of the magnetoresistive film 10. As shown in FIG. 7, a protective layer 43 is provided on the medium facing surface of the magnetoresistive film 10.

  The sense current for the magnetoresistive effect film 10 is energized in a direction substantially perpendicular to the film surface as indicated by the arrow A by the lower electrode 11 and the upper electrode 20 arranged above and below the magnetoresistive effect film 10. A bias magnetic field is applied to the magnetoresistive film 10 by a pair of bias magnetic field application films 41, 41 provided on the left and right. By this bias magnetic field, the magnetic anisotropy of the free layer 18 of the magnetoresistive effect film 10 is controlled to form a single magnetic domain, thereby stabilizing the magnetic domain structure and suppressing Barkhausen noise due to the domain wall movement. can do. Since the S / N ratio of the magnetoresistive film 10 is improved, high sensitivity magnetic reproduction is possible when applied to a magnetic head.

(Hard disk and head gimbal assembly)
The magnetic head shown in FIGS. 6 and 7 can be incorporated into a recording / reproducing integrated magnetic head assembly and mounted on a magnetic recording / reproducing apparatus.

  FIG. 8 is a main part perspective view illustrating a schematic configuration of such a magnetic recording / reproducing apparatus. That is, the magnetic recording / reproducing apparatus 150 of this embodiment is an apparatus of a type using a rotary actuator. In the figure, a magnetic disk 200 is mounted on a spindle 152 and rotated in the direction of arrow A by a motor (not shown) that responds to a control signal from a drive device control unit (not shown). The magnetic recording / reproducing apparatus 150 of this embodiment may include a plurality of magnetic disks 200.

  A head slider 153 that records and reproduces information stored in the magnetic disk 200 is attached to the tip of a thin-film suspension 154. The head slider 153 has a magnetic head including the magnetoresistive effect element according to any one of the above-described embodiments mounted near its tip.

  When the magnetic disk 200 rotates, the medium facing surface (ABS) of the head slider 153 is held with a predetermined flying height from the surface of the magnetic disk 200. Alternatively, a so-called “contact traveling type” in which the slider contacts the magnetic disk 200 may be used.

  The suspension 154 is connected to one end of the actuator arm 155. A voice coil motor 156, which is a kind of linear motor, is provided at the other end of the actuator arm 155. The voice coil motor 156 includes a drive coil (not shown) wound around a bobbin portion, and a magnetic circuit including a permanent magnet and a counter yoke arranged so as to sandwich the coil.

  The actuator arm 155 is held by ball bearings (not shown) provided at two positions above and below the spindle 157, and can be freely rotated and slid by a voice coil motor 156.

  FIG. 9 is an enlarged perspective view of the head gimbal assembly ahead of the actuator arm 155 as viewed from the disk side. That is, the assembly 160 has an actuator arm 155, and a suspension 154 is connected to one end of the actuator arm 155. A head slider 153 including a magnetic head including the magnetoresistive effect element according to any one of the above-described embodiments is attached to the tip of the suspension 154. The suspension 154 has a lead wire 164 for writing and reading signals, and the lead wire 164 and each electrode of the magnetic head incorporated in the head slider 153 are electrically connected. In the figure, reference numeral 165 denotes an electrode pad of the assembly 160.

  According to the present embodiment, by providing the magnetic head including the magnetoresistive element described above, it is possible to reliably read information magnetically recorded on the magnetic disk 200 at a high recording density.

(Magnetic memory)
Next, a magnetic memory equipped with the magnetoresistive effect element according to the embodiment of the present invention will be described. That is, by using the magnetoresistive effect element according to the embodiment of the present invention, a magnetic memory such as a random access magnetic memory (MRAM) in which memory cells are arranged in a matrix can be realized.

  FIG. 10 is a diagram showing an example of a matrix configuration of the magnetic memory according to the embodiment of the present invention. This figure shows a circuit configuration when memory cells are arranged in an array. In order to select one bit in the array, a column decoder 350 and a row decoder 351 are provided, and the switching transistor 330 is turned on by the bit line 334 and the word line 332 to be uniquely selected and detected by the sense amplifier 352. As a result, the bit information recorded in the magnetic recording layer (free layer) in the magnetoresistive film 10 can be read. When writing bit information, a magnetic field generated by applying a write current to a specific write word line 323 and bit line 322 is applied.

  FIG. 11 is a diagram showing another example of the matrix configuration of the magnetic memory according to the embodiment of the present invention. In this case, bit lines 322 and word lines 334 wired in a matrix are selected by decoders 360 and 361, respectively, and specific memory cells in the array are selected. Each memory cell has a structure in which a magnetoresistive element 10 and a diode D are connected in series. Here, the diode D has a role of preventing the sense current from bypassing in the memory cells other than the selected magnetoresistive effect element 10. Writing is performed by a magnetic field generated by supplying a write current to the specific bit line 322 and the write word line 323, respectively.

  FIG. 12 is a cross-sectional view showing the main part of the magnetic memory according to the embodiment of the present invention. FIG. 13 is a cross-sectional view taken along the line A-A ′ of FIG. 12. The structures shown in these drawings correspond to 1-bit memory cells included in the magnetic memory shown in FIG. This memory cell has a memory element portion 311 and an address selection transistor portion 312.

  The memory element portion 311 includes the magnetoresistive effect element 10 and a pair of wirings 322 and 324 connected thereto. The magnetoresistive effect element 10 is the magnetoresistive effect element (CCP-CPP element) according to the above-described embodiment.

  On the other hand, the address selection transistor portion 312 is provided with a transistor 330 connected via a via 326 and a buried wiring 328. The transistor 330 performs a switching operation according to the voltage applied to the gate 332, and controls opening and closing of a current path between the magnetoresistive effect element 10 and the wiring 334.

  Further, below the magnetoresistive effect element 10, a write wiring 323 is provided in a direction substantially orthogonal to the wiring 322. These write wirings 322 and 323 can be formed of, for example, aluminum (Al), copper (Cu), tungsten (W), tantalum (Ta), or an alloy containing any of these.

  In the memory cell having such a configuration, when writing bit information to the magnetoresistive effect element 10, a write pulse current is supplied to the wirings 322 and 323, and a composite magnetic field induced by these currents is applied to thereby apply the magnetoresistive effect element. The magnetization of the recording layer is appropriately reversed.

  When reading bit information, a sense current is passed through the wiring 322, the magnetoresistive effect element 10 including the magnetic recording layer, and the lower electrode 324, and the resistance value of the magnetoresistive effect element 10 or a change in the resistance value is measured. To do.

  The magnetic memory according to the embodiment of the present invention uses the magnetoresistive effect element (CCP-CPP element) according to the embodiment described above to reliably control the magnetic domain of the recording layer even if the cell size is reduced. Reliable writing can be ensured and reading can be performed reliably.

(Other embodiments)
Embodiments of the present invention are not limited to the above-described embodiments, and can be expanded and modified. The expanded and modified embodiments are also included in the technical scope of the present invention. The specific structure of the magnetoresistive film, and other shapes and materials of the electrode, bias application film, insulating film, etc., are appropriately implemented by those skilled in the art, and the present invention is similarly implemented. The effect of can be obtained. For example, when applying a magnetoresistive element to a reproducing magnetic head, the detection resolution of the magnetic head can be defined by providing magnetic shields above and below the element.

  The embodiment of the present invention can be applied not only to a longitudinal magnetic recording system but also to a perpendicular magnetic recording system magnetic head or magnetic reproducing apparatus. Furthermore, the magnetic reproducing apparatus of the present invention may be a so-called fixed type having a specific recording medium constantly provided, or a so-called “removable” type in which the recording medium can be replaced.

  In addition, all magnetoresistive elements, magnetic heads, magnetic storage / reproduction devices, and magnetic memories that can be appropriately designed and implemented by those skilled in the art based on the magnetic head and magnetic storage / reproduction device described above as embodiments of the present invention Are also within the scope of the present invention.

DESCRIPTION OF SYMBOLS 10 ... Magnetoresistance effect film | membrane, 11 ... Lower electrode, 12 ... Underlayer, 12a ... Buffer layer, 12b ... Seed layer, 13 ... Pinning layer, 14 ... Pin layer, 141 ... Lower pin layer, 142 ... Magnetic coupling layer, 143 ... upper pin layer, 15 ... lower metal layer, 16 ... spacer layer, 161 ... insulating layer, 162 ... current path, 17 ... upper metal layer, 18 ... free layer, 19 ... cap layer, 20 ... upper electrode, 21 ... current Path, 21 '... current path, 162 ... current path combining 21 and 21', 22 ... insulating layer, 22 '... insulating layer, 161 ... insulating layer combining 22 and 22'

Claims (10)

A magnetization pinned layer in which the magnetization direction is substantially pinned in one direction, a magnetization free layer whose magnetization direction changes corresponding to an external magnetic field, and provided between the magnetization pinned layer and the magnetization free layer and insulated In a method for manufacturing a magnetoresistive effect element having a layer and a spacer layer including a metal layer penetrating the insulating layer,
In forming the spacer layer,
Depositing a first metal layer;
Forming a second metal layer on the first metal layer;
The second metal layer is converted into a first insulating layer, and the first metal layer is sucked into the first insulating layer, whereby a first conductive layer penetrating the first insulating layer is formed. Perform a first conversion process to form,
Forming a third metal layer on the first insulating layer and the first conductive layer;
Converting the third metal layer into a second insulating layer and sucking the first conductive layer into the second insulating layer, so that the second insulating layer is only on the first conductive layer. By performing a second conversion process for forming a second conductive layer penetrating through the metal layer, the metal layer having the first conductive layer and the second conductive layer, and the first insulating layer and the second Forming the insulating layer having an insulating layer of
At least one of the first conversion process and the second conversion process is:
A first step of irradiating the film surface by ionizing or plasmaizing a gas containing at least one element of argon, xenon, helium, neon, and krypton;
A gas containing at least one of oxygen gas and nitrogen gas is supplied into an atmosphere obtained by ionizing or plasmaizing a gas containing at least one of argon, xenon, helium, neon, and krypton, and the ionized gas or A second step of performing at least one of an oxidation treatment and a nitridation treatment on the second metal layer or the third metal layer under an atmosphere of the gasified plasma;
A method for manufacturing a magnetoresistive effect element, comprising:   The fourth metal layer made of the same material as the first metal layer is formed after the first conversion process and before the formation of the third metal layer. The manufacturing method of the magnetoresistive effect element of description.   As the first metal layer, a material containing at least one element of Cu, Au, Ag, and Al is formed, and as the second and third metal layers, Al, Si, Mg, Ti, 3. The magnetoresistive effect element manufacturing method according to claim 1, wherein a material containing at least one element of Hf, Zr, Cr, Mo, Nb, and W is formed. Method.   3. The method of manufacturing a magnetoresistive element according to claim 2, wherein a material containing at least one element of Cu, Au, Ag, and Al is formed as the fourth metal layer.   The film thickness of the first metal layer is 0.1 to 1.5 nm, and the film thickness of the second and third metal layers is 0.3 to 1 nm. Manufacturing method of magnetoresistive effect element.   5. The method of manufacturing a magnetoresistive effect element according to claim 4, wherein the thickness of the fourth metal layer is 0.1 to 1.5 nm.   The magnetic layer according to any one of claims 1 to 6, wherein a metal layer containing at least one element of Cu, Au, Ag, and Al is formed after the second conversion treatment. A method of manufacturing a resistance effect element.   8. The method of manufacturing a magnetoresistive element according to claim 1, wherein at least one of the magnetization pinned layer and the magnetization free layer is formed of an alloy containing Co and Fe. .   The method of manufacturing a magnetoresistive element according to claim 1, wherein the magnetization pinned layer has a body-centered cubic structure.   The method for manufacturing a magnetoresistive element according to claim 1, wherein the magnetization free layer includes a layer formed of an alloy containing Ni and Fe.

Images