Classifications

• • H—ELECTRICITY
  • H03—ELECTRONIC CIRCUITRY
  • H03B—GENERATION OF OSCILLATIONS, DIRECTLY OR BY FREQUENCY-CHANGING, BY CIRCUITS EMPLOYING ACTIVE ELEMENTS WHICH OPERATE IN A NON-SWITCHING MANNER; GENERATION OF NOISE BY SUCH CIRCUITS
  • H03B28/00—Generation of oscillations by methods not covered by groups H03B5/00 - H03B27/00, including modification of the waveform to produce sinusoidal oscillations
• • B—PERFORMING OPERATIONS; TRANSPORTING
  • B82—NANOTECHNOLOGY
  • B82Y—SPECIFIC USES OR APPLICATIONS OF NANOSTRUCTURES; MEASUREMENT OR ANALYSIS OF NANOSTRUCTURES; MANUFACTURE OR TREATMENT OF NANOSTRUCTURES
  • B82Y25/00—Nanomagnetism, e.g. magnetoimpedance, anisotropic magnetoresistance, giant magnetoresistance or tunneling magnetoresistance
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01F—MAGNETS; INDUCTANCES; TRANSFORMERS; SELECTION OF MATERIALS FOR THEIR MAGNETIC PROPERTIES
  • H01F10/00—Thin magnetic films, e.g. of one-domain structure
  • H01F10/32—Spin-exchange-coupled multilayers, e.g. nanostructured superlattices
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01F—MAGNETS; INDUCTANCES; TRANSFORMERS; SELECTION OF MATERIALS FOR THEIR MAGNETIC PROPERTIES
  • H01F10/00—Thin magnetic films, e.g. of one-domain structure
  • H01F10/32—Spin-exchange-coupled multilayers, e.g. nanostructured superlattices
  • H01F10/324—Exchange coupling of magnetic film pairs via a very thin non-magnetic spacer, e.g. by exchange with conduction electrons of the spacer
  • H01F10/3254—Exchange coupling of magnetic film pairs via a very thin non-magnetic spacer, e.g. by exchange with conduction electrons of the spacer the spacer being semiconducting or insulating, e.g. for spin tunnel junction [STJ]
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01F—MAGNETS; INDUCTANCES; TRANSFORMERS; SELECTION OF MATERIALS FOR THEIR MAGNETIC PROPERTIES
  • H01F10/00—Thin magnetic films, e.g. of one-domain structure
  • H01F10/32—Spin-exchange-coupled multilayers, e.g. nanostructured superlattices
  • H01F10/324—Exchange coupling of magnetic film pairs via a very thin non-magnetic spacer, e.g. by exchange with conduction electrons of the spacer
  • H01F10/3254—Exchange coupling of magnetic film pairs via a very thin non-magnetic spacer, e.g. by exchange with conduction electrons of the spacer the spacer being semiconducting or insulating, e.g. for spin tunnel junction [STJ]
  • H01F10/3259—Spin-exchange-coupled multilayers comprising at least a nanooxide layer [NOL], e.g. with a NOL spacer
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01F—MAGNETS; INDUCTANCES; TRANSFORMERS; SELECTION OF MATERIALS FOR THEIR MAGNETIC PROPERTIES
  • H01F10/00—Thin magnetic films, e.g. of one-domain structure
  • H01F10/32—Spin-exchange-coupled multilayers, e.g. nanostructured superlattices
  • H01F10/324—Exchange coupling of magnetic film pairs via a very thin non-magnetic spacer, e.g. by exchange with conduction electrons of the spacer
  • H01F10/3268—Exchange coupling of magnetic film pairs via a very thin non-magnetic spacer, e.g. by exchange with conduction electrons of the spacer the exchange coupling being asymmetric, e.g. by use of additional pinning, by using antiferromagnetic or ferromagnetic coupling interface, i.e. so-called spin-valve [SV] structure, e.g. NiFe/Cu/NiFe/FeMn
• • H—ELECTRICITY
  • H03—ELECTRONIC CIRCUITRY
  • H03B—GENERATION OF OSCILLATIONS, DIRECTLY OR BY FREQUENCY-CHANGING, BY CIRCUITS EMPLOYING ACTIVE ELEMENTS WHICH OPERATE IN A NON-SWITCHING MANNER; GENERATION OF NOISE BY SUCH CIRCUITS
  • H03B15/00—Generation of oscillations using galvano-magnetic devices, e.g. Hall-effect devices, or using superconductivity effects
  • H03B15/006—Generation of oscillations using galvano-magnetic devices, e.g. Hall-effect devices, or using superconductivity effects using spin transfer effects or giant magnetoresistance

Definitions

• the present invention belongs to the field of radio-frequency spin-transfer oscillators, called STO oscillator ("Spin Torque Oscillator").
• STO oscillators are intended to operate in frequency ranges between a few hundred MHz and a few GHz.
• oscillator refers to a physical system of which at least one of the properties varies periodically or quasi-periodically in time.
• the role of an oscillator is to serve as a reference for time and frequency.
• VCO type oscillators for "Voltage Controlled Oscillator” are known for example: these devices are based on a resonant electronic circuit comprising a resistor R, an inductance L and a capacitance C, all forming an RLC circuit. VCO oscillators not only use this RLC architecture but also have a bias voltage to vary the values of L and C and thus change the frequency
• VCOs The electronics of new telecommunications products (mobile phones, for example) must be able to work over very wide frequency ranges.
• the VCOs must be capable of transmitting over several frequency ranges. For example, today's mobile phones have three or four frequency bands.
• nomadic technologies imposes additional constraints in terms of compactness of products.
• the first solution is the most suitable but can not be envisaged with a single VCO that has too little agility.
• the current solution involves the use of several VCOs, which poses a problem of congestion and adds interference phenomena between the different VCOs.
• a known solution capable of responding to the problems mentioned above is to turn to radiofrequency oscillators based on spintronic devices called STO oscillators.
• GMR giant magnetoresistance effects
• TMR Magnetoresistance Tunnel MagnetoResistance
• These structures consist of a stack of magnetic layers, the nature and arrangement of which are such that when an electric current passes through them, it is possible to obtain a variable resistance as a function of the applied magnetic field and / or the spin polarized current flowing through them.
• a device consists of the stack of two ferromagnetic layers (a magnetic layer called “trapped” whose magnetization is fixed direction and a so-called “free” layer whose magnetization is variable) separated by a non-magnetic layer (no magnetic) conventionally referred to as a spacer, made of metal for spin or oxide valves for magnetic tunnel junctions.
• spin electronics use the spin of electrons as an additional degree of freedom to generate new effects.
• the spin of electrons is at the origin of magnetoresistive phenomena in magnetic multi-layers, such as in particular giant magnetoresistance or tunnel magnetoresistance.
• the operating principle of the STO (Spin Torque Oscillator) consists in using the spin transfer torque to trigger a sustained precession of the magnetization.
• this precession causes an oscillation of the resistance and therefore the generation of an alternating voltage in the GHz range.
• the major advantage of STO oscillators is their high frequency agility since the resonant frequency changes over a very wide band depending on the polarization applied to the spintronic device.
• the main technical problem posed by the STOs remains the spectral purity of these oscillators, the linewidth typically being a few tens of MHz in the best case (see for example, Miizushima et al., Appl Phys Lett. 94, 152501, 2009, Georges et al., Phys Rev B 80, 060404 (R), 2009).
• This large line width is due to instabilities of the trajectory of the magnetization. This problem is not only related to thermal fluctuations but is also present at low temperature (see Georges et al., Phys Rev B 80, 060404 (R), 2009).
• the synchronization reduces the frequency tunability of all coupled oscillators.
• the powers emitted by these STOs are relatively low (at most a few thousand nV 2 / Hz - see, for example, Houssameddine et al., Appl Phys., Lett., 93, 022505, 2008 ).
• the principle described in this document is to take advantage of large current densities in the vicinity of the conductive bridges to locally generate oscillations of the magnetization without risking reversing the magnetization of the soft layer. Indeed, in the areas separating the conductive bridges, the current densities being much lower, the magnetization of these zones is not subjected to spin transfer torques and therefore remains almost at rest. Thus, the excitations generated at each conductive bridge are coupled together via spin waves. But, between each conductor bridge, there remains an intermediate zone in which the magnetization is not generally excited. These zones can be traversed by spin waves originating from neighboring conductive bridges but do not undergo magnetic excitations of great amplitude.
• the object of the present invention is to provide a spin-transfer oscillator having, compared with existing STOs, an increased transmitted power, a good quality factor, a very high spectral purity (ie a line width of narrow fundamental frequency), a reduced phase noise and the possibility of exploiting a plurality of harmonics in order to be able to exploit higher frequencies.
• the invention proposes a spin transfer oscillator comprising:
• a magnetic stack including at least two magnetic layers, of which at least one of said two magnetic layers, said oscillating layer, has a magnetization of variable direction;
• current supply means able to circulate a current of electrons perpendicular to the plane of said magnetic stack
• said magnetic stack comprising means capable of generating current inhomogeneities at the surface of said oscillating layer, said oscillator being characterized in that the intensity of current supplied by said supply means is chosen so that the magnetization said oscillating layer has a coherent magnetic configuration, said magnetic configuration oscillating as a whole at the same fundamental frequency.
• coherent magnetic configuration is meant a configuration close to the magnetic mono-domain or having magnetization rotations on a scale of length comparable to the lateral dimension of the surface of the oscillating layer: it may be a configuration of the type magnetic vortex, magnetic anti-vortex, multivortex or "C-state” or a combination of these configurations (for example several vortex or one or more vortex-vortex pair (s)).
• coherent micromagnetic configurations are described in the article "Stability of magnetic vortices in flat submicron permalloy cylinders" (M. Schneider et al., Journ.Appl.Phys.92 (2002) 1466). In the remainder of the description, we will speak of a vortex configuration for this coherent magnetic configuration, but the person skilled in the art can implement the invention for other types of coherent magnetic configurations.
• the oscillating layer may be a ferromagnetic single layer but also a synthetic antiferromagnetic layer comprising two ferromagnetic soft layers, possibly of different thicknesses, separated by an antiferromagnetic coupling layer.
• the oscillator according to the invention is based on the dynamics of collective and coherent excitations of the magnetization of the entire device (for example the dynamics of a magnetic vortex) induced by the spin transfer torque in a magnetic structure exhibiting strong lateral inhomogeneities in the current distribution passing through the structure via means capable of generating current inhomogeneities at the surface of the oscillating layer: these inhomogeneities may for example be created in a non-limiting manner by a set of nanowires which are conductive to through an insulating barrier that allow the localization of current lines. Typically, these inhomogeneities result in fluctuations of more than a factor 10 of conductivity between the most conductive zones and the less conductive zones.
• the present invention differs from the prior art (and in particular document US7504898) in that the inhomogeneous current structure is not used to generate magnetization precessions on localized zones interacting with each other by means of spin waves; on the contrary, surprisingly, the Applicant has found that a current intensity greater than a threshold determined as a function of the magnetic structure used to trigger a collective excitation of the whole magnetization of the oscillating layer of the oscillator, such as a vortex-type dynamic, whose frequency damentale is generally at a lower frequency (of the order of a few hundred MHz) than the fundamental frequency of known STOs.
• a threshold determined as a function of the magnetic structure used to trigger a collective excitation of the whole magnetization of the oscillating layer of the oscillator, such as a vortex-type dynamic, whose frequency damentale is generally at a lower frequency (of the order of a few hundred MHz) than the fundamental frequency of known STOs.
• the magnetization evolves over the entire surface of the excited layer of the oscillator (oscillating layer) causing a complete stirring of the magnetization of this layer.
• the present invention does not seek at all to avoid the reversal of the magnetization of the excited layer unlike the US7504898 patent but on the contrary to induce a collective motion of excitation of the coherent magnetic pattern of magnetization.
• the excited layer of the oscillator encompassing the entire surface thereof.
• the invention is therefore based on the use of the spin transfer torque generated by the inhomogeneous current distribution which, by an average effect due to the exchange interactions inside the oscillating layer, collectively excites the entire of the magnetization of this oscillating layer.
• the principle of the invention therefore consists in using an inhomogeneous current structure to simultaneously generate:
• the first effect can occur alone for a lower current intensity; we then have a static coherent magnetic configuration; by increasing the intensity of the current beyond a threshold intensity, the two effects are observed simultaneously (dynamic coherent magnetic configuration).
• the oscillating motion of the coherent magnetic configuration results in a cyclic temporal change in the average magnetization, and hence in a resistance oscillation due to the magnetoresistive effect. It is thus possible to generate an alternating voltage across the terminals of the device, sensing an intense fundamental mode (typically between 300 and 500 MHz) and many harmonics (typically 12 to 15).
• the oscillator according to the invention can operate either with a fixed magnetization layer (trapped layer) and a variable magnetization layer (oscillating layer) or with two oscillating layers (with or without a reference layer), the only condition being that the magnetic stack comprises at least two magnetic layers, one of which is an oscillating layer with a variable direction magnetization.
• the invention makes it possible to produce a frequency tunable oscillator, but of high power (several hundreds of thousands of nV 2 / Hz, ie one to two orders of magnitude greater than the known STO) and having a very high spectral purity (Q> 300, ⁇ ⁇ 2 ⁇ ).
• the fundamental frequency is typically between a few hundred MHz and a few GHz.
• the device according to the invention also makes it possible to generate a large series of harmonics (typically at least four harmonics whose amplitude remains lower than that of the fundamental frequency but sufficient to be largely exploitable): thus, according to the intended application, it will be possible to use the emitted fundamental frequency or some of the harmonics generated by the oscillator to increase in frequency.
• the oscillator according to the invention may also have one or more of the following characteristics, considered individually or in any technically possible combination:
• said means capable of generating current inhomogeneities at the surface of said oscillating layer are formed by a layer having strong inhomogeneities of conductivity able to generate in said oscillating layer strong lateral inhomogeneities of current;
• said layer having strong inhomogeneities of conductivity capable of generating in said oscillating layer strong lateral inhomogeneities of current is a layer characterized by fluctuations of more than a factor of conductivity between its most conductive zones and its zones the less conductive; said layer having high conductivity inhomogeneities is an insulating layer incorporating metal paths;
• said layer having high conductivity inhomogeneities is a tunnel junction integrating conductive paths obtained by exposing said junction to a suitable voltage (i.e. the voltage must be high enough to locally generate electrical breakdown phenomena);
• said layer having high inhomogeneities of conductivity is made based on a mixture of two materials having different conductivities and able to demix under the effect of annealing to form locally more conductive paths than on the rest of the layer ; in other words, said layer having high inhomogeneities of conductivity is made based on a mixture of two immiscible materials and having different electrical conductivities so as to form strong inhomogeneities of electrical conductivity under the effect of a annealing causing the demixing of the two constituent materials;
• said layer having high inhomogeneities of conductivity has a ratio greater than or equal to 10 conductivity between its most conductive zones and its less conductive zones; said layer having high conductivity inhomogeneities comprises in contact with at least one of its faces a nonmagnetic metal layer;
• said means capable of generating current inhomogeneities at the surface of said oscillating layer are formed by a plurality of nanocontacts
• said coherent magnetic configuration is a magnetic vortex, magnetic vortex, multivortex or "C-state” configuration or a combination of these configurations;
• said magnetic stack has a pillar shape whose section has one of the following forms:
• the diameter in the case of a circular, annular or flower-shaped section is the larger of the two diameters; in the case of a cell of elliptical shape, it is between 10 nm and 500 nm and preferably between 100 and 500 nm;
• said oscillating layer has a substantially circular section of radius R and a thickness L, the ratio L / R being chosen according to the operating frequency of the oscillator, if necessary iteratively using micromagnetic simulation software to determine the corresponding frequency at a given value of L / R;
• said magnetic stack comprises successively:
• said magnetic stack successively comprises two magnetic sub-stacks separated by a magnetic decoupling layer, each of the two sub-stacks integrating:
• said magnetic stack comprises a magnetic sub-stack successively including:
• said means capable of generating current inhomogeneities at the surface of said oscillating layer being located outside said sub-stack;
• said magnetic stack comprises two magnetic sub-stacks, each sub-stack including successively:
• said two sub-stacks being separated by a layer having high conductivity inhomogeneities
• said magnetic stack comprises two oscillating ferromagnetic layers of variable magnetization direction separated by said means capable of generating current inhomogeneities at the surface of said oscillating layers;
• said magnetic stack comprises:
• a magnetic sub-stacking successively including:
• a second oscillating magnetic layer of variable magnetization direction A second oscillating magnetic layer of variable magnetization direction
• first means capable of generating current inhomogeneities at the surface of said first oscillating layer
• FIGS. 1 to 3 illustrate the evolution of the magnetic configuration of the oscillating layer of the oscillator according to the invention as a function of the current injected into the oscillator;
• FIG. 4 illustrates a first embodiment of an oscillator according to the invention
• FIG. 5 illustrates a second embodiment of an oscillator according to the invention
• FIG. 6 illustrates a third embodiment of an oscillator according to the invention
• FIG. 7 illustrates a fourth embodiment of an oscillator according to the invention.
• FIG. 8 represents the excitation spectrum as a function of frequency for an oscillator according to the invention.
• FIG. 9 illustrates a fifth embodiment of an oscillator according to the invention.
• FIG. 10 illustrates a sixth embodiment of an oscillator according to the invention.
• FIG. 11 illustrates a seventh embodiment of an oscillator according to the invention
• FIG. 12 illustrates an eighth embodiment of an oscillator according to the invention.
• FIG. 13 illustrates a ninth embodiment of an oscillator according to the invention illustrates a fourth embodiment of an oscillator according to the invention
• FIG. 14 illustrates a tenth embodiment of an oscillator according to the invention
• FIG. 15 illustrates a particular section of an oscillator according to the invention.
• FIG. 4 schematically represents a first embodiment of an oscillator 30 according to the invention.
• Oscillator 30 comprises a magnetic stack E successively including:
• a ferromagnetic reference layer 34 with a fixed magnetization direction (here located in the plane of the reference layer), called a trapped layer;
• This stack E formed by the above set of layers is inserted between two electrodes 31 and 35 current supply for circulating an electric current of intensity I perpendicular to the plane of the stack of magnetic layers.
• the two electrodes 31 and 35 form current supply means able to circulate a current of electrons perpendicular to the plane of the magnetic stack E.
• This stack E has the shape of a pillar that can have different shapes of sections: circular, elliptical, annular, even more complex shapes such as flowers with 3 or 4 petals (see below about Figure 15).
• the characteristic diameter of the device is typically from a few tens to a few hundred nm (typically ⁇ 500 nm).
• the reference layer 34 may be a single layer made from an alloy based on Co, Fe, Ni, for example a CoFe layer of 3 nm.
• the reference layer 34 may also be a synthetic antiferromagnetic layer consisting of two ferromagnetic layers antiparallelly coupled through a thin antiparallel coupling layer, for example ruthenium of thickness 0.6 nm to 0.8 nm.
• This synthetic antiferromagnetic layer may itself be trapped by interaction with an antiferromagnetic spacer layer (for example ruthenium); an example of a synthetic antiferromagnetic entrapped layer is CoFe 2nm / Ru 0.6nm / CoFe 2.5 nm.
• the layer 33 with strong lateral inhomogeneities of electrical conductivity is a layer having in its plane strong lateral inhomogeneities of electrical conductivity (typically characterized by fluctuations of more than a conductivity factor of 10 between the most conductive zones and the least conductors of layer 33).
• This layer 33 with high lateral inhomogeneity electrical conductivity can be achieved by using an insulating layer pierced with small metal conductive holes 36 (called “pinholes” in English).
• the advantage of this separating layer is to allow confinement of the current lines at the nanoconstrictions.
• This layer structure 33 is called “confined current path” (CCP) for "Confined Current Path” ("current screen layer” in English).
• CCP confined current path
• This particular separating layer can be made for example by depositing an AlCu alloy with a few% copper which is then oxidized to form an alumina layer containing some metal clusters which will form the pinholes 36.
• the pinholes 36 typically have a size from 0.5nm to ten nanometers in diameter.
• the insulating layer containing the pinholes 36 may be formed by oxidation, nitriding or oxynitriding an initially metallic layer containing at least one element in the following group: Al, Hf, Zr, Ta, Ti, Mo, W, Nb, Si, mg. It typically has a thickness between 0.8 and 3.5nm.
• Pinholes 36, for their part, which may be non-magnetic or magnetic, comprise at least one of Cu, Ag, Au, Co, Fe or Ni elements.
• This insulating barrier 33 pierced with conducting holes 36 is most often made by depositing a layer of metal alloys comprising two elements having a strong difference in affinity for oxygen, for example of composition AI95Cu5. This layer is then oxidized as a tunnel barrier by natural oxidation or plasma oxidation. Due to the difference in oxygen affinity of aluminum and copper, aluminum oxidizes to alumina while copper coalesces in the form of copper aggregates to form conductive bridges across the aluminum layer. alumina.
• the current densities in the zones of high conductivity can be very important. (10 9 to 10 11 A / cm 2 ) so that spin transfer effects can be locally very significant in normal operation.
• Spin transfer is the reciprocal effect of spin filtering implemented in giant magnetoresistances. In the magnetoresistance phenomena, an electric current can be controlled by acting on the relative orientation of the magnetizations of the magnetic layers, according to a filtering effect analogous to a polarizer-optical analyzer assembly.
• the spin transfer effect corresponds to the interaction of the current on the filter: the spin polarized current, if sufficiently intense, can modify the orientation of the magnetization of the oscillating layer 32.
• the oscillating layer 32 is for example formed by a layer of Co or alloys Permalloy (Ni80Fe20) or alloys of Co and Fe rich in Co, for example Co90Fe10.
• the invention advantageously uses the current inhomogeneity created by the layer 33 to simultaneously generate a double effect at the level of the oscillating layer 32:
• This coherent micromagnetic configuration can for example be of the vortex, multivortex, vortex, or "C-state" type or a combination of these configurations;
• the coherent micromagnetic configuration of the magnetization is influenced by the ampere field induced by the current flowing through the structure.
• This inhomogeneous field favors the appearance of one or more vortices, or vortex-antivortex (or vortex or "C-state") pair (s).
• This confi- The particular magnetization configuration is obtained when the external magnetic field applied in the plane of the layers is low (typically ⁇ 20mT), provided that the structure has been magnetically prepared in a state where the oscillating layer is saturated under a field ( Typically 60mT) of opposite sign.
• the magnetic stack E constituting the active part of the device is of intermediate size, not too small for the vortex to have room to form, nor too large for that the dynamics of this vortex keeps its coherence (typically one dimension, the diameter in the case of the cylindrical form of the stack E, from 100 to 500 nm).
• the shape has a substantially cylindrical symmetry if one seeks to exploit the fundamental signal of excitations. But other forms are possible if one seeks to amplify the amplitude of the harmonic signals it will be as explained further with reference to FIG.
• FIGS. 1 to 3 respectively showing the magnetic configuration of the oscillating layer 32 with a zero intensity of current flowing through the zero structure (FIG. 1), an intermediate current intensity (FIG. current l 2 greater than h ( Figure 3).
• FIG. 1 shows the magnetic configuration 1 of the oscillating layer 32 in the absence of current flowing through the structure 30.
• This configuration 1 (the magnetization vectors are represented by arrows) has a magnetization having substantially the shape of ribs of sheet with a right magnetization 2 in the center of the layer 32 along a diameter of the layer 32 and a magnetization 3 slightly curved and symmetrical on either side of the magnetization 2 followed by a magnetization 4 always symmetrical on the one hand and other of the magnetization 2 and whose curvature is accentuated towards the edge of the layer 32 to follow the circular shape of the layer 32.
• Vortex core there is also a singularity in the center of the vortex 10 in a zone 1 1 called "vortex core" in which the magnetization can no longer rotate.
• the magnetization tends to come out of the plane of the layer, but this does not play a significant role in the context of this invention.
• the vortex core 1 1 coincides with the center C of the oscillating layer 32.
• a static coherent micromagnetic configuration appears beyond a threshold current I V ortex_static-
• a current of intensity greater than an intensity of the order of 5 mA will generate a static vortex for an oscillator according to the invention with a diameter of the order of 300 nm.
• the threshold current appearance of a coherent static micromagnetic configuration the V ort e x_statiq Ue is typically between 4 and 5 mA.
• the spin transfer effect can destabilize the magnetic configuration 10 and cause a cyclic vortex movement.
• a current of intensity greater than 17 mA will generate a dynamic vortex (ie set in motion the vortex).
• the threshold threshold of appearance of the vortex cyclic motion V ort e x_dynamic Ue is typically between 17 and 20 mA. It should be noted that the movement of the vortex then persists when the current is lowered below this threshold current: typically the vortex oscillation is triggered with an intensity greater than 17 mA and it then persists when the current is down to 6mA.
• the coherent magnetic configuration appearing above the first critical current is not a vortex of perfect cylindrical symmetry but a distorted vortex structure for example because of the magnetostatic field radiated by the magnetization of the reference layer. on the magnetization of the oscillating layer.
• this state of coherent magnetization has a magnetization which according to the magnetic history of the sample may be closer to the parallel orientation or closer to the antiparallel orientation with respect to the magnetization of the magnetization layer. reference.
• the resistance of the oscillator 30 in the coherent micromagnetic configuration is closer to the resistance of the antiparallel state AP (ie the average magnetization of the oscillating layer 32 is closer to the antiparallel state AP than to the parallel state P with respect to the magnetization of the reference layer), it is necessary that the current flows from the oscillating layer 32 to the reference layer 34 (ie that the electrons flow from the reference layer 34 to the oscillating layer 32 ).
• the spin transfer effect favors the parallel alignment of the magnetizations and destabilizes the AP state
• the resistance of the oscillator 30 in the coherent micromagnetic configuration is closer to the resistance of the state P (ie the average magnetization of the oscillating layer 32 is closer to the parallel state than to the antiparallel state compared to the magnetization of the reference layer)
• the current must flow from the reference layer 34 to the oscillating layer 32 (ie the electrons flow from the oscillating layer 32 to the reference layer 34).
• the spin transfer effect favors the antiparallel alignment of the magnetizations and destabilizes the P state.
• the initial micromagnetic state has a quasi-zero mean magnetization because of its cylindrical symmetry (case of the symmetrical vortex of FIG. 2), the polarity of the current does not matter.
• the cyclic movement of the vortex thus obtained results in a cyclic temporal modification of the average magnetization, and consequently in a resistance oscillation because of the magnetoresistive effect. It is thus possible to generate an alternating voltage across the oscillator 30, having an intense fundamental mode (typically between 300 and 500 MHz) and numerous harmonics (typically from 12 to 15).
• the amplitude of the fundamental mode is very high (> 500 000 nV 2 / Hz) and the line width of the order of 1 MHz.
• the amplitude of the harmonics decreases as the frequency increases; the amplitude of the harmonics 2f 0 and 3f 0 represents approximately 10% of the amplitude of the fundamental and the amplitude of the harmonic 4f 0 some% of the amplitude of the fundamental.
• FIG. 8 represents the excitation spectrum as a function of frequency: a fundamental mode power is observed at f 0 (400 MHz) of the order of 500 000 nV 2 / Hz, a thousand times higher than that obtained with the STO of the state of the art, for a power injected by the direct current only two to three times greater .
• 5f 0 2 GHz 3500 nV 2 / Hz 20 MHz
• the frequency of the vortex translation mode corresponding to an orbital movement of the vortex core, is directly related to the ratio L / R, where L is the thickness of the oscillating layer 32 and R is the radius of this oscillating layer.
• L is the thickness of the oscillating layer 32
• R is the radius of this oscillating layer.
• the natural frequency of the translation mode increases almost linearly with this geometric ratio. It is thus possible to envisage reaching frequencies of 1 to 2 GHz on nanometric samples in which the oscillating layer would have a thickness of 60 nm.
• the choice of materials can also make it possible to increase the fundamental frequency.
• the orbital motion of the vortex can be understood as the effect resulting on the one hand from the spin transfer torque that tends to cause a centrifugal movement of the vortex core and on the other hand the Amperian field which tends to create a attractive potential at the contacts of the layer 33 CCP. All the parameters allowing to reduce the size of the orbit make it possible to increase the frequency: this is why the frequency increases for example with the applied current.
• the frequency will increase with Gilbert's damping (see Mistral et al., Phys Rev. Lett 100, 257201, 2008). Now it is possible to increase the Gilbert damping by doping the ferromagnetic material used for the oscillating layer 32 with rare earth impurities such as Dy or Tb, for example.
• Layer 33 with high lateral inhomogeneity of conductivity based on pinhole structure (CCP) is only one exemplary embodiment.
• the layer 33 with high lateral inhomogeneity of conductivity is a layer having in its plane strong lateral inhomogeneities of electrical conductivity (typically characterized by fluctuations of more than a conductivity factor of 10 between the most conductors and the least conductive areas of layer 33).
• FIG. 5 illustrates a second embodiment of an oscillator 50 according to the invention.
• the oscillator 50 of FIG. 5 is identical to the oscillator 30 of FIG. 4 (the common elements have the same reference numbers), the only difference relating to the layer 53 with strong lateral inhomogeneities of different electrical conductivity of the layer 33 of Figure 4.
• the layer 53 with strong lateral inhomogeneities of electrical conductivity able to induce in the oscillating layer 32 strong current inhomogeneities are obtained starting from a magnetic tunnel junction to which is subjected to an electrical stress close to electrical breakdown of the junction.
• tunnel barriers of alumina or MgO with a thickness between 1 and 2 nm electrically clash under voltages of the order of 1 to 2V.
• the breakdown voltage depends in fact on the duration and the number of voltage taps and mainly the cumulative time during which the junction is exposed to the high voltage. It is known that before slamming completely irreversibly, the tunnel barrier can degrade locally under the effect of electrical stress leading to the appearance of "hot spots" 56 in the junction.
• hot spots 56 can be followed by measuring the electrical resistance of the junction, which is traversed by a current perpendicular to the interfaces.
• a total breakdown of the barrier leads to a fall in the electrical resistance of the barrier typically by a factor of 10 to 100 (for example a resistance drop of 5KiloOhms to a few hundred or tens of Ohms).
• pinhole primers hot spots
• the hot spots 56 thus formed can thus play the same role as the pinholes 36 of the CCP structure (FIG. 4).
• a more general way of creating layers generating strong lateral inhomogeneities of current consists in using alloys or composites of at least two materials having very different electrical conductivities and tending to demix. This is the case, for example, of certain semiconductor / metal alloys such as GeCu which, during annealing, tend to form conductive Cu 3 Ge clusters in a semiconductor Ge matrix. This is also the case of some diluted semiconductors (for example ZnOCo) tending to form clusters rich in Co in a matrix of ZnO. This is also the case for materials such as GeMn which tend to form conducting nanochannels of alloys rich in Mn, parallel to the direction of growth, embedded in a weakly conductive semiconductive matrix.
• semiconductor / metal alloys such as GeCu which, during annealing, tend to form conductive Cu 3 Ge clusters in a semiconductor Ge matrix.
• some diluted semiconductors for example ZnOCo
• materials such as GeMn which tend to form conducting nanochannels of alloys rich in Mn, parallel to the direction
• the inhomogeneity of the current can be created by performing top-down techniques of microelectronics or nanoindentation nanocontacts of size and position distributed at the base or at the top of the magnetoresistive pillar in the immediate vicinity of the surface of the oscillating layer. It can be seen that in this case, the means capable of generating current inhomogeneities at the surface of the oscillating layer are not created by a layer (CCP, layer with hot spots, etc.) but by a plurality of nanocontacts on the surface of the oscillating layer.
• the electrons may or may not have the capacity to maintain their spin during the crossing of these means.
• the electrons may keep a significant spin polarization at the crossing of the layer.
• the spin polarization is maintained as the electrons pass through this laterally inhomogeneous electrically conductive layer then this layer may be located between the reference layer and the oscillating layer.
• this layer In cases where the spin polarization of the electrons is destroyed during the crossing of this inhomogeneous lateral conductivity layer (case of GeMn, ZnOCo, CuGe layers), then this layer must be located outside the zone giving rise to the magnetoresistance of the structure, that is to say the zone comprising all or part of the oscillating layer, the spacer and all or part of the reference layer so as not to degrade the magnetoresistance signal of the structure.
• FIG. 6 schematically represents a third embodiment of an oscillator 40 according to FIG. the invention wherein the layer 42 with high lateral conductivity inhomogeneities is outside the area 48 giving rise to the magnetoresistance.
• the oscillator 40 comprises a magnetic stack E successively including:
• a ferromagnetic reference layer 45 having a fixed magnetization direction (here located in the plane of the reference layer), called a trapped layer;
• a non-magnetic (non-magnetic) decoupling layer 44 conventionally known as a spacer, for example an oxide
• This stack E formed by the set of layers above is inserted between two electrodes 41 and 46 of current supply for circulating an electric current of intensity I perpendicular to the plane of the stack of magnetic layers.
• the two electrodes 41 and 46 form current supply means able to circulate a current of electrons perpendicular to the plane of the magnetic stack E.
• the layer 42 of high lateral conductivity inhomogeneity is placed not between the reference layer 45 and the oscillating layer 43 but at another location, here at the back of the oscillating layer 43 is that is to say on the side of the oscillating layer 43 opposite the spacer 44 between the oscillating layer 43 and the reference layer 45.
• the layer 42 of high lateral conductivity inhomogeneity is placed not between the reference layer 45 and the oscillating layer 43 but at another location, here at the back of the oscillating layer 43 is that is to say on the side of the oscillating layer 43 opposite the spacer 44 between the oscillating layer 43 and the reference layer 45.
• nanocontacts on the back of the oscillating layer 45 or integrate the layer of strong lateral inhomogeneity of conductivity inside the oscillating layer.
• the electrons it is not necessary for the electrons to retain a large part of their spin polarization while passing through the layer 42 of high lateral conductivity inhomogeneity because they then repolarize while passing through the oscillating layer 43 or the layer. 45 (depending on the side on which was placed the layer 42 with high conductivity inhomogeneity) which allows the magnetoresistance to remain important.
• the spacer 44 separating the oscillating layer 43 from the reference layer 45 must ensure good magnetic decoupling between the oscillating layer 43 and the reference layer 45. This spacer 44 must retain a large part of the polarization spin electrons so that the magnetoresistance of the oscillator 40 is important.
• FIG. 7 schematically illustrates a fourth embodiment of an oscillator 100 according to the invention.
• Oscillator 100 comprises a magnetic stack E including:
• An antiferromagnetic trapping layer 107 An antiferromagnetic trapping layer 107;
• a ferromagnetic reference layer 106 with a fixed magnetization direction here located in the plane of the reference layer, called a trapped layer; this layer 106 is here by way of example a synthetic antiferromagnetic layer that is to say consisting of two ferromagnetic layers 106C and 106A antiparallelly coupled through a thin antiparallel coupling layer 106B; the reference layer 106 is here trapped by the antiferromagnetic layer 107;
• This stack E formed by the set of layers above is inserted between two electrodes 101 and 108 of current supply allowing circulating an electric current of intensity I perpendicular to the plane of the stack of magnetic layers.
• the two electrodes 101 and 108 form current supply means capable of circulating a current of electrons perpendicular to the plane of the magnetic stack.
• the layer 104 with high lateral conductivity inhomogeneity comprises on each side (it will be noted that this could only on one side or the other) one or more non-magnetic metal layers (for example copper) 103 and 105 respectively separating the layer 104 with a high conductivity inhomogeneity of the oscillating ferromagnetic 102 or reference 106 layers.
• These additional non-magnetic layers 103 and 105 make it possible to increase the magnetic decoupling between the oscillating and reference layers and can It will be appreciated that the oscillating layer 102 may also be formed of a combination of magnetic layers such as, for example, CoFe1 nm / NiFe 2nm.
• FIG. 9 schematically illustrates a fifth embodiment of an oscillator 200 according to the invention.
• Oscillator 200 comprises a magnetic stack E successively including:
• a first magnetic stack E1 successively comprising:
• second magnetic stack E2 successively comprising: o an oscillating ferromagnetic layer 204 of variable magnetization direction and having a magnetic configuration in accordance with the explanations given with reference to FIGS. 1 to 3
• ferromagnetic reference layer 202 of fixed magnetization direction.
• This stack E formed by the set of layers above is inserted between two electrodes 201 and 209 of current supply for circulating an electric current of intensity I perpendicularly to the plane of the stack of magnetic layers.
• the two electrodes 201 and 209 form current supply means able to circulate a current of electrons perpendicular to the plane of the magnetic stack E.
• the structure 200 comprises several oscillating layers (two in this case) and several layers with high lateral inhomogeneity conductivity (two in this case).
• the two oscillating ferromagnetic layers 206 and 204 are separated by the spacer 205 for reducing the magnetic coupling between the two oscillating ferromagnetic layers.
• the two layers with high conductivity lateral inhomogeneity 207 and 203 each respectively separate an oscillating ferromagnetic layer 206 and 204 from a ferromagnetic reference layer 208 and 202.
• the magnetoresistance comes from the two stacks E1 and E2.
• the layers with high conductivity inhomogeneity 207 and 203 retain the spin of the electrons so that the magnetoresistance is important.
• one of the oscillating layers 206 or 204 may be replaced by a synthetic antiferromagnetic layer consisting of two ferromagnetic soft layers separated by an antiferromagnetic coupling layer (for example ruthenium with a thickness of 0.6 nm as is well known to the human being). job).
• a synthetic antiferromagnetic layer consisting of two ferromagnetic soft layers separated by an antiferromagnetic coupling layer (for example ruthenium with a thickness of 0.6 nm as is well known to the human being). job).
• This sixth embodiment is illustrated in FIG. 10; in this case, the oscillator 300 comprises a magnetic stack E successively including:
• a first magnetic stack E1 successively comprising:
• a synthetic antiferromagnetic oscillating layer 312 having a magnetic configuration in accordance with the explanations given with reference to FIGS. 1 to 3 and comprising two ferromagnetic soft layers 308 and 306 separated by an antiferromagnetic coupling layer 307;
• a second magnetic stack E2 successively comprising:
• ferromagnetic reference layer 302 of fixed magnetization direction a ferromagnetic reference layer 302 of fixed magnetization direction.
• This stack E formed by the set of layers above is inserted between two electrodes 301 and 31 1 of current supply for circulating an electric current of intensity I perpendicularly to the plane of the stack of magnetic layers.
• the two electrodes 301 and 31 1 form current supply means capable of circulating a current of electrons perpendicular to the plane of the magnetic stack E.
• the thicknesses of the two ferromagnetic layers 308 and 306 constituting the resulting oscillating layer may then be advantageous. advantageously unsymmetrized so that the effect of the resulting amperene field acting on the ferromagnetic layer 308 in contact with the layer with high lateral conductivity inhomogeneity 309 has the right sign for the oscillations to be triggered in the ferromagnetic layer in contact with the layer to strong lateral inhomogeneity of current.
• the vortices that form in both layers 306 and 308 have opposite chirality because these two layers are antiferromagnetically coupled through layer 307 (i.e., a vortex rotates clockwise and the other turns in the opposite direction).
• the Amperian field created by the current enhances the formation of one of these two vortices but tends to prevent the formation of the other. So if the thicknesses of the two layers 306 and 308 are equal, the vortex will hardly form. It is therefore desirable that the layer in which the vortex has the common sense with respect to the amperian field be thicker than that in which the vortex is opposed to the amperian field so that the effect of the "good" layer (ie the layer in which the vortex has good sense compared to the ampere field) prevails.
• FIG. 11 illustrates a seventh embodiment of an oscillator 400 according to the invention.
• the two oscillating layers are each replaced by a synthetic antiferromagnetic layer consisting of two soft ferromagnetic layers separated by an antiferromagnetic coupling layer.
• the oscillator 400 comprises a magnetic stack E successively including:
• a first magnetic stack E1 successively comprising: a ferromagnetic reference layer 412 of fixed magnetization direction
• a synthetic antiferromagnetic oscillating layer 414 having a magnetic configuration in accordance with the explanations given with reference to FIGS. 1 to 3 and comprising two ferromagnetic soft layers 410 and 408 separated by an antiferromagnetic coupling layer 409;
• a second magnetic stack E2 successively comprising:
• a synthetic antiferromagnetic oscillating layer 415 having a magnetic configuration in accordance with the explanations given with reference to FIGS. 1 to 3 and comprising two ferromagnetic soft layers 406 and 404 separated by an antiferromagnetic coupling layer 405;
• ferromagnetic reference layer 402 of fixed magnetization direction.
• This stack E formed by the set of layers above is inserted between two electrodes 401 and 413 of current supply for circulating an electric current of intensity I perpendicular to the plane of the stack of magnetic layers.
• the two electrodes 401 and 413 form current supply means able to circulate a current of electrons perpendicular to the plane of the magnetic stack E.
• the thicknesses of the two ferromagnetic layers (respectively 410 and 408 and 406 and 404) constituting the synthetic antiferromagnetic sandwich may advantageously be unsymmetrized so that the effect of the resulting amperial field acting on the layer ferromagnetic (respectively 410 and 404) in contact with the layer with high lateral inhomogeneity conductivity (respectively 41 1 and 403) has the right sign for the oscillations are triggered in the ferromagnetic layer in contact with the layer with high lateral inhomogeneity of current .
• FIG. 12 illustrates an eighth embodiment of an oscillator 500 according to the invention.
• Oscillator 500 comprises a magnetic stack E successively including:
• a first sub-stack E1 successively comprising:
• a first ferromagnetic reference layer 508 of fixed magnetization direction (here located in the plane of the reference layer), called the trapped layer;
• a first non-magnetic layer 507 (non-magnetic) decoupling conventionally called spacers, for example oxide;
• a second sub-stack E2 successively comprising:
• This stack E formed by the set of layers above is inserted between two electrodes 501 and 509 of current supply for circulating an electric current of intensity I perpendicular to the plane of the stack of magnetic layers.
• the two electrodes 501 and 509 form current supply means able to circulate a current of electrons perpendicular to the plane of the magnetic stack E.
• the two oscillating ferromagnetic layers 506 and 504 are separated by the layer 505 with high inhomogeneity of conductivity.
• One and / or the other of the ferromagnetic layers can be replaced by a synthetic antiferromagnetic layer consisting of two ferromagnetic layers coupled antiferromagnetically.
• the oscillator according to the invention can also operate with two oscillating layers without a reference layer, the only condition being that the magnetic stack comprises at least two magnetic layers, one of which is an oscillating layer with a magnetization of variable direction.
• FIG. 13 represents an oscillator 600 according to a ninth embodiment of the invention.
• the oscillator 600 comprises a magnetic stack E successively including:
• This stack E formed by the set of layers above is inserted between two electrodes 601 and 606 current supply for circulating an electric current of intensity I perpendicularly to the plane of the stack of magnetic layers.
• the two electrodes 601 and 606 form current supply means able to circulate a current of electrons perpendicular to the plane of the magnetic stack E.
• the stack E comprises at least two oscillating layers 604 and 602 but no longer includes a reference layer.
• the electrons polarize by crossing a first oscillating layer and exert a spin transfer on the magnetization of the second oscillating layer through it.
• the oscillations result from a competition between the amperian field created by the current traversing the structure and spin transfer from one oscillating layer to the other.
• the stack E further comprises a spacer 603 consisting of a layer with a high lateral inhomogeneity of conductivity.
• the buffer layer 605 (for example of Ta, NiFeCr or CuN) is a layer intended to promote good growth of the stack E.
• the cover layer 607 is one or a set of functional layers necessary for the preparation and the engraving of the magnetoresistive pillar.
• FIG. 14 represents a tenth embodiment of an oscillator 700 according to the invention which constitutes a variant of the embodiment illustrated in FIG.
• Oscillator 700 comprises a magnetic stack E successively including:
• This stack E formed by the set of layers above is inserted between two electrodes 701 and 709 current supply for circulating an electric current of intensity I perpendicular to the plane of the stack of magnetic layers.
• the two electrodes 701 and 709 form current supply means able to circulate a neck. of electrons perpendicular to the plane of the magnetic stack E.
• the stack E comprises again at least two oscillating layers 706 and 704 but no longer has a reference layer.
• the electrons polarize by crossing an oscillating layer and exert a spin transfer on the magnetization of the other oscillating layer by crossing it.
• the oscillations result from a competition between the amperian field created by the current traversing the structure and spin transfer from one oscillating layer to the other.
• the high conductivity inhomogeneity layers 707 and 703 are located not between the two oscillating layers 706 and 704 but outside the oscillating ferromagnetic layers on the opposite side with respect to the magnetic decoupling spacer 705 separating them; it will be noted that the layers with a high conductivity inhomogeneity 707 and 703 could also be located inside the oscillating layers.
• one or more of the oscillating layers may also be synthetic antiferromagnetic layers.
• the shape of the oscillator according to the invention has a substantially cylindrical symmetry if one seeks to exploit the fundamental excitation signal. But other forms are possible if one seeks to amplify the signals of the harmonics.

Landscapes

• Engineering & Computer Science (AREA)
• Chemical & Material Sciences (AREA)
• Crystallography & Structural Chemistry (AREA)
• Power Engineering (AREA)
• Nanotechnology (AREA)
• Theoretical Computer Science (AREA)
• Hall/Mr Elements (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=43244987

Family Applications (1)

Country Status (4)

Families Citing this family (4)

Family Cites Families (5)

• 2010
  • 2010-06-09 FR FR1054530A patent/FR2961339B1/fr not_active Expired - Fee Related
• 2011
  • 2011-06-09 WO PCT/EP2011/059558 patent/WO2011154477A1/fr active Application Filing
  • 2011-06-09 EP EP11726745.0A patent/EP2580766A1/de not_active Ceased
  • 2011-06-09 US US13/702,413 patent/US8878618B2/en not_active Expired - Fee Related

Non-Patent Citations (2)

Also Published As

Similar Documents

Legal Events