A neuromagnet, a method for manufacturing a neuromagnet, and a neuromagnet sensor

Abstract

The disclosure provides a neural magnetic pole, a neural magnetic pole preparation method and a neural magnetic pole sensor, relates to the technical field of neural engineering, implantable biosensing and micro-electro-mechanical system cross, and the neural magnetic pole comprises a sensing unit and a reference unit; the sensing unit is located above a first area of a substrate and comprises a plurality of first magnetic tunnel junctions connected in series with each other; the first magnetic tunnel junctions are connected with a first interface through contacts and lead wires; the reference unit is located above a second area of the substrate and comprises a plurality of second magnetic tunnel junctions connected in series with each other; the second magnetic tunnel junctions are connected with a second interface through contacts and lead wires; wherein the first area and the second area are of the same shape and are arranged side by side or symmetrically along the axial direction of the substrate; the first magnetic tunnel junctions and the second magnetic tunnel junctions are of the same structure and each comprise a first functional stack, a barrier layer and a second functional stack stacked from bottom to top; and the contacts are located above the second functional stack.

Description

Technical Field

[0001] This disclosure relates to the interdisciplinary fields of neural engineering, implantable biosensing and microelectromechanical systems, and in particular to a neural magnetic pole, a method for preparing a neural magnetic pole, and a neural magnetic pole sensor. Background Technology

[0002] Neuronal action potentials are the basic units of neural information transmission and encoding. High spatiotemporal resolution and high fidelity in vivo detection of these potentials are key to revealing the working mechanisms of neural circuits and developing brain-computer interfaces and neurological disease diagnosis and treatment technologies.

[0003] Traditional electrophysiological recordings can only acquire scalar potential signals, failing to reflect the vector direction information of neural currents. They are also susceptible to the effects of tissue conductivity, electrode interface impedance, and glial scarring, resulting in poor long-term implantation stability and making it difficult to meet the demands of high-fidelity neural signal detection. While macroscopic magnetic detection devices offer high sensitivity, their large size and demanding operating conditions prevent miniaturization and implantable integration, hindering their use for in-situ measurements in deep brain regions and at the cellular level. Existing giant magnetoresistive (GMR) implantable neural poles suffer from low magnetoresistive ratios and insufficient sensitivity due to the intrinsic properties of the materials, making it difficult to achieve the picotesla level weak magnetic detection requirements for single-neuron action potentials. Furthermore, their single-ended structures often have limited ability to suppress environmental common-mode noise, resulting in low in vivo signal-to-noise ratios. Simultaneously, the ferromagnetic materials contained in magnetoresistive devices pose a risk of ion release in physiological environments. The devices have simple passivation protection, insufficient biocompatibility, and long-term implantation can easily induce inflammatory responses and tissue damage, failing to guarantee stable and reliable in vivo recordings.

[0004] It is evident that there are significant bottlenecks in related neural detection technologies: traditional electrophysiology cannot provide vector information and lacks long-term stability; macroscopic magnetic detection cannot achieve implantable microscale measurements; giant magnetoresistive magnetic electrodes have limited sensitivity and noise suppression capabilities, making it difficult to detect magnetic signals of single neuronal action potentials; at the same time, the biocompatibility and long-term implantation reliability of related devices cannot meet the requirements for in vivo applications. Summary of the Invention

[0005] In view of this, the present disclosure provides a neural magnetic pole, a method for preparing a neural magnetic pole, and a neural magnetic pole sensor.

[0006] A first aspect of this disclosure provides a neural magnetic pole, comprising: a sensing unit located above a first region of a substrate, including a plurality of first magnetic tunnel junctions connected in series, the first magnetic tunnel junctions being connected to a first interface via contacts and leads; and a reference unit located above a second region of the substrate, including a plurality of second magnetic tunnel junctions connected in series, the second magnetic tunnel junctions being connected to a second interface via contacts and leads; wherein the first region and the second region have the same shape and are arranged side-by-side or symmetrically along the axial direction of the substrate, the first magnetic tunnel junctions and the second magnetic tunnel junctions have the same structure, each including a first functional stack, a barrier layer and a second functional stack stacked from bottom to top, and the contacts being located above the second functional stack.

[0007] According to embodiments of this disclosure, the first functional stack includes a bottom electrode layer, a first antiferromagnetic layer, and a synthetic antiferromagnetic layer stacked from bottom to top; the second functional stack includes a free layer, a second antiferromagnetic layer, and a top electrode layer stacked from bottom to top.

[0008] According to embodiments of this disclosure, the materials of the bottom electrode layer and the top electrode layer are one or more of Ta, Au, Ag, Al, Cu, Pt, W, Ti, Mo, TaN, and TiN; the materials of the free layer are one or more of NiFe, CoFe, CoFeB, FeGaB, Co, Fe, NiFeCo, CoNbZr, CoFeB / NiFe, and CoFe / Ru / CoFe; the materials of the barrier layer are one or more of Al2O3, MgO, MgAl2O4, MgGa2O4, and MgZnO; the materials of the first antiferromagnetic layer and the second antiferromagnetic layer are one or more of IrMn, PtMn, FeMn, NiMn, and PtPbMn; the ferromagnetic layer material of the synthesized antiferromagnetic layer is one or more of CoFe, CoFeB, NiFe, and Co; and the nonmagnetic coupling layer material of the synthesized antiferromagnetic layer is one or more of Ru and Ta.

[0009] According to embodiments of the present disclosure, the neural magnetic pole further includes an insulating layer located above the substrate and covering the sides of the first magnetic tunnel junction and the second magnetic tunnel junction.

[0010] According to embodiments of this disclosure, the insulating layer is made of one or more of SiO2, Al2O3, HfO2, and parylene.

[0011] According to embodiments of this disclosure, the neural magnetic pole further includes a protective layer located above the insulating layer and covering the contacts and leads.

[0012] According to embodiments of this disclosure, the protective layer includes a first passivation layer and a second passivation layer, wherein the material of the first passivation layer is SiO2 and the material of the second passivation layer is Si3N4.

[0013] According to embodiments of this disclosure, the protective layer is made of one or more of Al2O3, HfO2, and parylene.

[0014] According to embodiments of this disclosure, the neural magnetic pole further includes a bio-coating located above the protective layer.

[0015] According to embodiments of this disclosure, the material of the bio-coating is one or more of hydrophilic proteins, heparin, and phospholipid monomolecules.

[0016] According to embodiments of this disclosure, the materials of the first interface, the second interface, the contacts, and the leads are one or more of Cr / Au, Ti / Au, Pt / Au, and Ta / Au.

[0017] According to embodiments of this disclosure, the substrate material is one or more of monocrystalline silicon, silicon oxide, silicon-on-insulator, flexible polyimide, glass, and sapphire.

[0018] According to embodiments of this disclosure, the number of first magnetic tunnel junctions and second magnetic tunnel junctions is equal and both are even; the even number of first magnetic tunnel junctions share a first functional stack in pairs, and any two first magnetic tunnel junctions that do not share the first functional stack are connected through contacts on the second functional stack to form a sensing unit; wherein, the contacts above the first magnetic tunnel junctions at both ends of the sensing unit are respectively connected to two first interfaces through leads; the even number of second magnetic tunnel junctions share a first functional stack in pairs, and any two second magnetic tunnel junctions that do not share the first functional stack are connected through contacts on the second functional stack to form a reference unit; wherein, the contacts above the second magnetic tunnel junctions at both ends of the reference unit are respectively connected to two second interfaces through leads.

[0019] A second aspect of this disclosure provides a method for fabricating a neural magnetic pole. This method includes: sequentially depositing a first functional stack, a barrier layer, and a second functional stack on a substrate surface; performing photolithographic patterning on a first region and a second region above the substrate, etching to the substrate surface to define a first functional stack structure for a sensing unit and a first functional stack structure for a reference unit, respectively; and performing photolithographic patterning on the surfaces of the second functional stacks in the first and second regions above the substrate, etching to the surface of the first functional stack to define a first magnetic tunnel junction for the sensing unit, respectively. The second magnetic tunnel junction structure of the structure and reference unit; an insulating layer is deposited on the substrate surface to cover the first and second magnetic tunnel junctions; photolithographic patterning is performed on the surface of the insulating layer to open vias on the top of the first and second magnetic tunnel junctions respectively; a metal layer is deposited on the surface of the insulating layer to connect multiple first magnetic tunnel junctions in series to form a sensing unit, and multiple second magnetic tunnel junctions are connected in series to form a reference unit, and the leads of the sensing unit and the leads of the reference unit are defined respectively; a protective layer is deposited on the surface of the metal layer; the protective layer covering the lead surface is etched to open part of the pad area to form the first interface and the second interface respectively.

[0020] According to embodiments of this disclosure, before etching the protective layer covering the lead surface to open a portion of the pad area and forming the first interface and the second interface, the neural magnetic pole preparation method further includes: thinning the substrate.

[0021] According to an embodiment of this disclosure, after etching the protective layer covering the lead surface to open a portion of the pad area and forming a first interface and a second interface respectively, the neural magnetic pole preparation method further includes: performing overall shape processing to obtain a probe-shaped neural magnetic pole.

[0022] A third aspect of this disclosure provides a neural magnetic pole sensor, comprising: the aforementioned neural magnetic pole, and an interface circuit connected to a first interface and a second interface of the neural magnetic pole; wherein the interface circuit comprises: a sensing signal processing module connected to the first interface, for processing the signal output by the sensing unit of the neural magnetic pole to obtain a sensing signal; a reference signal processing module connected to the second interface, for processing the signal output by the reference unit of the neural magnetic pole to obtain a reference signal; a differential amplification module connected to the sensing signal processing module and the reference signal processing module, for performing differential operations on the sensing signal and the reference signal to obtain a differential output signal; and a data acquisition module connected to the differential amplification module, for acquiring and storing the differential output signal.

[0023] According to embodiments of this disclosure: the sensing signal processing module includes a first Wheatstone bridge, a first amplifier, and a first filter connected in sequence; wherein, the first Wheatstone bridge is used to convert the resistance change of the sensing unit into a first differential voltage signal, the first amplifier is used to amplify the first differential voltage signal to obtain a first amplified signal, and the first filter is used to filter the first amplified signal to obtain a sensing signal; the reference signal processing module includes a second Wheatstone bridge, a second amplifier, and a second filter connected in sequence; wherein, the second Wheatstone bridge is used to convert the resistance change of the reference unit into a second differential voltage signal, the second amplifier is used to amplify the second differential voltage signal to obtain a second amplified signal, and the second filter is used to filter the second amplified signal to obtain a reference signal; wherein, the first amplifier is connected to a first feedback resistor, and the second amplifier is connected to a second feedback resistor, and by adjusting the resistance values ​​of the first feedback resistor and the second feedback resistor respectively, gain matching compensation is performed on the sensitivity difference between the sensing unit and the reference unit, so that the response amplitude of the first amplified signal and the second amplified signal to the same magnetic field environment is consistent.

[0024] According to embodiments of this disclosure, the first amplifier, the second amplifier, and the differential amplifier module are instrumentation amplifiers, programmable gain amplifiers, or dedicated integrated circuit chips for low-noise magnetic sensors.

[0025] According to embodiments of this disclosure, the data acquisition module is a data acquisition chip that integrates an amplifier and an analog-to-digital converter.

[0026] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0027] The above and other objects, features, and advantages of this disclosure will become clearer from the following description of embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0028] Figure 1 This is a schematic diagram of the neural magnetic poles according to an embodiment of the present disclosure.

[0029] Figure 2 This is a flowchart of a method for preparing neural magnetic poles according to an embodiment of the present disclosure.

[0030] Figure 3 This is a schematic diagram of film preparation according to an embodiment of the present disclosure.

[0031] Figure 4 This is a schematic diagram of etching a first functional stack according to an embodiment of the present disclosure.

[0032] Figure 5This is a schematic diagram of etching a second functional stack according to an embodiment of the present disclosure.

[0033] Figure 6 This is a schematic diagram of a deposited insulating layer according to an embodiment of the present disclosure.

[0034] Figure 7 This is a schematic diagram of an etched window according to an embodiment of the present disclosure.

[0035] Figure 8 This is a schematic diagram of a deposited metal layer according to an embodiment of the present disclosure.

[0036] Figure 9 This is a schematic diagram of a deposited protective layer according to an embodiment of the present disclosure.

[0037] Figure 10 This is a schematic diagram of substrate thinning according to an embodiment of the present disclosure.

[0038] Figure 11 This is a schematic diagram of etched pads according to an embodiment of the present disclosure.

[0039] Figure 12 This is a shape release diagram according to an embodiment of the present disclosure.

[0040] Figure 13 This is an interface circuit framework diagram according to an embodiment of the present disclosure.

[0041] Symbol explanation: 10-sensing unit; 20-reference unit; 30-first interface; 40-second interface; 1-substrate; 2-first functional stack; 3-barrier layer; 4-second functional stack; 5-metal layer; 6-insulating layer; 7-protective layer. Detailed Implementation

[0042] The embodiments of this disclosure will now be described with reference to the accompanying drawings. Various details of the embodiments of this disclosure are included to aid understanding and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0043] In the technical solutions disclosed herein, the collection, storage, use, processing, transmission, provision, disclosure, and application of data (including but not limited to user personal information) comply with the provisions of relevant laws and regulations, necessary confidentiality measures have been taken, and they do not violate public order and good morals.

[0044] For a long time, in vivo neural signal detection has mainly relied on traditional electrophysiological recording techniques. Electrophysiological methods acquire signals by detecting changes in extracellular potential and have a mature application foundation, but they have inherent limitations: electrical signals are scalar signals and cannot reflect the spatial direction and vector information of neural currents; detection results are easily affected by uneven conductivity of biological tissues, fluctuations in electrode-tissue interface impedance, and the configuration of reference electrodes, limiting spatial resolution and signal fidelity; during long-term implantation, glial scars easily form around the electrodes, leading to increased interface impedance and signal attenuation, making it difficult to meet the requirements for long-term stable in vivo recording.

[0045] To overcome the inherent limitations of traditional electrophysiological recordings, the detection of magnetic signals using the weak magnetic fields generated by neuronal activity has become an important development direction. Neuromagnetic signals possess advantages such as uniform penetration through biological tissues, no need for reference electrodes, immunity to tissue conductivity interference, and the natural carrying of magnetic field vector information. This enables the analysis of the direction of neural currents and spatial topology, representing a significant supplement and breakthrough to electrophysiological techniques.

[0046] The main methods for detecting related neuromagnetic signals are superconducting quantum interference devices and optically pumped magnetometers. Although they can achieve high-sensitivity magnetic field measurement, they have problems such as large size, harsh working conditions, and the need for low temperature or high-level magnetic shielding. They are only suitable for macroscopic magnetoencephalography detection outside the skull, and cannot be miniaturized or implanted, making it difficult to achieve in situ in vivo detection of deep brain regions and cells.

[0047] To achieve implantable, microscale neural magnetic signal detection, researchers combined magnetoresistive sensors with neural microprobes, proposing the concept of neural magnetic poles, analogous to neural electrodes. Existing neural magnetic poles mostly employ giant magnetoresistive (GMR) devices as sensing units, and have achieved preliminary recording of neural activity magnetic signals in in vitro and in vivo experiments, verifying the feasibility of magnetic pole detection. However, GMR devices have a low magnetoresistive ratio, limiting their ability to detect extremely weak magnetic signals corresponding to neuronal action potentials. Furthermore, traditional magnetic electrodes often employ single-ended detection structures, which cannot effectively suppress common-mode noise such as environmental magnetic field fluctuations and power frequency interference, resulting in low signal-to-noise ratios and poor signal stability in complex in vivo environments.

[0048] Furthermore, magnetoresistive sensors contain ferromagnetic materials such as Co, Fe, and Ni, which pose a risk of releasing metal ions in physiological environments, potentially triggering cytotoxicity and inflammatory responses in brain tissue. Existing devices generally employ simple single-layer passivation protection, which is insufficient for isolation and fails to meet the biocompatibility and stability requirements for long-term implantation. Traditional probe substrates are relatively thick, resulting in significant implantation trauma, further exacerbating tissue reactions and affecting detection accuracy and long-term reliability.

[0049] Therefore, there is an urgent need for an implantable neural magnetic signal detection technology that is highly sensitive, low-noise, biocompatible, and capable of long-term stable operation. Tunneling magnetoresistance (TMR)-based magnetic electrodes possess significant advantages such as high magnetoresistive ratio, high sensitivity, low noise, room temperature operation, and MEMS process compatibility, making them an ideal solution for detecting picotesla-level weak magnetic signals. This disclosure provides a neural magnetic pole, a method for fabricating the neural magnetic pole, and a neural magnetic pole sensor. Using TMR as the sensing element, it leverages its advantages of high magnetoresistive ratio, high sensitivity, and low noise to enhance the weak magnetic detection capability of the neural magnetic pole. By designing a spatial differential sensing structure, it improves detection sensitivity and noise suppression capabilities. Furthermore, by optimizing the biocompatible passivation layer, it enhances long-term implantation stability, providing a novel technical pathway for high-fidelity in vivo magnetic recording of neuronal action potentials.

[0050] Figure 1 This is a schematic diagram of a neural magnetic pole according to an embodiment of the present disclosure, wherein (a) is a partial cross-sectional schematic diagram of the neural magnetic pole, and (b) is a planar schematic diagram of the neural magnetic pole.

[0051] like Figure 1 As shown, the neural magnetic pole includes a sensing unit 10 and a reference unit 20. The sensing unit 10 is located above a first region of the substrate 1 and includes multiple first magnetic tunnel junctions connected in series. The first magnetic tunnel junctions are connected to a first interface 30 via contacts and leads. The reference unit 20 is located above a second region of the substrate and includes multiple second magnetic tunnel junctions connected in series. The second magnetic tunnel junctions are connected to a second interface 40 via contacts and leads. The first and second regions have the same shape and are arranged side-by-side or symmetrically along the axial direction of the substrate. The first and second magnetic tunnel junctions have the same structure, both including a first functional stack 2, a barrier layer 3, and a second functional stack 4 stacked from bottom to top. The contacts are located above the second functional stack 4.

[0052] This disclosure employs a tunneling magnetoresistive (TMR) device as the core sensing element, integrated into an implantable probe body, to capture weak magnetic signals generated by neuronal action potentials, thereby achieving magnetic detection of neural signals. Furthermore, this disclosure fabricates the TMR device into a differential structure for capturing and identifying weak magnetic signals from neurons, overcoming the limitation of traditional electrophysiological detection methods that cannot acquire neural current vector information. The differential structure used has a baseline spacing on the order of millimeters, which can effectively suppress far-field common-mode noise such as power frequency interference and slow drift of the geomagnetic field while preserving the local magnetic field of the target nerve. The common-mode suppression ratio is significantly better than that of single-ended detection structures.

[0053] In this embodiment, the neural magnetic poles are probe-shaped, which can be a long, straight rod, a forked type, or a U-shaped differential layout. When using a long, straight rod-shaped probe, the first region is the tip region of the probe, and the second region is the distal region of the probe. The first and second regions can be arranged side by side vertically along the longitudinal axis of the substrate. When using a forked probe or a U-shaped differential layout probe, the first and second regions can be arranged symmetrically horizontally along the longitudinal axis of the substrate.

[0054] In this embodiment, the number and size ratio of the sensing unit and the reference unit can be flexibly adjusted according to the probe size, and the sensor baseline spacing can also be adaptively changed, all of which can achieve spatial differential noise reduction effect. For example, the axial spacing between the sensing unit and the reference unit is 5mm.

[0055] In another embodiment of this disclosure, in addition to using multiple magnetic tunnel junctions (MTJs) connected in series, the sensing unit and the reference unit can also use a single large-size TMR device to achieve differential detection, or use a Wheatstone bridge integrated TMR chip to replace the discrete MTJ structure.

[0056] The intensity range of neuromagnetic signals is from tens of picoteslas (pT) to nanoteslas (nT). Conventional GMR devices struggle to achieve this level of detection sensitivity, while TMR devices can achieve a detection resolution better than 100 pT / √Hz at 1 kHz, which is sufficient for the effective resolution of action potential magnetic signals from single neurons. The implantable differential neuromagnetic electrode based on TMR provided in this disclosure utilizes a high-sensitivity TMR sensing unit to enhance weak magnetic detection capabilities and significantly suppresses common-mode noise and improves the detection signal-to-noise ratio through a spatial differential structure, enabling high-quality detection of neuronal action potential magnetic signals.

[0057] According to embodiments of this disclosure, the first functional stack 2 includes a bottom electrode layer, a first antiferromagnetic layer, and a synthetic antiferromagnetic layer stacked from bottom to top; the second functional stack 4 includes a free layer, a second antiferromagnetic layer, and a top electrode layer stacked from bottom to top.

[0058] In this embodiment, the bottom electrode layer and the top electrode layer are made of one or more of Ta, Au, Ag, Al, Cu, Pt, W, Ti, Mo, TaN, and TiN; the free layer is made of one or more of NiFe, CoFe, CoFeB, FeGaB, Co, Fe, NiFeCo, CoNbZr, CoFeB / NiFe, and CoFe / Ru / CoFe; the barrier layer is made of one or more of Al2O3, MgO, MgAl2O4, MgGa2O4, and MgZnO; the first antiferromagnetic layer and the second antiferromagnetic layer are made of one or more of IrMn, PtMn, FeMn, NiMn, and PtPbMn; the ferromagnetic layer material of the synthesized antiferromagnetic layer is one or more of CoFe, CoFeB, NiFe, and Co; and the nonmagnetic coupling layer material of the synthesized antiferromagnetic layer is one or more of Ru and Ta.

[0059] The synthesized antiferromagnetic layer adopts a sandwich structure or a multilayer composite structure in the form of ferromagnetic layer / nonmagnetic coupling layer / ferromagnetic layer. For example, the material for synthesizing the antiferromagnetic layer can be CoFe / Ru / CoFeB, Co / Ru / CoFeB, CoFeB / Ru / CoFeB, CoFe / Ru / CoFe, CoFeB / Ru / CoFe, or CoFeB / Co / Ru / CoFe / CoFeB.

[0060] For example, the bottom electrode layer is made of Ta / Ru / Ta / Ru / Ta / Ru; the first antiferromagnetic layer is made of PtMn; the synthetic antiferromagnetic layer is made of CoFe / Ru / CoFeB; the barrier layer is made of MgO; the free layer is made of CoFeB / NiFe; the second antiferromagnetic layer is made of IrMn; and the top electrode layer is made of Ru / Ta / Ru.

[0061] The bottom electrode layer uses a multilayer composite film of alternating Ta (tantalum) and Ru (ruthenium) to extract the tunneling current. The first antiferromagnetic layer uses PtMn (platinum-manganese alloy) as the bottom pinning source of the device. A strong exchange bias field is established through high-temperature annealing to fix the magnetization direction of the synthesized antiferromagnetic layer above, providing a stable reference. The synthesized antiferromagnetic layer uses a composite film of CoFe (cobalt-iron alloy), Ru (ruthenium), and CoFeB (cobalt-iron-boron alloy) as the reference layer of the device. The barrier layer uses MgO (magnesium oxide) as the insulating dielectric layer through which electrons tunnel, and uses the coherent spin filtering effect to achieve magnetoresistance changes. The free layer uses a composite film of CoFeB (cobalt-iron-boron alloy) and NiFe (nickel-iron alloy) as the core sensitive unit for sensing changes in the external magnetic field. Its magnetization direction rotates freely with the weak external magnetic field, converting the magnetic field signal into a change in resistance. The second antiferromagnetic layer uses IrMn (iridium-manganese alloy) as the top antiferromagnetic bias source, applying a weak exchange bias field to the free layer to achieve "soft pinning," eliminating the nonlinearity of the hysteresis loop near zero field and ensuring a linear relationship between the output and the magnetic field. The top electrode layer uses a Ru (ruthenium), Ta (tantalum), and Ru (ruthenium) multilayer composite film to extract the tunneling current.

[0062] Furthermore, to obtain better linearity and lower noise, a Ru layer can be inserted between the NiFe free layer and the IrMn second antiferromagnetic layer to precisely adjust the intensity of the exchange bias field and avoid excessive IrMn soft pinning that would cause a decrease in the sensitivity of the free layer.

[0063] According to an embodiment of this disclosure, the aforementioned neural magnetic pole further includes an insulating layer 6 located above the substrate 1 and covering the sides of the first magnetic tunnel junction and the second magnetic tunnel junction.

[0064] In this embodiment, the insulating layer 6 is made of one or more of SiO2, Al2O3, HfO2 and parylene.

[0065] For example, the SiO2 insulating layer has a thickness of 200 nm. This insulating layer fully covers the sidewalls of the MTJ junction region, effectively isolating the MTJ sidewalls from oxidation.

[0066] According to embodiments of this disclosure, the aforementioned neural magnetic poles further include: a protective layer 7, located above the insulating layer 6, and covering the contacts and leads.

[0067] In this embodiment, the protective layer 7 includes a first passivation layer and a second passivation layer. The first passivation layer is made of SiO2, and the second passivation layer is made of Si3N4. This embodiment employs a SiO2 / Si3N4 bilayer passivation structure to improve biocompatibility and long-term implantation stability, thereby overcoming the shortcomings of related technologies in terms of sensitivity, noise suppression, biocompatibility, and long-term reliability, and achieving high-quality, highly stable in vivo detection of neuronal action potential magnetic signals.

[0068] In another embodiment of this disclosure, the material of the protective layer 7 is one or more of Al2O3, HfO2 and parylene-C.

[0069] According to embodiments of this disclosure, the aforementioned neural magnetic poles further include a bio-coating located above the protective layer 7.

[0070] In this embodiment of the disclosure, the bio-coating material is one or more of hydrophilic proteins, heparin, and phospholipid monomolecules.

[0071] The embodiments disclosed herein further reduce the immune response after probe implantation by modifying the surface of the protective layer 7 with a biological coating such as a hydrophilic protein, heparin, or phospholipid monolayer.

[0072] In another embodiment of this disclosure, a Parylene-C fully encapsulated structure can be used to replace part of the inorganic passivation layer, thereby improving device flexibility while ensuring biocompatibility.

[0073] According to embodiments of this disclosure, the materials of the first interface, the second interface, the contacts, and the leads are one or more of Cr / Au, Ti / Au, Pt / Au, and Ta / Au.

[0074] In this embodiment, the interface, contacts, and leads are all metal layers 5. The metal layers 5 adopt a double-layer metal composite structure such as Cr / Au, Ti / Au, Pt / Au, and Ta / Au, which can achieve low-resistance interconnection.

[0075] According to embodiments of this disclosure, the substrate 1 is made of one or more of the following materials: single-crystal silicon, silicon oxide, silicon-on-insulator (SOI), flexible polyimide (PI), glass, and sapphire.

[0076] In this embodiment, the substrate 1 can be made of thinned single-crystal silicon, silicon oxide, SOI wafer, or directly made of flexible PI substrate, ultra-thin glass or sapphire substrate to prepare the neural magnetic electrode, further reducing the mechanical modulus and minimizing implantation damage.

[0077] According to embodiments of this disclosure, the number of first magnetic tunnel junctions and second magnetic tunnel junctions is equal and both are even.

[0078] An even number of first magnetic tunnel junctions share a first functional stack 2 in pairs, and any two first magnetic tunnel junctions that do not share the first functional stack 2 are connected through contacts on the second functional stack 4 to form a sensing unit 10; wherein, the contacts above the first magnetic tunnel junctions at both ends of the sensing unit 10 are connected to two first interfaces 30 respectively through leads.

[0079] An even number of second magnetic tunnel junctions share the first functional stack 2 in pairs, and any two second magnetic tunnel junctions that do not share the first functional stack 2 are connected through contacts on the second functional stack 4 to form a reference unit 20; wherein, the contacts above the second magnetic tunnel junctions at both ends of the reference unit 20 are connected to two second interfaces 40 respectively through leads.

[0080] The neural magnetic poles provided in this disclosure are fabricated using standard MEMS micro / nano fabrication processes. Figure 2 This is a flowchart of a method for preparing neural magnetic poles according to an embodiment of the present disclosure. Figures 3 to 12 These are schematic diagrams of each operation in the preparation process. The left diagram is a partial cross-sectional view, and the right diagram is a planar view.

[0081] like Figure 2 As shown, the method for preparing neural magnetic poles includes operations S201 to S210.

[0082] Operation S201: Sequentially deposit a first functional stack, a barrier layer, and a second functional stack on the substrate surface.

[0083] For example, firstly, a single-crystal silicon substrate 1 is selected as the carrier, and the wafer surface is subjected to standard cleaning (acetone, alcohol, deionized water ultrasonic cleaning and nitrogen drying) to remove surface contaminants. Subsequently, an MTJ functional film layer and an electrode layer are sequentially deposited on the surface of substrate 1. Specifically, the film stack from bottom to top includes: a bottom electrode layer, a first antiferromagnetic layer, a synthetic antiferromagnetic layer, a barrier layer, a free layer, a second antiferromagnetic layer, and a top electrode layer. To simplify the schematic diagram of the film structure, the portion below the barrier layer 3 is referred to as the first functional stack 2, and the portion above the barrier layer 3 is referred to as the second functional stack 4, as shown below. Figure 3 As shown.

[0084] Operation S202: Photolithographic patterning is performed in the first and second regions above the substrate, and etched to the substrate surface to define the first functional stacked structure of the sensing unit and the first functional stacked structure of the reference unit, respectively.

[0085] For example, photolithographic patterning is first performed, followed by oxygen plasma cleaning, adhesive treatment, and a standard photolithography process. The first functional stack 2 is etched using ion beam etching (IBE), simultaneously defining two independent first functional stack regions at axial positions 5 mm from the probe tip and distal end: one is the first functional stack of the TMR sensing unit (i.e., the first region), and the other is the first functional stack of the TMR reference unit (i.e., the second region), achieving the initial spatial separation layout of the differential structure. The etching process uses the complete disappearance of the characteristic peak lines of the lowest element in the optical emission spectrum (OES) as the endpoint detection mark, ensuring precise termination of etching on the surface of substrate 1, forming a structure as shown in the image. Figure 4 The independent first-function stacked structure is shown.

[0086] It is understood that this disclosure does not limit the specific spacing of reference units; the key is to achieve spatial decoupling of the detection signal from the background noise and to avoid crosstalk between signals from different brain regions.

[0087] Operation S203: Photolithographic patterning is performed on the surface of the second functional stack in the first region and the second region above the substrate, and etching is performed onto the surface of the first functional stack to define the first magnetic tunnel junction structure of the sensing unit and the second magnetic tunnel junction structure of the reference unit, respectively.

[0088] For example, based on the defined first functional stack 2 structure, the patterning and etching of the MTJ junction region continues. Standard photolithography steps are also performed, followed by IBE etching until the barrier layer 3 is completely removed. Since the barrier layer 3 uses MgO, the OES signal is monitored in real time during etching. Etching is stopped immediately when the characteristic peak of Mg is about to disappear to avoid secondary deposition problems and MTJ performance damage caused by over-etching. Independent MTJ junction regions for the tip sensing unit and the far-end reference unit are defined, forming... Figure 5 The MTJ unit structure shown ensures the structural consistency of the two sets of differential units.

[0089] In another embodiment of this disclosure, in addition to IBE ion beam etching, reactive ion beam etching (RIBE) or magnetron sputtering etching can also be used instead of MTJ etching.

[0090] Operation S204: Deposit an insulating layer on the substrate surface to cover the first magnetic tunnel junction and the second magnetic tunnel junction.

[0091] For example, after the MTJ junction region is etched, the resist removal process is completed in the shortest possible time, and the area is immediately transferred to an ICP-CVD device to deposit a 200nm thick SiO2 insulating layer 6. This insulating layer 6 completely covers the sidewalls and top surface of the MTJ junction region, effectively isolating the MTJ sidewalls from oxidation, while protecting the MTJ structure from mechanical or chemical damage in subsequent processes, preventing a decrease in magnetoresistivity, and providing a stable insulating environment for the differential cells, forming a structure like... Figure 6 The insulation protection structure shown.

[0092] In another embodiment of this disclosure, in addition to SiO2, the insulating layer may also be made of biocompatible films such as Al2O3, HfO2, and parylene-C, either alone or in combination.

[0093] Operation S205: Photolithographic patterning is performed on the surface of the insulating layer to create through holes at the top of the first magnetic tunnel junction and the second magnetic tunnel junction, respectively.

[0094] For example, photolithographic patterning is performed on the surface of insulating layer 6 to create independent interconnect vias at the top of the MTJ junction regions of the tip sensing unit and the far-end reference unit, ensuring complete electrical isolation between the two sets of differential units. Reactive ion etching (RIE) is used to create the vias, with the etching depth controlled at approximately 200 nm, providing a reliable channel for subsequent top electrode layer interconnection, forming a... Figure 7 The through-hole structure shown lays the foundation for the independent electrical connection of the differential unit.

[0095] Operation S206: Deposit a metal layer on the surface of the insulating layer, connect multiple first magnetic tunnel junctions in series to form a sensing unit, connect multiple second magnetic tunnel junctions in series to form a reference unit, and define the leads of the sensing unit and the leads of the reference unit respectively.

[0096] For example, a lift-off process is used to fabricate Cr / Au interconnect electrodes. After defining the electrode pattern by photolithography, a 30nm Cr / 200nm Au metal layer 5 is sputtered, and excess metal is removed by acetone lift-off to complete the electrode fabrication. In this step, multiple MTJ junctions in the tip region are sequentially connected in series to form a sensing unit, while multiple MTJ junctions in the distal region are sequentially connected in series to form a reference unit. The two sets of units are independent of each other and have no electrical crossover, ultimately forming a long-baseline spatial differential TMR sensing structure, achieving spatial decoupling between the target neural magnetic signal and environmental common-mode noise, as shown in the figure. Figure 8 The structure shown.

[0097] In another embodiment of this disclosure, the metal layer material, in addition to Cr / Au, can be replaced with structures such as Ti / Au, Pt / Au, or Ta / Au, all of which can achieve low-resistance interconnection.

[0098] Operation S207: Deposit a protective layer on the surface of the metal layer.

[0099] For example, after the electrode fabrication is completed, a 200nm SiO2 / 200nm Si3N4 double passivation layer is deposited using ICP CVD as a protective layer 7, covering the entire device surface. This protective layer 7 can effectively isolate the release of ions from ferromagnetic materials such as Co, Fe, and Ni in the MTJ, reduce local cytotoxicity and brain tissue inflammation after implantation, and improve biocompatibility and long-term implantation stability; at the same time, it optimizes the probe's electrical performance, reduces the risk of leakage, isolates different functional areas, and prevents short circuits or signal crosstalk, forming a protective layer 7. Figure 9 The double-layer protective structure is shown.

[0100] The SiO2 / Si3N4 double-layer passivation system used in this embodiment has the characteristics of high density, resistance to body fluid corrosion, and good metal ion barrier properties. Compared with a single passivation layer, it can significantly reduce the risk of leakage and ion release of the device under physiological environment and improve long-term implantation stability.

[0101] In another embodiment of this disclosure, in addition to SiO2 / Si3N4, the protective layer may also be made of biocompatible films such as Al2O3, HfO2, and parylene-C, either alone or in combination.

[0102] Operation S208: Thinning the substrate.

[0103] For example, after electrode stripping and before etching the passivation bilayer, the back substrate is thinned using a chemical mechanical polishing (CMP) process. To protect the front differential structure and facilitate die bonding and removal, a layer of photoresist is spin-coated onto the passivation bilayer surface as a temporary protective layer. Ultimately, the silicon substrate is thinned to 200 μm, significantly reducing mechanical damage to brain tissue during implantation, forming a structure like... Figure 10 The thin substrate structure shown.

[0104] This embodiment of the invention uses a SiO2 / Si3N4 double passivation layer to cover the magnetic pole surface to block the release of metal ions in the TMR device and reduce tissue inflammation. At the same time, the substrate thinning process reduces implantation trauma and meets the requirements for long-term in vivo biocompatibility.

[0105] Operation S209: Etch the protective layer covering the lead surface to open part of the pad area, forming the first interface and the second interface respectively.

[0106] For example, by removing the temporary protective adhesive layer and opening up part of the solder pad area through an etching process, a structure is formed as shown below. Figure 11 The structure shown provides an interface for subsequent electrical interconnections.

[0107] Operation S210: Perform overall shape processing to obtain probe-shaped neural magnetic poles.

[0108] For example, the probe body is shaped using a deep silicon etching (DRIE) process to release, such as Figure 12 The final implantable differential TMR magnetic electrode probe structure shown completes the device fabrication.

[0109] The device in this embodiment is fabricated using MEMS-compatible technology, enabling mass production and miniaturized integration. The probe size and thickness can be flexibly adjusted according to the implanted brain region, combining high detection performance with low implantation trauma, and is suitable for acute and chronic in vivo neural signal detection.

[0110] To achieve high signal-to-noise ratio detection of weak neural magnetic signals, this disclosure includes a two-stage dynamic matching differential interface circuit designed to compensate for TMR device process mismatch, suppress common-mode noise, and improve the signal-to-noise ratio of neural magnetic signal detection. This interface circuit is connected to the first and second interfaces of the neural magnetic poles, enabling high signal-to-noise ratio detection of weak neural magnetic signals. Figure 13This is an interface circuit framework diagram according to an embodiment of the present disclosure.

[0111] like Figure 13 As shown, the interface circuit includes: a sensing signal processing module, connected to the first interface, for processing the signal output by the sensing unit of the neural magnetic pole to obtain a sensing signal; a reference signal processing module, connected to the second interface, for processing the signal output by the reference unit of the neural magnetic pole to obtain a reference signal; a differential amplification module, connected to the sensing signal processing module and the reference signal processing module, for performing differential operations on the sensing signal and the reference signal to obtain a differential output signal; and a data acquisition module, connected to the differential amplification module, for acquiring and storing the differential output signal.

[0112] According to an embodiment of this disclosure, the sensing signal processing module includes a first Wheatstone bridge, a first amplifier, and a first filter connected in sequence; wherein, the first Wheatstone bridge is used to convert the resistance change of the sensing unit into a first differential voltage signal, the first amplifier is used to amplify the first differential voltage signal to obtain a first amplified signal, and the first filter is used to filter the first amplified signal to obtain a sensing signal.

[0113] Correspondingly, the reference signal processing module includes a second Wheatstone bridge, a second amplifier, and a second filter connected in sequence; wherein, the second Wheatstone bridge is used to convert the resistance change of the reference unit into a second differential voltage signal, the second amplifier is used to amplify the second differential voltage signal to obtain a second amplified signal, and the second filter is used to filter the second amplified signal to obtain a reference signal.

[0114] The first amplifier is connected to a first feedback resistor, and the second amplifier is connected to a second feedback resistor. By adjusting the resistance values ​​of the first and second feedback resistors respectively, gain matching compensation is performed on the sensitivity difference between the sensing unit and the reference unit, ensuring that the response amplitudes of the first and second amplified signals to the same magnetic field environment are consistent. After amplification, the sensing unit and the reference unit have consistent responses to the same magnetic field signal, eliminating sensitivity differences introduced by factors such as process and structure in device fabrication, and maintaining high sensitivity and high precision response.

[0115] In this embodiment, the TMR sensing unit and the TMR reference unit are respectively connected to two independent Wheatstone bridges. Each bridge includes an adjustable sliding potentiometer to compensate for device resistance deviations caused by micro-nano fabrication, achieve DC balance adjustment of the bridge, eliminate static offset voltage, and ensure that the two outputs remain consistent when there is no signal, thus providing a basis for subsequent differential operations.

[0116] In this embodiment, the output signals of the two bridge circuits are respectively fed into two independent instrumentation amplifiers. By adjusting the feedback resistors of the two amplifiers, dynamic sensitivity matching is achieved, ensuring that the response amplitudes of the sensing channel and the reference channel to a uniform ambient magnetic field are strictly equal, thereby maximizing the common-mode rejection ratio. This module can compensate for TMR sensitivity differences caused by film uniformity, etching deviations, and dimensional errors, ensuring the noise reduction effect of the differential structure.

[0117] In this embodiment of the disclosure, the signal after the first stage of amplification is sent to a bandpass filter. The filter adopts a fourth-order Butterworth structure to filter out DC drift, low-frequency 1 / f noise and high-frequency electromagnetic interference, so that the signal bandwidth matches the spectral range of the neuron's action potential, highlighting the effective signal and suppressing out-of-band noise.

[0118] The embodiments disclosed herein can compensate for deviations in MTJ resistance, magnetoresistive ratio, and sensitivity caused by micro-nano fabrication by dynamically matching the differential interface circuit, thereby avoiding a decrease in differential performance due to device mismatch and ensuring stable noise suppression capability even in complex in-situ environments.

[0119] According to embodiments of this disclosure, the first amplifier, the second amplifier, and the differential amplifier module are instrumentation amplifiers, programmable gain amplifiers, or dedicated integrated circuit chips for low-noise magnetic sensors.

[0120] In this embodiment, a two-stage instrumentation amplifier is used. The two filtered signals enter the second-stage instrumentation amplifier for differential operation. Since the two signals have the same response to the environmental common-mode magnetic field, the differential operation can significantly suppress common-mode components such as power frequency interference, geomagnetic field fluctuations, and system electromagnetic noise, retaining only the differential component caused by the magnetic signal of the local neuron at the needle tip, thus achieving weak signal extraction.

[0121] In another embodiment of this disclosure, in addition to employing a two-stage instrumentation amplifier architecture, the differential amplifier circuit can also use a programmable gain amplifier (PGA) or a dedicated low-noise magnetic sensor ASIC chip for signal conditioning. DC matching and sensitivity compensation can be achieved automatically via a digital potentiometer or an on-chip calibration algorithm, in addition to using an adjustable potentiometer. The filtering circuit can be replaced with multi-stage active filtering or switched-capacitor filtering depending on the signal frequency band, both of which can achieve noise suppression.

[0122] According to embodiments of this disclosure, the data acquisition module is a data acquisition chip that integrates an amplifier and an analog-to-digital converter.

[0123] In this embodiment of the disclosure, the differential output signal is further fed into a low-noise preamplifier for gain enhancement, and then analog-to-digital conversion is completed at a sampling rate of not less than 30kHz to realize digital acquisition and storage of the signal.

[0124] The embodiments disclosed herein achieve precise matching of the amplitude-frequency characteristics of the sensing channel and the reference channel through a dual-bridge structure, two-stage independent gain adjustment, and dynamic sensitivity matching. Combined with a spatial differential structure, common-mode noise can be significantly suppressed in complex electromagnetic environments, improving the system signal-to-noise ratio and ultimately achieving stable and high-fidelity detection of picotesla-level neuronal action potential magnetic signals.

[0125] In summary, this disclosure utilizes an implantable neuromagnetic electrode as a carrier to apply a TMR device for neuromagnetic signal detection. A proprietary fabrication process enables integrated device fabrication, while a complementary interface circuit design enhances detection performance. Biocompatibility design ensures long-term in vivo application, ultimately achieving high-sensitivity, high-stability, long-term in vivo detection of weak neuronal magnetic signals. This technology can be applied to brain science research, neural signal recording, and neural circuit analysis. Compared with related technologies, this disclosure offers the following advantages.

[0126] (1) Higher detection sensitivity is achieved by using TMR devices, which can capture weak magnetic signals at the single neuron level. Compared with existing GMR-based magnetic electrodes and traditional electrophysiological detection methods, this disclosure uses high magnetoresistance ratio TMR as the sensitive unit, which significantly improves the magnetic field detection sensitivity and can realize the detection of weak neural magnetic field signals at the picotesla level. It overcomes the limitations of insufficient sensitivity of GMR materials and the inability to identify the magnetic signals of single neuronal action potentials. At the same time, it makes up for the defect of electrophysiological detection that cannot obtain neural current vector information. The signal fidelity and information content are significantly improved.

[0127] (2) The differential structure, combined with the dynamic matching interface circuit, significantly improves the signal-to-noise ratio of in vivo detection. This disclosure adopts a differential structure with spatial separation between the probe tip sensing unit and the remote reference unit, and is equipped with a dual bridge, adjustable gain and two-stage differential amplification interface circuit. This can effectively compensate for the device parameter mismatch caused by the processing technology, significantly suppress common-mode noise such as power frequency magnetic field and environmental electromagnetic interference, and solve the problems of poor magnetic electrode noise suppression capability and insufficient in vivo signal stability of the existing single-ended structure. It can still achieve high signal-to-noise ratio and high reliability signal acquisition in complex physiological environments.

[0128] (3) Dedicated MEMS fabrication process and double-layer passivation design significantly improve biocompatibility and long-term stability. This disclosure ensures the consistency of device performance through precise etching control by spectrometer and integrated process, and effectively blocks the release of metal ions such as Co, Fe and Ni by using SiO2 / Si3N4 double passivation layer. Combined with substrate thinning to reduce implantation mechanical damage, it significantly reduces brain tissue inflammation and glial scar hyperplasia. Compared with the existing single-layer passivation and thick substrate magnetic electrodes, it achieves significant improvement in biocompatibility, long-term implantation reliability and signal stability, and is more suitable for chronic in vivo neural detection.

[0129] Those skilled in the art will understand that the features described in the various embodiments and / or claims of this disclosure can be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in this disclosure. In particular, the features described in the various embodiments and / or claims of this disclosure can be combined and / or combined in various ways without departing from the spirit and teachings of this disclosure. All such combinations and / or combinations fall within the scope of this disclosure.

[0130] The embodiments of this disclosure have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. The scope of this disclosure is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of this disclosure, and all such substitutions and modifications should fall within the scope of this disclosure.

Claims

1. A neural magnetic pole, characterized in that, The neural magnetic poles include: The sensing unit, located above a first region of the substrate, includes a plurality of first magnetic tunnel junctions connected in series, the first magnetic tunnel junctions being connected to a first interface via contacts and leads; The reference cell, located above a second region of the substrate, includes a plurality of second magnetic tunnel junctions connected in series, the second magnetic tunnel junctions being connected to a second interface via contacts and leads; The first region and the second region have the same shape and are arranged side by side or symmetrically along the axial direction of the substrate. The first magnetic tunnel junction and the second magnetic tunnel junction have the same structure, each including a first functional stack, a barrier layer and a second functional stack stacked from bottom to top, and the contact is located above the second functional stack.

2. The neural magnetic pole according to claim 1, characterized in that, The first functional stack includes a bottom electrode layer, a first antiferromagnetic layer and a synthetic antiferromagnetic layer stacked from bottom to top; the second functional stack includes a free layer, a second antiferromagnetic layer and a top electrode layer stacked from bottom to top.

3. The neural magnetic pole according to claim 2, characterized in that, The bottom electrode layer and the top electrode layer are made of one or more of Ta, Au, Ag, Al, Cu, Pt, W, Ti, Mo, TaN, and TiN; the free layer is made of one or more of NiFe, CoFe, CoFeB, FeGaB, Co, Fe, NiFeCo, CoNbZr, CoFeB / NiFe, and CoFe / Ru / CoFe; the barrier layer is made of one or more of Al2O3, MgO, MgAl2O4, MgGa2O4, and MgZnO; the first antiferromagnetic layer and the second antiferromagnetic layer are made of one or more of IrMn, PtMn, FeMn, NiMn, and PtPbMn; the ferromagnetic layer of the synthesized antiferromagnetic layer is made of one or more of CoFe, CoFeB, NiFe, and Co; and the nonmagnetic coupling layer of the synthesized antiferromagnetic layer is made of one or more of Ru and Ta.

4. The neural magnetic pole according to claim 1, characterized in that, The neural magnetic poles also include: An insulating layer is located above the substrate and covers the sides of the first magnetic tunnel junction and the second magnetic tunnel junction.

5. The neural magnetic pole according to claim 4, characterized in that, The insulating layer is made of one or more of SiO2, Al2O3, HfO2, and parylene.

6. The neural magnetic pole according to claim 4, characterized in that, The neural magnetic poles also include: A protective layer is located above the insulating layer and covers the contacts and the leads.

7. The neural magnetic pole according to claim 6, characterized in that, The protective layer includes a first passivation layer and a second passivation layer, wherein the material of the first passivation layer is SiO2 and the material of the second passivation layer is Si3N4.

8. The neural magnetic pole according to claim 6, characterized in that, The protective layer is made of one or more of Al2O3, HfO2, and parylene.

9. The neural magnetic pole according to claim 6, characterized in that, The neural magnetic poles also include: A bio-coating is located above the protective layer.

10. The neural magnetic pole according to claim 9, characterized in that, The bio-coating material is one or more of hydrophilic proteins, heparin, and phospholipid monomolecules.

11. The neural magnetic pole according to claim 1, characterized in that, The first interface, the second interface, the contact, and the lead are made of one or more of Cr / Au, Ti / Au, Pt / Au, and Ta / Au.

12. The neural magnetic pole according to claim 1, characterized in that, The substrate is made of one or more of the following materials: single-crystal silicon, silicon oxide, silicon-on-insulator, flexible polyimide, glass, and sapphire.

13. The neural magnetic pole according to claim 1, characterized in that, The number of the first magnetic tunnel junction and the number of the second magnetic tunnel junction are equal and both are even. An even number of the first magnetic tunnel junctions share a first functional stack in pairs, and any two first magnetic tunnel junctions that do not share a first functional stack are connected through contacts on a second functional stack to form the sensing unit; wherein, the contacts above the first magnetic tunnel junctions at both ends of the sensing unit are respectively connected to the two first interfaces through leads. An even number of second magnetic tunnel junctions share a first functional stack in pairs, and any two second magnetic tunnel junctions that do not share the first functional stack are connected through contacts on the second functional stack to form the reference unit; wherein, the contacts above the second magnetic tunnel junctions at both ends of the reference unit are respectively connected to two second interfaces through leads.

14. A method for preparing neural magnetic poles, characterized in that, The method for preparing neural magnetic poles is used to prepare neural magnetic poles as described in any one of claims 1 to 13, wherein the method for preparing neural magnetic poles comprises: A first functional stack, a barrier layer, and a second functional stack are sequentially deposited on the substrate surface; Photolithographic patterning is performed in the first and second regions above the substrate, and etched to the substrate surface to define the first functional stacked structure of the sensing unit and the first functional stacked structure of the reference unit, respectively. Photolithographic patterning is performed on the surface of the second functional stack in the first and second regions above the substrate, and etching is performed on the surface of the first functional stack to define the first magnetic tunnel junction structure of the sensing unit and the second magnetic tunnel junction structure of the reference unit, respectively. An insulating layer is deposited on the substrate surface to cover the first magnetic tunnel junction and the second magnetic tunnel junction; Photolithographic patterns are formed on the surface of the insulating layer to create through holes at the top of the first magnetic tunnel junction and the second magnetic tunnel junction, respectively. A metal layer is deposited on the surface of the insulating layer, multiple first magnetic tunnel junctions are connected in series to form a sensing unit, multiple second magnetic tunnel junctions are connected in series to form a reference unit, and the leads of the sensing unit and the leads of the reference unit are defined respectively. Deposit a protective layer on the surface of the metal layer; The protective layer covering the lead surface is etched to open up part of the solder pad area, forming the first interface and the second interface respectively.

15. The method for preparing neural magnetic poles according to claim 14, characterized in that, Before etching the protective layer covering the lead surface to open part of the pad area and form the first interface and the second interface respectively, the neural magnetic pole preparation method further includes: The substrate is thinned.

16. The method for preparing neural magnetic poles according to claim 14, characterized in that, After etching the protective layer covering the lead surface to open part of the pad area and forming the first interface and the second interface respectively, the neural magnetic pole preparation method further includes: The overall shape is processed to obtain probe-shaped neural magnetic poles.

17. A neural magnetic pole sensor, characterized in that, The neural magnetic pole sensor includes: The neuromagnetic poles as described in any one of claims 1 to 13, and An interface circuit connected to the first and second interfaces of the neural magnetic poles; The interface circuit includes: A sensing signal processing module, connected to the first interface, is used to process the signal output by the sensing unit of the neural magnetic pole to obtain a sensing signal. A reference signal processing module, connected to the second interface, is used to process the signal output by the reference unit of the neural magnetic pole to obtain a reference signal; A differential amplifier module, connected to the sensing signal processing module and the reference signal processing module, is used to perform differential operations on the sensing signal and the reference signal to obtain a differential output signal; The data acquisition module is connected to the differential amplifier module and is used to acquire and store the differential output signal.

18. The neural magnetic pole sensor according to claim 17, characterized in that: The sensing signal processing module includes a first Wheatstone bridge, a first amplifier, and a first filter connected in sequence; wherein, the first Wheatstone bridge is used to convert the resistance change of the sensing unit into a first differential voltage signal, the first amplifier is used to amplify the first differential voltage signal to obtain a first amplified signal, and the first filter is used to filter the first amplified signal to obtain a sensing signal. The reference signal processing module includes a second Wheatstone bridge, a second amplifier, and a second filter connected in sequence; wherein, the second Wheatstone bridge is used to convert the resistance change of the reference unit into a second differential voltage signal, the second amplifier is used to amplify the second differential voltage signal to obtain a second amplified signal, and the second filter is used to filter the second amplified signal to obtain a reference signal; The first amplifier is connected to a first feedback resistor, and the second amplifier is connected to a second feedback resistor. By adjusting the resistance values ​​of the first feedback resistor and the second feedback resistor respectively, gain matching compensation is performed on the sensitivity difference between the sensing unit and the reference unit, so that the response amplitude of the first amplified signal and the second amplified signal to the same magnetic field environment is consistent.

19. The neural magnetic pole sensor according to claim 18, characterized in that, The first amplifier, the second amplifier, and the differential amplifier module are instrumentation amplifiers, programmable gain amplifiers, or low-noise magnetic sensor dedicated integrated circuit chips.

20. The neural magnetic pole sensor according to claim 17, characterized in that, The data acquisition module is a data acquisition chip that integrates an amplifier and an analog-to-digital converter.

Images