Magnetoelectric antenna, implantable neuroscanner, and implantable neurostimulator

Abstract

There is provided a magnetoelectric antenna. The magnetoelectric antenna comprises a first electrode comprising a first metal contact, a second electrode comprising a second metal contact, a piezoelectric layer between the first electrode and the second electrode, and one or more magnetostrictive layers.

Description

[0001] MAGNETOELECTRIC ANTENNA, IMPLANTABLE NEUROSCANNER, AND IMPLANTABLE NEUROSTIMULATOR

[0002] Field of the Invention

[0003] The present invention relates to a magnetoelectric antenna and particularly, although not exclusively, to an implantable neuroscanner and / or an implantable neurostimulator including a magnetoelectric antenna.

[0004] Background

[0005] Implantable devices may be useable to scan and / or stimulate tissue according to need. However, such devices may require invasive procedures, such as surgical procedures, to implant the devices into the tissue of a subject.

[0006] Moreover, implantable devices may be bulky and require wiring to power these devices. This may make the implantation of such devices challenging.

[0007] In the context of neurostimulators, ultrasonic technology has emerged as an option for delivering stimulation signals to neural tissue. However, ultrasonic signals require some form of medium through which the signal can propagate, and they are limited in terms of the available frequencies and bandwidths with which the stimulation signals can be produced.

[0008] Meanwhile, radiofrequency (RF) antennas have also emerged as suitable candidates for scanning and stimulation in the context of implantable devices. RF antennas operate in the radiofrequency portion of the electromagnetic spectrum, typically ranging from 3 kHz to 300 GHz, and are capable of efficiently transmitting and receiving signals across various applications (e.g., telecommunications, broadcasting, radar, satellite communications, and the transmission of biological signals).

[0009] However, the size of RF antennas is inversely proportional to the transmission frequency of the antenna, meaning that a size of several centimetres may be needed for transmission frequencies on the order of 100 MHz or below. This may result in inconveniently large and / or bulk antennas that may not be suitable for use in the context of implantable devices.

[0010] The present invention has been devised in light of these considerations.

[0011] Summary of the Invention

[0012] In a general sense, described herein is a magnetoelectric antenna having a layered structure comprising one or more electrodes, a magnetostrictive portion for deforming in response to the application of an electromagnetic field thereto, and a piezoelectric portion for physically deforming with the magnetostrictive portion to induce the generation of a voltage. Such a magnetoelectric antenna may be used either as a stimulator (e.g., to deliver the generated voltage to a site for stimulation) or as a scanner (e.g., to transmit the generated voltage for analysis to infer information about the electromagnetic field applied to the magnetostrictive portion).8673170

[0013] 2

[0014] The magnetoelectric antennas described herein may beneficially provide a compact (in some cases, micro-scale) antenna that may be useable in a micro-scale implantable device.

[0015] In use, an electromagnetic signal that is received by the one or more magnetostrictive layers may induce a corresponding physical strain response in the one or more magnetostrictive layers. In the context of an oscillating (electro)magnetic signal, the physical strain response in the one or more magnetostrictive layers may be propagated through the magnetoelectric antenna has an acoustic wave. This acoustic wave may be propagated into the piezoelectric layer which, in response, generates a voltage signal in accordance with the acoustic wave (e.g., in accordance with a wave profile of the acoustic wave). This generated voltage signal may, for example, be used to stimulate tissue in the vicinity of the magnetoelectric antenna and / or may be transmitted to an external device for analysis of the electromagnetic signal received by the one or more magnetostrictive layers.

[0016] The piezoelectric layer and physically interface with the one or more magnetostrictive layers. In this way, the propagation of acoustic waves between the piezoelectric layer and the one or more magnetostrictive layers may be improved (e.g., losses may be reduced).

[0017] In a first aspect, there is provided a magnetoelectric antenna comprising: a first electrode comprising a first metal contact; a second electrode comprising a second metal contact; a piezoelectric layer between the first electrode and the second electrode; and one or more magnetostrictive layers.

[0018] The first metal contact may be formed from any suitable metal. For example, the first metal contact may be formed from any one or more of titanium, platinum, and / or nickel. For example, the first metal contact may be formed from a titanium / platinum alloy and / or a titanium / nickel alloy.

[0019] The first metal contact may have a thickness of 25 nm or more, 50 nm or more, 75 nm or more, or 100 nm or more. Alternatively, the first metal contact may have a thickness of 150 nm or less, 100 nm or less, 75 nm or less, or 50 nm or less. Alternatively, the first metal contact may have a thickness of between 25 and 150 nm, between 25 and 100 nm, between 25 and 100 nm, between 25 and 75 nm, between 25 and 50 nm, between 50 and 150 nm, between 50 and 100 nm, between 50 and 75 nm, between 75 and 150 nm, between 75 and 100 nm, or between 100 and 150 nm.

[0020] In some examples, the thickness of the first metal contact may be between 50 and 100 nm.

[0021] It has been found that providing the first metal contact with such a thickness may lead to an amplification of overtones in the acoustic response of the magnetoelectric antenna. The first electrode may further comprise an adhesion layer for adhering the first electrode to the piezoelectric layer. For example, the first metal contact may be formed from a suitable conductive material (e.g., a biocompatible metal such as platinum) and the adhesion layer may be formed from a different conductive material suitable for adhering to the piezoelectric layer. For example, the adhesion layer may be formed from titanium. The adhesion layer may have a thickness that is less than a thickness of the first metal contact.

[0022] The second metal contact may be formed from any suitable metal. For example, the second metal contact may be formed from any one or more of iron, gallium, iron-gallium (FeGa), platinum, and / or gold. For example, the second metal contact may be formed from a FeGa / platinum alloy and / or a gold / platinum8673170

[0023] 3

[0024] alloy. In cases where the second electrode is in direct contact with the piezoelectric layer and / or with the one or more magnetostrictive layers, the second electrode may further comprise respective adhesion layers for adhering the second electrode to the piezoelectric layer and / or one of the one or more magnetostrictive layers respectively. For example, titanium may be a suitable material for an adhesion layer to adhere the second electrode to the piezoelectric layer. Additionally or alternatively, iron or an iron alloy (e.g., FeGa) may be a suitable material for an adhesion layer to adhere the second electrode to at least one of the one or more magnetostrictive layers.

[0025] Platinum may be chosen to form at least part of the first and / or second metal contact for its high degree of biocompatibility.

[0026] The first and / or second metal contacts may be formed from alloys containing metals that are chosen to optimise the interface between each of the metal contacts and respectively adjacent layers of the magnetoelectric antenna.

[0027] The piezoelectric layer may be formed from any suitable piezoelectric material. For example, the piezoelectric layer may be formed from lead zirconate titanate (PZT). Alternatively, the piezoelectric layer may be formed from a metal-insulator compound having a crystalline structure. The crystalline structure may, for example, be a zincblende or wurtzite structure. Wurtzite crystal structures are known to exhibit strong piezoelectricity. Compounds having such a structure (and, in some cases, from which the piezoelectric layer may be formed) may include gallium nitride (GaN), indium nitride (InN), aluminium nitride (AIN) and zinc oxide (ZnO).

[0028] The piezoelectric layer may be formed of undoped or doped aluminium nitride (AIN) . In this way, the piezoelectric layer may have good biocompatibility.

[0029] The crystalline structure of the piezoelectric layer may have a c-axis orientation. The crystalline structure of the piezoelectric layer may have a grain size of at least 100 nm, e.g., at least 150 nm, and / or no more than 200 nm, e.g., no more than 180 nm. The crystalline structure of the piezoelectric layer may have a grain size of 167.8 nm. The piezoelectric layer may be formed of wurtzite AIN.

[0030] The piezoelectric layer may have a thickness of 10 pm or less, 5 pm or less, 1 pm or less, 500 nm or less, or 100 nm or less. Alternatively, the piezoelectric layer may have a thickness of 50 nm or more, 100 nm or more, 500 nm or more, 1 pm or more, or 5 pm or more. Alternatively, the piezoelectric layer may have a thickness between 100 nm and 10 pm, between 100 nm and 5 pm, between 100 nm and 1 pm, between 100 nm and 500 nm, between 500 nm and 10 pm, between 500 nm and 5 pm, between 500 nm and 1 pm, between 1 pm and 10 pm, between 1 pm and 5 pm, or between 5 pm and 10 pm.

[0031] In some examples, the thickness of the piezoelectric layer may be between 100 nm and 5 pm.

[0032] Each of the one or more magnetostrictive layers may be formed from a material that exhibits magnetostrictive behaviour - that is each of the one or more magnetostrictive layers may change shape in response to the application of a magnetic field, thereby generating strain therein that may be transferred to the piezoelectric layer for the generation of an electrical signal therefrom.8673170

[0033] 4

[0034] The material from which each of the magnetostrictive layers is formed may, for example, be a material having ferromagnetic, ferrimagnetic, or antiferromagnetic properties.

[0035] For example, each of the magnetostrictive layers may be formed from iron and / or an alloy of iron. The alloy may, for example, be an alloy of iron with one or more other materials, such as another transition metal (e.g., gallium, aluminium, and / or cobalt). For example, each of the one or more magnetostrictive layers may be formed from any one or more of: cobalt ferrite, iron gallium (FeGa, orgalfenol), iron gallium boron (FeGaB), iron cobalt (FeCo), and / or iron-aluminium alloys (Alfenol).

[0036] The one or more magnetrostrictive layers may be formed from Feo79Gao21. The one or more magnetostritive layers may have a BCC crystal structure, e.g., a (110)-oriented BCC crystal structure. For the purposes of the magnetoelectric antennas described herein, it has been found that when a magnetostrictive layer is formed from an FeGa alloy, the extent of magnetostriction in said layer may be maximised when the FeGa alloy has a gallium concentration between approximately 19% and 27%. Meanwhile, it has been found that when a magnetostrictive layer is formed from an FeCo alloy, the extent of magnetostriction in said layer may be maximised when the FeCo alloy has an approximately equiatomic composition (e.g., approximately 50% iron and approximately 50% cobalt).

[0037] Meanwhile, it has been found that when a magnetostrictive layer is formed from an FeAl alloy, the extent of magnetostriction in said layer may be maximised when the aluminium concentration of the FeAl alloy is between approximately 10% and 14%.

[0038] Additionally or alternatively, any of the magnetostrictive layers may be formed from an alloy of iron with one or more rare earth metals. For example, any of the one or more magnetostrictive layers may be formed from an alloy of iron with terbium and / or dysprosium. For example, the alloy may be iron-terbium (TbFe2), iron-dysprosium (DyFe2), or an iron-terbium-dysprosium alloy (TbxDyi-xFe2). For example, the alloy may be terfenol-D (TbxDyi-xFe2, where x is approximately 0.3). Terfenol-D is known to exhibit particularly strong magnetostrictive behaviour, with a magnetostriction coefficient 80 times greater than that of iron.

[0039] There may be a symmetry mismatch between the piezoelectric layer and the one or more magnetostrictive layers.

[0040] The piezoelectric layer may be formed of AIN, and the one or more magnetostrictive layers may be formed of FeGa, e.g., Feo79Gao21.

[0041] The symmetry mismatch between the piezoelectric layer and the magnetostrictive layer may boost scattering and internal friction within the magnetoelectric antenna, increasing the operating bandwidth of the magnetoelectric antenna.

[0042] The piezoelectric layer and / or the one or more magnetostrictive layer may be manufactured / deposited via sputtering techniques e.g., radiofrequency sputtering.

[0043] The one or more magnetostrictive layers may be physically interfaced with the piezoelectric layer. In use, an electromagnetic signal applied to the magnetostrictive layers may cause the one or more8673170

[0044] 5

[0045] magnetostrictive layers to deform (e.g., expand or contract). This may induce physical strain in the one or more magnetostrictive layers that may be transferred to the piezoelectric layer via the physical interface between the one or more magnetostrictive layers and the piezoelectric layer. The physical strain induced in the piezoelectric layer may cause the generation of a potential difference across the piezoelectric layer in accordance with the induced strain (which is itself in accordance with the received electromagnetic signal), thereby establishing a potential difference between the first and second electrode. The established potential difference causes the generation of an electrical signal in accordance with the induced strain that may be used, for example, to stimulate tissue in the vicinity of the magnetoelectric antenna and / or to analyse the received electromagnetic signal (e.g., in cases where the electromagnetic signal is received from tissue into which the magnetoelectric antenna has been implanted).

[0046] The magnetoelectric antenna as described herein may benefit from a very small size, making it suitable for implantation as part of an implantable device.

[0047] There may be a stiffness mismatch between the piezoelectric layer and the magnetostrictive layer of 1 :N, wherein N is at least 2, and no more than 10, or no more than 5. N may be 4.82.

[0048] There may be a Young’s modulus mismatch between the piezoelectric layer and the one or more magnetostrictive layers. For example, as mentioned, the piezoelectric layer may be formed of AIN, and the one or more magnetostrictive layer may be formed of Feo79Gao21. The Young’s modulus of the piezoelectric layer may be at least 2 times, or at least 4 times larger than that of the magnetostrictive layer. The Young’s modulus of the piezoelectric layer may be 4.8 times larger than that of the magnetostrictive layer.

[0049] Due to the Young’s modulus mismatch, in an electromagnetic field, the strain may be concentrated within the magnetostrictive layer, and the interfacial shear on the piezoelectric layer may be elevated. In this way, the acoustic impedance contrast between the layers and damping may be enhanced, and the operating bandwidth of the magnetoelectric antenna may be increased.

[0050] The magnetoelectric antenna may be formed on a substrate. The substrate may be a crystalline substrate. The crystalline substrate may, for example, comprise silicon, silicon carbide (SiC), sapphire, and / or gallium nitride (GaN). The crystalline substrate may be oriented in a suitable crystal plane, for example the (100) plane.

[0051] The magnetoelectric antenna may be configured to enhance higher order standing waves within the magnetoelectric antenna. In this way, the magnetoelectric antenna may have a wider operating bandwidth.

[0052] The substrate may be a polished substrate, e.g., a polished silicon substrate. The substrate may be a double-side polished substrate, e.g., a double-side polished silicon substrate.

[0053] In this way, the operating bandwidth of the magnetoelectric antenna may be increased.

[0054] A buffer layer may be provided between the substrate and the magnetoelectric antenna. The buffer layer may be provided to reduce a strain mismatch between the substrate and the magnetoelectric antenna.8673170

[0055] 6

[0056] The buffer layer may be formed from any suitable material. For example, the buffer layer may be formed from aluminium nitride, silicon nitride, silicon oxide, and / or silicon oxynitride.

[0057] The substrate may have any suitable thickness for supporting the depositing of the layers of the magnetoelectric antenna thereon. For example, the thickness of the substrate may be between 10 pm and 1mm.

[0058] The first electrode may comprise an acoustic Bragg reflector. The acoustic Bragg reflector may comprise a first set of Bragg layers formed form a first material and a second set of Bragg layers formed from a second material having a different acoustic impedance than the first material. Alternatively, the first electrode may be defined by the first metal contact, and the magnetoelectric antenna may further comprise an acoustic Bragg reflector (as described herein) as a part of the magnetoelectric antenna distinct from the first electrode.

[0059] An acoustic Bragg reflector is a structure defined by a series of alternating layers of the first material and the second material. The different acoustic impedances of the first material and the second material, and / or the thicknesses of each of the layers of the acoustic Bragg reflector may be selected such that the acoustic Bragg reflector is configured to reflect acoustic waves having a frequency within a predetermined bandwidth. This may minimise the losses associated with the magnetoelectric antenna because it may prevent acoustic waves propagating from the one or more magnetostrictive layers into the piezoelectric layer from escaping the layers of the magnetoelectric antenna. In this way, the efficiency of the magnetoelectric antenna may be improved. For example, a magnetoelectric antenna having an acoustic Bragg reflector as described herein may exhibit lower losses and, unlike antennas that rely on air gaps for acoustic damping, may be more easily sealed in a biocompatible manner for use in an implantable device.

[0060] The acoustic Bragg reflector may be arranged such that the layers of the first set of Bragg layers alternate with the layers of the second set of Bragg layers.

[0061] That is, the structure of the acoustic Bragg reflector may be defined, at least in part, by an alternating arrangement of the first set of Bragg layers and the second set of Bragg layers.

[0062] The acoustic Bragg reflector may have an overall thickness of 20 pm or less, 15 pm or less, 10 pm or less, or 5 pm or less.

[0063] The acoustic Bragg reflector may have an overall thickness of 1 pm or more, 5 pm or more, 10 pm or more, or 15 pm or more.

[0064] The acoustic Bragg reflector may have an overall thickness of between 1 and 20 pm, between 1 and 15 pm, between 1 and 10 pm, between 1 and 5 pm, between 5 and 20 pm, between 5 and 15 pm, between 5 and 10 pm, between 10 and 20 pm, between 10 and 15 pm, or between 15 and 20 pm.

[0065] In some examples, the overall thickness of the acoustic Bragg reflector may be between 1 and 15 pm. The acoustic Bragg reflector may, for example, be a symmetric acoustic Bragg reflector. That is each of the layers of the first set of Bragg layers may have a common first thickness. Further, each of the layers8673170

[0066] 7

[0067] of the second set of Bragg layers may have a common second thickness. The first thickness may be the same as, or different than, the second thickness.

[0068] Alternatively, the acoustic Bragg reflector may, for example, be an asymmetric acoustic Bragg reflector. That is one or more of the layers of the first set of Bragg layers may have a different thickness than at least one of the other layers of the first set of Bragg layers. Additionally or alternatively, one or more of the layers of the second set of Bragg layers may have different thickness than at least one of the other layers of the second set of Bragg layers. The relative thicknesses of each adjacent layer of the acoustic Bragg reflector may be selected to configure the acoustic Bragg reflector with a predetermined reflectivity at one or more predetermined acoustic wave wavelengths.

[0069] Asymmetric acoustic Bragg reflectors may benefit from reduced shear wave losses than symmetric acoustic Bragg reflectors. In some cases, to reduce shear wave losses, an asymmetric acoustic Bragg reflector may include a layer from amongst a given set of Bragg layers that has a thickness that is double the thickness of other layers from amongst said given set of Bragg layers.

[0070] For example, in an arrangement where the first material has a lower acoustic impedance than the second material. The first and second sets of Bragg layers may be arranged as a series of alternating layers. The first set of Bragg layers may comprise two layers and the second set of Bragg layers may comprise two layers. The first layer of the first set of Bragg layers may be distal from the first metal contact of the first electrode and may have a first thickness. The first layer of the second set of Bragg layers may be between the first layer of the first set and the first metal contact and may have a second thickness equal to half the first thickness. The second layer of the first set of Bragg layers may be between the second layer of the first set and the first metal contact and may have a third thickness equal to twice the first thickness. The second layer of the second set of Bragg layers may be between the second layer of the second set and the first metal contact and may have a fourth thickness equal to the second thickness. For the avoidance of doubt, the acoustic Bragg reflector may be formed from any number of suitable layers of the first and sets of Bragg layers having any suitable thicknesses to provide an acoustic Bragg reflector having reflectance properties selected by a user / manufacturerof the magnetoelectric antennas described herein.

[0071] The first material and / or the second material may be metal.

[0072] For example, the first material may be formed from a metal or an alloy, and the second material may be formed from a semiconductor or dielectric.

[0073] Alternatively, the first material may be formed from a semiconductor or dielectric, and the second material may be formed from a metal or an alloy.

[0074] Alternatively, both the first material and the second material may be a metal or alloy. That is, the acoustic Bragg reflector may be an all-metal acoustic Bragg reflector.

[0075] In other words, the first set of Bragg layers may be formed from metal. Additionally or alternatively, the second set of Bragg layers may be formed from metal.8673170

[0076] 8

[0077] As an example, the first material may be aluminium, and the second material may be tungsten. Any other suitable combination of materials may also be used, the selection of aluminium and tungsten is merely a suitable combination of materials that has been found to have an acoustic impedance ratio sufficiently high to effectively confine acoustic waves within the structure of the magnetoelectric antennas described herein.

[0078] In other examples, only one of the materials may be metal. For example, the first material may be silica (silicon dioxide, SiC>2), and the second material may be tungsten.

[0079] Using metal to form one or both of the sets of Bragg layers of the acoustic Bragg reflector may result in a compact acoustic Bragg reflector, thereby contributing to a reduced size for the overall magnetoelectric antenna, that may be manufactured more quickly (thereby improving the manufacturing efficiency of the magnetoelectric antenna). For example, when manufacturing an acoustic Bragg reflector using conventional manufacturing techniques (e.g., sputtering techniques such as plasma sputtering or RF sputtering), it has been found that the depositing of dielectric layers of the acoustic Bragg reflector may require a stable sputtering deposition over a time period of 12 to 24 hours. Meanwhile, using the techniques described herein, it has been found that metal layers of the acoustic Bragg reflector (e.g., aluminium and / or tungsten layers) may be deposited using sputtering-based deposition over a period of just 8 to 10 hours, representing a time-saving of up to 67%.

[0080] The first metal contact may be disposed between the acoustic Bragg reflector and the piezoelectric layer. That is, the metal contact of an electrode that comprises the acoustic Bragg reflector may be disposed between the acoustic Bragg reflector and the piezoelectric layer.

[0081] Having the metal contact in direct physical contact with the piezoelectric layer may improve the efficiency with which the potential difference induced in the piezoelectric layer is coupled into an electrical signal generated across the first and second electrode. Moreover, disposing the acoustic Bragg reflector distal from the piezoelectric layer may facilitate the reflection of acoustic waves that pass through the first metal contact back into the piezoelectric layer to induce further vibrations in the piezoelectric layer, and -consequently - improve the piezoelectric coupling within the piezoelectric layer.

[0082] The acoustic Bragg reflector may further comprise an intermediate layer arranged between the first metal contact and the first and second sets of Bragg layers.

[0083] In some cases, for example, cases where the first and / or second set of Bragg layers are formed from metal, there is an increased risk of electrical shorting as a result of a direct contact between the first metal contact and a stack of metal layers forming the acoustic Bragg reflector. The intermediate layer may, therefore, be formed from a dielectric or semiconductor with a lower electrical conductivity than a metal. The intermediate layer may be selected to have an acoustic impedance similar to the first material (from which the first set of Bragg layers are formed) or similar to the second material (from which the second set of Bragg layers are formed) to optimise the reflectance efficiency of the acoustic Bragg reflector. For example, in cases where the acoustic Bragg reflector is formed from alternating layers of tungsten and8673170

[0084] 9

[0085] aluminium, the intermediate layer may be formed from a suitable dielectric such as silicon oxynitride (SiON).

[0086] The intermediate layer may couple the acoustic Bragg reflector to the first metal contact. That is, the intermediate layer may acoustically couple acoustic waves propagating through the first metal contact into the acoustic Bragg reflector and vice versa. Additionally or alternatively, the intermediate layer may provide a physical interface between the acoustic Bragg reflector and the first metal contact.

[0087] Additionally or alternatively, the second electrode may comprise an acoustic Bragg reflector, referred to herein as a second acoustic Bragg reflector (the acoustic Bragg reflector of the first electrode may be referred to herein as a first acoustic Bragg reflector). The second acoustic Bragg reflector may comprise a first set of Bragg layers formed from a first material and a second set of Bragg layers formed from a second material having a different acoustic impedance than the first material.

[0088] The second acoustic Bragg reflector may have any of the properties or structures described above in relation to the first acoustic Bragg reflector.

[0089] In cases where the first electrode comprises the first acoustic Bragg reflector and the second electrode comprises the second acoustic Bragg reflector, the first acoustic Bragg reflector and the second acoustic Bragg reflector may have the same properties and / or structure. In such cases, the first acoustic Bragg reflector and second acoustic Bragg reflector may have any of the properties and / or structures described above in common.

[0090] Alternatively, the first acoustic Bragg reflector and the second acoustic Bragg reflector may have different properties and / or different structures. In such cases, the first acoustic Bragg reflector may have any of the properties and / or structures described above, and the second acoustic Bragg reflector may have any other of the properties and / or structures described above.

[0091] The second electrode may further comprise a third metal contact co-planar with the second metal contact. The second metal contact and the third metal contact may have different thicknesses. Alternatively, the second electrode may be defined, at least in part, by the second metal contact; and the magnetoelectric antenna may further comprise a third electrode defined, at least in part, by the third metal contact (as described herein). The second and third electrodes may be independently addressable. Alternatively, the second and third electrodes may be connected such that they are simultaneously addressable.

[0092] A surface of the second metal contact distal from the first metal contact may be co-planar from a corresponding surface of the third metal contact that is distal from the first metal contact. In this way, the different thicknesses of the second and third metal contact may result in a different spacing between the first metal contact and the second metal contact than the spacing between the first metal contact and the third metal contact. The spacing between the first electrode and the second electrode may be occupied by the piezoelectric layer. As such, a difference between the spacing between the first and second metal contacts and the spacing between the first and third metal contacts may result in a different thickness of piezoelectric material between the first and second metal contacts than a thickness of piezoelectric material between the first and third metal contacts.8673170

[0093] 10

[0094] In the context of piezoelectric materials, the resonant frequency of the piezoelectric layer is inversely proportion to the thickness of said layer. Accordingly, by providing different thicknesses of piezoelectric layer between different pairs of metal contacts, multiple different resonant frequencies in the magnetoelectric antenna can be achieved. For example, the spacing between the first and second metal contacts may yield a first piezoelectric resonant frequency, while the spacing between the first and third metal contacts may yield a second piezoelectric resonant frequency different from the first piezoelectric frequency. In this way, the sensitivity of the magnetoelectric antenna may be selectively increased at multiple different frequencies of interest to the user / manufacturer of the magnetoelectric antennas described herein.

[0095] Additionally or alternatively, the first electrode may comprise a fourth metal contact co-planer with the first metal contact. The first metal contact and the fourth metal contact may have different thicknesses.

[0096] The first metal contact may, for example, have a same thickness as the second metal contact.

[0097] Additionally or alternatively, the fourth metal contact may, for example, have a same thickness as the third metal contact. Alternatively, each of the metal contacts may have a respective thickness that may be the same as or different from any of the other metal contacts of the first and / or second electrode.

[0098] In this way, selective control of the number of resonant frequencies (and the value of each of those resonant frequencies) may be achieved with the first electrode as well as (or instead of) the second electrode. For example, in cases where the first electrode comprises the first metal contact and the fourth metal contact, and the second electrode comprises the second metal contact and the third metal contact, each of the metal contacts may be arranged such that a first portion of the first metal contact is aligned with a first portion of the second metal contact, a second portion of the first metal contact is aligned with a first portion of the third metal contact, a first portion of the fourth metal contact is aligned with a second portion of the second metal contact, and a second portion of the fourth metal contact is aligned with a second portion of the third metal contact. In this way, a magnetoelectric antenna may be provided with four (possibly different) piezoelectric resonant frequencies respectively determined by the thickness of piezoelectric layer occupying the spacings between (i) the first and second metal contacts, (ii) the first and third metal contacts, (iii) the fourth and second metal contacts, and (iv) the fourth and third metal contacts.

[0099] One or more of the metal contacts of the first electrode and / or second electrode may be a patterned metal contact.

[0100] In addition to the thickness of the piezoelectric layer, the resonant frequency (or frequencies) of the magnetoelectric antenna may be controlled by adjusting the geometry of one or more of the metal contacts of the first and / or second electrode. These geometries may be adjusted by patterning the corresponding metal contact appropriately to define an electrode profile of the electrode. In this way, both the resonant frequency and frequency-response profile (e.g., the width of the resonance peak) of the magnetoelectric antenna may be controlled according to the user / manufacturer’s needs.8673170

[0101] 11

[0102] The piezoelectric layer and / or the one or more magnetostrictive layers may be patterned in accordance with the patterned metal contact(s).

[0103] The patterning of each of the patterned layers from amongst the piezoelectric layer and / or one or more of the magnetostrictive layers may be aligned with the patterning of the one or more patterned contacts. This may simplify the manufacturing of the magnetoelectric antennas, as a single patterned mask may be useable to define the patterning / layout of each of the layers of the magnetoelectric antenna. Additionally, patterning the one or more magnetostrictive layers and / or the piezoelectric layer with the same pattern as one or more of the metal contacts of the magnetoelectric antenna may reduce the amplitude of any self-resonant responses within the magnetoelectric antenna, thereby improving the operational efficiency of the magnetoelectric antenna.

[0104] The one or more magnetostrictive layers may be disposed between the piezoelectric layer and the second electrode.

[0105] In some cases, it may be beneficial for a layer from amongst the one or more magnetostrictive layers to directly interface with the piezoelectric layer to improve the coupling of the induced strain from the one or more magnetostrictive layers into the piezoelectric layer.

[0106] Alternatively, the second electrode may be disposed between the piezoelectric layer and the one or more magnetostrictive layers.

[0107] For example, the second electrode may directly interface with the piezoelectric layer on a first side of the second electrode, and the second electrode may directly interface with a magnetostrictive layer of the one or more magnetostrictive layers on the second side of the second electrode.

[0108] In some cases, it may be beneficial for the second electrode to be arranged between the piezoelectric layer and the one or more magnetostrictive layers such that the second electrode directly interfaces with the piezoelectric layer to improve the efficiency with which the induced potential difference in the piezoelectric layer generates an electrical signal as a consequence of the potential difference established between the first and second electrode.

[0109] The magnetoelectric antenna may further comprise a capping layer on the one or more magnetostrictive layers.

[0110] Magnetostrictive materials (e.g., terfenol-D or alfenol) may be brittle or easily susceptible to fracture upon impact. A capping layer may therefore be beneficial to prevent damage to any of the one or more magnetostrictive layers.

[0111] In cases where the one or more magnetostrictive layers are disposed between the piezoelectric layer and the second electrode, the capping layer may facilitate an interface (e.g., an adhesion) between a layer of the one or more magnetostrictive layers and the second electrode.

[0112] The capping layer may be made from any suitable material. For example, the capping layer may be formed from aluminium. The capping layer may have a thickness of at least 2 nm and no more than 10 nm, e.g., 5nm.8673170

[0113] 12

[0114] Each of the one or more magnetostrictive layers may be encapsulated by an encapsulant. The encapsulant may be formed of Parylene, e.g., Parylene C. In this way, the biocompatibility of the magnetoelectric antenna may be improved.

[0115] The magnetoelectric antenna may further comprise an adhesion layer for adhering the one or more magnetostrictive layers to the piezoelectric layer.

[0116] The adhesion layer may promote the coupling of acoustic waves propagating between the one or more magnetostrictive layers and the piezoelectric layer.

[0117] The adhesion layer may be formed from any suitable material. For example, the adhesion layer may be formed from a metal having a similar atomic weight to at least one element from which the piezoelectric layer is formed. Further, the adhesion layer may be formed from a material having a similar acoustic impedance to one or both of the piezoelectric layer and the layer from amongst the one or more magnetostrictive layers that interfaces with the piezoelectric layer. For example, the adhesion layer may be formed from any one or more of titanium, nickel, chromium, tantalum, platinum, or alloys thereof (e.g., nichrome).

[0118] The adhesion layer may have a thickness of 20 nm or less, 15 nm or less, 10 nm or less, 5 nm or less, or 1 nm or less. Alternatively, the adhesion layer may have a thickness of 0.5 nm or more, 1 nm or more, 5 nm or more, 10 nm or more, or 15 nm or more. Alternatively, the adhesion layer may have a thickness between 0.5 and 20 nm, between 0.5 and 15 nm, between 0.5 and 10 nm, between 0.5 and 5 nm, between 0.5 and 1 nm, between 1 and 20 nm, between 1 and 15 nm, between 1 and 10 nm, between 1 and 5 nm, between 5 and 20 nm, between 5 and 15 nm, between 5 and 10 nm, between 10 and 20 nm, between 10 and 15 nm, or between 15 and 20 nm.

[0119] In some examples, the adhesion layer may have a thickness between 1 and 10 nm.

[0120] The one or more magnetostrictive layers may comprise a plurality of magnetostrictive layers defining a stack. Each adjacent pair of magnetostrictive layers in the stack may be separated by a laminating layer. The or each laminating layer may reduce the effects of eddy currents on the magnetostriction within each of the plurality of magnetostrictive layers.

[0121] The or each laminating layer may be formed from a suitable dielectric to reduce the eddy current effects, and to prevent short-circuiting effects within the magnetoelectric stack. For example, the or each laminating layer may be formed from aluminium nitride.

[0122] The magnetoelectric antenna may further comprise a casing arranged to encapsulate the first electrode, the second electrode, the piezoelectric layer, and the one or more magnetostrictive layers.

[0123] The casing may enclose the magnetoelectric antenna such that the magnetoelectric antenna is useable (together with any other components enclosed by the casing) as single device housed within the casing. The casing may be formed from a biocompatible material.8673170

[0124] 13

[0125] Biocompatibility of the casing may make a device, of which the magnetoelectric antenna may be part, suitable for use as an implantable device (e.g., an implantable scanner and / or stimulator).

[0126] The casing may, for example, be formed from platinum, stainless steel, titanium, ora cobalt-chromium alloy. Alternatively, the casing may be formed from a biocompatible polymer such as HDPE.

[0127] A thickness of each of the one or more magnetostrictive layers may be 80 nm or less.

[0128] In some cases, a thickness of each of the one or more magnetostrictive layers may be 70 nm or less, 60 nm or less, 50 nm or less, or 40 nm or less, or 30 nm or less. It has been found that such magnetostrictive layers are less susceptible to the formation of Bloch walls in their magnetic domains, instead the domain walls at such low thicknesses typically take the form of Neel walls. It may be beneficial for the domain walls to be Neel walls instead of Bloch walls, as Bloch walls require a stronger magnetic field to be applied to them to induce rotation of the magnetic domain. Magnetic domains with Neel walls therebetween are therefore more susceptible to magnetostriction with relatively weaker magnetic fields.

[0129] A thickness of each of the one or more magnetostrictive layers may be 20 nm or more.

[0130] In some cases, a thickness of each of the one or more magnetostrictive layers may be 30 nm or more, 40 nm or more, 50 nm or more, 60 nm or more, or 70 nm or more. It has been found that such magnetostrictive layers are less vulnerable to interfaces stresses which may physically weaken the layer, and to domain wall pinning which may inhibit magnetostriction in the layer.

[0131] A thickness of each of the one or more magnetostrictive layers may be between 20 nm and 80, between 20 nm and 70 nm, between 20 nm and 60 nm, between 20 nm and 50 nm, between 20 nm and 40 nm, between 20 nm and 30 nm, between 30 nm and 80 nm, between 30 nm and 70 nm, between 30 nm and 60 nm, between 30 nm and 50 nm, between 30 nm and 40 nm, between 40 nm and 80 nm, between 40 nm and 70 nm, between 40 nm and 60 nm, between 40 nm and 50 nm, between 50 nm and 80 nm, between 50 nm and 70 nm, between 50 nm and 60 nm, between 60 nm and 80 nm, between 60 nm and 70 nm, or between 70 nm and 80 nm.

[0132] In cases where the magnetoelectric antenna comprise a plurality of magnetostrictive layers (e.g., arranged to define a magnetostrictive stack), the combined thickness of all the magnetostrictive layers (including, for example, the thickness of any laminating layers therebetween) may be 10 pm or less, 5 pm or less, 1 pm or less, 500 nm or less, or 100 nm or less. Alternatively, the combined thickness may be 50 nm or more, 100 nm or more, 500 nm or more, 1 pm or more, or 5 pm or more. Alternatively, the combined thickness may be between 100 nm and 10 pm, between 100 nm and 5 pm, between 100 nm and 1 pm, between 100 nm and 500 nm, between 500 nm and 10 pm, between 500 nm and 5 pm, between 500 nm and 1 pm, between 1 pm and 10 pm, between 1 pm and 5 pm, or between 5 pm and 10 pm.

[0133] In some examples, the combined thickness of all the magnetostrictive layers (including, for example, the thickness of any laminating layers therebetween) may be between 100 nm and 5 pm.8673170

[0134] 14

[0135] In particular cases, the thickness of each of the one or more magnetostrictive layers may be between 30 nm and 60 nm.

[0136] The magnetoelectric antenna may have an active area of 0.05 mm2or less.

[0137] For example, the magnetoelectric antenna may have an active area of, 0.04 mm2or less, or 0.03 mm2or less.

[0138] That is, the side length of the magnetoelectric antenna’s active area may be on the order 5 pm or less (e.g., 2.5 pm or less).

[0139] Such small antennas may be suitable for use in micro-scale implantable devices. In this way they may be implanted via relatively non-invasive means (e.g., via injection as opposed to surgery), and may not require intrusive procedures for a patient to accommodate their implantation.

[0140] The magnetoelectric antennas described herein may be manufactured using any suitable manufacturing techniques. For example, the metal contacts of the first and second electrodes may be deposited via a deposition technique - e.g., vapour deposition such as plasma-enhanced chemical vapour deposition. Additionally or alternatively, one or more of the acoustic Bragg reflectors, the piezoelectric layers, and / or the magnetostrictive layers may be deposited via sputtering techniques - e.g., radiofrequency sputtering. As discussed, one or more features of the magnetoelectric antenna described herein may increase the operating bandwidth of the magnetoelectric antenna.

[0141] The magnetoelectric antennas may have an operating bandwidth of at least 200 MHz and / or no more than 15 GHz. The magnetoelectric antennas may have an operating bandwidth of at least 200 MHz and / or no more than 30 GHz. The magnetoelectric antennas may have an operating bandwidth of at least 20 GHz, e.g., 22.6 GHz.

[0142] The magnetoelectric antennas may be configured to operate (e.g., may be configured to generate an electrical signal, and / or to receive an electromagnetic signal) at a frequency of between 3 GHz and 4 GHz, e.g., between 3.0 GHz and 3.6 GHz. The magnetoelectric antennas may be configured to operate (e.g., may be configured to generate an electrical signal, and / or to receive an electromagnetic signal) at a frequency of 3.1 GHz, 3.3 GHz and / or 3.9 GHz. The magnetoelectric antennas may have a resonant frequency (e.g., a fundamental resonant frequency) between 3 GHz and 4 GHz, e.g., between 3.0 GHz and 3.6 GHz. The magnetoelectric antennas may have a resonant frequency (e.g., a fundamental resonant frequency) of 3.1 GHz, 3.3 GHz and / or 3.9 GHz.

[0143] In this way, the magnetoelectric antennas may establish more reliable wireless links with an ex vivo devices (e.g., ex vivo transmitters / receivers) when implanted in tissue (e.g., when the implantable device containing the magnetoelectric antennas is implanted in tissue).

[0144] The magnetoelectric antenna may comprise an active transduction layer. The active transduction layer may be formed of one or more stacked units, each unit comprising a respective piezoelectric layer and magnetostrictive layer. It will be appreciated that the piezoelectric layer may have any of the properties or structures described above for the piezoelectric layer, and the magnetostrictive layer may have any of the8673170

[0145] 15

[0146] properties or structures described above for the magnetostrictive layer. The thickness of the magnetostrictive layer may be at least 20 nm and no more than 30 nm, e.g., 25 nm. The thickness of the piezoelectric layer may be at least 1 nm and no more than 10 nm, e.g., 5nm. Each unit may be identical. The thickness of the active transduction layer may be at least 400 nm and no more than 500 nm, e.g., the thickness of the active transduction layer may be 453 nm. The active transduction layer may comprise [FeO.79GaO.21(25 nm) / AIN(5 nm)]z.ln another aspect, there is described an implantable neuroscanner comprising any of the magnetoelectric antennas described herein. In use, the magnetoelectric antenna is configured to record electrical signals emitted by a tissue into which the neuroscanner has been implanted and transmit the recorded electrical signals to an ex vivo device.

[0147] The implantable neuroscanner may comprise one or more of the magnetoelectric antennas as described herein.

[0148] The implantable neuroscanner may have an operating bandwidth of between 200 MHz and 15 GHz, or between 200MHz and 30 GHz. For example, the implantable neuroscanner may have an operating bandwidth of 22.6 GHz.

[0149] The reception strength of each magnetoelectric antenna of the neuroscanner may be within the range of -10 dBm to 10 dBm.

[0150] The magnetoelectric antenna(s) may be mounted on a PCB. The magnetoelectric antenna(s) may be mounted on a PCB via a silver nanoparticle ink (e.g., an ethylene glycol-based silver nanoparticle ink). The resistivity of the silver nanoparticle ink may be 7.9pQcm. In this way, the implantability of the implantable neuroscanner may be improved.

[0151] In use, moving ions in the vicinity of the implantable neuroscanner produce electromagnetic field signals that may excite the one or more magnetostrictive layers of each of the magnetoelectric antennas of the neuroscanner. This induces a physical (strain) response in the one or more magnetoelectric layers that is subsequently transferred (e.g., as an acoustic wave) into the piezoelectric layer of the corresponding antenna. The transferred strain may induce a potential difference within the piezoelectric layer that causes the establishment of a potential difference across the first and second electrodes of the antenna, thereby generating an electrical signal that may be transmitted to a device (e.g., an ex vivo device) for analysis of the scanned electromagnetic field signals.

[0152] For the avoidance of doubt, the implantable neuroscanners described herein need not be neuroscanners per se, but rather may be an implantable scanner for scanning any electrical signals in any suitable part of a subject’s anatomy.

[0153] In a further aspect, there is described an implantable neurostimulator comprising any of the magnetoelectric antennas described herein. In use, the magnetoelectric antenna is configured to receive a stimulation signal from an ex vivo device and stimulate neurons in a tissue into which the neurostimulator has been implanted in accordance with the received stimulation signal.

[0154] The implantable neurostimulator may comprise one or more of the magnetoelectric antennas as described herein.8673170

[0155] 16

[0156] The implantable neurostimulator may have an operating bandwidth of between 200 MHz and 15 GHz, or between 200MHz and 30 GHz. For example, the implantable neuroscanner may have an operating bandwidth of 22.6 GHz.

[0157] The transmission strength of each magnetoelectric antenna of the neurostimulator may be within the range of -10 dBm to 10 dBm.

[0158] The magnetoelectric antenna(s) may be mounted on a PCB. The magnetoelectric antenna(s) may be mounted on a PCB via a silver nanoparticle ink (e.g., an ethylene glycol-based silver nanoparticle ink). The resistivity of the silver nanoparticle ink may be 7.9pQcm. In this way, the implantability of the implantable neuroscanner may be improved.

[0159] In use, an external electromagnetic field may be transmitted to each magnetoelectric antenna of the neurostimulator (e.g., from an ex vivo device). This may may excite the one or more magnetostrictive layers of each of the magnetoelectric antennas of the neuroscanner. This induces a physical (strain) response in the one or more magnetoelectric layers that is subsequently transferred (e.g., as an acoustic wave) into the piezoelectric layer of the corresponding antenna. The transferred strain may induce a potential difference within the piezoelectric layer that causes the establishment of a potential difference across the first and second electrodes of the antenna, thereby generating an electrical signal that may be delivered to the subject as a stimulation signal via the first electrode, second electrode, or a further electrode that is electrically connected to the magnetoelectric antenna.

[0160] For the avoidance of doubt, the implantable neurostimulators described herein need not be neurostimulators per se, but rather may be an implantable stimulator for electrically stimulating any suitable tissue in a subject.

[0161] In use, the magnetoelectric antenna of the neurostimulator may be further configured to record electrical signals emitted by the tissue into which the neurostimulator has been implanted, and transmit the recorded electrical signals to the or another ex vivo device.

[0162] That is, an implantable device comprising one or more magnetoelectric antennas as described herein may be simultaneously configured with the functionality of both the implantable scanners described herein and the implantable stimulators described herein.

[0163] In such cases, the implantable device may comprise a first one or more magnetoelectric antennas configured (e.g., by virtue of their connecting circuitry) to operate as scanning antennas, and a second one or more magnetoelectric antennas configured (e.g., by virtue of their connecting circuitry) to operate as stimulating antennas. Additionally or alternatively, one or more of the magnetoelectric antennas of the implantable device may be configured to operate simultaneously or selectively as both scanning antennas and stimulating antennas.

[0164] In a further aspect, there is provided a system comprising one or more of the implantable devices (e.g., one or more neurostimulators and / or one or more neuroscanners) described herein and an ex vivo device, wherein the ex vivo device may be configured to transmit the electromagnetic signal to the implantable devices and / or to receive the electrical signal from the implantable devices. The ex vivo8673170

[0165] 17

[0166] device may be configured to transmit to the implantable devices an electromagnetic signal having a frequency of between 3 GHz and 4 GHz, e.g., a frequency between 3.0 GHz and 3.6 GHz e.g., a frequency of 3.1 GHz, 3.25 GHz, 3.3 GHz and / or 3.9 GHz. The ex vivo device may be configured to receive from the implantable devices an electrical signal having a frequency between 3 GHz and 4 GHz, e.g., a frequency between 3.0 GHz and 3.6 GHz e.g., a frequency of 3.1 GHz, 3.25 GHz, 3.3 GHz and / or 3.9 GHz.

[0167] In this way, the wireless links between the implantable devices and the ex vivo device may be more reliable when the implantable device is implanted in tissue, It will be understood that the benefits and advantages described above may relate to one embodiment or may relate to several embodiments. The embodiments are not limited to those that solve any or all of the stated problems or those that have any or all of the stated benefits and advantages. Variants should be considered to be included into the scope of the invention.

[0168] The invention includes the combination of the aspects and preferred features described except where such a combination is clearly impermissible or expressly avoided.

[0169] Summary of the Figures

[0170] Embodiments and experiments illustrating the principles of the invention will now be discussed with reference to the accompanying figures in which:

[0171] Figure 1 shows an example of an implantable device as described herein.

[0172] Figure 2 shows a layout of a magnetoelectric antenna as described herein.

[0173] Figure 3 shows a schematic of the layers of a magnetoelectric antenna.

[0174] Figure 4 shows a schematic of the layers of a magnetoelectric antenna.

[0175] Figure 5 shows a schematic of the layers of a magnetoelectric antenna.

[0176] Figure 6 shows a schematic of the layers of a magnetoelectric antenna.

[0177] Figure 7 shows a schematic of the layers of a magnetoelectric antenna.

[0178] Figure 8 shows a schematic of the layers of a magnetoelectric antenna.

[0179] Figure 9 shows a schematic of the layers of a magnetoelectric antenna.

[0180] Figure 10 shows a schematic of the layers of a magnetoelectric antenna.

[0181] Figure 11 shows a schematic of the layers of a magnetoelectric antenna.

[0182] Figure 12 shows a schematic of the layers of a magnetoelectric antenna.

[0183] Figure 13 shows a schematic of the layers of a magnetoelectric antenna.

[0184] Figure 14 shows a schematic of the layers of a magnetoelectric antenna.8673170

[0185] 18

[0186] Figure 15 compares the frequency responses of two magnetoelectric antennas. A. the frequency response of a magnetoelectric antenna having a single resonant frequency. B. the frequency response of a magnetoelectric antenna having at least two resonant frequencies.

[0187] Figure 16 illustrates patterning layouts for complementary pairs of first and second electrodes of the magnetoelectric antennas described herein.

[0188] Figure 17A illustrates a pBot federation wireless platform for bio-integrable communication consisting of packaged overtone ultrawideband (OUWB) magnetoelectric (ME) antennas with active area dominated by the circular components in green. A distributed network of ME nodes are embedded in tissue enabling wireless sensing / telemetry to an external receiver. The inset shows the oustic wave pathway in the polished substrate with lowered stress mounting the OUWB-ME antenna.

[0189] Figure 17B illustrates a microfabricated OUWB-ME antenna at the fingertip scale and mounted on a ceramic PCB (inset, scale bar: 1 mm) via ethylene glycol-based silver nanoparticle inks capable of RF transmission.

[0190] Figure 17C illustrates the device architecture consisting of AIN / Fe0.79Ga0.21 (110 BCC) engineered to stiffness mismatch of 1:4.82 to enhance bandwidth of antenna for audio-visual transmission of signals while miniaturizing size.

[0191] Figure 17D illustrates a packaged OUWB-ME antenna (pBot) connected to Universal Software Radio Peripheral (USRP) platform (scale bar: 1 cm). The precession of particles, under an external magnetic field Heff in the ferromagnetic layer is influenced by the conservative and dissipative magnetic torques of the magnetostrictive thin film.

[0192] Figure 17E shows a volumetric image projection post 7T-MRI of 6 pBots implanted in Agar.

[0193] Fig. 18A illustrates the ultrawideband reflection parameter of OUWB-ME antennas operating primarily in (i) overtone acoustic region, (ii) higher-order acoustic harmonic(s) and onset of self-resonance dominated by parasitic inductance and (iii) self-resonance wherein the electromagnetic field transmission dominates acoustic modes and is largely dominated by piezoelectric layer thickness.

[0194] Figures 18B-C illustrate the experimental and modeled impedance spectra in the overtone and higher order harmonic region of the OUWB-ME antennas that are utilized for extraction of acoustic properties of piezoelectric and magnetostrictive materials in the RF regime via the frequency and dynamic range. The absolute value of the experimental impedance is prone to manual artifacts such as probe landing and bond quality leading to mismatch with theoretical results in magnitude.

[0195] Figure 18D shows a comparison of reflection parameter between piezoelectric control resonators and OUWB-ME antennas highlighting the improvement in return loss. Shaded band represents ± 5% uncertainty error primarily due to calibration the trace is apparently thicker due to reproducible small oscillations of the trace as seen in Fig. 18B and Fig. 18C.8673170

[0196] 19

[0197] Figure 18E illustrates the reflection parameter of packaged OUWB-ME antennas (pBots) with nanoparticle-based inks prior to and after cyclic ex vivo testing under severed rat head, explanted rat brain, and human cortical autoptic tissue highlighting bonding stability.

[0198] Fig.19A shows normalized cell viability, measured as absorbance at 450 nm, of cells plated on the different substrates. N = 3-6 independent cultures. ****p < 0.0001 in comparison with the control condition at the same DIV using a two-way ANOVA (days in vitro vs. substrate) followed by Tukey’s multiple comparison test.

[0199] Figure 19B shows representative fluorescent images of primary neuronal cultures plated on glass, and silicon wafers coated with AIN, FeGa, and AIN / FeGa / Parylene. On top row, p3-Tubulin staining. On bottom row, neurons are represented in green (p3-Tubulin), astrocytes in magenta (GFAP), and nuclei in blue (DAPI). Scale bar = 50 pm.

[0200] Figure 19C illustrates a quantitative analysis of cell density, ratio between neurons and astrocytes, and relative area covered by the different cell types. N = 3-6 independent cultures. ***p < 0.001 and ****p < 0.0001 in comparison with the control condition, using a one-way ANOVA followed by Dunnett’s multiple comparison test.

[0201] Figure 19D shows representative calcium signals measured from cells growing on glass, AIN, and FeGa. Scale bar = 30 s.

[0202] Figure 19E shows a quantitative analysis of calcium transient duration, amplitude, and frequency (represented as inter-spike interval, ISI). N = 3-4 independent cultures. *p < 0.05 using a paired t-test between experimental condition and respective control.

[0203] Figure 20A shows a representative S12 (transmission) measurement for blank pBot (black trace) and pBot with fresh rat brain (blue trace). X-scale bar = 0.1 GHz. Y-scale bar = 10 dB. The same biological condition measured in the different pBots were normalized in respect to their respective blank pBot measurement, generating a fold-change (FC) plot.

[0204] Figure 20B shows a representative S22 (reflection) measurement for blank pBot (black trace) and pBot with fresh rat brain (blue trace). X-scale bar = 0.1 GHz. Y-scale bar = 2 dB. Similarly to S12 data, sample measurements were normalized in relation to their respective blank antenna measurement, creating the FC dataset.

[0205] Figure 20C shows the S22 signal denoised to quantify overtone strength,. X-scale bar = 0.2 GHz. Y-scale bar = 0.001 absolute magnitude. From the denoised data, the overtone amplitude was calculated and used to generate the overtone strength plot.

[0206] Figure 21A illustrates S12 fold-change (FC) for each biological condition. Data represented as mean ± SEM. N = 3-6 different pBots. Horizontal black lines represent significant differences between each condition and baseline.

[0207] Figure 21B shows changes in S12 at specific frequencies (3.1, 3.27, and 3.95 GHz). Data represented as mean ± SEM, with individual pBots measurements scattered as black dots. *p < 0.05, **p < 0.01 vs.8673170

[0208] 20

[0209] baseline using Kruskal-Wallis test followed by the two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli.

[0210] Figure 22A illustrates S22 fold-change (FC) for each biological condition. Data represented as mean ± SEM. N = 3-6 different pBots. Horizontal black lines represent significant differences between each condition and its control. All samples are mounted on non-conductive building blocks

[0211] Figure 23A shows representative overtone strength dynamics for baseline blank pBot (black trace), severed rat head (green trace), and explanted rat cold brain (purple trace) over the ME frequency range.

[0212] Figure 23B shows overall overtone strength frequency shift identified by cross-correlation over the ME range between each condition and its respective baseline.

[0213] Figure 23C shows an identification of main overtone strength features, consisting of two negative (N1 and N2) and two positive (P1 and P2) peaks.

[0214] Figure 23D shows a quantification of frequency shift (i) and overtone attenuation (ii) of the identified signal features. Data represented as mean ± SEM, with individual samples scattered in black dots. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 in relation to baseline using one-way ANOVA followed by Dunnet’s multiple comparison test.

[0215] Figure 23E shows a 2D representation of attenuation and frequency shifts caused by the different biological (and sham) samples.

[0216] Figure 23F shows a Decomposition of sample impact on the different identified signal features: (i) sham control and severed rat head, (ii) explanted rat brain samples, and (iii) autoptic human samples. The filled dot corresponds to N1, sequentially connected to P1 , N2, and P2.

[0217] Figure 24A illustrates Prototype link hardware demonstrating wireless audio-visual telemetry (A) Representational schematic of pBot (packaged OUWB-ME antenna) systems mounted between the transmitter and receiver front-ends (SI Note 1).

[0218] Figure 24B shows normalized received power versus rotation angle for ME-ME (blue), ME-RF (red) and RF-RF (black) configurations. Shaded bands indicate confidence intervales (darker shade: 68%, lighter shade: 95%).

[0219] Figure 24C shows receiver- power distributions for ME-RF (top) and RF-RF (bottom) under identical tilts; ME-RF clusters higher and tighter, indicating reduced sensitivity to out-of-plane misalignment.

[0220] Figure 24D shows a misalignment heat-map indicating the number of pBots, N, required to obtain RF-like stability. Cl half-width (dB re. peak) as a function of N and confidence multiplier, Z (1 = 68%, 1.96 = 95%) with the surface capturing 1A / averaging.

[0221] Figure 24E illustrates spectral fidelity of ME-ME (top) and ME-RF links (bottom) at (0° , 90° ) overlap (center frequency: 3.25 GHz) showing minimal angle dependent-distortion of the DVB-S / QPSK waveform. Overlaid spectra at 0° and 90° confirm that hybrid links preserve the transmitted spectrum while benefitting from the misalignment robustness in Figure 24C.8673170

[0222] 21

[0223] Figure 24F illustrates a real-time audio-visual sonogram streaming over a magnetoelectric link. The receiver is tuned to 3.25 GHz (span 4.0 MS / s); the left and right panels show the RF spectrum and waterfall for the transmitter and receiver, respectively. In both cases the waveform exhibits the same Root Raised Cosine shaped occupied band (0.675 MHz) centered at 3.25 GHz, with a stable vertical stripe in the waterfall indicating continuous lock. Center: decoded ultrasound sonogram with heart beat audio confirming end-to-end transport (SI Video 9, SI Note 2).

[0224] Figure 24G shows an MRI of OUWB-ME embedded in Agar. Left: Bruker PharmaScan 7T system and phantom placement. Right: XY, XZ and YZ axial slices with device locations circles after 10-hour scan. The OUWB-ME antennas produce small, localized contrast “voids” at the implant sites without gross geometric distortion or large-scale signal non- uniformity, indicating practical MRI compatibility. Scale bar: 0.4 mm.

[0225] Figure 25A shows the variation in effective transmission of OUWB-ME antennas during overtone with respect to device position (shown by green arrow) and corresponding strength of magnetic field applied during deposition. The difference reaches near 0 at the center of the sample where the field is minimum. The variation is subtracted from control piezoelectric device response and normalized by one-port reflection parameter of the horn antenna.

[0226] Figure 25B shows the raw received power peaking at ~3.2 GHz within the overtone range when the OUWB-ME antenna is connected as a receiver to a spectrum analyzer. The transmitting antenna is a horn antenna why the magnetic field oriented parallel to the easy-axis / in-situ DC field axis.

[0227] Figure 26 is a table showing RF mechanical properties of the materials composing the AIN / FeGa magnetoelectric antennas required for the modified Sittig’s model. indicates the properties that are not required in the first order model for determining the overtone frequency and coupling coefficients. The elastic and piezoelectric constants are in the longitudinal mode.

[0228] Figure 27 shows an XRD of sputtered FeGa (110) at 43.94° affiliated to increased magnetostriction due to co-existence of D03 and A2 phases.

[0229] Figure 28 shows an SEM of ME films with columnar grains deposited on low-stress polished Si.

[0230] Figure 29 shows reflection coefficient (S11) of piezoelectric resonators and OUWB-ME consisting of (i) high overtone region, (ii) spurious region and self-resonance, (iii) self-resonance. Inset highlighting the two distinct heterostructure stacks required for accurate modeling for derivation of RF electromechanical properties and stiffness.

[0231] Figure 30 shows in-plane and out-of-plane magnetic hysteresis loop of Feo.79Gao.21 film.

[0232] Figure 31 shows a comparison of as-diced OUWB-ME antenna and ink-bonded OUWB-ME antennas (pBot). The phase shift due to the conductive ink and PCB traces introduces additional resonance peaks outside the magnetoelectric overtone range of interest spanning 3-4 GHz (more primarily 3.1-3.6 GHz) Figure 32 illustrates biological samples used in the pBot transmission and reflection measurements. Figure 33 shows blank S12 and S22 measurements for all pBots.8673170

[0233] 22

[0234] Figure 34 shows isolated overtones (fast oscillation component) for open calibration and blank pBot measurements. X-scale bar = 0.2 GHz. Y-scale bar = 0.001 absolute magnitude.

[0235] Figure 35 shows representative S12 traces for all tested conditions. Respective blank measurement in black. X-scale = 0.1 GHz. Y-scale bar = 10 dB.

[0236] Figure 36 shows representative S22 traces for all tested conditions. Respective blank measurement in black. X-scale = 0.1 GHz. Y-scale bar = 2 dB.

[0237] Figure 37 shows detailed representative S22 traces for all tested conditions, highlighting overtones. Respective blank measurement in black. X-scale = 0.1 GHz. Y-scale bar = 0.5 dB.

[0238] Figure 38 shows the average frequency and strength of blank pBot overtones.

[0239] Figure 39 shows the decomposition of each biological tissue impact on the different signal features. Data presented as mean ± SEM. The filled dot represents N1 and is then sequentially connected to P1 , N2, and P2.

[0240] Figure 40 is a laser microscopy image of printed ethylene-glycol based nanoparticle bonding (scale bar: 20 urn).

[0241] Detailed Description of the Invention

[0242] Aspects and embodiments of the present invention will now be discussed with reference to the accompanying figures. Further aspects and embodiments will be apparent to those skilled in the art. All documents mentioned in this text are incorporated herein by reference.

[0243] Figure 1 shows an implantable device 100. The implantable device 100 may be operable as an implantable scanner (e.g., a neuroscanner) and / or as an implantable stimulator (e.g., a neurostimulator). The implantable device 100 comprises a casing 102 encapsulating one or more magnetoelectric antennas 200 and other control circuitry and components of the device 100 (other circuitry and components not shown).

[0244] The implantable device 100 may be powered by any suitable means. For example, the implantable device may be powered by an on-board battery and / or an antenna configured to generate power from received signals (e.g., received RF signals). For example, this antenna may be a magnetoelectric antenna 200 as described herein, or may be a rectenna having any suitable configuration. In cases where a magnetoelectric antenna 200 is used to harvest signals to power the implantable device, the control circuitry of the implantable device 100 may include one or more rectifiers to facilitate the conversion of the received signals into direct current for powering the implantable device 100. In some cases, e.g., in cases where the implantable device 100 is operable as a stimulator, the implantable device 100 may include one or more stimulation electrodes for stimulating tissue therewith. The one or more stimulating electrodes may extend from, be contiguous with, and / or form part of the casing 102.

[0245] The implantable device 100 is sized on the microscale for efficient and non-disruptive implantation into a subject / patient. For example, the implantable device may have a volume of 1 mm3or less, 0.1 mm3or8673170

[0246] 23

[0247] less, 0.01 mm3or less, or 0.001 mm3or less. Alternatively, the implantable device 100 may have a volume of 0.0001 mm3or more, 0.001 mm3or more, 0.01 mm3or more, or 0.1 mm3or more. Alternatively, the implantable device may have a volume between 0.0001 and 1 mm3, between 0.0001 and 0.1 mm3, between 0.0001 and 0.01 mm3, between 0.0001 and 0.001 mm3, between 0.001 and 1 mm3, between 0.001 and 0.1 mm3, between 0.001 and 0.01 mm3, between 0.01 and 1 mm3, between 0.01 and 0.1 mm3, or between 0.1 and 1 mm3.

[0248] For example, in cases where the implantable device 100 has a cubic body, the cubic body may have a side length of 1 mm or less, 500 pm or less, or 100 pm or less. Alternatively, the cubic body may have a side length of 50 pm or more, 100 pm or more, or 500 pm or more. Alternatively, the cubic body may have a side length between 50 pm and 1 mm, between 50 pm and 500 pm, between 50 pm and 100 pm, between 100 pm and 1 mm, between100 pm and 500 pm, or between 500 pm and 1 mm.

[0249] Figure 2 shows a layout of a magnetoelectric antenna 200. As can be seen from Figure 2, the total area of the magnetoelectric antenna is sub-mm. Meanwhile, the active area 202 of the magnetoelectric antenna may have an active area of 0.05 mm2or less, for example, 0.03 mm2. The active area 202 of the magnetoelectric antenna is an area of the magnetoelectric antenna 200 within which each layer of a stack of layers of the magnetoelectric antenna overlaps to facilitate the conversion of a received electromagnetic signal into a corresponding electrical signal (for transmission and subsequent analysis, and / or for stimulation). The layers of the magnetoelectric antenna 200 will be discuss below in relation to Figures 3 to 14.

[0250] Figure 3 shows a schematic of layers of a magnetoelectric antenna 300. The magnetoelectric antenna 300 comprises a first electrode, the first electrode comprising a first metal contact 302, a piezoelectric layer 304, an adhesion layer 306, a magnetostrictive layer 308, a capping layer 310 and a second electrode, the second electrode comprising a second metal contact 312.

[0251] The first metal contact 302 has a thickness of approximately 50 nm. This may lead to the amplification of overtones in the magnetoelectric antenna’s response. The first electrode may be formed from a suitably conductive, and optionally biocompatible, material such as platinum. The first metal contact 302 may further comprise a metal contact adhesion layer formed, for example, from titanium to adhere the first metal contact 302 to the piezoelectric layer 304.

[0252] The piezoelectric layer 304 may be formed from any suitable piezoelectric material. For example, the piezoelectric layer 304 may be formed from PZT, or any suitable crystalline piezoelectric material (e.g., GaN, InN, AIN, and / or ZnO). The piezoelectric layer 304 may have a suitable crystalline structure - e.g., a zincblende or wurtzite structure.

[0253] The adhesion layer 306 adheres the magnetostrictive layer 308 to the piezoelectric layer 304. The adhesion layer may be formed from any suitable adhering material, such as titanium, nickel, tantalum, platinum, or alloys thereof such as nichrome.8673170

[0254] 24

[0255] The magnetostrictive layer 308 may be formed from any suitable material that exhibits magnetostrictive behaviour. For example, the magnetostrictive layer may be formed from iron, FeCo, FeGaB, galfenol, alfenol or terfenol-D. The magnetostrictive layer 308 may have a thickness between 30 and 60 nm. The capping layer 310 may be formed from any suitable material that may reduce the risk of physical damage to the magnetostrictive layer 308. Additionally, the capping layer 310 may promote coupling between the magnetostrictive layer 308 and the second metal contact 312. The capping layer 310 may, for example, be formed from aluminium. That capping layer may, for example have any suitable thickness (e.g., between 1 and 20 nm - for example, 5 nm).

[0256] The second metal contact 312 has a thickness of approximately 50 nm. This may lead to an amplification of overtones in the magnetoelectric antenna’s response. The second electrode 312 may be formed from a suitably conductive, and optionally biocompatible, material such as platinum ora gold / platinum alloy. The second metal contact 312 may further comprise a metal contact adhesion layer formed, for example, from FeGa to adhere the second metal contact 312 to the magnetostrictive layer 308.

[0257] Figure 4 shows another schematic of layers of a magnetoelectric antenna 400. The magnetoelectric antenna 400 comprises a first electrode, the first electrode comprising a first metal contact 402, a piezoelectric layer 404, an adhesion layer 406, a magnetostrictive layer 408, a capping layer 410, and a second electrode, the second electrode comprising a second metal contact 412.

[0258] The magnetoelectric antenna 400 of Figure 4 differs from the magnetoelectric antenna 300 of Figure 3 in that, in the case of the magnetoelectric antenna 400 of Figure 4, the second metal contact 412 physically interfaces, directly, with the piezoelectric layer 404, whereas, in the case of the magnetoelectric antenna 300 of Figure 3, the adhesion layer 306, magnetostrictive layer 308 and capping layer 310 are provided between the piezoelectric layer 304 and the second metal contact 312.

[0259] In such cases, if the second metal contact 412 comprises a metal contact adhesion layer, the metal contact adhesion layer of the second metal contact 412 may be formed from the same material as described above in relation to the metal contact adhesion layer of the first metal contact 312 (described in relation to Figure 3) to promote adhesion of the second metal contact 412 to the piezoelectric layer 404.

[0260] Figure 5 shows another schematic of layers of a magnetoelectric antenna 500. The magnetoelectric antenna 500 comprises a first electrode, the first electrode comprising a first metal contact 502, a piezoelectric layer 504, an adhesion layer 506, a magnetostrictive stack 508, a capping layer 510, and a second electrode, the second electrode comprising a second metal contact 512.

[0261] The magnetoelectric antenna 500 of Figure 5 differs from the magnetoelectric antenna 300 of Figure 3 in that the magnetoelectric antenna 500 of Figure 5 comprises a plurality of magnetostrictive layers 508a-c arranged to define the magnetostrictive stack 508.

[0262] The magnetostrictive stack 508 comprises a plurality of magnetostrictive layers 508a-c. Each adjacent pair of magnetostrictive layers 508a-c is separated by a respective laminating layer 514a-b. In the example shown in Figure 5, the magnetostrictive stack 508 comprises three magnetostrictive layers8673170

[0263] 25

[0264] 508a-c and two laminating layers 514a-b, but other numbers of magnetostrictive and laminating layers are possible.

[0265] Each magnetostrictive layer 508a-c of the magnetostrictive stack 508 may have a thickness of between 30nm and 60 nm. Each magnetostrictive layer 508a-c may be formed from the same material. This material may be any suitable material that exhibits magnetostrictive behaviours - for example, iron, FeCo, FeGaB, galfenol, alfenol or terfenol-D.

[0266] Each laminating layer 514a-b of the magnetostrictive stack may be formed from a suitable dielectric (e.g., aluminium nitride) so as to reduce eddy current effects and prevent short-circuiting within the magnetoelectric stack 508. Each of the laminating layers 514a-b may have any suitable thickness (e.g., between 1 and 10 nm, for example 5 nm).

[0267] Figure 6 shows another schematic of layers of a magnetoelectric antenna 600. The magnetoelectric antenna 600 comprises a first electrode, the first electrode comprising a first metal contact 602, a piezoelectric layer 604, an adhesion layer 606, a magnetostrictive stack 608, a capping layer 610, and a second electrode, the second electrode comprising a second metal contact 612.

[0268] The magnetostrictive stack 608 comprises a plurality of magnetostrictive layers 608a-c. Each adjacent pair of magnetostrictive layers 608a-c is separated by a respective laminating layer 614a-b. In the example shown in Figure 6, the magnetostrictive stack 608 comprises three magnetostrictive layers 608a-c and two laminating layers 614a-b, but other numbers of magnetostrictive and laminating layers are possible.

[0269] The magnetoelectric antenna 600 of Figure 6 differs from the magnetoelectric antenna 500 of Figure 5 in that, the case of the magnetoelectric antenna 600 of Figure 6, the second metal contact 612 physically interfaces, directly, with the piezoelectric layer 604, whereas, in the case of the magnetoelectric antenna 500 of Figure 5, the adhesion layer 506, magnetostrictive stack 507 and capping layer 510 are provided between the piezoelectric layer 504 and the second metal contact 512.

[0270] Figure 7 shows another schematic of layers of a magnetoelectric antenna 700. The magnetoelectric antenna 700 comprises a first electrode, the first electrode comprising a first metal contact 702, a piezoelectric layer 704, an adhesion layer 706, a magnetostrictive layer 708, a capping layer 710, and a second electrode, the second electrode comprising a second metal contact 712.

[0271] The magnetoelectric antenna 700 is formed on a substrate 716, and interfaced to the substrate 716 by a buffer layer 718.

[0272] The first electrode includes (e.g., the first metal contact 702 is deposited on) an acoustic Bragg reflector 720. The acoustic Bragg reflector 720 interfaces with the first metal contact 702 via an intermediate layer 722.

[0273] The substrate 716 may be a crystalline substrate - e.g., a substrate oriented in the (100) crystal plane. The substrate 716 may be formed from any suitable material, such as silicon, SiC, sapphire and / or GaN.8673170

[0274] 26

[0275] The buffer layer 718 may reduce a strain mismatch between the substrate 716 and the magnetoelectric antenna 700 (e.g., the acoustic Bragg reflector 720). The buffer layer 718 may be formed from any suitable material for this purpose, e.g., aluminium nitride, silicon nitride, silicon oxide (or silica), and / or silicon oxynitride. The buffer layer may have any suitable thickness (e.g., between 50 and 500 nm, for example 100 nm).

[0276] The acoustic Bragg reflector 720 is formed from a series of alternating layers of first and second material, the first and second materials having respectively different acoustic impedances. In some cases, the thickness of the layers of the first and second materials may be uniform throughout the acoustic Bragg reflector, although in other cases the thicknesses may vary throughout the acoustic Bragg reflector as described above.

[0277] The acoustic Bragg reflector 720 may be an all-metal acoustic Bragg reflector. For example, the acoustic Bragg reflector 720 may be formed from alternating layers of aluminium and tungsten.

[0278] A first layer of the acoustic Bragg reflector 720 (proximal to the buffer layer 718) may be a 494 nm thick layer of aluminium. A second layer (deposited on the first layer) may be a 482 nm thick layer of tungsten. A third layer (deposited on the second layer) may be a 494 nm thick layer of aluminium. A fourth layer (deposited on the third layer) may be a 241 nm thick layer of tungsten. A fifth layer (deposited on the fourth layer) may be a 741 nm thick layer of aluminium. A sixth layer (deposited on the fifth layer) may be a 120.5 nm thick layer of tungsten. A seventh layer (deposited on the sixth layer) may be a 247 nm thick layer of aluminium. An eighth layer (deposited on the seventh layer) may be a 241 nm thick layer of tungsten. It has been found that this arrangement for the acoustic Bragg reflector demonstrates strong reflectance within the operating bandwidth of the magnetoelectric antennas described herein.

[0279] The intermediate layer 722 may couple the acoustic Bragg reflector 720 to the first metal contact 702 and / or the piezoelectric layer 704. The intermediate layer 722 may be formed from a suitable dielectric to reduce the risk of electrical shorting between the first metal contact 702 and the acoustic Bragg reflector 720. The intermediate layer 722 may also be formed from a material selected to have a similar acoustic impedance to one of the materials from which the acoustic Bragg reflector 720 is formed. For example, the intermediate layer 722 may be formed from silicon oxynitride.

[0280] The magnetoelectric antenna 700 of Figure 7 differs from the magnetoelectric antenna 300 of Figure 3 by the provision of the acoustic Bragg reflector 720 and the intermediate layer 722. In some cases, any of the magnetoelectric antennas described herein may be provided on a substrate, with which the antenna is interfaced via a buffer layer, such as the substrate 716 and buffer layer 718 described in relation to Figure 7.

[0281] Figure 8 shows another schematic of layers of a magnetoelectric antenna 800. The magnetoelectric antenna 800 comprises a first electrode, the first electrode comprising a first metal contact 802, a piezoelectric layer 804, an adhesion layer 806, a magnetostrictive stack 808, a capping layer 810, and a second electrode, the second electrode comprising a second metal contact 812.8673170

[0282] 27

[0283] The magnetostrictive stack 808 comprises a plurality of magnetostrictive layers 808a-c. Each adjacent pair of magnetostrictive layers 808a-c is separated by a respective laminating layer 814a-b. In the example shown in Figure 8, the magnetostrictive stack 808 comprises three magnetostrictive layers 808a-c and two laminating layers 814a-b, but other numbers of magnetostrictive and laminating layers are possible.

[0284] The magnetoelectric antenna 800 is formed on a substrate 816, and interfaced to the substrate 816 by a buffer layer 818.

[0285] The first electrode includes (e.g., the first metal contact 802 is deposited on) an acoustic Bragg reflector 820. The acoustic Bragg reflector 820 interfaces with the first metal contact 802 via an intermediate layer 822.

[0286] The magnetoelectric antenna 800 of Figure 8 differs from the magnetoelectric antenna 700 of Figure 7 in that the magnetoelectric antenna 800 of Figure 8 comprises a plurality of magnetostrictive layers 808a-c arranged to define the magnetostrictive stack 808.

[0287] Figure 9 shows another schematic of layers of a magnetoelectric antenna 900. The magnetoelectric antenna 900 comprises a first electrode, the first electrode comprising a first metal contact 902, a piezoelectric layer 904, an adhesion layer 906, a magnetostrictive layer 908, a capping layer 910, and a second electrode, the second electrode comprising a second metal contact 912.

[0288] The magnetoelectric antenna 900 is formed on a substrate 916, and interfaced to the substrate 916 by a buffer layer 918.

[0289] The first electrode includes (e.g., the first metal contact 902 is deposited on) an acoustic Bragg reflector 920. The acoustic Bragg reflector 920 interfaces with the first metal contact 902 via an intermediate layer 922.

[0290] The magnetoelectric antenna 900 of Figure 9 differs from the magnetoelectric antenna 700 of Figure 7 in that, in the case of het magnetoelectric antenna 900 of Figure 9, the second metal contact 912 physically interfaces, directly, with the piezoelectric layer 904, whereas, in the case of the magnetoelectric antenna 700 of Figure 7, the adhesion layer 706, the magnetostrictive layer 908 and capping layer 710 are provided between the piezoelectric layer 704 and the second metal contact 712.

[0291] Figure 10 shows another schematic of layers of a magnetoelectric antenna 1000. The magnetoelectric antenna 1000 comprises a first electrode, the first electrode comprising a first metal contact 1002, a piezoelectric layer 1004, an adhesion layer 1006, a magnetostrictive stack 1008, a capping layer 1010, and a second electrode, the second electrode comprising a second metal contact 1012.

[0292] The magnetostrictive stack 1008 comprises a plurality of magnetostrictive layers 1008a-c. Each adjacent pair of magnetostrictive layers 808a-c is separated by a respective laminating layer 1014a-b. In the example shown in Figure 10, the magnetostrictive stack 1008 comprises three magnetostrictive layers 1008a-cand two laminating layers 1014a-b, but other numbers of magnetostrictive and laminating layers are possible.8673170

[0293] 28

[0294] The magnetoelectric antenna 1000 is formed on a substrate 1016, and interfaced to the substrate 1016 by a buffer layer 1018.

[0295] The first electrode includes (e.g., the first metal contact 1002 is deposited on) an acoustic Bragg reflector 1020. The acoustic Bragg reflector 1020 interfaces with the first metal contact 1002 via an intermediate layer 1022.

[0296] The magnetoelectric antenna 1000 of Figure 10 differs from that of Figure 9 in that the magnetoelectric antenna 1000 of Figure 10 comprises a plurality of magnetostrictive layers 1008a-c arranged to define the magnetostrictive stack 1008.

[0297] Figure 11 shows another schematic of layers of a magnetoelectric antenna 1100. The magnetoelectric antenna 1100 comprises a first electrode, the first electrode comprising a first metal contact 1102, a piezoelectric layer 1104, an adhesion layer 1106, a magnetostrictive layer 1108, a capping layer 1110, and a second electrode, the second electrode comprising a second metal contact 1112.

[0298] The second electrode further comprises a third metal contact 1124 co-planar with the second metal contact 1112, having a thickness different from (e.g., greater than) the thickness of the second metal contact 1112, such that a spacing between the first metal contact 1102 and the third metal contact 1124 is different from (e.g., less than) the corresponding spacing between the first metal contact 1102 and the second metal contact 1112. In this way, as described above, the magnetoelectric antenna 1100 of Figure 11 may be operable with at least two resonant frequencies, whereas a magnetoelectric antenna 1100 having a second electrode comprising a single metal contact (e.g., the magnetoelectric antenna 300 of Figure 3) may be operable with a single resonant frequency.

[0299] The third electrode 1124 may be formed from a suitably conductive, and optionally biocompatible, material such as platinum ora gold / platinum alloy. The third metal contact 1124 may further comprise a metal contact adhesion layer formed, for example, from FeGa to adhere the second metal contact 1124 to the magnetostrictive layer 308.

[0300] The magnetoelectric antenna 1100 of Figure 11 differs from the magnetoelectric antenna 300 of Figure 3 in that the second electrode comprises both the second metal contact 1112 and the third metal contact 1124.

[0301] Figure 12 shows another schematic of layers of a magnetoelectric antenna 1200. The magnetoelectric antenna 1200 comprises a first electrode, the first electrode comprising a first metal contact 1202, a piezoelectric layer 1204, an adhesion layer 1206, a magnetostrictive stack 1208, a capping layer 1210, and a second electrode, the second electrode comprising a second metal contact 1212.

[0302] The magnetostrictive stack 1208 comprises a plurality of magnetostrictive layers 1208a-c. Each adjacent pair of magnetostrictive layers 1208a-c is separated by a respective laminating layer 1214a-b. In the example shown in Figure 12, the magnetostrictive stack comprises three magnetostrictive layers 1208a-c and two laminating layers 1214a-b, but other numbers of magnetostrictive and laminating layers are possible.8673170

[0303] 29

[0304] The second electrode further comprises a third metal contact 1224 co-planar with the second metal contact 1212, having a thickness different from (e.g., greater than) the thickness of the second metal contact 1212.

[0305] The magnetoelectric antenna 1200 of Figure 12 differs from the magnetoelectric antenna 1100 of Figure 11 in that the magnetoelectric antenna 1200 of Figure 12 comprises a plurality of magnetostrictive layers 1212a-c arranged to define the magnetostrictive stack 1208.

[0306] Figure 13 shows another schematic of layers of a magnetoelectric antenna 1300. The magnetoelectric antenna 1300 comprises a first electrode, the first electrode comprising a first metal contact 1302, a piezoelectric layer 1304, and adhesion layer 1306, a magnetostrictive layer 1308, a capping layer 1310, and a second electrode, the second electrode comprising a second metal contact 1312.

[0307] The magnetoelectric antenna 1300 is formed on a substrate 1316, and interfaced to the substrate 1316 by a buffer layer 1318.

[0308] The first electrode includes (e.g., the first metal contact 1302 is deposited on) an acoustic Bragg reflector 1320. The acoustic Bragg reflector 1320 interfaces with the first metal contact 1302 via an intermediate layer 1322.

[0309] The second electrode further comprises a third metal contact 1324 co-planar with the second metal contact 1312, having a thickness different from (e.g., greater than) the thickness of the second metal contact 1312.

[0310] The magnetoelectric antenna 1300 of Figure 13 differs from the magnetoelectric antenna 1100 of Figure 11 by the provision of the acoustic Bragg reflector 1320 and the intermediate layer 1322.

[0311] Figure 14 shows another schematic of layers of a magnetoelectric antenna 1400. The magnetoelectric antenna 1400 comprises a first electrode, the first electrode comprising a first metal contact 1402, a piezoelectric layer 1404, an adhesion layer 1406, a magnetostrictive stack 1408, a capping layer 1410, and a second electrode, the second electrode comprising a second metal contact 1412.

[0312] The magnetostrictive stack 1408 comprises a plurality of magnetostrictive layers 1408a-c. Each adjacent pair of magnetostrictive layers 1408a-c is separated by a respective laminating layer 1414a-b. In the example shown in Figure 14, the magnetostrictive stack 1408 comprises three magnetostrictive layers 1408a-cand two laminating layers 1414a-b, but other numbers of magnetostrictive and laminating layers are possible.

[0313] The magnetoelectric antenna 1400 is formed on a substrate 1416, and interfaced to the substrate 1416 by a buffer layer 1418.

[0314] The first electrode includes (e.g., the first metal contact 1402 is deposited on) an acoustic Bragg reflector 1420. The acoustic Bragg reflector 1420 interfaces with the first metal contact 1402 via an intermediate layer 1422.8673170

[0315] 30

[0316] The second electrode further comprises a third metal contact 1424 co-planar with the second metal contact 1412, having a different thickness from (e.g., greater than) the thickness of the second metal contact 1412.

[0317] The magnetoelectric antenna 1400 of Figure 14 differs from the magnetoelectric antenna 1300 of Figure 13 in that the magnetoelectric antenna 1400 of Figure 14 comprises a plurality of magnetostrictive layers 1408a-c arranged to define the magnetostrictive stack 1408.

[0318] In Figures 11 to 14, the magnetostrictive layer(s) and capping layer are provided between the piezoelectric layer and the second and third metal contacts of the respective magnetoelectric antenna. Alternatively, one or both of the second and third metal contacts may interface, directly, with the piezoelectric antenna (e.g., as described above in relation to Figures 4, 6, 9 and 10).

[0319] In such cases, if the third metal contact 1124, 1224, 1324, 1424 comprises a metal contact adhesion layer, the metal contact adhesion layer of the third metal contact 1124, 1224, 1324, 1424 may be formed from the same material as described above in relation to the metal contact adhesion layer of the first metal contact 312 (described in relation to Figure 3) to promote adhesion of the third metal contact 1124, 1224, 1324, 1424 to the piezoelectric layer 1104, 1204, 1304, 1404.

[0320] For the avoidance of doubt, each of the following may have the same configuration, arrangement and / or properties: (i) the first metal contacts 302, 402, 502, 602, 702, 802, 902, 1002, 1102, 1202, 1302, 1402 originally described above in relation to Figure 3; (ii) the piezoelectric layers 304, 404, 504, 604, 704, 804, 904, 1004, 1104, 1204, 1304, 1404 originally described above in relation to Figure 3; (iii) the adhesion layers 306, 406, 506, 606, 706, 806, 906, 1006, 1106, 1206, 1306, 1406 originally described above in relation to Figure 3; (iv) the magnetostrictive layers 308, 408, 508a-c, 608a-c, 708, 808a-c, 908, 1008a-c, 1108, 1208a-c, 1308, 1408a-c originally described above in relation to Figure 3; (v) the magnetostrictive stacks 508, 608, 808, 1008, 1208, 1408 originally described above in relation to Figure 5; (vi)the capping layers 310, 410, 510, 610, 710, 810, 910, 1010, 1110, 1210, 1310, 1410 originally described above in relation to Figure 3; (vii) the second metal contacts 312, 412, 512, 612, 712, 812, 912, 1012, 1112, 1212, 1312, 1412 originally described above in relation to Figure 3; (viii) the laminating layers 514a-b, 614a-b, 814a-b, 1014a-b, 1214a-b, 1414a-b originally described above in relation to Figure 5; (ix) the substrates 716, 816, 916, 1016, 1316, 1416 originally described above in relation to Figure 7; (x) the buffer layers 718, 818, 918, 1018, 1318, 1418 originally described above in relation to Figure 7; (xi) the acoustic Bragg reflectors 720, 820, 920, 1020, 1320, 1420 originally described above in relation to Figure 7; (xii) the intermediate layers 722, 822, 922, 1022, 1322, 1422 originally described above in relation to Figure 7; and / or (xiii) the third metal contacts 1124, 1224, 1324, 1424 originally described above in relation to Figure 11.

[0321] Figure 15 compares the frequency response of a single-resonance magnetoelectric antenna (e.g., the magnetoelectric antennas 300, 400, 500, 600, 700, 800, 900, 1000 described above in relation to Figures 3 to 10) to the frequency response of a multi-resonance magnetoelectric antenna (e.g., the magnetoelectric antennas 1100, 1200, 1300, 1400 described above in relation to Figures 11 to 14).8673170

[0322] 31

[0323] As can be seen from Figures 15A (single-resonance) and 15B (multi-resonance), the provision of multiple metal contacts in one or both of the first and second electrodes of the magnetoelectric antenna give rise to additional resonant peaks in the frequency response of the magnetoelectric antenna, making the magnetoelectric antenna more sensitive to a larger range of frequencies within the operating bandwidth of the magnetoelectric antenna.

[0324] Figure 16 illustrates patterning layouts for complementary pairs of first and second electrodes of the magnetoelectric antennas described herein. For example, the first metal contact may be patterned in accordance with pattern 1602 while the second / third metal contact may be patterned in accordance with pattern 1652, or vice versa. Alternatively, the first metal contact may be patterned in accordance with pattern 1604 while the second / third metal contact may be patterned in accordance with pattern 1654, or vice versa. Alternatively, the first metal contact may be patterned in accordance with pattern 1606 while the second / third metal contact may be patterned in accordance with pattern 1656, or vice versa.

[0325] Alternatively, the first metal contact may be patterned in accordance with pattern 1608 while the second / third metal contact may be patterned in accordance with pattern 1658, or vice versa. Alternatively, the first metal contact may be patterned in accordance with pattern 1610 while the second / third metal contact electrode may be patterned in accordance with pattern 1660, or vice versa. Alternatively, the first metal contact may be patterned in accordance with pattern 1612 while the second / third metal contact electrode may be patterned in accordance with pattern 1662, or vice versa. Alternatively, the first metal contact may be patterned in accordance with pattern 1614 while the second / third metal contact electrode may be patterned in accordance with pattern 1664, or vice versa.

[0326] One or more of the piezoelectric layers and the one or more magnetostrictive layers of the magnetoelectric antenna may also be patterned, with a pattern corresponding to the pattern 1602, 1604, 1606, 1608, 1610, 1612, 1614 of the first (or second / third) metal contacts. That patterning of the metal contacts may be chosen to select the resonant frequency (or frequencies) of the magnetoelectric antenna as described above.

[0327] Experimental Results

[0328] The following paragraphs describe a study conducted by the inventors using an implantable device according to the present disclosure. The implantable device is referred to as a " / jBot”.

[0329] OVERVIEW

[0330] Implantable and wearable devices require antennas that are both miniaturized and efficient, yet conventional designs are constrained by narrow bandwidth and orientation sensitivity. We report overtone ultra-wideband magnetoelectric (OUWB-ME) antennas that exploit higher order acoustic modes in polished silicon substrates to achieve a 22.6 GHz bandwidth in the 3-4 GHz range. Packaged into “pBots,” these magnetoelectric heterostructures bonded with silver nanoparticle inks maintain stable operation under biological loading. In vitro assays confirm the biocompatibility of AIN and the protective role of parylene encapsulation for FeGa. Ex vivo rat and human tissues reshape transmission spectra, identifying reproducible frequency windows near 3.3 and 3.9 GHz. pBots enable real-time audiovisual8673170

[0331] 32

[0332] telemetry using software-defined radios and exhibit compatibility with 7T MRI. By combining wideband response, robustness to misalignment, and biocompatible packaging, OUWB-ME pBots provide a scalable platform for wireless bio-integrated communication and telemetry

[0333] INTRODUCTION

[0334] In this work, the inventors present overtone ultra-wideband (OUWB) ME antenna, packaged by precision printing of silver-nanoparticle bonds (Fig. 17A-B). Rather than suppressing artefacts, OUWB-ME antennas deliberately harness typically undesirable, yet often unavoidable, overtone reflections arising in low-stress polished mechanical substrates and magnetoelectric thin-film stiffness mismatch (Fig. 17C) leading to an ultrawide 22.6 GHz bandwidth, the largest reported to date. Using double sided polished silicon, which exhibits a lowered residual stress and supports high-order acoustic overtones, the inventors leverage the electromechanical coupling of BCC Feo79Gao 21 and wurtzite AIN to access modes in the 3-4 GHz range. Operating in the 3-4 GHz band strikes a pragmatic balance between physics and physiology: wavelengths short enough for millimeter-scale, efficient antennas, yet long enough to maintain usable penetration through soft tissue. Compared with crowded sub-3 GHz ISM (Industrial, Scientific, and Medical) allocations, it offers wider contiguous bandwidth and a cleaner interference environment, enabling higher-throughput links for implants and wearables.

[0335] The packaged OUWB-ME antennas, herewith referred to as “pBots”, are interfaced with commercial transceivers to enable audio-visual transmission of sonogram data (Fig. 17D). Given their metallic and ferromagnetic constituents, we assess essential MRI compatibility at 7T. We further conduct systematic in vitro and ex vivo studies in rat and human tissues to confirm biocompatibility and quantify how tissue modulates transmission and overtone strength. Collectively, the analyses provide results towards the design and deployment of magnetoelectric antennas for wireless power and data transfer. Harnessing the standing wave conditions that excite higher-order overtones beyond the fundamental mode, an intrinsic large bandwidth emerges, accompanied by the coexistence of multiple bands that extend communication capabilities beyond conventional resonance modes.

[0336] RESULTS

[0337] Design and Characterization of OUWB-ME Antennas and Bots

[0338] The OUWB-ME antennas consist of 453 nm of AIN and Feo79Gao21 [(25 nm) / AIN(5 nm)]? as the active transduction layer along with a capping layer of 5 nm of Al exhibiting an active area of 0.038 mm2with an ultrawide bandwidth of 22.688 ± 0.256 GHz. The wideband characteristics of the devices can be divided into three distinct regions, (i) the primary overtone region, (ii) the spurious overtone and self-resonance onset followed by (iii) self-resonance as shown in Fig. 18A. Engineering a frequency separation between the magnetoelectric overtone resonances and the self-resonance increases transmission efficiency, mitigating the effects of the piezoelectric layer’s finite impedance. Furthermore, in the overtone region, application of an in-situ DC magnetic field enabled deterministic modulation of the OUWB-ME antenna transmission spectra across devices (Fig. 25A- B). The spatial variation across a single wafer, set by the static magnetic field applied during deposition, provides a practical route to fine-tune OUWB-ME8673170

[0339] 33

[0340] antennas without altering geometry. As seen in Fig. 25A, in the first overtone, the differential transmission metric rises from ~ 0 dB at the low-field centre (device position 8) to 1.23-1.47 dB toward one side (positions 4-5) and reaches 2.63 dB at the opposite edge (position 10). The 2.63 dB extremum corresponds to « 1.8* higher transmitted relative to the centre, yielding a wafer-intrinsic “gain ladder.” In practice, devices can be binned straight from the same sample, ~ 0 dB parts for nominal links, 1-1.5 dB parts to equalize arrays within ± 0.5 dB, and 2.63 dB “hot” parts where margin or misalignment tolerance is most demanding. Because this trimming derives from magnetization order rather than resonant geometry, it preserves the device’s wideband character, enabling deterministic, fabrication-level amplitude setting without re-matching. This magnetic control fosters the assembly of OUWB-ME antenna federations with programmable transmission power within a single fabrication run, thereby advancing scalable deployment. Integration of the spin-orbit-coupled layer, together with the dynamic terminating impedance imparted by the polished substrate, suppresses reflections and, in turn, broadens the operational bandwidth of the OUWB-ME antennas. The deliberate incorporation of a lossy yet low-stress substrate increases acoustic damping, thereby reducing antenna insertion loss. While techniques such as piezoresponse force microscopy and the Berlincourt method facilitate extraction of quasi-static material characteristics, we harness the overtone region to additionally resolve the RF properties of the materials via a modified multi-variable Sittig’s method, a critical factor in the design of OUWB-ME antennas and in predicting both their fundamental overtones and higher-order harmonics (Fig. 18B-C, Materials and Methods, Table Figure 26). The ferromagnetic crystal was optimized during sputtering towards the development of anisotropic (110) BCC Feo.79Gao.21, as confirmed by X-Ray Diffractometry (Fig. 27), attributed to ordered D03 and unordered A2 mixed phases which has been shown to enhance static magnetostriction. The CMOS-compatible piezoelectric AIN exhibits a c-axis formation of grain size of 167.8 nm. The confounding effects of symmetry mismatch and limited adatom mobility led to increased grain boundaries, yielding small, misoriented columnar grains, rough or partially amorphous interfaces, and residual-stress gradients that boost scattering and internal friction, contributing to the high bandwidth of the antennas (Fig. 28). In addition, the strong modulus mismatch (Feo79Gao21 = 59.95 GPa versus AIN = 289 GPa; Materials and Methods) concentrates misfit / thermal strain within the BCC layer, elevates interfacial shear on AIN, and visco-elastically loads the piezoelectric, enhancing acoustic impedance contrast and damping. This effect is accentuated in the reflection spectra; relative to piezoelectric controls, OUWB-ME antennas show more than a twofold suppression in the overtone region, as shown in Fig. 18D, yielding wider bandwidth while enabling magnetostrictive control. The net effect is further bandwidth broadening at the expense of Q-factor reduction (Fig. 29). Magnetometry measurements confirm a decrease in intrinsic coercivity and identify the hard axis as perpendicular to the sample plane, as indicated by a nearly hysteresis-free magnetization loop (Fig. 30). To improve implantability of the OUWB-ME antennas, bonding was carried out via an ethylene-glycol-based high viscosity silver conductive ink with resistivity of 7.9 pQ cm under micro-dispensing, mitigating the constraints towards on-chip implementation and implantation often affiliated with traditional non-planar wire-bonds. Owing to dielectric losses in both the PCB and the nanoparticle ink, the reflection parameter increased yet stayed below -10 dB throughout the overtone region as shown in Fig. 18E, with stable performance over8673170

[0341] 34

[0342] repeated ex vivo cycles. Phase shifts and parasitic resonances introduced by the metallic nanoparticle ink-bonded interfaces occur outside the primary magnetoelectric overtone band (3.0-3.6 GHz) and therefore do not perturb the OUWB ME antenna-pBot response within the band of interest (Fig. 31). In vitro biocompatability analysis of materials constituting the OUWB-ME Antenna

[0343] The active materials that compose the OUWB-ME antenna, in addition to Parylene, were used as substrates for primary cortical neuronal cultures, to ascertain their individual biocompatibility. Several biocompatibility aspects were targeted: cell viability, neuronal-glial balance within the culture, and neuronal function. Regarding cell viability, the level of mitochondrial activity was measured weekly over a month, using a colorimetric assay that relies on cellular metabolism, generating a colored formazan compound when exposed to NADH / NADPH. Thus, following incubation, the absorbance at 450 nm is a measure of cellular metabolism, acting as an indirect quantification of neuronal health. While AIN presented levels of cell viability equivalent to those of the control condition (glass coverslips), cells growing on FeGa substrates show a decrease in cell viability (Fig. 19A). The reduction in mitochondrial activity of cells growing on the FeGa substrate is linked to low cell survival. Indeed, cells were unable to form a monolayer on this substrate and only a small portion of neurons were able to survive, surrounded by astrocytes presenting an enlarged morphology, as seen in the immunostaining micrographs (Fig. 3B). All samples were subsequently stained for immunocytochemistry quantification (Fig. 3B, bottom row). In this case, no significant differences were found in terms of cell density (Fig. 3C). Analysis of the astrocyte-to-neuron coverage ratio revealed a significant shift only in cultures on FeGa, characterized by a reduction in neuronal area and a predominance of astrocytic coverage. The fact that cell density is slightly reduced in cultures growing on FeGa substrate but the area covered by astrocytes is equivalent to that of the control condition further corroborates the presence of hypertrophic astrocytes. This morphological change is a hallmark of the reactive astrocytic phenotype, which is associated with cellular stress responses and altered neuron-glia communication. Interestingly, although the cells cultured on AIN / FeGa / Parylene show a different network architecture, the balance between glial and neuronal populations is maintained and only a decrease in the area covered by neurons can be identified.

[0344] Since the aggregates do not allow for a deeper functional analysis of the cellular network, only cells cultured on the AIN and FeGa samples were functionally probed, using a fluorescent calcium indicator as a proxy fortheir intrinsic electrical activity. Calcium transients were detected in all experimental conditions (Fig. 19D) but, while cells growing on AIN presented calcium transient duration and amplitude similar to those cultured on glass coverslips, cells cultured on FeGa displayed longer calcium transients with significantly lower amplitude (Fig. 19E), which aligns with the finding that mostly only astrocytes were present. In all cases, the frequency of the calcium transients was equivalent to those plated on glass, as measured by the inter-spike interval. Finally, cells plated on AIN / FeGa / Parylene also displayed calcium transients, once more corroborating that their lower cell viability read-out and altered cellular architecture were due to lack of cellular adherence and not material toxicity, and reiterating the protective role of Parylene C encapsulation of materials such as FeGa.

[0345] Ex vivo transmission and reflection measurements of Bots8673170

[0346] 35

[0347] Transmission (S12) and reflection (S22) parameters were measured using the pBots in conjunction with a commercially available horn antenna. To examine the influence of biological tissue on these parameters, various biological samples (rat and human), summarized in Figure 32, were positioned directly over the pBots. Rat brain tissue was included alongside human samples to evaluate the pBot’s response across different tissue types and conditions, including fresh tissue and intact head preparations. Testing across diverse tissues is essential, as differences in composition, water content, and structural organization influence acoustic and electromagnetic properties. An inert modeling clay of comparable weight to the rat head was included as a sham control for sample weight.

[0348] Regarding both S12 (Fig. 20A) and S22 (Fig. 20B) measurements, traces from each sample were normalized as fold-change relative to their respective blank pBots measurements. This normalization allowed for comparison across pBots despite baseline variability (Fig. 33). For reflection overtone analysis, the S22 traces were detrended, allowing for the isolation of periodic components. Overtone strength was then calculated as the peak-to-peak amplitude of the denoised signal (Fig. 20C).

[0349] Importantly, the existence of overtones in the measurement is caused by the pBots and not a consequence of electrical / measurement noise, since overtones are not present in open calibration measures (Fig. 34).

[0350] Systematic analysis of the impact of biological tissue in pBot transmission (Fig. 21 A and Fig. 35) showed that most samples exhibited transmission peaks at similar bandwidths (3.1, 3.27, and 3.95 GHz).

[0351] Importantly, while the sham control also displayed an enhancement of transmission at these frequencies, the facilitation caused by biological tissue was generally stronger. Indeed, conditions involving fresh tissue (intact rat head, fresh brain, and cold brain, as well as unfixed human tissue) displayed a larger transmission peak around 3.1 GHz (Fig. 21 B). This peak disappeared or decreased following tissue fixation, for rat and human samples respectively, whereas another distinct peak at 3.27 GHz (present under all conditions) was slightly accentuated in the fixed tissue (Fig. 21 C). We hypothesize that fixation increases stiffness, imposing a fixed-constraint acoustic boundary, in the limiting case of a skull-intact rat head this produces a positive frequency shift, confirmed by overtone analysis (Fig. 23E), rather than the soft-boundary damping expected under compliant constraints. Since the acoustic resonance is responsible for magnetic field radiation, which is less impacted by the surface current, absorption in the soft tissue or bones indicates the optimal performance of the pBots at these frequencies. Finally, the peak found at 3.95 GHz followed the same pattern as that found at 3.1 GHz, with fresh tissue showing a larger increase when compared to their fixed counterparts (Fig. 21 D). These shifts may additionally reflect changes in dielectric properties due to tissue fixation and dehydration on account of the varying dielectric constant in air and water (1 versus 80) leading to variable tissue coupling to the atmosphere and constitute contributions in transmission arising from the parasitic effects of the conductive nanoparticle bonds, a previously unexamined but necessary component for the design of implantable magnetoelectric antennas. Furthermore, the cold brain exhibited a larger fold-change while maintaining the same dynamics, potentially due to the progressive warming and moving of the tissue during the measurement.8673170

[0352] 36

[0353] Reflection data (Fig. 22 and Fig. 36) revealed that the intact rat head, comprising brain, skull, and surrounding tissues, produced a noticeable damping effect but with a positive frequency shift, on account of increased stiffness. Hence, this was not exclusively due to sample weight, as tested using a similarly weighted sham control that showed no such effect. The other tested rat samples, including the isolated fresh brain, cold brain, and fixed brain, did not exhibit this damping, with the fresh brain slightly enhancing reflection. Among human samples, only the small fixed tissue showed a significant reflection attenuation, with the large fixed tissue trending similarly.

[0354] Detrended S22 traces highlighted the presence of overtones across the ME range (Fig. 37). A first comparison between biological samples and their baseline suggested that the presence of the different samples could induce a frequency shift in the overtone strength signal dynamics (Fig. 23A). To confirm this observation, the optimal lag for each condition was identified using cross-correlation analysis (Fig.

[0355] 23B), showing that the presence of the ex vivo severed rat head led to a shift toward higher frequencies, while the other biological samples showed a tendency to shift the signal toward lower frequencies.

[0356] Importantly, the tested sham controls, in both contact and no-contact configurations, did not lead to frequency shifts.

[0357] Given that cross-correlation analysis can only quantify overall frequency shift, four main overtone features were defined in the overtone strength dynamics (Fig. 23C) - two maxima (P) and two minima (N) peaks, occurring sequentially (N1 , P1 , N2, P2) and present in all pBots at comparable frequencies and with similar strength (Fig. 38). Pooling all features for each tested condition revealed that only the severed rat head led to a significant shift toward higher frequencies, while the cold rat brain and all human samples displayed a significant negative frequency shift (Fig. 23Di). In terms of overtone attenuation caused by the different samples, the same trends were observed, albeit with increased inter-pBot variability (Fig. 23ii). While all samples caused some degree of attenuation, only the explanted rat cold brain and the fixed small human tissue showed statistical significance. Noteworthily, the sham control in contact with the pBot also led to some overtone attenuation, which can be explained by its weight. The discrimination between weight per se and the impact of biological tissue will be explored below. Interestingly, the fresh rat brain did not impact overtone strength neither in frequency nor in attenuation.

[0358] Taken together, these results suggest that the severed rat head distinctly impacts overtone dynamics (Fig. 23E), likely reflecting the presence of skull and connective tissues. This emphasizes the importance of considering not only the target tissue but also the surrounding structures when assessing ME antenna performance in a biological environment. Moreover, the consistent negative frequency shift observed across all fixed samples suggests that fixation imposes common structural or dielectric changes, potentially linked to crosslinking of proteins and altered water mobility. The striking difference between explanted fresh and cold rat brain could stem from temperature- and fluid-related factors: the fresh brain remains warm and physiologically hydrated, while the cold brain contains additional aCSF and undergoes progressive warming during measurement, both of which affect its dielectric properties. The intermediate impact of the unfixed human tissue between these two conditions may also reflect a similar temperature-8673170

[0359] 37

[0360] and water-content dependence. These observations point toward hydration state and tissue temperature, as well as the presence of surrounding tissue, as key modulators of overtone dynamics.

[0361] Since different tissues might have a differential effect on the different signal features, each overtone feature was specifically evaluated (Fig. 23F and Fig. 38). Despite having identical weight, the sham control and severed rat head displayed divergent patterns (Fig. 23Fi), where the sham did not preferentially affect any feature, whereas the rat head induced stronger frequency shifts and attenuation at higher frequency features (N2 and P2). Regarding explanted rat brain samples, the fresh brain did not differentially impact the features, while both the cold and fixed brain samples showed a pronounced effect on N1 (Fig. 23ii) when compared to the other identified features.

[0362] Human autoptic samples, though heavier overall, induced smaller frequency shifts and attenuation than the rat samples (Fig. 23iii). Also, these samples differentially affected N2, with more pronounced negative frequency shifts. Interestingly, the unfixed human autoptic sample did not attenuate N2. Finally, although the large fixed human sample was one order of magnitude heavier than its small counterpart, its signal attenuation was not proportionally larger. While these insights are preliminary, they reveal a relevant and unexplored influence of biological tissues on OUWB-ME antenna overtone dynamics. Importantly, with this study, the inventors were able to experimentally observe higher frequency shift induced by skull proximity and lower frequency shift in intracranially implantable magnetoelectric antennas due to tissue proximity.

[0363] Data communication and MRI compatibility

[0364] To evaluate the wireless capabilities of pBots in a realistic biomedical communication scenario, we established a transmission experiment at 3.25 GHz in which a 1.96 MB sonogram video with heartbeat audio was delivered in real time using a software-defined radio (SDR) platform and a pair of transmitting and receiving pBots, for the first time (Fig. 24A). Transmission followed the Digital Video Broadcasting-Satellite (DVB-S) standard with Quadrature Phase Shift Keying (QPSK) modulation at a symbol rate of 500 kilo symbols per second. A Root Raised Cosine filter with a roll-off factor of 0.35 yielded an occupied bandwidth of 675 kHz. Unlike conventional electrical antennas of comparable size, which are typically constrained to narrowband operation and require tuning to their environment, the pBots exhibited intrinsically wideband characteristics, readily accommodating this modulated signal without distortion or detuning. This wideband response ensured that antenna behavior did not limit the experiment, allowing performance to be determined entirely by the digital modulation and SDR hardware. The baseband waveform was generated at eight samples per symbol, corresponding to a digital sampling rate of 4 megasamples per second. This oversampling enabled accurate pulse shaping, stable synchronization, and reliable filtering while remaining within the real-time streaming limits of the B205i SDR platform. The local oscillator for both transmitter and receiver was fixed directly at 3.25 GHz, and although this direct-on-frequency approach introduced a small DC artifact at baseband, the robustness of DVB-S QPSK demodulation rendered the effect negligible. The average received power was obtained by integrating the measured spectrum over a bandwidth b centered at fo under controlled in-plane and out-of-plane misalignments of the ME-RF link, and benchmarked against conventional RF-RF links. The RF antenna8673170

[0365] 38

[0366] was a log microstrip PCB with the magnetic field oriented in-plane. Under out-of-plane misalignment, a heterogeneous ME-RF pairing exhibited a markedly shallower power change than the RF-RF baseline. In the measurement shown (top two markers, ME-RF; bottom two, RF-RF), the ME-RF points cluster higher and tighter on the power axis, indicating lower sensitivity to out-of-plane tilt than an RF-RF link under identical conditions (Fig. 24B). We ascribe this robustness to the magnetic-field-dominated near-field coupling provided by the ME element, which mitigates the polarization and orientation penalties that degrade purely RF pairs.

[0367] In terms of angular misalignment, the contrast between ME-RF and RF-RF can be understood by examining their confidence intervals and the number of ME elements required to achieve comparable robustness. The per-angle standard deviation was calculated directly from the measured response at each angular position, quantifying the spread of values around the mean for both ME-RF and RF-RF curves. A single ME element exhibits a per-angle standard deviation of approximately 1.27 dB, whereas an RF element (RF-RF) shows a lower spread of about 0.43 dB, reflecting its inherently greater stability. Yet, because ink-bonded ME elements occupy only -1 mm2of printed circuit board (PCB) area compared to -147 cm2for the RF antenna, a reduction of almost five orders of magnitude, they can be densely arrayed, and their effective standard deviation decreases as 1 / A / , where N is the number of combined elements. From this scaling, about nine ME elements are sufficient to reduce ME-RF’s angular sensitivity to the same level as a single RF element. Importantly, the stability of nine ME elements is already close to that of an RF antenna, yet their combined footprint remains below 0.01 cm2. More stringent requirements highlight this advantage further: to confine the 68% confidence interval within ± 0.5 dB requires 6 elements, achieving ± 0.2 dB requires 40 elements, and ± 0.1 dB requires 160 elements. At the 95% confidence level, the corresponding numbers are 25, 154, and 616 elements for ± 0.5, ± 0.2, and ± 0.1 dB, respectively. Extending the analysis to the ME-ME case with the widest variation, the measured mean and 95th-percentile per-angle standard deviations (1.78 dB and 4.44 dB, respectively) imply 16 elements on average and 100 elements for worst-angle robustness to match the RF-RF trend, still only 0.16-1.0 cm2of ME footprint. These results demonstrate that although a single ME receiving element is more misalignment-sensitive than RF, the ability to integrate and average tens to hundreds of pBots within an area orders of magnitude smaller than that of a single RF antenna allows ME-RF not only to reach but to surpass RF-RF in robustness, while heterogeneous ME-RF pairings already display reduced out-of-plane sensitivity. Together, these findings establish a pathway for highly stable, miniaturized biomedical telemetry links where angular tolerance is critical.

[0368] Considering a federated application of pBots, MRI compatibility must be insured due to the presence of ferromagnetic materials, metallic layers and inks. Randomly dispersed clusters of OUWB-ME underwent 10-hour 7T MRI scans to gauge the movement of devices, if any. T1 -weighted images exhibited only minor edge artefacts around the chip, insufficient to confound interpretation of in vivo data. Segmentation masks delineated chips with the expected size and morphology, without intrachip artefacts. Difference (APos) images showed a net downward displacement between time points with no component along Bo, consistent with gravitational settling rather than magnetic forces. APos encodes the voxel-wise mask8673170

[0369] 39

[0370] difference (+1 , present only at the first scan; 0, unchanged; -1 , present only at the second) (SI Videos 5-8).

[0371] DISCUSSION

[0372] In this study, the inventors report OUBW-ME antennas with bandwidth of 22.6 GHz , established through fabrication, characterization and validation. AIN is biocompatible with dissociated rat neurons in vitro, while FeGa induces cell death unless isolated by encapsulants such as Parylene C. Regarding the ex vivo pBot measurement, although preliminary, these experiments constitute, to the inventors’ knowledge, the first systematic quantification of how biological tissue reshapes the transmission and reflection characteristics ME antennas. Across preparations, reproducible spectral features were observed that identify discrete frequency windows, most prominently near 3.3 and 3.9 GHz, likely to be preferential for reliable wireless links in tissue-embedded devices. The data further reveal both conserved and divergent responses across species and tissue-preparation states. In this context, overtone frequency shifts are governed primarily by tissue-induced damping and by acoustic reflections to and from the device, whereas overtone amplitudes provide a quantitative readout of loss, decreasing as damping increases. Collectively, these results offer a framework for selecting operational bands and interpreting overtone structure in biologically constrained ME systems.

[0373] Single-element analyses often cast ME antennas as orientation-tolerant because they couple through the magnetic near field and avoid platform currents, but this intuition breaks down in federated arrays. Each ME element acts as a magnetic dipole whose axis is fixed by the DC-bias. Coupling is strong when this axis aligns with the receive field and collapses when orthogonal, so even small inter-element bias scatter accumulates as polarization and phase mismatch. In a federation, those mismatches are further amplified, eroding the native stability improvement expected from an individual antenna on-chip and unpackaged. Tissue further imposes nonuniform acoustic loading, shifting overtone frequency and strength element by element, so the federation experiences unequal complex gains rather than a uniform misalignment penalty. This study provides the first insight into this behavior.

[0374] By combining compact form factor with wideband operation, magnetoelectric antennas overcome the limitations of conventional narrowband miniature antennas and provide a fundamentally different platform for wireless biomedical communication. The successful delivery of a sonogram video highlights their potential to enable next-generation wireless systems for minimally invasive diagnostics, implantable monitoring, and portable health technologies, where reliable and efficient transfer of imaging data is essential. Looking forward, the demonstrated link can be readily scaled to higher symbol rates, higher-order modulation formats, and multi-megabit data streams. For example, increasing the symbol rate to 2 mega symbols per second with QPSK would yield a throughput of 4 Mbps, sufficient to support full-motion ultrasound video or higher-fidelity diagnostic imaging.

[0375] The future of wireless telemetry with federations of micron-scale ME antennas depends on their wide 22.6-GHz usable band. With the available headroom, a federation of ME antennas can split the spectrum among nodes or take rapid turns sending short bursts, creating, when required, very wide effective signals that deliver few-millimeter ranging precision and tens-of-picoseconds timing, while still supporting8673170

[0376] 40

[0377] multi-gigabit data streams even if only a small fraction of the band is used at any moment. The breadth also buys resilience and the network can slide its center frequency to avoid tissue- or hardware-induced attenuation, run several parallel streams for robustness, and keep power and SAR low by duty-cycling and beamforming rather than pushing a single element hard. The ME federation can choose between precision, rate, and robustness on the fly, making practical, body-embedded telemetry realistic in ways narrowband implantable antennas cannot.

[0378] These results point toward the broader applicability of magnetoelectric antennas in enabling wideband biomedical data transfer and future clinical telemetric applications. Furthermore, by shedding light on tissue- and condition-dependent effects not only on antenna transmission and reflection, but also on acoustic magnetoelectric overtone frequency shift and signal attenuation, this study inspects this previously unexplored domain.

[0379] MATERIALS AND METHODS

[0380] Antenna fabrication and characterization

[0381] OUWB-ME antenna fabrication: All OUWB-ME antennas were fabricated on undoped (100) double sided polished Silicon wafers of resistivity 10000-100000 Qcm. The substrates were cleaned in Piranha followed by deposition of the bottom electrode (Ti / Pt; 10 / 70 nm). Aluminum Nitride and FeGa were deposited via RF sputtering in N2 / Ar and Ar, respectively. The deposition parameters in terms of sputtering power, pressure, and flow rates were optimized towards c-plane AIN and (110) FeGa, deposited under a peak in-situ magnetic field of 10 kA / m, measured via a Hall sensor (TLV493D-A1 B6; Infineon). Vias in AIN were etched using AZ400K developer and the top electrodes (Pt; 100 nm), patterned via lift-off in silicate developers. Prior to every deposition run, in-situ Argon etching was carried out for the removal of residual oxides and nitrides. Parallel permanent magnets provided DC in-plane domain-biasing during sputtering which was confirmed by on-chip two-port transmission measurements when normalized by the response of control piezoelectric resonators.

[0382] Packaging of OUWB-ME antennas towards gBots: Each antenna, initially protected with AZ4562 photoresist, was diced into 1 mm dies using a diamond saw (DISCO Corporation, Tokyo, Japan). The dies were mounted on ceramic Rogers PCBs with a non-conductive epoxy and the resist stripped in acetone followed by an IPA wash. Planar electrical interconnects were formed using a shear-thinning (non-Newtonian), glycol-based silver nanoparticle ink (CL85; XTPL, Poland) dispensed through a 20-pm-inner-diameter nozzle. To prevent oxidation of the Fe-based spintronic layer, the assemblies were heated at 180°C under continuous Ar flow. While the XTPL system is capable of consistently producing lines of sub-10 urn width and ~3 pm thickness, in this case we optimised for reliability and minimal resistances by performing multiple passes using excess ink.

[0383] Due to the repeated passes, the thickness of these lines was approximately 30 pm ± 10 pm, with a linewidth of 30 pm (Fig. S40).

[0384] RF Characterization of OUWB-ME antennas: On-chip reflection, transmission and impedance parameters were experimentally derived with a precision vector network analyzer (Keysight N5245B) following short-8673170

[0385] 41

[0386] open-load-thru (SOLT) calibration. The frequency resolution was carefully chosen to spectrally resolve the overtones and the IF bandwidth and power for all measurements were held constant at 1 kHz and 0 dBm, respectively. The one-port impedance parameters were derived according to

[0387]

[0388] characteristic impedance being 50 Q. The derived impedance data at the overtone and higher order harmonic region at ~8 GHz was utilized to determine the RF properties of the active materials via the modified Sittig’s model. The domain biasing in the magnetostrictive layer was exploited via the differencemodel between the magnetoelectric antennas and control piezoelectric devices making sure the spatial position of the probe stays the same between measurement. All two-port transmission parameters are normalized by the one-port reflection parameter of the commercial horn antenna (PowerLOG 70180; Aaronia) to account for noise. The horn antenna was connected to port 1 and the OUWB-ME Antenna to port 2.

[0389] Modified Sittig’s transmission line model: The input impedance of the overtone structure, Zin, is given by Eq. 1. The effective input impedance is the parallel combination of the impedances of A1 and A2 (Fig. S4). In overtone designs, acoustic waves disperse into the underlying layers, including the substrate, generating a ladder of overtones with spacing Af ~ v / 2ts, where v is the longitudinal sound velocity in the substrate and ts its thickness. The phase delay in the piezoelectric layer is = / piezo , with analogous treatment applied for the remaining layers. Acoustic impedance Zai follows Eq. 2, using thickness-weighted velocities and densities. In multilayer electrodes, properties are defined by relative layer fractions. A double-sided polished silicon wafer fundamentally alters the acoustic boundary wherein the smooth backside reflects rather than dissipates. This reflection disrupts assumptions of substrate loss, introduces additional phase terms, and reshapes standing wave conditions. Resonance shifts, impedance distortion, and Q degradation follow. The substrate is no longer passive, it becomes an active cavity and must be treated as such.

[0390] " >

[0391]

[0392] Where Zb — Zb / Zai, Z2 — Zl_ow / Zai, Yt°p—COdtop / Vtop, Ybottom—wd bottom / bottom, YSi—todsi / Vsi, Ztop — Aptop_electrodeVtop_electrode, Zbottom_electrode=Apbottom_electrodeVbotto _electrode, Zsi=ApsiVsi.

[0393] Vibrating Sample Magnetometry. Magnetic hysteresis loops of the samples were obtained using a Vibrating Sample Magnetometer (MicroSence VSM-EV9). A maximum magnetic field of + / -5 kOe was applied in the plane of the film at 0° and 90° concerning the short axis of the sample. Measurements were8673170

[0394] 42

[0395] also performed out-of-plane of the film with a maximum applied field of + / -20 kOe, to allow for full magnetization saturation.

[0396] X-ray Diffraction. The crystallography was via Empyrean, Malvern Panalytical, with Ka1 and Ka2 wavelengths at 1.540598 A and 1.544426 A, respectively, and of intensity ratio 0.5. The pulse height detection levels by 25% in Empyrean for accurate detection of Iron containing compounds that fluoresce with Cu radiation. The raw data was subsequently processed by carrying out background subtraction via the denoising methodology of Sonnerveld and Visser.

[0397] X-ray Photoelectron Spectroscopy. The surface properties of the sputtered films with the inclusion of the Al capping layer were analyzed via X-ray photoelectron spectroscopy (AXIS Supra+, Kratos Analytical) consisting of an Al Ka anode producing X-ray beams of energy 1.486 keV. The beam power is held at 450W leading to a penetration depth of ~ 9 nm. The deconvoluted fit of all spectra from the raw data is carried out via Voigt fitting of the Gaussian and Lorentzian functions.

[0398] In vitro biocompatibility

[0399] Materials. Poly(ethyleneimine) (PEI), Hank’s balanced salts (Ca2+ / Mg2+free) (HBSS), kynurenic acid, trypsin, deoxyribonuclease (DNase), bovine serum albumin (BSA), NaHCO3, MgSO4, HEPES, NaCI, KOI, Na2HPO4, heat-inactivated horse serum, D-glucose, paraformaldehyde (PFA), Triton X-100, rabbit anti-p3-tubulin primary antibody, mouse anti-GFAP primary antibody, and DAPI (0.5 mg / mL) were purchased from Merck KGaA (Darmstadt, Germany). Cell counting kit-8 (CCK-8) was acquired from Dojindo EU GmbH (Munich, Germany). Ethanol (70%), minimum essential medium (MEM), gentamycin (50 mg / mL), GlutaMAX, donkey anti-rabbit 488 secondary antibody, donkey anti-mouse 594 secondary antibody, and Oregon Green BAPTA-1 were obtained from ThermoFisher Scientific (Waltham, MA, USA). Fetal bovine serum (FBS) was purchased from Euroclone (Milan, Italy). D(-)-2-amino-5-phosphopentanoic acid (D-AP5) was acquired from Hello Bio (Dunshaughlin, Ireland). For cell culture experiments, two different solutions were used for the dissection protocol: dissection solution, consisting of 9.52 g L-1 HBSS, 4.2 mM NaHCO3, 33 mM D-glucose, 200 pM Kynurenic acid, 25 pM D-AP5, 0.025% (v / v) gentamycin, 0.03% (w / v) BSA, 12 mM MgSO4, and 12 mM HEPES; and digestion solution, composed of 137 mM NaCI, 5 mM KCI, 7 mM Na2HPO4, 25 mM HEPES, 200 pM Kynurenic acid, 25 pM D-AP5, and 4.2 mM NaHCO3. Similarly, two different cell media were used: H-MEM, consisting of MEM, supplemented with 5% (v / v) heat-inactivated horse serum, 20 mM glucose, 0.1% (v / v) gentamycin, and 100 pM GlutaMAX; and N-MEM, with the same composition as H-MEM, except with 50 pM GlutaMAX. Primary rat neuronal cultures. Before cell plating, all substrates were sterilized in 70% ethanol for 30 minutes, washed 3x with sterile ultrapure deionized water, and allowed to air dry. To enhance neuronal adherence, substrates were coated with 0.1% (w / v) polyethyleneimine (PEI) overnight at 37 °C, followed by 3x washes with sterile ultrapure deionized water and air drying. Primary neuronal cultures were prepared from the cortical neurons of P0-1 Wistar rats. Animals were sacrificed by decapitation, and their brains extracted and placed in 5 mL of dissection solution. Following the removal of the meninges, the cortical area was isolated and separated into small pieces in 5 mL of fresh dissection solution. The tissue was then moved under a laminar flow cabinet, washed with digestion solution, and exposed to 0.5% (w / v)8673170

[0400] 43

[0401] trypsin and 0.075% (w / v) DNase in digestion solution at 37°C for 5 minutes. The reaction was then stopped by incubating the tissue with 0.1% (w / v) trypsin inhibitor in dissection solution for 10 minutes at 4°C. Finally, the tissue was washed in dissection solution, placed in a solution of 0.06% (w / v) DNase in dissection solution, and mechanically dissociated. Cells were then centrifuged (100g, 5 minutes), after which the supernatant was discarded, and the pellet was resuspended in H-MEM. Following cell counting, the final cell suspension was plated on the prepared substrates or glass coverslips as the control condition, at a final density of 1500 cells mm-2in H-MEM, kept at 37°C in a humidified environment of 95% air-5% CO2. Every two days, half the volume of cell medium was replaced with fresh media.

[0402] Importantly, after 7 DIV, N-MEM was used. All experimental procedures were approved by local veterinary authorities and performed in accordance with Italian law (decree 116 / 96) and EU guidelines (86 / 609 / CE, 2007 / 526 / CE and 2010 / 63 / UE). Animal use was approved by the Italian Ministry of Health. Efforts were made to minimize suffering and the number of animals used.

[0403] Cell viability assay. At different time points (7, 14, 21, and 28 DIV), cell medium was renewed and supplemented with 10% (v / v) of CCK-8. After 2 hours of incubation, 100 pL of cell medium was collected and its absorbance was read at 450 nm using a microplate reader (Multiskan FC, Thermo Scientific). Immunostaining. At 14 DIV, cells were fixed with 4% (w / v) PFA in PBS for 30 minutes and washed 3x with PBS. Samples were permeabilized and blocked using 0.1% (v / v) Triton X-100 and 5% 577 (w / v) FBS in PBS for 1 hour, followed by overnight incubation at 4°C with primary antibodies (1:300). Then, samples were incubated with the respective secondary antibodies (1:500) for 2 hours and exposed to DAPI (1:500) for 30 minutes, both at room temperature and protected from light. Finally, samples were mounted on glass slides and imaged using an inverted fluorescence microscope with a 20x objective.

[0404] Calcium imaging. At 21 DIV, cell medium was removed and replaced with 5 pM Oregon Green BAPTA-1 in N-MEM. Following a 2-hour incubation at 37°C, samples were transferred to fresh medium and imaged with an inverted fluorescence microscope equipped with a 10x objective. Each sample was imaged at 5 different fields of view, with each recording lasting 180 s with a frame rate of 5 Hz.

[0405] Data analysis. For the cell viability assay, every independent experiment was performed with technical duplicates and at least three independent experiments were performed for each condition. Given the different surface area of the glass coverslips and the silicon wafers, the absorbance measurements were normalized to substrate surface area. For immunostaining experiments, five random fields were acquired for every condition in every independent biological replicate. Immunostaining quantification was performed using custom MATLAB scripts, where the percentage of cell coverage was determined by the ratio of positive pixels in each channel following image binarization. The cell density was calculated based on the number of nuclei found in each analyzed field. For calcium imaging experiments, the acquired videos were also analyzed using custom MATLAB scripts, where all individual cells were identified, and their mean fluorescence signal was used as the calcium transient trace.

[0406] Ex vivo transmission and reflection measurements8673170

[0407] 44

[0408] Experimental setup. The pBot was mounted on a PCB and placed at a fixed distance of 16 cm from a commercially available horn antenna (PowerLOG 70180; Aaronia), to be within the radiating near field region of the horn antenna. The pBot was placed in a custom isolation box interference with microwave absorbers to minimize external. The horn antenna was oriented in such a manner that it exhibited minimal electrical coupling to ground thus reducing parasitic effects while maintaining a high in-plane H-field. Transmission (S12) and reflection (S22) parameters were measured across the magnetoelectric overtone operational range (3-4 GHz), using six different pBots. VNA calibration was carried in the entire testing frequency range of 0.7-5 GHz, with a port power of 0 dBm and an IF bandwidth of 100 Hz. The horn antenna was connected to port 1 and the pBot integrated with the ex vivo samples to port 2.

[0409] Animal sample preparation. Postnatal day 28 (P28) Wistar rats were anesthetized and decapitated, after which the entire head was placed on the ME antenna with a thin layer of Parafilm for separation.

[0410] Transmission and reflection measurements were then acquired. Subsequent measurements were performed after sequentially extracting the brain, immersing it in ice-cold artificial cerebrospinal fluid (aCSF) for 10 minutes, and finally fixing it in 4% paraformaldehyde (PFA) in PBS for 24 hours. Four animals were used in total. All experimental procedures were approved by local veterinary authorities and performed in accordance with Italian law (decree 116 / 96) and EU guidelines (86 / 609 / CE, 2007 / 526 / CE and 2010 / 63 / UE). Animal use was approved by the Italian Ministry of Health. Efforts were made to minimize suffering and the number of animals used.

[0411] Human sample preparation. Access to human neocortical autoptic samples was obtained, upon approval by the Local Ethical Committee of the Univ, of Modena & Reggio Emilia, through the Forensic Medicine Institute of the same university. One sample was formalin-fixed post-autopsy and subdivided into a “large” and a “small” cortical piece. The second was harvested during autopsy and frozen immediately, only thawed prior to the experiment. Measurements were performed on both fixed and unfixed tissues.

[0412] Controls. As sham controls, an inert modeling clay of equivalent weight to the rat head was used. Both contact and non-contact configurations were tested to disambiguate effects due to weight or dielectric properties.

[0413] Signal processing and analysis. Reflection (S22) and transmission (S12) magnitudes were calculated as the absolute value of the complex (real + imaginary) signal components. S12 was normalized against the reflection magnitude (S11 ) of the horn antenna to rule out any external fluctuations associated with the horn antenna. The magnitude of each tested sample was converted in fold change in relation to the respective control, to minimize baseline differences between the different tested pBots. Differences in transmission and reflection were first identified by running a series of frequency-by-frequency t-tests between each condition and its control, with p-values adjusted by the False Discovery Rate (FDR) approach (Benjamini-Krieger-Yekutieli). Frequencies with consistent effects across samples were further analyzed using Kruskal-Wallis tests with FDR. To quantify reflection overtone strength throughout the tested frequency range, S22 traces were denoised by subtracting the low frequency component of the trace to remove the overall signal dynamics and isolate the overtones. Overtone strength was then calculated as the peak-to-peak amplitude of the denoised signal. Overtone strength cross-correlation8673170

[0414] 45

[0415] between each sample and its respective control was used to determine the overall frequency shift associated with each sample. Furthermore, from the overtone strength dynamics, the two major peaks and troughs were identified and separately analyzed to identify frequency and feature specific differences, using one-way ANOVA followed by Dunnet’s multiple comparison test.

[0416] Wireless telemetry of sonogram with heartbeat audio: To quantify how orientation affects data transfer, we measured link gain while rotating the pBot transmitter-receiver pair in azimuth (in-plane) and elevation (out-of-plane) connected to two software defined radio transceivers (SRP B205mini-i; Ettus Research). The antennas were mounted with their planes parallel at a fixed 10 mm separation and the transmitter normal defined 0°. For the azimuthal sweep, the receiver was rotated about its surface normal from 0° to 90° in 10° increments while transmit power and separation were held constant, and the received power was recorded in SDRangel. Continuous spectra and video were monitored to verify lock. The transmitter employed a low-pass filter (LPF) cutoff of 0.8-1.0 MHz, slightly exceeding the occupied bandwidth to preserve signal fidelity while suppressing spectral leakage. The receiver bandwidth was configured more broadly, at 1.0-1.2655 MHz, ensuring complete capture of the transmitted spectrum, including the filter roll-off and any minor frequency offsets, while still excluding excess noise. Transmit gain was set to 35-40 dB to avoid amplifier saturation and receive gain to 55-65 dB to maximize the signal-to-noise ratio.

[0417] MRI Compatibility test: Six OUWB-ME antennas were placed into 2% agar (MSK, UK) before it had reached its gelling temperature and left to set. Magnetic resonance images were acquired on a 7T PharmaScan scanner with a 1P T81021H volume coil (Bruker Biospin, Germany). 3D T1 weighted (T1w) fast low angle shot (FLASH) images were acquired, with two shorter scans at a lower resolution (TR / TE = 20 / 5 ms, FA = 10°, FOV = 30 x 21 x 21 mm, imaging matrix = 300 x 663210 x 210, Av = 2) acquired at an 10 hour, 8 minutes interval, and a high resolution scan (TR / TE 664 = 50 / 4.8 ms, FA = 10°, FOV = 26 x 17 x 20 mm, imaging matrix = 350 x 230 x 270 mm, Av = 66510). Images were processed in MATLAB (MathWorks, USA), with masks created by defining voxels that contained less than 10% of the maximal signal value for the entire image. Delta images were calculated by the difference of the initial image or mask minus the final image

[0418] ***

[0419] The features disclosed in the foregoing description, or in the following claims, or in the accompanying drawings, expressed in their specific forms or in terms of a means for performing the disclosed function, ora method or process for obtaining the disclosed results, as appropriate, may, separately, or in any combination of such features, be utilised for realising the invention in diverse forms thereof.

[0420] The order of the operations of the methods described herein is exemplary, but the steps may be carried out in any suitable order, or simultaneously where appropriate. Additionally, steps may be added or substituted in, or individual steps may be deleted from any of the methods without departing from the scope of the subject matter described herein. Aspects of any of the examples described above may be combined with aspects of any of the other examples described to form further examples without losing the effect sought.8673170

[0421] 46

[0422] While the invention has been described in conjunction with the exemplary embodiments described above, many equivalent modifications and variations will be apparent to those skilled in the art when given this disclosure. Accordingly, the exemplary embodiments of the invention set forth above are considered to be illustrative and not limiting. Various changes to the described embodiments may be made without departing from the spirit and scope of the invention.

[0423] For the avoidance of any doubt, any theoretical explanations provided herein are provided for the purposes of improving the understanding of a reader. The inventors do not wish to be bound by any of these theoretical explanations.

[0424] Any section headings used herein are for organizational purposes only and are not to be construed as limiting the subject matter described.

[0425] The terms “a” (or “an”), as well as the terms “one or more” and “at least one” can be used interchangeably herein.

[0426] The term “and / or” as used herein is to be taken as specific disclosure of each of specified listed features or components with or without one or more of the others. Thus, the term “and / or” as used in a phrase such as “A, B and / or C” encompasses each of: A and B and C; A and B; A and C; B and C; A or B or C; A or C; A or C; B or C; only A; only B; and only C.

[0427] The use of the term “comprise” and “include” to refer to the inclusion of integers, steps and / or operations nonetheless also encompasses aspects, examples and embodiments that may be analogously described with the term “consist” in respect of those integers, steps and / or operations.

[0428] Throughout this specification, including the claims which follow, unless the context requires otherwise, the word “comprise” and “include”, and variations such as “comprises”, “comprising”, and “including” will be understood to imply the inclusion of a stated integer or step or group of integers or steps but not the exclusion of any other integer or step or group of integers or steps.

[0429] It must be noted that, as used in the specification and the appended claims, the singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise. Ranges may be expressed herein as from “about” one particular value, and / or to “about” another particular value. When such a range is expressed, another embodiment includes from the one particular value and / or to the other particular value. Similarly, when values are expressed as approximations, by the use of the antecedent “about,” it will be understood that the particular value forms another embodiment. The term “about” in relation to a numerical value is optional and means for example + / - 10%.

Claims

1. 867317047Claims:

1. A magnetoelectric antenna comprising:a first electrode comprising a first metal contact;a second electrode comprising a second metal contact;a piezoelectric layer between the first electrode and the second electrode; and one or more magnetostrictive layers.

2. The magnetoelectric antenna according to claim 1 , wherein the first electrode comprises an acoustic Bragg reflector, the acoustic Bragg reflector comprising a first set of Bragg layers formed from a first material and a second set of Bragg layers formed from a second material having a different acoustic impedance than the first material.

3. The magnetoelectric antenna according to claim 2, wherein the acoustic Bragg reflector is arranged such that the layers of the first set of Bragg layers alternate with the layers of the second set of Bragg layers.

4. The magnetoelectric antenna according to claim 2 or 3, wherein the first material and / or the second material is metal.

5. The magnetoelectric antenna according to any of claims 2 to 4, wherein the first metal contact is disposed between the acoustic Bragg reflector and the piezoelectric layer.

6. The magnetoelectric antenna according to any of claims 2 to 5, wherein the acoustic Bragg reflector comprises an intermediate layer arranged between the first metal contact and the first and second sets of Bragg layers.

7. The magnetoelectric antenna according to any preceding claim, wherein the second electrode further comprises a third metal contact co-planar with the second metal contact, the second metal contact and the third metal contact having a different thickness.

8. The magnetoelectric antenna according to any preceding claim, wherein one or more of the metal contacts of the first electrode and / or the second electrode is a patterned metal contact.

9. The magnetoelectric antenna according to claim 8, wherein the piezoelectric layer and / or the one or more magnetostrictive layers is / are patterned in accordance with the patterned metal contact(s).

10. The magnetoelectric antenna according to any preceding claim, wherein the one or more magnetostrictive layers are disposed between the piezoelectric layer and the second electrode, or wherein the second electrode is disposed between the piezoelectric layer and the one or more magnetostrictive layers.86731704811. The magnetoelectric antenna according to any preceding claim, further comprising a capping layer on the one or more magnetostrictive layers.

12. The magnetoelectric antenna according to any preceding claim, further comprising an adhesion layer for adhering the one or more magnetostrictive layers to the piezoelectric layer.

13. The magnetoelectric antenna according to any preceding claim, wherein the one or more magnetostrictive layers comprise a plurality of magnetostrictive layers defining a stack, each adjacent pair of magnetostrictive layers in the stack being separated by a laminating layer.

14. The magnetoelectric antenna according to any preceding claim, further comprising a casing arranged to encapsulate the first electrode, the second electrode, the piezoelectric layer, and the one or more magnetostrictive layers.

15. The magnetoelectric antenna according to claim 14, wherein the casing is formed from a biocompatible material.

16. The magnetoelectric antenna according to any preceding claim, wherein a thickness of each of the one or more magnetostrictive layers is 80 nm or less.

17. The magnetoelectric antenna according to any preceding claim, wherein the magnetoelectric antenna has an active area of 0.05 mm2or less.

18. An implantable neuroscanner comprising a magnetoelectric antenna according to any preceding claim, wherein, in use, the magnetoelectric antenna is configured to record electrical signals generated in a tissue into which the neuroscanner has been implanted, and to transmit the recorded electrical signals to an ex vivo device.

19. An implantable neurostimulator comprising a magnetoelectric antenna according to any of claims 1 to 17, wherein, in use, the magnetoelectric antenna is configured to receive an electromagnetic signal from an ex vivo device, and to generate a stimulation signal in accordance with the received electromagnetic signal to stimulate a tissue into which the neurostimulator has been implanted.

20. The implantable neurostimulator according to claim 19, wherein, in use, the magnetoelectric antenna is further configured to record electrical signals emitted by the tissue into which the neurostimulator has been implanted, and to transmit the recorded electrical signals to the or another ex vivo device.

Images