METHOD FOR PRODUCING A SOFT OR PERMANENT MAGNET

Description

[0001] The invention relates to a method of manufacturing a permanent or soft magnet.

[0002] A permanent magnet is an object made of a ferromagnetic material. It has a natural magnetization present in the absence of an external magnetic field, as well as a high coercive field and remanence.

[0003] Many micro electromechanical systems hereinafter referred to as “MEMS” (i.e. the English acronym for “micro electro mechanical system”) comprise permanent magnets, the assembly forming magnetic micro electromechanical systems, hereinafter referred to with the abbreviation “MAGMEMS”.

[0004] MAGMEMS can be any type of MEMS sensor and actuator with electromagnetic transduction. For example, it can be a magnetic field sensor, an energy harvester or a relay.

[0005] These MAGMEMS are used in a wide variety of applications: in microbiology, particularly for microfluidic magnetic separation (e.g. biosensors), in telecommunications (e.g. radiofrequency microswitches), in automotive sensors.

[0006] Additionally, permanent magnets can be used in magneto-optical devices such as: magneto-optical sensors (e.g., fiber optic sensors, rotary sensors, and polarimetric sensors); deflectors; modulators (e.g., microwave modulators); non-reciprocal components (e.g., isolators and circulators).

[0007] Given the multitude of applications, permanent magnets have become essential objects in everyday life and thus represent a significant and promising market.

[0008] Until recently, the main source of materials for permanent magnets was rare earth alloys (e.g. scandium, yttrium and 15 lanthanides such as neodymium and samarium) which have the advantage of providing the best magnetic performance. A well-known example of a rare earth permanent magnet is the NdFeB magnet, also known as a "Neo magnet" which is composed of an alloy of neodymium (Nd), iron (Fe) and boron (B).

[0009] However, extracting rare earths is difficult and poses environmental problems. Therefore, the current trend is to move away from these materials and seek substitutes that perform just as well.

[0010] Furthermore, the manufacturing process for permanent magnets from rare earths is complex. For example, manufacturing an NdFeB magnet requires the following steps: melting, then grinding the raw materials or decrepitation with hydrogen, molding in a magnetic field, and sintering. Next, the magnet blanks must be machined and polished to the desired size and shape. Finally, a surface treatment, for example by galvanizing to coat the magnet with a nickel-copper-nickel layer, is necessary to prevent the magnet from oxidizing and decomposing into dust. Also, the limitations of machining do not allow permanent magnet sizes suitable for microelectronic applications to be obtained.

[0011] Furthermore, this manufacturing process requires a temperature of 1000°C and high pressures that are not compatible with MEMS production steps. This then requires manufacturing the magnet separately from the rest of the MEMS, then integrating it within the MEMS: this is tedious to implement in production lines and requires the magnet to be on a macroscopic scale. Indeed, microscopic scale magnets cannot be easily integrated into MEMS in this way: it requires very complex handling steps.

[0012] Barium and strontium hexaferrites, on the other hand, are the cheapest source of permanent magnet material. However, due to their weaker magnetization, they are not as magnetically effective as rare earth permanent magnets.

[0013] This is why we are always looking for new permanent magnet materials: free of rare earths, and which in addition have good magnetic properties, can be manufactured at low cost and easily in production lines, and can be integrated into MEMS or magneto-optical devices as detailed above during the production of these MEMS and devices. This requires that their integration be carried out at low temperature and in a manner compatible with mass production, have an appropriate dimensioning (namely in particular a microscopic scale) for possible integration in production lines during the manufacture of MEMS and magneto-optical devices.

[0014] In the field of microtechnology, different techniques are known for manufacturing permanent magnets in the form of magnetic layers on a substrate.

[0015] These techniques include: laser ablation deposition (also known as “Pulsed Laser Deposition”) which allows the production of a magnetic layer with a maximum thickness of 10 to 20 µm with a quantity of magnetic energy BHmax of 130 kJ / m 3< . However, the implementation temperature of this technique is 700°C; which is not compatible with MEMS. sputtering which allows the production of a magnetic layer with a quantity of magnetic energy BHmax of 400 kJ / m 3< but which also requires an annealing temperature between 400°C-700°C incompatible with MEMS and which has limitations as to the achievable thickness: of the order of a few micrometers, and at most 50 µm. Thus, greater thicknesses, for example of the order of 100-1000 µm, are not conceivable with this technique, even though they are highly sought after for MEMS, for example to create deflection.the encapsulation of isotropic powder in a polymer matrix. The resulting polymer can then be shaped by molding, spinning or screen printing. This allows the production of a magnetic layer with a maximum thickness of 400 µm with a quantity of magnetic energy BHmax of 53 kJ / m 3< . However, the significant dilution of the magnetic powders in the polymer leads to a significant reduction in the macroscopic properties of the resulting magnet. In addition, the screen printing deposition process conventionally used poses problems for large-scale industrialization. In addition, difficulties in homogeneity of the magnetic powder within the matrix can arise. All these disadvantages thus limit the interest of this technique.Physical Vapor Deposition (PVD), which produces a magnetic layer with a maximum thickness of 10 µm with a magnetic energy quantity BHmax of 120 kJ / m 3< . This technique is not fully satisfactory for certain applications, due to the limitation of the thickness of the magnetic layer to only 10 µm. Electrodeposition, which is a technique compatible with MEMS but which produces materials whose microstructure is not optimized for permanent magnets, and whose BHmax is limited to 30 kJ / m 3< . .

[0016] Thus, in addition to the fact that these different techniques can, for some of them (notably sputtering), be expensive to implement, they are not always fully satisfactory with regard to the thickness of the magnetic layers obtained and / or their magnetic properties for various applications in microtechnology as well as, where appropriate, due to the incompatibility of their production process with MEMS.

[0017] A soft magnet differs from a permanent magnet in that it has a weak coercive field while having a strong saturation magnetization.

[0018] Soft magnets can be used in microelectronics, in high-frequency components, such as inductors, common-mode filters, or even radio-frequency transformers, which generally include magnetic materials.

[0019] The interest in forming soft magnets based on spherical nanoparticles is, among other things, the possibility of having materials with properties in the high frequency range (>GHz) because their diameter is less than the skin thickness (which decreases when the frequency increases).

[0020] Dense assemblies of spherical nanoparticles made of a soft material (Fe, FeCo, FeNi, FeC, FeN, etc.) are interesting because they exhibit both strong magnetization (linked to the nature of the particles and their volume fraction in the assembly) and strong magnetic susceptibilities in the microwave domain.

[0021] For the purposes of this application, the term "soft material" means an object made from a ferromagnetic material which has a high magnetic susceptibility at the desired frequency and a high saturation magnetization.

[0022] Thus, as with permanent magnets, soft magnets have become essential objects in everyday life and represent a significant and promising market.

[0023] A known technique for manufacturing a soft magnet is that of depositing composite materials in thin layers by sputtering or electrodeposition. However, this technique has significant limitations due in particular to losses due to eddy currents, thermal and pressure stresses, the low magnetic content and their transfer, usually carried out by pick-and-place technique.

[0024] Thus, known soft magnet manufacturing techniques are not always satisfactory with regard to the density of the magnetic material obtained and / or with regard to their compatibility with micro-manufacturing techniques.

[0025] Furthermore, document US 2007 / 155025 A1 describes a method of manufacturing a device which comprises depositing a plurality of nanowire structures between a 1st and a 2nd magnetic microelectrodes which are arranged on a substrate, such that a majority of said nanowire structures are parallel to each other and electrically couple said 1st and 2nd magnetic microelectrodes.

[0026] The inventors of the present invention have overcome all these drawbacks detailed above regarding the manufacture of permanent or soft magnets for applications particularly in microtechnology. They have in fact developed a method for manufacturing a permanent or soft magnet which: is easy to implement, at room temperature; is economical in raw materials; offers a variety of sizes: the size of the resulting magnet can be between the micrometer and millimeter scale; for a permanent magnet, can be integrated both in a MEMS so as to obtain a MAGMEMS and in a magneto-optical device as detailed above, as well as in fluidic chips and biochips with magnetic function; for a permanent magnet, can be implemented during the manufacturing of a MEMS or a magneto-optical device or in fluidic chips and biochips with magnetic function or, if necessary, can be manufactured independently of the rest of the MEMS or the magneto-optical device or the fluidic chips and biochips with magnetic function and then be integrated within the MEMS or the magneto-optical device or the fluidic chips and biochips with magnetic function;for a soft magnet, allows to increase the density of the magnetic material obtained and is compatible with micro-fabrication techniques. ;

[0027] In addition, the permanent magnet obtained with the manufacturing process has magnetic properties that are ideally suited for the magneto-optical applications, as well as the microbiology, telecommunications, and automotive sensor applications detailed above. The soft magnet obtained with the manufacturing process has magnetic properties that are ideally suited for the microelectronic applications detailed above.

[0028] The invention therefore relates to a method for manufacturing a permanent or soft magnet which is characterized in that it comprises at least the following steps: a) there is provided: a solution containing at least one solvent in which are dispersed a set of objects which carry a permanent magnetic moment; a substrate on which are fixed to the surface or within a cavity which it may have at least a 1st pad and a 2nd pad which are made of a ferromagnetic material, said 1st pad having a face opposite a face which the 2nd pad has, said facing faces being parallel to each other; b) the solution is deposited on the surface of the substrate or, where appropriate, within the cavity, such that the 1st and 2nd pads are at least partly immersed in said solution;c) the substrate is placed in a magnetic field oriented in a direction perpendicular to the faces of the 1st plot and the 2nd plot which are opposite and parallel to each other so that at least a portion of the set of objects are grouped between said face of the 1st plot and said face of the 2nd plot which are opposite and parallel to each other and that each of these objects is oriented in the direction of the applied magnetic field, so as to form a permanent or soft magnet; d) optionally, at least one washing of the substrate is carried out with at least one solvent; e) optionally, the at least one dispersion solvent of the set of objects and, where appropriate, the at least one washing solvent are totally or partially evaporated. ;

[0029] Thus, during the manufacturing process, at least part of the set of objects which each carry a permanent magnetic moment are grouped by magnetophoresis between the face of the 1st plot and the face of the 2nd plot which are opposite and parallel to each other so as to obtain a macroscopic permanent or soft magnet between these two plots.

[0030] The principle of magnetophoresis-directed assembly is based on the movement of magnetic objects within a colloidal suspension in the presence of a non-uniform magnetic field. The objects are attracted to the positive magnetic field gradients, they assemble there and compact during the evaporation of the solvent. The application of the magnetic field therefore has two effects. The magnetic field creates a torque on the magnetic moment carried by the objects that aligns them in the direction of the field, and the positive magnetic field gradients created by the ferromagnetic pads on the support allow these objects to be moved and located in a dense manner.

[0031] Thanks to the two pads made of ferromagnetic material, at least a part of the set of objects that carry a permanent magnetic moment are structured between these two pads during step c) so that this part of the set of objects forms a permanent or soft magnet. In other words, the permanent or soft magnet obtained at the end of the manufacturing process results from the spatial structuring of at least a part of the set of objects that carry a magnetic moment. This spatial structuring is obtained thanks to the two pads made of ferromagnetic material which have two faces facing and parallel to each other and to the application of a magnetic field which is oriented in a direction perpendicular to these two faces.

[0032] In one embodiment of the invention, in step c), all of the objects are grouped together between said face of the 1st pad and said face of the 2nd pad which are opposite and parallel to each other and each of these objects is oriented according to the direction of the applied magnetic field, so as to form a permanent or soft magnet.

[0033] The manufacturing method according to the invention can be implemented within a device being manufactured, said device being able to be chosen from: MAGMEMS (e.g. for the applications described above), magneto-optical devices (e.g. those described above), fluidic chips and biochips with magnetic function (e.g. for cell sorting, separation of biological objects, capture of cells and molecules).

[0034] In other conceivable embodiments of the invention, the permanent or soft magnet is manufactured independently of the rest of the device in which it will then be integrated. The device in which the permanent magnet will be integrated may be one of those described just above.

[0035] Thus, the manufacturing method according to the invention has the advantage of being flexible in its implementation and can therefore be perfectly adapted to cases which require the permanent or soft magnet to be manufactured within the device in which it is integrated or, on the contrary, to cases for which it is necessary for the permanent or soft magnet to be integrated into the device after its manufacture. This flexibility of implementation is not always offered by the prior art permanent or soft magnet manufacturing techniques which have been recalled above.

[0036] Objects that carry a permanent magnetic moment can be chosen from a variety of objects.

[0037] As regards their material characteristics, these may be objects which are made of a metal chosen from cobalt, iron, nickel or platinum, the carbides of these metals, the nitrides of these metals, taken alone or as a mixture of these (for example, mixtures of iron and platinum or cobalt and platinum are possible, as well as the carbides or nitrides of mixtures of these metals).

[0038] For example, for the manufacture of a permanent magnet, we will preferably choose objects made of cobalt, iron, nickel or platinum, taken alone or in a mixture of these (for example, mixtures of iron and platinum or cobalt and platinum).

[0039] For example, for the manufacture of a soft magnet, we will preferably choose objects made of a soft magnetic material, such as iron, a mixture of iron and cobalt (FeCo), a mixture of iron and nickel (FeNi), iron carbides (FeC), iron nitrides (FeN), taken alone or in a mixture of these.

[0040] Regarding their sizing characteristics, preferably, said objects are nano-objects. This means that they have a size on the nanometric scale. For example, nano-objects can be nano-rods or nanoparticles such as isotropic particles with a diameter between 5 and 100 nm.

[0041] In one embodiment, said objects are cobalt nanorods and the magnet is a permanent magnet.

[0042] For example, cobalt nanorods may have been obtained from a synthesis process that is described in the 2009 publication by Soumare et al., entitled: "Kinetically controlled synthesis of hexagonally close-packed cobalt nanorods with high magnetic coercivity", Advanced Functional Materials, 19, 1971-1977. This synthesis consists of the reduction of cobalt carboxylates in a 1,2-butanediol solution containing sodium hydroxide at a temperature of 170°C for 15 minutes in the presence of ruthenium.

[0043] The synthesis of cobalt nanorods is thus perfectly within the reach of those skilled in the art.

[0044] In another embodiment, said objects are spherical iron carbide nanoparticles and the magnet is a soft magnet.

[0045] For example, iron carbide nanoparticles may have been obtained from a synthesis process that is described in the publication SS Kale et al., "Iron carbide or iron carbide / cobalt nanoparticles for magnetically-induced CO2 hydrogenation over Ni / SiRAlOx catalysts," Catal. Sci. Technol., vol. 9, no. 10, pp. 2601-2607, 2019.

[0046] The synthesis of iron carbide nanoparticles is thus perfectly within the reach of those skilled in the art.

[0047] Before their dispersion in the solution, the objects that carry a permanent magnetic moment (for example nano-objects) are advantageously washed in order to remove any residues, in particular residues from their synthesis (for example excess reagents or reaction side products). This makes it possible to obtain objects that carry a permanent magnetic moment free from impurities that could be detrimental to the device in which the permanent or soft magnet is integrated. The step of washing these objects carrying a permanent magnetic moment is perfectly within the reach of a person skilled in the art.

[0048] In one embodiment of the invention, when said objects are cobalt nanorods obtained at the end of the synthesis process detailed in the aforementioned publication by Soumare et al., they are advantageously washed, in order to eliminate the synthesis residues and the excess ligands. To do this, one or more successive washes can be carried out with a solvent, for example ethanol or chloroform. The solution thus obtained is mechanically stirred, then placed in an ultrasonic bath, so that the particles are redispersed. Said particles are then magnetically attracted and the supernatant is removed.

[0049] For example, in one embodiment of the invention, 540 mg of cobalt nanorods dispersed in 120 mL of 1,2-butanediol were obtained from the synthesis method described in the aforementioned publication by Soumare et al.. This corresponds to a molar concentration of cobalt of 0.08 mol / L. Then, a volume of between 4 and 10 mL of this cobalt nanorod solution is taken and washed in the following manner: a volume of 10 mL of solvent (ethanol or chloroform) is added. The solution obtained is mechanically stirred for 20 seconds, then placed in an ultrasonic bath for 5 minutes. The particles are magnetically attracted and the supernatant is removed. These steps are repeated between 4 and 6 times, for example between 2 and 3 times with ethanol and between 2 and 3 times with chloroform.

[0050] Objects that carry a permanent magnetic moment are dispersed in a solution that contains at least one solvent.

[0051] The dispersion solvent is advantageously chosen from those which have the following properties, taken alone or in combination: a good dispersal capacity for objects carrying a permanent magnetic moment; the following evaporation characteristics: complete evaporation at room temperature and atmospheric pressure or complete evaporation under vacuum and / or at a moderate temperature (i.e. below about 100°C), and this after a maximum of about one hour.

[0052] Furthermore, the dispersion solvent is preferably chosen from solvents which can be used during the manufacture of MEMS or magneto-optical devices, fluidic chips and biochips with magnetic function, without risk of damaging them.

[0053] This dispersion solvent may in particular be chosen from anisole, chloroform, toluene, chlorobenzene and mesitylene, taken alone or in mixtures thereof.

[0054] The concentration of objects that carry a permanent magnetic moment in the solution can be between 5.10 15< objects / L and 10 18< objects / L, preferably between 5.10 16< objects / L and 2.10 17< objects / L.

[0055] In one embodiment of the invention, when said objects which carry a permanent magnetic moment have a cylindrical shape of length 100 nm and diameter 15 nm, the volume fraction of said objects in the solution is between 0.01% and 10% and preferably between 0.1% and 2%.

[0056] In the example described above of cobalt nanorods, the volume of the dispersion solvent can be between 300 µL and 2.5 mL. This makes it possible to obtain a molar concentration of cobalt between 0.13 mol / L and 2.7 mol / L in the solution. In this embodiment of the invention, the solvent is preferably anisole.

[0057] The substrate can be of very different shapes, sizes and materials.

[0058] In one embodiment, the substrate may be a part of the device in which the permanent magnet obtained at the end of the method according to the invention is integrated. Thus, it may be a part of a MAGMEMS, a magneto-optical device, a fluidic chip or a biochip with magnetic function.

[0059] As will be explained below, the pads may be attached to the substrate by electrodeposition. Therefore, in embodiments of the invention, the substrate is selected from substrates suitable for electrodeposition.

[0060] The substrate is made of a non-ferromagnetic material that is compatible with the solvent in which the set of objects carrying a permanent magnetic moment are dispersed.

[0061] The substrate may be made of a material chosen from silicon (optionally functionalized with molecules such as octadecyltrichlorosilane or perfluorodecyltrichlorosilane), glass, polymers not soluble in organic solvents (for example polytetrafluoroethylene, hereinafter abbreviated as “PTFE”), metals and silica.

[0062] At least 2 pads are fixed on the substrate. As explained above, the 2 pads can be fixed to the surface of the substrate or within a cavity that the substrate may contain.

[0063] In one embodiment of the invention, a plurality of pads are fixed to the substrate such that each pad has at least one face opposite a face of at least one other pad, these opposite faces all being parallel to each other. This has the advantage of manufacturing several permanent magnets at the same time.

[0064] The two opposite faces of two pads may be separated by a distance of between 10 µm and 1 cm, preferably between 50 µm and 1 mm.

[0065] The studs are made of a ferromagnetic material. This ferromagnetic material may be chosen from nickel, cobalt, steel, nickel and iron alloys (for example permalloy), iron and cobalt alloys, iron and platinum alloys, cobalt and platinum alloys, nickel and iron alloys and nickel, cobalt, manganese and phosphorus alloys, taken alone or as a mixture thereof.

[0066] In a first embodiment of the invention, the pads are fixed to the surface of the substrate. For example, they can be fixed by electrodeposition, sputtering or atomic layer deposition.

[0067] In this 1st embodiment of the invention, the pads can be fixed to the surface of the substrate by means of a bonding layer. This bonding layer can be implemented in particular when the pads are fixed by electrodeposition to the surface of the substrate. For example, this bonding layer consists of the superposition of a layer of copper on a layer of titanium.

[0068] In this 1st embodiment of the invention in which the pads are fixed to the surface of the substrate: the substrate may have any three-dimensional shape. For example, the substrate may have a base with an area of between 4 mm 2 and 2,500 cm 2 , preferably between 4 cm 2 and 100 cm 2 and a height measured from this base of between 10 µm and 10 cm, preferably between 250 µm and 1 mm, the pads may have any three-dimensional shape. For example, the pads may have a base which is fixed to the surface of the substrate and whose area is between 50 µm 2< and 1 cm 2< , preferably between 2500 µm 2< and 1 mm 2< , and a height measured from this base of between 10 µm and 1 mm, preferably between 50 µm and 500 µm, the two faces facing and parallel to each other of the two pads are advantageously separated by a distance of between 10 µm and 1 cm, preferably between 50 µm and 1 mm.

[0069] In this 1st embodiment of the invention, the substrate may comprise between 2 and 100, preferably between 2 and 10, pads which are fixed to its surface, each pad having at least one face opposite a face of another pad, said opposite faces all being parallel to each other. This makes it possible to manufacture several permanent magnets at the same time.

[0070] As an example of this 1st embodiment of the invention, the substrate may be a cylindrical wafer (made of a material among those described above, for example silicon) and whose diameter is between 4 mm and 50 cm and whose height is between 50 µm and 1 mm. The pads are made of a ferromagnetic material (for example one of those described above, in particular nickel) and may have a parallelepiped shape whose height is 200 µm, the width of 500 µm and the length of between 100 µm and 1000 µm.

[0071] In a 2nd embodiment of the invention, the pads are fixed within a cavity presented by the substrate.

[0072] In a 1st variant of this 2nd embodiment: the substrate may have any three-dimensional shape. For example, the substrate may have a base whose area is between 1 mm 2< and 100 cm 2< , preferably between 25 mm 2< and 4 cm 2< , and a height measured from this base of between 1 mm and 10 cm, preferably between 5 mm and 2 cm, the substrate may have at least one cavity whose volume is between 1 mm 3< and 500 cm 3< , preferably between 125 mm 3< and 8 cm 3< , and a depth of between 1 mm and 10 cm, preferably between 5 mm and 2 cm, the pads which are fixed within the cavity of the substrate may have any three-dimensional shape.For example, the pads may have a base which is in contact with the substrate, the area of which is between 0.01 mm 2< and 1000 mm 2< , preferably between 0.05 mm 2< and 50 mm 2< and a height measured from this base of between 100 µm and 5 mm, preferably between 500 µm and 1 mm, the two faces facing and parallel to each other of the two pads are advantageously separated by a distance of between 100 µm and 10 mm, preferably between 500 µm and 1 mm.

[0073] In this 1st variant of the 2nd embodiment, the studs can be embedded in the wall of the substrate.

[0074] In these embodiments of the invention described just above, the permanent or soft magnet is manufactured outside the device in which it will be integrated.

[0075] For example, in this 1st variant of the 2nd embodiment, the substrate can be made of PTFE and the pads can have a cylindrical shape with a base in the shape of a disc which is fixed to the substrate within its cavity, said pads can be made of steel. The two faces facing and parallel to each other of the two pads therefore have the shape of a disc.

[0076] In these embodiments of the invention, the cavity may have a 1st face opposite a 2nd face, a plurality of pads being fixed on the 1st face of the cavity and each having a face opposite a face of one of the pads of the plurality of pads which is fixed on the 2nd face of the cavity, said opposite faces of the pads all being parallel to each other. This has the advantage of manufacturing several permanent or soft magnets at the same time.

[0077] In a 2nd variant of this 2nd embodiment in which the pads are fixed within a cavity presented by the substrate: the substrate can have any three-dimensional shape. For example, the substrate may have a base whose area is between 4 mm 2< and 2,500 cm 2< , preferably between 4 cm 2< and 100 cm 2< and a height measured from this base of between 10 µm and 10 cm, preferably between 250 µm and 500 µm, the substrate may have at least one cavity whose volume is between 500 µm 3< and 100 mm 3< , preferably between 125,000 µm 3< and 0.5 mm 3< , and a depth of between 10 µm and 1 mm, preferably between 50 µm and 500 µm, at least a 1st pad being fixed on a 1st face of the cavity and comprises a face opposite and parallel to a face of a 2nd pad fixed on a 2nd face of the cavity, said pads may have a base which is in contact with the substrate, the area of which is between 50 µm 2 and 1 cm 2, preferably between 2500 µm 2 and 1 mm 2, and a height measured from this base of between 10 µm and 1 mm,preferably between 50 µm and 500 µm, the two faces facing and parallel to each other of the two pads are advantageously separated by a distance of between 10 µm and 1 cm, preferably between 50 µm and 1 mm.

[0078] In this 2nd variant of the 2nd embodiment of the invention, a plurality of pads can be fixed on the 1st face of the cavity, each of these pads having a face opposite the face of one of the pads of the plurality of pads which are fixed on the 2nd face of the cavity, said opposite faces of the pads all being parallel to each other.

[0079] In this 2nd variant of the 2nd embodiment of the invention, the pads may be made of a ferromagnetic material chosen from those described above. The substrate may advantageously be made of a material chosen from those described above for the substrate.

[0080] In this 2nd variant of the 2nd embodiment of the invention, the pads may have been fixed on the 1st face and the 2nd face of the cavity by electrodeposition, cathode sputtering or atomic layer deposition.

[0081] In these embodiments of the invention, a continuous layer of ferromagnetic material may have been electrodeposited on at least a portion of the 1st face of the cavity and on at least a portion of the 2nd face of the cavity which faces and parallel to the 1st face, as well as on a face of the bottom of the cavity connecting these 1st and 2nd faces of the cavity.

[0082] In step b) of the method according to the invention, the solution is deposited on the surface of the substrate or, where appropriate, within its cavity in such a way that the pads are at least partially immersed. For example, the solution can be deposited with a dispenser, a pipette or a micropipette so as to deposit a determined volume of the solution.

[0083] The maximum height that the permanent or soft magnet obtained at the end of the manufacturing process according to the invention can reach is the height of the portion of the pads that will have been immersed in the solution. In other words, if the pads are completely immersed in the solution during step b), the maximum height that the magnet of the permanent or soft magnet can reach corresponds to the height of the pads.

[0084] Indeed, the height of the permanent or soft magnet obtained at the end of the manufacturing process according to the invention will depend in particular on the way in which steps b) to d) described more precisely below have been carried out, and in particular on the number of times they have been repeated. Of course, it is within the reach of those skilled in the art to implement these steps to obtain the desired height of the permanent or soft magnet.

[0085] The height of the permanent or soft magnet obtained with the manufacturing method according to the invention can be between 10 µm and 1 mm.

[0086] In step c) of the method according to the invention, the substrate is placed in a magnetic field such that at least part of the set of objects is grouped between two opposite faces of two pads towards the positive magnetic field gradients and is oriented in the direction of the magnetic field so as to form a permanent magnet. Thus, a magnetic field is applied to the entire substrate, magnetizing the pads of the substrate.

[0087] In step c), the intensity of the magnetic field may be between 100 mT and 1 T, preferably between 400 mT and 1 T. Most preferably, the intensity of the magnetic field is 1 T.

[0088] During step c), the magnetic field may be constant over time or may change over time. For example, the magnetic field may exhibit plateaus and / or ramps over time.

[0089] When the intensity of the magnetic field is not constant over time, for example if it has plateaus, this has the advantage of better densifying the permanent or soft magnet during its manufacture.

[0090] For example, the magnetic field strength can increase from 0 T to 1 T in 20 seconds with a ramp of 50 mT / s.

[0091] In another example, the magnetic field strength can increase from 0 T to 1 T in 400 s with 20 successive steps of 20 seconds duration, the strength increasing by 50 mT between each step.

[0092] The magnetic field can be produced by an electromagnet (e.g. an electromagnet that is used for electron paramagnetic resonance experiments).

[0093] The substrate may be placed in the magnetic field for a period of time between 1 minute and 10 hours, preferably between 5 min and 1 hour.

[0094] During step c), at least part of the solvent in the solution evaporates. Preferably, in step c), the magnetic field is maintained until the solvent in the solution has completely evaporated.

[0095] Then, at the end of step c), when the magnetic field has been stopped, optionally, one or more washes with at least one solvent can be carried out during a step d) of the process according to the invention. The solvent can be chosen from chloroform, toluene, anisole, mesitylene and chlorobenzene. This makes it possible to eliminate the dispersion solvent still present at the end of step c), in other words the dispersion solvent which would not have evaporated during step c).

[0096] Steps b) to c) or, where appropriate when the process comprises a washing step d), steps b) to d) as described above, may be repeated several times. For example, they may be repeated between 1 and 50 times, preferably between 1 and 10 times.

[0097] Preferably, before each new deposition of solution in accordance with step b), the magnetic field is stopped or reduced to an intensity of less than 10 mT so as not to reduce the magnetic properties of the permanent or soft magnet being manufactured.

[0098] At the end of step c), or where appropriate at the end of step d), it is optionally possible to carry out a step e) which consists of the total evaporation of the solvents (namely the solvent of the solution and any washing solvents).

[0099] Step e) of evaporation can be carried out by applying a magnetic field whose intensity is between 100 mT and 1 T, preferably between 400 mT and 1 T, and this for a duration of between 1 minute and 10 hours, preferably between 1 minute and 1 hour. The duration of the evaporation will depend on the solvent to be evaporated. In an embodiment of the invention in which the solvent to be evaporated is chloroform, the intensity of the magnetic field can be 1 T and the duration 5 minutes.

[0100] Step e) of evaporation can be controlled: by acting on the temperature of the substrate (which can be adjusted in particular with a Peltier module), by controlling the ambient pressure (for example by establishing a partial vacuum with a membrane pump for example).

[0101] Furthermore, the lower the boiling point of the dispersion solvent, the easier it will be to evaporate either during step c) or during step e) with the washing solvent.

[0102] In the embodiments of the invention in which the permanent or soft magnet is manufactured within a cavity between two pads which each have a base which is in contact with the substrate and whose area is between 0.01 mm 2 and 1000 mm 2 (preferably between 0.05 mm 2 and 50 mm 2 ) and a height measured from this base between 100 µm and 5 mm (preferably between 500 µm and 1 mm), the permanent or soft magnet can be recovered without difficulty at the end of the manufacturing method according to the invention. As explained above, in these embodiments of the invention, the permanent or soft magnet is manufactured outside the device in which it will be integrated.

[0103] In embodiments of the invention in which the permanent or soft magnet is intended to remain integrated on the substrate (for example, part of a MEMS or a magneto-optical device for a permanent magnet), a step of removing the pads may be implemented. For example, this step may consist of chemical etching, in particular if the pads have been fixed by electrodeposition.

[0104] Of course, removing the pads is not mandatory. Indeed, in cases where the permanent magnet is integrated within a MEMS or a magneto-optical device, the pads can be kept fixed on these types of substrate without this preventing their proper functioning.

[0105] It is important to note that the method for manufacturing a permanent magnet according to the invention is perfectly integrable into a conventional microelectronics method, in particular during manufacturing and within a MAGMEMS, a magneto-optical device, a fluidic chip or a biochip with magnetic function.

[0106] A device selected from a MAGMEMS, a magneto-optical device, a microelectronic component, a high-frequency component, a fluidic chip or a biochip with magnetic function is described, comprising: at least one permanent or soft magnet obtained according to the manufacturing method according to the invention as described above; at least one substrate; at least one 1st pad and one 2nd pad which are made of a ferromagnetic material and which are fixed to the surface or within a cavity which said substrate may have, said 1st pad has a face opposite and parallel to a face which the 2nd pad has, said permanent or soft magnet is located between said facing and parallel faces of the 1 er and 2 th plots.

[0107] The characteristics of the substrate, the 1st and 2nd pads may be those described above for the 1st embodiment and the 2nd variant of the 2nd embodiment of the manufacturing method according to the invention.

[0108] MAGMEMS can be chosen from any type of MEMS sensor and actuator with electromagnetic transduction. For example, it can be a magnetic field sensor, energy harvester or relay.

[0109] These MAGMEMS can be used in a wide variety of applications: in microbiology, particularly for microfluidic magnetic separation (e.g. biosensors), in telecommunications (e.g. radiofrequency microswitches) and in automotive sensors.

[0110] The magneto-optical device may be selected from magneto-optical sensors (e.g., fiber optic sensors, rotary sensors, and polarimetric sensors), deflectors, modulators (e.g., microwave modulators), and non-reciprocal components (e.g., isolators and circulators).

[0111] The invention will be better understood with the aid of the detailed description of experiments which are set out below with reference to the appended drawing representing results of experimental data relating to the manufacturing method according to the invention. DESCRIPTION OF FIGURES

[0112] [ FIG. 1 ] is a schematic perspective representation of a silicon substrate on which two nickel pads are fixed. [ FIG. 2 ] is a photograph taken using a scanning electron microscope (hereinafter abbreviated as “SEM”) of a 1st “nickel pad / cobalt permanent magnet / nickel pad” structure which was taken following a 1st implementation of the manufacturing method according to the invention. [ FIG. 3 ] is a graph of profilometric measurements of nickel and cobalt of this 1st structure. [ FIG. 4] is a photograph taken with an SEM of a 2nd structure “nickel pad / permanent cobalt magnet / nickel pad” which was taken after a 2nd implementation of the manufacturing method according to the invention. [ FIG. 5 ] is a graph of profilometric measurements of nickel and cobalt of this 2nd structure. [ FIG. 6 ] is a graph representing the hysteresis curves of these 1st and 2nd structures, as well as of a reference permanent magnet. FIG. 7 ] represents a graph of the evolution of the magnetic induction and the vibration amplitude of a MEMS resonator of the 1st structure. [ FIG. 8 ] represents a graph of the evolution of the magnetic induction and the vibration amplitude of a MEMS resonator of the 2nd structure. [ FIG. 9 ] is a schematic perspective representation of a silicon substrate on which nickel pads are fixed. [ FIG. 10] is a photograph taken using a SEM of the cobalt of a permanent magnet which was taken after a 3rd implementation of the manufacturing process according to the invention. [ FIG. 11 ] is a graph of profilometric measurements of nickel and cobalt of a 3rd structure “nickel pad / cobalt permanent magnet / nickel pad” obtained at the end of this 3rd implementation of the manufacturing method according to the invention. [ FIG. 12 ] is a schematic perspective representation of a PTFE substrate in which a cavity has been made and within which two steel studs are fixed. [ FIG. 13 ] is a graph representing the demagnetization curves of the permanent cobalt magnet obtained according to a 4th implementation of the manufacturing method according to the invention, as well as that of a reference permanent magnet. FIG. 14] represents a graph of the evolution of the magnetic induction and the vibration amplitude of a MEMS resonator of the permanent cobalt magnet obtained according to this 4th implementation of the manufacturing method according to the invention. [ FIG. 15 ] is a schematic perspective representation of a substrate which has a cavity within which pads are fixed. [ FIG. 16 ] is a sectional view along plane P of the figure 15 of the substrate. [ FIG. 17 ] is a photograph taken using an SEM of a “nickel pad / soft iron carbide magnet / nickel pad” structure which was taken after a 5th implementation of the manufacturing method according to the invention, [ FIG. 18 ] is a photograph taken using an SEM of the iron carbide nanoparticles of a soft magnet which was taken at the end of the 5th implementation of the manufacturing method according to the invention, [ FIG. 19] is a graph representing the hysteresis curves of the soft magnet obtained at the end of the 5th implementation of the manufacturing method according to the invention. EXPERIMENTAL PART A: 1st series of experiments: A - I - Carrying out step a) of the manufacturing process according to the invention : Preparation of cobalt nanorods:

[0113] 540 mg of cobalt nanorods dispersed in 120 mL of 1,2-butanediol were obtained from the synthesis process described in the aforementioned publication: Soumare et al., dating from 2009, entitled: “Kinetically controlled synthesis of hexagonally close-packed cobalt nanorods with high magnetic coercivity”, Advanced Functional Materials, 19, 1971-1977. This corresponded to a molar cobalt concentration of 0.08 mol / L.

[0114] A volume of 10 mL of solution was taken, then washed to remove synthesis residues and excess ligands.

[0115] More precisely, the washing was carried out as follows: a volume of 10 mL of solvent (ethanol or chloroform) was added. The resulting solution was mechanically stirred for 20 seconds, then placed in an ultrasonic bath for 5 minutes. The particles were magnetically attracted and the supernatant was removed. These steps were repeated 4 times: 2 times with ethanol, then 2 times with chloroform.

[0116] After washing, the cobalt nanorods were redispersed in 300 µL of anisole, so as to obtain an anisole solution containing cobalt nanorods with a molar concentration of 2.7 mol / L. Manufacturing and fixing of pads on substrates:

[0117] We had two substrates which were square silicon wafers with sides of 1 cm and a thickness of 500 µm.

[0118] On the 1st silicon wafer two nickel pads were fixed according to a 1st embodiment.

[0119] On the 2nd silicon wafer two nickel pads were fixed according to a 2nd embodiment. 1st form of construction of the plots:

[0120] The attachment of nickel pads to the silicon wafer according to the 1st embodiment was carried out as follows: A 50 nm thick titanium layer, then a 400 nm thick copper layer were deposited on the silicon wafer by sputtering.

[0121] Then, a layer of approximately 200 µm thickness of the resin marketed under the trade name WBR ™ < 2100 by the DuPont company, which is a photoresist resin, was deposited by lamination on the copper layer.

[0122] The resin was exposed by photolithography and developed to obtain wells corresponding to two rectangles: with a width “l” of 500 µm, a length “L 1” of 100 µm, spaced by a distance “d” of 500 µm.

[0123] Then, nickel was electroplated within these two rectangles, with the same thickness as the resin.

[0124] The resin was removed with an acetone bath.

[0125] The copper and titanium outside the two rectangles were etched in a mixture of hydrogen peroxide (diluted 1% by volume in deionized water) and sulfuric acid diluted 1% by volume in deionized water.

[0126] This resulted in a silicon wafer (i.e. the substrate) on which two parallelepiped-shaped nickel pads were fixed by means of a bonding layer consisting of the superposition of the 50 nm thick titanium layer and the 400 nm copper layer.

[0127] There figure 1schematically and in perspective represents this silicon wafer 1a on which are fixed the two parallelepiped-shaped nickel pads 2a which have a width “l”: 500 µm, a length L 1: 100 µm and a height “h1”: 168 µm and which are spaced by a distance “d” of 500 µm. The bonding layer made of titanium and copper is not shown on the figure 1 .

[0128] As can be seen on the figure 1 , the direction of the magnetic field B is perpendicular to the facing faces of the 1st and 2nd plots 2a. 2nd form of construction of the plots:

[0129] The 2nd embodiment of the pads was made in an identical manner to that described for the 1st embodiment except that the length was a length “L 2 ” of 1000 µm and the height was a height “h2 ” of 193 µm.

[0130] This resulted in a silicon wafer (i.e. the substrate) on which two parallelepiped-shaped nickel pads were fixed by means of a bonding layer consisting of the superposition of the 50 nm thick titanium layer and the 400 nm copper layer. A - II - Carrying out steps b) to e) of the manufacturing method according to the invention:

[0131] The silicon wafer with the 1st embodiment of the pads thus obtained was washed with ethanol and acetone, dried with a nitrogen flow and then placed at the bottom of a PTFE mold.

[0132] The mold was placed in an electromagnet that is used for electron paramagnetic resonance experiments and can generate a static magnetic field whose intensity can vary between 0 T and 1 T.

[0133] Then, the sequence of steps b) to d) of the manufacturing process was repeated 8 times in a row, as follows: in step b), 10 µL of the anisole solution containing the cobalt nanorods at a molar concentration of 2.7 mol / L were deposited on the silicon wafer. in step c), a magnetic field of 1 T oriented in a direction parallel to the length L 1 was applied for 5 minutes. During this step c), part of the anisole evaporated. in step d), the silicon wafer was washed with 1 mL of chloroform which was injected into the mold by lateral flow so as to remove the excess anisole which did not evaporate during step c) and the nanorods which did not align between the two nickel pads.

[0134] Before each new repetition (i.e., before each new implementation of step b), the magnetic field intensity was brought back to 0 T.

[0135] Then, at the end of these 8 repetitions of steps b) to d), in order to evaporate all of the anisole, a magnetic field of 1 T oriented in a direction parallel to the length L 1 was applied for 1 hour.

[0136] The manipulations described below were carried out in an identical manner for the 2nd silicon wafer on which the pads are fixed according to the 2nd embodiment, except that the sequences of steps b) to d) were repeated 7 times.

[0137] Thus, the manufacturing process according to the invention was carried out: according to a 1st implementation which resulted in a 1st structure consisting of a 1st permanent cobalt magnet surrounded by two nickel pads; according to a 2nd implementation which resulted in a 2nd structure consisting of a 2nd permanent cobalt magnet surrounded by two nickel pads.

[0138] The physical properties of these 1st and 2nd structures have been studied and are detailed below. A -III - Photographs and profilometric measurements:

[0139] THE figures 2 And 4 are photographs taken with the SEM of the 1st and 2nd structures as described above. These photographs were therefore taken after a 1st and 2nd implementation of the manufacturing method according to the invention.

[0140] The photograph of the figure 2 was taken with a magnification of 150 times and the photograph of the figure 4 with a magnification of 43 times with a SEM marketed by the company JEOL under the commercial name JSM-7800 F.

[0141] On the figure 2 , we can see a part of the first silicon wafer 1a, the nickel pads 2a, as well as the first permanent magnet 3a which results from the grouping of the cobalt nano-rods between these two pads 2a which are separated by a distance “d”.

[0142] On the figure 4 , we can see a part of the 2nd silicon wafer 1b, the nickel pads 2b, as well as the 2nd permanent magnet 3b which results from the grouping of the cobalt nano-rods between these two pads 2b which are separated by a distance “d”.

[0143] Profilometric measurements were carried out with a mechanical profilometer marketed by the company KLA TENCOR under the trade name P-17.

[0144] Profilometric measurements were carried out: on one of the 2a nickel plots at the location indicated by a continuous segment on the figure 2 ; on the 1st permanent magnet 3a at the location indicated by a dotted segment on the figure 2 ; on one of the 2b nickel plots at the location indicated by a continuous segment on the figure 4 ; on the 2nd permanent magnet 3b at the location indicated by a dotted segment on the figure 4 .

[0145] These measurements were thus carried out on the width of the nickel pads and the width of the two permanent magnets obtained at the end of the manufacturing process.

[0146] There figure 3 is a graph representing the evolutions of: the height of the nickel pad 2a as a function of the distance measured from a distance of 300 µm from the end of said pad 2a (continuous curve); the height of the 1st permanent magnet 3a as a function of the distance measured from a distance of 300 µm from the end of said magnet 3a (dotted curve).

[0147] There Figure 5 is a graph representing the evolutions of: the height of the nickel pad 2b as a function of the distance measured from a distance of 220 µm from the end of said pad 2b (continuous curve); the height of the 2nd permanent magnet 3b as a function of the distance measured from a distance of 220 µm from the end of said magnet 3b (dotted curve).

[0148] In view of the figures 3 And 5 , we note that the shape of the profiles of the 1st and 2nd permanent magnets 3a, 3b is similar to that of the profiles of the nickel pads 2a and 2b respectively.

[0149] Indeed, for the two permanent magnets 3a, 3b, we note: a slight reduction in their width (which is 430 µm) compared to that of the nickel pads which is 500 µm. a height of 144 µm for the 1st permanent magnet 3a close to that of the nickel pad 2a which is 168 µm, a height of 140 µm for the 2nd permanent magnet 3b close to that of the nickel pad 2b which is 193 µm. A-IV- Magnetic characterizations of the 1st and 2nd structures “nickel pad / permanent cobalt magnet / nickel pad”:

[0150] A magnetometric measurement was carried out to determine the magnetic properties at room temperature of the 1st and 2nd “nickel pad / cobalt permanent magnet / nickel pad” structures as described above.

[0151] There figure 6represents the hysteresis cycles of: the 1st structure (very tight dotted curve); the 2nd structure (dotted curve); a so-called “reference” magnet (continuous curve).

[0152] The reference magnet was obtained in a PTFE mold from the dispersion in 300 µL of chloroform of 2 mg of cobalt nanorods manufactured according to the synthesis described above, then dried at room temperature in a 1 T magnetic field generated by an electromagnet in a PTFE mold.

[0153] Unlike the permanent magnets of the 1st and 2nd structures, this reference magnet was not spatially structured. The cobalt nanorods were simply spread over the entire surface of the mold and aligned according to the direction of the applied magnetic field. There was therefore no structuring of the nanorods to form a structured volume as is the case with the manufacturing method according to the invention. Thus, the hysteresis cycle of the reference magnet corresponds to the intrinsic magnetic properties of the cobalt nanorods.

[0154] Table 1 below details the determined values of the coercive field H c (in kA / m) and the remanent magnetization µ 0 M r (in mT) for the 1st and 2nd structures, as well as for the reference magnet. [Table 1]

[0155] Table 1 detailing the magnetic properties of the 1st and 2nd structures and the reference magnet µ 0 M r (mT) Hc (kA / m) 1st structure 440 250 2nd structure 248 26 reference magnet 780 330

[0156] The significant differences in magnetic properties noted in Table 1 and on the figure 6 between the 1st and 2nd structures are mainly explained by the different quantity of nickel present in these structures. Nickel is a soft ferromagnetic material with a very low coercive field (0.1 kA / m). The 2b pads that were used for the manufacture of the 2nd permanent magnet 3b had a volume 10 times greater than the 2a pads; which contributed to significantly reducing the magnetic properties of this 2nd permanent magnet 3b.

[0157] In order to optimize the magnetic properties of such structures, it is therefore necessary to minimize the length L of the nickel pads. For technological reasons related to electroplating, the minimum value of the length "L" corresponds to a quarter of the height "h" of the nickel pad, i.e. for the case of this first series of experiments for a thickness of approximately 200 µm: a minimum length of 50 µm.

[0158] The magnetic properties thus obtained for the 1st and 2nd structures are very satisfactory, in particular for the integration of the permanent magnets they contain within devices such as MAGMEMS and magneto-optical devices.

[0159] Certainly, it is noted that the magnetic properties of the 1st and 2nd structures which were obtained at the end of the manufacturing process according to the invention are weaker than those of the reference magnet which, let us recall, correspond to the intrinsic properties of the cobalt nano-rods.

[0160] In order to improve the magnetic properties of these structures and thus bring them closer to those of the reference magnet, it would be possible to change the nature of the ferromagnetic material of the pads, for example by using an alloy of nickel and iron (e.g. a permalloy) or an alloy of cobalt, nickel, manganese and phosphorus. This would make it possible to modulate the radiated magnetic induction by adjusting the magnetization and coercivity of the pads.

[0161] Furthermore, the magnetic properties of these 1st and 2nd structures being reduced by the presence of nickel pads, the etching of the latter would make it possible to have nickel-free permanent magnets and therefore improve the magnetic properties.

[0162] Finally, it should be noted that the weaker magnetic properties of the 1st and 2nd structures compared to those of the reference magnet are also explained by the differences in shape.

[0163] Indeed, the hysteresis loop of the reference magnet corresponds to a thin layer given its lateral dimensions. The demagnetizing field, that is, the field produced by the magnet inside it, can therefore be considered zero. We then obtain the intrinsic properties of the material. On the other hand, this type of thin layer cannot be considered a magnet, because no field is radiated outside the layer.

[0164] In the case of the 1st and 2nd structures obtained at the end of the manufacturing process according to the invention, the sample is no longer only subjected to the magnetic field applied by the magnetometer, but also to its own demagnetizing field. This results in a reduction of the magnetization within the structure, but also in the generation of an externally generated field; which is perfectly appropriate and usable for MAGMEMS and magneto-optical devices. A - V- Evaluation of the magnetic performances of the permanent magnets obtained with the manufacturing process according to the invention A - V- a- Measurement of magnetic induction by Hall microprobe

[0165] The magnetic induction generated by the 1st and 2nd structures was measured using a magnetic field sensor.

[0166] More specifically, a Hall effect microprobe was instrumented using an electronic assembly and an NI-6341 acquisition card. This magnetic sensor operates thanks to the Hall effect. One pair of electrodes is continuously polarized. The presence of a magnetic field perpendicular to the surface of the Hall cross deflects part of the charges, creating a potential difference measured at the second pair of electrodes. The Hall cross has a width of 10 µm, which allows a local measurement of the magnetic induction generated by the permanent magnets.

[0167] The measurements taken are magnetic induction profiles, i.e. the magnet was brought close (up to a hundred micrometers) to the microprobe, using a displacement stage, at the center of the permanent magnet, i.e. at the position where the magnetic induction is maximum. Then, the sample was moved away from the microprobe. A magnetic induction value was retrieved using a LabView program for each position. A - V- b - Actuation of MEMS resonant devices

[0168] The performance of the 1st and 2nd structures was also measured via their ability to actuate MEMS devices. These were resonators, in the form of lever arms, which vibrate thanks to the Lorentz force resulting from the passage of an alternating current perpendicular to a static magnetic field. To do this, the magnet was brought close to this MEMS up to a limit distance of a hundred micrometers (to avoid any risk of damage).

[0169] The MEMS used consisted of a lever arm of width 285 µm and length 1430 µm.

[0170] The measurement of the MEMS resonance amplitude was carried out using piezoresistors implanted in the structure.

[0171] The deflection of the lever arm causes a maximum stress variation at this location. This changes the resistance of the piezoresistor, which is converted into voltage and measured using an electronic circuit and an acquisition card. A - V- c - numerical simulation

[0172] In addition, the magnetic induction of these 1st and 2nd structures was simulated using Comsol Multiphysics software, taking into account as parameters the cobalt volume fractions of 25%, 30% and 50%.

[0173] Comparing the evolution of the magnetic induction curve obtained experimentally with the Hall effect microprobe on the 1st and 2nd structures with those of the simulations obtained with different cobalt volume fractions allows us to have an estimate of the cobalt volume fraction in the permanent magnet of the structure considered. Indeed, when the magnetic induction curve obtained experimentally for a structure is close to a simulated magnetic induction curve with a given value of the cobalt volume fraction, this means that the cobalt volume fraction in the permanent magnet of this structure is close to this given value.

[0174] Also, the higher the cobalt volume fraction, the better the magnetic properties of the permanent magnet in the structure. In particular, a cobalt volume fraction of 50% is quite advantageous. Therefore, the values of 25%, 30% and 50% cobalt volume fractions were chosen to simulate magnetic induction. AVd-Results

[0175] On the figure 7 are represented: the evolution of the magnetic induction of the 1st structure as a function of the distance between the center of the permanent magnet 3a and the microprobe (“black filled squares”); the simulated evolution of the magnetic induction of the 1st structure taking into account a volume fraction of 25% of cobalt as a function of the distance between the center of the permanent magnet 3a and the microprobe (“continuous curve”); the simulated evolution of the magnetic induction of the 1st structure taking into account a volume fraction of 50% of cobalt as a function of the distance between the center of the permanent magnet 3a and the microprobe (“dot curve”); the evolution of the vibration amplitude of the MEMS resonator with the 1st structure as a function of the distance between the center of the permanent magnet 3a and the MEMS (“squares with black outline”).

[0176] On the figure 8 are represented: the evolution of the magnetic induction of the 2nd structure as a function of the distance between the center of the permanent magnet 3b and the microprobe (“black filled squares”); the simulated evolution of the magnetic induction of the 2nd structure taking into account a volume fraction of 30% of cobalt as a function of the distance between the center of the permanent magnet 3b and the microprobe (“continuous curve”); the simulated evolution of the magnetic induction of the 2nd structure taking into account a volume fraction of 50% of cobalt as a function of the distance between the center of the permanent magnet 3a and the microprobe (“dot curve”); the evolution of the vibration amplitude of the MEMS resonator with the 2nd structure as a function of the distance between the center of the permanent magnet 3b and the MEMS (“squares with black outline”).

[0177] In view of the figures 7 and 8, we note that the magnetic induction measured at 130 µm from the center of the magnet is significantly greater for the 1st structure compared to that of the 2nd structure (26mT vs 16mT) and it decreases to reach a similar value at 500 µm (5.5 mT vs 4mT).

[0178] Furthermore, these values are lower than those expected for a magnet comprising a volume fraction of 50%. This can be explained by the fact that a quantity of the anisole did not evaporate completely at the end of the manufacturing process according to the invention, which resulted in internal porosity during drying. This can be resolved by using other, more volatile solvents and by controlling the densification of the permanent magnet, for example with the pressure and temperature parameters.

[0179] Furthermore, it is noted that the vibration amplitude of the MEMS resonator decreases in the same way as the measured magnetic induction. This confirms that the cobalt permanent magnets of these 1st and 2nd structures provide sufficient energy to actuate MEMS devices.

[0180] The superposition of the magnetic induction curves (squares filled in black) and those of vibration amplitude (squares with black outline) validates that the actuation is indeed generated by the presence of a permanent magnet. B: 2nd series of experiments: B- I - Carrying out step a) of the manufacturing process according to the invention : Preparation of cobalt nanorods:

[0181] The nanorods were prepared in the same way as in part AI relating to the 1st series of experiments.

[0182] 3 mL of suspension were collected. After washing, the cobalt nanorods were redispersed in 360 µL of anisole, to obtain an anisole solution containing cobalt nanorods at a molar concentration of 0.66 mol / L. Manufacturing and fixing of the pads on a substrate :

[0183] We had a substrate which was a square silicon wafer with sides of 1 cm and a thickness of 500 µm.

[0184] 160 nickel pads were attached to the silicon wafer by electrodeposition. A 50 nm thick titanium layer and then a 400 nm thick copper layer were deposited on the silicon wafer by sputtering.

[0185] Then, a layer of approximately 30 µm thick of the resin marketed under the trade name AZ ®< 40 XT by the company MicroChemicals GmbH, which is a photoresist resin, was spin-coated onto the copper layer.

[0186] The resin was exposed by photolithography and developed to obtain wells corresponding to two rectangles: with a width “l” of 100 µm, a length “L 1” of 100 µm, spaced by a distance “d” of 100 µm.

[0187] Then, nickel was electroplated within these two rectangles, with the same thickness as the resin.

[0188] The resin was removed with an acetone bath.

[0189] The copper and titanium outside the two rectangles were etched in a mixture of hydrogen peroxide (diluted 1% by volume in deionized water) and sulfuric acid diluted 1% by volume in deionized water.

[0190] This resulted in a silicon wafer (i.e. the substrate) on which 160 parallelepiped-shaped nickel pads were fixed by means of a bonding layer consisting of the superposition of the 50 nm thick titanium layer and the 400 nm thick copper layer.

[0191] Each plot was parallelepiped in shape with a square base of 100 µm on each side and a thickness of 24 µm.

[0192] There figure 9 schematically represents in perspective the 1c silicon wafer on which 160 2c nickel pads are fixed.

[0193] More precisely, the 2c pads are distributed on the silicon wafer in two columns spaced 500 µm apart from each other. Each column comprises 8 rows of 10 2c pads which are spaced 100 µm apart from each other in each row. Each row of pads is spaced 500 µm apart. B- II - Carrying out steps b) to e) of the manufacturing process according to the invention:

[0194] Steps b) to e) were carried out in the same way as in the 1st series of experiments as detailed in part A-II above and were repeated twice.

[0195] Permanent magnets were thus obtained between the nickel pads depending on the direction of the magnetic field that was applied. B -III - Photographs and profilometric measurements:

[0196] There figure 10 is a photograph taken using the SEM with the same equipment described above at the end of the manufacturing process according to the invention with a magnification of 50,000 times which was taken at the level of one of the permanent magnets thus obtained.

[0197] In this photograph we can see the cobalt nanorods and see that they are correctly aligned.

[0198] Profilometric measurements were carried out with a mechanical profilometer as described in the 1st series of experiments.

[0199] These measurements were thus carried out on the width of two nickel pads in two consecutive rows and the width of the two permanent magnets obtained at the end of the manufacturing process from these two pads.

[0200] There figure 11 is a graph representing the evolutions of: the height of the two nickel studs as a function of the distance measured from a distance of 50 µm from the end of one of these studs (continuous curve); the height of the two permanent magnets as a function of the distance measured from a distance of 50 µm from the end of one of these magnets (dotted curve).

[0201] In view of the figure 11, we note that the shape of the profiles of the two permanent magnets is similar to that of the nickel pads. We note a very slight reduction in the width of the two permanent magnets compared to the nickel pads (90 µm versus 100 µm) while the thickness of the permanent magnet is almost identical to that of the nickel pads (23 µm versus 24 µm). C: 3rd series of experiments: C - I - Carrying out step a) of the manufacturing process according to the invention : Preparation of cobalt nanorods:

[0202] The nanorods were prepared in the same way as in part AI relating to the 1st series of experiments.

[0203] After washing, the cobalt nanorods were redispersed in 2.5 mL of chloroform, so as to obtain a chloroform solution containing cobalt nanorods with a molar concentration of 0.13 mol / L. Manufacturing and fixing the pads on a substrate:

[0204] As shown in the figure 12schematically and in perspective, we had a 1d substrate which was a PTFE mold in the center of which a 4d cavity was made.

[0205] The substrate 1d has a general parallelepiped shape of length 2 cm, width 2 cm and height 2 cm in the center of which is formed a cavity 4d of general parallelepiped shape whose dimensions are as follows: length 1 cm, width 1 cm and height 1 cm.

[0206] Cavity 4d has a 1st face 5d opposite a 2nd face 6d.

[0207] On the 1st face 5d is fixed a 1st plot 2d which is in the form of a cylindrical steel bar with a diameter of 1,500 µm and a length of 6 mm.

[0208] On the 2nd face 6d is fixed a 2nd plot 2d which is in the form of a cylindrical steel bar with a diameter of 1500 µm and a length of 6 mm.

[0209] The free ends of the two 2d plots are 2 mm apart.

[0210] The two steel bars are embedded in the substrate 1d. C- II - Carrying out steps b) to e) of the manufacturing process according to the invention:

[0211] The substrate 1d (i.e. the PTFE mold) with the two steel bars 2d was placed in an electromagnet.

[0212] 1.2 mL of the chloroform solution containing cobalt nanorods with a molar concentration of 0.13 mol / L was deposited within the cavity 4d so as to completely immerse the two steel bars 2d.

[0213] Then, a magnetic field of 1 T was applied. After 2 minutes, the excess chloroform was removed via a syringe, under a magnetic field.

[0214] The magnetic field was stopped after one hour, allowing the chloroform residue to completely evaporate.

[0215] The permanent magnet thus obtained was recovered using pliers, having first removed one of the steel bars.

[0216] It had a cylindrical shape with a length of 2 mm. Its diameter is almost constant: at the ends of the permanent magnet, it is 1.3 mm and at the center, it is 1.1 mm. C-III- Magnetic characterizations of the permanent magnet :

[0217] A magnetometric measurement was carried out in order to determine the magnetic properties at room temperature of the permanent magnet obtained with the manufacturing method according to the invention.

[0218] There figure 13 represents the demagnetization curves of: the permanent magnet obtained with the manufacturing process according to the invention (continuous curve); reference magnet as described in the 1st series of experiments (dotted curve).

[0219] Table 2 below details the determined values of the coercive field H c (in kA / m) and the remanent magnetization µ 0 M r (in mT) for each of these 2 magnets. [Table 2]

[0220] Table 2 detailing the magnetic properties of magnets according to the invention and reference obtained with the manufacturing process µ 0 M r (mT) Hc (kA / m) permanent magnet according to the invention 745 382 reference magnet 768 400

[0221] The values of the coercive field H c (in kA / m) and the remanent magnetization µ 0 M r (in mT) of the reference magnet are slightly different from those of the reference magnet of the 1st series of experiments. This is explained by the intrinsic variations in the nature of the cobalt nanorods used, as well as by the variations in their alignment to obtain the reference magnets during these 1st and 2nd series of experiments.

[0222] The drop in remanent magnetization (from 768 mT to 745 mT) and in the coercive field (from 400 kA / m to 382 kA / m) between the reference magnet and the permanent magnet obtained with the manufacturing method according to the invention is low. It is explained solely by taking into account the demagnetizing field in the permanent magnet according to the invention, which lowers its magnetization, but which allows the generation of a magnetic field outside of it. C - IV- Evaluation of the magnetic performances of the permanent magnets obtained with the manufacturing process according to the invention

[0223] On the figure 14 are represented: the evolution of the magnetic induction of the permanent magnet as a function of the distance between the center of the permanent magnet and the microprobe (“black filled squares”); the simulated evolution of the magnetic induction of the permanent magnet as a function of the distance between the center of the permanent magnet and the microprobe (“continuous curve”); the evolution of the vibration amplitude of the MEMS resonator with the permanent magnet as a function of the distance between the center of the permanent magnet and the MEMS (“squares with black outline”).

[0224] In view of the figure 14 , it is noted that the permanent magnet has a magnetic induction of 155 mT at 70 µm and reaches 52 mT at 560 µm.

[0225] These values are higher than those obtained on the 1st and 2nd structures of the 1st series of experiments. Indeed, increasing the size of the permanent magnet makes it possible to reduce the decrease in magnetic induction at the magnet output.

[0226] Furthermore, unlike the 1st and 2nd structures, the measurements were carried out for this permanent magnet in the absence of a soft ferromagnetic element (namely the pads) which reduces the magnetization of the permanent magnet.

[0227] The superposition of the magnetic induction curves (squares filled in black) and those of vibration amplitude (squares with black outline) validates that the actuation is indeed generated by the presence of a permanent magnet.

[0228] There figure 15 schematically and in perspective represents a substrate 1e which is a silicon wafer of parallelepiped shape. Within this wafer 1e is formed a cavity 4e.

[0229] There figure 16 is a sectional view along plane P of the figure 15 of the 1st substrate. On the figure 16, the 1st< 5th face and the 2nd< 6th face, the 7th bottom face of the 4th cavity, as well as two 2nd nickel pads are visible. The two 2nd nickel pads were obtained by electrodeposition of a layer of nickel on the 1st< 5th face, the 2nd< 6th face and the 7th bottom face of the 4th cavity.

[0230] The manufacturing method according to the invention can also be implemented with this substrate 1e which has a cavity within which two pads 2e are fixed. To do this, a solution containing a set of objects which carry a magnetic moment can be deposited within the cavity so as to totally or partially immerse the two pads 2e which have two facing faces. Then, a homogeneous magnetic field is applied so as to group the objects which carry a magnetic moment between these two pads 2e and to orient them according to the direction of the magnetic field. D: 4th series of experiments: D - I - Carrying out step a) of the manufacturing process according to the invention: Preparation of spherical iron carbide nanoparticles

[0231] Iron carbide nanoparticles Fe2.2C are synthesized as reported in [SS Kale et al., "Iron carbide or iron carbide / cobalt nanoparticles for magnetically-induced CO2 hydrogenation over Ni / SiRAlOx catalysts," Catal. Sci. Technol., vol. 9, no. 10, pp. 2601-2607, 2019]. They are spherical and have a diameter of 15 nm.

[0232] 80 mg of iron carbide nanoparticles were obtained from the synthesis process. 20 mg were taken with a spatula and dispersed in 1 mL of anisole, which corresponds to a molar concentration of approximately 3 mol / L. 1st form of construction of the plots:

[0233] In this first embodiment of the pads, a silicon wafer was used with a network of nickel pads. Each pad had a parallelepiped shape with the following dimensions: a width of 100 µm, a length of 100 µm and a height of 25 µm; the pads were spaced at a distance of 100 µm.

[0234] The silicon wafer was washed with ethanol and acetone, dried with a nitrogen flow and then placed on the bottom of a PTFE mold. 2nd form of realization of the plots:

[0235] In this second embodiment of the pads, a silicon wafer was used with two nickel pads facing each other. Each of the two pads had a parallelepiped shape with the following dimensions: a width of 500 µm, a length of 200 µm and a height of 150 µm; the two pads were spaced apart by a distance of 100 µm.

[0236] The silicon wafer was washed with ethanol and acetone, dried with a nitrogen flow and then placed on the bottom of a PTFE mold. D - II - Carrying out steps b) to e) of the manufacturing method according to the invention :

[0237] The PTFE mold with the silicon wafer carrying the 1st embodiment of the pads was placed in an electromagnet which is used for electron paramagnetic resonance experiments and which can generate a static magnetic field whose intensity can vary between 0 T and 1 T.

[0238] Then, the sequence of steps b) to d) of the manufacturing process was repeated twice in a row, as follows: in step b), 10 µL of the anisole solution containing the iron carbide nanoparticles at a molar concentration of 3 mol / L were deposited on the silicon wafer. in step c), a 1 T magnetic field oriented in a direction parallel to the length of the pads was applied for 5 minutes. During this step c), part of the anisole evaporated. in step d), the silicon wafer was washed with 500 µL of chloroform which was injected into the mold by lateral flow so as to remove the excess anisole which did not evaporate during step c) and the iron carbide nanoparticles which did not align between two nickel pads.

[0239] Before each new repetition (i.e., before each new implementation of step b), the magnetic field intensity was brought back to 0 T.

[0240] Then, after these 2 repetitions of steps b) to d), in order to evaporate all of the anisole, a magnetic field of 1 T was applied for 15 min.

[0241] We thus obtained a first structure of a “nickel pads / soft iron carbide magnets / nickel pads” network.

[0242] The manipulations described above were carried out in an identical manner for the silicon wafer bearing the 2nd embodiment of the pads, with the difference that the sequence of steps b) to d) of the manufacturing process was repeated 8 times in a row.

[0243] We thus obtained a second structure “nickel pad / soft iron carbide magnet / nickel pad”. D - III - Shots of the 1st network structure "nickel pads / soft iron carbide magnets / nickel pads"

[0244] There figure 17is a photograph taken with an SEM of the first network structure “nickel pads / soft iron carbide magnets / nickel pads” obtained at the end of the manufacturing process according to the invention as described above. This photograph was taken with a magnification of 160 times with an SEM marketed by the company JEOL under the trade name JSM-7800 F.

[0245] In the figure, we can see the silicon wafer 1f, the regular parallelepipeds corresponding to the nickel pads 2f of height 25 µm, width 100 µm and separated by a distance of 100 µm. We also see that the entire inter_plot space is filled with soft magnets 3c made of iron carbide nanoparticles, each soft magnet 3c having a height corresponding to the height of the nickel pads 2f.

[0246] There figure 18is a photograph taken with an SEM of a 3c soft magnet located between two nickel pads of the 1st network structure "nickel pads / iron carbide soft magnets / nickel pads" of the figure 17 This photograph was taken at a magnification of 300,000 times with a SEM marketed by JEOL under the trade name JSM-7800 F.

[0247] This photograph shows the organization of iron carbide nanoparticles and their high compactness. D - IV - Magnetic characterization of the 2nd structure “nickel pad / soft iron carbide magnet / nickel pad”

[0248] A magnetometric measurement was carried out in order to determine the magnetic properties at room temperature (300°K) of the 2nd structure “nickel pad / soft iron carbide magnet / nickel pad” obtained at the end of the manufacturing process according to the invention as described above.

[0249] There figure 19represents the magnetization hysteresis cycles of this 2nd structure “nickel pad / soft iron carbide magnet / nickel pad”, after removing the residual deposits at the end of the pads.

[0250] A coercive field Hc of 17 kA / m is obtained, showing the soft character of this magnet.

Claims

1. A method for producing a permanent (3a, 3b) or soft (3c) magnet, comprising at least the following steps: a) providing: - a solution containing at least one solvent in which a set of objects which bear a permanent magnetic moment is dispersed; - a substrate (1a, 1b, 1c, 1d, 1e, 1f) on which are fastened to the surface or within a cavity (4d, 4e) that it may comprise at least one first pad (2a, 2b, 2c, 2d, 2e, 2f) and a second pad (2a, 2b, 2c, 2d, 2e, 2f) which are made of a ferromagnetic material, said first pad (2a, 2b, 2c, 2d, 2e, 2f) include a face opposite a face of the second pad (2a, 2b, 2c, 2d, 2e, 2f), said opposite faces being parallel to each another; b) depositing the solution on the surface of the substrate (1a, 1b, 1c, 1f) or, where applicable, within its cavity (4d, 4e), such that the first and second pads (2a, 2b, 2c, 2d, 2e, 2f) are at least partially immersed in said solution; c) placing the substrate (1a, 1b, 1c, 1d, 1e, 1f) in a magnetic field (B) oriented in a direction perpendicular to the faces of the first pad and of the second pad (2a, 2b, 2c, 2d, 2e, 2f) which are opposite and parallel to each other, such that at least part of the set of objects becomes densely grouped between said face of the first pad (2a, 2b, 2c, 2d, 2e, 2f) and said face of the second pad (2a, 2b, 2c, 2d, 2e, 2f) which are opposite and parallel to each other, and each of said objects aligns with the direction of the applied magnetic field (B), so as to form a permanent (3a, 3b) or soft (3c) magnet; d) optionally, performing at least one wash of the substrate (1a, 1b, 1c, 1d, 1e, 1f) with at least one solvent; e) optionally, evaporating all or part of the at least one dispersion solvent of the set of objects and, where applicable, of the at least one washing solvent.

2. The production method according to claim 1, characterized in that the objects which bear a permanent magnetic moment are made of a metal selected from cobalt, iron, nickel, carbides of these metals, nitrides of these metals, taken alone or in combination thereof, or mixtures of iron and platinum or cobalt and platinum.

3. The production method according to claim 1 or 2, characterized in that the objects are nano-objects.

4. The production method according to claim 3, characterized in that said objects are cobalt nanorods and the magnet is a permanent magnet.

5. The production method according to claim 3, characterized in that said objects are spherical iron carbide nanoparticles and the magnet is a soft magnet.

6. The production method according to any one of claims 1 to 5, characterized in that the solvent in which the set of objects which bear a permanent magnetic moment is dispersed is selected from anisole, chloroform, toluene, chlorobenzene, and mesitylene, taken alone or in mixtures thereof.

7. The production method according to any one of claims 1 to 6, characterized in that the concentration of the objects which bear a permanent magnetic moment in the solution is comprised between 5.1015 objects / L and 1018 objects / L.

8. The production method according to any one of claims 1 to 7, characterized in that the substrate (1a, 1b, 1c, 1d, 1e, 1f) is made of a material selected from silicon, glass, polymers insoluble in organic solvents, metals, and silica.

9. The production method according to any one of claims 1 to 8, characterized in that the pads (2a, 2b, 2c, 2d, 2e, 2f) are made of a ferromagnetic material selected from nickel, cobalt, steel, nickel-iron alloys, iron-cobalt alloys, iron-platinum alloys, cobalt-platinum alloys, nickel-iron alloys, and nickel-cobalt-manganese-phosphorus alloys, used alone or in combination thereof.

10. The production method according to any one of claims 1 to 9, characterized in that the pads (2a, 2b, 2c) are fastened to the surface of the substrate (1a, 1b, 1c), and in that: - the substrate (1a, 1b, 1c) has a base with an area comprised between 4 mm2 and 2500 cm2 and a height measured from this base comprised between 10 µm and 10 cm, - the pads (2a, 2b, 2c) have a base which is fasten to the surface of the substrate (1a, 1b, 1c) with an area comprised between 50 µm2 and 1 cm2 and a height measured from this base comprised between 10 µm and 1 mm, - the two opposite and parallel faces of the two pads (2a, 2b, 2c) are separated by a distance comprised between 10 µm and 1 cm.

11. The production method according to claim 10, characterized in that the substrate (1a, 1b, 1c) comprises between 2 and 100 pads (2a, 2b, 2c), each pad having at least one face opposite a face of another pad (2a, 2b, 2c), said opposite faces all being parallel to each other.

12. The production method according to any one of claims 1 to 9, characterized in that the pads are fastened within a cavity (4d) formed in the substrate (1d), and in that: - the substrate (1d) has a base with an area comprised between 1 mm2 and 100 cm2 and a height measured from this base comprised between 1 mm and 10 cm, - the substrate (1d) has at least one cavity (4d) with a volume comprised between 1 mm3 and 500 cm3 and a depth comprised between 1 mm and 10 cm, - the pads (2d) have a base which is in contact with the substrate (1d) with an area comprised between 0.01 mm2 and 1000 mm2 and a height measured from this base comprised between 100 µm and 5 mm, - the two opposite and parallel faces of the two pads (2d) are separated by a distance comprised between 100 µm and 10 mm.

13. The production method according to claim 12, characterized in that the substrate (1d) is made of polytetrafluoroethylene and the pads (2d) have a cylindrical shape with a base in the form of a disk which is fastened to the substrate (1d) within its cavity (4d), said pads (2d) being made of steel14. The production method according to any one of claims 1 to 9, characterized in that the pads (2e) are fastened within a cavity (4e) formed in the substrate (1e), and in that: - the substrate (1e) has a base with an area comprised between 4 mm2 and 2500 cm2 and a height measured from this base comprised between 10 µm and 10 cm, - the substrate (1e) has at least one cavity (4e) with a volume comprised between 500 µm3 and 100 mm3 and a depth comprised between 10 µm and 1 mm, - at least one first pad (2e) being fastened to a first face (5e) of the cavity (4e) and includes a face opposite and parallel to a face of a second pad (2e) fasten to a second face (6e) of the cavity (4e), - said pads (2e) have a base which is in contact with the substrate (1e) with an area comprised between 50 µm2 and 1 cm2 and a height measured from this base comprised between 10 µm and 1 mm, - the two opposite and parallel faces of the two pads (2e) are separated by a distance comprised between 10 µm and 1 cm.

15. The production method according to any one of claims 1 to 14, characterized in that the height of the permanent magnet obtained at the end of said production method is comprised between 10 µm and 1 mm.

16. The production method according to any one of claims 1 to 15, characterized in that, at step c), the intensity of the magnetic field (B) is comprised between 100 mT and 1 T.