FR2842018A3 - MICRO-COMPONENT INCLUDING AN INDUCTIVE ELEMENT OF THE INDUCTANCE OR TRANSFORMER TYPE - Google Patents

Abstract

Micro-component including an inductive element of the inductance or transformer type, comprising a metal winding (42) of solenoid shape, connected to metallization levels (2) produced within the substrate (1) of the micro-component, and comprising a core magnetic (35) arranged in the center of the winding, characterized in that the magnetic core is formed by the successive stack of elementary layers of material of high magnetic permeability (26,28,30,32), and of elementary layers electrically insulators (25,27,29,31,33).

Description

-1

  MICRO-COMPONENT INCLUDING AN INDUCTIVE ELEMENT OF THE TYPE

INDUCTANCE OR TRANSFORMER.

  Technical field The invention relates to the field of microelectronics. More specifically, it relates to micro-components which may include integrated circuits, on which inductive components, of the inductance or transformer type, are made, comprising a metal winding of solenode shape. This guy

  component can in particular be used in applications of the radio frequency type, and for example in the field of telecommunications.

  The invention makes it possible to obtain inductors having higher performances than the existing components, in particular with regard to the evolution of the quality factors at high frequencies, of the order of a few

Gigahertz.

  Prior techniques In general, it is known, in particular from document EP 1 054 417, to produce inductors or transformers formed of a metallic winding of the solenode type, over an integrated circuit. This type of assembly makes it possible to connect an inductor or a transformer to an integrated circuit, by limiting the connection distance between the inductor and the integrated circuit, since the inductor or the transformer is then directly placed directly above the studs.

  interconnection made in the substrate of the integrated circuit.

  This type of mounting makes it possible not to use the substrate surface to implant passive components, such as inductors or transformers, since these are produced outside the substrate, according to a known technique

under the name of "ABO VE IC".

  -2 To improve the performance of such inductors, it is generally sought to increase the quality factor, corresponding to the quotient of

  coefficient of inductance by the resistance of the winding.

  To do this, one solution consists in equipping the inductance with a magnetic core, made from a material of high relative permeability. However, the structure and the materials used to form the magnetic cores known to date do not allow the operation of this type of inductance at high frequencies, higher than the GigaHertz. Indeed, at high frequency, there is the appearance of large eddy currents, as well as a degradation of the value of the relative permeability. In addition, the magnetic alloys generally used do not have anisotropy properties of their permeability. However, the permeability must remain constant up to high frequencies, typically of 3 GigaHertz, frequency range at which inductances operate in

  some radio frequency applications.

  In the American patent application, published under the number US 2000/0050906, the Applicant has described a solution making it possible to operate this type of inductor at particularly high frequencies. This solution consists in segmenting the magnetic core in a direction perpendicular to the axis of the solenode. However, such a solution has certain drawbacks for achieving the components at satisfactory cost prices, and it

  requires relatively complex production methods.

  One of the objectives of the invention is to allow the operation of high frequency inductors or transformers, while retaining

  satisfactory properties, in particular with regard to their quality factor.

  DESCRIPTION OF THE INVENTION The invention therefore relates to a micro-component including an inductive element of the inductance or transformer type. Such an inductive element comprises a metal winding of solenode shape, connected to metallization levels -3 produced within the substrate of the micro-component. In known manner, such a winding comprises a magnetic core disposed in the center of the winding. According to the invention, the magnetic core is formed by the successive stack of elementary layers of material of high magnetic permeability, and

  electrically insulating elementary layers.

  In other words, the magnetic core included in the metal winding has a laminated magnetic structure, which allows the inductor or the transformer to operate at particularly high frequencies, very strongly limiting the intensity of the eddy currents, due to resistance

  of each elementary magnetic layer.

  In practice, the space between the magnetic core and the metallic winding can be filled with an electrically insulating material, separating the winding from the magnetic core. In this case, the magnetic core may not be

  present only at the inside of the winding.

  In another embodiment, the space between the magnetic core and the metal coil can be free of material, so that the inductor and the core are separated by an air space. This arrangement improves the performance of the component by reducing the parasitic capacity which may exist between the core and the metal winding. In this case, the core being suspended within the winding, it extends beyond the ends of the winding, so as to rest on the substrate or on the layers covering it outside of

the metal winding.

  In practice, the elementary layers of high magnetic permeability can be produced from alloys obtained on the basis of elements chosen from the group comprising Cobalt, Tantalum, Zirconium, Palladium, Platinum, Rhenium, Ruthenium, Niobium, Cadmium and Hafnium. These alloys -4 have a substantially constant relative permeability up to frequencies

of the order of 3 GigaHertz.

  The alloys making it possible to produce the elementary layers of material of high permeability can also be obtained on the basis of elements chosen from the group comprising Cobalt, Zirconium, Tantalum, Cadmium, Hafnium, Lutetium and the rare earths chosen. among the lanthanides. This type of rare earth-based alloy has a substantially constant relative permeability up to frequencies of the order of one GigaHertz. This type of alloy can in particular be used to produce giant magnetic resistors, in particular used in magnetic sensors used in read heads. It can also be used to produce tunneling magnetoresistors, in particular used in fast access magnetic memory cells, or in devices

  actuation of systems based on micro electromechanical structures (MEMS).

  The alloys making it possible to obtain materials with high magnetic permeability having magnetostriction properties can also be obtained on the basis of elements chosen from the group comprising Cobalt, Iron, Silicon, Boron, Tantalum, Nickel, Zinc, Zirconium, Oxygen, Platinum, Palladium, Ruthenium, Rhenium, Lutetium, Manganese and

  elements chosen from lanthanides.

  In practice, the thickness of each layer of high permeability material

  can be between 250 and 10,000 Angstroms.

  In practice, the electrically insulating elementary layers, interposed between the layers of magnetic material of high permeability, can be produced from materials chosen from the group comprising the following elements: SiO2, HfAl309, Hf2ZrO4, Si3N4, benzocyclobutene, taken alone or

in combination.

  -5 In practice, it will be preferred that the component comprises at least one layer of material of low relative permittivity, interposed between the substrate and the layer of high relative permeability situated under the metal winding, in order on the one hand to reduce the parasitic capacitance between the winding and the semiconductor substrate and on the other hand that the magnetic field lines do not pass

  in the layers of the integrated circuit.

Brief description of the figures

  The manner in which the invention and the advantages derived therefrom

  will emerge clearly from the description of the embodiment which follows, given as

  non-limiting example, with the support of the appended figures 1 to 18, representing cross-sectional views of the micro-component in the area of implantation of the inductor, progressively during the stages of the production process. Figure 19 is a view in

  cross section of a micro-component produced according to an alternative embodiment.

  Of course, the thicknesses of the different layers illustrated in the figures are given only for the purpose of allowing the understanding of the invention,

  and are not always related to the actual thicknesses and dimensions.

  Manner of realizing the invention As already mentioned, the invention makes it possible to produce inductors or transformers of solenode shape above existing integrated circuits produced

within substrates.

  In the following description, the invention will be described more precisely in

  the case of the manufacture of an inductor, it being understood that the manufacture of a transformer can be deduced therefrom without difficulty for a person skilled in the art, in

  multiplying the number of windings required.

  Schematically, and as illustrated in Figure 1, the substrate (1) may include a metallization level (2) which can be connected to the rest of the integrated circuit -6 (not shown). This metallization level (2) is covered with a layer

  passivation (3) which is typically made of SiO2.

  The inductors produced in accordance with the invention can be produced according to a process linking the different stages described below, it being understood that some of these stages can be carried out in different ways, while obtaining similar results. Certain steps can also be considered useful, but not essential, and therefore as such be omitted

  without departing from the scope of the invention.

  In the first step, illustrated in FIG. 2, the passivation layer (3) is therefore etched, so as to form the interconnection hole (4),

  revealing the metallization layer (2).

  Subsequently, as illustrated in FIG. 3, a material of very low relative permittivity is deposited, also serving as a barrier layer to the diffusion of oxygen. This layer (5) can for example be made of carbide

silicon (SiC).

  Thereafter, and still as illustrated in FIG. 3, a material of low relative permittivity is deposited, with a thickness of a few microns. This layer (6) of material of low relative permittivity covers the entire surface of the component, including the interior of the interconnection hole (4). The material used can be chosen from the products sold under the names "CYCLOTEN" or "SiLK" by DOW CHEMICALS, "AURORA" by the company ASM, "BLACK DIAMOND" by the company APPLIED MATERIAL, or "CORAL" by the company NOVELLUS , or more generally all materials of low relative permittivity used as

  insulators of interconnection levels of transistors in integrated circuits.

  -7 Subsequently, a new layer (7) barrier to the diffusion of oxygen is deposited, which is then removed by etching directly below the hole

interconnection (4).

  Thereafter, as illustrated in FIG. 4, a new layer (8) of material of low relative permittivity, covering the entire component, is deposited. A layer of hard mask (10) is subsequently deposited,

  which can for example be based on TiN or Al.

  Subsequently, a lithography resin (9) is deposited, making it possible to define an opening on the hard mask layer (10), this opening corresponding substantially to the width of the bottom part of a turn of the future winding. . Thereafter, as illustrated in FIG. 5, the layers (8, 6) of material of low relative permittivity are removed by etching, so as to define a double damascene structure having a wide opening (11) at the level of the upper layer, and a smaller opening (12), directly above the

interconnection pad (4).

  Subsequently, and as illustrated in FIG. 6, a layer (15) barrier to the diffusion of copper is deposited, which occupies the various visible faces of the openings (11, 12) previously made. This layer (15) can be deposited by different techniques, in particular by PVD (for Physical Vapor Deposition), or IMT (for Ionized Metal Plasma) or CVD (for Chemical Vapor Deposition) or ALD (for Atomic Layer Deposition), or even by one

  technique known as E BEAM.

  The materials used to form this barrier layer can be based on materials chosen from the group comprising Ta (x, TaN, TiN, WN,) AN2, Os,

TiVFN, Mo, W, Ru, Rh.

  -8 Thereafter, a copper primer layer (16) is deposited. This

  primer layer (16) can be deposited by the various techniques already mentioned.

  This primer layer (16) is then protected, as illustrated in FIG. 7, by layers (17) of resin deposited on the periphery of the future lower part of the turns of the winding. Thereafter, one proceeds as illustrated in FIG. 8, a deposition of copper by electrolytic means. This deposit (20) fills the trench (12) by forming the interconnection pad connected to the metallization (2). This deposit also fills the opening (11) so as to form the lower part of the turns of the winding. This deposit may also be present above the upper layer (8) of material of low relative permittivity, and therefore form the bases

  (21) of the lateral zones of the future turns of the winding.

  Thereafter, we proceed to the elimination of the layers of resin having allowed

  to delimit the lower part of future winding turns.

  Thereafter, the priming layers and the diffusion barrier layer are removed outside the lower part of the turns of the future winding. This elimination is carried out by selective wet etching based on a bath composed of hydrofluoric acid, citric acid and peroxide.

  of hydrogen or ozone injected by bubbling.

  Thereafter, an annealing step is carried out which makes it possible to standardize the crystalline state of the copper between the priming layers and the electrolytic deposit so as to obtain the structure illustrated in FIG. 9. The crystallization step is composed of different temperature ramps in stages under successively nitrogen, hydrogen and then vacuum. When the optimum temperature for the growth of copper grains and diffusion towards the surface of impurities and other doping species of the metal is reached, an atmosphere is formed which is either hydrogen only or 10% hydrogen and made up of argon or -9 Nitrogen at a temperature between 350'C and 450'C but preferably

  between 3500C and 400'C and more precisely between 3800C and 3950C ..

  Thereafter, one proceeds as illustrated in FIG. 10, to deposit a layer (24) of an insulating material intended to support the future magnetic core, and to electrically separate this core from the metallic winding. This material can typically be benzocyclobutene, parylene, or even PSG (Phospho Silicate Glass) or BPSG (Bor Phosphore Silicate Glass), or even SiO2. Subsequently, as illustrated in FIG. 11, a first elementary layer 25 of electrically insulating material is deposited, which can be chosen from the group comprising SiO2, Hf2AI3O9, Hf2ZrO4, benzocyclobutene,

  or else a mixture of silicon dioxide and silicon nitride (Si3N4).

  This layer can be deposited according to various techniques, known under the acronyms of PECVD (for Plasma Enhanced Chemical Vapor Deposition), or alternatively ALD (for Atomic Layer Deposition). This layer has a

  thickness between 100 and 10,000 Amgstroms.

  Thereafter, a layer (26) of high relative permeability made of a material forming a magnetic alloy is deposited. The materials used to form this alloy can be chosen from the lists mentioned above. This layer of magnetic alloy can be deposited according to different techniques known under the acronyms of MOCVD (for Metal Organic Vapor Deposition) or even PVDUHV (for Physical Vapor Deposition

Ultra High Vaccum), or ALD.

  Thereafter, and as illustrated in FIG. 12, the steps of depositing insulating layers (25, 27, 29, 31, 33) and of magnetic alloy layers (26,

  28, 30, 32) to form a laminated structure (35).

  -10 The number of layers, their thickness and the type of material used for both the magnetic layers and the insulating layers can be chosen according to

of the desired application.

  By way of example, to produce giant magneto-resistances, it is possible to use magnetic alloys based on rare earth, and for example based on Holmium, Iron and Cobalt. Each of the magnetic layers has a thickness of the order of 2000 Angstroms, and are separated by insulating layers of SiO2, with a thickness of 750 Angstroms. Four magnetic layers can

be superimposed.

  For another type of application, and in particular for producing magnetoresistors with a tunnel effect, it is possible to use ten successive layers of magnetic material, produced on the basis of an alloy including Cobalt, Iron, Silicon and Boron, each layer having a thickness of 250 Angstroms, and being separated from the next magnetic layer by an insulating layer of a thickness

  of the order of 20 Angstroms, produced for example in HfZrO4.

  Similar results, for producing tunnel magnetoresistors can be used with the same number of magnetic layers made from the same magnetic alloy, but each having a thickness of 250 Angstroms, and being separated by insulating layers, a thickness of 20

  Angstroms, made for example in Hf2AI309.

  It is also possible to use amorphous alloys of soft magnetic behavior, produced for example by using a stack of five magnetic layers based on alloy of Cobalt, Tantalum, and Zirconium, with a thickness of 1000 Angstroms, separated by insulating layers with a thickness of 100 Angstroms, made from SiO2 and Si3N4. These alloys are more particularly intended for

  make inductors operating at frequencies above 2 GigaHertz.

  -11 Ferrite-based alloys can also be produced, for example produced by a stack of three magnetic layers based on nickel, Zinc, Iron, Oxygen and Zinc / Manganese alloys, with a thickness of approximately 3,500 Angstroms, separated by insulating layers with a thickness of 1000 Angstroms made of benzocyclobutene. Of course, all of the examples given above are only for information, and multiple variants remaining in the same spirit can be derived therefrom. Thereafter, and as illustrated in FIG. 13, etching steps are carried out eliminating the lateral portions of the stack forming the core, to define the shape of the final core. This etching is for example obtained chemically from a plasma source, or by wet treatment after a step of

  lithography to protect the upper face of the stack (23).

  Thereafter, and as illustrated in FIG. 14, a new layer (36) of insulating material, equivalent to the material (24) deposited above the lower part of the turns of the future winding, is deposited. This new layer (36) covers the entire core (35) so as to isolate it from the upper part of the

turns of the future winding.

  Thereafter, and as illustrated in FIG. 15, this layer is etched around the magnetic core, so as to define a trench (38) corresponding to the location of the lateral parts of the turns of the future winding. Subsequently, as illustrated in FIG. 16, a layer (39) barrier to the diffusion of copper is deposited and a layer (40) copper primer inside the trench (38) and above the layer (36) of insulating material, directly above the future upper part of the coil of the winding. These different

  deposits are made as mentioned above.

  -12 Thereafter, as illustrated in FIG. 17, a copper deposit is made (42) allowing the lateral part of one of the turns of the winding to be defined as

than the top.

  Thereafter, as illustrated in FIG. 18, the portions of the primer layer (40) and of the barrier layer (39) to the diffusion of copper are eliminated.

  lying outside the coil of the winding.

  Subsequently, it is also possible to deposit a layer of

  passivation to protect the metal coil.

  Of course, the various vertical trenches (38) intended to receive the lateral parts of the turns are implanted in order to define the

  succession of the different turns of the solenode winding.

  In an alternative embodiment illustrated in FIG. 19, the material filling the space (59) between the magnetic core and the winding can be

  eliminated so as to define an air space between the core and the winding.

  In this case, the core as illustrated in FIG. 19, protrudes laterally from the winding to come to rest by portions (58) on the layers of material of low relative permittivity (6), in order to be suspended from the inside the winding. It follows from the above that the inductors or transformers obtained in accordance with the invention have multiple advantages, and in particular the possibility of operating at frequencies significantly higher than the existing components, and typically at frequencies higher than GigaHertz, all retaining their satisfactory electrical properties,

  especially in terms of quality factor.

-13

Claims (7)

  1 / Micro-component including an inductive element of the inductance or transformer type, comprising a metal winding of solenode shape, connected to metallization levels (2) produced within the substrate (1) of the micro-component, and comprising a magnetic core (35) disposed in the center of the winding, characterized in that the magnetic core is formed by the successive stack of elementary layers (26, 28, 30, 32) of material of high magnetic permeability, and of electrically insulating elementary layers (25, 27, 29, 31, 33).   2 / Micro-component according to claim 1, characterized in that the space between the magnetic core and the metal winding is filled with a   electrically insulating material (36).   3 / Micro-component according to claim 1, characterized in that the space (59) between the magnetic core (35) and the metal winding (42) is free of material, the core extending (58) beyond ends of   the winding so as to rest on the substrate, or the layers (6) covering it.   4 / Micro-component according to claim 1, characterized in that the elementary layers of a material with high magnetic permeability are produced from alloys obtained based on elements chosen from the group comprising Cobalt, Tantalum, Zirconium, Palladium, Platinum, Rhenium,   Ruthenium, Niobium, Cadmium and Hafnium.   / Micro-component according to claim 1, characterized in that the elementary layers of a material with high magnetic permeability are produced from alloys obtained based on elements chosen from the group comprising Cobalt, Zirconium, Tantalum , Cadmium, Hafnium, Luthenium and an earth rare chosen from lanthanides.   -14 6 / Micro-component according to claim 1, characterized in that the elementary layers of a material with high magnetic permeability are produced from alloys obtained based on elements chosen from the group comprising Cobalt, Iron , Silicon, Boron, Tantalum, Nickel, Zinc, Zirconium, Oxygen, Platinum, Palladium, Ruthenium, Rhenium, Lutetium,   Manganese and elements chosen from lanthanides.   7 / Micro-component according to claim 1, characterized in that the electrically insulating elementary layers are produced from materials chosen from the group comprising SiO2, Hf2AI3O9, Hf2ZrO4, Si3N4,   benzocyclobutene, taken alone or in combination.   8 / micro-component according to claim 1, characterized in that the thickness of each elementary layer of high magnetic permeability is   between 250 and 10,000 Angstroms.   9 / micro-component according to claim 1, characterized in that it comprises at least one layer (6) of material of low relative permittivity interposed between the substrate and the stack of elementary layers of high   relative permeability, located under the metallic winding. DEPOSITOR: MEMSCAP   AGENT: LAURENT & CHARRAS Cabinet