EP2901473A1 - Assembly method, of the flip-chip type, for connecting two electronic components, assembly obtained by the method - Google Patents

Abstract

The invention relates to an assembly method for connecting two electronic components together, said components each having an assembly face, wherein the two assembly faces are moved together in what is known as an assembly direction X, and a given force F is applied to one and/or the other of the components, one and/or the other assembly face(s) having: -connection inserts made of rigid material having an elongate longitudinal shape in the assembly direction X; -connection tracks made of material having a hardness less than that of the inserts and having an elongate longitudinal shape transversely to the assembly direction X, wherein, in said method: -the inserts are aligned opposite corresponding tracks such that the inserts and the tracks form in pairs, after assembly, at least one approximately transverse intersection, -the force F is applied so as make the inserts penetrate into the tracks until the assembly is produced.

Description

 Method of assembling two electronic components, Flip-Chip type, assembly obtained by the process

 Technical area

 The present invention relates to a method of assembling two electronic components together, of the type with vertical connections, more commonly called "flip-chip" process in English or "flipped chip", according to which one of the two electronic components is returned to make electrical connections or assemblies vertically.

 The invention makes it possible to obtain an assembly between electronic components irrespective of their interconnection pitch and the spacing between these components.

 The invention more particularly relates to the assembly between a chip and a substrate, for example both made of silicon.

 The invention aims to reduce the force required to achieve the assembly between electronic components.

 The invention has wide applications in all microelectronic devices requiring vertical interconnections (in English "front to front") at very small steps.

 A particularly interesting application of the invention is the production of stacked 3D three-dimensional structures or multi-spectral heterogeneous imagers.

 By "assembly of two electronic components" is meant in the context of the invention as well an assembly of two components of different materials than a two-component assembly made of the same material. In particular, an assembly according to the invention may relate to an assembly between an electronic chip and a substrate, both of which may be made of silicon.

 By "no interconnection" is meant the distance between two connection tracks on the same electronic component.

 By "spacing between components" is meant the spacing between the two facing components defined by the interconnection height.

 State of the art

The so-called "Flip-Chip" technique is a well known technique for achieving mechanical and electrical interconnection, or assembly between two components, such as between a chip and a printed circuit substrate. This technique is called "Flip-Chip" because one of the components, usually the chip, which carries conductors, is returned to put the two components face-to-face to allow the operations of interconnections by bonding between the conductors and metal protuberances, known under the naming in English "bumps" and forming contacts made on the other component, generally a printed circuit substrate.

 According to this technique, it is constantly sought to reduce the spacing between components while allowing a number of connections always higher. However, the three major assembly categories currently known by this technique, namely brazing, thermocompression and the use of adhesives, such as the ACF (acronym for "Anisotropic Conductive Films"), each have their limits for reduce spacing.

 In particular, the low-temperature thermocompression by penetration of conductive inserts into protuberances, as described in the patent application WO 2006/054005, is limited by the very great force to be applied for a very large number of connections to be established and by the feasibility of actual realization of the inserts.

 Also, to overcome the limitations of this method, the applicant has proposed in the patent application WO 2009/115686, the production of conductive inserts in the form of blind tubes whose base is secured to the surface of a component. The conductive inserts have been the subject of a patent application EP2287904 perfection with at least one surface of the open end of an insert left free so as to allow the escape of the gases contained in the insert, when insertion. Various new shapes are provided in this application for conductive inserts, such as open bar, cross-section elements star, cross, with lobes ...

FIGS. 1, 2 and 3 show an assembly between two electronic components according to the above-mentioned application WO 2009/115686. An open and hollow insert 10 in the form of a blind tube of cylindrical section is secured by its base on the surface of a substrate 11 of a first electronic component. A ball 20, typically a solder ball, is secured to the surface of a substrate 21 of a second electronic component 2. The material of the ball 20 is of lower hardness than that of the insert 10. When the thermocompression assembly between the two components 1, 2, the insert 10 is aligned opposite the ball 20 and a substantially constant force F is applied according to the assembly direction X represented by the arrow (FIG. 1), until the assembly is obtained, that is to say the complete insertion of the insert 10 into the ball (FIGS. 2 and 2A) ).

 Although largely satisfactory in particular feasibility of making the inserts, the method according to these patent applications can be further improved.

 In particular, there is a need to reduce the intensity of the force to be applied to achieve the assembly by thermocompression between two microelectronic components, especially when the interconnections between components require the use of material (x) which must not pass to the liquid state during encapsulation steps (in English "packaging") subsequent to the actual assembly.

 It has been demonstrated in publications [1] and [2] that for the same insert made in the form of a microtube, the insertion force required depends not only on the mechanical properties of the ductile material itself (hardness, plasticity , ...) in which the stud is made but also of the insertion surface or in other words of intersection between the insert 10 and the pad of ductile material 20.

 The publication [3] has moreover explained that the insertion force of a tube is directly proportional to the insertion surface Si of the tube in the stud and therefore to its wall thickness and to the length L of its perimeter. . Thus, in the example shown in FIG. 3, the insertion surface Si of an insert in the form of a microtube with a cylindrical section of radius R and a wall thickness e1 is equal to the area of the section of the microtube 10. in a horizontal plane, ie in a plane orthogonal to the assembly direction X, ie Si = 2 * 7i * R * el. The insertion force F is therefore equal to F = k * 2 * 7i * R * el, with k constant.

With regard to the mechanical properties of the ductile material, when it is desired to perform an insertion at ambient temperature, it must be taken into account that the insertion force is, in first order approximation, inversely proportional to both the Young's modulus. the constituent metal of the connection pad and its yield strength at room temperature. On the other hand, of the known ductile metals, those with the lowest Young's modulus are those with melting points closest to ambient temperature. The list of ductile metals with low Young's modulus and low melting temperature is shown in Table 1 below. TABLE 1

 In addition to the above considerations on ductile metals, it is recommended, when thermocompression insertion assemblies take place between an electronic chip and a wafer, or between one chip and another chip, to make connections with materials that must not pass to the liquid state during the encapsulation steps ("packaging" in English) subsequent to the insertion. These packaging steps are up to temperatures of 230 ° C. This temperature of 230 ° C is the temperature at which the electronic components are soldered to an AgCuSn solder circuit board.

 Also, starting from Table 1, only aluminum can meet all the technical constraints. Indeed, indium has a melting temperature lower than 230 ° C, tin is likely to create so-called whiskers defects, lead is prohibited. Zinc (Zn) may be suitable but it remains less advantageous than aluminum (Al), the latter having a high Young's modulus.

 Aluminum nevertheless has a substantial Young's modulus, which has the disadvantages of:

 - require sizing complex insertion machines, especially in case of high number of connections: for example, a number of 106 connections to achieve according to the application WO2009 / 115686 requires the application of a 5000N insertion force;

to require durations dedicated to the insertion, high, these durations being proportional to the insertion force required; - Induce a risk of destruction of the active layers underlying the connection pads, due to the force exerted on the inserts during insertion.

 The inventor has therefore been confronted with the problem of reducing the insertion forces in a given ductile material, more particularly in aluminum.

 The general object of the invention is to propose an improvement of assembly method according to the aforementioned patent applications WO2009 / 115686 and EP2287904 and which overcomes at least some of the disadvantages of the state of the art mentioned above.

 A particular object of the invention is to propose a solution for reducing the insertion force to be applied to achieve the assembly between two microelectronic components according to a method of thermocompression at low temperature and to avoid any risk of electrical short circuit between the conductive inserts of one component and the connection tracks of the other component.

 Presentation of the invention

 To do this, the object of the invention is a flip-chip assembly method of two electronic components to one another, said components each comprising a so-called assembly face, according to which the two are brought closer together. assembly faces one of the other in a said assembly direction X and applies a given force F to one and / or the other of the components, one and / or the other face ( s) assembly comprising:

 connecting inserts of rigid material having an elongated longitudinal shape in the X assembly direction;

 - Connection tracks of material of lower hardness than the inserts and elongate longitudinal shape transverse to the X assembly direction;

 method according to which:

 the inserts are aligned opposite the corresponding tracks so that the inserts and the tracks form two by two, after assembly, at least one substantially transverse intersection,

 - Then apply the given force F to penetrate the inserts in the tracks until the assembly between the two components.

According to a preferred embodiment, in which the connection inserts of a component have a unit wall thickness (s) el and a unitary length of wall (s) L1 while the connection tracks of a component have a thickness unitary e2, the unit thicknesses e1, e2 being the smallest dimensions considered transversely to the assembly direction X, the length L 1 being the largest dimension of wall (s) transverse to the assembly direction X,

 method according to which prior to the alignment of the inserts facing the corresponding tracks, the dimensions el, e2, and Ll are determined so that each insertion surface Si between a given insert and a track is substantially equal to a multiple N of the produces unit thicknesses N = n * el * e2 where n is an integer, this multiple N being much smaller than the cross section S 1 of an insert considered transversely to the assembly direction SI = Ll * el.

 By "much smaller than the cross section SI of an insert", it is meant that a very small portion of the cross section of an insert meets the ductile connecting material during insertion, this very small part being otherwise sufficient to ensure the mechanical rigidity of each connection made. Advantageously, the insertion surface Si is at least 40% less than the cross section SI.

 An insert according to the invention may comprise one or more walls. Also by

"Unitary length of wall (s) L1" means the perimeter delimited by the wall or walls. For example, for a star-shaped wall, the unit length is the sum of all the lengths of the branches of the star, which constitutes the perimeter.

 A connection track according to the invention consists of a continuous metal deposit in all possible forms. Thus, a single insert can be inserted in several parts of the same track, and therefore the insertion surface Si to consider according to the invention comprises all the intersections between a thickness of a track and an insert.

 In other words, the invention essentially consists in very significantly reducing the cross section of the ductile material of a component that each insert of the other component penetrates during connection assembly. The invention is simple to implement, since it is only necessary to make tracks of very thin ductile material without there being any need to modify either the shape or the dimensions or the hard material constituting the inserts or their method of production. known from the aforementioned patent applications WO2009 / 115686 and EP2287904.

Thanks to the invention, the insertion forces to be applied to make the connections can be considerably reduced. As a corollary, we thus eliminate everything complex dimensioning of the insertion machines and considerably reduces the durations necessary for the insertion steps themselves.

 The unit thickness of walls and inserts may be between 0.1 and 1 μιη, preferably between 0.1 and 0.5 μιη.

 The unit thickness of the tracks e2 may be between 0.05 and 2 μιη, preferably between 0.1 and 1 μιη.

 According to an alternative embodiment, the constituent material of the tracks is preferably a ductile metallic material chosen from aluminum Al, indium In, Au gold, tin Sn, lead, Pb, bismuth, antimony Sb, zinc Zn, aluminum-copper alloy AlCu, alloys of SnAgCu, SnAg, AgCu, SnCu. Tracks made of ductile metallic material may advantageously be made by photolithography, of the additive or subtractive type, or by electrolysis of the metal or alloy.

 According to an alternative embodiment, the material constituting the tracks may be a hard metal material selected from Cu copper, Ti titanium, Ti titanium nitride, W tungsten, WN tungsten nitride, Mo molybdenum, or Au, chromium Cr, nickel Ni, platinum Pt. The connection tracks of hard metal material can be made by photolithography, additive or subtractive type. Au, Cu or Ni tracks can be made by electrolytic growth.

 The tracks made of hard metal material are advantageously made according to the same technique as the inserts.

 The inserts according to the invention are preferably blind micro-tubes whose base is secured to one of the components. The inserts and, where appropriate, the connecting tracks are thus advantageously manufactured as described in the patent application WO2009 / 115686 or in the application EP 2287904. When the connecting tracks are produced according to the technique described in these applications, they may have a very large thickness e2, advantageously of the order of 0.1 μιη or 0.05μιη, which allows to reduce even more significantly the forces required for the insertion of the inserts within them. In general, the conductive inserts according to the invention may have any shape: micro-tube, tip, open bar, element cross-section star, cross, with lobes ....

According to an advantageous variant, and in order to better distribute the mechanical stresses during the insertion, the connection track parts to be penetrated by a The same tube is in the form of at least three branches symmetrically distributed around a point of symmetry, preferably at the number of four branches distributed at 90 from one another.

 The force F applied per insert can be very small, preferably less than 5mN, preferably less than 0.8mN, typically equal to 0.5mN.

 Advantageously, the alignment and the application of the force F are carried out at ambient temperature.

 The spacing between the two components corresponding to the height H is preferably between a ratio p / 20 and a ratio equal to p / 2, p being the interconnection step between two connection tracks of a component.

 The interconnection pitch p between two connection tracks of a component may be less than or equal to 50 μιη.

 The spacing between the two components corresponding to the height H is advantageously less than 20 μιη, typically equal to 1 μιη.

 According to an advantageous embodiment, one of the components is a chip and the other component is a printed circuit substrate.

 detailed description

 Other advantages and features of the invention will emerge more clearly from a reading of the detailed description of the invention, given by way of illustration and without limitation with reference to the following figures among which:

 - Figure 1 is a schematic sectional view of two electronic components at an insert and a connection pad according to the prior art, before assembly;

 - Figure 2 is a schematic sectional view of two electronic components at an insert and a connection pad according to the prior art, once assembled;

 FIG. 3 is a view from above of FIG. 2;

- Figure 4 is a schematic top view of two electronic components at an insert and a connecting track according to an example of the invention, once assembled; - Figure 5 is a schematic top view of two electronic components at an insert and a connecting track according to another example of the invention, once assembled;

 - Figures 6 and 6A are views respectively from above and in section of two electronic components at an insert and a connection track, showing in detail their assembly;

 - Figures 7 to 10 are schematic top views of two electronic components at an insert and a connecting track according to four other examples of the invention, once assembled.

 For the sake of clarity, the same references designating the same elements of an electronic component according to the state of the art and of an electronic component according to the invention are used for all of Figures 1 to 10.

 It is specified that the various elements, in particular the connection tracks, according to the invention are represented solely for the sake of clarity and that they are not to scale.

 Figures 1 to 3 relating to an insertion assembly according to the state of the art have already been commented in the preamble. They are not described here in detail.

 FIGS. 4 to 10 show a connection insert 10 inserted in a connection track 20 each belonging to one of two electronic components 1, 2, such as microchips hybridized by means of a press tool. support on the upper component.

 The component 1, which is the one returned, comprises a substrate 11 on which are secured by their base conductive inserts 10 in the form of a blind tube, the inserts all having a height h. The choice of heights h inserts advantageously takes into account the minimum pitch p between interconnections to perform. Thus, preferably the minimum height h is of the order of p / 20 in order to allow non-flatness to be accepted between the components 1, 2 to be assembled. Preferably, the maximum height h is of the order of p / 2 in order to limit the so-called buckling effects that can occur beyond.

To make these conductive inserts 10, it has advantageously been described as described in the patent application WO2009 / 1 15686. Each insert 10 has a wall length L1 and a unit wall thickness el. The unit thickness is here referred to as the average dimension, or average width, of the wall of the insert in one direction. transverse to the longitudinal direction thereof. The directions of thickness, length and height locally forming an orthogonal reference. The unit thickness el of a tube 10 is for example equal to 0.2μιη. Each insert tube 10 may have any shape in cross section as shown diagrammatically in FIG. 4. It may be a square section tube (FIG. 5), with a circular section (FIGS. 7 to 10).

 The component 2 comprises meanwhile a substrate 21 on which connection tracks 20 of the same height H have been made. The choice of heights H of the tracks 20 advantageously takes account of the minimum pitch p between interconnections to be made. Thus, preferably the minimum height H is of the order of p / 20 in order to make it possible to accept non-flatness between the chips to be assembled and the maximum height H is equal to p / 2 in order to allow the complete insertion of a tube 10 of maximum height h. The height H of the tracks 20 is calculated so that it is greater than that of the inserts 10 in order to prevent the hard metal of the inserts 10 from abutting on the circuit (s) below the track (s) 20. insertion. Each track 20 has a unit thickness e2. The unit thickness e2 is for example equal to Ιμιη. Here the unit thickness e2 is the average dimension, or average width, of the wall of the track in a direction transverse to the longitudinal direction thereof. The directions of thickness, length and height locally forming an orthogonal reference.

 Tracks 20 are in the form of linear and vertical patterns. Each track 20 may be in the form of a single elongate strip (FIGS. 4 and 9), of a tube for example of rectangular or square section (FIG. 10), of a cross (FIGS. 5, 7 and 8) with the branches joined by a via.

 According to an alternative embodiment, the constituent material of the tracks 20 is a ductile metal material selected from Al aluminum, indium In, Au gold, Sn tin, lead, Pb, bismuth, antimony Sb, AlCu aluminum-copper alloy, SnAgCu, SnAg, AgCu, SnCu alloys. The ductile metal material tracks may be made by photolithography, additive or subtractive type, or by electrolysis of the metal or alloy.

According to an alternative embodiment, the material constituting the tracks is a hard metal material selected from copper Cu, titanium Ti, titanium nitride TiN, tungsten W, tungsten nitride WN, molybdenum Mo, gold Au, Cr chromium, nickel Ni, platinum Pt. Connection tracks of hard metal material can be made by photolithography, additive or subtractive type. Au, Cu or Ni tracks can be made by electrolytic growth.

According to the invention, the dimensions el, e2, and Ll are determined by calculation so that each insertion surface Si, Sii + Si ₂ + Si ₃ ... between an insert and a given track is substantially equal to one multiple N of the product of the unit thicknesses N = n * el * e2 where n is an integer, this multiple N being much smaller than the cross section SI of an insert considered transversely to the assembly direction SI = Ll * el.

 Thus, by considerably reducing the insertion section compared to that when inserting the complete section of a tube according to the state of the art as shown in FIG. 3, the force of constant insertion that it is necessary to apply between an insert 10 and a connection track 20.

 In other words, according to the invention, the insertion section between a track and an insert is minimized while maintaining it sufficient to obtain the mechanical rigidity of the desired interconnection contact. Depending on the circuit shapes, tracks can be made according to several possible configurations and therefore have a greater or lesser number of unit wall thicknesses e2 intersected by the same insert 10. Thus, the same insert 10 can be inserted into a only thickness e2 (FIGS. 4 and 9), in two thicknesses e2 (FIG. 10), four thicknesses e2 (FIGS. 5 and 8), eight thicknesses e2 (FIG. 7) ...

 The various assembly steps according to the invention are now described.

Step 1: the two components 1, 2 are approached and aligned so as to have each insert 10 facing a part of a connection 20.

 Step 2: a force F is applied in the orthogonal assembly direction X to the faces of the substrates supporting the inserts 10 and tracks 20. The force F is applied to a press tool resting on the upper component 1, which causes the insertion of the inserts 10 in the tracks 20. If denotes the cross section intersected by each insert 10 and is substantially equal to a multiple N of the product of the unit thicknesses N = n * el * e2. The section Si is very small compared to the force applied, the stress generated is very high and each track 20 is then plastically deformed. There is insertion of each insert 10 by plastic deformation of each corresponding track 20.

Step 3: the force F is applied until the insertion of the inserts 10 on their total height h in the connection tracks 20. Step 4: Unload and remove the press tool. The two components 1, 2 are assembled (hybridized) with the electrical connection established between each connection track 20 and each conductive insert 10.

 FIG. 6A shows in detail an assembly at an insert 10 and the corresponding track 20, obtained according to the assembly method which has just been described.

 The insertion force required according to the invention is proportional to the common insertion section between each insert 10 and each track 20. Thus, for example, by choosing a track 20 of unit width e2 much lower than the perimeter circular section of a tube of radius R, an insertion of a tube 10 of radius R into a single track 20 according to the invention (FIG. 9) requires a much lower insertion force than an insertion of the same tube in all its circumference as in the state of the art (Figure 3), in a ratio equal to ε2 / 2 * π * R.

 Indeed, an insertion according to Figure 9 is on a section equal to el * e2 while an insertion according to Figure 3 is on a section equal to el * 2 * 7i * R.

 In configurations where the interconnections are subjected to significant thermomechanical stresses, insertions can be made in two parts of tracks symmetrical with respect to a point. One can thus choose for example to make several intersections between an insert 10 and a track 20 symmetrically around the center of the insert (Figures 5, 7 and 8), that is to say in a counterbalanced manner.

 It is possible to multiply the number of parts of tracks 20 to be intercepted by a tube

10, especially in order to use tracks of unit thickness e2 very thin, typically of the order of the unit wall thickness and an insert. Thus, by way of example, it is possible to insert a tube in a number equal to eight unit wall thicknesses as shown in FIG. 7. It is thus possible to produce very thin thicknesses e1, e2, typically equal to 0.1. μιη, the insertion section equal to 8 * el * e2 then being very small typically equal to 8 * 0.1 * 0.1 or 0.08μιη ² , and the insertion force required also.

 For example, we consider connections to be made in a step equal to ΙΟμιη with a tube 10 of radius R = 2.5 μιη.

According to the state of the art, such a tube 10 requires an insertion force Fl equal to 4 mN for insertion of its total circumference in an aluminum connection pad, of diameter equal to 7μιη. According to the invention, in order to reduce the insertion force considerably, a symmetrical cross-shaped aluminum track 20 is formed (FIG. 8), and an aluminum track 20 in the form of an elongated strip (FIG. 9). of unit thicknesses e2 equal to 1 μιη. It should be noted that in the example shown, the cross 20 is made with a resumption of contact on a via opening 22 of diameter equal to 2 μιη, in the form of a round stud 23 of diameter equal to 3μιη surmounted by the four branches. 20 of unit width e2.

 Preferably, in order to better distribute the mechanical stresses during the insertion, the track portions 20 to be intercepted by a tube 10 are in the form of at least three branches symmetrically distributed around a point of symmetry. It may therefore be for example three branches distributed at 120 ° to one another, four branches distributed at 90 ° to each other forming a symmetrical cross (Figure 8), eight branches grouped two by two with one group being distributed at 90 ° from another also forming a symmetrical cross (Figure 7). This gives an isostatic mechanical connection between inserts and tracks in all directions.

 For the same depth of insertion, the insertion forces to be respectively applied to the cross of Figure 8 and the band of Figure 9 are compared to the force Fl according to the state of the art, equal to:

 F2 = Fl * 4 * e / 2nR, equal to 0.25 * Fl;

 F3 = Fl * e / 2nR, equal to 0.06 * F1.

 Thus, a very substantial reduction in the insertion force required with a track 20 according to the invention is obtained compared to a total circumferential insertion in a ductile stud made of the same material according to the state of the art.

 The necessary insertion forces were computationally compared by varying the tube diameter of the inserts 10 and the unit thickness of the tracks 20 with the insertion surface Si.

 The results, between an insertion configuration according to the state of the art (FIG. 3), an insertion according to the invention with a single elongated track (FIG. 9) and an insertion according to the invention with a track 20 in the form of of a rectangular cross section tube (Figure 10) are given respectively in Tables 2 to 4 following.

It is specified that the studs according to the state of the art and the tracks 20 according to the invention are made of aluminum and that the insertion force calculated at constant pressure for insertion of the total circumference of a tube 10 into a stud. according to the state of the art is equal to 5mn. The force required for the two configurations according to the invention is then compared in order to have the same pressure.

 TABLE 2

 (ASSEMBLY TUBE 10 / CONNECTION PLATE 20 ACCORDING TO FIGURE 3)

 From these tables 2 to 4, it appears that the insert insertion force can be considerably reduced by the invention, from 84 to 97%, in this example.

 In conclusion, compared to thermocompression assembly methods according to the state of the art, such as those described in the aforementioned patent applications WO2009 / 1 15686 and EP2287904, the invention makes it possible to considerably reduce the constant insertion force. for a given ductile material.

 An interesting advantage of the invention is that it makes it possible to multiply the number of hybridized points with constant insertion force for a given ductile material.

According to the invention, it is possible to produce a stack of two assemblies each obtained according to the reduced insertion force assembly method which has just been described. The invention has wide applications in all microelectronic devices, intended to operate at high operating temperatures, requiring vertical interconnections (in English "front to front") at very small steps.

 A particularly interesting application of the invention is the production of stacked 3D three-dimensional structures or multi-spectral heterogeneous imagers.

 Many other applications may be provided for the invention, more particularly for the matrices:

 heterogeneous detection devices, of large size, with a large number of insertion connections (cooled IRCMOS, X detectors, etc.),

 - temperature sensitive, hybridized to "cold", that is to say at room temperature,

 - sensitive to mechanical stresses.

 For example, it is possible to produce such matrices by proposing ductile aluminum tracks in accordance with the numerical example given above: the same force of the order of 0.5 mN can be applied to make a connection according to the invention. in an aluminum track 20 as shown in FIG. 9 only for a connection according to the state of the art in an indium pad 20 as shown in FIG. 3. However, applying a weak force to effect insertion into a track in FIG. aluminum is advantageous because it is much easier and less expensive to make an aluminum track 20 as shown in FIG. 9 by a subtractive photolithography technique (etching) than an indium pad 20 as shown in FIG. additive photolithography (in English "lift-off") or electrolysis.

 Other variants and improvements may be provided without departing from the scope of the invention.

The invention is not limited to the examples which have just been described; it is possible in particular to combine with one another characteristics of the illustrated examples within non-illustrated variants. References cited

 [1]: .B. Goubault de Brugière, F.Marion, M. Fendler et al. "Micro tube insertion into indium, copper and other materials for 3D applications. Proc 60th Electronic Components and Technology Conf, Las Vegas, NV, 2010 pl757;

 [2]: B. Goubault de Brugiere, F. Marion, M. Fendler et al "A ΙΟμηι pitch interconnection technology using micro tube insertion into Al-Cu for 3D applications. Proc 61th Electronic Components and Technology Conf, Orlando, FL, 2011 pl400;

 [3]: D. Saint-Patrice, F. Marion, M. Fendler et al. "New Reflow Soldering and Tip in Buried Box (TB2) Techniques for Ultrafine Pitch Megapixel Imaging Array," Proc. 58th Electronic Components and Technology Inc., Orlando, FL, 2008p 46-53.

Claims

 1. Flip-chip assembly method of two electronic components (1, 2) to each other, said components each having a face called assembly face, according to which the two assembly faces are brought together one of the other in a direction X said assembly and applies a given force F to one and / or the other of the components, one and / or the other face (s) of assembly comprising:  - Connection inserts (10) of rigid material having an elongated longitudinal shape in the X assembly direction;  - Connection tracks (20) of material of lower hardness than inserts and elongate longitudinal shape transverse to the X assembly direction;  method according to which:  the inserts are aligned opposite the corresponding tracks so that the inserts and the tracks form two by two, after assembly, at least one substantially transverse intersection,  - Then apply the given force F to penetrate the inserts in the tracks until the assembly between the two components.  2. Assembly method according to claim 1; the connection inserts of a component having a unitary wall thickness (s) el and a unitary length of wall (s) L1 while the connection tracks of a component have a unit thickness e2, the unit thicknesses e1, e2 being the smallest dimensions considered transversally to the assembly direction X, the length L 1 being the largest dimension of wall (s) transverse to the assembly direction X, method according to which before the alignment of the inserts facing the corresponding tracks, the dimensions el, e2, and Ll are determined so that each insertion surface Si, Sii + Si ₂ + Si ₃ ... between an insert and a given track is substantially equal to a multiple N of the product of the unit thicknesses N = n * el * e2 where n is an integer, this multiple N being much smaller than the cross section SI of an insert considered transversely to the direction of assembly SI = Ll * el. 3. Assembly method according to claim 2, wherein the insertion surface Si is at least 40% less than the cross section S 1. 4. The assembly method according to claim 2 or 3, wherein the unit wall thickness and inserts is between 0.1 and 1 μιη, preferably between 0.1 and 0.5 μιη.  5. Assembly method according to one of claims 2 to 4, wherein the unit thickness of the tracks e2 is between 0.05 and 2 μιη, preferably between 0.1 and 1 μιη.  6. A method of assembly according to one of the preceding claims, wherein the constituent material of the tracks is a ductile metal material selected from aluminum Al, indium In, gold Au, tin Sn, lead , Pb, Bismuth, Sb antimony, AlCu aluminum-copper alloy, SnAgCu, SnAg, AgCu, SnCu alloys.  7. An assembly method according to claim 6, wherein the ductile metal material tracks are carried out by photolithography, additive or subtractive type, or by electrolysis of the metal or alloy.  8. The assembly method according to one of claims 1 to 5, wherein the constituent material of the tracks is a hard metal material selected from Cu copper, Titanium Ti, titanium nitride TiN, tungsten W, nitride Tungsten WN, Mo molybdenum, Chrome Cr, Nickel Ni, Platinum Pt.  9. The assembly method according to claim 8, wherein the hard metal material connection tracks are made by photolithography, additive or subtractive type.  10. The assembly method according to claim 8, wherein Au, Cu or Ni tracks are made by electrolytic growth.  11. The assembly method according to claim 8, wherein the tracks of hard metal material are made according to the same technique as the inserts.  12. The assembly method according to any one of claims, wherein the inserts are blind micro-tubes whose base is secured to one of the components. 13. An assembly method according to one of the preceding claims, wherein the connecting track portions (20) to be penetrated by the same tube (10) are in the form of at least three branches symmetrically distributed. around a point of symmetry, preferably at the number of four branches distributed at 90 from one another.  14. The assembly method according to one of the preceding claims, wherein the force F applied per insert is less than 5mN, preferably less than 0.8mN, typically equal to 0.5mN.  15. The assembly method according to one of the preceding claims, wherein the alignment and the application of the force F are carried out at ambient temperature.  16. A method of assembly according to one of the preceding claims, wherein the spacing between the two components corresponding to the height H is between a ratio p / 20 and a ratio equal to p / 2, p being the step of interconnection between two connection tracks of a component (1, 2).  17. The assembly method according to one of the preceding claims, wherein the interconnection step p between two connection tracks of a component (1, 2) is less than or equal to 50 μιη.  18. The assembly method according to one of the preceding claims, wherein the spacing between the two components corresponding to the height H is less than 20 μιη, typically equal to 1 μιη.  19. The assembly method according to one of the preceding claims, wherein one of the components (1) is a chip and the other component (2) is a printed circuit substrate.