Direct hybrid chip-to-plate bonding process

Abstract

Title: Direct Hybrid Chip-to-Wave Bonding Method The invention relates to a direct hybrid chip (10)-to-wafer (20) bonding method comprising the following steps: Providing at least one chip (10) comprising a first copper pad (11) and a first silicon oxide layer (12), Providing a wafer (20) comprising a second copper pad (21) and a second silicon oxide layer (22), Manipulating the chip (10) so as to position the face (100) of the chip (10) opposite the chip receiving area (210) on the wafer (20), aligning the first and second pads (11, 21), Depositing at least one drop of water (30) in the chip receiving area (210) and / or on the face (100) of the chip (10), Applying pressure to the chip (10) to form a water film from the water droplet (30) (31) between the face (100) and the receiving area (210) of the chip (10). Figure for the abbreviation: Fig. 2

Description

Title of the invention: Direct hybrid chip-to-plate bonding method technical field

[0001] The present invention relates to 3D integration, more particularly to direct chip-to-plate bonding and more particularly to direct hybrid chip-to-plate bonding. STATE OF THE ART

[0002] In the field of 3D integration of semiconductor components, methods are being developed to transfer and assemble one wafer or chip onto another wafer. Among these methods, direct bonding and, more specifically, hybrid bonding allow the assembly of two integrated circuits in a stacked configuration, and / or the establishment of a direct electrical connection between these two integrated circuits via the bonding interface. This makes it possible to increase the interconnection density of components manufactured by 3D integration.

[0003] In particular, hybrid copper / oxide bonding is an industrial bonding technique typically consisting of joining two surfaces composed of copper pads surrounded by silicon oxide. These surfaces, composed of two materials, are thus called hybrid. The objective of the bonding is to establish electrical contact between the copper pads of the two facing surfaces. The copper pads are typically a few microns in size. It is therefore necessary to perfectly align the two surfaces prior to direct bonding. Initially developed for "board-to-board" bonding, this hybrid bonding is now being developed to perform "chip-to-board" bonding using so-called "pick and place" machines (which can be translated as "pick and place," "pick and drop," or "pick and position," depending on the common meaning used by those skilled in the art).This allows for a combination of high placement accuracy and high throughput. The paper "Die to Wafer Direct Hybrid Bonding Demonstration with High Alignment Accuracy and Electrical Yields, A. Jouve et al., in 2019 International 3D Systems Integration Conference (3DIC), pp. 1-7" discloses a direct hybrid chip-to-wafer bonding process exhibiting high alignment accuracy with high yield. One drawback of this process is that D bonding defects can appear on some IOK(b) chips, as illustrated in [Fig. 1].

[0004] There is therefore a need to reduce, or even eliminate, bonding defects for direct bonding in general, and more particularly for hybrid direct chip-to-wafer bonding. One objective of the invention is to meet this need. SUMMARY

[0005] To achieve this objective, according to one embodiment, a direct chip-to-board bonding process is provided, comprising at least the following steps: - Provide at least one chip with a first flat face, preferably including initial alignment marks. - Provide at least one plate with a second flat face including a chip reception area and preferably second alignment marks, - Manipulate the chip so that the first face of the chip is positioned opposite the chip reception area on the second face of the board, aligning the chip with the board, preferably using the first and second alignment marks, - Bring the first side of the chip into contact with the chip reception area, so as to stick the chip to the second side of the board.

[0006] Advantageously, the method comprises, before contact and preferably before handling the chip, depositing at least one drop of water in the chip reception area on the second side of the board and / or on the first side of the chip. The water drop does not cover the alignment marks at this stage.

[0007] Advantageously, the method also includes, during the handling of the chip and after deposition of the water droplet, an application of pressure on the chip configured to form, from the deposited water droplet, a water film between the first face and the receiving area of ​​the chip.

[0008] According to another embodiment, a direct hybrid chip-to-plate bonding process is provided, comprising at least the following steps: - To provide at least one chip comprising at least one first copper pad and one first silicon oxide layer, said chip having a first flat face formed by exposed parts of the first copper pad and the first silicon oxide layer, - Provide at least one wafer comprising at least a second copper pad and a second layer of silicon oxide, said wafer having a second flat face comprising a chip-receiving area, said area being formed by exposed portions of the second copper pad and the second layer of silicon oxide, - Manipulate the chip so that the first face of the chip is positioned opposite the chip reception area on the second face of the board, aligning the first and second copper pads, - Bring the first side of the chip into contact with the chip reception area, so as to stick the chip to the second side of the board.

[0009] Advantageously, the method includes, before contact and preferably before handling the chip, a deposit of at least one drop of water in the receiving area of ​​the chip on the second face of the plate and / or on the first face of the chip.

[0010] Advantageously, the method also includes, during the handling of the chip and after deposition of the water droplet, an application of pressure on the chip configured to form, from the deposited water droplet, a water film between the first face and the receiving area of ​​the chip.

[0011] In the course of developing the present invention, it was found that the formation of the water film by depositing a drop of water and applying pressure to said drop of water advantageously made it possible to eliminate bonding defects between the chip and the plate.

[0012] In general, it is inadvisable and / or counterintuitive to form a liquid interface, particularly a water-based one, when performing precise chip alignment with the wafer. It is presumed that the presence of water between the chip and the wafer will alter the initial alignment typically achieved by precision industrial "pick and place" equipment.

[0013] In the context of the present invention, it was surprisingly observed that applying pressure to the water droplet, which allows the formation of a thin film of water, maintains the initial alignment achieved by the equipment. Thus, the formation of a thin film of water prevents misalignment of the chip with respect to the plate. This presence of water in a thin film reduces or eliminates bonding defects.

[0014] To avoid bonding defects, a person skilled in direct hybrid bonding could consider more complex solutions, for example, performing the handling and contact under vacuum or under a controlled helium-based atmosphere, or even bending the chips during bonding. These solutions impose more constraints on industrial equipment and are more expensive. They have not been adopted in the context of the present invention.

[0015] The process according to the invention thus advantageously makes it possible to reduce or eliminate bonding defects in direct bonding and in hybrid direct bonding, while controlling costs. BRIEF DESCRIPTION OF THE FIGURES

[0016] The aims, objects, features and advantages of the invention will become clearer from the detailed description of an embodiment thereof, which is illustrated by the following accompanying drawings in which:

[0017] [Fig. 1] Fig. 1 is an acoustic microscopy image illustrating, for the purpose of above the bonding defects between a plate and certain chips assembled by direct hybrid bonding, according to the prior art.

[0018] [Fig.2][Fig.3][Fig.4][Fig.5][Fig.6] Figures 2, 4 and 6 schematically illustrate in Cross-section of the direct hybrid bonding steps, according to an embodiment of the present invention. Figures 3 and 5 schematically illustrate, in top view, the direct hybrid bonding steps, according to an embodiment of the present invention.

[0019] [Fig.7] Fig.7 is an acoustic microscopy image illustrating, for the purpose of above the chips assembled on a plate by direct hybrid bonding, according to an embodiment of the present invention.

[0020] The drawings are given by way of example and are not limiting of the invention. They constitute schematic representations of the principle intended to facilitate understanding of the invention and are not necessarily to scale with practical applications. In particular, in the schematic diagrams, the thicknesses and / or dimensions of the different layers, patterns and reliefs are not representative of reality. DETAILED DESCRIPTION

[0021] Before proceeding with a detailed review of embodiments of the invention, optional features that may be used in combination or alternatively are listed below:

[0022] According to one example, the first flat face and the receiving area of ​​the chip are based on the same material, for example Si, Ge, AsGa, InP, GaN, SiC, A12O3, diamond, SiO2, Si3N4, SiCN, A12O3, TiN, TaN, WN, Cu, Ti, Ni, Au, W.

[0023] According to one example, the chip comprises at least a first pad based on a first material and a first layer based on a second material, and the first planar face is formed by exposed parts of the first pad and the first layer.

[0024] According to one example, the plate comprises at least a second pad based on the first material and a second layer based on the second material, and the chip receiving area is formed by exposed parts of the second pad and the second layer,

[0025] According to one example, the process is a direct hybrid chip-to-plate bonding process.

[0026] According to one example, the first material is chosen from copper, titanium, nickel, gold, tungsten.

[0027] According to one example, the second material is chosen from SiO2, Si3N4, SICN, A12O3, TiN, TaN, WN.

[0028] According to one example, the water film spreads over the second side, remaining within the chip reception area.

[0029] According to one example, the second alignment marks are located in the chip's receiving area. Alignment is typically achieved by inserting a microscope objective between the second alignment marks in the chip's receiving area and the first alignment marks on the first flat face of the chip. After alignment, the microscope objective is removed, and the chip and the chip's receiving area are brought into contact via the water film. According to one example, the deposition of at least one water droplet is configured so that the at least one water droplet does not completely cover said first and / or second alignment marks.

[0030] In one example, the water film extends over the second surface, outside the chip receiving area. It is not necessary for the water film to be confined to the chip receiving area. In particular, it is not necessary to provide hydrophobic areas around the chip receiving area. This avoids the need to specifically prepare the second surface of the board and / or the first surface of the chip for bonding.

[0031] According to one example, the second face has a so-called peripheral zone surrounding the chip receiving area, both the peripheral zone and the chip receiving area being hydrophilic. This allows the water film to extend freely beyond the chip receiving area. This enables excess water to be drained during the formation of the water film between the first face of the chip and the chip receiving area on the wafer.

[0032] According to one example, the water film has a thickness less than or equal to 10 pm, preferably less than 1 pm and more preferably less than 100 nm or less than 50 nm. This prevents the chip from being carried by the water out of its alignment position with the chip reception area.

[0033] According to one example, the deposited water droplet initially occupies, before the application of pressure, a surface area S30 on the chip's receiving area that is much smaller, i.e., at least half the size, of the surface area S20 corresponding to the chip's receiving area. The water droplet can typically be centered on the chip's receiving area. It is not necessary to cover the entire surface of the chip's receiving area when depositing the water droplet. According to one example, the water droplet is deposited on the first flat face of the chip and initially occupies, before the application of pressure, a surface area S30 on the first flat face that is much smaller, i.e., at least half the size, of the surface area of ​​said first flat face of the chip.

[0034] According to one example, applying pressure allows the water droplet to spread over and outside the chip's receiving area, forming a water film between the first surface and the chip's receiving area. This pressure is applied voluntarily silently, typically by "pick and place" equipment. This pressure is not solely due to the weight of the chip or the surface tension forces of the water droplet.

[0035] According to one example, the second side and / or the first side include alignment marks. This allows the chip to be precisely aligned with the chip receiving area, typically when using industrial "pick and place" type equipment.

[0036] In one example, the first face includes first alignment marks. In another example, the second face includes second alignment marks in the chip's receiving area. In another example, the first and / or second alignment marks are configured to align the chip with the chip's receiving area, by means of a microscope or semi-transparent optics interposed between the chip and the chip's receiving area.

[0037] According to one example, the deposited water droplet has a volume chosen such that, after natural spreading of the water droplet on the receiving area of ​​the chip, the water droplet does not cover the second alignment marks. According to another example, the deposited water droplet has a volume chosen such that, after natural spreading of the water droplet on the first face of the chip, the water droplet does not cover the first alignment marks.

[0038] According to one example, the deposited water droplet has a volume between 1 pL and 100 pL, preferably between 8 nL and 10 pL.

[0039] According to one example, the deposited water droplet has a volume less than or equal to 100 pL. According to another example, the deposited water droplet has a volume greater than or equal to 8 nL. According to another example, the deposited water droplet has a volume less than or equal to 10 pL.

[0040] According to one example, the applied pressure is maintained for a duration of between 100 ms and 10 s, preferably for 1 s. This allows the formation of a water film, reducing the risk of this water film carrying the chip out of its initial alignment position. This also makes it possible to achieve a placement rate compatible with industrial requirements.

[0041] As an example, the relative humidity of the atmosphere is controlled to be greater than or equal to 80% during the handling of the chip. This limits the evaporation of the deposited water droplet before pressure is applied to it. The evolution of the water droplet's volume is thus better controlled. The volume of the water droplet can therefore be reduced. This makes it possible to apply the process to very small chips, for example, with a first surface area on the order of 100 x 100 pm².

[0042] According to one example, the process further comprises a drying step after ap Pressure is applied, configured to remove the water film. Contact, i.e., direct or hybrid direct bonding, is thus achieved. This drying can be carried out by simple storage under ambient atmosphere. Alternatively, it can be carried out under a dry atmosphere, in a desiccator for example, or under an atmosphere of inert gas, for example, nitrogen, argon, or helium. A relative humidity of less than 1% can thus be obtained. It is also possible to perform this drying by placing the plate under vacuum, for example, at 20 mbar pressure and 21°C. It is preferable not to go below the saturated vapor pressure of water. It is also possible to increase the temperature, for example, to 75°C. It is preferable not to exceed the boiling point of water.

[0043] According to one example, the process further includes an annealing step intended to improve or strengthen the bonding, after contacting.

[0044] Within the scope of the present invention, a method for transferring and bonding one or more chips onto a single board is described. This method is preferably intended for industrial implementation, for transferring and bonding a plurality of chips onto each board of a plurality of boards. It falls within the field of direct hybrid bonding. "Hybrid" means that the bonding surfaces are composed of at least two materials. "Direct" means that the bonding interface, after final bonding, corresponds directly to the two bonding surfaces, without any bonding layer, such as a polymer adhesive, interposed between the two bonding surfaces.

[0045] In this application, a "chip" typically means an integrated circuit comprising microelectronic or optoelectronic components, or microelectromechanical systems (MEMS). A "wafer" typically means a substrate comprising or bearing a plurality of chips. By extension, a chip can also refer to a piece of wafer without components. Alignment can be achieved using alignment marks or, in the latter case, simply with the accuracy of the machine's movement. Alignment is typically achieved with an accuracy of 100 µm or better.

[0046] The chip receiving area is also referred to as the sticking area in the following.

[0047] It is specified that, within the framework of the present invention, the terms “on”, "Overcomes", "covers", "underlying", "opposite" and their equivalents do not necessarily mean "in contact with". Thus, for example, the depositing, transfer, gluing, assembly or application of a first layer on a second layer does not necessarily mean that the two layers are directly in contact with each other, but means that the first layer at least partially covers the second layer by being either directly in contact with it, or by being separated from it by at least one other layer or at least one other element.

[0048] A layer may also be composed of several sub-layers of the same material or of different materials.

[0049] A substrate, film, or layer "based" on a material A is understood to mean a substrate, film, or layer comprising only that material A or that material A and possibly other materials, for example dopant elements or alloying elements.

[0050] A water droplet according to the invention is preferably composed of pure or deionized water (DI water). In particular, it does not contain particles larger than 20 nm. It preferably has a resistivity greater than 1 MOhm.

[0051] The hybrid surface of the chip or plate can be made of different materials: Copper, titanium, nickel, gold, tungsten for example (these metals may or may not be oxidized on the surface) SiO2, Si3N4, SICN, A12O3, TiN, TaN, WN for example.

[0052] The surface of the chip may not be hybrid and may be composed of a single material (Si, Ge, AsGa, InP, GaN, SiC, A12O3, diamond, SiO2, Si3N4, SiCN, A12O3, TiN, TaN, WN, Copper, titanium, nickel, gold, tungsten for example).

[0053] The surface of the plate may not be hybrid and may be composed of a single material (Si, Ge, AsGa, InP, GaN, SiC, A12O3, diamond, SiO2, Si3N4, SiCN, A12O3, TiN, TaN, WN, Copper, titanium, nickel, gold, tungsten for example).

[0054] Several embodiments of the invention implementing successive steps of the manufacturing process are described below. Unless explicitly stated, the adjective "successive" does not necessarily imply, although this is generally preferred, that the steps follow each other immediately; intermediate steps may separate them.

[0055] Furthermore, the term "step" refers to the execution of a part of the process, and can designate a set of sub-steps.

[0056] Furthermore, the term "step" does not necessarily mean that the actions carried out during a step are simultaneous or immediately successive. Certain actions of a first step may, in particular, be followed by actions related to a different step, and other actions from the first step may be repeated later. Thus, the term "step" does not necessarily imply unitary actions that are inseparable in time and in the sequence of phases of the process.

[0057] A preferably orthonormal coordinate system, comprising the x, y, z axes, is shown in the accompanying figures. When only one coordinate system is shown on the same sheet of figures, this coordinate system applies to all the figures on that sheet.

[0058] In this patent application, the term "thickness" will preferably be used for a layer or film. The thickness is measured in a direction normal to the plane. of principal extension of the layer or film. Thus, a layer or film typically has a thickness along z. The relative terms "on", "overlies", "under", "underlying" refer to positions taken along the z direction.

[0059] An element located "in line with" or "directly above" another element means that these two elements are both located on the same line perpendicular to a plane in which extends mainly a lower or upper face of a substrate, that is to say on the same line oriented vertically on the cross-sectional figures.

[0060] Figure 1 shows an acoustic microscopy image obtained after direct hybrid bonding of chips 10 onto a plate 20, according to a standard direct hybrid bonding method. Of the nine chips 10 bonded to this area of ​​plate 20, six 10 KB chips exhibit normal bonding, without apparent defects, and three 10 KB chips exhibit bonding defects. The method according to the invention aims to eliminate these bonding defects.

[0061] One principle of the method according to the invention is to place a drop of water between the chip and the board and to apply pressure to this drop of water so as to form a thin film of water between the chip and the board. Surprisingly, the alignment of the chip with respect to the board is maintained in the presence of this thin film of water. When the water film dries, the chip comes into direct contact with the board and the bonding is flawless.

[0062] Figures 2 to 6 illustrate certain steps of the direct hybrid bonding process according to the invention.

[0063] As illustrated in [Fig. 2], a chip 10 comprising copper pads 11 at least partially embedded within a silicon oxide layer 12 is brought by a pick-and-place device 1 opposite a plate 20. The face 100 of the chip 10 is flat and consists in particular of the exposed portions of the copper pads 11 and the exposed portions of the silicon oxide layer 12. The face 100 is therefore a composite or hybrid face having a surface formed partly by the protruding copper pads 11 and the exposed portions of the layer 12.

[0064] The face 200 of the plate 20 opposite the chip 10 is substantially identical to the face 100, at least in the area 210 where the chip 10 is received. In this area 210, the face 200 is flat and consists in particular of the exposed portions of the copper pads 21 and the exposed portions of the silicon oxide layer 22. In the area 210, the face 200 is therefore, like the face 100, a composite or hybrid face having a surface formed partly by the protruding copper pads 21 and the exposed portions of the layer 22.

[0065] The "Pick and Place" device 1 allows the chip 10 to be manipulated so as to align the pins 11 of the chip 10 with the pins 21 of the plate 20. This Alignment can typically be achieved using first alignment marks positioned on the chip 10 (not shown) and / or second alignment marks 40 positioned on the plate 20, in the area 210, as shown in the top view in [Fig. 3]. Alignment is typically performed using a microscope placed between the chip 10 and the plate 20, allowing simultaneous observation of the first and second alignment marks. After alignment, the microscope is removed, and the chip 10 is lowered until it contacts the receiving area 210, crushing the water droplet to form a film of water interposed between the chip 10 and the area 210. Pick-and-place devices are generally equipped with such a microscope. This is the case, for example, with the NEO HB device marketed by SET. In [Fig.3], the first and second alignment marks are in the form of crosses which overlap after alignment.Alternatively, the first and second alignment marks can be complementary. For example, the second alignment marks 40 consist of a cross (as in [Fig.3]) and the first alignment marks consist of four disjoint squares which, after alignment, are positioned to complement the cross and form a square.

[0066] Before or during the manipulation of the chip 10 by the "Pick and Place" device 1, a drop of water 30 is deposited on at least one of the faces 100, 200 opposite each other. In one possibility, the drop of water 30 is deposited only in the area 210, for example, in the center of the area 210. In another possibility, the drop of water 30 is deposited or formed only on the face 100 of the chip 10, for example, in the center of the face 100. In yet another possibility, a first drop of water 30 is deposited or formed on the face 100, and a second drop of water 30 is deposited or formed on the area 210. The principle of this process step is to interpose at least one drop of water between the face 100 and the area 210 intended to come into contact with each other.

[0067] The volume of the water droplet 30 is preferably quite small. Generally, this volume can be between 1pL and 100pL, preferably between 8nL and 100pL. These values ​​typically correspond to a droplet volume obtained in a single application. The droplet application step can be repeated several times to increase the total volume applied.

[0068] If the drop is placed before alignment, the volume of the water drop 30 is to be adapted according to the surface S2io of the bonding area, so that the surface S2io of the bonding area is partially covered at least on 1%, or even 10% or 25% of its surface, or more particularly at least 50% of its surface by the drop taking into account its natural spreading but must not cover the alignment marks in order to allow alignment.

[0069] If the drop is arranged after alignment, there is less constraint on its volume.

[0070] In all cases, after the water droplet is crushed, the water film 31 thus formed can cover the entire S2io surface of the bonding area (as illustrated in [Fig.4] and [Fig.5]) or only a part (10% or even 25% or 50%) of this surface.

[0071] The smaller the volume of the water droplet 30, the faster it will evaporate. Handling the chip 10 and precisely aligning it with the area 210 can take some time, for example, a few seconds or tens of seconds. The choice of the volume of the water droplet 30 can take into account this handling time during which the droplet 30 begins to evaporate. Preferably, the volume of the water droplet 30 is large enough to compensate for water loss through evaporation and to cover the entire surface S2io or at least 1%, or more particularly at least 10, 25, or 50%, of the bonding area under the pressure exerted by the chip 10 during the next step of the process.

[0072] A compromise in choosing the volume of the water droplet 30 can therefore be found depending on the size of the chip 10, and / or depending on the distance between the initial position of the droplet and the alignment marks, and / or depending on the distance between two adjacent chip reception zones 210, and / or depending on the handling and alignment time. For example, for a 3 x 3 mm² chip 10 and alignment marks 40 located 100 pm inside the corners or edges of zone 210, a reasonable volume for the water droplet 30 is approximately 520 nL. Alternatively, the volume of the droplet to be dispensed is chosen based on the surface area S30 occupied by the droplet 30 on zone 210 during deposition and after its natural spreading. The surface S30 occupied by the drop 30 after its natural spreading without specific support is preferably at least twice less than the surface S2i0 of the zone 210, or even at least ten times less.

[0073] It is possible to slow down the evaporation rate of the water droplet 30 by increasing the relative humidity of the atmosphere in the pick-and-place machine. This reduces the required droplet volume. For example, in an atmosphere with a relative humidity RH of around 90%, the droplet 30 can have a volume of only a few tens of picoliters (pL). The size of the droplet 30 is thus reduced. This makes it possible to implement the process for very small chips, for example, on the order of 100 x 100 pm².

[0074] When the drop is deposited or formed and the alignment is carried out, the "Pick and Place" device 1 is lowered towards the area 210. In the case illustrated in [Fig.2], the face 100 therefore first comes into contact with the drop 30, which has the effect of spreading the drop 30 over the area 210.

[0075] The "Pick and Place" device 1 then applies pressure to the drop 30 so that the drop 30 spreads over the entire area 210, or at least over 1% or more particularly on 50% of zone 210, to form a water film 31 between face 100 and zone 210, as illustrated in Figures 4 and 5. When this pressure is applied, some of the water is typically ejected outside the bonding zone 210. The water film 31 extends beyond zone 210. This does not hinder the invention. It is not necessary to confine the water or the water film 31 within zone 210. Zone 220 at the periphery of zone 210 can be hydrophilic, like zone 210. This avoids the need to specifically prepare the peripheral zone 220. In particular, it is not necessary to provide hydrophobic areas in the peripheral zone 220, nor even a particular topography such as a step that could confine the droplet within zone 210. The implementation of the method is facilitated.

[0076] The pressure exerted by the "Pick and Place" device 1 must be sufficient to obtain a water film 31 of low thickness e3b. The thickness e3i of the residual water film 31 is typically less than 5 pm, preferably less than 1 pm, and more precisely less than 100 nm or less than 50 nm. To obtain such a thickness e3i, a force between 0.1 N and 1 kN, preferably between 1 N and 300 N, is applied to the "Pick and Place" device 1. A support pressure of a few tens to a few thousand pascals (Pa) is thus exerted on the water film 31.

[0077] According to an example, the pressure exerted on the water film 31 is on the order of 104 to 3.106 Pa for a square chip of 10mm*10mm subjected to a force of 1 to 300 N. For a chip of 1 mm*1 mm, the pressure varies from 106 to 3.108 Pa.

[0078] The pressure exerted on the water film 31 is maintained for a duration of between 100 ms and 10 s, preferably between 500 ms and 5 s. In one example, the pressure exerted on the water film 31 is maintained for a duration of approximately 1 s. This duration is perfectly compatible with the industrial implementation of the process. The "Pick and Place" device 1 is then removed, typically to handle another chip and perform another bonding operation according to the process. Advantageously, during the removal of the "Pick and Place" device 1, the alignment between the chip 10 and the area 210 is maintained in the presence of the thin film 31.

[0079] As illustrated in [Fig. 6], the water film 31 disappears upon drying, and the chip 10 is in direct contact with the area 210. The direct hybrid bonding of the chip 10 to the area 210 is thus at least partially achieved. The drying of the water film 31 can be achieved by simple storage in an atmosphere with a relative humidity (RH) below 100%, and preferably below 50%. It is also possible to carry out the drying in a very dry atmosphere (RH < 1%), and / or in an atmosphere of neutral gas such as nitrogen, argon, or helium, for example. It is also possible to obtain dry air with a desiccant. Alternatively, this drying can be performed by placing the board under vacuum in an atmosphere with a pressure lower than ambient pressure, for example, at 20 mbar pressure and an ambient temperature of 21°C. It is preferable not to go below the saturated vapor pressure of water. It is also possible to increase the drying temperature, for example up to 75°C. It is preferable not to exceed the boiling point of water.

[0080] After drying, conventional annealing for direct hybrid bonding can be carried out, for example at 300°C for 2 hours.

[0081] According to a particular embodiment of the process, all steps are carried out in a cleanroom at 21 °C and 45% relative humidity. The hybrid surfaces of face 100 and area 210 are composed of copper pads surrounded by silicon oxide, typically according to an embodiment described in the document "Die to Wafer Direct Hybrid Bonding Demonstration with High Alignment Accuracy and Electrical Yields, A. Jouve et al., in 2019 International 3D Systems Integration Conference (3DIC), pp. 1-7". The size of each chip 10 is 3 x 3 mm². A NEO HB pick-and-place machine from SET is used. It is modified to include a NanoJet water droplet dispenser (with an NJ-K-4010 piezoelectric valve) from Microdrop. Before starting the alignment of chip 10 and plate 20, a drop of water 30 of 520 nL is deposited on plate 20 in the center of the receiving area 210 of chip 10.The water droplet 30 does not cover the alignment marks 40, which are located beyond the corners of the bonding zone 210. Alignment is performed in less than 5 seconds, and the chip 10 is brought into contact with the water droplet 30. A force of 20 N is applied to the chip 10, corresponding to a pressure of approximately 2.2 GPa. The pressure is maintained for 1 second. The process is repeated to bond a plurality of chips 10 to the plate 20. The plate 20 is then removed from the pick-and-place machine and stored in the cleanroom atmosphere for 24 hours. After storage, it is annealed at 300°C for 2 hours in an oven.

[0082] Figure 7 is an image obtained using a scanning acoustic microscopy (SAM) microscope on the plate 20 bearing the chips 10, in order to characterize the quality of the direct hybrid bonding. It is clear that all the chips 10, 10OK are perfectly bonded without interface defects by the process according to the invention.

[0083] In view of the preceding description, it is clear that the proposed method offers a particularly efficient solution for direct hybrid chip-to-plate bonding.

[0084] The invention is not limited to the embodiments previously described.

Claims

1.

2. Demands A method for direct bonding of a chip (10) to a board (20) comprising at least the following steps: - Provide at least one chip (10) having a first flat face (100), - Provide at least one plate (20) having a second flat face (200) including a chip (10) reception area (210), - Manipulate the chip (10) so as to position the first face (100) of the chip (10) opposite the receiving area (210) of the chip (10) on the second face (200) of the plate (20), aligning the chip with the plate, - Bring the first face (100) of the chip (10) into contact with the receiving area (210) of the chip (10), so as to stick the chip (10) onto the second face (200) of the plate (20), - before contact, a deposit of at least one drop of water (30) in the receiving area (210) of the chip (10) on the second face (200) of the plate (20) and / or on the first face (100) of the chip (10), and - during the handling of the chip (10) and after the deposition of the water droplet (30), an application of pressure on the chip (10) configured to form, from the deposited water droplet (30), a water film (31) between the first face (100) and the receiving area (210) of the chip (10), the process being characterized in that the second face (200) has a zone (220) called peripheral around the zone (210) of receiving the chip (10), said peripheral zone (220) and the zone (210) of receiving the chip (10) both being hydrophilic. A method according to the preceding claim, wherein: • the chip (10) comprises at least a first pad (11) based on a first material and a first layer (12) based on a second material, and the first flat face (100) is formed by exposed parts of the first pad (11) and the first layer (12), • the plate (20) includes at least one second stud (21) at base of the first material and a second layer (22) based on the second material, and the receiving area (210) of the chip (10) is formed by exposed parts of the second pad (21) and the second layer (22), the process being a direct hybrid chip (10) to plate (20) bonding process.

3. A method according to the preceding claim wherein the first material is selected from copper, titanium, nickel, gold, tungsten, and wherein the second material is selected from SiO2, Si3N4, SiCN, Al2O3, TiN, TaN, WN.

4. A method according to any one of the preceding claims wherein the first planar face (100) comprises first alignment marks and the second planar face (200) comprises second alignment marks (40), and wherein the alignment of the chip with respect to the plate is carried out via said first and second alignment marks.

5. A method according to the preceding claim wherein the second alignment marks (40) are arranged in the receiving area (210) of the chip (10), and wherein the deposition of at least one water droplet (30) is configured so that the at least one water droplet (30) does not totally cover said second alignment marks (40).

6. A method according to any one of the preceding claims in which the water film (31) extends over the second face (200), outside the area (210) of receiving the chip (10).

7. A method according to any one of claims 1 to 4 in which the water film (31) extends over the second face (200), remaining within the area (210) of receiving the chip (10).

8. A method according to any one of the preceding claims wherein the water film (31) has a thickness e3i less than 5 pm, preferably less than Ipm and more preferably less than 100 nm or less than 50 nm.

9. A method according to any one of the preceding claims, wherein the deposited water droplet (30) initially occupies, before application of pressure, a surface area S30 on the chip (10) receiving area (210) that is much smaller, i.e., at least twice smaller, than the surface S2io corresponding to the area (210) of receiving the chip (10), and the application of pressure allows the said water droplet (30) to be spread on and outside the area (210) of receiving the chip (10), to form the water film (31) between the first face (100) and the area (210) of receiving the chip (10).

10. A method according to any one of the preceding claims wherein the deposited water droplet (30) has a volume less than or equal to 100 pL.

11. A method according to any one of the preceding claims wherein the deposited water droplet (30) has a volume between 8 nL and 10 pL.

12. A method according to any one of the preceding claims wherein the applied pressure is maintained for a duration of between 100 ms and 10 s.

13. A method according to any one of the preceding claims wherein the relative humidity of the atmosphere is controlled so as to be greater than or equal to 80% during the handling of the chip (10).

14. A method according to any one of the preceding claims further comprising a drying step after application of pressure, configured to remove the water film (31).

Images