Method of manufacturing structure comprising at least two chips on substrate
By forming a dummy donor substrate on the support substrate and adjusting the block density, the polishing inhomogeneity problem during layer transfer is solved, and a high-density and uniform chip distribution is achieved, which is suitable for microelectronics, optical or optoelectronics fields.
Patent Information
- Application Number
- CN202380032847.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2022-04-04
- Filing Date
- 2023-04-03
- Publication Date
- 2024-11-15
AI Technical Summary
In the microelectronics, optical or optoelectronics, in the design of multi-layer structures, the polishing inhomogeneity problem during the layer transfer process due to the difference in size of the donor substrate and the receiving substrate, especially when the block density is low, There is a large distance between the sheets on the donor substrate and the sheets on the receiving substrate, resulting in different polishing heights and excessive material consumption, affecting the performance of electronic, optical or optoelectronic components.
Chemical mechanical polishing and material removal are performed by placing the block of the donor substrate on the support substrate and bonding it to the receiving substrate, so that the block at least partially covers different areas of interest of the receiving substrate. , adjust the size of the block part to form a chip, increase the block density, reduce the block spacing, and optimize the polishing effect.
It effectively reduces polishing inhomogeneity, improves the block density and coverage, avoids excessive material consumption caused by polishing, ensures the uniformity and density of the chip, and is suitable for forming high-quality electronic, optical or optoelectronic components.
Smart Images

Figure CN118974913A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a method of manufacturing a structure comprising at least two chips located on a substrate. Background Art
[0002] In the field of microelectronics, optics or optoelectronics, the design of multilayer structures sometimes requires the transfer of tiles in the form of parts of a layer of a donor substrate to a receiving substrate.
[0003] This type of processing is often referred to as tiling and involves the partial transfer of a layer removed from a donor substrate to form one or more tiles that are arranged in a pattern or at predetermined locations on a receiving substrate.
[0004] Such tiling may be necessary due to the size difference between the donor substrate and the receiving substrate. In particular, due to the size difference, it is not possible to transfer the entire layer of the donor substrate to the receiving substrate.
[0005] A known layer transfer process is Smart Cut TM A process in which a layer is transferred from a donor substrate to a receiving substrate by implanting atomic species in a donor substrate to form a weakened zone defining the layer to be transferred, bonding the donor substrate to a receiving substrate, and separating the donor substrate along the weakened zone. However, this process assumes that the donor and receiving substrates are of the same size.
[0006] However, while silicon substrates are available in relatively large sizes (typically 300 mm in diameter), other materials of interest currently exist only in the form of bulk substrates of smaller sizes, such as 10 or 15 cm in diameter. In addition, these materials of interest are sometimes particularly expensive, so it is desirable to minimize potential waste created during transfer. This is particularly true for III-V semiconductor materials, including nitrides (e.g., indium nitride (InN), gallium nitride (GaN), and aluminum nitride (AlN) for binary compounds), arsenides (e.g., indium arsenide (InAs), gallium arsenide (GaAs), and aluminum arsenide (AlAs) for binary compounds), and phosphides (e.g., indium phosphide (InP), gallium phosphide (GaP), and aluminum phosphide (AlP) for binary compounds).
[0007] Based on Smart Cut TM The solution of the process does not transfer the entire donor substrate layer, but forms an intermediate substrate (called a pseudo-donor substrate), takes one or more pieces from at least one donor substrate, and transfers the pieces to a support substrate with the same diameter as the receiving substrate. The process then includes forming a weakened zone in each piece by implanting atomic substances, bonding the pseudo-donor substrate to the receiving substrate via the piece, and separating each piece along the weakened zone to transfer part of each piece to the receiving substrate.
[0008] During this process, after forming the dummy donor substrate and / or after transferring a portion of the tile to a receiving substrate, it is usually necessary to polish the tile. This polishing is usually a chemical mechanical polishing (CMP) aimed at planarizing the free surface of the tile or a portion of the tile.
[0009] A particular difficulty arises when the density of the tiles is not too high, i.e. when the coverage of the tiles on the receiving substrate (corresponding to the ratio of the total surface area of the tiles to the total surface area of the receiving substrate) is typically less than 50% or, in practice, less than 30%. In this case, there are relatively large distances between the tiles on the pseudo-donor substrate and the tiles on the receiving substrate, typically of the order of a few millimeters.
[0010] However, the polishing cloth applied to the tiles can deform at the grooves formed by the spaces between the tiles. This deformation leads to differences in polishing height between the tiles and excessive material consumption at the edges of the tiles compared to the center of the tiles.
[0011] These non-uniformities caused by polishing are detrimental to the subsequent use of the wafer to form electronic, optical or optoelectronic components. Summary of the invention
[0012] The object of the present invention is to devise a method for manufacturing a structure comprising at least two chips on a supporting substrate, making it possible to overcome the above-mentioned polishing drawbacks.
[0013] To this end, the present invention proposes a method for manufacturing a structure comprising at least two chips located on a receiving substrate, comprising:
[0014] - forming a dummy donor substrate by arranging at least one piece of at least one donor substrate on a support substrate;
[0015] - bonding the dummy donor substrate to the receiving substrate via the tiles such that each tile at least partially covers at least two different areas of interest of the receiving substrate;
[0016] - transferring portions of the tile to the receiving substrate;
[0017] - at least one chemical mechanical polishing step performed on said piece of said dummy donor substrate and / or on the part of the piece transferred to said receiving substrate;
[0018] - after the at least one chemical mechanical polishing step, removing material from the tile parts in order to divide the respective tile parts into at least two chips, each arranged on a corresponding region of interest.
[0019] Adjusting the size of the tile portions after transferring them to a receiving substrate to form a chip makes it possible to increase the density of tiles on the pseudo-donor substrate and reduce the distances between tiles and between transferred tile portions. Thus, polishing defects associated with these excessive distances between tiles or between transferred tile portions are minimized.
[0020] After the etching step, the coverage of the chips on the receiving substrate, ie the ratio of the total surface area of the chips to the total surface area of the receiving substrate, is advantageously less than 50%, preferably less than 30%.
[0021] Particularly advantageously, the coverage of the patches on the pseudo-donor substrate, ie the ratio of the total surface area of the patches to the total surface area of the support substrate, is greater than 50%, preferably greater than 75%.
[0022] Each piece may have a side length greater than or equal to 5 mm, preferably greater than or equal to 8 mm, and preferably greater than or equal to 10 mm.
[0023] Furthermore, each chip can have a side length of less than or equal to 5 mm, preferably less than or equal to 3 mm, and preferably less than or equal to 1 mm.
[0024] Particularly advantageously, the distance between two adjacent segments is less than or equal to 3 mm, preferably less than or equal to 1 mm.
[0025] On the other hand, the distance between two adjacent chips may be greater than or equal to 3 mm, preferably greater than or equal to 5 mm.
[0026] In certain embodiments, each tile at least partially covers two different regions of interest of a receiving substrate, and a chip is formed from the tile on each corresponding region of interest.
[0027] In other embodiments, each tile covers at least corresponding corners of four different regions of interest arranged in a square or rectangular shape of the receiving substrate, and a chip is formed by the tile at each corner of the corresponding region of interest.
[0028] In other embodiments, each tile is in the form of a strip that at least partially covers at least three aligned regions of interest of the receiving substrate, and a chip is formed from the tile on each corresponding region of interest.
[0029] Particularly advantageously, the support substrate and the receiving substrate have substantially the same diameter, which is greater than the diameter of the donor substrate.
[0030] The pseudo-donor substrate may be bonded to the receiving substrate by molecular adhesion.
[0031] In a preferred embodiment, the transfer of each piece portion includes, in sequence: forming a weakened zone by implanting atomic substances in each piece of the pseudo-donor substrate to define the piece portion to be transferred, joining the pseudo-donor substrate to the receiving substrate via the piece, and separating the pieces along the weakened zone.
[0032] In other embodiments, the transferring of the individual tile portions includes thinning the individual tiles of the dummy donor substrate by grinding and / or etching.
[0033] Each piece may have a thickness between 20 μm and 1000 μm, preferably between 100 μm and 700 μm, and the transferred portion of each piece may have a thickness between 30 nm and 1.5 μm.
[0034] Each tile may include:
[0035] a semiconductor material, such as a III-V material, in particular indium nitride (InN), gallium nitride (GaN), aluminum nitride (AlN), indium arsenide (InAs), gallium arsenide (GaAs), aluminum arsenide (AlAs), indium phosphide (InP), gallium phosphide (GaP) or aluminum phosphide (AlP), or a IV or IV-IV material, in particular germanium or silicon carbide (SiC),
[0036] - Piezoelectric materials, such as lithium tantalate (LiTaO3), lithium niobate (LiNbO3), potassium sodium niobate (K x Na 1-x NbO3 or KNN), barium titanate (BaTiO3), quartz, lead zirconate titanate (PZT), lead magnesium niobate and lead titanate (PMN-PT), zinc oxide (ZnO), aluminum nitride (AlN) or aluminum scandium nitride (AlScN), and / or
[0037] - Electrically insulating materials such as diamond, strontium titanate, yttria-stabilized zirconia or sapphire.
[0038] In some embodiments, the method comprises a step of forming at least one epitaxial layer on the respective tile portion by epitaxy, the step of epitaxy being followed by a material removal to provide the respective tile portion as a chip.
[0039] In some embodiments, the material removal includes:
[0040] - forming a mask comprising a protective film and at least one opening defining a space to be formed between two chips, the protective film partially covering the tile portions in a pattern defining the chips in each tile portion;
[0041] - etching the respective tile portions through the respective openings of the mask to form the spaces for separating the chips from each other.
[0042] The etching is performed using an etching solution, and the mask is formed of a photoresist.
[0043] Alternatively, the etching may be performed using an ion beam, and the mask may be formed of metal.
[0044] In other embodiments, material removal is performed selectively by a focused ion beam that is controlled to scan only the area of the tile portion to be removed. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] Other features and advantages of the present invention will become apparent from the following detailed description taken in conjunction with the accompanying drawings, in which:
[0046] - Figure 1 is a top view of the receiving substrate with the region of interest depicted,
[0047] - Figure 2 The steps of removing the piece from the donor substrate are illustrated,
[0048] - Figure 3 The step of placing the piece on a supporting substrate to form a pseudo donor substrate is illustrated,
[0049] - Figure 4 illustrating the step of forming a weakened area in the block,
[0050] - Figure 5 illustrates the step of bonding a dummy donor substrate to a receiving substrate via a tile,
[0051] - Figure 6 illustrates the step of transferring part of the tile to a second support substrate after the tile is separated along the weakened area,
[0052] - Figure 7 The step of forming a mask on a portion of a tile to define a chip to be formed is illustrated.
[0053] - Figure 8 The steps of etching the unmasked area of the block to form a chip are illustrated.
[0054] - Fig. 9 The step of removing the protective film after etching is illustrated.
[0055] - Fig. 10A An embodiment is illustrated in which each tile partially covers two adjacent regions of interest, and Fig. 10B illustrates an example of the distribution of chips formed from the tiles,
[0056] - Fig.11A An embodiment is illustrated in which each patch partially covers a plurality of aligned regions of interest, and Fig. 11B illustrates an example of the distribution of chips formed from the tiles,
[0057] - Fig. 12A An embodiment is illustrated in which each tile partially covers four regions of interest arranged in a square, and Fig. 12B An example of distribution of chips formed from the tiles is illustrated.
[0058] For the sake of clarity, the drawings are not necessarily drawn to scale. In addition, the number and shape of the regions of interest, tiles and chips shown in the figures are given by way of illustration only. DETAILED DESCRIPTION
[0059] The invention proposes forming a structure comprising at least two chips on a receiving substrate using a substrate called a pseudo-donor substrate, wherein polishing non-uniformities can be avoided even when the chips are relatively widely spaced and / or not very densely distributed on the receiving substrate, the substrate comprising tiles intended to form at least two chips respectively, the chips being distributed on at least two different areas of interest of the receiving substrate. The pseudo-donor substrate is bonded to the receiving substrate via the tiles to transfer parts of the tiles to the receiving substrate. Next, the material on the individual tiles, in particular the material between the chips, is removed so that the individual chips can be formed.
[0060] The expression "not very densely packed" is to be understood in this context as a coverage of the chips on the receiving substrate, which corresponds to a ratio of the total surface area of the chips to the total surface area of the receiving substrate, typically less than 50%, even less than 30%, and / or a distance between two adjacent chips greater than or equal to 3 mm, even greater than or equal to 5 mm.
[0061] The expression "area of interest" is understood herein as an area of a receiving substrate which is intended to subsequently form an electronic circuit different from the electronic circuits formed in adjacent areas. In particular, in a subsequent step of manufacturing the electronic circuit, the receiving substrate can be cut to form individual chips.
[0062] The regions of interest are not necessarily visible on the receiving substrate; they can in particular be defined on a plane indicating the positions of the various components on the receiving substrate at the various steps of the manufacturing process. The positions of the regions of interest on the receiving substrate are connected to the plane via a reference system of the receiving substrate (i.e. a coordinate system attached to the receiving substrate), which makes it possible to define the positions of various points on the surface of the receiving substrate. For example, the reference system can include a reference point that can be located on the edge of the receiving substrate (e.g. a notch commonly used in the semiconductor industry), and two orthogonal axes extending in the plane of the substrate, one of which extends radially from the notch and the other intersecting the first axis in the center of the substrate.
[0063] Figure 1 An example of the distribution of regions of interest on a receiving substrate is illustrated. The receiving substrate 3 has a plurality of regions of interest Z1, Z2, Zn (n is an integer between 1 and the total number of regions of interest), which are delimited by dashed lines. These regions are not necessarily marked on the surface of the receiving substrate, but the position of each region of interest in the reference system of the receiving substrate is known.
[0064] exist Figure 1 In the illustrated example, the regions of interest all have the same rectangular shape and are arranged in rows and columns on the receiving substrate to form a grid. However, in other embodiments not illustrated, the regions of interest may have different sizes and / or shapes.
[0065] The transfer process involves subjecting the tile to at least one chemical mechanical polishing. This polishing can be performed after the tile is placed on a support substrate to form a pseudo-donor substrate, before being bonded to a receiving substrate, on the portion of the tile transferred to the receiving substrate, or even on both the tile of the pseudo-donor substrate and the portion of the tile transferred to the receiving substrate.
[0066] Since the tiles are sized to form at least two chips, the coverage of the tiles on the dummy donor substrate can be increased and / or the distance between the tiles can be reduced relative to the coverage of the chips on the receiving substrate and the distance between the chips.
[0067] Thus, the coverage of the tiles on the pseudo-donor substrate, i.e. the ratio of the total surface area of the tiles to the total surface area of the support substrate, is advantageously greater than 50%, preferably greater than 75%. The diameter of the receiving substrate may be substantially the same as the diameter of the support substrate. In this case, this coverage will also be the coverage of the portion of the tiles transferred to the receiving substrate.
[0068] Advantageously, the distance between two adjacent patches is less than or equal to 3 mm, preferably less than or equal to 1 mm.
[0069] Furthermore, the pseudo-donor receiver and receiving substrate may have the shape of a relatively thin disk, typically a few hundred micrometers thick (commonly referred to as a "wafer" in the microelectronics field), or a rectangle or any geometric shape.
[0070] Thus, the tiles (and transferred tile portions) are advantageously sufficiently densely distributed so that chemical mechanical polishing does not produce thickness non-uniformities between tiles or transferred tile portions. Specifically, deformation of the polishing cloth between adjacent tiles is minimized so that polishing is substantially uniform over the entire surface of the tile or tile portion.
[0071] Only after this polishing step is the chip formed by removing material from the portion of the block transferred to the receiving substrate. This material removal leads to a reduction in the density of the chip relative to the density of the block, but does not entail the risk of losing uniformity in chip thickness and shape, as long as the chip does not need to be subjected to a further chemical-mechanical polishing step.
[0072] The removal of material can be carried out by various techniques known per se.
[0073] Some techniques use a mask to protect the surface of the chip and expose the surface of the portion of the tile to be removed by applying an etchant.
[0074] Particularly advantageously, the mask is formed by photolithography, and then the protective film is formed by a photoresist resistant to an etchant, the etchant being a chemical etching composition. Since photolithography is very precise, it is possible to ensure that the chip has the desired shape and position.
[0075] Alternatively, the etchant is an ion beam (a technique known as "sputtering"). The mask may include a metal protective film or photoresist that protects the material of the tile in the areas covered by the protective film from the ion beam.
[0076] After all or part of the thickness of the tile portion has been etched, the mask is removed.
[0077] It is noted that if a selective etching process is used, in which a controlled focused etching ion beam is scanned over only the area of the tile portion to be removed, then a mask may not be used.
[0078] The dummy donor substrate can be produced by placing individual tiles or small groups of tiles on a support substrate, for example by a "Pick and Place" process; on the other hand, the transfer of multiple tiles from the dummy donor substrate to the receiving substrate can be performed simultaneously.
[0079] Thus, the process can subdivide the tile into a plurality of smaller chips extending over a plurality of different regions of interest, which are preferably adjacent. Thus, the process has the advantage of being able to form chips of very small size without the need to place them individually on a receiving substrate, while avoiding the problems of non-uniformity due to chemical mechanical polishing (if the chips are not very densely distributed).
[0080] For example, each piece initially has an edge with a length greater than or equal to 5 mm, preferably greater than or equal to 8 mm, preferably greater than or equal to 10 mm.
[0081] Advantageously, the tiles are made of a material that is not available on the market in the form of a donor substrate of large dimensions. Thus, the diameter of the donor substrate may be less than 30 cm, for example about 10 cm or 15 cm.
[0082] This is particularly true for III-V semiconductor materials, including nitrides (e.g., indium nitride (InN), gallium nitride (GaN), and aluminum nitride (AlN) for binary compounds), arsenides (e.g., indium arsenide (InAs), gallium arsenide (GaAs), and aluminum arsenide (AlAs) for binary compounds), and phosphides (e.g., indium phosphide (InP), gallium phosphide (GaP), and aluminum phosphide (AlP) for binary compounds).
[0083] The same is true for IV or IV-IV semiconductor compounds such as germanium and silicon carbide.
[0084] The blocks can also be made of piezoelectric materials, such as lithium tantalate (LiTaO3) or lithium niobate (LiNbO3), potassium sodium niobate (K x Na 1-x NbO3 or KNN), barium titanate (BaTiO3), quartz, lead zirconate titanate (PZT), a compound of lead magnesium niobate and lead titanate (PMN-PT), zinc oxide (ZnO), aluminum nitride (AlN) or aluminum scandium nitride (AlScN) (a non-limiting list).
[0085] The tiles may also be made of electrically insulating materials such as diamond, strontium titanate (SrTiO3), yttria stabilized zirconia (YSZ), or sapphire.
[0086] Figures 2 to 9 The formation of a dummy donor substrate is schematically illustrated, including arranging a tile on a first support substrate and transferring the tile parts to a receiving substrate, and then dividing each tile part into at least two chips.
[0087] Reference Figure 2, the pieces P1-P3 are cut from the donor substrate 2. Any technique known to the person skilled in the art can be used to cut the pieces. It can in particular be carried out by sawing and / or cleaving, or by laser cutting. It can also be combined, for example, with a step of partial plasma etching of the cutting lines, a technique also known as "plasma dicing".
[0088] Figure 3 Illustrated Figure 2 The pieces taken away from the donor substrate are placed on the support substrate 1 to form a dummy donor substrate 10.
[0089] Placement may also be performed using a "pick and place" technique, in which a robot grabs a tile or group of tiles previously cut from a donor substrate and places them at a predetermined location on a support substrate.
[0090] In some embodiments, the individual pieces are attached to the support substrate by molecular adhesion. To this end, the pieces and / or the support substrate may be surface treated in advance to promote good molecular adhesion. These treatments may include, in particular, cleaning, deposition of a bonding layer such as silicon oxide (SiO2), plasma activation prior to bonding, and annealing.
[0091] In other embodiments (not illustrated), bonding of the die to the supporting substrate may involve an intermediate bonding layer, such as a polymer bonding layer, a eutectic bonding layer, or a ceramic bonding layer.
[0092] The diameter or size of the support substrate 1 is preferably larger than the diameter or size of the donor substrate 2. For example, the support substrate 1 may have a disk shape with a diameter of 300 mm.
[0093] Figure 4 The formation of a weakened zone in the tiles P1-P3 is illustrated to define the surface portion of the tiles to be transferred to the receiving substrate. As indicated by the arrows, the weakened zone 11 is preferably formed by implanting atomic substances (e.g. hydrogen and / or helium) in the tiles, the depth of which corresponds to the thickness of the layer to be transferred.
[0094] Alternatively, the dummy donor substrate may be the subject of chemical mechanical polishing of the tile surface.
[0095] Figure 5 Illustrated Figure 3 The pseudo-donor substrate in is bonded to the receiving substrate via the tile.
[0096] During this bonding process, the two substrates are positioned relative to each other in such a way that individual patches of the dummy donor substrate 10 at least partially cover at least two different areas of interest of the receiving substrate 3 .
[0097] In the example illustrated, each tile partially covers four regions of interest arranged in a square or rectangle, i.e. the four regions of interest belong to two adjacent rows and two adjacent columns, the centers of the four regions of interest forming the four vertices of the square or rectangle. Each tile thus covers not only part of the surface of the regions of interest, but also the surface of the receiving substrate located between the regions of interest, which surface has a cross shape in this example. Typically, in order to optimize the surface of the receiving substrate and minimize material losses, the distance between the regions of interest is minimized. The surface of each region of interest covered by the tile is adjusted according to the size of the component to be formed in or on the chip or region of interest.
[0098] Particularly advantageously, the individual tiles are attached to the receiving substrate 3 by molecular adhesion. To this end, the tiles and / or the receiving substrate may be subjected to surface treatments beforehand to promote good molecular adhesion. These treatments may include in particular cleaning, deposition of a bonding layer (e.g. silicon dioxide (SiO2)), plasma activation before bonding, polishing and annealing, preferably at low temperatures (generally meaning below 300°C).
[0099] The supporting substrate 1 and the receiving substrate 3 advantageously have substantially the same diameter, for example about 300 mm.
[0100] Next, refer to Figure 6 , the pieces are separated along the weakened zone 11 in order to transfer the portions P'1 , P'2, P'3 delimited by said weakened zone to a receiving substrate.
[0101] The blocks arranged on the support substrate typically have a thickness between 20 μm and 1000 μm, preferably between 100 μm and 700 μm. The transferred portion of each block typically has a thickness between 30 nm and 1.5 μm.
[0102] According to the Smart Cut just described TM Alternatively, after bonding the dummy donor substrate to the receiving substrate, the tile portions can be obtained by thinning the individual tiles of the dummy donor substrate by grinding and / or etching. TM One advantage of the process is that material loss is minimized and the pseudo-donor substrate can be selectively recycled and used to transfer tile portions again to the same receiving substrate or another receiving substrate.
[0103] Optionally, after transferring the tile portion to a receiving substrate, the surface of the tile portion may be subjected to chemical mechanical polishing.
[0104] Figure 7 It is illustrated that a mask is formed, for example, by photolithography, and includes protective films M11, M12, M13, M14 partially covering the respective tile portions and openings extending between portions of the protective films.
[0105] The protective film is configured to define a pattern corresponding to the geometry and location of each chip on each corresponding region of interest.
[0106] The openings generally correspond to the surface of the receiving substrate between the regions of interest. Thus, in the example illustrated, the openings are cross-shaped and delimit four rectangular portions M11 - M14 of the protective film on the respective tile portions.
[0107] According to one embodiment, the protective film does not extend to the edge of the tile portion located in the area of interest. This allows, in a subsequent etching step, also to remove the peripheral area of the tile portion, for example to adjust the size of said portion. However, in order to minimize material losses, the size of the tile is preferably designed to minimize the peripheral area to be removed.
[0108] The film can be made of any suitable resist. In particular in the field of microelectronics, photoresists sold by Shipley or AZ Electronic Materials, for example, are dispensed in a viscous state and spread on a substrate with the aid of a spinner (known as a "spin coating" operation) and then annealed. After development, a UV exposure step can retain or conversely eliminate the exposed areas. According to a variant known as a "dry film photoresist", these resists can also be applied by laminating thick films, which typically have a thickness between 15 and 50 μm and are provided in roll form.
[0109] Reference Figure 8 , apply the etching solution to Figure 7 The composition of the etching solution depends on the material of the block. For example, in the case of blocks made of InP, wet etching in a bath of hydrochloric acid (HCl) and phosphoric acid (H3PO4) can be used. Plasma etching based on halogenated gases (based on, for example, chlorine or on a methane / hydrogen (CH4 / H2) mixture) can also be used.
[0110] Advantageously, the etching solution does not or hardly attacks the material of the receiving substrate 3 .
[0111] Alternatively, etching by ion beam sputtering (for example argon ion beam sputtering) can also be used. In this case, the protective film is advantageously metallic.
[0112] At the end of etching, the parts of the block not covered by the protective film have been removed (over the entire or partial thickness of the block) to separate the chips.
[0113] Reference Fig. 9 , removing the protective film to release the surface of the chip. The structure is thus ready for subsequent manufacturing steps, such as depositing or epitaxying one or more additional layers on the chip, forming electronic components in or on the chip, etc.
[0114] It is important to note that chemical mechanical polishing is not performed after chip formation in order to avoid creating non-uniformities within the chip. In other words, any chemical mechanical polishing step is performed on a pseudo-donor substrate, or on a receiving substrate after transferring a tile portion but before the material removal leading to chip production. Thus, in all cases, chemical mechanical polishing is performed on tiles or tile portions having relatively large dimensions, which are larger than the dimensions required to produce the target chip. More specifically, the dimensions of each tile are several times larger than the target size of the chip during manufacturing. For example, the size of each tile can be two or four times the target size.
[0115] A variety of geometries are available depending on the intended use of the final structure and the placement of the area of interest.
[0116] In certain embodiments, each tile at least partially covers two different regions of interest of a receiving substrate, and a chip is formed from the tile on each corresponding region of interest.
[0117] For example, Fig. 10A As illustrated, each tile and the transferred tile portion P'1 can be arranged to cover portions of two adjacent regions of interest and the surface of the receiving substrate located between the regions of interest. Material removal can be performed to remove material from each tile portion facing the surface separating the regions of interest and, if necessary, to remove material from each tile portion facing each region of interest to conform to the desired geometry of each chip p11, p12 (see Fig. 10B ).
[0118] according to Fig.11A In another example, each tile and the transferred tile portion P'1, P'2, P'n may be in the form of an elongated strip partially covering the aligned multiple regions of interest. Then, one or more chips p11, p12, p1n may be formed aligned in each region of interest belonging to the same row or column (see Fig. 11B ).
[0119] In other embodiments, Fig. 12A ( Figure 5 As illustrated in the enlarged view of FIG. 1 , each tile and the transferred tile portion P'1 can be arranged to cover the corresponding corners of four different regions of interest arranged in a square on the receiving substrate, and from the tile, chips p11, p12, p13, p14 are formed at the corners of the corresponding regions of interest (see FIG. Fig. 12B ).
[0120] Of course, these various configurations may be combined. For example, differently oriented tile portions (eg, pairs of perpendicular ones) may be shifted depending on the area of interest to be covered.
[0121] The process thus provides a great deal of latitude for forming the chip depending on the components to be formed from it.
[0122] Application Examples
[0123] The invention proposes different cases of particularly advantageous application, in particular in the field of microelectronics.
[0124] This process makes it possible to arrange the chips in a less dense manner on a receiving substrate with a diameter of 300 mm. The receiving substrate may be a silicon substrate or a substrate of the silicon-on-insulator (SOI) type.
[0125] In photonic applications, the active layer of the receiving substrate may include photonic circuits with passive or active devices, such as one or more waveguides, one or more multiplexers, one or more microresonators, etc. The chip transferred to this layer may be made of InP, which is a material more suitable than silicon for epitaxial growth of III-IV material stacks to form lasers.
[0126] In radio frequency (RF) applications, the active layer of the receiving substrate may include components operating at relatively low frequencies, while the chip, advantageously made of InP or GaN, may include components operating at higher frequencies.
[0127] In micro-LED applications, the size of the GaN chip is advantageously less than 50 μm.
Claims
1. A method for manufacturing a structure comprising at least two chips (p11, p12, p13, p14) located on a receiving substrate (3), include: - forming a dummy donor substrate (10) by arranging at least one piece (P1, P2, P3) of at least one donor substrate (2) on a support substrate (1); - bonding the pseudo-donor substrate (10) to a receiving substrate (3) via the tiles, such that each tile at least partially covers at least two different regions of interest (Z1, Z2, Zn) of the receiving substrate (3); - transferring a portion (P'1, P'2, P'3) of said patch (P1, P2, P3) to said receiving substrate (3); - at least one chemical mechanical polishing step performed on said pieces (P1, P2, P3) of said dummy donor substrate and / or on the parts of the pieces transferred to said receiving substrate; - After the at least one chemical mechanical polishing step, material is removed from the tile parts (P'1, P'2, P'3) to divide each tile part into at least two chips (p11, p12, p13, p14) respectively arranged on the corresponding area of interest (Z1, Z2, Zn).
2. The method according to claim 1, in, After the etching step, the coverage of the chips (p11, p12, p13, p14) on the receiving substrate (3), ie the ratio of the total surface area of the chips to the total surface area of the receiving substrate, is less than 50%, preferably less than 30%.
3. The method according to any one of claims 1 and 2, in, The coverage of the patches (P1, P2, P3) on the pseudo-donor substrate (10), that is, the ratio of the total surface area of the patches to the total surface area of the support substrate, is greater than 50%, preferably greater than 75%.
4. The method according to any one of claims 1 to 3, in, Each piece (P1, P2, P3) has a side length greater than or equal to 5 mm, preferably greater than or equal to 8 mm, and more preferably greater than or equal to 10 mm.
5. The method according to any one of claims 1 to 4, in, Each chip (p11, p12, p13, p14) has a side length less than or equal to 5 mm, preferably less than or equal to 3 mm, and more preferably less than or equal to 1 mm.
6. The method according to one of claims 1 to 5, in, The distance between two adjacent patches (P1, P2, P3) is less than or equal to 3 mm, preferably less than or equal to 1 mm.
7. The method according to one of claims 1 to 6, in, The distance between two adjacent chips (p11, p12, p13, p14) is greater than or equal to 3 mm, preferably greater than or equal to 5 mm.
8. The method according to one of claims 1 to 7, in, Each tile at least partially covers two different regions of interest of the receiving substrate, and a chip is formed by the tile on each corresponding region of interest.
9. The method according to one of claims 1 to 7, in, Each tile covers at least corresponding corners of four different regions of interest arranged in a square or rectangle on the receiving substrate, and a chip is formed by the tile at each corner of the corresponding region of interest.
10. The method according to one of claims 1 to 7, in, Each tile is in the form of a strip, at least partially covering at least three aligned regions of interest of the receiving substrate, and a chip is formed from the tile on each corresponding region of interest.
11. The method according to one of claims 1 to 10, in, The support substrate (1) and the receiving substrate (3) have substantially the same diameter, which is larger than the diameter of the donor substrate (2).
12. The method according to one of claims 1 to 11, in, The pseudo-donor substrate (10) is bonded to the receiving substrate (3) by molecular adhesion.
13. The method according to one of claims 1 to 12, in, The transfer of each piece part (P'1, P'2, P'3) includes in sequence: forming a weakened zone (11) by implanting atomic substances in each piece (P1, P2, P3) of the pseudo donor substrate (10) to define the piece part (P'1, P'2, P'3) to be transferred, bonding the pseudo donor substrate (10) to the receiving substrate (3) via the piece (P1, P2, P3), and separating each piece along the weakened zone (11).
14. The method according to one of claims 1 to 12, in, The transfer of the individual tile portions (P'1, P'2, P'3) comprises thinning the individual tiles (P1, P2, P3) of the pseudo-donor substrate by grinding and / or etching.
15. The method according to one of claims 1 to 14, in, Each patch (P1, P2, P3) has a thickness between 20 μm and 1000 μm, preferably between 100 μm and 700 μm, and the transferred portion of each patch has a thickness between 30 nm and 1.5 μm.
16. The method according to one of claims 1 to 15, in, Each block (P1, P2, P3) includes: a semiconductor material, such as a III-V material, in particular indium nitride InN, gallium nitride GaN, aluminum nitride AlN, indium arsenide InAs, gallium arsenide GaAs, aluminum arsenide AlAs, indium phosphide InP, gallium phosphide GaP or aluminum phosphide AlP, or a IV or IV-IV material, in particular germanium or silicon carbide SiC, - Piezoelectric materials, such as lithium tantalate LiTaO 3 、LiNbO 3 、Potassium sodium niobate K x Na 1-x NbO 3 or KNN, barium titanate BaTiO 3 , quartz, lead zirconate titanate PZT, a compound of lead magnesium niobate and lead titanate PMN-PT, zinc oxide ZnO, aluminum nitride AlN or aluminum scandium nitride AlScN, and / or - Electrically insulating materials such as diamond, strontium titanate, yttria-stabilized zirconia or sapphire.
17. The method according to one of claims 1 to 16, comprising the step of forming at least one epitaxial layer on the respective tile parts by epitaxy, the step of epitaxy being followed by material removal in order to provide the respective tile parts as chips.
18. The method according to one of claims 1 to 17, in, The material removal includes: - forming a mask comprising a protective film and at least one opening defining a space to be formed between two chips, the protective film partially covering the tile portions in a pattern defining the chips in each tile portion; - etching the respective tile portions through the respective openings of the mask to form the spaces for separating the chips from each other.
19. The method according to claim 18, in, The etching is performed using an etching solution, and the mask is formed of a photoresist.
20. The method according to claim 18, in, The etching is performed using an ion beam, and the mask is formed of metal.
21. The method according to one of claims 1 to 17, in, The material removal is performed selectively by a focused ion beam which is controlled to scan only the areas of the tile portion to be removed.
