Transfer method and structure of a semiconductor device

Abstract

Title: Method and Structure for Transferring a Semiconductor Device. The invention relates to a method for transferring a semiconductor device (100) by IR laser ablation. An active layer is transferred from a donor substrate to a silicon transfer substrate (3) by bonding it with a cleavage layer (20) between the active layer and the transfer substrate (3). The semiconductor device (100) is then formed from the active layer on the transfer substrate (3). The semiconductor device (100) is then assembled onto a receiving substrate (2) and subsequently released under infrared radiation through the transfer substrate (3). The cleavage layer (20) advantageously comprises a metallic release layer (21) and a mineral absorption layer (22). The invention also relates to a transfer structure enabling the implementation of the transfer method. Figure for the abstract: Fig. 13

Description

Title of the invention: Method and structure for transferring a semiconductor device technical field

[0001] The present invention relates to the field of technologies for microelectronics and optoelectronics. It finds a particularly advantageous application in the mass transfer of semiconductor devices, for example, GaN-based silicon light-emitting diodes. STATE OF THE ART

[0002] A semiconductor device typically comprises at least one semiconductor layer, referred to as the "active" layer, which plays a role in the operation of said device. A semiconductor device may refer, for example, but not limited to, a CMOS circuit, i.e., a circuit based on complementary metal-oxide-semiconductor transistors (CMOS being the English acronym for "Complementary Metal Oxide Semiconductor"), a MEMS microelectromechanical system (English acronym for "MicroElectroMechanical Systems"), or an optoelectronic device such as an LED (English acronym for "Light Emitting Diode"). In the following, one or more LEDs or micro-LEDs are chosen as examples to illustrate the transfer process and structure, without this being a limiting factor.

[0003] Typically, to form a self-emissive display screen comprising a plurality of RGB (English acronym for "Red Green Blue") pixels emitting their own light, different LEDs are assembled on a screen support, which includes, for example, control electronics.

[0004] The LEDs are typically formed on a donor substrate or growth substrate, then individualized—that is, separated from one another—before being transferred to the display support or a receiving substrate. Not all the LEDs from the donor substrate are necessarily transferred to the receiving substrate.

[0005] The LED growth substrate is preferably silicon-based, in order to take advantage of the production capabilities of microelectronics industry equipment. To collectively transfer a large number of LEDs from a silicon-based growth substrate, one solution is to use a glass transfer substrate onto which the LEDs are bonded with an organic adhesive. The growth substrate is removed, and the LEDs are then individually placed in contact with the receiving substrate. UV laser radiation then degrades the organic adhesive through the glass transfer substrate, freeing the LEDs and assembling them at the Receiving substrate. This UV laser ablation method requires a glass transfer substrate that is poorly compatible with microelectronics production equipment (substrate detection, mechanical deformation). Furthermore, organic adhesives are temperature-sensitive. They can cause significant wafer deformation and degrade under the influence of temperature. Consequently, performing intermediate technological steps on the LEDs after removal of the growth substrate becomes more complicated.

[0006] To overcome these drawbacks, document WO2022111141 Al discloses a solution involving mineral bonding, for example, oxide-oxide bonding, between the transfer substrate and the devices on the donor substrate. In this solution, a silicon transfer substrate with an inorganic cleavage layer, typically aluminum-based, is used. IR laser radiation is then used to degrade the cleavage layer through the silicon transfer substrate, freeing the devices assembled on the recipient substrate. In practice, this solution requires a very energetic IR laser and can partially degrade the devices to be transferred.

[0007] The present invention aims to overcome at least partially the drawbacks of the solutions mentioned above.

[0008] In particular, one object of the present invention is to provide an optimized method for transferring a semiconductor device. Another object of the present invention is to provide a transfer structure for a semiconductor device, enabling the implementation of the transfer method.

[0009] The other objects, features, and advantages of the present invention will become apparent from an examination of the following description and accompanying drawings. It is understood that other advantages may be incorporated. In particular, certain features and advantages of the transfer method may be applied mutatis mutandis to the transfer structure, and vice versa. SUMMARY

[0010] To achieve the objectives mentioned above, one aspect relates to a process for transferring at least one semiconductor device, from a first substrate called donor to a second substrate called receiver, using a silicon-based transfer substrate.

[0011] The process comprises at least the following steps: - to provide the donor substrate comprising at least one active layer intended to be integrated into the semiconductor device, at the level of a first face, - to provide the silicon-based transfer substrate, featuring a second face, - form a cleavage layer on at least one of the first face and the second face, the cleavage layer being inorganic, - assemble the donor substrate with the transfer substrate, via the cleavage layer, - remove the donor substrate, retaining at least one active layer on the transfer substrate, and form at least one semiconductor device from at least one active layer, - provide the receiving substrate, - assemble at least one semiconductor device with the receiving substrate, - illuminate, by infrared radiation, the cleavage layer through the transfer substrate, so as to eliminate at least part of the cleavage layer, to decouple the at least one semiconductor device from the transfer substrate.

[0012] Advantageously, the formation of the split layer includes the formation of a metal release layer on the side of the semiconductor layer, and the formation of a mineral absorption layer on the side of the transfer substrate, so that the split layer includes the metal release layer and the mineral absorption layer, the mineral absorption layer being configured to absorb at least 20% of the infrared radiation during illumination.

[0013] Thus, the cleavage layer is in the form of a bilayer which advantageously allows the functions of each of the metallic release and mineral absorption layers to be separated.

[0014] The metallic release layer is designed to melt or vaporize under the effect of the heat imparted by the absorption layer, which absorbs IR radiation. The metallic release layer also protects at least one semiconductor device from infrared radiation by reflecting the IR radiation back to the absorption layer.

[0015] The mineral absorption layer efficiently absorbs infrared radiation, producing sufficient heat to partially or completely melt or vaporize the metallic release layer. The energy of the IR radiation can thus be reduced.

[0016] During the development of the present invention, it was observed that the use of a simple aluminum layer as a cleavage layer, while theoretically feasible, requires very high laser power in practice due to the reflectivity of aluminum. A picosecond (ps) pulsed IR laser is not practical.

[0017] The addition of a mineral absorption layer, for example based on a refractory transition nitride, makes it possible to sufficiently reduce the energy required for the melting / vaporization of the metal release layer, to allow the use of a pulsed IR laser delivering an energy less than or equal to 10 microjoules (pj).

[0018] Document WO2022111141 Al provides alternative options for the splitting layer: either an aluminum layer or a light-to-heat conversion layer. In the first case, as indicated, the energy required is very high due to the reflectivity of aluminum. In the second case, the reflectivity of the layer decreases, and the devices are no longer protected from IR radiation.

[0019] Practical developments carried out during the development of the present invention have led to considering a combination of absorption and liberation / reflection layers that is not considered in the alternatives proposed by document WO2022111141 AL

[0020] This combination advantageously reduces the energy required delivered by the pulsed IR laser while protecting the devices to be transferred.

[0021] Another aspect concerns a transfer structure comprising, in stacking order, the following: - a first substrate called the donor, - at least one active layer intended to be integrated into at least one semiconductor device, - a splitting layer, - a silicon-based transfer substrate,

[0022] The stack is configured so that, under illumination by infrared (IR) radiation through the transfer substrate, the cleavage layer is at least partially eliminated, so as to decouple the transfer substrate from the stack.

[0023] Advantageously, the splitting layer comprises: - a metallic release layer on the side of the active layer, - a mineral absorption layer on the side of the transfer substrate, the layer mineral absorption being configured to absorb at least 20% of infrared radiation during illumination.

[0024] Such a transfer structure advantageously allows the transfer process described above to be implemented.

[0025] The layer stack does not include organic layers. This stack can therefore be advantageously processed by conventional microelectronic process steps, for example to form a semiconductor device from the active layer, after removal of the donor substrate and before decoupling from the transfer substrate. BRIEF DESCRIPTION OF THE FIGURES

[0026] The aims, objects, features and advantages of the invention will be apparent better than the detailed description of the methods of implementation of the latter which are illustrated by the following accompanying drawings in which:

[0027] [Fig. 1A][Fig. 1B][Fig. 2][Fig. 3][Fig. 4][Fig. 5][Fig. 6] Figures 1A to 6 illustrate schematically, the steps of a transfer process according to an embodiment of the present invention.

[0028] [Fig.7][Fig.8][Fig.9][Fig.l0][Fig.ll][Fig.l2][Fig.l3][Fig.l4] Figures 7 to 14 schematically illustrate steps of a transfer process according to another embodiment of the present invention.

[0029] The drawings are given by way of example and are not limiting of the invention. They constitute schematic representations of principle intended to facilitate understanding of the invention and are not necessarily to scale with practical applications. In particular, the dimensions of the various layers and parts of the transfer structures and LEDs are not necessarily representative of reality. DETAILED DESCRIPTION

[0030] Before beginning a detailed review of embodiments of the invention, it is recalled that the invention according to its first aspect includes in particular the following optional features which may be used in combination or alternatively:

[0031] According to one example, the at least one semiconductor device comprises a plurality of semiconductor devices each having a characteristic dimension less than or equal to 10 pm, and preferably less than or equal to 5 pm.

[0032] According to one example, the illumination is configured to decouple only certain semiconductor devices from the plurality of semiconductor devices. The method thus allows a localized transfer of micro-components.

[0033] According to one example, the semiconductor devices of the plurality of semiconductor devices are separated from one another before illumination, the semiconductor devices remaining held together by the cleavage layer on the transfer substrate. The semiconductor devices are typically individualized before separation by laser illumination.

[0034] According to one example, the illumination is performed by a pulsed IR laser producing IR radiation with an energy between 0.5 pJ and 10 pJ, preferably between 1 pJ and 10 pJ, for a duration between 1 picosecond (ps) and 20 ps. Such a laser advantageously emits "low" energy radiation that can be shaped to target individual semiconductor devices.

[0035] According to one example, the IR radiation has a wavelength less than or equal to 2.5 pm. According to one example, the illumination is carried out by a pulsed IR laser producing IR radiation in the form of pulses, and the energy of a pulse is between 0.1 pJ and 100 pJ. According to one example, the duration of a pulse is between 100 femtoseconds (fs) and 50 picoseconds (ps).

[0036] According to one example, the mineral absorption layer has a resistivity between 50 p.ohm.cm (micro-ohm-centimeter) and 1 m.ohm.cm (milli-ohm-centimeter). Such an absorption layer conducts the charge carriers produced during the absorption of IR radiation. These charge carriers are advantageously conducted to the metallic release layer to increase the heat transferred to the metallic release layer.

[0037] According to one example, the mineral absorption layer is chosen to be made of a refractory material with a melting point greater than or equal to 1500°C. This allows the implementation of most, if not all, thermal processes in the microelectronics industry without degrading the mineral absorption layer. Thermal degradation under IR illumination is thus limited to the metal release layer.

[0038] According to one example, the mineral absorption layer is chosen to be a refractory material based on a transition metal, preferably based on Ti, V, Zr, Ta, Hf, Nb, W.

[0039] According to one example, the mineral absorption layer is chosen from transition metals such as Ti, V, Zr, Ta, Hf, Nb, W.

[0040] According to one example, the mineral absorption layer is selected from transition metal-based refractory transition nitrides such as TiN, VN, ZrN, TaN, HfN, NbN, or their alloys. According to another example, the mineral absorption layer is selected from transition metal-based refractory transition carbides. According to yet another example, the mineral absorption layer is selected from transition metal-based refractory transition borides.

[0041] According to one example, the mineral absorption layer has a thickness between 10 nm and 100 nm.

[0042] According to one example, the mineral absorption layer has an extinction coefficient k > 1 for an illumination wavelength less than or equal to 2.5 pm.

[0043] According to one example, the metal release layer has an infrared reflectivity greater than or equal to 97%. This makes it possible to protect the semiconductor device from IR radiation during illumination.

[0044] According to one example, the metal release layer has a resistivity between 2.5 p.ohm.cm (micro-ohm-centimeter) and 50 p.ohm.cm (micro-ohm-centimeter).

[0045] According to one example, the metal release layer has a thickness between 30 nm and 200 nm.

[0046] According to one example, the metal release layer is based on a simple metal such than aluminum, copper, gold, silver.

[0047] According to one example, the metal release layer has a melting temperature less than or equal to 1100°C.

[0048] According to one example, the mineral absorption layer is made of a refractory material having a melting temperature greater than or equal to 2500°C.

[0049] According to one example, the mineral absorption layer is based on a transition refractory nitride taken from among TiN, VN, ZrN, TaN, HfN, NbN, or an alloy of said transition refractory nitrides.

[0050] Except where incompatibility exists, technical features described in detail for a given embodiment may be combined with technical features described in the context of other embodiments described by way of example and without limitation, so as to form another embodiment which is not necessarily illustrated or described. Such an embodiment is obviously not excluded from the invention.

[0051] In the present invention, the method is particularly dedicated to the transfer of semiconductor devices, especially devices with micrometer dimensions such as light-emitting diodes (LEDs). Individual LEDs or semiconductor devices typically have dimensions, when projected onto an xy basis plane, ranging from 2.5 pm x 2.5 pm to 50 pm x 50 pm.

[0052] The invention can be implemented more broadly for various microelectronic or optoelectronic devices, or even for MEMS electromechanical devices or microsystems. For example, the invention can be implemented in laser or photovoltaic devices.

[0053] Unless explicitly stated otherwise, it is specified that, within the framework of the present invention, the relative disposition of a third layer intercalated between a first layer and a second layer does not necessarily mean that the layers are directly in contact with each other, but means that the third layer is either directly in contact with the first and second layers, or separated from them by at least one other layer or at least one other element.

[0054] Thus, the terms and phrases "to take support" and "to cover" or "to cover" do not necessarily mean "in contact with".

[0055] The steps of the process as claimed are understood in a broad sense and may possibly be carried out in several sub-steps.

[0056] In this patent application, the terms "light-emitting diode", "LED" or simply "diode" are used synonymously. An "LED" may also be understood to mean a "micro-LED" or even a smart LED, as appropriate.

[0057] A portion or element described as "sacrificial" means that this element is intended to be "sacrificed", that is to say, removed during a later process step.

[0058] A substrate, layer, or device "based" on a material M is understood to be a A substrate, a layer, or a device comprising only this material M, or this material M and possibly other materials, for example, alloying elements, impurities, or dopant elements. Thus, a GaN-based diode typically comprises GaN and AlGaN or InGaN alloys.

[0059] A coordinate system, preferably orthonormal, comprising the x, y, z axes is shown in some of the accompanying figures. This coordinate system can be applied by extension to the other figures on the same sheet of figures.

[0060] In this patent application, the terms thickness for a layer and height for a structure or device will be preferred. The thickness is taken along a direction normal to the principal extension plane of the layer, and the height is taken perpendicular to the xy base plane. Thus, a layer typically has a thickness along z when it extends mainly along an xy plane, and a protruding element, for example a device, has a height along z. The relative terms "on," "under," and "below" preferentially refer to positions taken along the z direction.

[0061] Dimensional values ​​are understood to be within manufacturing and measurement tolerances.

[0062] The terms "approximately," "about," and "in the order of" mean, when referring to a value, "within 10%" of that value, or, when referring to an angular orientation, "within 10°" of that orientation. Thus, a direction substantially normal to a plane means a direction having an angle of 90+10° with respect to the plane.

[0063] In this patent application, the optical absorption coefficient, denoted a or a, is defined as the ratio between the absorbance and the optical path length traveled by electromagnetic radiation in a given medium (expressed in m² or cm²)

[0064] The extinction coefficient (also called the attenuation coefficient) of the medium, denoted k, measures the energy loss of electromagnetic radiation passing through that medium. Analogous to the absorption coefficient, it also takes into account, in addition to absorbance, the effects due to scattering and luminescence. It depends on the material and the wavelength. It is the imaginary part of the complex refractive index: n = n' + in" with k = n". Transparent materials have a low extinction coefficient, while opaque materials have a high extinction coefficient.

[0065] Within the framework of the present invention, the illumination is carried out in the infrared wavelength range, preferably in the near-infrared range, for wavelengths between 1 pm and 2.5 pm.

[0066] An object of the invention is to transfer, via a transfer structure and an IR laser ablation process, a semiconductor or optoelectronic device onto a substrate receiver, for example, a screen support. One principle of the invention consists of performing transfers and bonding via a separation layer comprising a metallic release layer and a mineral absorption layer. The metallic release layer is a sacrificial layer in the transfer process. In particular, the absorption layer absorbs IR light and converts it into heat. This heat diffuses, notably towards the metallic release layer. This typically induces, at the interface between the two layers, partial melting or partial vaporization of the metallic release layer. The interface between the two layers is then degraded, and the two layers can separate. A portion of the release layer that has not melted may remain on the side of the device being transferred.

[0067] Figures IA and IB illustrate two variants in which the cleavage layer 20 is respectively formed directly on the transfer substrate 3 ([Fig.1A]) or directly on the donor substrate 1 ([Fig.1B]).

[0068] According to a first example illustrated in [Fig. 1A], a donor substrate 1 typically comprising an active layer 10 surmounted by a mineral bonding layer 11 is provided. The donor substrate 1 is preferably silicon-based.

[0069] The active layer 10 can be, for example, based on GaN (p-GaN and / or n-GaN), and / or AlGaN, and / or InGaN epitaxially grown on the donor substrate 1. It typically constitutes an active layer, comprising, for example, a PN junction or quantum wells, configured to form a semiconductor device, for example, one or more LEDs. The active layer 10 can also comprise three-dimensional structures arranged side by side, for example, based on InGaN, such as nanowires or nanopyramids. The active layer 10 can also include optionally structured sublayers intended to form, for example, electrical contacts, for example, on p-GaN and on n-GaN, for example, based on a transparent conductive oxide (TCO).

[0070] The mineral bonding layer 11 is typically based on silicon dioxide SiO2. In particular, it allows the active layer 10 to be planarized so as to obtain a flat face 201 suitable for mineral bonding, and allowing direct SiO2-on-SiO2 bonding.

[0071] The transfer substrate 3 is silicon-based. In this example, the cleavage layer 20 is formed on a face 202 of the transfer substrate 3. The exposed face 203 of the cleavage layer 20 is typically flat and suitable for mineral bonding. Mineral bonding can, for example, be molecular bonding. The mineral bonding between the transfer substrate 3 and the donor substrate 1 occurs here between faces 201 and 203.

[0072] The cleavage layer 20 comprises at least one mineral absorption layer 22, and at least one metallic release layer 21. The mineral absorption layer 22 is formed on the side of the transfer substrate 3, for example directly on the face 202. The metallic release layer 21 is formed on the mineral absorption layer 22, such that the mineral absorption layer 22 is intercalated between the transfer substrate 3 and the metallic release layer 21. The exposed face 203 of the cleavage layer 20 corresponds here to a face of the metallic release layer 21. According to a possibility not shown, a silicon oxide layer is formed on face 203 of the cleavage layer 20. The bonding then occurs via SiO2-SiO2 molecular bonding.

[0073] The mineral absorption layer 22 is typically at least partially electrically conductive and preferably has a resistivity between 50 pΩ·cm and 1 mΩ·cm. This optimizes the electronic heat transfer between the mineral absorption layer 22 and the metallic release layer 21. The mineral absorption layer 22 preferably has a thickness along the z-axis between 10 nm and 100 nm. It typically has an optical extinction coefficient k > 1 in the near-infrared range, for wavelengths less than or equal to 2.5 pm. This allows it to absorb at least 20%, or even at least 30% or at least 40%, of near-infrared radiation. The mineral absorption layer 22 can be based on a transition metal, for example Ti, Ta, W, Zr, Nb, or Hf.It can alternatively be based on a nitride of these transition metals, called a transition refractory nitride, for example TiN, TaN, VN, ZrN, HfN, NbN. Alternatively, it can be based on a carbide of these transition metals, called a transition refractory carbide, for example TiC, TaC, VC. Alternatively, it can be based on a boride of these transition metals, called a transition refractory boride, for example TiB2, TaB, TaB2, VB2, HfB2. The mineral absorption layer 22 can be based on a combination or alloy of these materials, for example C5HfTa4. The mineral absorption layer 22 preferably has a melting point greater than or equal to 1500°C. This prevents the mineral absorption layer 22 from thermally degrading upon illumination. Thus, after illumination, the integrity of the mineral absorption layer 22 is preserved.

[0074] The metallic release layer 21 is typically electrically conductive and preferably has a resistivity between 2.5 p.ohm.cm and 50 p.ohm.cm. The metallic release layer 21 preferably has a thickness along z between 30 nm and 200 nm. It typically has an optical reflectivity greater than or equal to 97% in the near-infrared range, for wavelengths less than or equal to 2.5 pm. When illuminated through the transfer substrate 3, the infrared radiation transmitted by the mineral absorption layer 22 is thus reflected by the metallic release layer 21. This protects the semiconductor layer 10 and / or a semiconductor device from IR radiation. formed from the semiconducting layer 10. The metallic release layer 21 can be based on a single metal, for example Al, Cu, Au, Ag, Ti. The metallic release layer 21 preferably has a melting point less than or equal to 1100°C. The difference in melting point between the metallic release layer 21 and the mineral absorption layer 22 is thus sufficiently large to guarantee partial (not shown) or complete melting / vaporization of the metallic release layer 21 while preserving the integrity of the mineral absorption layer 22.

[0075] The characteristics of the different layers 10, 11, 21, 22, in particular of the mineral absorption layer 22 and the metallic release layer 21, are common to the other embodiments described below.

[0076] According to a second example illustrated in [Fig. 1B], the cleavage layer 20 is formed on a face 201 of the donor substrate 1. The exposed face 204 of the cleavage layer 20 corresponds here to a face of the mineral absorption layer 22. The mineral bonding between the transfer substrate 3 and the donor substrate 1 occurs here between faces 202 and 204. According to a possibility not illustrated, one or more layers of silicon oxide are formed on face 204 of the cleavage layer 20 and on face 202 of the transfer substrate 3. The bonding then occurs by molecular bonding of the SiO2-SiO2 type.

[0077] In this example, the metallic release layer 21 is formed on the side of the donor substrate 1, for example directly on the face 201. The mineral absorption layer 22 is formed on the metallic release layer 21, so that the metallic release layer 21 is intercalated between the donor substrate 1 and the mineral absorption layer 22.

[0078] As illustrated in [Fig.2], regardless of which face 201 or 202 is formed on the cleavage layer 20, the donor substrate 1 and the transfer substrate 3 are assembled via the cleavage layer 20, typically by molecular bonding.

[0079] As illustrated in [Fig.3], the donor substrate 1 is then removed, for example partly by grinding and partly by chemical-mechanical polishing (CMP) or by selective chemical attack with respect to the active layer 10. The active layer 10 and the oxide layer 11 are retained on the transfer substrate 3.

[0080] As illustrated in [Fig. 4], the active layer 10 and the oxide layer 11 are then structured to form and individualize semiconductor devices 100, for example, LEDs. This structuring may include annealing steps at temperatures of several hundred degrees Celsius. Advantageously, the layers 21, 22 of the cleavage layer 20 can withstand these temperatures. The structuring can therefore be carried out after transferring the initial layers 10, 11 onto the transfer substrate 3. The transfer process is thus less restrictive for the sequence of steps in forming the devices 100. In the example illustrated in [Fig. 4], the semiconductor devices 100 comprise an insulating and transparent part 111, an active part 110 typically configured to emit light, and contact pads 112. The formation of individual devices 100 may include one or more anisotropic etchings along z. The etching of the oxide layer 11 is typically selective with respect to the metallic release layer 21.

[0081] As illustrated in [Fig. 5], after the devices 100 are formed, the transfer substrate 3 is brought into contact with the receiving substrate 2, for example, a screen support, intended to receive at least some of the devices 100. The face 205 of the receiving substrate 2 typically has contact pads 211 at the locations intended for receiving the devices 100. After alignment of the contact pads 112 and 211, bonding is performed, for example, by heat compression or by Cu-Cu bonding between the contact pads 112 and 211. The cleavage layer 20 is then illuminated by IR radiation through the transfer substrate 3. This IR radiation is typically localized only above the devices 100 to be transferred. At least 20%, preferably at least 30% or at least 40%, of the IR radiation is typically absorbed by the layer 22.The energy of the IR radiation thus absorbed is transferred as heat to the metallic release layer 21.

[0082] As illustrated in [Fig. 6], the metallic release layer 21 melts or vaporizes under the effect of the heat induced by IR radiation, thus releasing devices 100a assembled on the receiving substrate 2. Since the IR radiation is localized, other devices 100b can be retained on the transfer substrate 3, via unmelted portions 21p of the metallic release layer 21. The transfer substrate 3 can then be moved to another location to perform another transfer, for example, for the device(s) 100b.

[0083] IR radiation typically has a wavelength between 1 pm and 2.5 pm, preferably between 1.5 pm and 2 pm. It is preferably produced in pulses by a pulsed laser. The energy of an IR radiation pulse is preferably between 0.1 pJ and 100 pJ, preferably between 1 pJ and 20 pJ, preferably between 5 pJ and 10 pJ, for example, on the order of 7 or 8 pJ. The duration of an IR radiation pulse is preferably between 100 fs and 50 ps, ​​preferably between 1 ps and 20 ps. A single IR radiation pulse can be sufficient to detach the device 100a from the transfer substrate 3.

[0084] Figures 7 to 14 illustrate another embodiment of the transfer process. In this embodiment, a first transfer of the semiconductor layer 10 is carried out from the growth substrate la to an intermediate substrate 1b, for example, silicon-based. After growth of the active layer 10 on the substrate la ([Fig. 7]), the intermediate substrate 1b is bonded to the active layer 10 via one or more oxide layers 11, for example, oxide-based. silicon ([Fig.8]). The growth substrate is then removed, for example partly by trimming and partly by chemical mechano-polishing (CMP) or by selective chemical attack with respect to the active layer 10, so as to expose a face 200 of the active layer 10 ([Fig.9]).

[0085] As illustrated in [Fig. 10], the intermediate substrate 1b bearing the active layer 10 is bonded to the transfer substrate 3 via the cleavage layer 20. The cleavage layer 20 can be pre-formed on face 200 of the active layer 10 or on face 202 of the transfer substrate 3. The intermediate substrate 1b here corresponds to the donor substrate 1 of the embodiment described above. The mineral absorption layer 22 is formed so as to be positioned on the side of the transfer substrate 3. The metallic release layer 21 is formed so as to be positioned on the side of the active layer 10.

[0086] As illustrated in [Fig. 11], after bonding the intermediate substrate 1b is removed, for example by trimming and / or polishing and / or chemical etching. The oxide layer(s) 11 are preferably also removed so as to expose the active layer 10, for example by polishing and / or chemical etching.

[0087] As illustrated in [Fig. 12], the active layer 10 can then be structured to form the individual devices 100. In this example, the devices 100 typically comprise an active portion 110, for example configured to emit light, and contact pads 112. As before, the formation of the individual devices 100 may include one or more anisotropic etchings along z. The etching of the active layer 10 is typically selective with respect to the underlying metallic release layer 21.

[0088] As before, after the devices 100 have been formed, the transfer substrate 3 is brought into contact with the receiving substrate 2. The contact pads 112 and 211 are aligned and bonded together. Localized illumination of the cleavage layer 20 is then performed by IR radiation through the transfer substrate 3 ([Fig. 13]). The metallic release layer 21 melts or vaporizes locally under the effect of the heat induced by the IR radiation, thus releasing the devices 100a assembled on the receiving substrate 2, and retaining other devices 100b on the transfer substrate 3 ([Fig. 14]). The transfer substrate 3 can then be moved to another location to perform another transfer, for example, for the remaining device(s) 100b.

[0089] As illustrated through the preceding examples, the structure and transfer method according to the invention therefore advantageously allows the formation of unitary semiconductor devices during the transfer, after removal of the donor substrate, and their localized transfer onto the receiving substrate.

[0090] The invention is not, however, limited to the embodiments previously described.

[0091] In particular, the number, shape and arrangement of semiconductor devices can be adapted according to the intended applications.

Claims

Demands

1. A method for transferring at least one semiconductor device (100) from a first substrate (1, 1b), called the donor, to a second substrate (2), called the receiver, using a silicon-based transfer substrate (3), said method comprising at least the following steps: • provide said donor substrate (1, 1b) comprising at least one active layer (10) intended to be integrated into said semiconductor device (100), at the level of a first face (201, 200), • provide said silicon-based transfer substrate (3), having a second face (202), • form a cleavage layer (20) on at least one of said first face (201, 200) and said second face (202), said cleavage layer (20) being inorganic, • assemble said donor substrate (1, 1b) with said transfer substrate (3), via said cleavage layer (20), • remove said donor substrate (1, 1b), retaining said at least one active layer (10) on the transfer substrate (3), and form the at least one semiconductor device (100) from said at least one active layer (10), • provide said receiving substrate (2), • assemble said at least one semiconductor device (100) with said receiving substrate (2), • illuminate, by infrared (IR) radiation, the cleavage layer (20) through said transfer substrate (3), so as to eliminate at least part of said cleavage layer (20), in order to decouple said at least one semiconductor device (100, 100a) from said transfer substrate (3), said process being characterized in that the formation of said cleavage layer (20) comprises the formation of a metallic release layer (21) on the side of said active layer (10), and the formation of a mineral absorption layer (22) on the side of said transfer substrate (3), such that said cleavage layer (20) comprises said metallic release layer (21) and said mineral absorption layer (22), said mineral absorption layer (22) being configured to absorb at least 20% of said infrared radiation during illumination, said metallic release layer (21) being configured to melt or vaporize locally upon illumination by infrared (IR) radiation.

2. A method according to the preceding claim wherein said at least one semiconductor device (100) comprises a plurality of semiconductor devices (100a, 100b) each having a characteristic dimension less than or equal to 10 pm, and wherein said illumination is configured to decouple only certain semiconductor devices (100a) from said plurality.

3. A method according to the preceding claim wherein said semiconductor devices of said plurality of semiconductor devices (100a, 100b) are separated from each other before illumination, said semiconductor devices remaining held together by said cleavage layer (20) on said transfer substrate (3).

4. A method according to any one of the preceding claims wherein said illumination is carried out by a pulsed IR laser producing IR radiation with an energy between 1 pJ and 10 pJ, for a duration between 1 picosecond (ps) and 20 ps.

5. A method according to any one of the preceding claims wherein said mineral absorption layer (22) has a resistivity between 50 p.ohm.cm (micro-ohm-centimeter) and 1 m.ohm.cm (milli-ohm-centimeter).

6. A method according to any one of the preceding claims wherein said mineral absorption layer (22) is selected from a refractory material having a melting temperature greater than or equal to 1500°C.

7. A method according to any one of the preceding claims wherein said mineral absorption layer (22) is selected from a refractory material based on a transition metal, preferably based on Ti, V, Zr, Ta, Hf, Nb, W.

8. A method according to any one of the preceding claims wherein said mineral absorption layer (22) is selected from transition metal-based refractory nitrides such as TiN, VN, ZrN, TaN, HfN, NbN, or their alloys.

9. A method according to any one of the preceding claims wherein said metallic release layer (21) has an infrared reflectivity greater than or equal to 97%.

10. A method according to any one of the preceding claims wherein said metallic release layer (21) has a resistivity between 2.5 p.ohm.cm (micro-ohm-centimeter) and 50 p.ohm.cm (micro-ohm-centimeter).

11. A method according to any one of the preceding claims wherein said metallic release layer (21) has a melting temperature less than or equal to 1100°C.

Images