Process for manufacturing a composite structure

Abstract

Process for manufacturing a composite structure with the following phases: a) providing a donor substrate (10) and a carrier substrate (30), the donor substrate having a first surface (20); b) forming a weakened zone (40) in the donor substrate (10), the weakened zone (40) defining a working layer (50) with the first surface (20) of the donor substrate (10); c) joining the carrier substrate (30) and the donor substrate (10); d) breaking the donor substrate along the weakened zone such that the working layer (50) is transferred to the carrier substrate (30); e) thinning the working layer (50) to form a thinned working layer (51), wherein the thinning consumes a non-uniform thickness of the working layer (50); wherein the manufacturing process is characterized in that phase b) comprises an implantation step with a non-uniform dose over the extent of the weakened zone, so that the working layer (50) transferred to the carrier substrate (30) exhibits a thickness profile after completion of phase d), wherein the thickness profile is suitable to at least partially compensate for the non-uniformity in the consumption of the working layer (50) during phase e), and wherein the thinned working layer (51) has a substantially uniform thickness after completion of all phases.

Description

FIELD OF THE INVENTIONThe present invention relates to a process for making a composite structure.BACKGROUND OF THE INVENTIONA process for producing a composite structure known from the prior art and illustrated in FIG. 1 comprises the following phases: a) providing a donor substrate 1 having a first surface 2 and a carrier substrate 3; b) forming a weakening zone 4 in the donor substrate 1, wherein the weakening zone 4 delimits a working layer 5 with the first surface 2 of the donor substrate 1; c) joining the carrier substrate 3 and the donor substrate 1; d) breaking the donor substrate along the weakening zone such that the working layer 5 is transferred to the carrier substrate 3; e) thinning the working layer 5 such that a thinned working layer 6 is obtained.The essential drawback associated with this prior art manufacturing process is that the thinned working layer 6 has non-uniformity of thickness.This is because thinning phase e) generally comprises a phase of partial oxidation of working layer 5, followed by removal of the oxidized part of working layer 5.The oxidation partially oxidizes the working layer 5 to an uneven thickness. This is reflected by a variation in the thickness of the thinned working layer 6 after the oxidized part of the working layer 5 is removed. The variation in thickness of the thinned working layer 6 may exceed 1 nm after completion of all phases, where 1 nm is 10 Å.This is particularly disadvantageous for the production of silicon-on-insulator structures for which control of the variation of the thickness of the silicon layer to + / -5 Å is required.Thus, controlling the thickness of the thinned working layer 6 after completion of all phases of the manufacturing process remains very difficult.It is thus an object of the invention to provide a process for the production of structures in which better control of the variation of the thickness of the thinned working layer 6 is possible.BRIEF DESCRIPTION OF THE INVENTIONThe present invention aims to overcome the above disadvantages and relates to a process for the production of a composite structure having the following phases: a) providing a donor substrate and a carrier substrate, the donor substrate having a first surface; b) forming a weakening zone in the donor substrate, the weakening zone delimiting a working layer with the first surface of the donor substrate; c) joining the carrier substrate and the donor substrate; d) breaking the donor substrate along the weakening zone such that the working layer is transferred to the carrier substrate; e) thinning the working layer such that a thinned working layer is formed, the thinning consuming an uneven thickness of the working layer; wherein in said manufacturing process it is to be noted that phase b) is carried out such that the working layer transferred to the supporting substrate has a thickness profile at the end of phase d), said thickness profile being suitable to at least partially compensate for the non-uniformity in the consumption of the working layer during phase e), said thinned working layer having substantially a uniform thickness at the completion of all phases.Thus, the uniformity of the thickness profile of the working layer at the completion of the phase d) makes it possible to compensate for the nonuniformity of the thinning phase. In particular, the uniformity of the thinned working layer at the termination of the phase e) is improved.According to one embodiment, thinning phase e) comprises a phase of oxidizing the working layer so as to form a layer of oxide, the thickness of said layer of oxide being non-uniform, followed by a phase of removing said layer of oxide.According to one embodiment, phase b) is carried out by implantation of at least one of the two species H or He.Thus, the implantation of at least one of the two types of hydrogen or helium makes it possible to form a weakened zone.According to an embodiment, the total dose of the implanted species is non-uniform over the extension of the zone of weakness, the non-uniformity of the dose of the implanted species being suitable for creating the thickness profile of the working layer at the end of the fracture d phase).Thus, the implantation conditions make it possible to form a zone of weakness and in particular to condition the thickness profile of the working layer at the end of phase d).This is because the applicant has unexpectedly found that the implantation of a total dose of species that is non-uniform over the entire extension of the zone of weakness makes it possible to adjust the variations in the thickness of the working layer at the completion of phase d). An excess thickness of a part of the working layer after phase d) is observed when said part is contained, upon completion of phase b), in a region of the zone of weakness having a dose of implanted species greater than in the rest of the zone of weakness.According to an embodiment, the dose of implanted hydrogen ions is uniform over the entire extent of the zone of weakness and the dose of implanted helium ions is non-uniform over the extent of the zone of weakness.Thus, the thickness profile of the working layer at the completion of phase d) is conditioned by the dose of implanted helium ions.According to one embodiment, the layer of oxide formed during phase e) exhibits a greater thickness in its central part and a lesser thickness towards the annular edge region of the layer of oxide, and the dose of implanted species is greater in the central part of the zone of weakness and decreases towards the annular edge of the zone of weakness.Therefore, the excess dose of implanted species in the central part of the zone of weakness makes it possible to obtain a rounded thickness profile of the working layer at the end of phase d). Therefore, the thickness profile of the thinned working layer is substantially constant over the entire extent of the thinned layer.According to one embodiment, the layer of oxide formed during phase e) presents a lesser thickness in its central part and a greater thickness towards the annular edge of the layer of oxide, and the dose of species implanted during phase b) is lower in the central part of the zone of weakness and increases towards the annular edge of the zone of weakness.Therefore, the excess dose of implanted species of the annular edge region of the weakening zone makes it possible to obtain a trough-shaped thickness profile of the working layer at the end of phase d). Therefore, the thickness profile of the thinned working layer is substantially constant over the entire extent of the thinned layer.According to one embodiment, phase b) is performed in two phases:a first implantation of species according to a first implantation energy, wherein the dose of the first implantation of species is non-uniform over the extension of the weakened zone.a second implantation of species corresponding to a second implantation energy lower than the first implantation energy, wherein the dose of the second implantation of species is non-uniform across the extension of the weakened zone.The second implantation energy is greater than 90% of the first implantation energy and the dose of the first implantation of species and the dose of the second implantation of species are complementary over the entire extent of the zone of weakness, the non-uniformity of the dose of the first implantation of species and the non-uniformity of the dose of the second implantation of species being suitable for generating the thickness profile of the working layer at the end of the phase of fracture d).According to an embodiment, the species implanted during the first implantation and the second implantation contain hydrogen ions.According to an embodiment, a dielectric layer is formed on the first surface of the donor substrate before phase b).According to one embodiment, the dielectric layer comprises at least one of the following materials: silicon oxide or silicon nitride.According to one embodiment, the donor substrate comprises at least one of the following materials: silicon, germanium or a silicon / germania alloy.BRIEF DESCRIPTION OF THE DRAWINGSOther characteristics and advantages will become apparent from the following description of the embodiments of a process for the production of a composite structure according to the invention, given by way of non-limiting example with reference to the accompanying drawings, in which:FIG. 1 is a schematic illustration of a process for the production of a composite structure according to known prior art techniques;FIGS. 2 and 3 are schematic representations of a process for producing a composite structure according to the invention;FIGS. 4 aand 4 bare schematic representations of an implantation phase and a breakage phase according to the invention;FIG. 5 illustrates a thickness profile of a working layer along a diameter, the layer being obtained after the breaking phase according to the invention;FIG. 6 shows a thickness profile of a working layer along a diameter, the layer being obtained after the breaking phase according to the invention.DETAILED DESCRIPTION OF EMBODIMENTS OF THE INVENTIONThe same reference numerals are used for identical elements or elements playing the same role in the different embodiments to simplify the description.FIGS. 2 and 3 schematically show a process for the production of a composite structure.To simplify the illustration, the respective thicknesses of the different layers are not shown to scale.Stage a) of the process for making the composite structure comprises providing a donor substrate 10 having a first surface 20, and a carrier substrate 30.The donor substrate 10 provided in the phase a) may include one of the materials selected from: silicon, germanium or silicon / germanium alloy.The carrier substrate 30 provided in phase a) may be constructed from the materials commonly used in microelectronics, optical, optoelectronic or photovoltaic industries.In particular, the carrier substrate 30 comprises at least one of the following materials selected from the following group: silicon, silicon carbide, silicon / germanium, glass, ceramic and a metal alloy.According to an advantageous embodiment, as shown in FIG. 3, a dielectric layer 53 may be formed on the first surface 20 of the donor substrate 10 between the phase a) and the phase b).The dielectric layer 53 may include at least one of silicon oxide or silicon nitride.For example, the dielectric layer 53 may be formed by thermal oxidation of a donor substrate 10, wherein the donor substrate 10 comprises silicon. The thermal oxidation may be carried out at a temperature between 800° C.-1100° C. in an oxidizing atmosphere. For example, thermal oxidation of a donor substrate 10 comprising silicon at a temperature of 800° C. in an oxygen atmosphere makes it possible to form a layer of silicon oxide. A low pressure chemical vapor deposition technique and a plasma enhanced chemical vapor deposition technique are also techniques of choice to form the dielectric layer 53 comprising silicon oxide or silicon nitride.Phase b) of the process of making the composite structure includes forming a zone of weakness 40 in the donor substrate 10. The working layer 50 is intended to be transferred to the carrier substrate 30. According to a particularly advantageous embodiment, the weakening zone 40 is formed by the implantation of at least one of the species selected from: hydrogen ions or helium ions. The formation of the weakened zone 40 is described in detail below in the description of the different embodiments.Phase b) is followed by a phase c) which comprises the joining together of the carrier substrate 30 and the donor substrate 10. The joining can be carried out, for example, by molecular bonding. According to a first embodiment illustrated in FIG. 2, the assembly phase may be carried out by bringing the first surface 20 of the donor substrate into contact with the support substrate 30. Alternatively, as illustrated in FIG. 3, when the dielectric layer 53 is formed on the first surface 20 of the donor substrate 10, the step of joining may be performed by bringing the dielectric layer 53 and the support substrate 30 into contact.The phase c) is followed by a phase d) of breaking the donor substrate 30 along the zone of weakness. At the end of the breaking phase, the working layer 50 is transferred to the carrier substrate 30. Advantageously, the dielectric layer 53 has been formed on the first surface 20 of the donor substrate 10, and at the end of the breaking phase d), the working layer 50 and the dielectric layer 53 are transferred to the support substrate 30. The step of breaking d) may be carried out by a heat treatment at a temperature between 300-500°C for a period of time from 30 minutes to 24 hours.The phase d) is followed by a phase e) of thinning the working layer 50. Thinning of the working layer 50 results in a thinned working layer 51.Thinning of the working layer 50 is advantageously performed by a sacrificial oxidation phase. This is a surface treatment aimed at creating a layer of oxide 52 on the surface and in the upper part of the working layer 50. The oxidation of a working layer comprising silicon is therefore generally carried out between 800-1000° C.In this process, particular water vapor (the term "wet oxidation" is used in this case) or molecular oxygen (the term "dry oxidation" is then used) can be used. The corresponding reactions are:The wet oxidation is faster and makes it possible to obtain an oxide of lower quality than that obtained by the dry oxidation.The oxidizing gas may further contain a few percent hydrochloric acid (HCl) to neutralize the metal ions that may occur therein.The layer of oxide 52 is formed by both the consumption of the working layer 50 and the oxygen contributed by the ambient gas.Depending on the thickness desired for the layer of oxide 52 and the oxidation temperature, the time for oxidation is generally between a few minutes and a few hours.The total thickness of the formed layer of oxide 52 is generally between 50 and 500 nm, typically between 100 and 400 nm.Thermal oxidation of the silicon is often carried out using furnaces containing one or more tubes into which the composite structures to be treated are loaded. For large diameter silicon-on-insulator type composite structures, oxidation is more uniform when carried out in vertical tube ovens in which the composite structures are placed horizontally, one below the other.However, it has been found that the thickness of the layer of oxide 52 is not uniform throughout its extent.Thus, for example, in the case of a working layer 50 comprising silicon, a layer of oxide 52 whose thickness in its central part is of the order of 1500 Å has a thickness in the order of 1485 Å at its annular edge part, and such a layer whose thickness in its central part is of the order of 500 Å has a thickness in the order of 485 Å at its annular edge part.Such variations in thickness can be observed, for example, using an ellipsometer.Removal of the layer of oxide 52 or "sacrificial deoxidization" is an etch that is generally performed according to a wet process sequence.In the case of removing a layer of silicon oxide, the means providing the thermal etch is based on hydrofluoric acid (HF). After removal of the layer of silicon oxide 52, a thinned working layer 51 made of silicon is obtained.It should be noted that under certain conditions it is also possible to obtain the opposite, i.e. a layer of oxide 52 thicker at its edge portion than at its centre (i.e. concave). However, this case is much less frequent than the previous case.The parameters capable of causing thicker oxidation in the middle, or vice versa at the edge, include, for example, the partial pressure and the flow rate of the different gases introduced during the oxidation, the optional temperature ramp during the oxidation (the cause of an optional temperature gradient between the edge and the middle of the wafers), or the position in the furnace.A description of the different embodiments of phase b) of the generation of the zone of weakness 40 will now be given. Phase b) is carried out so that the working layer 50 transferred to the supporting substrate 20 has a thickness profile at the completion of phase d), said thickness profile being suitable for at least partially compensating for the non-uniformity in the consumption of the working layer 50 during phase d), the thinned working layer 51 having substantially a uniform thickness at the completion of all phases.In particular, the thickness profile of the working layer 50 at the end of the fracture d phase) is suitable to compensate for the non-uniformity of a sacrificial oxidation phase performed in the e phase).The term "compensating for non-uniformity" is understood to mean generating a thickness profile of the working layer 50 such that upon completion of the thinning phase, the thinned working layer 51 exhibits a substantially constant thickness throughout its extent.First Embodiment of Phase b)According to a first embodiment, the weakening zone 40 is produced by implantation of at least one of the species selected from: hydrogen ions or helium ions. The total dose of implanted species is non-uniform throughout the extent of the zone of weakness 40, and the diversity of the dose of implanted species is suitable to produce the non-uniform thickness profile of the working layer 50 upon termination of the fracture d phase).The term "dose of implanted species" is understood to mean the amount of implanted species per unit of surface area of the zone of weakness 40. The dose of an implanted species is measured in atoms per cm 2.The implantation phase is advantageously carried out by a single-slice implantation device. Unlike wafer stack implantation tools, a single wafer implanter makes it possible to implant a non-uniform dose of a species throughout the extent of the weakening zone 40.The filing company has found that the thickness profile of the working layer 50 at the completion of the fracture phase depends on the uniformity of the dose of an implanted species throughout the extent of the zone of weakness 40.An excess thickness of a portion of the working layer 50 after phase d) is observed when this portion is contained, upon completion of phase b), in a region of the zone of weakness 40 that comprises a dose of the implanted species in excess with respect to the remainder of the zone of weakness 40. Therefore, an excessive dose of implanted species in a region of the zone of weakness 40 makes it possible to produce an excessive thickness of the working layer 50 upon completion of phase d) in said region of the zone of weakness 40.As is thus shown in FIG. numbers 4 aand 4 b, an area A of the weakened zone 40 has an excess dose of implanted species compared to the rest of the weakened zone 40. The part 54 of the working layer 50 comprised in said region A in phase b) will have a greater thickness than the rest of the working layer 50 upon completion of phase d).For example, FIG. 5 shows the thickness profile of a silicon layer that has been transferred to a carrier substrate. In this particular case, the implantation conditions are as follows:implanting hydrogen ions at an energy of about 20 keV and a dose of about 1 x 10 16 atoms / cm 2, which is constant over the entire extension of the weakening zone.implantation of helium ions at an energy of about 30 keV with an overdose of 3% over the central part over a diameter of the order of 100 mm.It is then found that, after the breaking phase d), the working layer 50 made of silicon has a greater thickness in the middle than at its annular edge zone. The thickness profile of the working layer is then described as a rounded profile.For example, if the layer of oxide 52 formed during thinning phase e) has a rounded thickness profile, it is advantageous to implant a larger dose of a species in the central portion of the zone of weakness (40) and a smaller dose of a species towards the annular edge portion of the zone of weakness (40). The dose of the implanted species in the central part of the weakened zone 40 can be between 2 and 9%, preferably between 3 and 6%, greater than the dose of the implanted species in the annular edge region of the weakened zone. Thus, the thinned working layer 51 will have improved uniformity of thickness as compared to the prior art. The term "improved uniformity in thickness" is intended to mean minor variations in thickness.On the other hand, if the layer of oxide 52 formed during thinning phase e) exhibits a trough-shaped thickness profile, it is advantageous to implant a larger dose of a species in the annular edge region of the weakening zone 40 and a smaller dose of the species in the central part of the weakening zone 40. The dose of implanted species in the annular edge region of the weakening zone 40 may be between 2 and 9%, preferably between 3 and 6%, greater than the dose of implanted species in the central part of the weakening zone 40. The term "improved uniformity of thickness" is understood to mean minor variations in thickness.It is particularly advantageous if the dose of hydrogen ions implanted into the weakening zone 40 is constant or constant over the entire extent of the weakening zone 40 and the dose of implanted helium ions is non-uniform over the extent of the weakening zone 40. Thus, the thickness profile of the working layer is conditioned by the non-uniformity of the dose of the implanted helium ions.Second Embodiment of Phase b)According to a second embodiment, phase b) is carried out in two phases:first implantation of a species corresponding to a first implantation energy, wherein the dose of the first implantation of the species is non-uniform over the extension of the weakening zone 40.second implantation of a species corresponding to a second implantation energy lower than the first implantation energy, the dose of the second implantation of a species being non-uniform across the extension of the weakening zone 40.The second implantation energy is greater than 90% of the first implantation energy. The dose of the first implantation of a species and the dose of the second implantation of a species are complementary over the entire extent of the zone of weakness 40. The dose non-uniformity of the first implant of a species and the dose non-uniformity of the second implant of a species are suitable for creating the thickness profile of the working layer 50 after completion of the fracture phase d).The dose of the implanted species is measured in atoms per cm 2.The implantation phase is advantageously carried out by means of a single-slice implantation device. Unlike wafer stack implanters, a single wafer implanter makes it possible to implant a non-uniform dose of a species throughout the extent of the weakening zone 40.For example, the first implantation is an implantation of hydrogen ions having an implantation energy equal to 24.07 keV. The dose of hydrogen ions implanted during the first implantation is equal to about 1×10 16 atoms / cm 2 in a central part of the weakening zone 40 having a diameter of 200 mm, and is about 0.5×10 16 to 0.6×10 16 atoms / cm 2 in the rest of the weakening zone 40. The dose of hydrogen ions during the second implantation is 0.5×10 16 to 0.6×10 16atoms / cm 2 in a central part of the weakening zone 40 having a diameter of 200 mm, and is about 1×10 16 atoms / cm 2 in the rest of the weakening zone 40. The thickness profile of the working layer is then described as a trough-shaped profile.In a particularly advantageous manner, the non-uniformity of the thinning phase e) can be determined before the implementation of the process according to the invention. For a thinning phase e) carried out by sacrificial oxidation, this may consist of thermally oxidizing a donor substrate 10 or a working layer 50 and measuring the thickness profile generated therewith in a means for measuring the thickness of layers, for example an ellipsometer. Knowing the thickness profile of the layer of oxide produced during the thinning phase then makes it possible to adjust the phase of forming the zone of weakness for mass production of the composite structures.More specifically, furnaces for thermal oxidation nowadays exhibit a relatively high thermal stability. Consequently, a phase of thermal oxidation repeated on different occasions will produce layers of oxide always exhibiting substantially the same thickness profile. The process according to the present invention is then advantageously used for the production of composite structures such as a silicon-on-insulator.Furthermore, the process according to the invention makes it possible to produce composite structures, and in particular silicon-on-insulator structures, for which the thinned working layer 51 has smaller variations in thickness compared to the prior art.

Claims

A process for making a composite structure comprising the steps of: a) providing a donor substrate (10) and a carrier substrate (30), the donor substrate having a first surface (20); b) forming a zone of weakness (40) in the donor substrate (10), the zone of weakness (40) defining a working layer (50) with the first surface (20) of the donor substrate (10); c) joining the carrier substrate (30) and the donor substrate (10); d) breaking the donor substrate along the zone of weakness such that the working layer (50) is transferred to the carrier substrate (30); e) thinning the working layer (50) such that a thinned working layer (51) is formed, the thinning consuming a non-uniform thickness of the working layer (50); wherein the manufacturing process is characterized in that phase b) comprises an implantation step with a non-uniform dose over the extension of the zone of weakness such that the working layer (50) transferred to the carrier substrate (30) exhibits a thickness profile after completion of phase d), wherein the thickness profile is suitable to at least partially compensate for the non-uniformity in the consumption of the working layer (50) during phase e), and wherein the thinned working layer (51) exhibits a substantially uniform thickness after completion of all phases.Manufacturing process according to Claim 1, in which thinning comprises a phase of oxidation of the working layer (50) so as to form a layer of oxide (52), the thickness of the layer of oxide (52) being non-uniform, followed by a phase of removal of the layer of oxide (52).Production process according to claim 2, in which phase b) is carried out by implantation of at least one of the two species H or He.The manufacturing process according to claim 3, wherein the total dose of the implanted species is non-uniform throughout the extension of the zone of weakness (40), and wherein the non-uniformity of the dose of the implanted species is suitable to generate the thickness profile of the working layer (50) after the completion of the fracture d phase).The manufacturing process according to claim 4, wherein step b) is performed by implanting hydrogen ions and helium ions, and wherein the dose of implanted hydrogen ions is uniform over the entire extension of the weakening zone (40) and the dose of implanted helium ions is non-uniform over the extension of the weakening zone (40).Manufacturing process according to Claim 4 or 5, in which the layer of oxide (52) formed during phase e) has a greater thickness in its central part and a lesser thickness towards the annular edge region of the layer of oxide (52), and in which the dose of implanted species is greater in the central part of the zone of weakness and decreases towards the annular edge region of the zone of weakness.Manufacturing process according to Claim 4 or 5, in which the layer of oxide (52) formed during phase e) has a smaller thickness in its central part and a greater thickness towards the annular edge region of the layer of oxide (52), and in which the dose of a species implanted during phase b) is smaller in the central part of the zone of weakness and greater towards the annular edge region of the zone of weakness.Manufacturing process according to claim 3, in which phase b) is carried out in two phases: - a first implantation of one of the species according to a first implantation energy, the dose of the first implantation of one of the species being non-uniform over the extension of the weakening zone (40), - a second implantation of one of the species according to a second implantation energy less than the first implantation energy, the dose of the second implantation of one of the species being non-uniform over the extension of the weakening zone (40), the second implantation energy being greater than 90% of the first implantation energy, wherein the dose of the first implantation of one of the species and the dose of the second implantation of one of the species are complementary over the entire extent of the zone of weakness (40), and wherein the non-uniformity of the dose of the first implantation of one of the species and the non-uniformity of the dose of the second implantation of one of the species are suitable for generating the thickness profile of the working layer (50) after completion of the phase of fracture d).The manufacturing process of claim 8, wherein the species implanted during the first implantation and the second implantation comprises hydrogen ions.The manufacturing process according to any one of claims 1 to 9, wherein a dielectric layer (53) is formed on the first surface of the donor substrate before phase b).The manufacturing process according to claim 9, wherein the dielectric layer (53) comprises at least one of silicon oxide or silicon nitride.Production process according to one of Claims 1 to 11, in which the donor substrate (10) comprises at least one of the following materials: silicon, germanium or a silicon / germanium alloy.

Images