JP4846363B2 - Reuse of wafer with buffer layer after thin layer removal - Google Patents

Abstract

Recycling a donor wafer after having taken at least one useful layer of material chosen from semiconductor materials, comprises removal of substance on side of wafer where taking-off took place. Substance is removed by mechanical means so that, after removal, at least part of buffer structure remains, this part capable of being reused as buffer structure during subsequent useful layer taking-off. Recycling a donor wafer (10) after having taken at least one useful layer of a material chosen from semiconductor materials, the donor wafer comprising successively a substrate (1), a buffer structure and, before taking-off, a useful layer, comprises removal of substance on the side of the donor wafer where the taking-off took place. The removal of substance comprises employing mechanical means so that, after removal of substance, at least part of the buffer structure remains, this at least part of the buffer structure capable of being reused as a buffer structure during a subsequent useful layer taking-off. Independent claims are also included for the following: (a) production of a donor wafer intended to provide a useful layer by taking-off and capable of being recycled after taking-off, comprising: formation of a first part of a buffer structure on a substrate; formation of a protective layer on the first part of the buffer structure, in a material chosen from crystalline materials; formation on the protective layer of the second part of the buffer structure, such that it has a lattice parameter in the vicinity of the protective layer the same as that of the first part of the buffer structure in the vicinity of the protective layer; (b) taking off a useful layer on a donor wafer to be transferred to a receiving substrate, comprising: bonding the donor wafer to the receiving substrate; detaching a useful layer bonded to the receiving substrate from the donor wafer; and recycling the donor wafer complying with the method of recycling; (c) application of the method of cyclically taking-off for producing a structure comprising the receiving substrate and the useful layer, the useful layer comprising at least one of the following materials: silicon-germanium (SiGe), strained Si, Ge, an alloy belonging to the IU-V family, the composition of which is respectively chosen from the possible (aluminum, gallium, indium)-(nitrogen-phosphorus-arsenic) (Al,Ga,In)-(N,P,As) combinations; and (d) a donor wafer having supplied a useful layer by taking-off and capable of being recycled complying with the method of recycling, comprising a substrate and a remaining part of the buffer structure.

Description

  The present invention relates to the reuse of a donor wafer after a thin semiconductor layer is transferred from a donor wafer having a buffer layer to a receiving substrate.

  The term “buffer layer” usually refers to a transition layer, which describes a first crystal structure, such as a substrate, and material properties such as structural or stoichiometric properties, or atomic surface recombination properties. It exists between the second crystal structure mainly having the function to change.

  In the special case of a buffer layer, the buffer layer provides a second crystal structure with a lattice constant that is significantly different from the lattice constant of the substrate.

  For this purpose, the buffer layer can have a composition that varies gradually with thickness, and the gradual change in the composition of the buffer layer is related to the gradual change in its lattice constant.

  That is, the composition may change at a certain rate of change, the sign of the rate of change may be reversed, or the composition may change suddenly, or a constant composition layer containing defects may be provided.

  Described here is a modified (buffer) layer of a modified form such as modified epitaxy.

  To form a particular structure, a layer or stack formed on the buffer layer is peeled from the donor wafer for transfer to the receiving substrate.

  One of the main uses for transferring thin layers formed on a buffer layer relates to the formation of a strained silicon layer.

  If the lattice constant of the material in the plane of the interface is greater than its nominal lattice constant, a layer of “distorted” material is formed by stretching, and if the former is smaller than the latter, a layer of “distorted” material is formed by compression. .

  Also, if it is very close to the nominal lattice constant of the “relaxed” material, a layer of “relaxed” material is formed, which is the lattice constant of the material in equilibrium in the bulk state.

  When a strained silicon layer is formed by stretching, some properties, such as the electron mobility of the material, are clearly improved.

  Other materials, such as SiGe, can be stripped in substantially the same way.

Transferring such layers onto the receiving substrate is known to those skilled in the art, particularly in a process called Smart-cut ^(c), and can form structures such as SOI (Semiconductor On Insulator).

  For example, when a layer is peeled from relaxed SiGe, the resulting structure functions as a support for growing silicon.

  Since the nominal lattice constant of SiGe (depending on the germanium content) is larger than the nominal lattice constant of silicon, growing silicon on an SGOI (Silicon-Germanium On Insulator) pseudo-substrate forms a strained silicon layer due to stretching. .

  An example of such a process is L.L. J. et al. Huang et al. IBM's document ("SiGe-On-Insulator pre-prepared by wafer bonding and layer transfer for high-performance field-effect transformers No. 1, Applied Phys. 9 / Applied Physics. Here, a method for manufacturing a Si / SGOI structure is introduced.

  Another example of such a process is described in document US2002 / 007481.

  Modified growth can be applied to others, and in particular to III-V semiconductors.

  Therefore, the transistor is generally manufactured by a technology based on GaAs or InP.

  Regarding the behavior of electrons, InP is much superior to GaAs. In particular, when an InP layer and an InGaAs or InAlAs layer are combined, the electron mobility is improved.

  However, in terms of putting elements using InP technology on the market, the cost, versatility, mechanical brittleness, and bulk substrate size (the maximum diameter of InP is usually There is a limit to 4 inches) compared to 6 inches of GaAs.

  A solution to this problem is likely to be in the receiving substrate, the InP layer that is stripped and obtained by modified epitaxy of the buffer layer on the GaAs substrate.

  In a stripping process such as etch back, the remaining portion of the substrate and the buffer layer may be destroyed during stripping.

In other stripping processes such as Smart-cut ^(c) , the substrate is reused but the buffer layer is lost.

  However, degenerative manufacturing techniques are complex.

  Therefore, optimizing and manufacturing such a buffer layer is time consuming and includes difficult and costly processes.

  Furthermore, the internal strain due to the composition change easily causes crystal defects such as dislocations and point defects.

  These internal strains, and hence crystal defects, can be made difficult to occur, particularly by increasing the thickness at which the lattice constant changes.

  This is the main reason why the buffer layer is usually thick, typically 1 to a few micrometers.

  However, due to economic and technical regulations, important characteristics of the buffer layer, such as thickness and some structural complexity, are degraded.

  In particular, for these reasons, it may be better to always avoid completely forming the buffer layer before reusing the substrate.

  In a first aspect, the present invention is a method for reusing a donor wafer after peeling off at least one useful layer of a material selected from a plurality of semiconductor materials, in order to achieve this object, The donor wafer comprises a substrate, a buffer structure, and a useful layer prior to stripping, and the method comprises removing material on the side of the donor wafer where the stripping has occurred, and after removing the material, At least a part of the buffer structure remains, and at least a part of the buffer structure can be reused as a buffer structure for peeling off the next useful layer.

  In a preferred aspect of the recycling method of the present invention, a protective layer is further present in the donor wafer, at least a part of the buffer structure is present under the protective layer, and an etching power for the material of the protective layer of the substance removing means is high. The material of the protective layer is selected from a plurality of crystalline materials so that the material can be selectively removed, which is substantially different from the etching power for at least one material in two neighboring regions.

  According to a second aspect, the present invention provides a method for producing a donor wafer that provides a useful layer by stripping and is reusable after stripping by the preferred recycling method, wherein the first part of the buffer structure on the substrate. And a protective layer is formed on the first part of the buffer structure by a material selected from a plurality of crystal materials, and the lattice constant of the second part in the vicinity of the protective layer is in the vicinity of the protective layer. Forming a second portion of the buffer structure on the protective layer so as to be substantially the same as a lattice constant of the first portion of the buffer structure; A manufacturing method is provided.

  According to a third aspect, the present invention is a method for stripping a useful layer on a donor wafer for transfer to a receiving substrate, wherein the donor wafer is bonded to the receiving substrate and bonded from the donor wafer to the receiving substrate. There is provided a method for peeling off a useful layer, comprising the step of peeling off the useful layer and reusing the donor wafer according to the recycling method.

  According to a fourth aspect, the present invention provides a method for reusing a useful layer from a donor wafer, comprising a plurality of steps of peeling the useful layer according to the above-described peeling method. .

  According to a fifth aspect, the present invention provides a SiGe, Si, III-V alloy, and (Al, Ga, In)-for producing a structure comprising the receiving substrate and the useful layer. The reusable peeling method or an application of the peeling method is provided, wherein the useful layer comprises at least one material selected from possible combinations of (N, P, As).

  According to a sixth aspect, the present invention provides each donor wafer according to each method of the present invention.

  Other aspects and advantages of the invention will become apparent upon reading the following detailed description of preferred method implementations, given non-limiting examples, with reference to the accompanying drawings.

  The main object of this invention is to reuse a wafer after at least one useful layer has been peeled from a wafer with a buffer structure (ie, any structure that acts as a buffer layer) to integrate the useful layer into a semiconductor structure. This reuse includes at least partially restoring the buffer structure so that it can be used again for subsequent stripping.

  Therefore, the reuse process must include an appropriate process that does not damage at least a part of the buffer structure.

  In fact, the buffer structure usually contains crystal structure defects such as dislocations and, importantly, when energy is applied to it, it propagates and increases in size, which energy can be heat treated, chemical processes or It can be brought about by a mechanical process.

  For example, when a SiGe buffer structure is heated at a temperature of 350 ° C., 450 ° C. or 550 ° C., the state of the structure changes depending on the selected temperature (eg, Journal of Crystal Growth, vol. By Re et al. 227-228, pp. 749-755, document “Structural charac- terization and stability of Sil-xGex / Si (100) heterostructures grown-by molecular beam epitaxy” in July 2001. As the temperature increases, the buffer structure tends to decrease its internal stress, which is caused by structural relaxation on the sliding surface, stacking faults, and other types of structural relaxation. This can cause any problems at the interface with the useful layer being formed. Therefore, it is important to keep these internal stresses in the buffer structure.

  Therefore, reuse should be adapted to this in order to prevent and control the spread of these crystal stresses in the buffer structure which degrades the properties of the buffer structure and the useful layers formed thereon. Must be used.

  Effectively, the buffer structure has a crystalline structure that is substantially relaxed and / or free of a significant number of structural defects on the surface.

  The “buffer layer” is as already generally described in this document.

Effectively, the buffer layer is provided in the buffer structure and has at least one of the following two functions.
1. Reduce the defect density of the upper layer,
2. Match the different lattice constants of the two crystal structures.

  With regard to the second function of the buffer layer, the buffer layer is an intermediate layer between the two structures, has a first lattice constant approximately equal to the lattice constant of the first structure around one of its surfaces, and the other surface Has a second lattice constant approximately equal to the lattice constant of the second structure.

  In the remainder of this document, the buffer layer or buffer structure will disclose the above function as this latter buffer layer.

  However, the invention relates to any buffer layer or buffer structure that is more general in this document.

  Further disclosed below is an example of the method of the present invention involving the reuse of a donor wafer consisting initially of a support substrate and a buffer structure, where the useful layer is stripped.

  Referring to FIG. 1, a donor wafer (a thin layer donor to be peeled) 10 included in the prior art includes a support substrate 1 and a buffer structure I.

  As an application of the donor wafer 10 in the present invention, the useful layer is peeled off from the portion 4 of the buffer structure I and / or from the upper layer (not shown in FIG. 1) formed on the surface of the buffer structure I. Is performed to integrate the structure into a structure such as an SOI structure.

  The support substrate 1 of the donor wafer 10 includes at least one semiconductor layer having a first lattice constant at the interface with the buffer structure I.

  In one particular embodiment, the support substrate 1 is made of a single semiconductor having a first lattice constant.

  In the first embodiment of the buffer structure I, the buffer structure I consists of the buffer layer 2.

  The buffer layer 2 located on the support substrate 1 can in this case have a second lattice constant which differs greatly from the first lattice constant of the support substrate 1 on its surface, and therefore in the same donor wafer 10. The two layers 1 and 4 have different lattice constants.

  Furthermore, in certain applications, the buffer layer 2 can prevent the upper layer from being exposed to high defect density and / or high stress.

  Further, in some applications, the buffer layer 2 can improve the surface state of the upper layer.

  Usually, the buffer layer 2 has a lattice constant that gradually changes according to the thickness so as to transition between two lattice constants.

  Such a layer is usually referred to as a modified layer.

  This gradual change of the lattice constant is continuously performed within the thickness of the buffer layer 2.

  Alternatively, it can be done "in stages", each stage being a thin layer with an almost constant lattice constant that is different from the lattice constant of the lower stage, and changing the lattice constant discontinuously at each stage. .

  It can also be a more complex aspect, such as a change in composition with a change rate, a reversal of the sign of the change rate, or a discontinuous sudden change in composition.

  The change in the lattice constant in the buffer layer 2 appears by gradually increasing the concentration of at least one atom not included in the substrate 1 starting from the substrate 1.

  Thus, for example, the buffer layer 2 formed on the substrate 1 made of a single material can be formed of two, three, four or more types of materials.

  Thus, for example, the buffer layer 2 formed on the substrate 1 made of a two-element material can be made of three, four or more element materials.

  The buffer layer 2 can be effectively grown on the support substrate 1 by epitaxy or the like using known techniques such as CVD (Chemical Vapor Deposition) and MBE (Molecular Beam Epitaxy).

  Usually, for example, in order to obtain the buffer layer 2 made of an alloy of various atoms, the buffer layer 2 can be formed by other known methods.

  The incidental process for the surface treatment of the substrate 1 under the buffer layer 2 can be performed, for example, by CMP polishing before the buffer layer 2 is formed.

  In the second embodiment of the buffer structure I, referring to FIG. 1, the buffer structure I comprises a buffer layer 2 (substantially the same as the buffer layer 2 of the first embodiment) and an additional layer 4.

  The additional layer 4 may be provided between the substrate 1 and the buffer layer 1 or on the buffer layer 1 as seen in FIG.

  In the first special case, this additional layer 4 comprises a second buffer layer, such as a buffer layer, which contains defects and thus can improve the crystal quality of the layer formed on the buffer structure I. May be.

  The additional layer 4 is preferably formed of a semiconductor having a constant material composition.

  The selection of the composition and thickness of the buffer layer 4 so formed is a particularly important criterion for obtaining this property.

  Thus, for example, structural defects in an epitaxially grown layer usually decrease gradually within the thickness of this layer.

  In the second special case, the additional layer 4 is located on the buffer structure I and functions as the upper layer of the buffer layer 2.

  Therefore, it determines the second lattice constant.

  In the third special case, the additional layer 4 is located on the buffer layer 1 and plays a role in the peeling performed in the donor wafer 10, for example at this level.

  The additional layer may further have various functions such as a function selected from these three cases.

  In an advantageous embodiment, the additional layer 4 is located on the buffer layer 2 and has a second lattice constant different from the first lattice constant of the support substrate 1.

  In certain special cases of this latter embodiment, the additional layer 4 is made of a material relaxed by the buffer layer 2 and has a second lattice constant.

  The additional layer 4 can be formed by being effectively grown on the buffer layer 2 by, for example, epitaxial growth by CVD or MBE.

  In the first embodiment, the growth of the additional layer 4 is performed in situ immediately after the formation of the lower buffer layer 2, and in this case, the formation of the buffer layer 2 is also effectively performed by the layer growth.

  In the second embodiment, the growth of the additional layer 4 is performed after an additional step for the surface treatment of the lower buffer layer 2, and this surface treatment is performed by, for example, CMP polishing, heat treatment or other planarization technique. Any slip planes, stacking faults, and so on, so that dislocations and other defects contained in the buffer layer 2 are not propagated, the size is not increased, and the quality of the final buffer structure I formed is deteriorated. In order to prevent the formation of other defects.

Peeling of the useful layer from the donor wafer 10 is performed according to one of the following main modes.
(1) The useful layer to be peeled is a part of the additional layer 4.
(2) The useful layer to be peeled off is a part of the upper layer (not shown in FIG. 1), and this upper layer is already formed on the buffer structure I by epitaxial growth after surface treatment of the buffer structure I, for example. It may be what you have.

  The donor wafer 10 thus functions as a substrate for upper layer growth.

  The upper layer may consist of one or more thin layers, depending on the mode of stripping used.

  Furthermore, it effectively has a lattice constant approximately equal to the lattice constant of the relaxed material of the free surface of the buffer structure I, for example the crystalline structure of the same material, or all or part of it is stretched Or a layer of other materials that are distorted by compression or a combination of these two materials.

  In some embodiments of the donor wafer 10, one or more intermediate layers are further inserted between the buffer structure I and the top layer. In this case, this or these intermediate layers are not removed.

  (3) The useful layer to be peeled is a part of the additional layer 4 and the upper layer (formed in substantially the same manner as described in the second peeling mode).

  Regardless of which stripping mode is selected and referring to FIG. 2, after stripping, and in many cases, protruding portions 7a and / or irregular portions 7b appear on the stripping surface of the remaining donor wafer 10.

  This “relieved” release surface belongs to the post-release layer 7 located above the buffer layer 2.

  Depending on the post-release mode selected from the three post-release modes already mentioned, the post-peeling layer 7 is constituted by one or more intermediate layers and all or some part of the layer 4 such as a part of the upper layer. Is done.

  The raised portions 7a and 7b on the surface of the post-peeling layer 7 mainly appear depending on the technique used between the post-peeling mode and the post-peeling.

  Therefore, for example, the post-peeling mode that is currently used in the industry is that the useful layer is peeled from only a part of the surface of the donor wafer 10 (usually almost the central part) instead of the entire surface of the donor wafer 10. Leave a protruding part as quoted in 7a above. These protruding portions usually exist integrally around the surface of the donor wafer 10, and all the protruding portions are known as “peeling rings” in the industry.

Thus, for example, known release techniques that will be discussed further later in this document, such as the previously described Smart-cut ^(c) , may result in a roughened surface as quoted at 7b.

  Once stripped, the invention is reused to restore the donor wafer 10.

  Recycling typically consists of two steps that remove material and restore at least a portion of donor wafer 10.

  The first step of recycling of the present invention consists of removing at least the relief portions 7a, 7b (seen in FIG. 2).

  The removal of the material of the invention is carried out such that at least a part of the buffer structure I that can be used again during the subsequent stripping of the new useful layer remains after removal.

  Therefore, unlike the known reuse in the prior art, the remaining portion of the buffer structure I after the material removal is reused.

  In the first special case of recycling, and with respect to the second exfoliation mode (2), so that no material is removed from the intact buffer structure I, so that the entire buffer structure I is retained. The thickness of the upper layer can be selected so that after the peeling, the remaining part of the upper layer (this is the layer 7 after peeling) is removed by standard mechanical means for removing substances such as polishing means or CMP. It is effective.

  The thickness of material removed during reuse by standard mechanical means such as polishing is typically about 2 micrometers, even though the thickness currently reaches as high as 1 micrometer.

  In a second special case of recycling, and with respect to the second exfoliation mode (2), so that no material is removed from the intact buffer layer 2 and thus the entire buffer layer 2 is retained. After peeling, the upper layer and the upper layer so that the remaining part of the upper layer (which is the post-peeling layer 7) and at least part of the additional layer 4 are removed by polishing means or standard mechanical means for material removal such as CMP. It is effective to select the thickness of the additional layer 4. In other special cases of reuse, the removal of the substance effectively uses a means for chemically attacking the material, such as chemical etching.

  The etching may be purely chemical, electromechanical, photoelectrochemical, or other equivalent etching, such as etching performed during chemical mechanical polishing.

  One effective etching mode is selective etching.

  Thus, it is possible in particular to use an etching fluid (ie gas or solution) suitable for selectively etching the material removed from the material to be reused, and these two materials are in adjacent layers. The material to be reused forms an etch stop layer and strips the portion to be effectively removed while protecting the layer to be reused from chemical etching.

The selective property between the two materials is obtained, for example, by at least one of the following cases.
The two materials are different, or
The two materials contain approximately equivalent atoms, except at least one atom, or
The two materials are approximately equivalent, but at least one atom of one material has an atomic concentration greatly different from the atomic concentration of the same atom of the other material, or
The two materials have different porosity densities.

For example, Si is etched with a solution containing a compound such as KOH (potassium hydride, selection ratio: about 1: 100), NH ₄ OH (ammonium hydroxide, selection ratio: about 1: 100) or TMAH (tetramethyl ammonium hydroxide). In doing so, it is known that SiGe functions as a stop layer.

  For example, if the germanium concentration of SiGe is higher than or equal to 25%, etching a germanium concentration lower than or equal to 20% with a solution containing a compound such as TMAH will function as a stop layer. It has been.

For example, a solution containing a compound such as EDP (ethylene diamine pyrocatechol), KOH or N ₂ H ₂ (hydrazine) suitably doped with Si with a selected concentration of doping atoms such as boron higher than 2 × 10 ¹⁹ cm ⁻³ It is known that when undoped Si material is etched using, it functions as a stop layer.

For example, it is known that porous Si is etched selectively with respect to non-porous crystalline Si using a solution containing a compound such as KOH or HF + H ₂ O ₂ .

  Therefore, the additional layer 4 can be selectively etched with respect to the buffer layer and / or if there is an upper layer, the upper layer can be etched with respect to the additional layer 4 or the intermediate layer (if present).

  Removal of the material by chemical means may use mechanical means or other means to attack the substrate.

  In particular, it is also possible to perform CMP polishing using a selective chemical etching solution.

  This chemical etching may be performed after or before removal of the material by mechanical means that corrodes the material, such as by polishing, grinding, attack or other means.

  Typically, the removal of the material is any additional means of attacking the material that can remove the material without completely exfoliating the buffer structure I and without damaging at least a portion of the buffer structure I. It may be used.

Accordingly, one of the following material removal modes is used.
(A) removing a part of the post-peeling layer 7 provided with at least the relief parts 7a, 7b, or
(B) Remove the entire layer 7 after peeling, or
(C) The entire layer 7 after peeling and a part of the buffer layer 2 are removed.

  In the case where the post-peeling layer 7 includes a part of the original upper layer, it is preferable that the upper layer portion is completely peeled off in the material removal mode (a).

  Referring to FIG. 3, a portion of the original buffer structure remaining after material removal is quoted with I '.

This consists of:
If the material removal mode (a) is used and no part of the additional layer 4 is stripped in this mode, the entire original buffer structure I.
Alternatively, when the material removal mode (a) is used and a part of the additional layer 4 is peeled off in this mode, the buffer layer 2 and a part of the additional layer 4.
Or the buffer layer 2 when the substance removal mode (b) is used.
Or a part of the buffer layer 2 when the material removal mode (c) is used.

  In the second reuse step after the first reuse step involved in material removal, at least some of the multiple layers that were peeled off during the first step are reformed.

  First, in some cases, it is preferable to finish the surface of the donor wafer 10 from which material has been removed during the first recycling step to remove any irregularities that may have appeared during material removal.

  For this purpose, for example, CMP polishing, heat treatment or other planarization techniques are used to prevent the dislocations and other defects contained in the buffer structure I from propagating and to increase in size and to the buffer structure. Prevent any slip planes, stacking faults or other defects from degrading the quality of I.

  When a part of the original buffer structure I is removed during the first reuse process, the buffer structure I is restored from the buffer structure I ′ remaining in the second reuse process.

  In restoring the buffer structure I, once formed, it is effective if the restored is substantially the same as the original buffer structure I.

  However, in some embodiments, some manufacturing parameters may be altered slightly to obtain a slightly different buffer structure I. For example, the concentration of certain compounds in the material is slightly changed.

  When a part of the original buffer layer 2 is cut off during the first reuse process, the removed part of the buffer layer 2 is re-formed in the restoration of the buffer structure I.

  If all or part of the original additional layer 4 is cut off during the first reuse step, all or part of the additional layer 4 is reconstructed in the restoration of the buffer structure I.

  In this case, it is possible to form the additional layer 4 having a thickness substantially the same as or substantially different from the original.

  Once the buffer structure I is restored, an upper layer can be formed thereon, the upper layer at least partially comprising a new useful layer to be stripped, and further comprising the buffer structure I and the upper layer. One or more intermediate layers can be provided in between.

  The multiple layers that can be formed during this second recycling step are effectively grown on each lower layer by, for example, CVD or MBE epitaxial growth.

  In the first case, if at least one of these layers is grown there immediately following the formation of the lower growth support member, the latter is also effectively formed by layer growth in this case.

  In the second case, at least one of these layers is grown after a sub-step of finishing the surface of the lower growth support member, this finish being used, for example, by CMP polishing, heat treatment or other planarization techniques, In order to prevent dislocations and other defects contained in the buffer structure I from propagating, so as not to increase in size, and to prevent any slip planes, stacking faults and other defects from deteriorating the quality of the buffer structure I. To do.

  A donor wafer 10 is ultimately obtained that is substantially the same as the original donor wafer, ie, the donor wafer 10 shown in FIG.

  The donor wafer 10 obtained in this way comprises at least a part of the original buffer structure I and thus at least a part of the original buffer layer 2, so that time, as in the known recycling method, is complete, It is possible to avoid costly and costly reformation.

  A donor wafer 10 that can be reused by a special processing mode of the above recycling method is described in the remainder of this document, and effectively protects at least a portion of the buffer structure I, especially during a compatible reuse.

  Each of the donor wafers 10 seen in FIGS. 4, 5 and 6 includes a substrate 1 and a buffer structure I, similar to the donor wafer 10 seen in FIG.

  Each of these donor wafers 10 further includes a protective layer 3 located in a portion located on the same side of the buffer structure I, and the buffer structure I serves as an interface with the substrate 1 of the protective layer 3.

  The protective layer 3 defined by the present invention is formed of a material selected from a crystalline material such as a semiconductor, and has a main function of protecting a part of the donor wafer 10 below the protective layer 3 and is reused during reuse. At least a part of the buffer structure I is formed during at least one of the multiple substance removal processes to be performed.

  The protective layer 3 is effective when a layer is grown on the lower growth support member by, for example, CVD or MBE epitaxial growth.

  In this embodiment and in the first case, the protective layer 3 is grown in situ immediately after the formation of the lower layer. In this case, the lower layer is also effectively formed by the layer growth.

  In the second case, the protective layer 3 is grown after a sub-process that finishes the surface of its lower layer, and this finish is included in the buffer structure I using, for example, CMP polishing, heat treatment or other planarization techniques. To prevent dislocations and other defects from propagating, to prevent the size from increasing, and to prevent the formation of any slip surfaces, stacking faults or other defects that degrade the quality of the buffer structure I.

  There is at least one means for removing a substance having a function of attacking the material forming the protective layer 3, and a material substantially different from at least one material in two regions in the vicinity of the protective layer 3 is used as the protective layer 3. Choose as material.

  Therefore, there is a function of selectively removing the substance.

  The selective material removal in the protective layer 3 is performed by at least one of the following selective material removal modes.

  In the region near the protective layer 3 and with respect to the protective layer 3, the material located on the side of the already removed useful layer is selectively removed, and the protective layer 3 forms a layer that stops substance removal. .

  The material of the protective layer 3 is selectively removed, and a layer in the vicinity of the protective layer 3 and on the same side as the protective layer 3 of the substrate 1 is formed to stop the substance removal.

  Furthermore, in a special process with selective material removal, two selective material removal modes for the same protective layer 3 are combined in succession.

  Therefore, the layer on the protective layer 3 and the protective layer 3 are selectively removed.

  Which selective material removal mode is selected as a process during the first recycling step and stops material removal even for removing a portion of the donor wafer located on the side of the removed useful layer Layer (a protective layer 3 in the case of the first selective substance removal, and a region located on the same side as the protective layer 3 of the substrate 1 in the vicinity of the protective layer 3 in the case of the second selective substance removal). Exists.

  Accordingly, the stop layer functioning as a barrier for the substance attack similarly protects the material of a part of the lower part of the protective layer 3 (at least constituting part of the buffer structure I).

  In some cases, it is desirable that the protective layer 3 does not substantially hinder the crystal structure of the neighboring layer, and in particular, does not hinder crystal growth of the lower layer to be formed. In many cases, substantially corresponds to the lattice constant of the lower portion of the protective layer 3.

  This last point is particularly important when the protective layer 3 is located in the buffer structure I (shown in FIG. 4).

  This result is obtained by several embodiments of the protective layer 3 described below.

  In the first embodiment of the protective layer 3, the protective layer 3 is limited to those having substantially the same lattice constant as the lattice constants of the two regions in the vicinity of the protective layer 3. This is true even though the nominal lattice constant of 3 is substantially different.

  Therefore, two main conditions must be satisfied to perform this process.

  In order to prevent defects (dislocations, local strains, etc.) from appearing in the protective layer 3, the lattice constants of the protective layer 3 and the lower region of the protective layer 3 do not have greatly different values.

  The protective layer 3 must be sufficiently thin so that strain within the thickness of the protective layer 3 does not gradually relax and / or defects do not occur. For this purpose, the thickness of such a protective semiconductor layer 3 made of a distorted crystal material must be less than the critical thickness known to the person skilled in the art, in particular the material forming it, the material of the neighboring layers And depends on the technique of forming the strained layer. Typical critical thickness is a few hundred angstroms or less.

  Some examples of “standard critical thickness” can be found in “High-Mobility Si and Ge structures” of Friedrich Schaffler (“Semiconductor Science Technology” 12 (1997) 1515-1549).

  In the second embodiment of the protective layer 3, the material of the protective layer 3 is selected so as to have a nominal lattice constant that is approximately close to the nominal lattice constant of the material that forms the region near the protective layer 3.

  Therefore, unlike the first embodiment, in this case, the crystal structure of the protective layer 3 is relaxed.

  For this purpose and in addition, a material is selected, for example, for the protective layer 3 and its construction so as to satisfy the selectivity criteria during substance removal performed during the first recycling step. At least one of the atoms is different from the constituent atoms of the neighboring material, while maintaining a lattice constant close to the lattice constant of the neighboring region, so this constituent atom has selectivity for the neighboring layer in question. The main atom to be determined.

  In certain special cases, the constituent atoms of the material of the protective layer 3 are not found in the material forming the neighboring region involved in the removal of the selective substance, and therefore these materials are completely different.

  In some other special cases, each different constituent atom of the protective layer 3 relative to the neighboring region involved in selective material removal may be an additional atom or an atom lost from the neighboring layer in question.

  For example, even if the protective layer 3 having substantially the same lattice constant as the neighboring region is doped, the lattice constant after doping can be substantially prevented from being affected.

  When the protective layer 3 is made of the same material as the region included in the selective substance removal and close to the protective layer 3, this doping atom becomes an atom that determines the selection ability.

  When doping the protective layer 3, if it is not desired that defects such as dislocations, in particular helical dislocations, appear, the protective layer 3 must be thinner than a certain critical thickness known to those skilled in the art.

  In 3rd embodiment of the protective layer 3, the surface of the layer already formed in order to form a porous layer is made porous.

  This porosification is carried out, for example, by anodization, atomic species implantation or other porosification techniques described in document EP0849788A2.

  This layer of porous material forms the protective layer 3 when at least one nearby material has received a selective substance attack determined by a suitable attack means.

  This protective layer 3 is preferably between two neighboring layers, i.e. between a layer whose surface is made porous and a layer formed on this porous material layer, which are substantially the same. Material.

  Since porous formation does not substantially hinder the crystal structure of these two neighboring layers, such a porous layer 3 does not substantially hinder the crystal structure of the donor wafer 10.

  Therefore, a crystal structure that is very close to or substantially the same as the crystal structure of the region in the vicinity of the protective layer 3 is given to the protective layer 3, and thus the protective layer 3 does not interfere with the crystal of the peripheral structure. .

  However, in some cases, the protective layer 3 can also have some influence on the lattice constant of the surrounding structure, and the complete or relatively distorted or relaxed state that the protective layer 3 can give to neighboring layers is In these cases, it exhibits characteristics that contribute little to subsequent applications.

  Various material removal techniques are used in the protective layer 3.

  In the first substance removing technique, a frictional force is applied to the protective layer 3 in order to peel at least a part of the substance to be removed.

  These frictional forces can be applied by the polishing plate, for example, a grinding action and / or a chemical action may be combined.

  The material forming the protective layer 3 is selected from a crystalline material so that there is a mechanical substance attack method having a mechanical attack force, and the mechanical attack force of the material forming the protective layer 3 is two in the vicinity of the protective layer 3. It is substantially different from the mechanical attack force of at least one material in one of the regions and is capable of performing at least one selective mechanical attack method.

The selective mechanical attack method is one of the following mechanical attack methods.
-Selective mechanical attack on the material in the region in the vicinity of the protective layer 3 and on the side of the peeled useful layer with respect to the protective layer 3.

  Therefore, the material of the protective layer 3 has a characteristic that can withstand a very large mechanical attack by a mechanical attack in the region above the protective layer 3.

  For this purpose, for example, the protective layer 3 can be hardened on the upper layer so as to be suitable for the mechanical attack method chosen to remove the upper region.

  Therefore, for example, it is known that carbonated Si containing C having a concentration of typically between 5% and 50% is harder than non-carbonated Si.

  -In the selective mechanical attack of the protective layer 3 material, an etching stop layer is formed in the vicinity of the protective layer 3 and on the same side of the substrate as the protective layer 3.

  The material of the protective layer 3 has a mechanical attack that is substantially small due to the mechanical attack in the region above the protective layer 3, and in particular has the property to withstand corrosion.

  For example, the protective layer 3 can be softened with respect to the lower layer so as to be suitable for the material removal technique selected to remove the protective layer 3.

  The second material removal technique consists of chemically etching the material to be removed.

  Wet etching can be performed using an etching solution suitable for the material to be removed.

  Dry etching can be performed to remove substances by plasma etching or sputtering.

  Furthermore, the etching may be purely chemical, electromechanical, or photoelectrochemical.

  The material of the protective layer 3 is a material that forms the protective layer 3 that is substantially different from at least one of the two regions in the vicinity of the protective layer 3, and an etching fluid (gas or Solution) and selected from crystalline materials that can be subjected to at least one selective etching method.

The selective etching method is one of the following etching methods.
The material located in the vicinity of the protective layer 3 and on the side of the useful layer already peeled from the protective layer 3 is selectively removed, and the protective layer 3 forms an etching stop layer.

  The material of the protective layer 3 is selectively removed, and an etching stop layer is formed in a region near the protective layer 3 and on the same side as the protective layer 3 of the substrate 1.

  Which selective etching method is selected as the process during recycling and for removing a portion of the donor wafer located on the same side as the useful layer that has been stripped, is also used as an etch stop layer (first etching method). In the case of the second selective etching method, there is a protective layer 3 in the case, and a region in the vicinity of the protective layer 3 located on the same side as the protective layer 3 of the substrate 1 is present.

  Accordingly, the stop layer functions as a barrier for chemical etching, and in the same way, protects the material of a part of the lower part of the protective layer 3 (which constitutes at least a part of the buffer structure I).

As described above, the selective characteristics for removing the substance between the material of the protective layer 3 and the material in the neighboring region included in the selective etching can be obtained from the following facts.
The two materials are different, or
The two materials contain approximately equivalent atoms, except at least one atom, or
The two materials are approximately equivalent, but at least one atom of one material has an atomic concentration greatly different from the atomic concentration of the same atom of the other material, or
The two materials have different porosity densities.

  Referring to FIG. 4, the protective layer 3 is in the buffer structure I, and the donor wafer 10 includes a support substrate 1, a bottom 2 ′ of the buffer structure I, a protective layer 3, and an upper part 4 ′ of the buffer structure I.

  Here, the protective layer 3 can protect the bottom 2 ′ of the buffer structure I.

  In such reuse of the donor wafer 10 of the present invention during the first step, all parts located on the same side as the protective layer 3 of the upper 4 ′ of the buffer structure I are peeled off.

  The selective material removal in the protective layer 3 is performed by at least one of the following selective material removal modes.

  The material is removed from the region of the portion 4 ′ in the vicinity of the protective layer 3, and the protective layer 3 forms a layer that stops substance removal.

  The material is removed from the protective layer 3, and a layer in which the region 2 ′ near the protective layer 3 stops removing the substance is formed.

  Any selective material removal mode is performed as a process during reuse, and the layer for stopping the material removal (first selective material removal of the first selective material removal) is also used for removing the portion 4 ′ in the vicinity of the protective layer 3. In the case of the second selective substance removal, there is a region 2 ′ in the vicinity of the protective layer 3 and functions as a substance attack or etching barrier, and in the same way Protect the material of the bottom 2 'of structure I.

  Therefore, the crystal structure of the buffer structure I is not substantially hindered, and this kind of protective layer 3 must have its own crystal structure, which is substantially in the crystal structure near the buffer structure I. It must be manufactured according to an embodiment which is the same and can obtain this material property, such as one of the above three embodiments.

  After the above removal of the material on the bottom 2 ′ of the buffer structure I, on reuse, a new top 4 ′ of the buffer structure I, and if possible (during the second selective material removal mode above) It is advantageous to form a new layer of the protective layer 3 that has already been removed (or by a treatment suitable for removing this layer 3).

  These layers 3, 4 ′ are grown in situ or after an additional step for surface treatment of the donor wafer 10 where the growth takes place, which surface treatment can be performed, for example, by CMP polishing, heat treatment or other Any slip planes, stacking faults, etc. that are performed by planarization techniques, do not propagate dislocations and other defects contained in the buffer structure I, prevent its size from increasing, and degrade the quality of the buffer structure I This prevents the defect from being formed.

  Thus, the bottom 2 'of the buffer structure I remains intact during reuse, which is different from the conventional method.

  In a special and effective embodiment of the donor wafer 10, the bottom 2 ′ of the buffer structure I is a buffer layer and the top 4 ′ of the buffer structure I is a buffer, as shown in FIG. It is an additional layer to the layer.

  This particular embodiment has the effect of protecting the buffer layer 2 ', i.e. a part of the buffer structure I which is usually the most difficult, long and costly to manufacture.

  The additional layer 4 ′ is usually formed by epitaxial growth combining fixed parameters (atom concentration for epitaxial growth, temperature, pressure, atmosphere, growth rate, rate, etc.), and is itself peeled off during the step of peeling the layer from the donor wafer 10. Therefore, it seems that there is no need to protect this additional layer 4 ′ by the protective layer 3 during reuse.

  However, in another particular embodiment of the donor wafer 10, the protective layer 3 can be arranged in the additional layer 4 'in order to protect at least a part of the additional layer 4'.

  In another particular embodiment of the donor wafer 10, the protective layer 3 is formed in the buffer layer 2 'in order to protect only a part of the buffer layer 2', for example the most difficult part to form.

  Referring to FIG. 5, the second donor wafer 10 of the present invention is different from the donor wafer 10 shown in FIG. 2, and the protective layer 3 is no longer seen in the buffer structure I, but just above the buffer structure I.

  Furthermore, the upper layer 5 is present on the protective layer 3 and at least a part of the useful layer is peeled off in transferring the layer from the donor wafer 10.

  The composition and crystal structure of the upper layer 5 are selected depending on the desired physical, electrical and / or mechanical properties desired in the transferred structure.

  The material of this upper layer 5 may have, for example, a nominal lattice constant substantially the same as that of the portion of the buffer structure I adjacent to the protective layer 3 in order to maintain a substantially relaxed structure.

  Furthermore, the material of this upper layer 5 has, for example, a nominal lattice constant substantially different from the part of the buffer structure I adjacent to the protective layer 3 and the lattice constant of the part of the buffer structure I adjacent to the protective layer 3. The thickness may be sufficiently small to maintain the thickness, and therefore may be distorted.

  Here again, the material of the upper layer 5 may be selected to have, for example, an intermediate structure between a distorted structure and a relaxed structure.

  In an advantageous embodiment, the upper layer 5 is formed by layer growth, for example by CVD or MBE epitaxial growth.

  In this manner and in the first case, the upper layer 5 is grown immediately after the formation of the upper portion of the buffer structure I. In this case, the buffer structure I is also effectively formed by the layer growth.

  In the second case, the upper layer 5 is grown after the sub-process of finishing the upper surface of the lower buffer structure I, and this finish is performed using, for example, CMP polishing, heat treatment or other planarization techniques, The dislocations and other defects involved are not propagated, the size is not increased, and any slip planes, stacking faults or other defects that degrade the quality of the buffer structure I are prevented.

  With respect to the protective layer 3, its function is in this case to protect substantially all of the lower part of the buffer structure I and the substrate 1 from material removal which takes place during the first recycling step.

  In the reuse of the donor wafer 10 after the useful layer has been peeled from the donor wafer 10 in the upper layer 5, it is substantially the same as the protective layer 3 of the upper layer 5 that is located on the same side during the first step. To peel all parts.

  The selective material removal in the protective layer 3 is performed by at least one of the following selective material removal modes.

  The material of the upper layer 5 in the vicinity of the protective layer 3 is selectively removed, and the protective layer 3 forms a layer that stops substance removal.

  The material of the protective layer 3 is removed, and a layer in which the region of the buffer structure I near the protective layer 3 stops substance removal is formed.

  Any selective material removal mode is performed as a process during reuse, and the layer for stopping the material removal, even for removing the remaining upper layer 5 (in the case of the first selective material removal) In the case of the second selective material removal, there is an upper region of the buffer structure I in the vicinity of the protective layer 3 and functions as a barrier for material etching or attack. Similarly, the buffer structure I Protect material.

  Further, in order to maintain the influence of the structure of the buffer structure I on the structure of the grown upper layer 5, so that the protective layer 3 does not substantially disturb the crystal structure of the buffer structure I immediately below, and above It is effective to prevent the crystal growth of the upper layer 5 from being substantially inhibited, and therefore, it is effective to manufacture in accordance with the three embodiments already described.

  After the removal of the material above the buffer structure I, upon reuse, a new upper layer 5 and a protective layer 3 (suitable for removing this layer 3 during or during the second selective material removal mode described above). If already removed), it is effective to form a new protective layer 3.

  These layers 5, 3 are grown in situ or after an additional step for the surface treatment of the donor wafer 10 where the growth takes place, this surface treatment being for example CMP polishing, heat treatment or other planar Any slip planes, stacking faults, and other defects that are made by the technology, so that dislocations and other defects contained in the buffer structure I are not propagated, the size is not increased, and the quality of the buffer structure I is deteriorated It is intended to prevent the formation of defects.

  Referring to FIG. 6, the third donor wafer 10 of the present invention differs from the donor wafer 10 seen in FIG. 3 mainly in that there is an intermediate layer 8 between the buffer structure I and the protective layer 3.

  The composition and crystal structure of the intermediate layer 8 are selected depending on the desired physical, electrical and / or mechanical properties.

  In order to maintain a substantially relaxed structure, the material of the intermediate layer 8 may have, for example, a nominal lattice constant substantially the same as the buffer structure I in a portion adjacent to the interface. In this case, the intermediate layer 8 is an extension of the buffer structure I, which can further increase the crystal rigidity of the growth surface of the upper layer 5, for example.

  Furthermore, the material of the intermediate layer 8 has, for example, a nominal lattice constant substantially different from the buffer structure I in a portion adjacent to the interface, and maintains the lattice constant of the buffer structure I in a portion adjacent to the protective layer 3. It may be as thin as possible and therefore distorted.

  In an advantageous embodiment, the intermediate layer 8 is formed by layer growth, for example by CVD or MBE epitaxial growth.

  In this embodiment and in the first case, the layer in question is grown in situ immediately after the formation of the lower layer, in which case the lower layer is also effectively formed by layer growth.

  In the second case, the layer in question is grown after a sub-step of finishing the upper surface of the lower layer, and this finish is included in the buffer structure I, for example using CMP polishing, heat treatment or other planarization techniques. To prevent dislocations and other defects from propagating, to prevent the size from increasing, and to prevent the formation of any slip surfaces, stacking faults or other defects that degrade the quality of the buffer structure I.

  With respect to the protective layer 3, its function is in this case to protect substantially the entire lower intermediate layer 8, the entire buffer structure I and the substrate 1 from material removal which takes place during the first recycling step. It is.

  The selective material removal in the protective layer 3 is performed by at least one of the following selective material removal modes.

  The material of the upper layer 5 in the vicinity of the protective layer 3 is removed, and the protective layer 3 forms a layer that stops substance removal.

  The material of the protective layer 3 is removed, and a layer in which the region of the intermediate layer 8 in the vicinity of the protective layer 3 stops the substance removal is formed.

  Any selective material removal mode is performed during reuse, and the layer for stopping the material removal, even for removing the remaining upper layer 5 (protection in the case of the first selective material removal mode) In the case of the second selective material removal mode, there is an intermediate layer 8 region in the vicinity of the protective layer 3 and functions as a barrier for material etching or attack. Protect material.

  Further, in order to maintain the influence of the structure of the intermediate layer 8 on the structure of the grown upper layer 5, so that the protective layer 3 does not substantially disturb the crystal structure of the intermediate layer 8 immediately below, and above it. It is effective that the crystal growth of the upper layer 5 is not substantially inhibited, and therefore, it is effective to manufacture in accordance with the two embodiments already described.

  After the removal of the material on top of the intermediate layer 8, upon reuse, a new top layer 5 and a protective layer 3 (suitable for removing this layer 3 during or during the second selective material removal mode described above). If already removed), it is effective to form a new protective layer 3.

  These layers 5, 3 are grown in situ or after an additional step for the surface treatment of the donor wafer 10 where the growth takes place, this surface treatment being for example CMP polishing, heat treatment or other planar Any slip planes, stacking faults, and other defects that are made by the technology, so that dislocations and other defects contained in the buffer structure I are not propagated, the size is not increased, and the quality of the buffer structure I is deteriorated It is intended to prevent the formation of defects.

  Referring to FIGS. 7a to 7f, various steps of a method for peeling a thin layer from a donor wafer 10 provided with a protective layer 3 and reusing the donor wafer 10 are shown and substantially as described with reference to FIG. In general, a donor wafer 10 having the same layer structure is used, and therefore, referring to FIG. 7a, it comprises a substrate 1 and a buffer structure I, in which a protective layer 3 is present.

  In the example to be studied, the buffer layer 2 and the additional layer 4 in the buffer structure I are separated by the protective layer 3.

  In this method example of the invention, an upper layer 5 is added on the additional layer 4.

  The removal performed during this process relates to the stripping of part of the additional layer 4 and the upper layer 5.

  Similarly, in another embodiment of the donor wafer 10, there are various upper layers, and the peeling is related to these upper layers and, in some cases, the portion of the additional layer 4, or there is no upper layer and the peeling is performed. Relates only to the portion of the additional layer 4.

  Furthermore, the protective layer 3 that is considerably thin is often required. As described above, if the protective layer 3 is too thick, the crystal characteristics of the buffer structure I, such as the occurrence of defects such as dislocations or lattice constant changes, are affected. Can be given.

  Therefore, in this case, the thickness of the protective layer 3 should be smaller than the critical thickness beyond which an undesirable effect may appear.

  These four layers 2, 3, 4 and 5 are formed by epitaxial growth using a known technique such as CVD or MBE.

  In the first case, at least one of these four layers is grown in situ immediately following the formation of the lower growth support member, in which case the lower growth support member is also effectively formed by layer growth. Is done.

  In the second case, at least one of these four layers is grown after the sub-step of finishing the upper surface of the lower growth support member, and this finish can be achieved, for example, by CMP polishing, heat treatment or other planarization techniques. Used to prevent dislocations and other defects contained in the buffer structure I from propagating, not to increase in size, and to form any slip surfaces, stacking faults and other defects that degrade the quality of the buffer structure I Do not be.

  A method of peeling the thin layer is shown in FIGS. 7b and 7c.

  The first preferred stripping step of the present invention consists of forming a fragile region in the additional layer 4, which is done for subsequent separation and separation of the desired layers.

  Various techniques for forming such fragile regions are shown here.

The first technique, called Smart-cut ^(c) , known to those skilled in the art (and found in many studies covering techniques for thinning wafers) is the first technique In the process, atomic species having a certain energy (such as hydrogen ions) are implanted, and thus the weak region is formed.

  The second technique consists of forming at least one porous layer to form a fragile interface, as described in document EP-A-0845788.

  In the exemplary method of the present invention, the fragile region effectively formed by one of these two techniques is formed between the upper layer 5 and the additional layer 4 or in the additional layer 4.

  If the upper layer 5 is sufficiently thick, a fragile region may be formed therein, which is the case when the upper layer 5 comprises a laminate.

  Referring to FIG. 7b, the second step for peeling off the thin layer consists of depositing the receiving substrate 6 on the surface of the upper layer 5.

  The receiving substrate 6 forms a mechanical support member that supports the upper layer 5 that is peeled from the donor wafer 10 and has sufficient rigidity to protect the upper layer 5 from any mechanical strain from the outside. Have.

  The receiving substrate 6 may be made of, for example, silicon, quartz, or other types of materials.

  The receiving substrate 6 is adhered to the upper layer 5 by adhering it to the upper layer 5 and bonded thereto. Here, molecular adsorption is effectively performed between the substrate 6 and the upper layer 5.

  This coupling technique includes, in particular, modified examples. Y. Tong, U. It is described in a document named “Semiconductor Wafer Bonding” by Gosele and Wiley.

  If necessary, the appropriate pretreatment of each surface to be bonded and / or the supply of thermal energy and / or the application of further binders may be combined with the bonding.

  Therefore, for example, when the heat treatment is performed during or immediately after bonding, the bonded portion can be strengthened.

  Further, the bonding can be controlled by inserting a bonding layer such as silica having a particularly high molecular bonding force between the upper layer 5 and the receiving substrate 6.

  In order to form the SOI structure from the layer to be peeled off, it is effective to select an electrically insulating material as a material for forming the bonding surface of the receiving substrate 6 and / or a bonding layer that can be formed. The upper semiconductor layer 5 is transferred together with or without a part of the additional layer 4.

  Once the receiving substrate 6 is bonded, a portion of the donor wafer 10 is peeled away from the pre-formed fragile region.

In the case of the first technique (Smart-cut ^(c) ), in the second step, in order to peel off a part of the donor wafer 10 in the fragile region, the implantation region (which forms the fragile region) is heated and / or mechanically. Treat or supply other energy.

  In the case of the second technique, heat and / or mechanical treatment is applied to the fragile layer or other energy is supplied to peel a portion of the donor wafer 10 from the fragile layer.

  By stripping in the fragile region by one of these two techniques, the majority of the wafer 10 can be stripped to obtain a structure that can include the buffer structure I, the top layer 5, various bonding layers, and the receiving substrate 6. it can.

  In the exfoliated layer, the step of finishing the surface of the formed structure may include any surface roughness, thickness non-uniformity and / or unnecessary layer by, for example, chemical mechanical polishing, CMP, etching, or at least heat treatment. Effectively done to get rid of.

  The post-peeling layer 7 after peeling forms a portion where the protective layer 3 remains, and the entire wafer is reused to form a donor wafer 10 ′ to be used again when peeling another layer thereafter.

  The reuse process is shown in FIGS. 7d, 7e and 7f.

  Referring to FIG. 7d, the first recycling step corresponds to removing substantially all of the post-peeling layer 7 and, in some cases, removing the protective layer 3 as well.

  A mechanical or chemical mechanical attack or a suitable treatment is first performed in order to remove a part of the remaining part of the additional layer 4 of the layer 7 after peeling, for example by abrasion, polishing, CMP, chemical Any slip surface that is performed by an attack by etching, heat treatment or planarization, so that dislocations and other defects contained in the buffer layer 2 do not propagate, the size does not increase, and the quality of the buffer structure I deteriorates. In order to prevent the formation of stacking faults and other defects.

  Furthermore, some of these material etching or attack techniques can be combined or performed one after another, for example, chemical etching and CMP.

  In all cases, the first recycling step uses at least one of the selective material removal modes described above.

  Referring to FIGS. 7e and 7f, the second reuse step corresponds to the restoration of the same layer and the formation of the additional layer 4 'and the upper layer 5' that were practically present before peeling.

  Further, this restoration includes forming the removed protective layer 3.

  Forming the layers by a technique substantially equal to one of those described in detail above effectively restores those layers.

  The resulting donor wafer 10 ″ ′ layers 4 ′, 5 ′ need not be the same as donor wafer 10 layers 4, 5, and the donor wafer shown in FIG. 7 d functions as a substrate for other types of layers. Also good.

  In the example of the method of the invention detailed above, delamination involves a part of the additional layer 4 and the upper layer 5.

  At the same time, this example can be applied to delamination involving only a part of the additional layer 4 (donor wafer 10 without the upper layer 5).

  At the same time, this example can be applied to delamination involving only a part of the upper layer 5, the reuse consists of removing the remaining part of the upper layer 5.

  In the example of the method of the present invention described in detail above, the protective layer 3 is located between the buffer layer 2 and the additional layer 4.

  In this example, it is obvious that the present invention can also be applied when the protective layer 3 is located in the buffer layer 2 or the additional layer 4.

  In general, this example also applies when the protective layer 3 is located in the buffer structure I.

  Using the donor wafer 10 seen in FIG. 2, the description of the method of the present invention with reference to FIGS. 7a-7f is applicable to each donor wafer 10 seen in the following drawings.

  In FIG. 5, instead of arranging the protective layer 3 in the buffer structure I, the protective layer 3 is arranged between the buffer structure I and the upper layer 5, so that the protective layer 3 is peeled off by the upper layer 5 and reused. The peeling of the material is completed by selective etching of the upper layer 5 with respect to the protective layer 3 and / or selective etching of the protective layer 3 with respect to the buffer structure I.

  In FIG. 6, the intermediate layer 8 is disposed between the buffer structure I and the protective layer 3 by adding the intermediate layer 8 to the donor wafer 10, and the intermediate layer 8 is peeled off by the upper layer 5 for reuse. The material peeling is terminated by selective etching of the upper layer 5 with respect to the protective layer 3 and / or selective etching of the protective layer 3 with respect to the intermediate layer 8.

  After the donor wafer 10 of the present invention is reused, the useful layer peeling method can be performed again.

  In the effective context of the present invention, the following steps are repeated and each is repeated to perform a repeated method of peeling the useful layer from the donor wafer 10 according to the present invention.

Exfoliation Mode, and Repeating Method in the Present Invention Prior to performing the exfoliating method, the method for forming donor wafer 10 of the present invention can be implemented with one or more techniques for forming a thin layer on a substrate as described above. .

  In the remainder of this document, an example of an embodiment of a donor wafer 10 with a buffer structure I that can be processed by the method of the present invention is shown.

  In particular, materials that can be used effectively for such donor wafers are shown.

  As already indicated, the buffer structure I formed on the substrate 1 having the first lattice constant often has the main function of having the second lattice constant on its free surface.

  The buffer structure I therefore comprises a buffer layer 2 that allows such matching of lattice constants.

A technique often employed to obtain a buffer layer 2 having this characteristic is to obtain a buffer layer 2 consisting of several atoms,
At least one atom included in the composition of the substrate 1, and
It is at least one atom which is not contained or hardly contained in the substrate 1 and whose concentration gradually changes within the thickness of the buffer layer 2.

  When the concentration of this atom gradually changes in the buffer layer 2, this is the main cause, and the lattice constant in the buffer layer 2 gradually changes like a modifying action.

  Therefore, in this embodiment, the buffer layer 2 is mainly an alloy.

  The atoms constituting the substrate 1 and the buffer layer 2 may be selected from group IV such as Si or Ge.

  In this case, for example, SiGe whose Ge concentration gradually changes with a thickness between the substrate 1 made of Si and a value close to zero at the interface with the substrate 1 and a specific value on the other surface of the buffer layer 2. The buffer layer 2 is obtained.

  As another scenario, the substrate 1 and the buffer layer 2 may be configured with a group of III-V atoms such as (Al, Ga, In)-(N, P, As) as a possible combination.

  In this case, for example, the substrate 1 made of ASGa, and As and / or Ga and at least one other atom, the other atom has a value close to zero at the interface with the substrate 1 and the other surface of the buffer layer 2. Thus, the buffer layer 2 in which the concentration gradually changes with a thickness between a certain value of the above can be obtained.

  As a possible combination, the substrate 1 and the buffer layer 2 may be configured by a group of II-VI group atoms such as (Zn, Cd)-(S, Se, Te).

The following are some examples of such embodiments.
The first three examples relate to a donor wafer 10 comprising a substrate 1 made of Si, a buffer layer 2 made of SiGe, and another layer made of Si and SiGe.

  These wafers 10 are particularly useful for stripping strained SiGe and / or Si layers to form SGOI, SOI, or Si / SGOI structures.

  In this situation, the etching solution used depends on the material to be etched (Si or SiGe). Therefore, the etching solutions capable of etching these materials are classified into categories by assigning an identifier included in the following list of each category.

S1: a solution containing at least one of KOH, NH ₄ OH (ammonium hydroxide), TMAH, EDP or HNO ₃ H, or HNO ₃ , HNO ₂ H described on page 9 of document WO 99/53539 A combination of chemical agents such as ₂ O ₂ , HF, H ₂ SO ₄ , H ₂ SO ₂ , CH ₃ COOH, H ₂ O ₂ , and H ₂ O, such as the solution currently under consideration, SiGe Etching solution for Si against

Selective etching solution for SiGe against Si, such as S2: HF: H ₂ O ₂ : CH ₃ COOH (selectivity is about 1: 1000) or HNA (hydrogen fluoride-nitric acid-acetic acid solution)

  A selective etching solution for SiGe having a Ge concentration substantially lower than or equal to 20% relative to SiGe having a Ge concentration equal to or higher than about 25%, such as a solution comprising Sc1: TMAH or KOH

Sd1: of undoped Si against Si doped with boron at a level preferably higher than 2 × 10 ¹⁹ cm ⁻³ , such as a solution containing EDP (ethylenediamine pyrocatechol), KOH or N ₂ H ₂ (hydrazine) Selective etching solution for

  Example 1: After reuse, the donor wafer 10 has a substrate 1 made of Si, a buffer structure I made of SiGe having a buffer layer 2 and an additional layer 4, and the upper layer 5 remaining after peeling a part of the upper layer 5. And a post-peeling layer 7 made of Si or SiGe forming a part.

  The buffer layer 2 is preferably a layer in which the Ge concentration gradually increases from the interface with the substrate 1 so that the SiGe lattice constant changes as described above.

  Its thickness is typically between 1 and 3 micrometers so that good structural relaxation is obtained at the surface and defects with different lattice constants are embedded.

  The additional layer 4 is substantially relaxed by the buffer layer 2 and is made of SiGe which is uniform and has a Ge concentration substantially equal to that of the buffer layer 2 near the interface with the buffer layer 2.

  The germanium concentration in the silicon in the relaxed SiGe layer is typically between 15% and 30%.

  This 30% limit is a typical limit in current technology, but will change within a few years.

  The thickness of the additional layer 4 varies greatly from case to case and is typically between 0.5 and 1 micron.

  When the layer 7 after peeling is made of Si, selective etching of Si on the additional layer 4 made of SiGe is effectively performed using an S1 type etching solution to remove Si.

  If the post-peeling layer 7 is made of SiGe, and the Ge concentration of the post-peeling layer 7 is substantially lower than or equal to 20% and the Ge concentration of the additional layer 4 is substantially equal to or higher than 25%, an additional SiGe layer Selective etching of the post-peeling layer with respect to 4 is performed using a Sc1 type etching solution to remove the post-peeling layer.

  In all cases, the final portion of the layer 7 after stripping is completely removed by chemical means, and an etch stop in the protective layer 3 forms a stop layer, protecting the lower layer that is desired to remain.

  Example 2: After recycling, the donor wafer 10 is substantially equal to that of Example 1, the difference being that there is a protective layer 3 in the wafer 10.

  The protective layer 3 is made of strained Si, SiGe, or Si doped with boron.

  If the protective layer 3 is made of distorted Si, it is recognized again that the thickness of the protective layer 3 should not exceed the critical thickness here.

  Thus, for example, the critical thickness of the protective layer 3 made of strained Si inserted between two SiGe layers each having a Ge concentration approximately equal to 20% is typically equal to about 20 nanometers.

  In the first case, the protective layer 3 is arranged between two SiGe layers.

  This is particularly the case when the protective layer 3 is arranged between two layers of the buffer structure I, between the buffer structure I and the post-peeling layer 7 made of SiGe, or in the post-peeling layer made of SiGe. is there.

  Therefore, various types of etching are performed depending on the material of the protective layer 3.

  When the protective layer 3 is made of distorted Si, the upper part made of SiGe is selectively etched using an S2 type solution and / or after the layer 7 is removed after the peeling, the protective layer 3 is made of an S1 type solution. Is selectively etched.

  If the Ge concentration is approximately equal to or higher than 25% and the protective layer 3 is formed, and if the Ge concentration of the upper layer is substantially lower than or equal to 20%, the upper SiGe portion uses the Sc1 type solution. Selectively etched.

  If the Ge concentration is substantially lower than or equal to 20% and the protective layer 3 is made of SiGe, and if the Ge concentration of the lower layer is approximately equal to or higher than 25%, after the post-peeling layer 7 is removed The protective layer 3 is selectively etched using a Sc1 type solution.

  In the second case, the protective layer 3 is arranged between the lower SiGe layer and the upper Si layer.

  This is particularly true between the buffer structure I and the post-peeling Si layer, or between the SiGe intermediate layer 8 and the post-peeling Si layer 7, or post-peeling layer 7 between the SiGe layer and the Si layer. This is a case where the protective layer 3 is disposed inside.

  Various types of etching are performed depending on the material of the protective layer 3.

When the protective layer 3 is made of B-doped Si, the upper Si portion is selectively etched using an Sd1 type solution,
When the protective layer 3 is made of SiGe, the upper Si portion is selectively etched using an S1 type solution.

  If the Ge concentration is substantially lower than or equal to 20% and the protective layer 3 is made of SiGe, and if the Ge concentration of the lower layer is approximately equal to or higher than 25%, after the post-peeling layer 7 is removed The protective layer 3 is selectively etched using a Sc1 type solution.

  In the third case, the protective layer 3 is arranged between two Si layers.

  This is especially the case when the protective layer 3 is arranged between the Si intermediate layer and the post-peeling Si layer 7 or within the post-peeling Si layer 7.

  Various types of etching are performed depending on the material of the protective layer 3.

When the protective layer 3 is made of B-doped Si, the upper Si portion is selectively etched using an Sd1 type solution,
If the protective layer 3 is made of SiGe, the upper Si portion is selectively etched using an S1 type solution and / or
After the post-peeling layer 7 is removed, the protective layer 3 is selectively etched using an S2 type solution.

  Example 3: After reuse, the donor wafer 10 is made of a substrate 1 made of Si, a buffer structure I having a SiGe buffer layer 2 and an additional Ge layer 4, and a remaining part of the upper layer 5 after stripping a part of the upper layer 5. A post-peeling AsGa layer 7 to be formed and a protective AlGaAs layer 3 disposed on the post-peeling layer 7 are provided.

  The buffer layer 2 is preferably a layer in which the Ge concentration gradually increases from the interface with the substrate 1 so that the lattice constant changes between the lattice constant of the Si substrate 1 and the lattice constant of the additional Ge layer 4. .

  For this purpose, the Ge concentration in the buffer layer 2 is increased from about 0 to about 100%, more precisely about 98%, so that the theoretical lattice between the two materials is perfectly matched.

The first scenario uses an etching solution such as a solution containing citric acid (C ₆ H ₈ C ₇ ) and hydrogen peroxide having a pH between about 6 and 7 (selectivity is typically 20). Then, almost all of the remaining post-peeling layer 7 can be peeled off by selective chemical etching of the post-peeling layer 7, where the protective layer 3 behaves like an etching stop layer.

  As a second scenario, after removing a part of the post-peeling layer 7 on the top of the protective layer 3, when the aluminum concentration of the protective layer 3 is more than 20%, dilute hydrochloric acid (between about 9% and 48%) (selection factor is typically Almost all of the protective layer 3 can be peeled off by selective chemical etching of the protective layer 3 with a selective etching solution such as a solution containing between 350 and 10,000), where the post-peeling lower layer 7 is an etching stop layer. Behaves like

  As a third scenario, two etchings can be subsequently performed in order to strip at least a part of the layer 7 after stripping and the protective layer 3.

  Thus, the buffer structure I is saved and completely reused.

  Example 4: After reuse, the donor wafer 10 comprises a substrate 1 with at least one AsGa portion at the interface with the buffer structure I, at least a portion of the buffer structure I made of III-V material, and a portion of the upper layer 5. And a post-peeling layer 7 made of a III-V material that forms the remaining portion of the upper layer 5 after peeling.

  The main advantage of this buffer structure I is that the lattice constant of the material of the upper layer 5 (nominal value is about 5.87 angstroms) matches the lattice constant of AsGa (nominal value is about 5.65 angstroms).

  In bulk III-V materials, and when comparing bulk InP and bulk AsGa, the latter is less expensive, more readily available in the semiconductor market, is not mechanically fragile, and uses technology that makes contact by the backside The material is well known and its size is large (typically 6 inches compared to 4 inches for bulk InP).

  All the advantages provided by such a donor wafer 10 can be seen here because the active layer of III-V material to be transferred can be formed, in particular, with high quality and properties, It is close to the characteristics seen when a III-V material is formed as a bulk product.

  As a special aspect of the donor wafer 10 before peeling, InP from which the upper layer 5 before peeling is peeled off is provided.

  Since the size of the bulk InP is normally limited to 4 inches, the donor wafer 10 is a measure for forming the InP layer with a size of 6 inches, for example.

  The buffer structure I for forming such an upper layer 5 typically requires a thickness greater than 1 micrometer and is thicker if reused according to the present invention.

  The epitaxial growth techniques normally performed to form such a buffer structure I are further particularly difficult and costly, and it is therefore preferable if the buffer structure I can be at least partially restored after peeling off the useful layer.

  It is effective that the buffer structure I includes a buffer layer 2 made of InGaAs whose In concentration varies between 0 and about 53%.

  The buffer structure I may further comprise an additional layer 4 made of a III-V group material such as InGaAs or InAlAs with a substantially constant atomic concentration.

  In the case of a special exfoliation, a part of the InP upper layer 5 and the additional layer 4 is exfoliated and transferred to the receiving substrate.

  Thus, any electrical properties that exist between the two exfoliated materials will benefit.

  That is, when a part of the peeled additional layer 4 is made of InGaAs or InAlAs, the electron mobility of the peeled layer is increased by the discontinuity of the electronic band between the latter material and InP.

  Other embodiments of donor wafer 10 comprising other III-V group compounds such as InAlAs are possible.

  Typical applications for such delamination include HEMT (“High-Electron Mobility Transistor”) or HBT (“Heterojunction Bipolar Transistor”).

  A chemical etching solution that can selectively remove one III-V material relative to another III-V material is effectively used during the first recycling step.

  Therefore, for example, in order to remove the post-InP stripping layer 7 without removing the lower InGaAs layer, InP selective etching is effectively performed using a solution containing concentrated HCl.

Example 5: After reuse, the donor wafer 10 has a substrate 1 with AsGa at the interface with the buffer structure I, a buffer structure I with InAsGa at the interface with the post-peeling layer 7, and an upper part after stripping a part of the upper layer 5. The InP post-peeling layer 7 that forms the remaining part of the layer 5, and is arranged between the post-peeling layer 7 and the buffer structure I or in the post-peeling layer 7, and is made of In _x Ga _1-x As _y P _{1 -y} And a protective layer 3.

  This type of donor wafer 10 (without the release layer 3) has already been described in Example 4.

  As a first scenario, almost all of the remaining post-peeling layer 7 can be peeled off by selective chemical etching of the post-peeling layer 7 using a selective etching solution such as a solution containing HF, where the protective layer 3 Behaves like an etch stop layer.

As a second scenario, a part of the post-peeling layer 7 on the protective layer 3 is removed, and then the protective layer 3 is selectively chemically etched with a selective etching solution such as a solution containing Ce ^IV H ₂ SO _4. 3 can be peeled off, where the lower layer of the protective layer 3 behaves like an etch stop layer.

  As a third scenario, two etchings can be subsequently performed in order to strip at least a part of the layer 7 after stripping and the protective layer 3.

  Thus, the buffer structure I is saved and completely reused.

  Example 6: After reuse, the donor wafer 10 comprises a substrate 1 comprising AsGa at the interface with the buffer structure I, a buffer structure I comprising InGaAs, and an InP protective layer 3 disposed on or in the InGaAs.

As a first scenario, almost all of this material on top of the protective layer 3 can be stripped by selective chemical etching of InGaAs on the top of the protective layer 3 with a selective etching solution such as a solution containing Ce ^IV H ₂ SO ₄ , Here, the protective layer 3 behaves like an etching stop layer.

  As a second scenario, after removing InGaAs on the protective layer 3, almost all of the protective layer 3 can be peeled off by selective chemical etching of the protective layer 3 using a selective etching solution such as a solution containing HF, Here, InGaAs under the protective layer 3 behaves like an etching stop layer.

  As a third scenario, two etchings can be subsequently performed to strip a portion of InGaAs and the protective layer 3.

  The semiconductor layer indicated in this document contains other components such as carbon whose carbon concentration is substantially lower than or equal to 50%, or in particular, lower than or equal to 5% concentration in this semiconductor layer. Can be added.

  Finally, the present invention applies not only to the buffer structure I, the intermediate layer 8 or the upper layer 5 made of the material shown in the above example, but also to IV-IV, III-V, II-VI group alloys. .

  These alloys may be binary, ternary, quaternary, or higher alloys.

  The present invention is not limited to the reusable buffer layer 2 or buffer structure I which has the main function of matching the lattice constant between two adjacent structures to each different lattice constant, but is most commonly defined in this document, and And any buffer layer 2 or buffer structure I that can be reused according to the invention.

  The structure finally obtained after peeling is not limited to any of SGOI, SOI, Si / SGOI structure, or HEMT and HBT transistor structures.

It is a figure which shows the donor wafer of a prior art. It is a figure which shows the donor wafer after peeling. It is a figure which shows the donor wafer after a 1st reuse process. It is a figure which shows the 1st donor wafer by this invention. It is a figure which shows the 2nd donor wafer by this invention. It is a figure which shows the 3rd donor wafer by this invention. FIG. 3 shows a step of the method of the invention in which a thin layer is stripped from a donor wafer and the donor wafer is reused after stripping. FIG. 3 shows a step of the method of the invention in which a thin layer is stripped from a donor wafer and the donor wafer is reused after stripping. FIG. 3 shows a step of the method of the invention in which a thin layer is stripped from a donor wafer and the donor wafer is reused after stripping. FIG. 3 shows a step of the method of the invention in which a thin layer is stripped from a donor wafer and the donor wafer is reused after stripping. FIG. 3 shows a step of the method of the invention in which a thin layer is stripped from a donor wafer and the donor wafer is reused after stripping. FIG. 3 shows a step of the method of the invention in which a thin layer is stripped from a donor wafer and the donor wafer is reused after stripping.

Claims (19)

A method of reusing a donor wafer after peeling off at least one useful layer of a material selected from a plurality of semiconductor materials, wherein the donor wafer comprises a substrate, a buffer structure, and before peeling. And the method comprises the step of removing material at a side of the donor wafer where the delamination has occurred, wherein one of the plurality of surfaces of the buffer structure is the same as the lattice constant of the substrate. And the other surface has a second lattice constant identical to the lattice constant of the useful layer, and at least a portion of the buffer structure remains after the material is removed, At least a portion of the buffer structure is reused as a buffer structure for peeling off the next useful layer, and after removing the material from the donor wafer, the side of the donor wafer where the material was removed To form a layer for reproducing the donor wafer,
There is a further protective layer in the donor wafer, at least a part of the buffer structure is in the lower part of the protective layer, and the etching power for the material of the protective layer of the substance removing means is at least one material in two neighboring regions Unlike the etching power for, the material of the protective layer is selected from a plurality of crystal materials so that the material can be selectively removed,
The buffer structure has a buffer layer and an additional layer before peeling,
The recycling method according to claim 1, wherein the protective layer is in the additional layer . A method of reusing a donor wafer after peeling off at least one useful layer of a material selected from a plurality of semiconductor materials, wherein the donor wafer comprises a substrate, a buffer structure, and before peeling. And the method comprises the step of removing material at a side of the donor wafer where the delamination has occurred, wherein one of the plurality of surfaces of the buffer structure is the same as the lattice constant of the substrate. And the other surface has a second lattice constant identical to the lattice constant of the useful layer, and at least a portion of the buffer structure remains after the material is removed, At least a portion of the buffer structure is reused as a buffer structure for peeling off the next useful layer, and after removing the material from the donor wafer, the side of the donor wafer where the material was removed To form a layer for reproducing the donor wafer,
There is a further protective layer in the donor wafer, at least a part of the buffer structure is in the lower part of the protective layer, and the etching power for the material of the protective layer of the substance removing means is at least one material in two neighboring regions Unlike the etching power for, the material of the protective layer is selected from a plurality of crystal materials so that the material can be selectively removed,
The recycling method, wherein the protective layer is in the buffer structure . A method of reusing a donor wafer after peeling off at least one useful layer of a material selected from a plurality of semiconductor materials, wherein the donor wafer comprises a substrate, a buffer structure, and before peeling. And the method comprises the step of removing material at a side of the donor wafer where the delamination has occurred, wherein one of the plurality of surfaces of the buffer structure is the same as the lattice constant of the substrate. And the other surface has a second lattice constant identical to the lattice constant of the useful layer, and at least a portion of the buffer structure remains after the material is removed, At least a portion of the buffer structure is reused as a buffer structure for peeling off the next useful layer, and after removing the material from the donor wafer, the side of the donor wafer where the material was removed To form a layer for reproducing the donor wafer,
The buffer structure has a buffer layer and an additional layer before peeling,
The buffer structure has a thickness that can include defects, and / or has a surface lattice constant different from the surface lattice constant of the substrate,
There is a further protective layer in the donor wafer, at least a part of the buffer structure is in the lower part of the protective layer, and the etching power for the material of the protective layer of the substance removing means is at least one material in two neighboring regions Unlike the etching power for, the material of the protective layer is selected from a plurality of crystal materials so that the material can be selectively removed,
The recycling method, wherein the protective layer is in the buffer layer . A method of reusing a donor wafer after peeling off at least one useful layer of a material selected from a plurality of semiconductor materials, wherein the donor wafer comprises a substrate, a buffer structure, and before peeling. And the method comprises the step of removing material at a side of the donor wafer where the delamination has occurred, wherein one of the plurality of surfaces of the buffer structure is the same as the lattice constant of the substrate. And the other surface has a second lattice constant identical to the lattice constant of the useful layer, and at least a portion of the buffer structure remains after the material is removed, At least a portion of the buffer structure is reused as a buffer structure for peeling off the next useful layer, and after removing the material from the donor wafer, the side of the donor wafer where the material was removed To form a layer for reproducing the donor wafer,
The buffer structure has a buffer layer and an additional layer before peeling,
The buffer structure has a thickness that can include defects, and / or has a surface lattice constant different from the surface lattice constant of the substrate,
There is a further protective layer in the donor wafer, at least a part of the buffer structure is in the lower part of the protective layer, and the etching power for the material of the protective layer of the substance removing means is at least one material in two neighboring regions Unlike the etching power for, the material of the protective layer is selected from a plurality of crystal materials so that the material can be selectively removed,
The recycling method, wherein the protective layer is in the additional layer . A method of reusing a donor wafer after peeling off at least one useful layer of a material selected from a plurality of semiconductor materials, wherein the donor wafer comprises a substrate, a buffer structure, and before peeling. And the method comprises the step of removing material at a side of the donor wafer where the delamination has occurred, wherein one of the plurality of surfaces of the buffer structure is the same as the lattice constant of the substrate. And the other surface has a second lattice constant identical to the lattice constant of the useful layer, and at least a portion of the buffer structure remains after the material is removed, At least a portion of the buffer structure is reused as a buffer structure for peeling off the next useful layer, and after removing the material from the donor wafer, the side of the donor wafer where the material was removed To form a layer for reproducing the donor wafer,
There is a further protective layer in the donor wafer, at least a part of the buffer structure is in the lower part of the protective layer, and the etching power for the material of the protective layer of the substance removing means is at least one material in two neighboring regions Unlike the etching power for, the material of the protective layer is selected from a plurality of crystal materials so that the material can be selectively removed,
The protective layer is present in the buffer structure;
The recycling method, wherein the protective layer is doped . A method of reusing a donor wafer after peeling off at least one useful layer of a material selected from a plurality of semiconductor materials, wherein the donor wafer comprises a substrate, a buffer structure, and before peeling. And the method comprises the step of removing material at a side of the donor wafer where the delamination has occurred, wherein one of the plurality of surfaces of the buffer structure is the same as the lattice constant of the substrate. And the other surface has a second lattice constant identical to the lattice constant of the useful layer, and at least a portion of the buffer structure remains after the material is removed, At least a portion of the buffer structure is reused as a buffer structure for peeling off the next useful layer, and after removing the material from the donor wafer, the side of the donor wafer where the material was removed To form a layer for reproducing the donor wafer,
There is a further protective layer in the donor wafer, at least a part of the buffer structure is in the lower part of the protective layer, and the etching power for the material of the protective layer of the substance removing means is at least one material in two neighboring regions Unlike the etching power for, the material of the protective layer is selected from a plurality of crystal materials so that the material can be selectively removed,
The protective layer is present on the buffer structure;
The recycling method, wherein the protective layer is doped . The material removing step is a method of recycling according to any one of claims 1 to 6, characterized in that it comprises a chemical etching process. The recycling method according to claim 7 , wherein the material removing process from the donor wafer includes a selective chemical etching process. An etch selectivity between the material being etched and the stop material that stops etching uses a determined etch species; and
The two materials are different, or
One of the two materials is doped, or
The two materials are the same, but the concentration of at least one atom in one material is different from the concentration of the same atom in the other material, or
The recycling method according to claim 8 , characterized in that it is obtained from the fact that the two materials have different porous densities. The material removing step is a method of recycling according to any one of claims 1 to 9, characterized in that it comprises the step of removing a portion of said buffer structure remaining after the peeling. The recycling method according to claim 1 , wherein the substance removing step includes a step of removing at least a part of the additional layer remaining after the peeling. Recycling method according to claim 1 or 9 wherein the material removing step is characterized by having a step of removing a portion of the buffer structure. 2. The method according to claim 1, wherein the donor wafer includes an upper layer having the useful layer to be peeled before peeling, and the material removing step includes a step of removing the remaining upper layer after peeling. The reuse method according to any one of 12 above. The upper layer is
(A) a material selected from the group consisting of SiGe and strained silicon;
(B) a material selected from the group consisting of AsGa and germanium;
(C) a group III-V element InP or other alloy;
The reuse method according to claim 13 , further comprising: (A) the substrate includes silicon, and the buffer structure includes a SiGe buffer layer having a germanium concentration that increases with thickness, and a relaxed SiGe layer on the buffer layer;
(B) The substrate contains AsGa, the buffer structure has a buffer layer, the buffer layer contains an alloy of three or more III-V elements, and the elements constituting the alloy are III Selected from the possible combinations of at least two further elements selected from the group consisting of the group and the group V, the two further elements having a concentration that varies gradually depending on the thickness of the buffer layer;
(C) The recycling method according to claim 1, wherein the donor wafer has at least one layer containing carbon having a carbon concentration of about 50% or less. The recycling method according to claim 6 , wherein a material of an upper region of the protective layer is selectively removed, and the protective layer becomes a stop layer for removing the substance. The recycling method according to claim 16 , wherein the material of the protective layer is selectively removed, and a lower region of the protective layer becomes a stop layer for the material removal. The recycling method according to any one of claims 1 to 17 , wherein the mechanical attack of the protective layer is performed in combination with selective chemical etching so that selective chemical mechanical planarization can be performed. Donor wafer according to any one of claims 1 to 18 wherein the protective layer is equal to or distorted telescopically by the lower layer of the protective layer.

Images