阅读说明：本技术 晶圆的减薄方法 (Wafer thinning method ) 是由 高博 黄伯宁 万玉喜 于 2021-07-06 设计创作，主要内容包括：本申请实施例提供一种晶圆的减薄方法,涉及半导体技术领域,用于解决SiC功率器件良率低、工艺复杂、制备成本高的问题。晶圆或者理解为复合衬底,包括层叠设置的第一碳化硅层、介质层以及第二碳化硅层。晶圆具有相对的第一侧和第二侧,第二碳化硅层远离第一碳化硅层的一侧为晶圆的第一侧。晶圆的减薄方法,包括：在第二侧将临时衬底载片与晶圆临时键合；从第一侧对晶圆进行激光照射,使激光的能量在第二碳化硅层与介质层的界面处进行聚焦烧蚀,使得第二碳化硅层与介质层分离。(The embodiment of the application provides a wafer thinning method, relates to the technical field of semiconductors, and is used for solving the problems of low yield, complex process and high preparation cost of SiC power devices. The wafer, or composite substrate as it is understood, comprises a first silicon carbide layer, a dielectric layer and a second silicon carbide layer arranged in a stack. The wafer has opposite first and second sides, and the side of the second silicon carbide layer away from the first silicon carbide layer is the first side of the wafer. The thinning method of the wafer comprises the following steps: temporarily bonding the temporary substrate carrier to the wafer at the second side; and carrying out laser irradiation on the wafer from the first side, so that the energy of the laser is focused and ablated at the interface of the second silicon carbide layer and the dielectric layer, and the second silicon carbide layer is separated from the dielectric layer.)

1. The method for thinning the wafer is characterized in that the wafer comprises a first silicon carbide layer, a dielectric layer and a second silicon carbide layer which are arranged in a stacked mode; the wafer is provided with a first side and a second side which are opposite, and the side of the second silicon carbide layer far away from the first silicon carbide layer is the first side of the wafer;

the thinning method of the wafer comprises the following steps:

temporarily bonding a temporary substrate carrier to the wafer at the second side;

and carrying out laser irradiation on the wafer from the first side, so that the energy of laser is focused and ablated at the interface of the second silicon carbide layer and the dielectric layer, and the second silicon carbide layer is separated from the dielectric layer.

2. The wafer thinning method according to claim 1, wherein the second silicon carbide layer has an absorption coefficient for laser light smaller than an intrinsic absorption coefficient.

3. The method for thinning the wafer according to claim 1, wherein the laser is nonlinearly absorbed at the interface of the second silicon carbide layer and the dielectric layer.

4. The method for thinning the wafer according to claim 1, wherein the bonding temperature of the temporary bonding is less than or equal to 300 ℃.

5. The wafer thinning method according to claim 1, wherein the melting point of the temporary substrate carrier is greater than the bonding temperature of the temporary bonding.

6. The method of thinning a wafer of claim 1, wherein temporarily bonding a temporary substrate carrier to the wafer at the second side comprises:

and temporarily bonding the temporary substrate slide and the wafer at the second side by adopting temporary bonding glue or paraffin.

7. The method for thinning the wafer according to claim 1, wherein the laser irradiation of the wafer from the first side comprises:

and irradiating the wafer from the first side by using infrared laser.

8. The method for thinning the wafer according to claim 1, further comprising:

and processing the surface of the first silicon carbide layer far away from the temporary substrate slide to remove the residues of the dielectric layer.

9. The method for thinning the wafer according to claim 8, wherein the step of treating the surface of the first silicon carbide layer away from the temporary substrate carrier comprises:

and processing the surface of the first silicon carbide layer far away from the temporary substrate slide by at least one of wet etching, dry etching or cleaning.

10. The method for thinning the wafer according to any one of claims 1 to 9, wherein the second side of the wafer is further provided with a switch function component;

temporarily bonding a temporary substrate carrier to the wafer at the second side, comprising:

and temporarily bonding the temporary substrate slide with the switch functional component.

Technical Field

The application relates to the technical field of semiconductors, in particular to a wafer thinning method.

Background

Because the silicon carbide (SiC) material has the excellent physical characteristics of wide forbidden band width, high critical breakdown field strength, large heat conductivity and the like, the SiC power device has the characteristics of high pressure resistance, high temperature resistance, high switching speed, small switching loss and the like, and has wide application in the fields of aerospace aviation, smart power grids, rail transit, new energy power generation, electric automobiles, industrial power supplies and the like.

In order to reduce the on-resistance of the SiC power device, the performance of the device is improved. In general, after a film layer structure of a SiC power device is prepared, a SiC substrate needs to be thinned. However, on one hand, a large amount (for example, approximately 200-300 um) of SiC is purely grinded and wasted, and on the other hand, because the hardness of the SiC material almost reaches the level of diamond, the traditional mechanical thinning rate is low, the grinding head damage is large, and the risk of cracking of the SiC substrate is extremely high. Therefore, the preparation cost and the sale price of the SiC power device are high, and the popularization and the application of the SiC power device in various fields are seriously limited.

Disclosure of Invention

The embodiment of the application provides a method for thinning a wafer, which is used for solving the problems of low yield, complex process and high preparation cost of SiC power devices.

In order to achieve the purpose, the technical scheme is as follows:

in a first aspect, a method for thinning a wafer is provided, which can be applied to thinning a composite substrate in a semiconductor assembly. The wafer or composite substrate comprises a first silicon carbide layer, a dielectric layer and a second silicon carbide layer which are arranged in a stacked mode; the wafer is provided with a first side and a second side which are opposite, and the side of the second silicon carbide layer far away from the first silicon carbide layer is the first side of the wafer; the thinning method of the wafer comprises the following steps: temporarily bonding the temporary substrate carrier to the wafer at the second side; and carrying out laser irradiation on the wafer from the first side, so that the energy of the laser is focused and ablated at the interface of the second silicon carbide layer and the dielectric layer, and the second silicon carbide layer is separated from the dielectric layer.

According to the thinning method of the wafer provided by the embodiment of the application, the wafer is of a sandwich laminated structure, the sandwich laminated structure comprises a dielectric layer, and the refractive indexes of the dielectric layer and the second silicon carbide layer are different. Therefore, after laser light is irradiated into the wafer from the surface of the second silicon carbide layer far away from the first silicon carbide layer, the laser light is easily focused and absorbed at the interface of the dielectric layer and the second silicon carbide layer. The second silicon carbide layer can be stripped from the first silicon carbide layer by increasing the laser energy to reach the melting point of the dielectric layer, so that the wafer is thinned. The stripped second silicon carbide layer can be recycled for preparing a new wafer. Therefore, the wafer is thinned by the wafer thinning method provided by the embodiment of the application, and the preparation cost of the semiconductor device can be reduced.

In one possible embodiment, the second silicon carbide layer has an absorption coefficient for the laser light that is less than the intrinsic absorption coefficient. The second silicon carbide layer has the characteristic of low absorption coefficient or transparency to laser, so that the loss of laser energy of the second silicon carbide layer is reduced, and the focusing ablation effect is improved.

In one possible embodiment, the laser light is non-linearly absorbed at the interface of the second silicon carbide layer and the dielectric layer.

In one possible embodiment, the bonding temperature of the temporary bond is less than or equal to 300 ℃. In this way, it is avoided that the temporary bonding process is too hot, which may affect the switching function in the semiconductor device (e.g. cause melting of the metal structure in the switching function).

In one possible embodiment, the melting point of the temporary substrate carrier is greater than the bonding temperature of the temporary bond. Therefore, the situation that the temporary substrate carrier is dissolved and cannot support the semiconductor assembly in the temporary bonding process can be avoided.

In one possible embodiment, temporarily bonding a temporary substrate carrier to the wafer at a second side comprises: temporarily bonding a temporary substrate carrier to the wafer at the second side with a temporary bonding glue. The process is mature and the cost is low.

In one possible embodiment, laser irradiation of a wafer from a first side comprises: the wafer is laser irradiated from the first side with infrared laser. The process is mature and the cost is low.

In a possible embodiment, the method for thinning the wafer further includes: and processing the surface of the first silicon carbide layer far away from the temporary substrate slide to remove the residue of the dielectric layer. The surface of the first silicon carbide layer far away from the temporary substrate slide glass is treated, so that the surface of the first silicon carbide layer far away from the temporary substrate slide glass meets the requirements of roughness and the like, and the subsequent manufacture of structures such as metal electrodes is facilitated.

In one possible embodiment, treating a surface of the first silicon carbide layer remote from the temporary substrate carrier includes: and processing the surface of the first silicon carbide layer far away from the temporary substrate slide by at least one of wet etching, dry etching or cleaning.

In a possible embodiment, the second side of the wafer is further provided with a switch function component; temporarily bonding a temporary substrate carrier to the wafer at a second side, comprising: and temporarily bonding the temporary substrate slide with the switch functional component. The thinning process is performed after the switch functional components are formed on the wafer. Therefore, the wafer thinning method provided by the embodiment of the application can be compatible with a high-temperature process in the semiconductor device preparation process.

Drawings

Fig. 1 is an application scenario diagram of an uninterruptible power supply system according to an embodiment of the present application;

fig. 2 is a schematic structural diagram of a metal-oxide semiconductor field effect transistor according to an embodiment of the present application;

fig. 3 is a diagram of a thinning process of a SiC homoepitaxial layer provided in an embodiment of the present application;

FIG. 4 is a diagram illustrating another process of thinning a SiC homoepitaxial layer according to an embodiment of the present disclosure;

fig. 5A is a schematic structural diagram of a semiconductor device according to an embodiment of the present disclosure;

fig. 5B is a schematic structural diagram of another semiconductor device according to an embodiment of the present disclosure;

fig. 6A is a flowchart of a method for thinning a wafer according to an embodiment of the present disclosure;

fig. 6B is a process diagram of a method for thinning a wafer according to an embodiment of the present disclosure;

fig. 7A is a flowchart of another method for thinning a wafer according to an embodiment of the present disclosure;

fig. 7B is a process diagram of another wafer thinning method according to an embodiment of the present disclosure;

fig. 8A is a structural diagram of a semiconductor device according to an embodiment of the present disclosure after a wafer is thinned;

fig. 8B is a process diagram of a semiconductor device according to an embodiment of the present disclosure in a subsequent process.

Reference numerals:

1-a substrate; 2-a semiconductor layer; 3-a well region; a 4-source region; 5-a contact zone; 6-insulating film 7-interlayer insulating film; 10-a first silicon carbide layer; 20-a dielectric layer; 30-a second silicon carbide layer; 41-a first bonding layer; 42-a transition layer; 43-a second bonding layer; 44-a third bonding layer; 50-temporary substrate slide; 100-a wafer; 200-a switch function; 300-a switch function component; 1000-semiconductor component.

Detailed Description

The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments.

Hereinafter, the terms "first", "second", and the like are used for descriptive convenience only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first," "second," etc. may explicitly or implicitly include one or more of that feature. In the description of the present application, "a plurality" means two or more unless otherwise specified.

In the embodiments of the present application, unless otherwise specifically stated or limited, the term "electrically connected" may be a direct electrical connection or an indirect electrical connection through an intermediate.

In the embodiments of the present application, words such as "exemplary" or "for example" are used to mean serving as an example, instance, or illustration. Any embodiment or design described herein as "exemplary" or "e.g.," is not necessarily to be construed as preferred or advantageous over other embodiments or designs. Rather, use of the word "exemplary" or "such as" is intended to present concepts related in a concrete fashion.

In the embodiment of the present application, "and/or" describes an association relationship of an associated object, indicating that three relationships may exist, for example, a and/or B may indicate: a exists alone, A and B exist simultaneously, and B exists alone, wherein A and B can be singular or plural. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship.

In the embodiments of the present application, directional indications such as up, down, left, right, front, and rear, etc., used for explaining the structure and movement of the various components in the present application, are relative. These indications are appropriate when the components are in the positions shown in the figures. However, if the specification of the position of the element changes, then the direction indications will change accordingly.

The embodiment of the application provides an electronic device, which can be, for example, a charging pile, an Uninterruptible Power System (UPS), a photovoltaic inverter, a motor drive power supply, and the like. The embodiment of the present application does not particularly limit the specific form of the electronic device.

A metal-oxide semiconductor field effect transistor (MOSFET) device is a semiconductor device, and has the advantages of low power consumption, stable performance, strong radiation resistance, convenient control mode, small size, light weight, long service life, strong anti-interference capability, high operating frequency, simple bias, and the like, so that the MOSFET device is widely used in analog circuits and digital circuits.

Taking an electronic device as an example for the USP, the UPS is a component for supplying power to a load such as a computer that requires a continuous supply of electric power. As shown in fig. 1, a schematic diagram of a UPS and its peripheral structure is illustrated. The UPS comprises an input end and an output end, wherein the input end of the UPS is connected with the power system, and the output end of the UPS is connected with the load so as to realize uninterrupted power supply to the load.

The power system may be, for example, a power plant, a substation, a utility transmission line, etc. In a normal state of the power system, a portion of power supplied by the power system is transmitted to a load via the UPS, and a portion of power supplied by the power system is stored in the UPS. In the abnormal state of the power system, the power system cannot transmit power to the load, and at this time, the power stored in the UPS is transmitted to the load.

A load consumes power supplied from a power system, and the load may be, for example, an electrical device in a plant; the load may also be a server, a processor, a memory, etc. of the data center.

A UPS is an automatic system configured to supply power immediately without interruption in the event of a power interruption or failure supplied by a power system. The UPS stably supplies power if the voltage or frequency of power supplied from the power system varies, or the supply of power from the power system is momentarily interrupted or changed, thereby reducing the possibility of damage, loss, or deletion of load data, and reducing the possibility of shutdown or malfunction of the control apparatus.

In which the UPS includes components such as a power device and a bidirectional switch for implementing the above functions of the USP, and the MOSFET device can be used as the power device in the above USP. It should be understood that when the MOSFET device is used as a power device, the electronic device provided by the embodiment of the present application is not limited to the USP shown in fig. 1, and any electronic device that needs to use a power device belongs to the application scenario of the embodiment of the present application.

A structure of a MOSFET device is exemplified, and as shown in fig. 2, the MOSFET device includes a substrate 1, a semiconductor layer 2, a well region 3, a source region 4, a contact region 5, an insulating film 6, a gate (gate, G), a source (S), an interlayer insulating film 7, and a drain (D).

Among them, the semiconductor layer 2, the well region 3, the source region 4, the contact region 5, the insulating film 6, the gate electrode G, the source electrode S, the interlayer insulating film 7, and the drain electrode D may be referred to as a switching function part 200 of the MOSFET device. As the name implies, the switching function 200 of the MOSFET device is a combination of structures for implementing the switching function of the MOSFET device, and the substrate 1 is used to carry the switching function 200 of the MOSFET device.

Taking the MOSFET device shown in fig. 2 as an example, the switching functional unit 200 of the MOSFET device includes structures located on two opposite sides of the substrate 1, and in the embodiment of the present application, the combination of the structures located on the same side in the switching functional unit 200 and including the semiconductor layer 2 is referred to as a switching functional component 300. Taking fig. 2 as an example, the switching functional element 300 includes a semiconductor layer 2, a well region 3, a source region 4, a contact region 5, an insulating film 6, a gate electrode G, a source electrode S, and an interlayer insulating film 7. In the embodiment of the present application, a structure in which the switching functional element 300 is provided over the substrate 1 is referred to as a semiconductor element 1000.

The working principle of the MOSFET device is as follows: the region in the well region 3 between the source region 4 and the semiconductor layer 2 serves as a conduction channel of the MOSFET device. When the voltage of the grid G is larger than the threshold voltage of the MOSFET device, the conductive channel is opened, and electrons of the source S flow through the channel under the action of the voltage of the drain D, flow downwards to the substrate 1 and reach the drain D to form source-drain current. When the voltage of the grid electrode G is smaller than the threshold voltage of the MOSFET device, the conducting channel is turned off, and the source-drain current is immediately turned off.

Because the silicon carbide (SiC) material has the excellent physical characteristics of wide forbidden band width, high critical breakdown field strength, large heat conductivity and the like, the power device adopting SiC as the substrate has the characteristics of high pressure resistance, high temperature resistance, high switching speed, small switching loss and the like, and has wide application in the fields of aerospace, smart power grids, rail transit, new energy power generation, electric automobiles, industrial power supplies and the like.

In order to realize the reduction of the thickness of the SiC substrate, the on-resistance of the SiC power device is reduced, and the performance of the device is improved.

A method for thinning a homogeneous epitaxial layer of SiC is provided, as shown in FIG. 3, a laser beam transparent to single crystal SiC is irradiated to an ingot of SiC and focused on an inner layer, thereby forming a modified layer. The SiC wafer is grown from the single crystal SiC ingot by cutting along the modified layer under stress.

By the thinning method of the SiC homogeneous epitaxial layer, the cutting loss of the SiC crystal ingot can be effectively reduced, the cutting efficiency of the SiC crystal ingot is improved, and the cost of the SiC substrate is reduced.

However, the above method for thinning the SiC homogeneous epitaxial layer mainly focuses on reducing the cutting loss of the SiC ingot and improving the cutting efficiency, does not involve peeling and recycling of the SiC wafer, and cannot solve the problem of substrate waste caused by thinning the SiC substrate.

Also provided is a method for thinning a SiC homoepitaxial layer, as shown in FIG. 4, a defect layer is manufactured near the surface of the SiC substrate by an ion implantation method; repairing defects on the surface of the SiC substrate caused by ion implantation through annealing treatment; after an epitaxial layer is epitaxially grown on the surface of the SiC substrate in a homoepitaxy mode, laser irradiates the SiC substrate, is focused below the defect layer, and the SiC substrate is separated from the defect layer to be thinned.

By the thinning method of the homogeneous epitaxial layer, the stripped residual SiC substrate can be used for growing the SiC epitaxial layer again after being polished, so that the utilization rate of the SiC substrate can be effectively improved, and the production cost of the epitaxial layer is reduced.

However, in the above method for thinning the SiC homogeneous epitaxial layer, since the SiC substrate undergoes epitaxial growth before laser lift-off, the defect layer induced by ion implantation is annealed and repaired in the epitaxial high-temperature growth process. Then, it is difficult for only the remaining defect layer to achieve laser absorption at the time of subsequent laser irradiation, so that the success rate of separation of the SiC substrate from the defect layer is very low. Therefore, the method for thinning the SiC homogeneous epitaxial layer cannot be compatible with a high-temperature process.

In order to solve the above problems, an embodiment of the present application further provides a method for thinning a wafer, and by the method for thinning a wafer, the problems of small mechanical thinning rate, large damage to a grinding head, and great risk of breakage of a SiC substrate in the above process for performing round thinning on the SiC substrate can be solved, and the problem that the process cannot be compatible with a high temperature process can also be solved.

Next, a description will be given of an example in which the wafer thinning method provided in the embodiments of the present application is applied to a semiconductor device, and first, a structure of the semiconductor device will be briefly described.

As shown in fig. 5A, the semiconductor device 1000 includes a wafer 100 (or a composite substrate referred to as a semiconductor device 1000) and a switching function device 300 disposed on the wafer 100. The wafer 100 includes a first silicon carbide layer 10, a dielectric layer 30, and a second silicon carbide layer 30, which are stacked, the wafer 100 has a first side and a second side opposite to each other, a side of the second silicon carbide layer 30 away from the first silicon carbide layer 10 is a first side of the wafer, and the switch function component 300 is disposed on a side of the first silicon carbide layer 10 away from the second silicon carbide layer 30 (a second side of the wafer 100). The semiconductor module 1000 has a first surface a1 and a second surface a2 opposite to each other, and the surface of the second silicon carbide layer 30 away from the first silicon carbide layer 10 is the first surface a1 of the semiconductor module 1000. That is, the surface of the wafer 100 away from the switch function module 300 is the first surface a1 of the semiconductor device 1000, and the surface of the switch function module 300 away from the wafer 100 is the second surface a2 of the semiconductor device 1000.

In some embodiments, the first silicon carbide layer 10 is a high quality (P-grade) silicon carbide structure relative to the second silicon carbide layer 30, and the second silicon carbide layer 30 is a low quality (D-grade) silicon carbide structure relative to the first silicon carbide layer 10. Illustratively, the second silicon carbide layer 30 has a local or global defect density that is greater than the defect density of the first silicon carbide layer 10.

In some embodiments, the lattice direction of the first silicon carbide layer 10 is deflected in an interval greater than or equal to 0 ° and less than or equal to 8 ° along the <0001> lattice direction.

For example, the first silicon carbide layer 10 is not deflected (deflected by 0 °) in the <0001> lattice direction, the first silicon carbide layer 10 is deflected by 1 ° in the <0001> lattice direction, the first silicon carbide layer 10 is deflected by 2 ° in the <0001> lattice direction, the first silicon carbide layer 10 is deflected by 3 ° in the <0001> lattice direction, the first silicon carbide layer 10 is deflected by 4 ° in the <0001> lattice direction, the first silicon carbide layer 10 is deflected by 5 ° in the <0001> lattice direction, the first silicon carbide layer 10 is deflected by 6 ° in the <0001> lattice direction, and the first silicon carbide layer 10 is deflected by 7 ° in the <0001> lattice direction.

In some embodiments, the thickness of the first silicon carbide layer 10 is made less than or equal to 350 um.

For example, the thickness of the first silicon carbide layer 10 is 300um, 250um, 200um, 150um, 100 um.

In some embodiments, the material of the first silicon carbide layer 10 is single crystal silicon carbide.

In some embodiments, the second silicon carbide layer 30 is also deflected in the interval greater than or equal to 0 ° and less than or equal to 8 ° along the <0001> lattice direction.

Illustratively, the second silicon carbide layer 30 has a lattice direction deflection angle that is the same as the lattice direction deflection angle of the first silicon carbide layer 10.

In some embodiments, the thickness of the second silicon carbide layer 30 is less than or equal to 3000 um.

For example, the thickness of the second silicon carbide layer 30 is 2500um, 2000um, 1500um, 1000um, 500um, etc.

The second silicon carbide layer 30 may support the first silicon carbide layer 10, and the thickness of the second silicon carbide layer 30 is too thick, which may cause waste of resources and increase the thickness of the wafer 100.

In some embodiments, the material of the second silicon carbide layer 30 comprises single crystal silicon carbide or polycrystalline silicon carbide.

In some embodiments, as shown in fig. 5A, the first silicon carbide layer 10 is bonded in direct contact with the second silicon carbide layer 30.

Based on this, it is understood that a first bonding layer 41 is naturally formed during the direct contact bonding of the first silicon carbide layer 10 and the second silicon carbide layer 30. By controlling the bonding time of the first silicon carbide layer 10 and the second silicon carbide layer 30, the thickness of the first bonding layer 41 can be controlled. In this case, the dielectric layer 20 in the wafer 100 is the first bonding layer 41 naturally formed during the direct contact bonding of the first silicon carbide layer 10 and the second silicon carbide layer 30.

In some embodiments, the dielectric layer 20 has a thickness of less than or equal to 5 nm. For example, the dielectric layer 20 has a thickness of 1nm, 2nm, 3nm, or 4 nm.

Since the thickness of the dielectric layer 20 is too large, it has an effect on both the electrical and thermal conductivity properties of the wafer 100. Therefore, the thickness of the dielectric layer 20 is controlled to be less than or equal to 5nm, and the thickness of the dielectric layer 20 is reduced as much as possible on the basis of ensuring the direct contact bonding of the first silicon carbide layer 10 and the second silicon carbide layer 30.

In other embodiments, as shown in fig. 5B, the first silicon carbide layer 10 and the second silicon carbide layer 30 are bonded through the transition layer 42.

The material of the transition layer 42 may include, for example, an insulating medium such as SiO2 (silicon oxide), Si3N4 (silicon nitride), and Al2O3 (aluminum oxide), or a conductive medium such as Si (silicon) and SiC (silicon carbide), or a metal such as Al (aluminum), Cu (copper), Pt (platinum), Ni (nickel), Ti (titanium), Au (gold), and Cr (chromium), or a composite multilayer material including these materials.

In the process of bonding the first silicon carbide layer 10 and the second silicon carbide layer 30 through the transition layer 42, the second bonding layer 43 is naturally formed when the first silicon carbide layer 10 is bonded to the transition layer 42. The second silicon carbide layer 30, when bonded to the transition layer 42, will naturally form a third bonding layer 44.

Based on this, it can be understood that the dielectric layer 20 in the wafer 100 includes the second bonding layer 43, the transition layer 42, and the third bonding layer 44.

In some embodiments, the thickness of the dielectric layer 20 (the sum of the thicknesses of the transition layer 42, the second bonding layer 43, and the third bonding layer 44) is less than or equal to 100 nm. For example, the sum of the thicknesses of the transition layer 42, the second bonding layer 43, and the third bonding layer 44 is 90nm, 80nm, 70nm, and 60 nm.

Since the sum of the thicknesses of the transition layer 42, the second bonding layer 43 and the third bonding layer 44 is too large, both the electrical and thermal conductivity properties of the wafer are affected. Therefore, the sum of the thicknesses of the transition layer 42, the second bonding layer 43, and the third bonding layer 44 is controlled to be less than 100nm, and the sum of the thicknesses of the transition layer 42, the second bonding layer 43, and the third bonding layer 44 is minimized on the basis of ensuring stable bonding of the back surface a1 of the first silicon carbide layer 10 and the front surface b2 of the second silicon carbide layer 30.

As shown in fig. 6A, taking the second side of the wafer 100 further provided with the switch functional element 300 as an example, the method for thinning the wafer according to the embodiment of the present application is described, where the method for thinning the wafer includes:

s10, as shown in fig. 6B, the temporary substrate slide 50 is temporarily bonded to the wafer 100 at the second side of the wafer 100.

In the case where the second side of the wafer 100 is provided with the switching function assembly 300, the temporary substrate carrier 50 is temporarily bonded to the switching function assembly 300 located at the second side of the wafer 100 to achieve temporary bonding of the temporary substrate carrier 50 to the wafer 100. That is, the temporary substrate chip 50 is temporarily bonded to the switching function module 300 at the second surface a2 of the semiconductor module 1000.

That is, the temporary substrate carrier 50 is located on the side of the switch function assembly 300 away from the wafer 100, and the temporary substrate carrier 50 is temporarily bonded to the semiconductor assembly 1000.

Temporary bonding, as used herein, is understood to mean bonding that can be restored. The temporarily bonded temporary substrate slides 50 may be debonded from the semiconductor assembly 1000 during subsequent processing as needed.

In some embodiments, the bonding temperature for temporarily bonding the temporary substrate slide 50 to the wafer 100 is less than or equal to 300 ℃.

In this way, it is avoided that the temporary bonding process temperature is too high to affect the above-mentioned switch function module 300 in the semiconductor assembly 1000 (e.g. to melt the metal structure in the switch function module 300).

The material of the temporary substrate carrier 50 is not limited by the embodiments of the present application, and in some embodiments, the melting point of the temporary substrate carrier 50 is greater than the bonding temperature of the temporary bonding.

In this way, it is avoided that the temporary substrate carrier 50 dissolves during the temporary bonding process and cannot support the semiconductor assembly 1000.

With respect to the method of temporarily bonding the temporary substrate slide 50 to the wafer 100, in some embodiments, the temporary substrate slide 50 is temporarily bonded to the wafer at the second side of the wafer 100 using a temporary bonding glue or wax.

That is, in some embodiments, the temporary substrate slide 50 is temporarily bonded to the semiconductor assembly 1000 at the second surface a2 of the semiconductor assembly 1000 using a temporary bonding paste.

In other embodiments, the temporary substrate slide 50 is temporarily bonded to the semiconductor package 1000 at the second surface a2 of the semiconductor package 1000 using paraffin wax.

S20, as shown in fig. 6B, laser irradiation is performed on the wafer 100 from the first side of the wafer 100, so that the energy of the laser is focused and ablated at the interface between the second silicon carbide layer 30 and the dielectric layer 20, so that the second silicon carbide layer 30 is separated from the dielectric layer 20.

That is, the wafer 100 is irradiated with laser light from the first surface a1 of the semiconductor device 1000, so that the energy of the laser light is focused and ablated at the interface between the second silicon carbide layer 30 and the dielectric layer 20, so that the second silicon carbide layer 30 is separated from the dielectric layer 20.

Among them, when light propagates through a medium, a phenomenon in which the intensity of light attenuates with the propagation distance (penetration depth) is referred to as absorption of light. The absorption of light follows the law of absorption (Beer-Lambert law), the absorption coefficient being a constant in Beer-Lambert law (Beer-Lambert law), denoted α, and is called the absorption coefficient of the medium for this monochromatic light. The greater the absorption coefficient, the more significant the attenuation of the light.

The intrinsic absorption coefficient is the optical absorption coefficient of the optical wavelength of which the photon energy corresponds to the forbidden bandwidth of the medium in the medium, the optical absorption coefficient of the optical wavelength of which the photon energy is greater than the forbidden bandwidth of the medium is greater than the intrinsic absorption coefficient, and the optical absorption coefficient of the optical wavelength of which the photon energy is less than the forbidden bandwidth of the medium is less than the intrinsic absorption coefficient.

Therefore, the second silicon carbide layer 30 has a low absorption coefficient or is transparent to laser light, so that the loss of laser energy in the second silicon carbide layer 30 is reduced, and the effect of focus ablation is improved.

Illustratively, the absorption coefficient of the second silicon carbide layer 30 for the laser light may be made smaller than the intrinsic absorption coefficient by adjusting the wavelength of the laser light selected.

In some embodiments, the wafer 100 is laser irradiated from a first side of the wafer 100 using an infrared laser, an ultraviolet laser, or the like.

That is, the wafer 100 is irradiated with laser light from the first surface a1 of the semiconductor module 1000 using infrared laser light, ultraviolet laser light, or the like.

In some embodiments, the laser light is non-linearly absorbed at the interface of the second silicon carbide layer 30 and the dielectric layer 20.

Nonlinear absorption refers to the phenomenon that under the action of strong light, the absorption coefficient of a medium to light is far larger than that of the medium.

Based on this, after the energy of the laser is focused and ablated at the interface of the second silicon carbide layer 30 and the dielectric layer 20, the dielectric layer 20 is naturally separated from the second silicon carbide layer 30, so that the second silicon carbide layer 30 is separated from the first silicon carbide layer 10 in the wafer 100.

The second silicon carbide layer 30 separated from the wafer 100 may be reused after being cleaned, and used to manufacture a new wafer 100.

Considering that the energy of the laser is focused and ablated at the interface of the second silicon carbide layer 30 and the dielectric layer 20, there may be residues of the dielectric layer 20 on the surface of the first silicon carbide layer 10 away from the switch functional component 300.

In some embodiments, as shown in fig. 7A, the method for thinning a wafer further includes:

s30, as shown in fig. 7B, the surface of the first silicon carbide layer 10 remote from the temporary substrate slide 50 is treated to remove the residue of the dielectric layer 20.

In some embodiments, the surface of the first silicon carbide layer 10 remote from the temporary substrate carrier 50 is treated by at least one of wet etching, dry etching, or cleaning to remove the residue of the dielectric layer 20.

After the residue of the dielectric layer 20 is removed, the thinning of the silicon wafer of the wafer 100 is completed. And entering the subsequent process flow according to the requirement.

In some embodiments, after treating the surface of the first silicon carbide layer 10 away from the temporary substrate 50, the roughness of the surface of the first silicon carbide layer 10 away from the temporary substrate 50 is less than or equal to 0.5 nm.

In this way, formation of other structures on the surface of the first silicon carbide layer 10 remote from the temporary substrate 50 may be facilitated.

Taking the MOSFET device shown in fig. 2 as an example, the structure obtained by thinning the wafer 100 and surface-cleaning the first silicon carbide layer 10 includes the first silicon carbide layer 10 and the switching functional element 300 as shown in fig. 8A. Subsequent processes, such as forming a metal electrode (drain D), de-bonding the temporary substrate carrier 50 to the semiconductor assembly 1000, and the like, may be performed to finally complete the semiconductor device, as shown in fig. 8B.

The surface of the first silicon carbide layer 10 far away from the temporary substrate slide 50 is treated, so that the surface of the first silicon carbide layer 10 far away from the temporary substrate slide 50 meets the requirements of roughness and the like, and the subsequent manufacture of structures such as metal electrodes is facilitated.

The following describes a method for thinning a composite substrate-based wafer in a semiconductor device according to an embodiment of the present application, with two specific examples.

Illustratively, in the semiconductor assembly 1000, the wafer 100 includes a 2um thick high quality single crystal first silicon carbide layer 10, a 20nm thick SiO2 dielectric layer 20, and a 350um thick low quality single crystal second silicon carbide layer 30.

And selecting an Al2O3 substrate with the same size as the temporary substrate slide 50, and temporarily bonding the temporary substrate slide 50 and the functional device 200 of the semiconductor assembly 1000 at the bonding temperature of 150 ℃.

Scanning irradiation is carried out on the surface (the surface of the second silicon carbide layer 30) of the wafer 100 far away from the functional device 200 by using laser with the wavelength of 1064nm, and the laser energy is nonlinearly absorbed and focused ablated at the interface of the second silicon carbide layer 30 and the SiO2 medium layer 20, so that the second silicon carbide layer 30 in the wafer 100 is stripped from the SiO2 medium layer 20, and the wafer 100 is thinned.

And corroding and cleaning the surface of the first silicon carbide layer 10, removing the residue of the SiO2 dielectric layer 20, and entering the subsequent process flow.

The second silicon carbide layer 30 thus peeled off is etched and cleaned, and is reused to produce the wafer 100.

Alternatively, in the illustrated semiconductor assembly 1000, the wafer 100 includes a 2um thick high quality single crystal first silicon carbide layer 10, a 20nm thick Pt dielectric layer 20, and a 350um thick low quality single crystal second silicon carbide layer 30.

The Si substrate with the same size is selected as the temporary substrate slide 50, and the temporary substrate slide 50 and the functional device 200 of the semiconductor assembly 1000 are temporarily bonded, wherein the bonding temperature is 210 ℃.

Scanning irradiation is carried out on the surface (the surface of the second silicon carbide layer 30) of the wafer 100 far away from the functional device 200 by using infrared laser with the wavelength of 980nm, nonlinear absorption and focused ablation are carried out on the laser energy at the interface of the second silicon carbide layer 30 and the SiO2 medium layer 20, the second silicon carbide layer 30 in the wafer 100 is stripped from the SiO2 medium layer 20, and therefore the wafer 100 is thinned.

And corroding and cleaning the surface of the first silicon carbide layer 10, removing the residue of the SiO2 dielectric layer 20, and entering the subsequent process flow.

The second silicon carbide layer 30 thus peeled off is etched and cleaned, and is reused to produce the wafer 100.

In the thinning method of the wafer according to the embodiment of the present application, in the semiconductor device 1000, the wafer 100 has a sandwich laminated structure, which includes the dielectric layer 20, and the refractive indexes of the dielectric layer 20 and the second silicon carbide layer 30 are different. Therefore, after laser light is irradiated into the wafer 100 from the surface of the second silicon carbide layer 30 away from the first silicon carbide layer 10, the laser light is easily absorbed in focus at the interface between the dielectric layer 20 and the second silicon carbide layer 30. By increasing the laser energy to reach the melting point of the dielectric layer 20, the second silicon carbide layer 30 can be peeled off from the first silicon carbide layer 10, so that the wafer 100 can be thinned. The stripped second silicon carbide layer 30 may be recycled for use in preparing a new wafer 100. Therefore, the composite substrate in the semiconductor assembly is thinned by the wafer thinning method provided by the embodiment of the application, so that the preparation cost of the semiconductor device can be reduced.

In addition, the thinning process of the wafer 100 is performed after the formation of the switch function component 300. Therefore, the wafer thinning method provided by the embodiment of the application can be compatible with a high-temperature process in the semiconductor device preparation process.

The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.