JP2010010317A - Method of manufacturing diffused wafer, and diffused wafer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method of manufacturing a diffused wafer for spin-etching a diffused wafer having a non-diffused layer and a diffused layer, thereby removing part of the non-diffused layer to make the thickness of the non-diffused layer almost uniform. <P>SOLUTION: The method includes the spin-etching step of removing the part 23 of the non-diffused layer 21 of the diffused wafer 20 having the non-diffused layer 21 and the diffused layer 22 by sheet spin-etching. The spin-etching step uses a spin-etching device movable in the in-plane direction D2 of the diffused wafer 20 and having a liquid supply port for supplying an etchant to the non-diffused layer 21 of the diffused wafer 20. The liquid supply port is positioned respectively on an in-plane peripheral part 20c and an in-plane central part 20d of the diffused wafer 20 to control the wafer in-plane distribution of an etched-off amount of the non-diffused layer 21 based on the thickness-oriented shape of the non-diffused layer 21 when performing the spin-etching. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a method for manufacturing a diffusion wafer having a non-diffusion layer and a diffusion layer, and a diffusion wafer.

  In order to manufacture power devices for discrete applications such as high-power transistors, diodes, and rectifiers, a diffusion wafer having two layers of a low resistivity diffusion layer and a high resistivity non-diffusion layer in which a high concentration of dopant is diffused. Is used.

The thickness of the non-diffusion layer in the diffusion wafer affects the characteristics of the semiconductor device. Therefore, the non-diffusion layer is required to be formed with high accuracy so that the thickness does not vary between wafers and does not vary in the in-plane direction of the wafer (direction perpendicular to the wafer thickness direction). The For example, since the drain-source breakdown voltage (VDSS) characteristic and the thickness of the non-diffusion layer are highly correlated, in order to obtain a semiconductor device having a uniform VDSS characteristic in the in-plane direction, the thickness is uniform in the in-plane direction. A diffusion wafer having a non-diffusion layer is required.
By the way, the non-diffusion layer in the diffusion wafer is etched by, for example, single wafer spin etching (see, for example, Patent Document 1 below).

  An example of a conventional diffusion wafer manufacturing method will be described with reference to FIGS. FIG. 11 is a flowchart showing an example of a conventional diffusion wafer manufacturing method. 12A and 12B are cross-sectional views showing a procedure for obtaining two diffusion wafers from one silicon wafer.

As shown in FIG. 11, an example of a conventional method for manufacturing a diffusion wafer includes the following steps S1 to S10.
(S1) Slicing Step A silicon ingot is sliced to obtain a silicon wafer (hereinafter also simply referred to as “wafer”).

(S2) Chamfering process A chamfering process (beveling) is performed on the wafer that has undergone the slicing process S1. Specifically, the edge is rounded by applying a rotating grindstone to the edge of the side end surface of the wafer. Thereby, the crack of a wafer, chipping prevention, etc. are achieved.

(S3) Lapping process The wafer that has undergone the chamfering process S2 is set in a lapping apparatus, and lapping is performed on the surface of the wafer. Thereby, damages such as scratches and irregularities on the surface of the wafer caused by slicing are removed and the wafer is flattened.

(S4) Alkaline Etching Step The wafer flattened through the lapping step S3 is immersed in an alkaline etching solution, and the wafer is subjected to alkali etching. Thereby, the processing strain layer on the surface of the wafer is removed.

(S5) Diffusion process The wafer that has undergone the alkali etching process S4 is loaded into a boat and exposed to an atmosphere (such as in a diffusion furnace) where the temperature is high and the dopant exists in that state for a predetermined time. As a result, as shown in FIG. 12A, diffusion layers 122 and 122 are formed on the wafer 110 on both surface sides of the wafer 110 with the non-diffusion layer 121 interposed therebetween.

(S6) Two-divided process As shown in FIGS. 12A and 12B, the wafer 110 that has undergone the diffusion process S5 is moved along the center line in the thickness direction D1 (the one-dot chain line in FIG. 12A). The wafer 110 is cut into two in the thickness direction D1 by cutting with a wire saw or the like. Thereby, two diffusion wafers 120 are obtained from one wafer 110. The diffusion wafer 120 has a non-diffusion layer 121 and a diffusion layer 122, respectively.

(S7) Surface Grinding Step The diffusion wafer 120 formed in the two-part dividing step S6 is set on a surface grinding machine, and a grinding wheel is pressed against the surface 121a of the non-diffusion layer 121 of the diffusion wafer 120 so that the surface of the diffusion wafer 120 is flat. Apply grinding. Thereby, the surface 121a of the non-diffusion layer 121 of the diffusion wafer 120 is planarized.
Instead of performing the two-divided step S6, instead, the wafer 110 can be surface ground by a surface grinder to obtain the diffusion wafer 120 in which the surface 121a of the non-diffusion layer 121 is planarized.

(S8) Chamfering process Chamfering (beveling) is performed on the diffusion wafer 120 that has undergone the surface grinding process S7. Specifically, the edge is rounded by applying a rotating grindstone to the edge of the side end face of the diffusion wafer 120. Thereby, cracking of the diffusion wafer 120, prevention of dust generation, and the like are achieved.

(S9) Spin Etching Step Etching solution is supplied to each part of the surface 121a of the non-diffusion layer 121 in the diffusion wafer 120 that has undergone the chamfering step S8, and the diffusion wafer 120 is spin etched one by one in a single wafer manner. The Thereby, a part of the non-diffusion layer 121 in the diffusion wafer 120 is removed.

(S10) Cleaning Step The diffusion wafer 120 that has undergone the spin etching step S9 is cleaned.
The diffusion wafer 120 that has undergone the cleaning step S10 is sent to the next step and subjected to the next process.

By the way, in the diffusion wafer 120, as described above, it is required that the thickness of the non-diffusion layer 121 is uniform in the in-plane direction (direction orthogonal to the thickness direction of the wafer) D2.
However, in the above-described conventional diffusion wafer manufacturing method, it is difficult to make the thickness of the non-diffusion layer 121 uniform in the in-plane direction D2. FIGS. 13A and 13B are cross-sectional views showing a procedure for obtaining two diffusion wafers from one silicon wafer having different diffusion layer shapes.

  More specifically, as shown in FIG. 13A, in the wafer 110, the diffusion layer 122 is formed thinner in the central portion in the in-plane direction D2 than in the peripheral portion in the in-plane direction D2. This phenomenon is mainly caused by the in-plane tendency of the temperature in the wafer 110 and the in-plane uniformity of the dopant gas in the diffusion step S5.

  Accordingly, as shown in FIG. 13B, in the diffusion wafer 120 obtained by cutting the wafer 110 along the center line in the thickness direction D1 (the one-dot chain line in FIG. 13A), the diffusion layer 122 has a surface The in-plane direction central portion 120d is formed thinner than the inward peripheral portion 120c. On the other hand, the surface 121a of the non-diffusion layer 121 is flat in the in-plane direction D2. That is, the non-diffusion layer 121 is thicker in the in-plane direction center portion 120d than in the in-plane direction peripheral portion 120c, contrary to the diffusion layer 122. Thus, the thickness of the non-diffusion layer 121 is not uniform in the in-plane direction D2.

  In addition, it may be possible to make the thickness of the non-diffusion layer 121 uniform by making the thickness of the diffusion layer 122 uniform (preventing variation in thickness), but the formation of the diffusion layer 122 on the wafer 110 is extremely difficult. In a high temperature diffusion furnace, etc., the diffusion process is performed for a long time. Under industrial conditions, the thickness of the diffusion layer 122 can be made uniform by further improving the accuracy of the diffusion process. It is difficult to plan.

JP 2007-103857 A JP-A-5-28397

  Therefore, the present invention makes the thickness of the non-diffusion layer substantially uniform by performing spin etching on a diffusion wafer having a non-diffusion layer and a diffusion layer with a non-uniform thickness, and removing a part of the non-diffusion layer. It is an object of the present invention to provide a diffusion wafer manufacturing method that can be used, and a diffusion wafer obtained by the manufacturing method.

  Although not related to a diffusion wafer, Japanese Patent Application Laid-Open No. 5-28397 (Patent Document 2) describes as a technique related to a method for manufacturing a semiconductor device (semiconductor device), “on the semiconductor substrate, the center of the wafer is the periphery of the wafer. In the method of manufacturing a semiconductor device, after providing an epitaxial layer having a thickness distribution thicker than a portion, the epitaxial layer is etched to a predetermined thickness at a reaction rate using a predetermined etching solution. When etching the wafer, the diffusibility of the etching solution is enhanced in the peripheral portion of the wafer rather than in the central portion of the wafer by using a spin etching apparatus and contacting the etching solution while rotating the wafer. A method for manufacturing a semiconductor device is disclosed.

  However, the technique described in Patent Document 2 is a technique in which the present invention removes a part of a non-diffusion layer in a diffusion wafer by spin etching, whereas a part of an epitaxial layer provided on a semiconductor substrate is removed. This is a technique of removing by spin etching. That is, the technique disclosed in Patent Document 2 and the present invention differ in the layer (epitaxial layer / non-diffusion layer) to be removed by spin etching.

  In addition, the technique described in Patent Document 2 and the present invention are greatly different in the magnitude (thickness) of the machining allowance removed by spin etching. That is, the technique described in Patent Document 2 is a technique that removes a machining allowance of about 1 μm or less from the epitaxial layer, whereas the present invention is about 10 μm or more from the non-diffusion layer as described later. This is a technique for removing the machining allowance, and the size (thickness) of the machining allowance differs by about two orders of magnitude in the μm order.

  (1) A diffusion wafer manufacturing method according to the present invention includes a non-diffusion layer and a diffusion wafer having a spin etching process in which a part of the non-diffusion layer in a diffusion wafer having a diffusion layer is removed by spin etching in a single wafer mode. The spin etching step uses a spin etching apparatus that is movable in an in-plane direction of the diffusion wafer and includes a liquid supply port that supplies an etching solution to the non-diffusion layer in the diffusion wafer, A liquid supply port is positioned at each of the in-plane direction peripheral portion and the in-plane direction center portion of the diffusion wafer, and an etching solution is supplied from the liquid supply port to the non-diffusion layer. Based on the shape, the distribution in the wafer plane of the etching removal amount of the non-diffusion layer during the spin etching is controlled. It is characterized in.

  According to the invention of (1), spin etching is performed on a diffusion wafer having a non-diffusion layer and a diffusion layer with a non-uniform thickness, and the thickness of the non-diffusion layer is reduced by removing a part of the non-diffusion layer. It can be uniform.

  (2) In the invention of (1), the diffusion layer has a concave shape in which the non-diffusion layer side is recessed from an in-plane direction peripheral portion of the diffusion wafer toward an in-plane direction center portion, It is preferable that a part of the non-diffusion layer removed from the diffusion wafer has a convex shape in which the diffusion layer side swells from the periphery in the in-plane direction toward the center in the in-plane direction.

  According to the invention of (2), when the non-diffusion layer side of the diffusion layer has a concave shape that is recessed from the peripheral portion in the in-plane direction toward the central portion in the in-plane direction (diffusion having such a concave shape). In general, the portion of the non-diffused layer that is removed from the diffusion wafer (that is, the machining allowance) has a convex shape that swells from the periphery in the in-plane direction toward the center in the in-plane direction. Therefore, the shape of the non-diffusion layer in the thickness direction becomes a bowl after spin etching. Therefore, the thickness of the non-diffusion layer tends to be substantially uniform.

  (3) In the invention of (1) or (2), before the spin etching step, a non-diffusion layer thickness measurement step of measuring the thickness of the non-diffusion layer in the diffusion wafer by an FT-IR method, The spin etching process controls the distribution of the non-diffusion layer in the wafer surface during the spin etching based on the thickness distribution of the non-diffusion layer measured in the non-diffusion layer thickness measurement process. It is preferable to do.

  According to the invention of (3), the spin etching step is performed based on the non-diffusion layer thickness distribution measured by the non-diffusion layer thickness measurement step. Control internal distribution. Therefore, the thickness uniformity of the non-diffusion layer after spin etching is further improved.

  (4) Moreover, the diffusion wafer of this invention is manufactured by the manufacturing method of the said diffusion wafer, The difference of the maximum thickness in the said non-diffusion layer and minimum thickness is less than 3 micrometers, It is characterized by the above-mentioned.

  According to the present invention, spin etching is performed on a diffusion wafer having a non-diffusion layer and a diffusion layer with a non-uniform thickness, and a part of the non-diffusion layer is removed to make the thickness of the non-diffusion layer substantially uniform. The manufacturing method of the diffusion wafer which can be manufactured, and the diffusion wafer obtained by this manufacturing method can be provided.

  Hereinafter, an embodiment of a method for producing a diffusion wafer according to the present invention will be described with reference to the drawings. FIG. 1 is a flowchart showing an embodiment of a method for producing a diffusion wafer according to the present invention. 2A and 2B are cross-sectional views showing a procedure for obtaining two diffusion wafers from one silicon wafer. 3A and 3B are cross-sectional views showing a procedure for removing a part of the non-diffusion layer in the diffusion wafer.

As shown in FIG. 1, the manufacturing method of the diffusion wafer of this embodiment is provided with following process S1-S8, S11, S12, and S10. The following steps S1 to S8 and S11 are substantially the same as steps S1 to S8 and S11 in the conventional diffusion wafer manufacturing method described above, but will be described again.
(S1) Slicing Step A silicon ingot is sliced to obtain a silicon wafer (hereinafter also simply referred to as “wafer”).

(S2) Chamfering process A chamfering process (beveling) is performed on the wafer that has undergone the slicing process S1. Specifically, the edge is rounded by applying a rotating grindstone to the edge of the side end surface of the wafer. Thereby, the crack of a wafer, chipping prevention, etc. are achieved.

(S3) Lapping process The wafer that has undergone the chamfering process S2 is set in a lapping apparatus, and lapping is performed on the surface of the wafer. Thereby, damages such as scratches and irregularities on the surface of the wafer caused by slicing are removed and the wafer is flattened.

(S4) Alkaline Etching Step The wafer flattened through the lapping step S3 is immersed in an alkaline etching solution, and the wafer is subjected to alkali etching. Thereby, the processing strain layer on the surface of the wafer is removed.

(S5) Diffusion process The wafer that has undergone the alkali etching process S4 is loaded into a boat and exposed to an atmosphere (such as in a diffusion furnace) where the temperature is high and the dopant exists in that state for a predetermined time. As a result, as shown in FIG. 2A, diffusion layers 22 and 22 are formed on the wafer 10 on both surface sides of the wafer 10 with the non-diffusion layer 21 interposed therebetween.

As shown in FIGS. 2A and 2B, the wafer 10 that has undergone the diffusion step S5 is cut by a wire saw or the like along the center line in the thickness direction D1 (the one-dot chain line in FIG. 2A). The wafer 10 is divided into two in the thickness direction D1. Thereby, two diffusion wafers 20 are obtained from one wafer 10. The diffusion wafer 20 has a non-diffusion layer 21 and a diffusion layer 22, respectively.
The front surface 21 a side of the non-diffusion layer 21 of the diffusion wafer 20 is referred to as the front surface 20 a of the diffusion wafer 20, and the front surface 22 a side of the diffusion layer 22 of the diffusion wafer 20 is referred to as the back surface 20 b of the diffusion wafer 20.

Here, for the reason described above, the non-diffusion layer 21 has its diffusion layer 22 side away from the non-diffusion layer 21 from the peripheral portion 20c in the in-plane direction D2 of the diffusion wafer 20 toward the center portion 20d in the in-plane direction D2. It has a convex shape that swells toward the diffusion layer 22.
On the other hand, the diffusion layer 22 has a concave shape in which the non-diffusion layer 21 side is recessed from the non-diffusion layer 21 toward the diffusion layer 22 as it goes from the in-plane direction peripheral portion 20c of the diffusion wafer 20 toward the in-plane direction center portion 20d. have.

(S7) Surface grinding process The diffusion wafer 20 formed in the two-part process S6 is set in a surface grinding machine, and a grinding stone is pressed against the surface 20a of the diffusion wafer 20 (the surface 21a of the non-diffusion layer 21) to diffuse. Surface grinding is performed on the wafer 20. Thereby, the surface 21a of the non-diffusion layer 21 of the diffusion wafer 20 is planarized.

(S8) Chamfering process Chamfering (beveling) is performed on the diffusion wafer 20 that has undergone the surface grinding process S7. Specifically, by applying a rotating grindstone to the edge of the side end surface of the diffusion wafer 20, the edge is rounded. As a result, the diffusion wafer 20 can be prevented from cracking, dust generation and the like.

(S11) Non-diffusion layer thickness measurement process The thickness of the non-diffusion layer 21 is set to FT using the FT-IR (Fourier transform infrared spectroscopy) apparatus 31 (refer FIG. 4) with respect to the diffusion wafer 20 which passed chamfering process S8. -Measured by IR method. The spin etching system 30 including the FT-IR apparatus 31 will be described later.

(S12) Spin Etching Step An etching solution is supplied to the surface 20a of the diffusion wafer 20 that has undergone the non-diffusion layer thickness measurement step S11, and the diffusion wafer 20 is spin etched one by one in a single wafer manner. Thereby, as shown in FIGS. 3A and 3B, a part 23 of the non-diffusion layer 21 in the diffusion wafer 20 is removed. Hereinafter, the part 23 to be removed in the non-diffusion layer 21 is also referred to as “removal”.
Details of the spin etching apparatus 40 that performs the spin etching step S12 will be described later. The spin etching system 30 includes an FT-IR device 31, a spin etching device 40, and the like.

(S10) Cleaning Step The diffusion wafer 20 that has undergone the spin etching step S12 is cleaned.
The diffusion wafer 20 that has undergone the cleaning step S10 is sent to the next step and subjected to the next process.

  Next, the spin etching system 30 will be described. FIG. 4 is a functional block diagram showing a spin etching system including an FT-IR apparatus and a spin etching apparatus. FIG. 5 is a perspective view schematically showing a spin etching apparatus. 6A and 6B are plan views schematically showing a spin etching apparatus, in which FIG. 6A is a diagram showing a state in which a liquid supply port is located in the in-plane peripheral portion of the diffusion wafer, and FIG. 6B is a liquid supply port. It is a figure which shows the state which is located in the in-plane direction center part of a diffusion wafer.

As shown in FIG. 4, the spin etching system 30 includes an FT-IR device 31 and a spin etching device 40.
The FT-IR apparatus 31 can measure the thickness of the non-diffusion layer 21 in the diffusion wafer 20 by the FT-IR method.

  As shown in FIGS. 4 and 5, the spin etching apparatus 40 includes a spin chuck 41, a liquid supply port 42, a rotating arm 43, a CPU 44, and a memory 45.

  As shown in FIG. 5, the spin chuck 41 can rotate the diffusion wafer 20 while holding the non-diffusion layer 21 upward (on the side opposite to the upper surface of the spin chuck 41). The rotational speed (angular speed) of the spin chuck 41, that is, the rotational speed of the diffusion wafer 20 held by the spin chuck 41 can be appropriately set according to the etching conditions and the like. The rotational speed of the spin chuck 41 is not particularly limited, but can be set to 500 to 900 rpm, for example.

The liquid supply port 42 can supply the etching liquid to the non-diffusion layer 21 in the diffusion wafer 20 rotated by the spin chuck 41. The supply form of the etching liquid to the diffusion wafer 20 is not particularly limited, and may be, for example, dripping, jetting, or flowing down. The supply rate (mL / s) of the etching solution to the diffusion wafer 20 can be appropriately set according to the etching conditions and the like.
Since the liquid supply port 42 is connected to the rotating arm 43 as described later, the liquid supply port 42 can move in the in-plane direction D2 of the diffusion wafer 20.

As the etching solution, one having an appropriate viscosity (surface diffusibility) can be used depending on etching conditions and the like, and is not particularly limited. For example, phosphoric acid-based mixed acid (H ₃ PO ₄ · HNO ₃ · H ₂ SO ₄ .HF) is used.

  The diffusion rate of the etching solution (the rate of spreading on the surface 20a of the diffusion wafer 20) is appropriately set according to the etching conditions and the like. The diffusion rate of the etching solution can be changed by changing the viscosity of the etching solution, the supply rate of the etching solution, the rotation speed of the spin chuck 41, and the like.

  One end of the rotating arm 43 is connected to the liquid supply port 42, and the other end is pivotable about the other end. According to the rotating arm 43, for example, as shown in FIG. 6A, the liquid supply port 42 is positioned in the in-plane peripheral portion 20c of the diffusion wafer 20 rotated by the spin chuck 41 for a predetermined time. Alternatively, as shown in FIG. 6B, it can be positioned at the center 20d in the in-plane direction of the diffusion wafer 20 rotated by the spin chuck 41 for a predetermined time.

  Alternatively, according to the rotating arm 43, the liquid supply port 42 is, for example, from the in-plane direction peripheral portion 20c of the diffusion wafer 20 shown in FIG. 6A to the in-plane of the diffusion wafer 20 shown in FIG. It can be moved to the direction center portion 20d, or from the in-plane direction center portion 20d of the diffusion wafer 20 shown in FIG. 6B to the in-plane direction peripheral portion 20c of the diffusion wafer 20 shown in FIG. Can be moved.

The liquid supply port 42 can move in the almost radial direction of the spin chuck 41 (that is, in the almost radial direction of the diffusion wafer 20) as the rotation arm 43 rotates. The moving speed of the liquid supply port 42 along the radial direction of the spin chuck 41 is appropriately set according to the etching conditions and the like, and can be set to 0 to 20 (mm / s), for example.
The liquid supply port 42 can supply the etching liquid E to the diffusion wafer 20 while moving along the radial direction of the spin chuck 41.

  The CPU 44 controls the entire spin etching system 30. In particular, the CPU 44 performs predetermined control on the spin chuck 41, the liquid supply port 42, the rotating arm 43, and the like. For example, the CPU 44 determines the rotational speed of the spin chuck 41 and the etching liquid from the liquid supply port 42 based on the shape of the diffusion layer 22 in the thickness direction D1 and the thickness of the non-diffusion layer 21 measured by the FT-IR device 31. The E supply speed, the rotation speed of the rotation arm 43 (the movement speed of the liquid supply port 42 along the radial direction of the spin chuck 41), the stop position of the rotation arm 43, and the like are controlled.

  The memory 45 stores predetermined data. For example, the memory 45 stores various function programs for executing the spin etching system 30, an etching condition DB 45a, and the like. The etching condition DB 45 a is a database that stores etching conditions according to the thickness of the non-diffusing layer 21 measured by the FT-IR apparatus 31.

Next, the spin etching step S12 will be described in detail.
In the spin etching step S12, the spin etching system 30 is used to control the in-wafer distribution of the etching removal amount of the non-diffusion layer 21 during the spin etching. Specifically, in the spin etching step S12, the liquid supply port 42 is positioned at each of the in-plane direction peripheral portion 20c and the in-plane direction center portion 20d of the diffusion wafer 20, and the etching solution E is supplied from the liquid supply port 42 to the non-diffusion layer. 21.

  Specifically, the liquid supply port 42 is positioned in the in-plane peripheral portion 20 c of the diffusion wafer 20 for a predetermined time, and the etching solution E is supplied from the liquid supply port 42 to the surface 21 a of the non-diffusion layer 21. Thereafter, the liquid supply port 42 is moved from the in-plane direction peripheral portion 20 c to the in-plane direction center portion 20 d of the diffusion wafer 20. In the meantime, the etching solution E is supplied from the solution supply port 42 to the surface 21a of the non-diffusion layer 21. The liquid supply port 42 is positioned at the center 20d in the in-plane direction of the diffusion wafer 20 for a predetermined time, and the etching solution E is supplied from the liquid supply port 42 to the surface 21a of the non-diffusion layer 21. In this way, based on the shape of the non-diffusion layer 21 in the thickness direction, the in-wafer distribution of the etching removal amount of the non-diffusion layer 21 during spin etching is controlled.

  In addition, the spin etching step S12 is based on the thickness distribution of the non-diffusion layer 21 measured in the non-diffusion layer thickness measurement step S11 together with the shape in the thickness direction of the non-diffusion layer 21. The in-wafer surface distribution of the etching removal amount 21 is controlled. For example, when the thickness of the non-diffusion layer 21 in the diffusion wafer 20 is measured by the FT-IR apparatus 31, the CPU 44 accesses the etching condition DB 45 a in the memory 45 and performs etching suitable for the measured thickness of the non-diffusion layer 21. Get conditions. Based on the etching conditions, predetermined control is performed on the spin chuck 41, the liquid supply port 42, the rotating arm 43, and the like to perform predetermined spin etching on the diffusion wafer 20.

  As a result, as shown in FIG. 3, the allowance 23 of the non-diffusion layer 21 is removed in the diffusion wafer 20 that was flat before the spin etching step S12. Then, the machining allowance 23 has a convex shape that swells from the non-diffusion layer 21 toward the diffusion layer 22 as it goes from the in-plane direction peripheral portion 20c of the diffusion wafer 20 toward the in-plane direction center portion 20d.

  Thus, according to the manufacturing method of this embodiment, the diffusion wafer 20 having the non-diffusion layer 21 and the diffusion layer 22 having a non-uniform thickness is obtained through the steps S1 to S8, S11, S12, and S10. On the other hand, the thickness of the non-diffusion layer 21 can be made substantially uniform by performing spin etching and removing a part (the machining allowance 23) of the non-diffusion layer 21. As a result, for example, a diffusion wafer 20 in which the difference between the maximum thickness and the minimum thickness in the non-diffusion layer 21 is within 3 μm is obtained.

Below, the various graphs regarding the effect of this invention are demonstrated.
FIGS. 7A to 7C are graphs showing measured values of the machining allowance when the target value of the machining allowance of the non-diffusion layer is varied. FIG. 7A is a graph showing a case where a machining allowance for a flat shape is aimed at, and FIG. 7B is a graph showing a case where a machining allowance having a difference in unevenness (unevenness gap) of 1.5 μm is aimed at. 7 (c) is a graph showing a case where a machining allowance with a difference of unevenness of 2 μm is aimed.

  In FIGS. 7A to 7C, the one-dot chain vertical line passing through the central portion in the horizontal axis direction indicates the in-plane direction central portion in the machining allowance. The horizontal axis indicates the distance in the in-plane direction from the center in the in-plane direction at the machining allowance. The right end and the left end of the graph each indicate a position of 75 mm in the in-plane direction from the center in the in-plane direction. The vertical axis represents the thickness of the machining allowance (μm). A solid line indicates an actual measurement value of the machining allowance when the machining allowance thickness is 11 μm at the center in the in-plane direction of the wafer. A broken line indicates an actual measurement value of the machining allowance when the machining allowance thickness is 10 μm at the center in the in-plane direction of the wafer. A two-dot chain line indicates an actual measurement value of the machining allowance when the machining allowance thickness is 9 μm at the center in the in-plane direction of the wafer.

  According to FIGS. 7A to 7C, it can be seen that the difference in the unevenness of the machining allowance can be varied (0 μm = flat shape, 1.5 μm, 2 μm) by varying the spin etching conditions. In addition, by making the target value of the thickness of the non-diffusion layer different (11 μm, 10 μm, 9 μm), the measured value of the thickness of the non-diffusion layer can be made different, and the overall shape of the machining allowance is irrespective of the target value of the thickness It can be seen that the shapes are almost the same.

FIG. 8 is a graph showing the thickness distribution of the non-diffusion layer. The vertical axis represents the thickness distribution (maximum value-minimum value) (μm) in the in-plane direction of the non-diffusion layer. The left side shows the distribution in the conventional example, and the right side shows the distribution in the present invention. Squares mean a 20% distribution relative to the median. The horizontal bars above and below the square mean a 75% distribution with respect to the median value.
According to FIG. 8, it can be seen that the thickness distribution of the non-diffusion layer is smaller in the present invention than the conventional example, that is, the uniformity of the thickness of the non-diffusion layer is improved.

  FIGS. 9A and 9B are graphs showing the relationship between the supply position of the etching solution and the thickness of the removal allowance of the non-diffusion layer. FIG. 9A shows the case where the etching solution is supplied at the peripheral portion in the in-plane direction of the wafer and the central portion in the in-plane direction, and the case where the etching solution is supplied only at the central portion in the in-plane direction of the wafer. It is a graph. FIG. 9B is a graph showing the time when the supply time of the etching solution is varied in the case where the etching solution is supplied only in the peripheral portion in the in-plane direction of the wafer.

9A and 9B, the horizontal axis indicates the distance in the in-plane direction D2 from the center in the in-plane direction at the machining allowance. The vertical axis represents the thickness of the machining allowance (μm).
In FIG. 9A, the solid line shows the case where the etching solution is supplied at the peripheral portion in the in-plane direction and the central portion in the in-plane direction of the wafer. A broken line shows the case where the etching solution is supplied only in the center in the in-plane direction of the wafer.
In FIG. 9B, a solid line, a broken line, and a two-dot chain line indicate a case where the etching solution is supplied only in the peripheral portion in the in-plane direction of the wafer. A solid line indicates a case where the etching solution supply time is 1 second, a broken line indicates a case where the etching solution supply time is 2 seconds, and a two-dot chain line indicates that the etching solution supply time is 3 seconds. Indicates a case.

  As shown in FIG. 9A, when the etching solution is supplied only at the center portion in the in-plane direction of the wafer (broken line), the thickness of the machining allowance becomes too large at the center portion in the in-plane direction of the wafer. That is, the difference in the thickness of the machining allowance becomes too large between the central portion in the in-plane direction of the wafer and the peripheral portion in the in-plane direction. Therefore, depending on the shape of the diffusion layer on the non-diffusion layer side, the allowance cannot be made a desired shape.

  Further, as shown in FIG. 9B, when the etching solution is supplied only in the peripheral portion in the in-plane direction of the wafer (solid line, broken line, two-dot chain line), the machining allowance is in the central portion in the in-plane direction of the wafer. There is almost no thickness (the non-diffusion layer 21 is hardly removed), and the thickness of the machining allowance increases as it goes to the periphery in the in-plane direction of the wafer. Further, the longer the supply time of the etching solution, the greater the thickness of the machining allowance.

On the other hand, according to FIG. 9A, in the case where the etching solution is supplied at the peripheral portion in the in-plane direction of the wafer and the central portion in the in-plane direction (solid line), The difference in thickness of the machining allowance between the inner peripheral portion and the inner circumference is moderate, and the machining allowance is formed (a part of the non-diffusion layer is removed).
As described above, according to FIG. 9, by supplying the etching liquid both in the in-plane direction peripheral part and in-plane direction center part of the wafer, a machining allowance having a desired shape in the thickness direction can be formed. Recognize.

FIGS. 10A to 10D show different supply times of the etching liquid at the peripheral portion in the in-plane direction of the wafer when the etching liquid is supplied at the peripheral portion in the in-plane direction and the central portion in the in-plane direction of the wafer. It is a graph which shows the thickness of the machining allowance at the time.
10A to 10D, the horizontal axis indicates the distance in the in-plane direction from the center in the in-plane direction at the machining allowance. The vertical axis represents the thickness of the machining allowance (μm).

  In the example shown in FIGS. 10A to 10D, after supplying the etching solution for a predetermined time at the peripheral portion in the in-plane direction of the wafer, the etching solution is supplied for 70 seconds at the central portion in the in-plane direction. ing. In the example shown in FIGS. 10A, 10B, 10C, and 10D, the supply time of the etchant in the peripheral portion in the in-plane direction of the wafer is 0 second, 1 second, 2 seconds, and 3 seconds, respectively. It is.

  According to FIG. 10, the longer the supply time of the etching solution in the in-plane peripheral portion of the wafer, the thicker the machining allowance in the in-plane peripheral portion of the wafer. It turns out that becomes small. On the other hand, the shorter the supply time of the etching solution in the peripheral area in the in-plane direction of the wafer, the smaller the thickness of the allowance in the peripheral area in the in-plane direction of the wafer. I understand.

Although one embodiment and one embodiment of the present invention have been described above, the present invention is not limited to the above-described one embodiment and embodiment.
For example, in the above-described embodiment, the etching solution is supplied in the in-plane direction central portion 20d of the diffusion wafer 20 after the etching solution is supplied in the in-plane direction peripheral portion 20c of the diffusion wafer 20. However, the etching solution may be supplied to the peripheral portion 20c in the in-plane direction of the diffusion wafer 20 after the etching solution is supplied to the center portion 20d of the diffusion wafer 20 in the in-plane direction.

  In the above-described embodiment, the etching solution E is supplied from the liquid supply port 42 even while the liquid supply port 42 moves between the in-plane direction peripheral portion 20 c and the in-plane direction center portion 20 d of the diffusion wafer 20. However, the present invention is not limited to this, and while the liquid supply port 42 moves between the in-plane direction peripheral portion 20c and the in-plane direction center portion 20d of the wafer 10, the etching solution E is supplied from the liquid supply port 42. May be stopped.

  In the above-described embodiment, a part 23 (tolerance) on the non-diffusion layer 21 side of the diffusion layer 22 extends from the non-diffusion layer 21 toward the in-plane direction center 20d from the in-plane direction peripheral portion 20c of the diffusion wafer 20. Although it has a convex shape that swells toward the diffusion layer 22, it may have other shapes.

  In the embodiment, the diffusion wafer 20 is obtained by dividing the wafer 10 having the diffusion layers 22 on the both sides in the thickness direction into two in the thickness direction D1, but the invention is not limited to this. For example, in the thickness direction D1 For the wafer 10 having the diffusion layers 22 and 22 on both sides, the diffusion wafer 20 may be obtained by removing substantially half of the thickness direction D1 in the diffusion layer 22 on one side and the non-diffusion layer 21 by grinding. it can.

It is a flowchart which shows one embodiment of the manufacturing method of the diffusion wafer of this invention. (A) And (b) is sectional drawing which shows the procedure of obtaining two diffusion wafers from one silicon wafer. (A) And (b) is sectional drawing which shows the procedure which removes a part of non-diffusion layer in a diffusion wafer. It is a functional block diagram showing a spin etching system including an FT-IR apparatus and a spin etching apparatus. It is a perspective view which shows a spin etching apparatus typically. It is a top view which shows a spin etching apparatus typically, (a) is a figure which shows the state in which a liquid supply port is located in the in-plane direction peripheral part of a diffusion wafer, (b) is a figure where a liquid supply port is a diffusion wafer. It is a figure which shows the state located in an in-plane direction center part. (A)-(c) is a graph which shows the measured value of the magnitude | size of machining allowance in case the target value of the machining allowance magnitude of a non-diffusion layer is varied. It is a graph which shows distribution of the thickness of a non-diffusion layer. (A) And (b) is a graph which shows the relationship between the supply position etc. of an etching liquid, and the thickness of the machining allowance of a non-diffusion layer. (A) to (d) in the case where the etching solution is supplied at the peripheral portion in the in-plane direction of the wafer when the etching solution is supplied at the peripheral portion in the in-plane direction and the central portion in the in-plane direction of the wafer. It is a graph which shows the thickness of the machining allowance. It is a flowchart which shows an example of the manufacturing method of the conventional diffusion wafer. (A) And (b) is sectional drawing which shows the procedure of obtaining two diffusion wafers from one silicon wafer. (A) And (b) is sectional drawing which shows the procedure which obtains two diffusion wafers from one silicon wafer from which the shape of a diffusion layer differs.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Wafer 20 Diffusion wafer 20c In-plane direction peripheral part 20d In-plane direction center part 21 Non-diffusion layer 22 Diffusion layer 40 Spin etching apparatus 42 Liquid supply port D1 Thickness direction D2 In-plane direction E Etching liquid S11 Non-diffusion layer thickness measurement process S12 Spin etching process

Claims (4)

A diffusion wafer manufacturing method comprising a non-diffusion layer and a spin etching step of removing a part of the non-diffusion layer in a diffusion wafer having a diffusion layer by spin etching in a single wafer type,
The spin etching step uses a spin etching apparatus that is movable in an in-plane direction of the diffusion wafer and includes a liquid supply port that supplies an etching solution to the non-diffusion layer in the diffusion wafer, and the liquid supply port is Based on the shape of the non-diffusion layer in the thickness direction by supplying the non-diffusion layer with the etching liquid from the liquid supply port located in the in-plane direction peripheral part and in-plane direction center part of the diffusion wafer, A method for producing a diffusion wafer, comprising: controlling in-wafer distribution of an etching removal amount of the non-diffusion layer during the spin etching. The diffusion layer has a concave shape in which the non-diffusion layer side is recessed from an in-plane direction peripheral portion of the diffusion wafer toward an in-plane direction center portion,
A part of the non-diffusion layer removed from the diffusion wafer has a convex shape in which the diffusion layer side swells from an in-plane direction peripheral portion to an in-plane direction center portion of the diffusion wafer. The manufacturing method of the diffusion wafer of Claim 1 to do. Before the spin etching step, comprising a non-diffusion layer thickness measurement step of measuring the thickness of the non-diffusion layer in the diffusion wafer by FT-IR method,
The spin etching process controls the distribution of the non-diffusion layer in the wafer surface during the spin etching based on the thickness distribution of the non-diffusion layer measured in the non-diffusion layer thickness measurement process. The method for producing a diffusion wafer according to claim 1 or 2, wherein:   A diffusion wafer manufactured by the method for manufacturing a diffusion wafer according to claim 1, wherein a difference between a maximum thickness and a minimum thickness in the non-diffusion layer is within 3 μm.

Images