JPH01114044A - Manufacture of semiconductor substrate - Google Patents

Abstract

PURPOSE:To reduce material loss and to decrease a cost by dividing by slicing a substrate formed with diffused regions on both side faces in its thicknesswise direction, and providing two pairs. CONSTITUTION:A starting wafer 10 having 1200mum of thickness is prepared, and diffused region 11 are formed on both side faces. Then, the wafers 10 which has finished diffusing steps are superposed with the orientation flat parts as reference of aligning them, adhered with paraffin, thereby forming an ingot 12. Reinforcing dummy wafers 13 are adhered to both side faces of the ingot 12. They are adjusted at positions by position detecting sensors 16 adhered onto a slice carbon base 14, and sequentially sliced at a predetermined feeding pitch P by a diamond blade. Then, the paraffin is removed, the wafers are exfoliated, the sliced faces are polished, and mirror-polished. Thus, material loss can be reduced, and its cost is decreased.

Description

[Detailed Description of the Invention] [Object of the Invention] (Industrial Application Field) The present invention relates to a method for manufacturing a semiconductor substrate used in power semiconductor devices, etc. The present invention relates to a method of manufacturing the formed semiconductor substrate.

(Prior Art) A semiconductor substrate (hereinafter referred to as a wafer) used in power semiconductor devices, etc. has a low-resistance diffusion region formed on one side, and the elements are diffused on the other side. After the formation, a metal layer is formed on the diffusion region on one side so that the electrode can be taken out. One S L is used for such purposes.
IDE Lapped wafers) are conventionally manufactured from starting wafers that are lapped to a thickness of around 600 μm. Since this OSL wafer is manufactured from a crystal rod (ingot) through processes such as slicing and lapping, in order to manufacture a wafer with a thickness of 600 μm, a material with a thickness of 1200 μm, which is approximately twice that thickness, is required.

FIGS. 6(a) to d) (≠ Cross-sectional views sequentially showing each process according to a conventional method. First, a wafer 30 with a thickness of 600 μm is prepared (FIG. 6(a)). This wafer 30 is an ingot. The wafer is sliced from the state of is 550 μm.Next, diffusion regions 31 were formed on both sides of the wafer 30 (FIG. 6(b)), and then one side was polished with a grinder, and the other diffusion region was removed by polishing by lapping, etc. After (Figure 6 (
c)) The removed surface is mirror-polished to form a mirror-finished wafer (FIG. 6(d)). Note that the total material loss due to grinding, lapping, and mirror polishing is about 300 μm.

In the conventional method, the starting wafer has a thickness of 600 μm, whereas the manufactured OSL wafer has a thickness of 300 μm, and half of the material is discarded. This cannot be avoided with conventional methods. Regarding the cost allocation for O8L wafers, commercially available ones have a profit margin of 10% and indirect costs of 15%, with the remaining 75% being direct costs.

Furthermore, the raw wafer costs accounted for 56.4 of the total direct costs.
%become. Therefore, the cost influence of chemical wafers in O8L wafers is very large, and is a major barrier to reducing the cost of O8L wafers.

However, to obtain a starting wafer of 600 μm, slicing,
It cannot be completed unless two steps of lapping are performed, and these steps result in a material loss of approximately 550 μm. This has a diameter of 125 mm and results in a material loss of approximately 16 g or more.

(Problems to be Solved by the Invention) As described above, in the conventional method, the final wafer is finished by removing large portions from the processed wafer, which has the drawback of wasting material and increasing costs.

The present invention has been made in consideration of the above circumstances, and its purpose is to provide a method of manufacturing a semiconductor substrate that can manufacture a semiconductor substrate at low cost.

[Configuration of the Invention (Means for Solving Problems)] The method for manufacturing a semiconductor substrate of the present invention includes a step of diffusing impurities on both sides of a semiconductor substrate to form a diffusion region, and a step of diffusing the substrate in two directions in the thickness direction. The process consists of a step of dividing the substrate into pieces, and a step of polishing the surface of each divided substrate opposite to the diffusion region to a mirror finish.

(Function) In the method of the present invention, a substrate with diffusion regions formed on both sides is sliced and divided in the thickness direction to obtain two pieces, thereby reducing feed loss per substrate compared to the conventional method.

(Example) Hereinafter, an example method of the present invention will be described with reference to the drawings. FIGS. 1(a) to 1(e) are sectional views sequentially showing each step according to the method of the present invention.

First, a starting wafer 10 having a thickness of 1200 μm, which is twice the thickness of the conventional one, is prepared (FIG. 1(a)). As in the past, this wafer 10 is sliced from an ingot and then lapped to have a desired thickness. This wafer 10 has a diameter of, for example, 125 mm.
±0.5mm, resistivity is 40 to 50ΩCm,
N-type P (phosphorus) is introduced as an impurity, the surface is JIS #1000 lapping finish, and the parallelism is 5.
It is less than μm. Also, the slice fee is 35 yen as before.
The thickness is about 0 μm, and the wrap margin is about 200 μm. Therefore, the material loss of the starting wafer is 550 μm, as in the conventional case.

Next, diffusion regions 11 are formed on both sides of the starting wafer 10 (
Figure 1(b)). This diffusion is performed, for example, in a 50-piece formation using 50 starting wafers as 10 groups. First, the starting wafers are pretreated with an alkaline cleaning solution to improve surface cleanliness, and then placed on a quartz boat in units of 10 tons. Subsequently, after being inserted into a diffusion furnace in an oxygen atmosphere at a temperature of 1200° C., the steam obtained by evaporating POc, f3 with N2 gas is mixed with oxygen gas and supplied into the furnace. After maintaining the state for a predetermined period of time, the supply of POcOsO4 gas is stopped and the boat is taken out of the furnace. Next, by removing the phosphorus glass formed on the surface with an HF solution, so-called deposition, in which regions containing impurities at a high concentration are deposited on both sides of the starting wafer 10, is completed.

After deposition, the surface impurity concentration of some wafers is evaluated and confirmed to be 10"/cm3 or higher.

After deposition, the wafers are stacked closely together in units of 10 t on a quartz boat.

In this case, efficiency is increased by stacking 50 pieces at the same time. Stuck boat at temperature 1250
It is inserted into a furnace in an oxygen atmosphere at ℃ and subjected to slumping for a predetermined period of time. After slapping, take the boat out of the furnace,
The phosphorus glass formed on the surface is removed again with an HF solution. After this, the depth of the diffusion region 11 of some wafers is measured and confirmed that it is within 150 μm ± 5 μm,
Record the representative value for each lot.

Next, the wafers 1O that have undergone the diffusion process are bonded together as shown in the side view of FIG. 2 to form an ingot 12. At this time, paraffin having a melting point of 70° C. is used as the adhesive. At this time, each wafer is heated in advance on a heater,
By repeating the process of applying paraffin to the surface and stacking another wafer on top of it, all the wafers are adhered to form the ingot 12. At this time, the orienteering flat portions provided on each wafer are used as alignment standards to overlap each other. Further, dummy wafers 13 for reinforcement are bonded to both end faces of the ingot 12. This dummy wafer 13 also has a diameter of 125 mm±0.5 mm, a thickness of, for example, 600 μm, and a surface of JIS #1.
Finished with 000 laps. After these sea urchins 11 are adhered, a pressure of 100 g/cm2 or more is applied to remove excess paraffin and cooled. Wipe off excess paraffin on the surface with triclean. The total length of the bonded ingot was measured, and the thickness of the paraffin used as the adhesive was calculated per 81 sea urchins and recorded.

Next, the ingot 12 is sliced as described above to divide the wafer 10 into two wafers 18 as shown in FIG. 1(C). This slicing is performed as follows. First, an ingot is bonded onto a carbon base for slicing using adhesive resin (for example, epoxy resin). The ingot glued onto the base is Maycr & B
It is attached to a TS-27 type diamond blade type slicing device manufactured by Urge. The diamond blade used at this time has a two-inch outer diameter, a blade thickness of 340 μm, and a base metal thickness of 150 μm.

Also, the cutting distance when cutting with this blade is 400μ
m.

Next, this slicing process will be explained in detail using the side views shown in FIGS. 3(a) to 3(d). First, a dummy wafer 15 for position correction is bonded in front of the ingot 12 bonded onto the carbon base 14 for slicing. This dummy wafer 15 has a thickness of, for example, 3000 μm. Next, a sensor 16 for position detection is placed to face the dummy wafer 15 for position correction installed on the entire surface of the ingot 12 bonded on the carbon base 14 for slicing (see Fig. 3).
a)). Note that 17 is a diamond blade of the slicing device. Next, carbon base 14 for slicing
The dummy wafer 15 located in front of the ingot 12 bonded thereon is sliced by the diamond blade 16 (FIG. 3(b)). Next, remove the sliced dummy wafer 15 and remove the remaining dummy wafer 15.
Adjust the position of sensor IB so that it is parallel to the plane of (
Figure 3 (C)). Next, after removing the remaining dummy wafer 15, the position of the slicing carbon base 14 is adjusted so that the plane of the reinforcing dummy wafer 13 on the end face of the ingot 12 is parallel to the sensor 16. This results in
The alignment between the plane of the wafer and the diamond blade 1B is completed.

After this, before turning on the automatic feed mechanism in the slicing machine, manually cut the (cutting allowance/2) -400 μm.
/2-200 μm in advance, the automatic feeding mechanism is turned on and slices are sequentially performed at a feeding pitch P. P-T+
ω, T is the average thickness of each wafer lO, and ω is the thickness of paraffin per wafer. Here, if T is 1200 μm and ω is 2 μm, then the feed pitch P in this case is 1202 μm. At this time, the slicing speed was 50 mm/min, and the number of rotations of the blade was 200 Or
pm, and city water was used as the cooling water. The side view of FIG. 4 shows the state after such a slicing process is completed. Each of the wafers 10 is divided into two wafers 18.

Next, after each of the two divided wafers 18 is removed from the carbon base for slicing, the adhesive resin and paraffin are removed using a chemical solution, the wafer is peeled off, and the surface is cleaned. Thereafter, the sliced surface of each wafer 18 is polished using an automatic grinder (FIG. 1(d)). The polishing stock at this time is determined according to the finished thickness of the so-called intrinsic layer in which impurities left in the final step are not introduced, and for example, the polishing stock is set to 180 μm.

For example, the polishing device used at this time is 5VG manufactured by Renzan Kikai.
-502 can be used, and the grindstone is # for rough grinding.
#400 and #1200 were used for finish grinding, and the rough grinding speed was 150 μm/min and the finish grinding speed was 10 μm/min.

Thereafter, the polished surface of each wafer 18 is mirror polished to have a mirror surface. This step is performed after each wafer 18 is bonded onto the polishing plate using wax. As the polishing agent, a colloidal polisher made of 5i02 is used, and the mirror polishing distance is 20 μm. After polishing, the wafer is peeled off from the plate and cleaned with chemicals to remove wax, etc., and a diffusion region 11 is formed on one side.
The O3L wafer 19 is completed (FIG. 1(e)).

FIG. 5 is a diagram showing a comparison of material loss occurring in each process between the conventional method and the method of the present invention. As mentioned above, raw wafers account for a high proportion of the cost of O3L wafers manufactured by conventional methods, and cost reductions in raw wafers are directly linked to cost reductions in OSL wafers. In the raw wafer manufacturing process, a total of 550 μm of material loss occurs, 200 μm per lap of 350 μm during slicing up to the starting wafer, but this does not depend on the thickness of the wafer. Therefore, this loss is the same value for both the conventional method and the method of this invention.

Furthermore, in the method of the present invention, a slice loss of 400 μm occurs because one wafer is sliced into two after diffusion, but this loss does not occur in the conventional method. Further, while the conventional method requires a loss of 300 μm during grinding and mirror polishing after diffusion, the method of the present invention can reduce this loss to 200 μm by eliminating the need to polish the diffusion region. In the conventional method, a total of 850 μm of material loss occurs for each wafer as described above. However, in the method of this invention, the loss shown in FIG. 5 occurs every two wafers, so the material loss per wafer can be reduced to 575 μm, which is 275 μm less than the 850 μm of the conventional method. can.

As a result, waste of materials can be reduced and O3L wafers can be manufactured at low cost.

[Effects of the Invention] As explained above, according to the present invention, it is possible to provide a method for manufacturing a semiconductor substrate that can manufacture a semiconductor substrate at low cost.

[Brief Description of the Drawings] Fig. 1 is a sectional view sequentially showing each step according to the method of the present invention, Fig. 2 is a side view of an intermediate step in the above method, and Fig.
The figure is a cross-sectional view showing detailed steps in the middle of the above method, FIG. 4 is a side view of the final step of the above method, FIG. 5 is a diagram showing material loss occurring in each step, and FIG. 6 is a conventional method. FIG. lO... Starting wafer, 11... Diffusion region, 12...
・Ingot, 13... Dummy wafer for reinforcement, 1
4... Carbon base for slicing, 15... Dummy wafer for position correction, 1B... Sensor for position detection, 17... Diamond blade, 18. 19...
・Wafer. Applicant's agent Patent attorney Takehiko Suzue (a) 1st V! J Fig. 1 Fig. 20 Fig. 4 Fig. 5 (a) (C) (d) Fig. 6 Fig. 1, Indication of case Patent applicant (307) Toshiba Corporation 4, agent USE Pill 7, 3-7-2 Kasumigaseki, Chiyoda-ku, Tokyo;
Contents of amendment (1) The scope of claims is corrected as shown in the attached sheet. (a) Delete the sentence "grind" on page 5, line 18. (3) Correct "grinder" on page 12, line 4 to read "grinder." 2. Scope of Claims (1) A step of diffusing impurities on both sides of a semiconductor substrate to form a diffusion region, a step of dividing the substrate into two in the thickness direction, and a step of forming a diffusion region on each of the divided substrates on the 11th side opposite to the diffusion region. (The step of dividing the 211m board into two pieces in the thickness direction,
2. The method of manufacturing a semiconductor substrate according to claim 1, wherein the method is performed using an ingot made by bonding a plurality of substrates each having an enlarged area formed thereon. (3) The method for manufacturing a semiconductor substrate according to claim 1, wherein the step of dividing the substrate into two pieces in the thickness direction is performed by a slicing method using a diamond blade. (4) The method for manufacturing a semiconductor substrate according to claim 3, wherein slicing is performed with reinforcing dummy substrates bonded to both end faces of the ingot. (5) Eleven correction substrates are placed on one end surface of the ingot to correct the position between the plane of the ingot and the plane of the diamond blade! 3. The method of manufacturing a semiconductor substrate according to claim 3, wherein 1T slicing is performed. Applicant's agent Patent attorney Takehiko Suzue

Claims (5)

[Claims] (1) A step of diffusing impurities on both sides of a semiconductor substrate to form a diffusion region, a step of dividing the substrate into two in the thickness direction, and polishing the opposite side of each divided substrate from the diffusion region. 1. A method for manufacturing a semiconductor substrate, comprising a step of mirror-finishing the substrate. (2) The method for manufacturing a semiconductor substrate according to claim 1, wherein the step of dividing the substrate into two in the thickness direction is carried out in the form of an ingot made by bonding a plurality of substrates each having one diffusion region formed thereon. . (3) The method for manufacturing a semiconductor substrate according to claim 1, wherein the step of dividing the substrate into two pieces in the thickness direction is performed by a slicing method using a diamond blade. (4) The method for manufacturing a semiconductor substrate according to claim 3, wherein slicing is performed with reinforcing dummy substrates bonded to both end faces of the ingot. (5) The slicing is performed with a correction substrate installed on one end surface of the ingot for correcting the position between the plane of the ingot and the plane of the diamond blade. A method for manufacturing a semiconductor substrate.