JPS5850424B2 - How to divide a semiconductor substrate - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a semiconductor substrate dividing method for dividing a semiconductor substrate on which a plurality of semiconductor elements are fabricated to obtain individual semiconductor element substrates.

A silicon semiconductor element substrate is usually obtained by dividing a silicon substrate on which a large number of semiconductor elements are fabricated, but if the area of the silicon semiconductor element substrate to be fabricated is fixed, the number of silicon semiconductor elements to be fabricated is In order to increase the silicon substrate utilization rate, consideration is given to making the distance between individual elements as narrow as possible.

Incidentally, a scribing method using a diamond point is usually used to divide a silicon semiconductor element substrate, but these methods form a strained layer with a certain width on the silicon substrate, and as described above, the element spacing can be reduced. If consideration is given to making the area narrower, even a slight shift in the dividing position may have an adverse effect on the individual elements, and the characteristics of the divided silicon semiconductor elements may be impaired.

For example, in a planar transistor, the channel stopper region is created as a wide region including a dividing line, and the oxide film formed on the channel stopper region is removed, and the division is performed using the opening formed as a mark. .

That is, a method is generally adopted in which the center of the opening (one center of the channel stopper area) is visually measured using an optical microscope, but it is difficult to determine the correct dividing position, and deviations from the correct dividing position are likely to occur.

That is, the channel stopper region usually has a width of several plus micrometers or more, and it is extremely difficult to position the diamond point exactly in the center of the channel stopper region.

The present invention has been made in view of the above-mentioned drawbacks, and includes manufacturing a plurality of elements in a silicon substrate by selective diffusion treatment using an insulating film such as a silicon dioxide film (hereinafter referred to as Si02 film) as a mask, and Opposing diffusion regions are created on both sides of the micro region corresponding to the dividing line of the silicon substrate using the above selective diffusion process, and then S is formed on the opposing diffusion regions and on the micro region in between.
By selectively removing the i02 film and dividing the silicon substrate using the step formed between the silicon substrate surface of the diffused part and the non-diffused part as a mark, the selection accuracy of the dividing position is increased. , which attempts to obtain a highly reliable silicon element substrate.

Next, a method for dividing a semiconductor substrate on which planar transistors are formed according to an embodiment of the present invention will be described with reference to the drawings.

As shown in FIG.
After forming the mold base region 2, a photo-etching process is used to form a window 4 for emitter diffusion in the SiO2 film 3 covering the base region, and an Sio film covering the upper portion of the silicon substrate located between the base regions is formed. Windows 5 and 6 for forming channel stopper regions are formed in the two films 3.

When forming these windows 5 and 6, S is left between them.
The iO2 film portion 7 is located on the dividing line, and its width is approximately equal to the width of the diamond needle, for example, approximately 35μ.
Care should be taken to ensure that the distance is around m.

Next, as shown in FIG. 2, n+ type emitter regions 8 are formed in each of the plurality of base regions 2 by diffusing an n type impurity such as phosphorus.

During this diffusion process, phosphorus is diffused through windows 5 and 6, and n
+ type channel stopper regions 9, 10, and 11 are formed at the same time, and between these regions, that is, under the SiO2 film 7, there is a non-diffusion region 1 narrower than the width of the Sio2 film portion 7.
2 is also formed.

Also, at this time, a thin SiO2 film 1 is formed on the regions 9, 10, and 11.
3°14.15 is formed.

After this, an etching process is performed to open a window for electrode formation and to remove the SiO2 films 13, 14°15.7 formed on the channel stopper regions 9, 10, and 11 and on the non-diffusion region 12. .

FIG. 3 shows SiO2 films 7, 13°, 14. after the above treatment.
15 is an enlarged cross-sectional view showing the vicinity of the channel stopper region portion from which the channel stopper region portions 9 and 10 have been removed due to thermal oxidation in the emitter diffusion process.
．． New 8102 films 13, 14, 15 formed by oxidizing the silicon substrate are formed on top of 11, and by removing these new SiO2 films 13, 14, 15, channel stopper regions 9, 10, . A step is formed between the upper surface of the non-diffusion portion 11 and the upper surface of the non-diffusion portion 12 .

This step can be sufficiently identified under an optical microscope, and as shown in FIG. 3, the non-diffusion region portion 12, that is, the dividing line, is identified by the step, and a diamond needle 13, for example, is set in this portion. After this, scribing is performed and further pressing force is applied to achieve extremely accurate division.

That is, in FIG. 3, the step is the channel stopper area 9.
, 10, and the diamond needle 13 can be easily and accurately set in this portion.

The total width of the channel stopper regions 9 and 10 may be approximately several micrometers, but it is desirable that the interval therebetween be as small as the width of the dividing blade.

With this method, the minute width dividing line can be formed at the same time as the channel stopper region is formed by simply changing the mask for diffusion without increasing the number of man-hours.
Using this dividing line as a mark, it is possible to easily perform accurate division.

Although only transistors have been described here, the same applies to diodes, and the formation of steps does not necessarily require the use of a diffusion process for forming an emitter region; for example, a base diffusion process or an independent diffusion process. You can also do this by

As explained above, the method of the present invention removes an insulating film such as a silicon oxide film formed on the substrate surface by diffusion treatment, and then utilizes the step difference between the non-diffused region and the diffused region. By forming a minute dividing line and optically recognizing it, the dividing blade is applied to the line to perform the division, and it is possible to perform extremely accurate division.

[Brief explanation of drawings]

FIG. 1 is a cross-sectional view of a silicon substrate showing a state in which a diffusion mask for forming an emitter region and a channel stopper region is formed when dividing by the method of the embodiment of the present invention, and FIG. 2 shows the state after the same diffusion. 3 is an enlarged sectional view of the channel stopper region after the oxide film is removed. 1... N-type silicon substrate, 2... P-type base region, 3... Silicon oxide film, 4...
...Emitter diffusion window, 5, 6...Channel stopper formation window, 7...Silicon oxide film portion, 8...N+ type emitter region, 9, 10 ,1
1... Channel stopper area, 12...
...Non-diffusion region, 13...Diamond needle.

Claims (1)

[Claims] 1. A plurality of elements are fabricated into a semiconductor substrate by selective diffusion using an insulating film as a mask, and the above-mentioned selective diffusion or another selective diffusion is applied to both sides of the region corresponding to the dividing line of the semiconductor substrate. Two opposing diffusion regions are created with a distance between them, and then the insulating film on the opposing diffusion regions and on the fine region between them is selectively removed, and the resulting diffusion and non-diffusion regions are separated. A method for dividing a semiconductor substrate, characterized in that the substrate is divided using a step on the surface of the semiconductor substrate as a mark.