JPS63272048A - Manufacture of isolation region of semiconductor device - Google Patents

Abstract

PURPOSE:To improve the form of a field oxide film and reduce a conversion difference proportional to a bird's beak and, further, suppressing the deteriora tion in device characteristics by a method wherein silicon is evaporated on a pad silicon film and nitrogen ions are implanted into the pad silicon film simultaneously to form a silicon nitride film and then a field oxide film is formed in an aperture. CONSTITUTION:A pad oxide film 2 is formed on a silicon substrate 1 and a pad silicon film 3 is formed on the pad oxide film 2. Then, silicon is evaporated on the pad silicon film 3 and, at least simultaneously, nitrogen ions are implanted into the pad silicon film 3 with an acceleration voltage with which the maximum range of the nitrogen ion does not reach the substrate 1 and under the conditions above the required implantation rate to form a silicon nitride film 4. Then, after at least the silicon nitride film 4 is selectively etched, a field oxide film 6 is formed in the etched part 5 in a high temperature oxidiz ing atmosphere. After that, the films 2-4 including at least the remaining silicon nitride film 4 other than in the field oxide film 6 and in the silicon substrate 1 are removed by etching.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a method of manufacturing an isolation region of a semiconductor element, which forms an isolation region of a semiconductor element.

(Prior art) Conventionally, selective oxidation method CL has been used as a method for manufacturing element isolation regions.
OCO8 method) is mainly adopted.

Here, Yupinhan and Pingma; Pin
Han and BingMa; Extend
ed abstract84-IJ, Electro
chem, Soc,, 98.1984 l5OLA
TION PROCESS USING POLYSI
LICON BUFFERLAYERFOR5CALE
D MO8/VLSI”) etc. &LOC
A method of manufacturing an isolation region of a semiconductor device, which is an improved version of the O8 method, will be described with reference to the cross-sectional view of FIG.

A silicon oxide film 12 consisting of a silicon oxide film StO is grown to a thickness of 500λ on a P-type silicon substrate 11 on a crystal surface (100) in a high temperature oxidation atmosphere (see FIG. 2(&)).
reference). Next, LP-CVD (Low Pressure Chemical Peno!-Deposition; Low Press
ure Chemi-cal Vapor Dep
polysilicon 13 to 1500
It is grown to a thickness of λ [see FIG. 2(b)]. Then the same <
A silicon nitride film 14 is grown to a thickness of 1500λ using LP-CVD (see FIG. 2(c)). Next, regular photolithography is used to perform renost patterning to form an element region (active region) and an element isolation region (field region). After this patterning, the silicon nitride film 14 and polysilicon 13i are etched using a resist (not shown) as a mask. The opening 15 formed by this etching becomes an element isolation region. Next, a channel stopper for preventing inversion is placed, for example, at a B + acceleration voltage of 40 KeV and an implantation amount of I x 1013 IONS/
, 1 (see Figure 2 (d)). Next, 1
The field oxide film 16 is grown to a thickness of 6000λ by performing oxidation for about 2 hours in a wet high temperature oxidation atmosphere at 000°C (
(See Figure 2(e)). Next, the silicon nitride film 14 is removed using hot phosphoric acid, and then the polysilicon 13 is removed by dry etching or wet etching.

Further, the pad oxide film 12 is wet-etched with a hydrofluoric acid solution to expose the surface of the silicon substrate 11 and form an element isolation region in which a nine-element region and a field oxide film 16 are formed (see FIG. 2(f)). ).

(Problems to be Solved by the Invention) However, even with the methods described above, the finished reverse shape of the field oxide film 16 is stepped, as shown in FIG. In the photolithography process for patterning these materials, the focus at the stepped portion may be incorrect. It may be difficult to obtain the desired dimensions.

In order to reduce this stepped shape, it is possible to make the rad oxide film 12 thicker, the polysilicon 13t thinner, or the silicon nitride film 14 thinner. The advantage of the above-mentioned manufacturing method for improving the disadvantage, that is, preventing the increase in the density of the semiconductor element, is lost, and conversely, in order to further suppress the occurrence of bird's beak and achieve higher density of the semiconductor element. If the pad oxide film 12 is made as thin as possible and the silicon nitride film 14 is made as thick as possible, crystal defects will occur and device characteristics will deteriorate, making it impossible to obtain a technically satisfactory product.

This invention improves the shape of the field oxide film described above, and furthermore, ((finishing dimension) - (patterning dimension) =
An object of the present invention is to provide a method for manufacturing an isolation region of a semiconductor device, which can reduce the conversion difference determined by 2X (bird's beak) and suppress deterioration of device characteristics.

[Means for solving problems]

The method for manufacturing an isolation region of a semiconductor device according to the present invention includes forming a pad oxide film and a pad silicon film in this order on a silicon substrate, and then depositing silicon and implanting nitrogen ions toward the primary silicon film. At least simultaneously, a silicon nitride film is formed, then at least the silicon nitride film is selectively etched, a field oxide film is formed on this etched portion in a high temperature oxidation atmosphere, and then a silicon substrate and a field oxide film are formed in a high temperature oxidation atmosphere. Films other than the field oxide film are removed by etching.

(Function) The method for manufacturing an isolation region of a semiconductor element according to the present invention is as follows:
The silicon nitride film formed by performing at least simultaneous vapor deposition and ion implantation becomes guar as nitrogen enters the inside, and this silicon nitride film is also oxidized at a slow rate during the formation of the field oxide film. A smooth field oxide film shape can be obtained.

(Example) Hereinafter, an example of the present invention will be described in detail based on the drawings. FIGS. 1A to 1E are cross-sectional views of an embodiment of the present invention, and the film on the back surface of the silicon substrate is omitted. First, the P-type silicon substrate l on the crystal surface (ioo)
300^Grad oxide film 2 is applied on top in a high-temperature oxidizing atmosphere.
Grow thickly (see Figure 1 <&)).

Next, using a device capable of simultaneous vacuum deposition and ion implantation in the same reaction tank, pad silicon film 3 is first deposited on pad oxide film 2 to a thickness of approximately 1000λ at a growth rate of 40175y by EB deposition. (See Figure 1(b)). Even when the thickness of the vaporized M film of the pad silicon film 3 has grown to about 1000λ, silicon vapor deposition is continued (however, the vapor deposition rate may be changed) and nitrogen ions are simultaneously implanted.

For example, acceleration voltage 20KeV, nitrogen ion implantation amount l×1
Nitrogen ions are implanted under the conditions of 017ION8/, :sj. This IvD method (ion beam and vacuum evaporation method: Ion Beam
and Vacuum & aporationMeth
A silicon nitride film 4 is formed by the reaction between the deposited silicon and nitrogen ions (od), and the nitrogen ions implanted into the pad silicon film 3 react with silicon molecules to be modified into the silicon nitride film 4.

However, the silicon nitride film 4 formed in this way is not necessarily stoichiometric silicon nitride (Six N
4) Instead of ・Gutsudosilicon@3, the maximum range R
Since the accelerating voltage is set so as to have p (Projection Range) t-, the further inside from Rp, the silicon nitride film becomes smaller in the number of nitrogen atoms, and the further inside it becomes, the silicon nitride film has no nitrogen atoms implanted. The film becomes a film 3 (see FIG. 1(C)).

Next, the silicon nitride film 4 and the grading silicon film 3 are selectively etched using a patterned renost (not shown) as a mask using the usual IJ rough-height technique. The opening 5 as an etched portion formed by this etching becomes an element isolation region in which a field oxide film will be formed later.

Next, a channel straggler for inversion prevention is installed, for example, at a B + acceleration voltage of 40 KeV and an ion implantation density of lX1013.
Inject under the condition of 10Ns/, 1. When the renost is removed after this injection, the result is as shown in FIG. 1(d).

Next, oxidation is performed for about 2 hours in a wet high temperature oxidation atmosphere at 1000° C. to grow a field oxide film 6 to a thickness of 6000λ in the area including the opening 5 (see FIG. 1(e)).
.

Next, the entire silicon nitride film 4 is removed by etching with hot phosphoric acid.

Subsequently, the entire pad silicon film 3 is removed by wet or dry etching.

Then, by etching and removing all of the notched oxide film 2, an element region 7 in which the surface of the silicon substrate 1 is exposed and an element isolation region in which a field oxide film 6 is formed are formed (FIG. 1(f)). reference).

The following semiconductor device manufacturing process is the same as the normal conventional process.

In the above embodiment, the silicon nitride film 4 and the pad silicon film 3 are etched, but the silicon oxide film 2 or the silicon substrate 1 may also be etched. Alternatively, etching may be delayed up to the silicon nitride film 4.

Also, the multilayer film is etched using lenost as a mask, but lenost is removed after etching only the upper layer.
- The lower layer film can also be etched using the 4-turned upper layer film as a mask.

Further, in the above embodiment, when etching the pad silicon film 3, the pad silicon film 3 can be oxidized to form a silicon oxide film, and then this silicon oxide film can be removed.

Furthermore, in the above embodiment, the growth of the pad silicon film 3 and the growth of the silicon nitride film 4 are performed in the same apparatus.
After growing the /4' silicon film 3 using LP-CVD etc., the silicon nitride film 4 is grown using IVDII7ct.
It is also possible to form

Friend, in the above example, it is appropriate to set Rp so that the nitrogen ions are in the notched silicon film 3.
Even if Rp is set, it may be set as long as the amount of ions entering the silicon substrate l is negligible. However, in this case, nitrogen ions are necessarily included in the rad silicon film. That is, the pad silicon film 3 does not exist between the silicon nitride film 4 and the etched oxide film 2, and instead, the silicon-rich silicon nitride film plays this role.

(Effects of the Invention) As described above in detail, according to the present invention, a pad oxide film is formed, and after a /hood silicon film is formed, silicon is formed by simultaneous vacuum evaporation of silicon and nitrogen ion implantation. Since a nitride film is formed, the following four matters can be expected.

(2) Since the composition of the silicon nitride film gradually becomes richer in nitrogen guar and silicon toward the silicon substrate, stress can be relaxed, crystal defects can be suppressed, and device characteristics can be improved.

■ The unreacted pad silicon suppresses the lateral diffusion of the oxidizing agent during field oxide film growth, and the pad silicon itself consumes the oxidizing agent and turns into a silicon oxide film, which prevents lateral bird's beak. It is possible to suppress generation, reduce conversion differences, and achieve higher density.

■ The silicon nitride film, which is nitrogen guar near the pad silicon side, has a slow rate, but it is oxidized during field oxidation, so a smooth field oxide film shape can be obtained, which prevents the interconnection material from breaking later, and is also useful for photolithography. defocus can be prevented.

■ Furthermore, when growing a pad silicon film and a silicon nitride film continuously in the same equipment, the formation time is shortened which leads to cost reduction, and foreign matter can be easily removed from the wafer due to movement between the equipment or manual handling. The chance of adhesion to the wafer can be reduced, and the occurrence of element defects can be suppressed.

[Brief explanation of drawings]

FIGS. 1(a) to 1(f) are sectional views showing each process according to an embodiment of the present invention, and FIGS. 2(a) to 2(f) are sectional views showing each process of a conventional example. l...Silicon M plate, 2...Nine oxidized JI! ,
3... Pad silicon film, 4... Silicon nitride film, 5... Opening, 6... Field oxide film. Each technical field 1 field 1m by special exception Figure 1 Figure 1 Phase 21 J3 each process meresen Figure 2 Conventional III Each craftsman L = nl! V'' side view Figure 2

Claims (1)

[Claims] A first step of forming a pad oxide film on a silicon substrate, and a second step of forming a pad silicon film on the pad oxide film.
and nitriding by performing at least simultaneously vapor deposition of silicon toward the pad silicon film and implantation of nitrogen ions at an acceleration voltage that does not allow the maximum range of nitrogen ions to reach inside the silicon substrate and at a predetermined implantation amount or more. Third step to form a silicon film
a fourth step of selectively etching at least the silicon nitride film; a fifth step of forming a field oxide film on the etched portion in a high-temperature oxidizing atmosphere; and a step of etching the silicon substrate and the other than the field oxide film. a sixth step of removing the film including at least the remainder of the silicon film by etching.