JPH08321496A - Film preparing method - Google Patents

Abstract

PURPOSE: To form a film having about the same film thickness as the one on the upper surface of a base, even in a recessed part thereof, likewise, by forming the film by using a primary activation energy possessed by active molecules generated by an optical excitation and a secondary activation energy which is possessed by a formed surface. CONSTITUTION: A reactive gas passes through a flow rate system 28 and a valve 27 in a doping system 12, reaches a gas outlet nozzle 18 and is made to enter a reaction chamber 20. On the occasion when the gas flows in directions 19, it is excited by an ultraviolet light supplied to the reaction chamber 20, makes a vapor-phase reaction such as an activation reaction or a decomposition reaction and formes a film on a surface of a base 30 whereon the film is be formed. Active molecules generated by the activation by the light oscillate on the surface at this time, but the distance of the movement is very long. In this formation of the film, however, the effect of a mean free path in the vapor phase of the reactive gas is sufficiently small compared with that of the oscillator on the surface. Accordingly, a product can be filled also in a hole or a cavity of the recessed part.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention uses a photo-CVD method (exciting a reactive gas by light to form a film-like product on a surface to be formed), in which a portion to be formed on a substrate constitutes a recess, By filling the recess with the product, the buried field insulator in the semiconductor integrated circuit is formed, and the over-etched region is filled, that is, the element is formed inside the substrate provided in the shape of a hole or a cave. To do.

[0002]

2. Description of the Related Art Conventionally, as a CVD method (vapor phase method), a reduced pressure vapor phase method and a plasma CVD method are known. This technique is widely used in the manufacturing process of semiconductor integrated circuits.

[0003]

These are the film thickness of the film formed on the side surface of the concave portion or the convex portion with respect to the film thickness (dt) of the film formed on the upper surface of the convex portion where the formation surface is uneven. (ds)
Dt / ds <1 is always in the ratio with the above, and how to approach this value to 1 was a problem to be solved in the field of miniaturization of VLSI.

Particularly, in a hole-shaped or cave-shaped recess provided in a semiconductor substrate, a recess having a depth (depth) greater than or equal to the width of the upper portion of the opening is sufficiently recessed. It was completely impossible to form a product having a uniform film thickness inside.

In the trench method, which has become a hot topic in recent years, the depth is 3 times or more, preferably 5 times or more as large as the width of the opening of the recess. Therefore, it has been impossible at all to form a coating having the same or substantially the same film thickness as that of the surface of the convex portion of the substrate even inside such a deep hole or cave.

Further, in contrast to the structure in which the cavity is formed inside the concave portion in which the internal opening is wider than the width (size) of the opening above the concave portion of the substrate, the coating is formed over the inside. It has never been possible to generate.

That is, when a conventionally known CVD method is performed, LP
In the CVD method (reduced pressure CVD method), the reactive gas is activated by thermal energy in the gas phase and reacts or decomposes. Then, this is diffused into the gas phase by the thermal energy given to the molecule and adheres to the surface to be formed to form a film. Therefore, by lengthening the mean free path of the active reaction product molecules in the gas phase, that is, when the film is formed under reduced pressure compared to normal pressure, even if the surface to be formed has irregularities, Alternatively, the adhesion to the bottom surface can be promoted. Then, for example, in the case of a pattern having a pitch of 2 μm and a depth of 2 μm and a film thickness of, for example, 1 μm is formed on the upper surface,
On the side surface, it is possible to form a film up to about 0.7 μ,
The bottom surface is only 0.6 μ and the step coverage ds
The limit was / dt ≈ 0.7.

This step coverage is exactly the same as the LPCVD method in that even in the plasma CVD method, the active reaction product diffuses to the surface to be formed and the film is formed there. Another film forming method, vacuum deposition, is ds
Only / dt ≈ 0.3 to 0.5 can be obtained.

[0009]

According to the present invention, the recesses provided in these semiconductor substrates may have a structure in which the inside is deep enough or the inside is wider than the opening at the top. , It is based on the discovery that it is possible to form a coating film of approximately the same thickness on the inner surface of the substrate as with the upper surface of the substrate by using the photo-CVD method. And as long as the width of the opening (distance between both walls) is narrower inside, the reaction product formed by the photo-CVD method is cusp into this inside as if pouring the liquid. It has been found that it can be filled without remaining.

The present invention is based on the fact that such a photo-CVD method requires a completely different film formation theory from the film formation theory which has been considered in the vapor phase method up to now, and the theory.

That is, in the photo-CVD method, the active gas excited by light can be stored for a very long time when it becomes active. This activation energy is called primary activation energy. Furthermore, the energy that sustains this active state is
It is also substantially present in the product already formed on the surface to be formed by the photo-CVD method (the upper surface of the product). This activation energy is called secondary activation energy.

Regarding the film formation, the primary active energy of this active molecule (here, it is intended to be atomic or molecular ultrafine particles chemically decomposed or reacted by photoexcitation) and the two of the surface to be formed have The active molecule oscillates on the surface to be formed in the region where the sum of the secondary activation energy is smaller.

The active molecule moves positively on the surface to be formed,
When a region in which a film is less formed is found rather than a region in which a film is already formed, the region is first attached to the region to form a film. This theory is based on the assumption that the active state of an active molecule that has once been activated by light and decomposed or reacted is extremely long. This premise could be fully proved in Example 2 described below.

If the active molecules are moved by a positive motion on the surface on which the active molecules are formed, and the surface can be positively moved without any light irradiation, the active molecules will be 2 mm even on the substrate surface.
You can move that distance. Experimentally, there is a sign of film formation due to surface positivity even at a long distance of about 5 mm. Therefore, as shown in the present invention, in the concave portion on the substrate, compared to the opening in the upper portion of the concave portion, a film having the same film thickness as the surface is formed on the inner surface in the shape of a cave which is wider in the inside where no light irradiation is performed. be able to. As a practical pattern for forming a concave portion, however, it is necessary to provide a concave portion in which the product formed by the photo-CVD method is filled so that the light irradiation is not performed and the active molecule moves to the surface up to 2 mm. It's safe.

On the other hand, as another model, the molecules having the primary active energy tend to repel each other with the coating having the secondary active energy. Therefore, the molecules having the primary activation energy move to the surface to be formed up to the region having the smaller secondary activation energy. It may be considered that the film is formed in a region where the secondary activation energy is smaller and is formed as a part of the film.

The present invention will be described below based on examples.

[0017]

Example 1 FIG. 1 shows an outline of the photo-CVD method used in the present invention.

In the drawing, a reaction system (10), an exhaust system (11),
It has a doping system (12). The substrate (30) is held in the prechamber (14) and the holder (17) and moved to the reaction chamber (20) via the gate valve (31). At this time, the preliminary chamber (14) and the reaction chamber (20) are both evacuated by the vacuum pumps (24) and (25). The reaction system is made ultra-high vacuum by a turbo molecular pump (23).

A low pressure mercury lamp (21) which emits ultraviolet light having wavelengths of 184 nm and 254 nm was arranged in the light source chamber (22).

Here, the light is emitted upward, and the heating chamber (16)
The substrate (30) is heated from the back side by the halogen heater (15).

The reactive gas is flowed by the doping system (12).
(28), through the valve (27) to reach the gas blowing nozzle (18). Then, during the flow of (19) and (19 '), the reaction chamber (20) is excited by ultraviolet light to be activated or decomposed, and a gas phase reaction of the reaction is performed, and the surface to be formed of the substrate (30) is It is formed as a film.

As a reactive gas, polysilane (SinH ₂ n ₂ n
≧ 2), polyfluorosilane (SinF ₂ n + ₂ n ≧ 2) was used.
On the other hand, by mixing with ammonia (NH ₃ ), silicon nitride is produced as a reaction product. If necessary, P or N
The impurities are added to form only a polysilane or a carrier gas to form a silicon film.

Further, an oxide gas was introduced in addition to the silicide gas to form silicon oxide, phosphorus glass, or boron glass.

Further, a metal alkylated product was introduced to form a metal, or a metal alkylated product and polysilane were mixed and supplied to form a metal and a silicide film.

[0025]

[Embodiment 2] In this embodiment, a silicon single crystal substrate (30 ′) (size 15 × 20 mm, thickness 380 μ) in the apparatus of FIG. 1 is used.
It is arranged at the position. And Si ₂ H ₆ and NH ₃
With NH ₃ / Si ₂ H ₆ = 10, the substrate temperature was about 200 ° C., and ultraviolet light was irradiated from the lower side. The pressure is 3 torr. Then, as shown in FIG. 2 (in the drawing, the coating is emphasized to be thick),
Of course, a film (33) having a film thickness of 1000Å can be formed on the lower side of the substrate (30 '). In addition, a film thickness of 1000Å can be obtained on the side surface (33 ') as well. However, it should be noted that even on the back side that is not illuminated (33``)
As shown in Fig. 3, the coating is formed on the inside over a length of about 5 mm (35). And at least 2m
The inside of up to m is that the same 1000 Å thick coating as the bottom surface is formed.

This film is a silicon nitride film, and the same applies to an amorphous silicon film.

What is more interesting is that the surface positive distance due to this wraparound seems to be hardly affected by the temperature of the substrate. It should be noted that the silicon nitride film can be formed at a temperature in the range of room temperature to 400 ° C.

Needless to say, no film is formed including the surface (33) even if the temperature is close to 300 ° C. when no ultraviolet light irradiation is performed.

This fact is due to the fact that the active molecule activated by light
Although ultrafine particles having a reactive or decomposed molecular or atomic active state are called active molecules, the distance is unimaginable for a long time. The effect of the mean free path in the vapor phase of the reactive gas on the film formation is extremely small compared to the surface positivity.

From this fact, as is known in the art, when the primary active energy of the active molecule simply collides with the surface to be formed, the energy is lost at that portion, and a film is formed by the pyrolysis reduced pressure gas phase method or plasma gas phase. A theory completely different from the film formation mechanism in the method is required.

In addition to the active primary energy of the active molecule in the photo-CVD method, the present inventor also has active secondary energy on the surface to be formed. Therefore, the active molecule has this active secondary energy. As long as the primary energy of the active gas does not reach the surface to be formed, the length of the active gas continues to oscillate for about 5 mm and is generated as a film component on the surface to be formed under the condition of more stable low energy level. It came to the theory that it would be done.

The present invention is based on the technical idea of "the secondary energy of the surface to be formed is dominantly promoting the surface positive motion of the active molecule".

[0033]

[Embodiment 3] FIG. 3 shows a silicon semiconductor substrate of the present invention provided with a U-shaped groove having a concave portion for a trench, and a U-shaped groove is provided inside the groove.
The product is filled by the CVD method.

The process will be briefly described with reference to FIG.

In FIG. 3A, a silicon nitride film (2) is formed on a silicon semiconductor (1). Using this as a mask, a hole (also referred to as a trench groove) (3) having a recess was formed at a depth of about 4 to 5 μ (width of the upper part was 1.5 μ).

After that, the silicon nitride film is removed, and the structure shown in FIG.
As shown in (B), a silicon oxide film was formed using the apparatus of Example 1.

Then, the active silicon oxide film is surface-migrated on the surface to be formed, and is also formed in the recess, which is a region having a lower secondary activation energy. As a result, in the initial state (4
-1), the upper surface (35-1) of the formation surface and the inner surface (35'-1) of the recess have approximately the same thickness. Similarly, the continuously formed (35-2), (35'-2), (35-3), and (35'-3) are also formed to have substantially the same thickness.

As a result of these, the recesses have their inner surfaces (36),
Since the product is formed on both sides of (37), the product can be filled with the product sufficiently faster than the upper surface of the substrate. As a result, the upper surface of the substrate can be made substantially the same flat surface or a smooth continuous surface as shown in FIG. 3 (B).

If necessary, as shown in FIG. 3C, only this upper surface may be selectively removed by isotropic etching, and (8) and (8 ') may be made flat. ) The board
It may be provided so as to coincide with the upper surface of (1).

In FIG. 3, the silicon oxide film is filled with one kind of insulator. However, only the upper surface of the film 4-1 may be thermally oxidized silicon oxide or silicon nitride. In this case, the inside may be considered as the concave portion in FIG.

However, a semiconductor having such a recess is likely to generate lattice defects on the bottom surface of the recess at high temperature. Therefore optical CV
The D method can form a film at about 300 ° C, and its product is extremely dense and has the same properties as a film formed at 1000 ° C or higher, so it is extremely preferable without inducing such lattice defects in the recesses. there were.

[0042]

EXAMPLE 4 In this example, the film formation in the process of Example 3 was produced with a different film.

For example, a polycrystalline or amorphous silicon film in which phosphorus is added as an impurity to a silicon nitride film or a silicon oxide film (4) or a metal conductor film (5) such as titanium chloride or tungsten on a silicon semiconductor, and silicon oxide ( 6) is laminated as a product. A capacitor is formed between the semiconductor substrate (1) and the polycrystalline single crystal silicon film (5) and the substrate.

[0044]

Embodiment 5 This embodiment shows the process in FIG. In order to increase the capacity of the capacitor, a cavity is provided inside the recess and this is used as the capacitor. The process is abbreviated below.

In FIG. 5A, a recess is formed to a depth of 5 μm as in the third embodiment. The width of the upper part is about 2.5 μ and the inside is about 1.5 μ. Further, thereafter, silicon nitride films (45) and (45 ′) are formed by the photo CVD method. Then, as shown in (A), a film can be formed sufficiently uniformly inside the recess.

Further, as shown in FIG. 5C, an anisotropic plasma etching method is used. Then, only (45 ') and the bottom region (39) of the recess can be selectively removed. Further thereafter, anisotropic etching was performed on the silicon substrate. Silicon can be selectively etched laterally according to the degree of etching. And you can make a cave (40).

Thereafter, a silicon oxide film (41) having a thickness of about 1000 Å was formed on the silicon formation surface by photo CVD. Still another photo-CVD method was used to form a first phosphorus-doped silicon film or titanium silicide (42). A body film (4) such as a silicon oxide film or a silicon nitride film is formed by an optical CVD method. Further, the conductive film (5) is formed of, for example, a phosphorus-added polycrystalline silicon film or a conductor such as titanium silicide. After that, a silicon oxide film (46) for filling is formed. Contact (4
4) and (43) are provided as a pair of electrodes of the capacitor.

By doing so, the capacitance of the capacitor can be increased.

[0049]

As is apparent from the above description, the present invention is based on the theory that the surface positivity of active molecules is important in the film formed by the photo-CVD method and the experimental fact thereof. As a result, it is now possible to fill the holes or cavities of the recesses, which were previously impossible with the thermal CVD method, with the product.

The same applies not only to silicon as a semiconductor but also to other compound semiconductors.

Along with the future development of the photo-CVD method, a larger application is expected.

[Brief description of drawings]

FIG. 1 shows a plan view of a photo-CVD apparatus used in the present invention.

FIG. 2 is a semiconductor cross-sectional view according to a basic experimental example showing the principle of the photo-CVD method of the present invention.

FIG. 3 shows an embodiment of the present invention.

FIG. 4 shows an embodiment of the present invention.

FIG. 5 shows an embodiment of the present invention.

Claims (2)

[Claims] 1. A step of forming a recess or a cave having at least one inner surface on the surface, a step of supplying a reactive gas containing at least a metal alkylated product, and a step of CVD using the reactive gas to form the recess. Alternatively, when the thickness of the coating on the inner surface of the cave is ds and the thickness of the coating on the surface is dt, substantially ds /
A method for producing a coating film, characterized in that a layer containing at least one metal is formed on the surface so that dt = 1. 2. A step of forming a recess or a cave having at least one internal surface on the surface, a step of supplying a reactive gas containing at least a metal alkylated product, and a CVD method from the reactive gas, respectively. When the thickness of the coating film on the inner surface of the recess or the cave in the layer is ds and the thickness of the coating film on the surface is dt, the metal is made substantially ds / dt = 1. A method for producing a coating film, comprising forming a plurality of layers including the same on the surface.