JPS59182520A - Optical cvd method - Google Patents

Abstract

PURPOSE:To improve the growing speed significantly and prevent the exhaust system from contamination by a method wherein the gas, which is the same as used for the emission source of a light source, or the gas, which has light absorption near the wavelength of the emitted light, is introduced into a reaction vessel. CONSTITUTION:An Si wafer 3 is put in a reaction tube 2 and the reaction tube 2 is exhausted. Then Xe gas is introduced through a gas introducing pipe 5. At that time, the wafer 3 is heated by a heater 4. Then NH3 is introduced through a gas introducing pipe 6. After an Xe-mercury lamp of a light source 1 is lit, SiH4 is supplied into the reaction tube 2 through an introducing pipe 7. After that, the supply of the gases is discontinued in the order of SiH4 and NH3 and the lamp is put out. With this constitution, as the growing speed is improved significantly, the intensity of the light source need not be so high. Because mercury vapor is not introduced into the reaction tube, the exhaust system is not contaminated.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to chemical vapor deposition (CVD) using light, and particularly to a photoCVD method suitable for lowering the temperature and improving the growth rate.

[Background technology]

In conventional optical CVD methods, it is customary to use a light source with high light intensity, such as an ArF or KrF excimer laser. When a laser beam is used, it is difficult to form a desired film on the entire surface of a substrate wafer at once because the beam does not spread. However, due to the intra-beam intensity distribution peculiar to laser light,
Make the thickness uniform +7) There is a problem with K (J, vac
, Sci, Technol, 21(1) 23.19
82). There are examples of using a mercury lamp to avoid these problems, but in this case it is necessary to introduce mercury vapor into the reaction tube as a sensitizer. Even if such a sensitizer is added, the rate of 5iCh formation by 5iH402, for example, is slow at 2 to 4 people/s. Although the influence of this mercury vapor on the quality of the formed film is unknown, the inflow of mercury vapor into the exhaust system is expected to form an amalgam with metals, which induces deterioration of the exhaust system. In the unlikely event that mercury vapor leaks outside the equipment system, there are concerns that it may affect the human body. The conventional method had the following drawbacks (Integrated Circuit Symposium S57.6゜4, p. 90)
).

[Purpose of the invention]

An object of the present invention is to provide a safe and effective photo-CVD method that overcomes the above-mentioned drawbacks of the conventional photo-CVD method.

[Summary of the invention]

In order to eliminate the above-mentioned drawbacks of the photo-CVD method, the mercury vapor introduced into the reaction tube may be replaced with another substance to achieve the same or better effect. To do this, it is necessary to use a gas other than mercury vapor that absorbs the light from the mercury lamp, or to change the light source used and emit it.
A gas used for the c-form is introduced into the reaction tube as a sensitizer. Furthermore, in order to increase the efficiency of the chemical reaction and the rate of film formation, a gas different from the introduced gas is introduced.

This exposes the negative pole to its catalytic role.

[Embodiments of the invention]

An embodiment of the present invention will be described below with reference to FIG.

The thoroughly cleaned 7-recon wafer 3 is placed in the reaction tube 2, and the inside of the reaction tube is sufficiently evacuated using the exhaust system 9. The degree of wall thickness achieved at this time was about I x 10'-3.

After this, about 0.0% of xenon scum is added from the gas introduction pipe 5. OI
Introduce Torr. At this time, the silicon wafer 3 is heated to about 400C by the heater 4. Then, ammonia (NH3) is introduced from the gas conduit 6 at a pressure of 0.7 Torr. Then the xenon-mercury rung of light source 1 (intensity 0)
．． 5 mW/cm2), then mono 7 runs'
(S 1H4) was fed into the 0.3 Torr reactor from the inlet tube 7. After 10 minutes, the gas supply is stopped in the order of 5i) (4, NH3), and the lamp is also turned off.The horn gas is stopped with K, and nitrogen is supplied into the reaction tube from the introduction tube 8, and the inside of the reaction tube is sufficiently replaced with nitrogen. Afterwards, the silicon wafer was removed.

In this way, a silicon nitride film with a refraction of -*1.99 and a thickness of 4,000 yen was obtained. The film growth rate was about twice to 400 nm/mm compared to the conventional photo-CVD method.

In addition to the above examples, N2, He, Ar I
When L or 1 gas was introduced at 0.01 Torr, the growth rate was even higher, at 9,600 people/min.

〔Effect of the invention〕

According to the present invention, the growth rate can be significantly improved, so there is no need to use a light source with a very strong intensity t. Furthermore, since mercury vapor is not introduced into the reaction tube, the exhaust system is not contaminated and is safe.

A similar effect was found in the case of the reactant gas SiH+ alone, and although it was difficult to obtain amorphous silicon with a thickness of 400 Å or more in a single experiment using the conventional photoCVD method, the growth rate was 100 Å/100 Å. We were able to obtain a thickness of 1000 people with rr, in.

[Brief explanation of the drawing]

FIG. 1 is a system diagram showing the concept of an optical CVD apparatus. 1...Light source, 2...Reaction tube, 3...Substrate Unino・
-14... Heater for heating the substrate, 5 to 8... Gas system, 9... Exhaust system. Agent Patent Attorney Akio Takahashi No. 1 Satoshi

Claims (1)

[Scope of Claims] 1. A photo-CVD method characterized in that a gas having light absorption at or near the emission wavelength of the gas used as the emission source of the light source is introduced into the reaction vessel. 2. The photo-CVD method according to claim 1, wherein an inert gas is introduced into the reaction vessel in addition to the activated gas.