JPS60107840A - Manufacture of semiconductor element - Google Patents

Abstract

PURPOSE:To obtain an oxide film having uniform film thickness with excellent reproducibility by heating a semiconductor substrate in an oxidizing atmosphere gas containing a very small amount of water content in quantity more than maximum water content discharged from an oxidizing atmosphere system and less than 1,000ppm when the oxide film is formed on the surface of the substrate. CONSTITUTION:A quartz jig 3 on which a large number of Si substrates 1 are erected at intervals is received in a quartz pipe 2 surrounded by a heating furnace 12, oxygen to which a very small amount of water content is added is flowed into the quartz pipe 2 and the substrates 1 are heated at approximately 1,000 deg.C, and SiO2 films are formed on the surfaces of the substrates. In the constitution, a purifier 4 for oxygen is disposed outside the quartz pipe 2, oxygen from the purifier is bubbled in a quartz vessel 6 receiving pure water 7 fitted in a thermostatic chamber 5, and water content is regulated by a water content densitometer while a very small amount of water content is made to be contained in oxygen and oxygen is flowed into the quartz pipe 2. Water content is regulated in quantity more than maximum water content discharged from the quartz pipe 2 and less than 1,000ppm.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a method for manufacturing a semiconductor device, and more particularly to a method for oxidizing a semiconductor substrate with good uniformity and reproducibility of oxide film thickness.

[Background of the invention]

In recent years, as MO8/LSIs have become more highly integrated and faster, gate oxide films have become thinner. As is well known, the gate oxide film thickness is an important parameter that determines the electrical characteristics of the MOS, such as the threshold voltage (VTII), and variations in the oxide film thickness can be directly reflected in VTR variations. It will be done. This is particularly noticeable when the oxide film is thin, and has become a major problem in the gate oxide film formation process in MO8/LSI. As a method for forming a gate oxide film, dry oxidation is generally used, which is a so-called thermal oxidation method in which a semiconductor substrate is heat-treated in an oxidizing dry atmospheric gas. This is because the film thickness uniformity and reproducibility are superior to other methods. In order to reduce the thickness of the film in thermal oxidation methods, common methods include lowering the oxidation temperature and mixing non-oxidizing gas such as nitrogen into oxygen to lower the oxygen partial pressure in the oxidizing atmosphere. However, this method has a problem in that it is insufficient in terms of film thickness uniformity and reproducibility in response to MO8/LSI, which is becoming more highly integrated and faster.

[Purpose of the invention]

Therefore, an object of the present invention is to provide a method for manufacturing a semiconductor element with good uniformity and reproducibility of oxide film thickness.

[Summary of the invention]

The present invention is characterized in that a semiconductor substrate is heat-treated in an atmospheric gas in which a controlled amount of moisture is added to a conventionally used oxidizing dry atmospheric gas.

The present invention will be explained in more detail below.

Generally, a gate oxide film of MO8/LSI or the like is formed by heat-treating a semiconductor substrate in oxygen or a mixed gas of oxygen and nitrogen. However, this method has the problem that the film thickness varies from oxidation lot to lot, and the reproducibility of uniformity is insufficient. A detailed investigation into the cause revealed that trace amounts of moisture contained in the oxidizing atmosphere were involved. The trace amount of moisture contained in the oxidizing atmosphere is thought to be contained in the introduced oxygen or nitrogen, and released from the quartz tube, the semiconductor substrate, and the jigs supporting the semiconductor substrate during heating. As a result of analyzing the moisture concentration in the former introduced oxygen or nitrogen, it was found to be 0.5 to 0.6 pI)m
It was confirmed that the value remained constant. Therefore, the moisture released from the quartz tube, the semiconductor substrate, and the support for the semiconductor substrate during oxidation is considered to be the cause of the variation in film thickness. It is difficult to quantify this released moisture because there is no way to directly analyze it. Therefore, we added a small amount of water to oxygen and investigated the relationship between water content and film thickness in detail. The results are shown in FIG. As can be seen from the figure, there is a linear relationship between the oxide film thickness tow and the water concentration Nvx2o (toxoc 10gNa2o) in the range where there is a lot of moisture, but about 2p
It varies below pm. From this result, it can be inferred that the moisture released from the quartz tube, the semiconductor substrate, and the support for the semiconductor substrate described above varies from lot to lot within a maximum range of 2 pI)m or less. It is extremely difficult to eliminate these released moisture and keep it at a constant value. Therefore, it is necessary to add moisture to the oxygen in advance to the extent that even if the uncontrollable released moisture fluctuates, it will hardly affect the film thickness.
If oxidation is performed within a range where the relationship ox6c logN2G holds true, variations will disappear.

The present invention was devised based on this idea.

Note that the maximum value of moisture released from an oxidizing atmosphere system such as a quartz tube, a semiconductor substrate, and a support for a semiconductor substrate is - for example, about 2
1) Although it was stated that it was pm, this value is based on the inner diameter of the quartz tube,
It varies depending on various factors such as length, material and drying state or drying state of the semiconductor substrate, quantity of input, and flow rate of introduced gas. Therefore, the amount of water that should be added to the introduced oxidizing gas is:
The effects of the present invention cannot be obtained unless the amount of water released is equal to or greater than the maximum value in each oxidation system. In addition, if the amount of added water is large, it becomes difficult to handle thin films with a high rate of oxide film formation, so it is practically desirable that the amount of water is 11,000 pp or less.

As explained above, according to the present invention, the influence of uncontrolled released water is eliminated, so the uniformity of the thickness of the produced oxide film is greatly improved.

[Embodiments of the invention]

Examples of the present invention will be described below.

FIG. 2 shows the entire oxidation system of one embodiment of the present invention. The semiconductor substrates used were silicon wafers 51 (20 pieces) having a plane orientation of (100), n-type conductivity, resistivity of 20 Ωm, diameter of 76 w, and thickness of 500 μm. First, silicon Ueno S1 is placed in a quartz tube 2 kept at 900C and a quartz jig 3 is placed inside the quartz tube 2.
Insert with. After flowing 3 tons of oxygen containing 100 ppm of moisture into the quartz tube 2 and holding it at 100, Silicon Ueno-1 was taken out from the quartz tube 2. A quartz tube 10 is placed inside a heating furnace 12 . In addition, moisture 100 [)
I)m was added as follows. First, oxygen is passed through the gas purifier 4 to a moisture concentration of about 0.5 ppm.

Next, this dry oxygen is branched into two parts, one of which goes directly into the quartz tube 2, and the other part of which is kept in a constant temperature bath 5 for 10 minutes.
The dry oxygen was allowed to pass through the surface of the pure water 7 in the quartz container 6 which was maintained at a temperature of 2.5 °C. In front of the quartz tube 20, there is a moisture concentration meter 8.
Now I checked the water content and adjusted the water concentration to 1100 pp with valve 9-1.

An ellipsometer was used to evaluate the thickness of the oxide film formed on Silicon Ueno-1. The film thickness was measured at a total of 5 points for each wafer: 1 point in the center and 4 points on the periphery (points 10 mm from the edge of the Crycon wafer and at 90° angles to each other).
It is a point. All wafers (20 wafers) were measured in this manner. Furthermore, similar experiments were conducted on 10 lots, and the reproducibility was confirmed. The results are shown by curve A in FIG. For comparison, 10 lots were also conducted using the conventional method without adding water (water concentration in introduced oxygen was 0.5 to 0.6 ppm), and are shown as curve B in FIG. In order to match the oxide film thickness with the example of the present invention, in the experiment using the conventional method, 900G, 1
A heat treatment of 20- was carried out. As a result, curves A and B in Figure 3
As is clear from the above, the variation between velocities in the example of the present invention was extremely small compared to the conventional method, and the effect of the present invention was confirmed.

In addition, silicon wafers were heated at 100 OC in an atmosphere containing 50 ppm of water in a mixed gas of oxygen and nitrogen (oxygen flow rate: 0.3 t/-, nitrogen flow rate: 2.7 t/m).

The results of heat treatment in the order of 100 (other conditions were the same as in the previous example) are shown by curve C in FIG. Moreover, the curved line in FIG. 4 shows the result of the conventional method in which heat treatment was performed at 1000C and 115-degrees. As is clear from these results, the variation in the examples of the present invention was extremely small compared to the conventional method, and the effects of the present invention were confirmed. Furthermore, even when experiments were carried out by varying the oxidation temperature, time, oxygen partial pressure, and amount of added water, the results were similar, confirming the effect of the present invention.

In the above example, a bypass method using manual pulp was mainly used as a simple method to add a small amount of moisture to dry oxygen or a mixed gas of oxygen and nitrogen, but it is possible to control the amount of small amount of moisture with precision. The effects of the present invention can be achieved by any other method as long as it is possible.

Fifth. Figure 6 shows an example of a trace moisture control system.

First, FIG. 5 will be explained. The system shown in FIG. 5 includes a gas purification device 4, a moisture addition device 13, and a moisture sensor 14. The gas purification device 4 removes impurities such as hydrocarbons and unnecessary moisture from the p1 source gas, and then the moisture addition device 1
Add a predetermined amount of water to step 3. The moisture concentration of the gas exiting the moisture addition device 13 is detected by a moisture sensor 14, and feedback is applied to the moisture addition device 13 to accurately control the added moisture so that the desired moisture concentration is achieved.

Next, FIG. 6 will be explained. The system shown in FIG. 6 includes a gas purification device 4, a moisture addition device 13, a moisture removal device 15, and a moisture sensor 14. First, the raw material gas is purified in the same manner as in the case of FIG. 5, and then a certain amount of water at a concentration higher than the desired water concentration is added using the water addition device 13. The gas emitted from the moisture addition device 13 is passed through the moisture removal device 15,
Remove a predetermined amount of moisture. By detecting the moisture concentration of the gas emitted from the moisture removal device 15 with the moisture sensor 15 and controlling the amount of moisture removed by applying feedback to the moisture removal device 15, the moisture concentration can be precisely controlled to a desired moisture concentration. There is. Further, in FIG. 6, feedback may also be applied to the water addition device 13 to control the amount of water added at the same time.

〔Effect of the invention〕

In this way, according to the present invention, an oxidation method with extremely good film thickness uniformity and reproducibility can be realized even in areas where the oxide film is thin.
The characteristics and manufacturing yield of O8/LSI etc. can be greatly improved.

[Brief explanation of drawings]

FIG. 1 is a diagram showing the relationship between the moisture concentration contained in introduced oxygen and the oxide film thickness of silicon and wafers heat-treated in this atmospheric gas, and FIG. 2 is a diagram showing the oxidation system in one embodiment of the present invention. FIGS. 3 and 4 are diagrams showing variations in oxide film thickness in an example of oxidation according to the present invention and variations in the conventional method, and FIG. FIG. 6 is a diagram showing an example of a trace moisture control system. DESCRIPTION OF SYMBOLS 1... Silicon wafer, 2... Quartz tube, 3... Wafer holder, 4... Gas purification device, 5... Constant temperature chamber, 6... Quartz container for pure water, 7... Pure water, 8... Moisture permeability meter, 9-11... Pulp, 12... Heating furnace,
13... Moisture addition device 1st eye 4\both (in the system +: -3: +It 3L'if** N
Hzo (F'Pm) also 3rd mouth, soto 1115 0t person do 3 Ryo 6 figure

Claims (1)

[Scope of Claims] 1. A method for manufacturing a semiconductor device, which comprises heating a semiconductor substrate in an oxidizing atmosphere to which a trace amount of moisture is added to form an oxide film on the semiconductor substrate. 2. In item 1, the trace amount of water added is greater than the maximum amount of water released in the oxidizing atmosphere, and is 11,000 pp.
A method for manufacturing a semiconductor device, characterized in that the amount is as follows.