JPS6283400A - Method of improving cylinder for vapor growth of organometallic compound - Google Patents

Abstract

PURPOSE:To obtain a thin crystal film having uniform composition by providing a porous diffuser at the lower part of the dip tube of a dopant of the cylinder for the vapor growth of an organometallic compd. and uniformize the amt. of the dopant to be injected. CONSTITUTION:A porous diffuser such as a plastic ball filter and a cell ball disk having pores of 1-100mu size is provided at the lower part of the dip tube of a dopant of the cylinder for the vapor growth of an organometallic compd.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to metal organic vapor phase epitaxy.
icChemical Vapor [)eposit
ion hereafter referred to as HO-CV [)] 10-C
This article concerns a method for improving VD cylinders.

[Conventional technology]

Currently, it is well known that HO-CVD is performed in order to increase the production efficiency of semiconductor lasers.

)10-CV[) is to send organic metal compounds as gas onto a wafer under high temperature, cause these gases to react on the wafer, and grow a single crystal thin film of the compound produced by this reaction. It is.

More specifically, for example, (Ct-13)3Ga and AS)-13 gas are reacted on a wafer under high temperature, and the following reaction formula (CHa)aG is obtained.
a+AS)-Ll-)GaAs+3C
This is a method of growing a Ga AS single crystal thin film using H+.

In this case, other organometallic compounds such as (C
) la > 3Zn and doping with Zn is often practiced.

FIG. 1 shows a schematic diagram of the No-CV[) device. That is, the wafer 1 is heated to a high temperature by the heater 2, and in the above example, AsHa is in a gaseous state, so a cylinder is not required, but (CH3) 3 G is present on the wafer 1.
Since a and (C)fa)32rl as a dopant are liquids, they are stored in a cylinder 7, and an inert gas 3)-1e is blown into the cylinder through a dip tube 8 while controlling the flow rate to form an organometallic compound. A gas is generated and sent onto the wafer, causing a chemical reaction and growing a single crystal thin film on the wafer.

(Problem to be solved by the invention) NO-CVD Ni# (+NT, (CHa) 3 Ga
Although the main raw material gas such as 1As 83 is relatively large at a gas inlet of approximately 200 m/sec, the amount of inert gas introduced into the dopant is usually extremely small at 1 to 20 rIiIl/min, so it is difficult to prevent bubbles from entering the dip tube. The generation conditions are extremely poor, making it difficult to flow the gas at a constant flow rate.As a result, the amount of dopant implanted becomes undefined, making it impossible to form a thin crystal film with a uniform composition.

(Means for Solving the Problems) Therefore, as a result of various studies, the present inventors installed a bolus disperser not only at the dip tube for the main raw material, but especially at the tip of the dip tube for the dopant. It has been found that even better 10-CVD can be achieved.

(Function) Therefore, in the manner described in the following test example, a dip filter (NUPRO FILTER) (Nupro Company: 4800 East 345) was used as a dispersion device.
44094 Street, Willaukbai, Ojai
4800 East 345th 5treet
.

Willonohby, 0hio 44094) S S
- Various types of 2F series sintered filters were installed, and the state of foaming from the disperser was observed.

Test 1゜ Using the apparatus shown in Figure 2, test liquid 1 was placed in beaker 15.
Ethanol as 4 (liquid temperature 28℃, 1, Oc, p,)
(room temperature: 29°C), and as a filter for the disperser 13, use a filter of 2.7.15.60μ from NUPROCo's S S-2F series,
The generation of bubbles in the filter was observed at an NZ rate of 10 d/min.

The results showed that all the foaming conditions were good.

Test 2゜Under the same conditions as Test 1, only test solution 14 was replaced with ethanol and propatool (liquid temperature 28℃, 1.76C, p)
The state of foaming was also observed under the same wave conditions (liquid temperature: 0°C, 4.6c, p). The results were almost the same as Test 1, and it was found that the foaming condition was extremely good.

In this way, the mesh size of the filter is 15~
It was found that about 60μ is good. NUPROF in the exam
Although using ILTER, but not limited to,
For example, it may be a plastic ball filter or a Cellpol disk, or it may be made of stainless steel or ceramics, or it may be fibrous.
As long as it is durable and has a uniform bolus, a fine bolus disperser of 1 to 100μ is installed, and N
It was found that o-CVD was sufficient.

Various organometallic compounds such as dialkylzinc, dialkylcadmium, and dialkyltellurium are considered as dopants.

〔Example〕

Example 1゜(CH3) 3 Ga and AsH3 as main raw materials on the substrate
When performing GaAs@cVD using (CI-13)zZn as a dopant, He was used as a carrier gas and the flow rate to the dopant was set at 15 d/mi.
n and put NUPROFILTER in the dip tube.
S S -2F series 7μ and 60μ were used, and a good thin film was obtained using io-CVDu. However, in the case without the filter, the film formation was non-uniform.

[Brief explanation of drawings]

Figure 1 is a schematic diagram of a 10-CVD apparatus, in which 1.
... Wafer 2 ... Heater 3 ... Inert gas 4.5.6 ... Organometallic compound 7 ... Cylinder 8 ... Dip tube 2nd
The figure is an explanatory diagram of the main parts of the apparatus used in the method of the present invention, in which 9... Inert gas 10... Pressure gauge 11... Flow meter 12... Dip tube 13... Dispersion Container 14...Test liquid 15...Beaker Figure 1 Procedural amendments November 13, 1985 Official of the Bureau of Corrections Urimanto aUS 1, Indication of the case 1985 Patent application No. No. 219680 No. 2, Title of the invention: Method for improving cylinders for organometallic vapor phase growth (Tokyo Kyobashi Building) Name: Toyo Straf Co., Ltd. 7-Chemical Specification Column Detailed explanation of the invention Contents of amendment 1: Specification r (CH
:l) 3 ZnJ and Alwor (CH3) 22nJ
Corrected each. 2. On page 3, line 7, the text "r200ml/sec" was corrected to "r200d/min." 3. In the 5th line of page 4 of the same page, the word ``Uiraukubai'' has been changed to ``[Kuilauhubai'']. 4. On the 7th line of the same page, rWillonqhby j and must be corrected to rWillonqhbyj and rWillonqhbyj.

Claims (1)

[Claims] (1) An improved method for a cylinder for organometallic vapor phase epitaxy, characterized in that a bolus disperser of 1 to 100 μm is attached to the lower part of the dopant dip tube in the cylinder for organometallic vapor phase epitaxy.