JPH08255777A - Semiconductor wafer cleaning method - Google Patents

Abstract

PURPOSE: To provide a cleaning method to obtain a II-VI compd. semiconductor wafer having a flatness and good growth pane suited for the epitaxial growth. CONSTITUTION: A method of cleaning a II-VI compd. semiconductor wafer, characterized in that after exposing a wafer to H excited by RF discharge, microwave discharge, ECR discharge or d-c discharge, a beam of atoms of an element of which the wafer is composed, or molecules of a compound containing such an element is applied to the wafer.

Description

Detailed Description of the Invention

[0001]

The present invention relates to ZnS, ZnTe, C
dS, CdSe, CdTe, HgTe, HgSe, Hg
The present invention relates to a method for cleaning the surface of a II-VI group compound semiconductor substrate such as S and mixed crystals thereof.

[0002]

2. Description of the Related Art Conventionally, when a II-VI group compound semiconductor is epitaxially grown on a II-VI group compound semiconductor substrate by the MBE method, a method for cleaning the substrate in a growth apparatus is, for example, in the case of a ZnSe-based substrate. Cleaning by heating, etching by sputtering with an inert gas, or reactive ion etching (RIE, Reactive Ion Etching) with a reactive gas has been performed.

[0003]

A method of cleaning by heating when growing by the MBE method using a II-VI group compound semiconductor substrate is described in J. Cryst. Growth 86 (19).
88) p.342-347. In this method, when the substrate is placed in a vacuum and the temperature is raised, a spot pattern is observed by high-speed backscattered electron diffraction (RHEED) observation. This indicates that the surface is rough at the atomic level, and flatness suitable for epitaxial growth has not been obtained. In addition, since the oxide film of several atomic layers and the carbon-based adsorbate remain on the outermost surface layer, good epitaxial growth cannot be performed.

Next, a method of performing sputtering etching with Ar gas is described in J. Vac. Sci. Technol B3 (1985) p.
1637-1640. This is because no suitable chemical etchant such as sulfuric acid and hydrogen peroxide-based mixed acid in GaAs has been found in II-VI group semiconductors.
Etching is performed by sputtering. In this case, since the surface is sputtered by the kinetic energy of Ar ions, the processing strain layer remaining during polishing can be removed, but a new damaged layer is formed by sputtering, the surface morphology is rough, and only a spot pattern can be obtained by RHEED observation. Therefore, good epitaxial growth cannot be performed.

The method of RIE is described in J. Vac. Sci.
Technol B9 (1991) p.1934-1938.
This is to minimize the drawback of sputtering etching, that is, damage due to sputtering, and to chemically dry-etch the surface by using active species generated from plasma using BCl ₃ as a reactive gas. Is. In this case, the RHEED observation has a streak pattern showing flatness at the atomic level. However, the surface morphology becomes rough due to the residual damage due to the kinetic energy of the ions.

Therefore, the present invention solves the above problems and provides a substrate cleaning method for obtaining a II-VI group compound semiconductor substrate which is suitable for epitaxial growth and has excellent flatness and surface morphology without damage. It is what

[0007]

The present invention has solved the above problems by irradiating a hydrogen gas excited to a high energy state and a beam of an element or a compound containing the element which constitutes the substrate II- This is a method for cleaning a Group VI compound semiconductor substrate, and the details are as follows.

(1) RF for a II-VI group compound semiconductor substrate
(Radio Frequency) discharge, microwave discharge, ECR (El
ectron Cyclotron Resonance) discharge, or after irradiating with hydrogen excited by direct current discharge, a beam of the element constituting the substrate or a compound beam containing the element is irradiated. Method for cleaning compound semiconductor substrate.

(2) The irradiation start temperature of the excited hydrogen gas is adjusted to a temperature between room temperature and 700 ° C.
(1) A method for cleaning the surface of a II-VI group compound semiconductor substrate as described above.

(3) The irradiation start temperature for irradiating the beam of the element constituting the substrate or the compound containing the element is 100 to
Above (1) characterized by adjusting the temperature between 700 ℃
A method for cleaning a II-VI group compound semiconductor substrate according to the description.

(4) The II-VI group compound semiconductor substrate is II
Any one of the above (1) to (3), characterized in that the group element is a binary compound of two or more elements selected from Zn, Cd and Hg and the group VI element selected from S, Se and Te. Method of cleaning a II-VI group compound semiconductor substrate according to item 6.

[0012]

In the present invention, a plasma state is created by discharging hydrogen gas, hydrogen is excited to a high-energy state of atoms or molecules, and the II-VI group compound semiconductor substrate is irradiated with the hydrogen, so that Zn-O The bond with oxygen is completely reduced to remove the oxide. Further, the carbon-based residue is converted to C—H-based hydrocarbon and evaporated. Since these series of reactions are carried out particularly without giving kinetic energy, it is possible to form a flat clean surface without a damaged layer. And
In order to prevent the evaporation of the constituent elements from the substrate at these series of process temperatures, the beam of the constituent elements is irradiated to suppress the deterioration of the surface.

Although the following examples describe ZnSe substrates, the present invention is not limited to these.
ZnS, ZnTe, CdS, CdSe, CdTe, Hg
The same is also valid for Te, HgSe, HgS and mixed crystal semiconductor substrates thereof. Further, high frequency discharge was used in the following examples, but microwave discharge, ECR discharge, and direct current discharge can be used as well.

[0014]

EXAMPLE After the surface of the ZnSe substrate was irradiated with hydrogen excited by 13.56 MHz high frequency discharge to clean the surface, ZnSe was epitaxially grown. The above-mentioned substrate cleaning and epitaxial growth were performed by the MBE apparatus shown in FIG. This MBE device is provided at the bottom of the ultra high vacuum chamber 1,
A raw material evaporation cell 2 and a discharge cell 3 are additionally provided, a semiconductor substrate is fixed to a holder 4 at a position corresponding to the cell, and a substrate heating heater 7 provided on the back surface thereof is heated to a predetermined temperature. A cell shutter 6 is provided for each cell and a main shutter 8 is provided on the surface of the substrate to adjust the beam irradiation time. A hydrogen gas 10 supply pipe was connected to the discharge cell 3, and a liquid nitrogen shroud was arranged near the inner wall of the ultra-high vacuum chamber 1.

A ZnSe (100) wafer was used as the substrate. This wafer had G
e (400) asymmetric diffraction in two-crystal diffraction measurement ZnSe
The (400) diffraction half-value width is about 10 seconds, which shows that the crystallinity of the substrate is good and there is no polishing damage. It should be noted that the theoretical full width at half maximum of diffraction is about 8 to 9 seconds in a perfect crystal. When there is damage due to polishing or when the crystallinity of the substrate is poor, the half-width of diffraction becomes as large as 20 seconds or more.

This wafer was degreased by organic cleaning with acetone, trichloroethylene and isopropyl alcohol, and then placed in the above MBE apparatus without chemical wet etching. After that, when examined by RHEED observation, a 1 × 1 pattern was confirmed even at room temperature, and it was found that the substrate characteristics were good and the single crystal was formed up to the outermost surface. When the outermost surface of the wafer is polycrystallized or amorphous and damage remains, RHEED exhibits a ring-shaped or halo-shaped pattern.

Then, hydrogen gas purified through the Pd permeable film was introduced into the ultra-high vacuum chamber through the high frequency discharge cell at a rate of 1.0 sccm. The ultimate vacuum in the chamber at this time was about 1 × 10 ⁻⁶ Torr. With a stable flow rate, 100 W of 13.56 MHz high frequency was applied to cause plasma discharge. As a result of plasma spectroscopy, it was found that the emission of hydrogen atoms was dominant.

After the plasma discharge is stabilized, the substrate temperature is raised to 500 ° C. at a temperature rising rate of 10 to 20 ° C./min according to the temperature program of FIG. 2 while flowing excited hydrogen under the above conditions.
Heated up. The above substrate temperature is the temperature near the heater of the manipulator. 1x1 with increasing temperature
The intensity of the pattern gradually increased, and the Zn-stabilized surface of the C (2x) structure was observed as a streak pattern at 450 to 500 ° C. FIG. 3 shows RHEE observed when an electron beam with an accelerating voltage of 15 kV is incident from <100> of the epitaxial layer and three directions equivalent thereto, that is, four directions in total.
It is a photograph of the D pattern. Diffraction at the position indicated by the arrow appears at the position of 1/2 of the normally observed 1 × 1 pattern. With <110> incidence orthogonal to this, only a 1 × 1 pattern appeared. This is <10
The 0> direction shows four-fold symmetry and shows a long period structure that is twice as long as a normal lattice arrangement, and it can be seen that a so-called C (2 × 2) structure is formed.

X-ray electron spectroscopy (XPS) measurement of the surface was carried out without taking out the wafer from the chamber.
The O bond and the carbon signal were below the measurement limit. Thus, as a result of the removal of oxygen from the surface of the substrate, the most stable rearrangement at this temperature was achieved, indicating that the Zn is in a stabilized state. Further, the diffraction pattern of RHEED becomes a spot when the wafer is not flat, but when the wafer is flat, the spots are connected to show a streak pattern. Therefore, as described above, the streak pattern of the C (2 × 2) structure is shown. This means that it is flat at the atomic level.

For comparison, the substrate was cleaned by the temperature program of FIG. 1 in the same manner as in Example 1 while flowing hydrogen without discharging, but the temperature was raised from 500 ° C. to 700 ° C. ( Other conditions are the same, only the (1 × 1) pattern of FIG. 4 was obtained, and the C (2 × 2) pattern (FIG. 3) was not obtained. Note that FIG. 4 is a photograph of the RHEED pattern observed when an electron beam is incident from <100> and <110>. In this case, it was found that only 1 × 1 pattern was observed regardless of the direction of incidence, and surface rearrangement did not occur. Also, 700
When the temperature was raised above 0 ° C, the dissociation of Zn and Se from the surface became remarkable, and the surface could deteriorate, so that it could not be used as a substrate. In this state, when the surface was examined by in-situ XPS measurement, Zn—O bond and carbon signal were observed,
It was confirmed that a clean surface was not obtained.

As described above, the excited hydrogen is irradiated and heated, and when the atomic level flat and clean surface is obtained, the irradiation of the excited hydrogen is stopped, and immediately among the constituent elements of the substrate, S
The irradiation of e was started. At the same time as this irradiation, the substrate temperature is raised to 10
The growth temperature was lowered to 350 ° C. at a rate of ° C./min. This is to suppress deviation of stoichiometry on the surface when the oxide film on the surface of the substrate is removed. FIG. 5 is a photograph of a RHEED pattern observed when an electron beam is incident from <110> of the epitaxial layer. Diffraction at the position indicated by the arrow appears at the position of 1/2 of the normally observed 1 × 1 pattern. Further, in <100> incidence and <110> incidence orthogonal to this, only 1 × 1 pattern appeared. This indicates a long-period structure in which the <110> direction is two-fold symmetric and is twice as long as a normal lattice arrangement.
1) It can be seen that the structure is formed.

Next, epitaxial growth of undoped ZnSe was performed on the surface prepared as described above. The flux is Se 2 × 10 ^-6 Torr and Zn 1 × 10 ^-6 T
orr. After 2 hours from the start of Se irradiation, Z
Irradiation with n started the epitaxial growth. A streak pattern was formed immediately after the start of growth, and it was found that two-dimensional growth started from the initial stage. RHEE growing
The D pattern showed a Se-stabilized 2 × 1 pattern. Three
When the substrate was taken out after the time growth, the epitaxial layer thickness was 1.5 μm.

When Hg-Cd laser light is irradiated by the photoluminescence method and the fluorescence spectrum at 4.2K is observed, the emission of free excitons is dominant, and GaAs
Neither the separation of the free exciton peak due to the strain nor the Y emission due to the defect as observed in the above heteroepitaxial growth was observed. Further, in order to examine the crystallinity, epitaxial growth with a thickness of about 4 μm was carried out, and the above-mentioned two-crystal X-ray diffraction measurement was carried out. Was there.

For comparison, only epitaxial hydrogen cleaning was carried out, and epitaxial growth was carried out in the same manner as above, but without Se beam irradiation. From the state where a clean Zn-stabilized surface of C (2 × 2) appeared in RHEED observation by excited hydrogen irradiation, the manipulator temperature was raised to 350 ° C. which is a growth temperature.
The temperature was lowered at a rate of ° C / min. In this state, a C (2 × 2) Zn stabilization pattern was still observed by RHEED observation.

As a method of starting the growth, there are three methods of first irradiating Se or Zn or simultaneously irradiating them. In any method, RHEED is spotted when the growth is about 5 nm from the start of the growth. It grows in three dimensions and becomes a streak pattern after that. After growth,
The half-width of the two-crystal X-ray diffraction was about 20 seconds, which was about the same as that of Se irradiation. However, when the cross section was observed with a transmission electron microscope, the defect density at the growth interface was 1 compared with the case of Se irradiation. It turned out to be an order of magnitude larger.

In case of not irradiating with excited hydrogen, as shown in FIG.
The C (2 × 2) plane did not appear as shown in FIG. When an epitaxial growth is attempted on this surface, a single crystal grows, but initially grows three-dimensionally, the half-width of the two-crystal X-ray diffraction after the growth is about 50 seconds, and the defect density of the interface is not irradiated with Se. It was found to be about two orders of magnitude higher than in the case of hydrogen treatment, and it was found that the crystallinity was inferior even when compared with the above case.

[0027]

According to the present invention, the substrate surface is cleaned by irradiating it with excited hydrogen and heating it, and by irradiating the beam of the constituent elements, a good growth interface can be formed.

[Brief description of drawings]

FIG. 1 is a conceptual diagram of an MBE device used in an example.

FIG. 2 is a diagram showing a temperature program of the embodiment.

FIG. 3 is an RHEED photograph showing a C (2 × 2) streak pattern of a substrate after irradiation with excited hydrogen.

FIG. 4 is a RHEED photograph showing a (1 × 1) spotting pattern of a substrate not irradiated with excited hydrogen.

FIG. 5 is an RHEED photograph showing a (2 × 1) streak pattern before growth.

Claims (4)

[Claims] 1. An RF for a II-VI group compound semiconductor substrate
Illuminating with hydrogen excited by any one of discharge, microwave discharge, ECR discharge, or direct current discharge, and then irradiating with a beam of the element constituting the substrate or a compound containing the element II -Method for cleaning group VI compound semiconductor substrate. 2. The method for cleaning the surface of a II-VI group compound semiconductor substrate according to claim 1, wherein the irradiation start temperature of the excited hydrogen gas is adjusted to a temperature between room temperature and 700 ° C. 3. An irradiation start temperature for irradiating a beam of an element constituting the substrate or a compound containing the element is 100 to 7
II-V according to claim 1, wherein the temperature is adjusted to 00 ° C.
Method for cleaning group I compound semiconductor substrate. 4. The group II-VI compound semiconductor substrate is a binary compound of two or more elements in which the group II element is selected from Zn, Cd and Hg and the group VI element is selected from S, Se and Te. The method for cleaning a II-VI group compound semiconductor substrate according to any one of claims 1 to 3, wherein: