JPH05178687A - Method for growing semiconductor crystal - Google Patents

Abstract

PURPOSE:To provide the method for growing a semiconductor single crystal having the excellent controllability of the absolute film thickness. CONSTITUTION:A dummy GaAs substrate 2 having about 2mm size is used, is set at the center of a substrate holder 1 and is subjected to RHEED vibration measurement. As a result, the RHEED vibration measurement can be made in the same place at all times and, therefore, the influence of the local nonuniformity of a Ga beam 5 can be eliminated. As a result, the measurement of the growth speed with 0.3 to 0.6% accuracy is possible. The controllability of the film thickness is greatly improved as compared with the conventional beam flux measurement if the desired crystal growth is executed after this RHEED vibration measurement.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a semiconductor crystal growth method, and more particularly to a semiconductor crystal growth method using molecular beam epitaxy (MBE).

[0002]

2. Description of the Related Art The crystal growth method by MBE is widely used for the growth of crystals used in quantum well lasers, high-mobility transistors, etc., because it enables crystal growth of ultrathin films with good controllability.

[0003]

However, according to the conventional MBE method, although the film thickness can be precisely controlled in the crystal growth of one wafer, the absolute film thickness controllability in RUN to RUN is It wasn't always good. For example, in the example of crystal growth of a compound semiconductor of GaAs, there is a drawback that the absolute film thickness has a variation of about 2 to 3%.

Therefore, for example, when a semiconductor multilayer film (usually referred to as a DBR reflective film) is crystal-grown by the MBE method as a reflective film for a surface emitting laser, the absolute film thickness is 2 to 2.
When it fluctuates by about 3%, the central wavelength fluctuates by about 20 to 30 nm, which causes a problem that the controllability of the laser oscillation wavelength is poor. Therefore, considering the case of growing such a surface emitting laser, the controllability of the wavelength needs to be 10 nm or less. For this reason, the controllability of the absolute film thickness requires an accuracy of 1% or less.

An object of the present invention is to provide a semiconductor crystal growth method excellent in absolute film thickness controllability.

[0006]

In order to achieve the above object, a method for growing a semiconductor crystal according to the present invention comprises a step of measuring a growth rate and growing a semiconductor crystal using a molecular beam epitaxy apparatus. Therefore, in the process of measuring the growth rate, prior to the desired crystal growth, the growth is performed on the dummy semiconductor substrate fixed to the center of the substrate holder, and the change in the high-speed electron beam diffraction intensity at this time is measured to obtain the desired value. Is a step of correcting the growth rate during crystal growth, and the size of the dummy semiconductor substrate is within 3 mm.

[0007]

In the conventional crystal growth by MBE, the beam speed as a material is usually measured by an ion gauge before the desired growth to correct the growth rate. However, depending on the day, the measurement using the ion gauge may result in 2
~ 3% variation. For this reason, the accuracy of measuring the growth rate becomes a variation in the grown film thickness, resulting in a 2-3% variation in the film thickness.

The crystal growth method of the present invention uses high-speed backscattered electron diffraction (RHEED). That is,
When RHEED measurement is performed during crystal growth by MBE,
It is known that the diffraction intensity of the electron beam fluctuates periodically with time, which corresponds to the deposition rate of one atomic layer. Such changes in the diffraction intensity of RHEED
It is called D vibration. Therefore, the growth rate can be accurately known by measuring the RHEED oscillation in the initial stage of growth.

However, in the ordinary MBE growth, the beam intensity changes by about ± 20% within the film surface of the wafer. Therefore, the growth rate must be accurately measured due to the variation in the location where the RHEED vibration measurement is performed. It is difficult.

In the method of the present invention, RHEED vibration measurements are always made at the same location, so that prior to desired growth,
It is a method of growing on a dummy substrate and measuring the growth rate by RHEED vibration measurement.

Further, in our experiments, it was found that the size of the dummy substrate is an important factor. That is, it has been found that the smaller dummy substrate has a smaller variation in location and the measurement accuracy is improved. FIG. 2 shows the relationship between the size of the dummy substrate and the variation in measurement, which was measured by our device.

In this experiment, in the MBE apparatus, a dummy GaA fixed at the center of a molybdenum substrate holder was used.
GaAs was crystal-grown on the s substrate and RHEED vibration measurement was performed. RHEED repeated 3 times under the same conditions
The vertical axis shows the variation of the measured values when the vibration measurement is performed.

As a result of comparing the sizes of the dummy substrates between 3 mm and 2 mm, the average of the measurement variations is 0.
The variation was 6% and 0.2%, and the smaller dummy substrate was better. Therefore, when growing the DBR reflective film as described above, it is necessary to suppress the variation in RUN to RUN to 1% or less, but it is understood that a dummy wafer having a size of 3 mm or less is sufficient. Also, 2 mm
It is understood that the growth rate can be measured with an accuracy of 0.5% or less at the size of.

[0014]

Embodiments of the present invention will now be described in detail with reference to the drawings. FIG. 1 shows the RHEED of one embodiment of the present invention.
It is a schematic diagram of vibration measurement. In the figure, 1 is a substrate holder, 2
Is a GaAs dummy substrate, 3 is Ga, 4 is a Ga shutter, 5 is a Ga beam, 6 is As, 7 is an As shutter,
8 is an As beam. 9 is a heater, 10 is RHEED
An electron beam, 11 is a fluorescent screen, 12 is a lens, 13 is a photodiode, 14 is a recorder, and 15 is an MBE chamber.

The size of the dummy substrate 2 is 2 mm,
Since it is fixed to the center of the substrate holder 1, even if the substrate 2 is rotated, RHEED vibration measurement can be performed at the same place. In order to perform RHEED vibration measurement on this dummy GaAs substrate 2, first, the substrate holder 1 is set at the growth position (the position of the substrate holder 1 in FIG. 1) with the Ga shutter 4 and the As shutter 7 closed. ..

Next, as shown in FIG. 1A, the As shutter 7 is opened and the As beam 8 is directed to the dummy GaAs substrate 2.
To irradiate. In this state, the heater 9 is energized to gradually raise the temperature of the dummy GaAs substrate 2.

When the temperature is raised to around 630 ° C., the dummy GaA
The GaAs oxide film on the surface of the s substrate 2 can be evaporated. This is a process called As thermal cleaning that is often performed in MBE growth. After such thermal cleaning of As, as shown in FIG. 1B, the Ga shutter 4 is opened, and the Ga beam 5 is irradiated to the dummy GaAs substrate 2 to start the growth of GaAs on the dummy GaAs substrate 2. ..

Next, RHEED vibration measurement is started. That is, the RHEED electron beam 10 accelerated at about 10 to 20 kV is irradiated and the time change of the diffraction intensity is measured. The diffraction intensity can be recorded in the recorder 14 by converting the light of the fluorescent screen 11 for observing the diffraction pattern of the RHEED electron beam 10 into an electric signal by the photodiode 13.

In an ordinary MBE apparatus, the intensity of the Ga beam 5 is not stable for about 3 minutes immediately after opening the Ga shutter 4, so it is necessary to measure the cycle of RHEED vibration after the Ga shutter 4 is opened and a sufficient time has passed. There is.

Since the growth rate can be accurately corrected by the RHEED vibration measurement as described above, the dummy G
In the desired crystal growth immediately after the growth of the aAs substrate, the controllability of the absolute film thickness was significantly improved.

FIG. 3 shows the results of comparison between the case where the dummy GaAs substrate 2 is grown and the case where the conventional beam flux measurement is performed. FIG. 3 shows a surface emitting laser D
It is a plot of the set value of the center wavelength when the BR reflective film is MBE-grown and the measured value. In the conventional beam flux measurement, the variation of the measured value with respect to the linear relationship between the set value and the measured value is 23 nm at the maximum.
It can be seen that the RHEED vibration measurement can grow with a variation of maximum 6 nm.

This variation amount depends not only on the measurement accuracy of the growth rate but also on the temporal stability of the beam intensity.
Therefore, as a result of further improving the accuracy of the temperature regulator for further stabilizing the temperature, a value of 3 nm has recently been obtained as the maximum deviation amount. The variation for this set wavelength is
When converted into a film thickness, the absolute film thickness accuracy is 0.3%.

As described above, it was found that by using a small dummy substrate of about 2 to 3 mm, the growth rate can be accurately measured by the RHEED vibration measurement, so that the absolute film thickness accuracy is improved. From this result, it is expected that the measurement accuracy will be further improved by using a smaller dummy substrate. However, it has been found that a dummy substrate smaller than 2 mm (for example, about 1 mm) cannot obtain a sufficient electron beam diffraction intensity, so that the measurement accuracy is rather lowered.

In the beam system (~ 0.1 mm) of the electron beam source used in the current RHEED measuring device, it is considered that the optimum beam diameter is about 2 mm. Furthermore, it is expected that by using a high-intensity electron beam, sufficient RHEED vibration measurement can be performed even with a smaller dummy substrate, and the measurement accuracy will be further improved.

In the above embodiments, the growth rate of GaAs was measured, but the present invention is not limited to this, and the present invention is also applicable to MBE growth of AlGaAs, other semiconductors, InP, Si and the like.

[0026]

As described above, according to the present invention, since the growth rate can be measured with high accuracy, crystal growth of a semiconductor with good controllability of absolute film thickness can be realized.

[Brief description of drawings]

1A and 1B are schematic views of a semiconductor crystal growth method according to an embodiment of the present invention.

FIG. 2 is a diagram showing the relationship between the size of a dummy substrate and variations in measured growth rates when the present invention is applied.

FIG. 3 is a diagram showing experimental values of set wavelengths and measured wavelengths when a DBR reflective film is grown.

[Explanation of symbols]

 1 substrate holder 2 dummy GaAs substrate 3 Ga 4 Ga shutter 5 Ga beam 6 As 7 As shutter 8 As beam 9 heater 10 RHEED electron beam 11 fluorescent screen 12 lens 13 photodiode 14 recorder 15 MBE chamber

Claims (1)

[Claims] 1. A method of growing a semiconductor crystal by using a molecular beam epitaxy apparatus, which comprises a step of measuring a growth rate, wherein the step of measuring the growth rate comprises a step of forming a substrate holder prior to desired crystal growth. Is a step of performing growth on a dummy semiconductor substrate fixed at the center of the dummy semiconductor substrate, and measuring the change in high-speed electron beam diffraction intensity at this time to correct the growth rate at the time of desired crystal growth. Is within 3 mm, a semiconductor crystal growth method.