JPS6235635A - Measurement of film thickness distribution and measuring device thereof - Google Patents

Abstract

PURPOSE:To measure the in-plane distribution of a film thickness with high accuracy and in a short time by a method wherein a device for measuring the film thickness distribution in provided with a means which makes an electron beam incident in a crystal growth plane, a means which deflects the electron beam in the crystal growth plane, a means which receives a reflected and diffracted beam and displays a separated luminescent spot and a detecting means which converts light at the luminescent spot into an electrical signal. CONSTITUTION:A superhigh vacuum container 1 is provided with plural pieces of evaporation sources, a molecular beam is generated from each evaporation source, the molecular beams are simultaneously applied to a heated substrate 4 to form a single crystal thin film on the substrate through molecular beam epitaxial growth and an in-plane film thickness distribution is measured during the growth of the single crystal thin film. In that case, an electron beam is incided on the crystal growth plane of the single crystal thin film, and moreover, the incident electron beam is deflected to the prescribed direction in the crystal growth plane and a reflected electron beam diffraction image from the crystal growth plane is measured to measure the in-plane distribution of growth rate of the single crystal thin film during the growth of the film. By this way, a film thickness and the in-plane distribution of the film thickness can be found with high accuracy and in a short time.

Description

Detailed Description of the Invention [Industrial Application Field] The present invention provides a method in which a plurality of evaporation sources are arranged in an ultra-high vacuum container, molecular beams are generated from each of the evaporation sources, and these molecular beams are heated. The present invention relates to a method and apparatus for simultaneously growing a single crystal thin film on the substrate by molecular beam epitaxial growth and measuring the in-plane film thickness distribution during the growth of the single crystal thin film.

[Prior Art 1 Compound semiconductors such as GaAs are widely used as constituent materials for semiconductor lasers and high-speed semiconductor devices. One of the essential techniques for producing these elements is single-crystal thin film formation technique.

As the performance of devices increases, the requirements for thin film technology become stricter, and thin film technology exhibits rapid compositional changes within the range of a single atomic layer.
It is also required to be able to form an electrically and optically defect-free interface.

Molecular beam epitaxial growth (NBK) has been shown to meet these requirements and is now
It is used to form thin films of not only semiconductors of the ■-■ group, but also multi-component systems such as the IF-Vl group and ][-N-V group, and single-element semiconductors such as Si.

In principle, the NBK method is an advanced form of the vacuum evaporation method, in which multiple evaporation sources called molecular beams are placed in an ultra-high vacuum chamber, and the molecular beams generated from these evaporation sources are directed into the vacuum evaporation method. A thin film is deposited on the substrate by irradiating the substrate placed in the container.

For example, as shown in FIG. 1, the MBE apparatus is equipped with a vacuum pump 2 for evacuating a vacuum container 1, a plurality of molecular beam sources 3 are attached to the vacuum container 1, and A substrate holder 5 supporting the substrate 4 and having a built-in heater for heating the substrate 4 is arranged, and an electron gun 6 for irradiating the formed film with an electron beam in order to examine the crystallinity of the film. A reflected electron beam diffraction device is provided which includes a deflection electrode 7 and a fluorescent screen 8 that projects the diffraction pattern of the reflected electron beam from the deflection electrode 7.

Although the MBE device has the above-mentioned features, there was no effective means for observing the in-plane distribution of film thickness.
After film growth, the film thickness distribution was obtained only by separating the sample and etching its cross section. This method is a destructive method, and the accuracy is only a few percent, which is not very good.

Therefore, developing a means for determining film thickness distribution using a non-destructive method and with high precision is an important issue from both an industrial and academic perspective. Regarding the former, when manufacturing a large number of lasers and FETs from one wafer, if the film thickness distribution is known in advance, the yield of devices can be improved. In the latter case, when investigating the physical properties of a superlattice, since the physical properties of a superlattice strongly depend on the period of the superlattice, if
This is because if the film thickness is non-uniform within the plane, different experimental results will be obtained depending on where the wafer is taken out, making it difficult to confirm the physical properties correctly.

[Problems to be Solved by the Invention] Accordingly, one object of the present invention is to appropriately use a backscattered electron beam diffraction device installed in a IIIBF apparatus to obtain an in-plane measurement of the film thickness without destroying the sample during film growth. An object of the present invention is to provide a method for measuring film thickness distribution that can measure the distribution during film growth.

Another object of the present invention is to provide a film thickness distribution measuring device that can measure the in-plane film thickness distribution with high precision and in a short time without destroying the sample during film growth.

[Means for solving the problems] In order to achieve such an objective, the measurement method of the present invention has the following features:
Multiple evaporation sources are placed in an ultra-high vacuum container, each of which generates molecular beams, and these molecular beams are simultaneously irradiated onto a heated substrate to grow a single crystal thin film on the substrate using molecular beam epitaxial growth. To measure the in-plane film thickness distribution during the growth of a single crystal thin film, an electron beam is incident on the crystal growth surface of the single crystal thin film, and the incident electron beam is swayed in a predetermined direction on the crystal growth surface. , is characterized in that the in-plane distribution of the growth rate of a single crystal thin film is measured during film growth by measuring a reflected electron beam diffraction image from the crystal growth surface.

The measurement device of the present invention has a plurality of evaporation sources arranged in an ultra-high vacuum container, generates molecular beams from each of the evaporation sources, and simultaneously irradiates these molecular beams onto a heated substrate to detect Z on the substrate. In an apparatus that grows a single crystal thin film by molecular beam epitaxial growth and then determines the in-plane film thickness distribution during the growth of the single crystal thin film, a collimated beam is generated and a laser beam is applied to the crystal growth surface of the single crystal thin film. A means for making the electron beam incident, a means for swinging the electron beam continuously or stepwise on the crystal growth surface, and a means for receiving the reflected diffraction beam from the crystal growth surface to project continuous or plural bright spots. The present invention is characterized by comprising means having a phosphor screen for detecting bright spots, and a plurality of detecting means for receiving light from a bright spot and converting it into an electrical signal.

[Function] According to the present invention, firstly, differences in growth rates can be measured using a high-resolution method, and by observing about 100 vibrations, it is possible to identify differences in growth rates down to 0.1%. Second, if the measurement time is extremely short and one cycle is 1 second, about 10 seconds is sufficient.

Thirdly, by installing a plurality of detectors and using a composite wave of a staircase wave or a triangular wave and a pulse as the modulated wave, the in-plane distribution can be determined in one measurement.

[Example] The present invention will be described in detail below with reference to the drawings.

According to recent MBE research, when a backscattered electron diffraction image is observed during film growth, a phenomenon in which the light intensity of the diffraction image periodically fluctuates was found, and this phenomenon occurs during the process of film growth layer by layer. It was revealed that the vibration period corresponds to the growth rate of the film (Appl, Phys,
A, 19B3, 31.1).

Therefore, if the vibration period of a film can be observed during film growth, the growth rate can be determined during film growth. Film thickness is expressed as the product of growth rate and growth time, so
Knowing the growth rate during film growth is sufficient to know the film thickness.

However, it is not possible to determine the in-plane distribution of film thickness by this method, because the electron beam is focused narrowly to cause electron diffraction, so the growth rate determined by the above method is This is because it only represents the growth rate in a very small area (about 1*+n in diameter).

By the way, as shown in FIG. 1, the electron beam diffraction apparatus is provided with a deflection electrode 7 for deflecting a well-collimated electron beam, and a bias applied to the deflection electrode 7 causes a deflection to occur on the sample surface of the substrate 4. Adjust the position of the irradiated electron beam.

Therefore, in the present invention, the irradiation position of the electron beam on the sample surface is changed by appropriately changing the bias applied to the deflection electrode 7 over time. That is, in the present invention, a bias in which a DC bias from a DC bias source 9 and a triangular wave, rectangular wave, or step-like staircase current from a superimposed radio wave source 10 are superimposed is applied to the deflection electrode 7.

Here, first, an appropriate DC bias is applied from the DC bias source 9 to irradiate an electron beam to a desired position iA on the film. If the electron beam is selected and directed to another desired point B, the electron beam will reciprocate between A and B. If the superimposed signal is a rectangular wave, the electron beam has two points, A and B.
Move only the point back and forth.

At this time, it is necessary to make the frequency of the superimposed signal sufficiently larger than the frequency of the vibration of the light intensity of the diffraction image, and to make it oscillate at high speed on the crystal growth surface of the incident electron beam. Since the frequency of the light intensity oscillation of the diffraction image is about IHz when the frequency is 0.3 Hz/h, it is necessary to set the frequency of the superimposed signal to about 10 to 100 times that value, that is, 10 Hz or more. As the growth rate of the film increases, the frequency of the light intensity oscillation in the diffraction image also increases, so the frequency of the superimposed signal must also increase accordingly.
The practical film growth rate in H is generally 0.1 to
Since it is in the range of 10 IL■/h, even in the case of 10 IL■/h, the frequency of the superimposed signal should be at least about 1oOHz. However, if the frequency of this superimposed signal is higher than this, the light intensity oscillation of the diffraction image will occur. Measurements can be taken in sufficient response.

Now, with these settings, if two photodetectors are placed on the fluorescent screen 8 at positions where reflected images or diffraction images from points A and B appear, vibrations at both points can be detected simultaneously. The difference in the growth rate between the two points can be determined from the difference in the vibration period between the two measurable points.For the 0th order, fix the starting point A and repeat the same measurement by sequentially changing the amplitude of the rectangular signal to be modulated. By creating a calibration curve between the amplitude and the distance from the starting point A, the growth rate, in other words, the in-plane distribution of the film thickness can be obtained.

FIG. 1 shows an embodiment of the measuring device of the present invention for implementing the measuring method of the present invention described above.

Here, the vacuum container l is heated to 10τ art using pump 2.
After evacuation to an ultra-high vacuum as described below, Ga and As beams are irradiated from the molecular beam source 3 onto the GaAs substrate heated to about 830"O, thereby depositing GaAsWJr on the substrate 4.
was formed. During film growth, when an electron beam accelerated to about 10 keV is applied to the film surface from a direction almost perpendicular to the molecular beam at an angle of 2 to 3 degrees, an electron beam diffraction pattern is projected onto the screen 8. It will be done. This pattern is photographed by a television camera 11 and displayed on a cathode ray tube 12. The light intensity of the projected pattern was detected using two photodetectors, such as solar cells 13.

In this example, two pairs of opposing coils were used as the deflection electrode 7, and the electron beam was deflected by a magnetic field generated by applying a bias to these coils. Here, a bias obtained by superimposing a modulation signal current from a superimposed current source 10 on a DC current from a DC bias source 8 was supplied to the coil 7.

2(A), (B) to FIG. 5(A), (B) show the correlation between the waveform of this superimposed current and the shape of the spot on the phosphor screen 8.

That is, when the superimposed current is a triangular wave as shown in FIG. 2(A), the spot on the phosphor screen 8 traces the locus of a bar-shaped bright line.

When the superimposed current is a rectangular wave as shown in FIG. 3 (Todoroki), the spots on the phosphor screen 8 become two bright spots corresponding to two levels of the rectangular wave.

When the superimposed current is a staircase wave as shown in FIG. 4(A), the spots on the phosphor screen 8 become a plurality of bright spots corresponding to each step of the staircase wave.

When the superimposed current is a composite wave of a triangular wave and a pulse as shown in FIG. 5(A), the spots on the phosphor screen 8 become a plurality of bright spots corresponding to the level of each pulse of the composite wave.

In order to detect one of these points, in addition to directly detecting the bright spot on the phosphor screen 8 with a photodetector,
A similar spot shape can be obtained by using a cathode ray tube 12.
It can also be formed on top. On the screen surface of such a cathode ray tube 12, solar cells 13 are arranged at a plurality of points in the part where the spots are generated, signals corresponding to the spots are simultaneously extracted from each spot, and the outputs are recorded on the recorder 14. I can do it.

Next, we will discuss experimental results when a rectangular wave is used as the modulation signal current. Figure 3 shows the state of vibration ■ and ■ at two different points when the frequency of the rectangular wave is 100 Hz. The 0 period is approximately 1.1 seconds. The periods of vibration and I at both points are as follows. There was a slight difference, with a difference of 0.2 seconds in the time required for the same cycle of lO. When this value is converted into a relative ratio of growth rate, it corresponds to about 2%.

Figure 4 shows the results of determining the rate of change in growth rate by fixing one point and changing the amplitude of the rectangular wave. was less than 1%, but it can be seen that the rate of change monotonically increases as the distance from the wafer center increases. The in-plane distribution of the film thickness obtained by cutting this sample after S was prepared and using the stain-etch method almost coincided with the in-plane distribution obtained from FIG. 4.

[Effects of the Invention] As explained above, according to the present invention, the film thickness and the in-plane distribution of the film thickness can be determined with high accuracy and in a short time during film growth in an MBE apparatus.

That is, according to the present invention, firstly, the difference in growth rate can be measured with high resolution, and if approximately 100 vibrations are observed, the difference in growth rate can be measured with high resolution.
．． Differences in growth rate of up to 1% can be identified. Secondly, the measurement time is extremely short; if the period is 1 second, about 10 seconds is sufficient. Thirdly, by installing a plurality of detectors and using a composite wave of a staircase wave or a triangular wave and a pulse as the modulated wave, the in-plane distribution can be determined in one measurement.

Note that the present invention is extremely effective for use in measuring film thickness distribution not only in MBE apparatuses that require ultra-high vacuum, but also in ordinary vacuum evaporation apparatuses.

[Brief explanation of the drawing]

Figure 1 is MB! Schematic cross-sectional views showing one embodiment of the measuring device of the present invention incorporated into a device, FIGS. 2(A) and (B), FIGS. 3(A) and (B), and FIG. 4(A) and ( B) and FIGS. 5A and 5C are explanatory diagrams schematically showing the correlation between the superimposed current that modulates the electron beam and the shape of the diffraction image appearing on the fluorescent screen. Figure 6 is a characteristic diagram of experimental results showing how the light intensity oscillates in the diffraction image obtained when diffraction images corresponding to two different points on the film surface are observed simultaneously, and Figure 7 is a characteristic diagram of the experimental results using the present invention. FIG. 3 is a characteristic diagram showing the in-plane distribution of the rate of change in growth rate obtained by the method of FIG. DESCRIPTION OF SYMBOLS 1... Vacuum container, 2... Vacuum pump, 3... Molecular beam source, 4... Substrate, 5... Substrate holder with built-in heater, 6... Electron gun, 7... Deflection electrode, 8... Fluorescent screen, 9... DC bias source, lO... Superimposed current source, 11... Television camera, 12... Cathode ray tube, 13... Solar cell, 14...・Recorder. Patent applicant: Agent of Nippon Telegraph and Telephone Corporation
Patent Attorney Yoshi Tani - Figure 5 Figure 6 o 10 S giant Xi (nm) of Jm o゛5 in wafer Figure 7

Claims (1)

[Claims] 1) A plurality of evaporation sources are arranged in an ultra-high vacuum container, each of the evaporation sources generates a molecular beam, and a heated substrate is irradiated with these molecular beams simultaneously to produce the substrate. To grow a single-crystal thin film by molecular beam epitaxial growth on the single-crystal thin film and measure the in-plane film thickness distribution during the growth of the single-crystal thin film, an electron beam is incident on the crystal growth surface of the single-crystal thin film, and the incident electrons are The method is characterized in that the in-plane distribution of the growth rate of the single crystal thin film is measured during film growth by swinging a beam in a predetermined direction on the crystal growth surface and measuring a reflected electron beam diffraction image from the crystal growth surface. Film thickness distribution measurement method. 2) Multiple evaporation sources are placed in an ultra-high vacuum container, each of which generates molecular beams, and a heated substrate is irradiated with these molecular beams simultaneously to form a single crystal thin film on the substrate. In an apparatus for performing molecular beam epitaxial growth and measuring in-plane film thickness distribution during the growth of the single crystal thin film, there is provided a means for generating a collimated electron beam and making it incident on the crystal growth surface of the single crystal thin film; beam,
means for causing the crystal growth surface to fluctuate continuously or stepwise; means having a fluorescent screen that receives the reflected diffraction beam from the crystal growth surface and projects continuously or divided into a plurality of dots; A film thickness distribution measuring device characterized by comprising a plurality of detection means for receiving the light at the point (1) and converting it into an electrical signal.