JP2834632B2 - Method for determining growth rate and method for forming single crystal thin film - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for determining a growth rate of a single crystal thin film formed by epitaxial growth and a method for forming a single crystal thin film.

[0002]

2. Description of the Related Art Conventionally, an example in which a single crystal thin film formed by epitaxial growth is applied to a device is described in reference I.
(Compound semiconductor device [I], Industrial Research Council, 1984
Year, P. There is a single crystal thin film having a multilayer structure disclosed in 142).

Referring to FIG. 4, the structure of a single crystal thin film layer used in a GaAs / AlGaAs high electron mobility transistor (hereinafter referred to as HEMT) disclosed in Document I will be briefly described. explain.

A conventional layer structure is a semi-insulating GaAs substrate 1
A high-purity GaAs layer 12 having a thickness W ₃ of 0.8 μm and formed by molecular beam epitaxy (MBE) is provided on the substrate. Further, a high-purity Al _1-x Ga _x A having a film thickness W ₂ of 60 A ° (the symbol of A ° represents Angstroms) formed on the high-purity GaAs layer 12 by the MBE method.
s layer 14 and Si doped Al having a thickness of 0.1 μm W _{1
A 1-x} Ga _x As layer 16 is provided.

Further, a high-purity GaAs layer 12 and an Al ₁ -XG
The interface between a _X As layer 14, the area of the two-dimensional electron gas 18 electrons gather is formed, this region has little impurity ions to lower the carrier mobility because of high purity, therefore, the electrons high mobility Have a degree.

It is necessary to know the growth rate in order to control the thickness of the above-mentioned single crystal thin film. As a method for determining this growth rate, reference II (Reference II: “Japan”
seJournal of Applied Physs
ics ", Tadao Isibasi et al.
Vol. 20, NO. 9, PP. L623 to L626,
1981).

[0007] A method for determining the growth rate disclosed in Document II will be described with reference to FIG.

FIG. 5 shows an X-ray diffraction pattern obtained by forming a superlattice structure composed of an AlAs layer and a GaAs layer on the (100) plane of a GaAs substrate by molecular beam epitaxy and measuring the X-ray diffraction of the (400) plane. 3 shows a rocking curve. In the figure, the horizontal axis represents the angle Δθ (second), and the vertical axis represents the intensity.

The diffraction peak Pa is (40) of the GaAs substrate.
0) represents the diffraction peak of the plane. Also, the diffraction peak P
b, Pc, and Pd represent the 0th, -1st, and + 1st order satellite peaks, respectively. The peaks having the measured peaks are represented by A, B, C, and D, and the respective diffraction peaks are represented by Pa, Pb, Pc, and Pd.

The angle difference between the diffraction peak Pa and the diffraction peak Pb is defined as Δθ _{0, s,} and from this angle difference Δθ _{0, s} , the average value of the composition of Al (aluminum) constituting a one-period superlattice is determined. be able to.

Next, the angle difference between the diffraction peak Pb and the diffraction peak Pd is defined as Δθ _{0, + 1,} and the average film thickness of one period of the superlattice layer is obtained from the angle difference Δθ _{0, + 1} . Therefore, the respective film thicknesses of the GaAs layer and the AlAs layer during one period of the superlattice structure can be determined. In the document II, the thicknesses of the GaAs layer and the AlAs layer are set to 29 A and 31 A, respectively. One-period average GaAs layer and AlA of superlattice
By dividing the thickness of the s layer by the respective growth time, G
The growth rates of aAs and AlAs are required.

[0012]

However, in the conventional MBE method, it is known that beam overshoot occurs when a shutter of a K cell (also called a Knudsen cell) is opened. If crystals are grown by opening and closing the shutters of the Ga cell and the Al cell regularly and repeatedly during the formation of the AlAs / GaAs superlattice, the thickness of the superlattice layer becomes large due to the effect of overshoot. Therefore, Ga
Since the growth rate of As is greatly estimated, Ga
There is a problem that the growth rate of As cannot be accurately determined.

Further, conventional AlAs formed by MBE method
In the / GaAs superlattice, the lattice mismatch ratio of AlAs to GaAs is as small as about 0.13%. Therefore, the Al composition (or Ga composition) cannot be determined accurately. Therefore, the growth rate of each film thickness could not be obtained with high accuracy. Hereinafter, the reason will be described with reference to Reference III (Reference III: “Philips Jo
urnal of Research "Vol. 4
1, No. 3, 1986, p. 271).

The 0th-order satellite peak Pb and GaAs in the X-ray diffraction rocking curve of the (400) plane in FIG.
The angle difference Δθ _{0, s} between the diffraction peak Pa of the substrate and the superlattice 1
The value x of the average Al composition during the cycle is given by the following equation.

X = {a _GaAs / (a _AlAs− a _GaAs )} × (1−ν / 1 + ν) × (−cot θ) × Δθ _{0, s} (1) where a _GaAs is a lattice constant of GaAs and a _AlAs is AlA
The lattice constant of s, ν is Poisson's ratio, and θ is the black angle.

However, since the lattice mismatch ratio of AlAs to GaAs is small (lattice mismatch ratio is about 0.13%), {a _GaAs / (a _AlAs -a _GaAs )} in equation (1) becomes large. The reason is that the reciprocal of the lattice mismatch ratio with respect to the lattice mismatch ratio of AlAs with _GaAs is ｛a _GaAs / (a
_AlAs- a _GaAs )}. At the same time _, the error of the value x of the Al composition with respect to the error of Δθ _{0, s} also increases. For this reason, there has been a problem that the growth rate of GaAs cannot be determined accurately.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned problems, and accordingly, an object of the present invention is to provide a method for determining a growth rate of a single-crystal thin film that can accurately determine the growth rate of the single-crystal thin film. Is to provide. Also,
An object of the present invention is to provide a method for forming a single crystal thin film in which a single crystal thin film having a desired thickness is grown according to a determined growth rate.

[0018]

In order to achieve this object, according to the method for forming a single crystal thin film of the present invention, a first material and a lattice having a large lattice mismatch with respect to the base are formed on the base. A superlattice structure composed of a single-crystal thin film is formed by periodically epitaxially growing a second substance having a small mismatch ratio or the same as the base material, and measuring the X-ray diffraction of the superlattice structure. It is characterized in that the growth rates of the first and second substances are determined.

In practicing the present invention, the first material is preferably made of InGaAs or InAs.
And the second material is preferably GaAs.

In practicing the present invention, the epitaxial growth is preferably performed by MBE.

In practicing the present invention, preferably, when the superlattice structure is formed on the underlayer by using the MBE method, the shutter of the second material K cell is kept open, It is good to repeatedly open and close the shutter of the In K cell included in the above.

In practicing the present invention, preferably, a first material having a large lattice mismatch and a second material having a small lattice mismatch or the same as the base are periodically formed on the base. A superlattice structure composed of a single crystal thin film is formed by epitaxial growth, and the growth rates of the first and second substances are determined by measuring the X-ray diffraction of the superlattice structure. Preferably, a second material is deposited on the lower surface at the determined growth rate of the second material to form a single crystal thin film having the second material.

[0023]

According to the above-described method for forming a single crystal thin film of the present invention, the first material having a large lattice mismatch ratio and the second material having a small lattice mismatch ratio or the same as the underlayer are formed on the underlayer. A substance is periodically epitaxially grown to form a superlattice structure composed of a single crystal thin film. An X-ray diffraction rocking curve is obtained by measuring the X-ray diffraction of the thin film having the superlattice structure, and the growth rates of the first substance and the second substance are determined from the data of the obtained X-ray rocking curve. At this time, the first material is InGaAs with respect to the base.
Alternatively, InAs is used, and the second material is GaAs. When a single-crystal thin film is formed on a base, epitaxial growth is performed by MBE.

The value of the lattice mismatch ratio for GaAs in the above equation (1) can be changed depending on the material constituting the superlattice layer in this manner. For example, the lattice mismatch ratio of GaAs to InAs
Therefore, it is about 1/52 of the reciprocal of the value of the lattice mismatch ratio for GaAs in (1).
The values of the respective lattice constants are those disclosed in Document IV (see "IV-V Group Semiconductor Mixed Crystal", Corona, 1988, pp. 39).

Next, InGaAs is used as the first substance,
The reason why the growth rate of GaAs can be obtained with high accuracy when GaAs is used as the substance will be described.

In the present invention, the conventional A
A superlattice structure is formed on a GaAs substrate using InAs having a lattice mismatch ratio of about 52 times larger than lAs,
The growth rate of GaAs was determined by measuring the X-ray diffraction of an As / GaAs superlattice. here,
Conventional AlAs / GaAs superlattice and InGaAs / Ga
Assuming that the Poisson ratio of the As superlattice is equal, Δθ of both
_In order to make _{0 and s} equal, from equation (1) described above, I
The average In composition y in one cycle of the nGaAs / GaAs superlattice is about 1/52 of the average Al composition x in one cycle of the conventional AlAs / GaAs superlattice. Therefore, AlAs /
The composition of Ga in the GaAs superlattice is 1-x (where x is Al
The composition of ), And the average Ga composition in one period of the InGaAs / GaAs superlattice is 1-y = (1-x) / 5
2 (where y is the composition of In).

In addition, AlAs / GaAs superlattice and InG
Assuming that the average GaAs film thickness in one cycle with the aAs / GaAs superlattice is equal, the average GaAs film thickness in one cycle is {(1-x) and {(1-y) =}, respectively. (1
−x) / 52 °. Therefore, considering the composition error due to the error of Δθ _{0, s} as the error rate of the average film thickness of GaAs in one cycle, the AlAs / GaAs superlattice is −Δx /
(1-x), the InGaAs / GaAs superlattice is −Δy /
(1−y) = − Δx / (52−x). Therefore, I
The nGaAs / GaAs superlattice is more compatible with the conventional AlAs /
The average film thickness of GaAs during one period can be obtained with (52-x) / (1-x) times higher accuracy than the GaAs superlattice. For example, if X ≒ 0.52 is substituted, an accuracy of about 107 times the growth rate can be expected.

Therefore, by forming a superlattice structure composed of a single crystal thin film as in the embodiment of the present invention,
The thickness of a material (for example, GaAs) can be accurately determined. Therefore, the growth rate of GaAs can be determined with high accuracy.

Further, the second material is grown on a new base using the growth rate determined by X-ray diffraction.
A single crystal thin film containing a substance is formed. In this manner, the thickness of the single crystal thin film can be controlled accurately.

When forming a superlattice structure, molecules are emitted while the K-cell shutter of the second substance is kept open, and the K-cell shutter of In contained in the first substance is repeatedly opened and closed. In this manner, a superlattice made of a single crystal thin film is formed with the K-cell shutter of the second material kept open until the growth of the superlattice made of the first material and the second material is completed. For this reason, the influence of beam overshoot can be eliminated.

[0031]

Embodiments of the present invention will be described below with reference to the drawings. It should be noted that the drawings used in the description of this embodiment merely schematically show the shapes, sizes, and arrangements of the components so that the present invention can be understood.
Further, each diffraction peak value of the X-ray diffraction is obtained by merely plotting a representative peak, and many other peak values actually appear.

First, a method of forming a superlattice structure used for X-ray diffraction measurement according to the present invention will be described with reference to FIGS.

FIG. 1 is a cross-sectional view of a main part in which a first material and a second material are periodically epitaxially grown to form a superlattice structure.

As the substrate 20, a (100) plane GaAs substrate is used.

Next, on the upper surface of the substrate 20, for example, MBE
The GaAs buffer layer 22 is formed by using (molecular beam epitaxy). Here, the GaAs substrate 20 and G
The aAs buffer layer 22 is generally referred to as an underlayer 23. A superlattice consisting of 25 periods of the first material layer 24 and the second material layer 26 is grown on the underlayer 23 by using, for example, the MBE method under any suitable conditions. In the embodiment of the present invention, the growth conditions were set as follows.

The growth temperature is set to, for example, 450 ° C., and the ratio of group V element / group III element is made sufficiently large to maintain the stabilized state of arsenic (As).

The first material layer 24 is formed of a material having a large lattice mismatch with respect to the underlayer, for example, an InGaAs layer. Hereinafter, the first material is referred to as InGaAs. At this time, the growth rate of the InGaAs layer 24 is set to, for example, 0.5 μm.
/ H (micrometer / hour).

On the other hand, the second material layer 26 is formed of the same material as the underlayer 23, such as a GaAs layer, having a small lattice mismatch with respect to the underlayer. The growth rate at this time is, for example, about 0.4 μm / h. Hereinafter, the second material is GaAs.
Called.

Next, with reference to FIG. 2, the opening / closing control of a K-cell (K-cell is an abbreviation of a nu-cent cell) shutter when forming a superlattice by the MBE method will be described.

Each K cell used in the MBE apparatus (not shown) has a second material such as arsenic (As) or gallium (Ga), and a first material such as indium (In) or gallium. (Ga) source is supplied. The shutters attached to the K cell are referred to as an As cell shutter, a Ga cell shutter, and an In cell shutter, respectively.

First, in a region I where the GaAs buffer layer 22 is formed, the As cell shutter and the Ga cell shutter are opened, and the In cell shutter is closed.

Next, in the time domain II for forming the InGaAs / GaAs superlattice, the As cell shutter and the G
Open the a-cell shutter and set the In-cell shutter to the period T ₁
And repeatedly opened and closed in a period T ₂ of the cycle. In this embodiment, for example, the period T ₁ 10 seconds, and the period T ₂ to open and close the shutter 120 seconds cycle.

Therefore, unlike the conventional case, the Ga cell shutter is not opened and closed, so that the influence of the Ga beam overshoot can be eliminated.

Next, X-ray diffraction was measured using the structure formed in FIG. 1, and a rocking curve was drawn in FIG. In the figure, the horizontal axis represents the angle (second) and the vertical axis represents the intensity.

This X-ray diffractometer was measured using a pentacrystal X-ray diffractometer (manufactured by Philips, Model PW1754). The first to fourth crystals use the (220) plane of germanium (Ge), and the fifth crystal uses InGaAs /
A GaAs superlattice is used. The X-ray source is CuKα ₁
Is used.

FIG. 3 shows the diffraction peak A on the (400) plane of the GaAs substrate, the 0th-order satellite peak B of the InGaAs / GaAs superlattice, the -3rd-order satellite peak C, and-.
A secondary satellite peak D, a −1st order satellite peak E, a + 1st order satellite peak F and a + 2nd order satellite peak G are observed.

Diffraction peak A and zero-order satellite peak B
And the angle difference Δθ _{0, s} is −245.8 seconds. From this angle difference Δθ _{0, s} , the average In composition y in one period of the superlattice is obtained. At this time, the In composition y is 0.013.

Next, when the angle difference Δθ _{-1, + 1} between the satellite peaks E and F is measured, it is 1935.1 seconds. The average film thickness Λ of one period of the superlattice is determined from the angle difference Δθ _{-1, + 1} . That is, the average film thickness of one period of the superlattice Λ = 19
5.7 A °.

From this, in the InGaAs / GaAs superlattice, the average In composition y in one cycle is obtained, and the average film thickness の of one cycle is obtained.
Of the film thickness d _G of InAs and the film thickness d _I of InAs are obtained. That is, d _G = 192.8 A ° and d _I = 2.7 A °. Therefore, the GaAs film thickness d _G during one cycle is set to
While growing the cycle, the growth rate of GaAs can be calculated by dividing by the period T ₁ + T ₂ (130 seconds) during which the Ga cell shutter is open.

The calculated GaAs growth rate is 192.8 / 0.0361 = 5339.1 A ° / hour = 0.534 μm / hour.

Next, the InGaAs / PGA of the embodiment of the present invention will be described.
Composition error of GaAs superlattice and conventional literature II mentioned above
And the composition error of the AlAs / GaAs superlattice disclosed in US Pat.

[0052] a is calculated the composition x of a conventional _{Al x = L B / (L} B + L Z) = 31A ° / (31A ° + 29A °) ≒ 0.52. However, L _B is of AlAs thickness, L _Z is GaA
s represents the film thickness.

Since Δθ _{0, s} = −192 seconds, substituting the values of x and Δθ _{0, s} into the above equation (1) gives {a _GaAs / (a _AlAs −a _GaAs )} × {(1- ν) / (1 + ν)｝ × (−cot θ) = − 2.7 × 10 ⁻³ / sec (2)

Assuming that the error of θ _{0, s} is ± 10 (seconds), θ _{0, s} = −192 ± 10 (seconds). Therefore,
From the expressions (1) and (2), x ≒ 0.52 ± 0.027.

On the other hand, the value of the Ga composition (1-x) is 1-x ≒ 0.48 ± 0.027.

Here, when the error of the conventional Ga composition is calculated, ± 0.027 / 0.48 = ± 0.056 (± 5.6%
).

However, in the embodiment of the present invention, since the reciprocal of the lattice mismatch ratio of InAs to GaAs in the equation (1) is about 1/52 as compared with the conventional case, it is assumed that θ _{0, s}
= −192 ± 10 (seconds) are the same, the composition y of In of the first substance is y = (− 2.7 × 10 ⁻³ /52)×(−192±10)=0.00996±0 .00052.

On the other hand, the value of the composition (1-y) of Ga is as follows: 1-y = 0.9900 ± 0.00052.

Therefore, when calculating the error of the composition (1-y) of Ga, ± 0.00052 / 0.9900 = ± 0.0005 (± 0.0005)
0.05%).

The error of the Ga composition of the embodiment of the present invention calculated as described above is the G error of the conventional AlAs / GaAs lattice.
a It can be seen that the accuracy is about 100 times or more higher than the composition error. Therefore, the accuracy of the growth rate obtained from the thickness of the superlattice can be improved. For this reason, when, for example, a GaAs layer is deposited on a new underlayer, the GaAs growth rate determined from the embodiment of the present invention is used for the growth of the GaAs layer.
When a single crystal thin film having an aAs layer is formed, the film thickness can be controlled very accurately.

In this embodiment, the first material is InGaAs.
Although a mixed crystal was used, for example, InAs, In
A compound selected from the group of Sb, AlSb, GaSb and the like may be used.

[0062]

As is apparent from the above description, according to the method for forming a single crystal thin film of the present invention, the first material having a large lattice mismatch with respect to the base and the lattice mismatch are formed on the base. A superlattice structure composed of a single-crystal thin film is formed by epitaxially growing a small or the same second material as the underlayer. The superlattice thus formed can have a lattice mismatch ratio about 52 times as large as the conventional one.

The X-ray diffraction of the superlattice structure is measured to determine the film thickness of each of the first and second substances from the X-ray diffraction rocking curve. Is found. The film thickness can be easily controlled by forming a GaAs single crystal thin film using the growth rate thus determined.

Further, the first material is made of InGaAs with respect to the underlayer.
s or InAs, and the second material is preferably GaAs. By using such a substance, the lattice mismatch rate can be increased.

The epitaxial growth is performed by the MBE method. When a superlattice structure is formed on an underlayer by using the MBE method, the K-cell shutter of Ga and As contained in the second material is kept open, and the K-cell of In contained in the first material is left open. Open and close the shutter repeatedly. As described above, since the super lattice structure is formed with the Ga K cell shutter kept open, there is no danger of overshoot. Therefore, since the thickness of the superlattice does not become unnecessarily thick, it can be expected that the growth rate can be improved.

Thereafter, a single crystal thin film having the second substance is formed on a new base using the growth rate of the second substance determined by X-ray diffraction. Therefore, the thickness of the single crystal thin film can be controlled more accurately than before, and a desired thickness can be obtained.

[Brief description of the drawings]

FIG. 1 is a cross-sectional view for explaining a method of forming a superlattice structure according to the present invention.

FIG. 2 is a diagram for explaining opening / closing control of a K cell shutter by an MBE method.

FIG. 3 is an X-ray diffraction rocking curve diagram of the superlattice structure of the present invention.

FIG. 4 is a diagram of a conventional semiconductor device having a multilayer structure for realizing high electron mobility.

FIG. 5 is an X-ray diffraction rocking curve diagram of a conventional superlattice structure.

[Explanation of symbols]

 20: GaAs substrate 22: GaAs buffer layer 23: underlayer 24: first material (InGaAs layer) 26: second material (GaAs layer) 28: InGaAs / GaAs superlattice

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁶ identification code FI H01L 29/812 (58) Field surveyed (Int.Cl. ⁶ , DB name) C30B 1/00-35/00 C23C 14/06 H01L 21/203 H01L 21/338 H01L 29/778 H01L 29/812

Claims (5)

(57) [Claims] 1. A single crystal of a single crystal, wherein a first material having a large lattice mismatch and a second material having a small lattice mismatch or the same as the underlayer are periodically epitaxially grown on the underlayer. Forming a superlattice structure comprising a thin film, and measuring the X-ray diffraction of the superlattice structure to obtain the first
And a growth rate of the second substance. 2. The method according to claim 1, further comprising:
The substance is InGaAs or InAs, and the second substance is G
aAs determination method. 3. A method for determining a growth rate, wherein the epitaxial growth according to claim 1 is performed by an MBE method. 4. When the superlattice structure is formed on an underlayer using the MBE method according to claim 3, the second material has a K
A method for determining a growth rate, wherein the shutter of the indium (In) K cell contained in the first material is repeatedly opened and closed while the shutter of the cell is kept open. 5. A single crystal of a single crystal, wherein a first material having a large lattice mismatch ratio and a second material having a small lattice mismatch ratio or the same as the underlayer are periodically epitaxially grown on the underlayer. Forming a superlattice structure comprising a thin film, and measuring the X-ray diffraction of the superlattice structure to obtain the first
And determining the growth rate of the second material, and then depositing the second material on a new substrate at the determined growth rate of the second material to form a single crystal thin film having the second material. A method of forming a single crystal thin film.