JPS6350012A - Substrate holder mechanism for molecular epitaxy equipment - Google Patents

Abstract

PURPOSE:To uniformize temperature distribution over the surface of a semiconductor substrate without using in, by providing a radiation intercepting ring for pressing the peripheral edge of the substrate received in a substrate receiving seat, the ring having an inner diameter substantially equal to that of the substrate receiving seat. CONSTITUTION:A ring-shaped substrate holder 3 having an L-shaped cross section supports a semiconductor substrate 1 at the peripheral edge of the bottom face thereof. Since the holder 3 has a wide aperture, the substrate 1 is heated directly by radioactive heat from a heater 4. A radiation intercepting ring 5 is provided to press the peripheral edge of the opposite side of the substrate 1. The radiation intercepting ring 5 has an inner diameter substantially equal to that of the substrate receiving seat 11 of the holder 3. In this manner, the peripheral edge of the substrate is shielded from the heat radiation by the interception of the receiving seat 11 and, therefore, no heat is lost from the peripheral edge of the substrate. Thus, temperature can be distributed substantially uniformly over the whole surface of the substrate 1.

Description

DETAILED DESCRIPTION OF THE INVENTION (7') Technical Field The present invention relates to a substrate holder mechanism for a molecular beam epitaxy apparatus.

Molecular beam epitaxy Molecular Beam E
Pitaxy is one of the techniques for growing an epitaxial layer on a semiconductor substrate.

An apparatus is used that combines a molecular beam crystal growth chamber with a subsequent sample preparation room or analysis room. The substrate (wafer) is transported within the apparatus while attached to the holder. Further, in this state, epitaxial growth is performed.

The molecular beam crystal growth chamber is maintained at an ultra-high vacuum of approximately 10" Torr. The substrate fixed to the holder is attached to a manipulator in the center of the molecular beam crystal growth chamber. A heater installed here allows The substrate is heated from the back side.

A plurality of molecular beam source cells are provided diagonally below the molecular beam crystal growth chamber. The molecular beam source cell includes a crucible, a heater, a heat shield, a thermocouple, etc. Raw materials are placed in a crucible to form a melt, which is then evaporated. Because it is in an ultra-high vacuum, the mean free path is long, and it becomes a molecular beam and travels in a straight line. This hits the surface of the substrate and forms an epitaxially grown layer.

The present invention relates to an improvement in the mechanism of a substrate holder that holds a substrate in such an MBE apparatus.

(A) Prior Art (I) FIGS. 3 and 4 are a plan view and a vertical sectional view showing the structure of a conventional substrate holder.

A substrate holder 13 having a flat plate portion and a cylindrical portion following the flat plate portion.
Second, the substrate 1 is attached using a low melting point metal 6.

The substrate holder 13 is fixed to the manipulator by the structure of the cylindrical portion, but this is not shown.

In the molecular beam crystal growth chamber, when the substrate is turned on by a manipulator, the result is as shown in FIG. 4, and there is a heater 4 on the back side to heat the substrate.

The heating temperature varies depending on the material of the substrate and the type of molecular beam source. If the substrate is GaAs, for example, 600″C~70
Heat to 0°C.

The low melting point metal 6 is, for example, In (156° C.).

The substrate 1 can be attached to the entire surface of the flat plate portion of the holder 13.

Most conventional MBEs have used this type of holder structure. Since no mechanical impact is applied to the substrate, it is suitable for holding thin and fragile substrates. Moreover, since the heat of the heater once heats the substrate holder 3 and this heat is transmitted to the substrate, the substrate can be heated uniformly.

Furthermore, since the entire surface is exposed and the molecular beam hits the entire surface, a molecular beam epitaxial growth layer can be formed over the entire surface of the substrate.

However, this structure had the following problems.

First, the substrate must be uniformly attached to the holder using In, but this is difficult. In melts and becomes fluid, and just by placing the substrate on the holder, the In spreads out, but sometimes a cavity may remain. The thickness may not be uniform.

This causes disturbance in the temperature distribution on the substrate.

In other words, uneven heating occurs.

Furthermore, after the epitaxial growth is finished, the substrate holder must be taken out from the MBE apparatus and the substrate must be peeled off from the substrate holder.

The substrate holder is heated to melt the In. Push the board parallel to the holder surface. The board then slides on the holder surface. The board comes off from the edge of the holder. In this way, the substrate can be peeled off.

Next, In adhering to the back surface is removed to clean the back surface.

As described above, the holder structure using In has the drawback that processes such as adhesion and peeling are troublesome.

Another shortcoming has become apparent.

It has been found that when molecular beam crystal growth is performed using a holder structure using In, CO is generated as an impurity. This was discovered through QMAS analysis.

CO is contained in In as an impurity, and it is thought that when it is heated under an ultra-high vacuum, it comes out from inside the In.

Although the causal relationship is not clear, it appears that In is responsible for the generation of CO.

When CO is generated, it attaches to the epitaxial surface,
It causes the crystal structure to be disturbed.

Because of these difficulties, there is a demand for a holder structure that does not use In.

(1) Prior Art) FIGS. 5 and 6 are a plan view and a vertical sectional view showing a conventional holder structure that does not use In.

This holder 3 is not flat but ring-shaped. It has a large opening in the center. The ring-shaped holder 3 has an L-shaped cross section, and the substrate 1 is placed on the L-shaped step.
A stopper 2 holds down the surface of the substrate 1.

The heater 4 directly heats the back surface of the substrate 1 using radiant heat.

In this way, the substrate 1 is held by the holder 3 and the stopper 2 at the peripheral edge having a width of about several yttm.

The substrate 1 may be GaAs or Si. In any case, it passes infrared rays well and absorbs little, so
Difficult to heat up due to radiant heat.

Therefore, the back side of the substrate is sometimes coated with TiN fz and a substance with high infrared absorption.

This kind of structure does not use In, so it is easy to attach, peel,
There are no problems with CO emissions.

However, it has become clear that the temperature distribution within the substrate surface is non-uniform.

The temperature is lower at the peripheral edge of the substrate with a width of about 1 cm.

The temperature at the periphery is about 10°C compared to the center of the board.
To a lesser extent.

Molecular beam crystal growth is performed while the substrate is heated to, for example, 600° C. to 700° C. If the in-plane temperature distribution varies by as much as 10°C, the quality of the epitaxially grown layer will deteriorate.

FIG. 7 is a graph showing the temperature distribution within the substrate.

In this case, an area of about 1 mm around the periphery of the substrate cannot be used for making electronic devices.

Even if the substrate is a 3-inch wafer (approximately 7.6 eIn diameter), the yield will be extremely poor if an area of 2σ in diameter is wasted.

The in-plane non-uniform temperature distribution occurs because the stepped portion of the substrate holder 3 holds the periphery of the back surface of the substrate 1. Since this stepped portion blocks radiant heat from the heater, the temperature at the periphery of the substrate does not rise easily.

The lower stepped portion of this holder will be referred to as a substrate seat.

The portion of the substrate holder where the substrate is held is only slightly wide. However, the width of the low-temperature part is about one cell.

The reason why the temperature becomes low at the periphery is thought to be as follows.

Since the board seat 11 is provided, radiant heat from the heater is blocked at the peripheral edge of the board.

Heat from the intermediate portion of the substrate is transmitted to the open space at the periphery of the substrate by thermal conduction. Furthermore, the holder 30 board seat 11
There is also heat conduction from the

Although there is such heat conduction, the temperature at the periphery is still lower than that at the middle.

00 Purpose It is an object of the present invention to provide a substrate holder structure that does not use In and allows the temperature distribution of the substrate to be uniform within the plane.

(6) Structure In the substrate holder shown in FIGS. 5 and 6, the temperature at the periphery of the substrate is lower than that at the middle part because the radiation from the heater 4 is blocked by the substrate seat 11. be.

However, since the board seat is at a high temperature, the heat from the board should be transferred to the periphery of the board. Even with this heat, the temperature at the periphery is still low because there is radiation from the surface of the periphery.

Since the substrate 1 is at a high temperature, heat is radiated from the surface.

The temperature of the substrate becomes constant when the absorption of heat from the heater and the release of heat from the surface by heat radiation are balanced.

The surface of the substrate is exposed in the same way at both the intermediate and peripheral portions.

However, only the peripheral edge of the surface of the substrate is blocked by the substrate seat of the substrate holder. Therefore, less heat is absorbed in the peripheral portion, and naturally less heat is radiated. For this reason, the temperature at the periphery is lower than that at the middle.

Therefore, the inventor thought that it would be sufficient to suppress the radiation of heat at the peripheral portion. If there is no radiation of heat at the periphery, the only heat dissipation at the periphery is conduction. Since heat transfer through thermal conduction is slow, almost no heat is lost at the periphery.

If this is done, the balance of heat will be balanced, the temperature at the periphery will not be lower than the other parts, and the temperature distribution within the plane will be uniform.

This will be explained using drawings.

FIG. 1 is a plan view showing the structure of a substrate holder mechanism of a molecular beam epitaxy apparatus according to the present invention, and FIG. 2 is a longitudinal sectional view.

The semiconductor substrate 1 is made of GaAs, InP, or S.
It is a wafer such as i. A ring-shaped substrate holder 3 having an L-shaped cross section supports the periphery of the back surface of the substrate 1.

The width of the portion of the substrate seat 11 that corresponds to the periphery of the substrate supporting the periphery of the substrate is about 1 to 2 min.

The thickness of the substrate holder 3 is, for example, 3πM to 4πm.

The diameter depends on the size of the substrate. The material can withstand high temperatures and does not outgas. The condition is MO or Ta.

Since the substrate holder 3 has a wide opening, the substrate 1 is directly heated by the radiant heat from the heater 4.

As mentioned above, the substrate does not absorb much infrared rays, so
A substance that increases absorption, such as TiN, is sometimes deposited on the back side of the substrate.

The structure of the heater 4 is arbitrary.

Conventionally, many heaters for heating a substrate have heater wires meandering radially from the center to the outside and from the outside to the center. There is also a double spiral heater, which looks like a combination of two mosquito coils.

Equal heat must be radiated to all parts of the board. Although the heater is fixed, the substrate holder rotates, so the amount of heat is averaged in the rotation direction.

Since the thermal acid distribution must also be averaged in the radial direction, various heater wire distributions have been devised.

The present invention can be applied to heaters of any shape.

The novelty of the present invention is that an annular radiation blocking ring 5 is provided to press the opposite periphery of the board, and the inner diameter Δ of this ring is made almost the same as the inner diameter Σ of the board seat. did,
That's what it means.

The radiation blocking ring 5 and the substrate holder 3 are fixed by making a vertical hole, passing a wire 21 through the hole, and twisting both ends of the wire.

Due to the high temperature of about 700°C, the connection between the screw hole and the screw often seizes and cannot be removed. Even if it comes out, it cannot withstand being used again.

Therefore, the ring 5 and the substrate holder 3 are fixed together using the wire 21.

In FIG. 2, the substrate seat 11 of the holder is shown close to the heater 4, and the radiation blocking ring 5 is located far from the heater 4.

This may be the other way around.

That is, the arrangement may be such that the holder 3 is inverted. The arrangement is such that the ring 5 is close to the heater and the substrate seat 11 is far from the heater.

n> Effect Comparing the present invention with the prior art shown in FIGS. 5 and 6,
The only difference is whether the board is held down by a ring or by three small stoppers 2.

Heat is lost from the periphery of the substrate due to radiation, and in order to prevent this, a radiation blocking ring 5 is used in the present invention.

The condition that the inner diameter Δ of the radiation blocking ring 5 and the inner diameter Σ of the substrate seat 11 of the substrate holder 3 are approximately equal will be explained.

The peripheral edge of the back surface of the substrate is blocked from radiation from the heater by the substrate seat 11. This area has low heat income.

By suppressing heat dissipation due to radiation, it is possible to suppress the temperature drop at the peripheral edge of the substrate. Therefore, in the present invention, a ring 5 for preventing radiation is provided at the peripheral edge.

The condition Δ-Σ means that heat loss due to thermal radiation will not occur from the peripheral edge of the substrate, which is blocked by the substrate seat and does not receive thermal radiation.

In this way, it is possible to prevent the temperature from decreasing due to heat dissipation at the periphery, so that the temperature distribution becomes substantially uniform over the entire surface of the substrate.

In fact, when conducting experiments in this manner, the difference in temperature between the peripheral portion and the middle portion was reduced to about 1 to 3°C.

Next, the limitations of IA-Σ1 will be described. Since the purpose is to reduce the balance of radiant heat from the periphery of the substrate to zero, it does not have to be strictly A-Σ.

Let W be the diameter of the substrate.

As already mentioned, the difference between W and Σ is about 2 to 4 yttm.

If A is smaller than a value between W and Σ, there is a radiation blocking effect.

The lower limit of A is determined by the extent to which the area of the region where epitaxial growth is not possible is allowed. It is determined by something unrelated to radiation.

The wider the region that can be epitaxially grown, the better.

Therefore, the limit of the difference between Δ and Σ is It can be evaluated as a degree of

(G) Effect: The difference in heating temperature between the middle part and the peripheral part of the substrate becomes extremely small. Therefore, the conditions for epitaxial growth are uniform over the entire surface of the substrate. Therefore, it is possible to form an epitaxially grown film with high uniformity.

Since the periphery of the surface of the substrate is covered, no epitaxial growth film is formed on this portion. Therefore, this part cannot be used to make an element and is wasted. but,
This is about the width of one or two fights on the periphery, and is something to behold.

Since In is not used, there is no need for processes such as pasting and peeling, and no CO gas is emitted. High purity epitaxial layers can be grown.

[Brief explanation of drawings]

FIG. 1 is a plan view showing the substrate holder mechanism of the present invention. Figure 2 is a longitudinal sectional view of the same thing. FIG. 3 is a plan view of a conventional substrate holder. FIG. 4 is a longitudinal sectional view of the same thing as FIG. 3. FIG. 5 is a plan view of another conventional substrate holder. FIG. 6 is a longitudinal sectional view of FIG. 5. FIG. 7 is a temperature distribution diagram on the substrate in the example of FIGS. 5 and 6. １・・・・・・・・・・・・Board 2 songs stopper 3・・・・・・・・・・・・・・・Board holder 4・・・・・・
... Heater 5 ... Radiation blocking ring 6 ...
・・・・・・・・・Low melting point metal 11・・・・・・・・・・
...Osamu Makabe, inventor of circuit board seat

Claims (4)

[Claims] (1) In order to perform molecular beam crystal growth, a substrate holder mechanism that supports a semiconductor substrate includes a substrate seat 11 that has an inner diameter Σ smaller than the substrate diameter W and holds the peripheral edge of one side of the substrate.
an annular substrate holder 3 having an L-shaped cross section; a radiation blocking ring that holds the periphery of the substrate 1 fitted in the substrate seat 11 of the substrate holder 3 and is to be fixed to the substrate holder 3; The inner diameter Λ of the radiation blocking ring 5 and the board seat 11 are
A substrate holder mechanism for a molecular beam epitaxy apparatus, characterized in that the inner diameters Σ of the substrates are substantially the same. (2) A substrate holder mechanism for a molecular beam epitaxy apparatus according to claim (1), characterized in that the radiation blocking ring 5 and the substrate holder 3 are connected by a wire. (3) There are mathematical formulas, chemical formulas, tables, etc. between the absolute value of the difference between the inner diameter Σ of the board seat 11 and the inner diameter Λ of the radiation blocking ring 5, and the outer diameter W of the board and the inner diameter Σ of the board seat. Claim No. 1 (1) characterized in that the relationship ▼ holds true.
Substrate holder mechanism of the molecular beam epitaxy apparatus described in ). (4) The substrate holder 3 and the radiation blocking ring 5 are made of Ta or Mo.
A substrate holder mechanism for a molecular beam epitaxy apparatus according to claim 1, characterized in that it is manufactured by.