JPH06224123A - Semiconductor crystal growth method - Google Patents

Abstract

(57) [Summary] [Object] With regard to a method for growing a semiconductor crystal suitable for forming a sharp interface, an object thereof is to provide a method for growing a semiconductor crystal capable of effectively reducing segregation. A method for growing a semiconductor crystal, in which a second semiconductor layer having a partially different composition is grown on a first semiconductor layer, the step of covering the surface of the first semiconductor layer with a substantially uniform hydrogen atom layer. And a step of supplying the raw material of the second semiconductor layer while holding the hydrogen atomic layer, and growing at least one atomic layer of the second semiconductor layer on the surface of the first semiconductor layer by segregation of hydrogen atoms. Including.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for growing a semiconductor crystal, and more particularly to a method for growing a semiconductor crystal suitable for forming a sharp interface.

With the recent demand for higher performance of semiconductor devices, the demand for heterojunctions is also increasing. A plurality of semiconductor layers forming a heterojunction can be arbitrarily selected,
Moreover, it is required that the composition distribution at the interface is steep. Further, it is required that impurity doping can be performed at any place with any profile.

For example, as a high speed and low power consumption semiconductor device, an integrated circuit of a complementary field effect transistor using a compound semiconductor is drawing attention. It is promising to use a Ge layer as the channel material of the p-channel field effect transistor and a III-V group compound semiconductor layer as the channel material of the n-channel field effect transistor.

It is desired to form these channel layers with a steep composition profile and with a highly pure semiconductor material.

[0005]

2. Description of the Related Art Semiconductor crystal growth methods are roughly classified into a liquid phase method and a vapor phase method. The vapor phase method is chemical vapor deposition (CV
D), physical vapor deposition (PVD) and the like. Currently, metal-organic vapor phase epitaxy (MOC) is used for vapor phase crystal growth of compound semiconductors.
VD) and molecular beam epitaxy (MBE) are often used.

MBE can be controlled in atomic layer units, and is widely used for manufacturing semiconductor elements utilizing superlattices and optical semiconductor devices having a heterojunction. MBE
In semiconductor crystal growth using, a molecular beam source for generating a molecular beam is prepared according to the substance to be grown,
A molecular beam source is selectively used to grow the desired material layer.

[0007]

In the case of growing a second semiconductor layer having a different composition at least on the first semiconductor layer, a combination of the first and second semiconductor layers,
It is difficult to prevent the material of the first semiconductor layer, at least a part of which is segregated in the second semiconductor layer, depending on the growth method thereof. That is, it was difficult to form a sharp interface by mixing the two at the interface.

For example, in the case of Ge and III-V group compound semiconductors, the effect of this phenomenon is remarkable. Ge is a group IV semiconductor and serves as a dopant in the group III-V compound semiconductor. In group IV semiconductors such as Ge, III, V
Group atoms also serve as dopants.

Therefore, when forming a heterojunction between a group IV semiconductor and a group III-V compound semiconductor, the doping profile becomes difficult to control unless a sharp heterojunction can be formed.

Not only in the case of the heterojunction of the base semiconductor, but also in the case of doping the semiconductor, the desired characteristics cannot be obtained unless the desired doping profile is obtained. However, the dopant was segregated depending on the combination of the semiconductor material and the dopant, and the desired doping profile could not be obtained.

The concept of segregation and diffusion used in this specification will be described. FIG. 2A shows the concept of diffusion. An interface 8 is formed between the first semiconductor layer 6 and the second semiconductor layer 7.
Grow continuously. After the growth, the constituent material of the first semiconductor layer 6 and the constituent material of the second semiconductor layer 7 exchange their positions. This phenomenon is called diffusion and occurs randomly at a given activation energy. Therefore, after a predetermined time and a predetermined temperature have passed since the interface was formed, the hetero-interface has some disturbance.

FIG. 2B shows the phenomenon of segregation. Segregation means that when an attempt is made to grow the second semiconductor layer 7 having the second composition on the first semiconductor layer 6 having the first composition, the constituent material of the first semiconductor layer 6 disposed below is It is a phenomenon of moving to the surface side.

Similar to diffusion in that the constituent materials of the first semiconductor layer 6 and the second semiconductor layer 7 are replaced, segregation preferentially occurs between the growing atomic layer and the layer immediately below. The difference occurs.

That is, the second semiconductor is formed on the first semiconductor layer 6.
When the mono-atomic layer is grown on the semiconductor layer 7 of
The semiconductor layer 7 has a neighboring atom only on the lower side and does not have a neighboring atom on the upper side. Therefore, segregation is performed by incorporating the constituent atoms of the second semiconductor layer, which are easy to move.

Therefore, the activation energy of segregation is generally lower than the activation energy of diffusion, and segregation is more likely to occur than diffusion. An object of the present invention is to provide a method for growing a semiconductor crystal capable of effectively reducing segregation.

[0016]

A method for growing a semiconductor crystal according to the present invention is a method for growing a semiconductor crystal, in which a second semiconductor layer having a partially different composition is grown on a first semiconductor layer. A step of covering the surface of the first semiconductor layer with a substantially uniform hydrogen atom layer; and supplying the raw material of the second semiconductor layer while holding the hydrogen atom layer, and segregating hydrogen atoms to the surface of the first semiconductor layer. Growing a second semiconductor layer of approximately two atomic layers.

[0017]

[Function] On the first semiconductor layer covered with the hydrogen atom layer,
When the raw material of the second semiconductor layer is supplied, the raw material of the second semiconductor layer enters under the hydrogen atom layer to grow the second semiconductor layer on the first semiconductor layer.

Since the second semiconductor layer is covered with the hydrogen atom layer, it does not become the uppermost atom layer. Therefore, the first
Segregation between the semiconductor layer and the second semiconductor layer can be prevented. Further, by supplying the raw material of the second semiconductor layer, if the second semiconductor layer becomes almost a diatomic layer, segregation can be prevented even if the hydrogen atomic layer is removed.

[0019]

EXAMPLES The present invention was made based on the fact that hydrogen is an element which is extremely prone to segregate. FIG. 1 is a schematic view for explaining a method for growing a semiconductor crystal according to a basic embodiment of the present invention.

As shown in FIG. 1A, first, the first semiconductor layer 1 is grown. Next, as shown in FIG. 1B, a hydrogen atom layer 3 having a substantially uniform distribution is formed on the surface of the first semiconductor layer 1.

Hydrogen atoms are preferentially adsorbed on the surface of the underlying semiconductor and are not adsorbed much on the surface on which hydrogen has already been adsorbed, so that a hydrogen layer of a monoatomic layer is formed. Although the case where a monoatomic hydrogen atom layer is formed is described here, the hydrogen atomic layer does not necessarily have to be completely monoatomic layer formed. If the hydrogen atom layer that covers the entire surface of the first semiconductor layer 1 substantially uniformly is formed, the function of the hydrogen atom layer on the first semiconductor layer is exhibited. For example, if about 50% of the surface of the first semiconductor layer, which is the underlying surface, is covered with a substantially uniform hydrogen atom layer, it can be said that a substantially uniform hydrogen atom layer is formed.

As shown in FIG. 1C, the constituent molecules of the second semiconductor layer are supplied from above the first semiconductor layer 1 on which the monoatomic hydrogen atom layer 3 is formed. Even if the constituent molecules of the second semiconductor layer are deposited on the hydrogen atom layer 3, the hydrogen atoms are extremely likely to segregate. Therefore, the hydrogen atoms move to the surface and the deposited constituent molecules of the second semiconductor layer are: Move onto the first semiconductor layer 1.

In this way, the hydrogen atom layer 3 behaves as if it were a filter, and the constituent atoms of the second semiconductor layer supplied from above are deposited on the surface of the first semiconductor layer.

This phenomenon continues continuously while the hydrogen atom layer 3 covers the surface, and the constituent atoms of the second semiconductor layer 2 supplied from the surface side pass through the hydrogen atom layer 3 to form 1
As shown in (D), the second semiconductor layer 2 is grown.

When the second semiconductor layer 2 grows by two atomic layers or more, the constituent atoms of the first semiconductor layer 1 at the interface and the second atomic layer
The constituent atoms of the semiconductor layer 2 face each other with a background of two or more atomic layers. Therefore, even if mutual substitution of atoms contacting each other at the interface occurs, the phenomenon is not segregation,
It becomes diffusion.

Although the case where the second semiconductor layer is a diatomic layer has been described here, it is not always necessary to form a complete diatomic layer. If the second semiconductor layer is formed so as to cover the entire surface of the first second semiconductor layer substantially uniformly, the surface is stabilized.

For example, if about 50% of the surface of the first semiconductor layer is covered with the substantially uniform second semiconductor layer, segregation can be effectively prevented even if the hydrogen atom layer is removed. In this case, although it is about 1.5 atomic layers, it can be functionally considered to be almost two atomic layers.

After that, as shown in FIG. 1E, even if the hydrogen atom layer 3 is extinguished, segregation at the interface does not occur. Even if segregation occurs on the surface, since the same substance is used, the composition profile does not sag.

Since the activation energy of diffusion is higher than that of segregation, atom transfer is limited. In this way, segregation is effectively prevented and a sharp interface is realized.

FIG. 3 shows the structure of the MBE apparatus used in the embodiment of the present invention. In the figure, 11 is an ultrahigh vacuum chamber, which is evacuated by a vacuum pump 22 such as a turbo molecular pump. The semiconductor substrate 12 is placed on the substrate holder 13 and heated by the substrate heater 14.

In the ultra-high vacuum chamber 11, a plurality of Knudsen (K) cells 16 for metal sources are arranged.
A shutter 15 is arranged in front of each K cell 16 to block the raw material molecules emitted from the K cell.

The radical beam cell 21 is connected to a gas cylinder 19 filled with hydrogen via a flow control device 20. 13.56 MHz for the radical beam cell 21
Is connected to a high-frequency power source to generate hydrogen radicals by converting hydrogen gas into plasma. The generated hydrogen radicals are
The substrate 12 is irradiated with a hydrogen radical beam.

The shutter 15 turns on / off the material beam.
It can be opened and closed. The combination of materials necessary for growing each semiconductor layer is performed by opening and closing the shutter 15, and a substance having a desired composition is deposited on the substrate 12. In the figure, the liquid nitrogen shroud 17 covers the inner surface of the ultrahigh vacuum chamber 11 to prevent re-evaporation of deposits.

Hydrogen molecules may be used instead of hydrogen radicals. However, in the case of hydrogen molecules, after attaching to the surface in the form of hydrogen molecules, it is converted into the form of hydrogen atoms before terminating the surface. Therefore, the efficiency of termination is higher when hydrogen radicals are supplied rather than hydrogen molecules.

FIG. 4 shows a material supply timing when a layer of a second semiconductor material is continuously grown on a layer of a first semiconductor material according to an embodiment of the present invention. The supply of the material of the first semiconductor is completed, and after a certain period of time, the supply of the material of the second semiconductor is started. Here, before the supply of the material of the first semiconductor is completed, the supply of hydrogen is started and
After the supply of the semiconductor material is started, the supply of hydrogen is finished.

The time during which hydrogen is supplied before the end of the supply of the first semiconductor material is τ1, the time during which no raw material is supplied and only hydrogen is supplied is τ2, and the material of the second semiconductor is Let τ3 be the time during which the supply of hydrogen and the supply of hydrogen are performed.

Τ1 is the uppermost layer of the first semiconductor material is V
In the case of a material having a high vapor pressure such as a group element or a group VI element, the surface is covered with a hydrogen atom layer to prevent re-evaporation of the group V or group VI element. It may be omitted if the uppermost layer of the first semiconductor material layer is a material having a low vapor pressure.

Τ2 is the time for covering the surface of the first semiconductor layer with hydrogen before the material of the second semiconductor flies. When τ1 is provided, the surface of the first semiconductor layer may be covered with hydrogen during both τ1 and τ2.

It is preferable that the surface of the first semiconductor layer is completely covered with the hydrogen atom layer, but if the surface is almost uniformly covered with hydrogen, the effect of hydrogen covering occurs. For example, 1 for all surfaces
A hydrogen atom layer corresponding to / 2 may be formed.

In τ3, the second semiconductor layer is 1.5 atomic layer,
Preferably, the surface is covered with a hydrogen atom layer until the growth of two atomic layers. This is to prevent a state in which the material of the second semiconductor becomes the uppermost layer and the layer below it becomes the material of the first semiconductor.

A more specific embodiment will be described below. The first semiconductor is Ge and the second semiconductor is GaAs.
And Filling K cell with Ge, Ga and As respectively,
The temperature of each K cell is 1200 ° C, 1000 ° C, and 200 ° C. At this time, the growth rate of Ge is, for example, 0.5 μm.
/ H, the growth rate of GaAs is 0.8 μm / h.

The Ge substrate is introduced into the growth chamber and the growth chamber is evacuated by the turbo molecular pump. The Ge substrate is heated to 600 ° C. to remove the natural oxide film. Then, the substrate temperature is set to 200
Then, the Ge layer is grown to a thickness of 3000Å, and a GaAs layer is grown to 2000Å on the Ge layer.

The depth-direction profile of each atomic concentration of the sample grown without hydrogen supply according to the conventional technique is shown in FIG.
become that way. Since Ge atoms segregate in the GaAs layer, Ge is deeply present inside the GaAs layer grown on the GaAs layer even though the Ge supply is cut off.

In this embodiment, the Ge substrate is introduced into the growth chamber and heated to 600 ° C. to remove the natural oxide film, then the substrate temperature is lowered to 200 ° C. and the Ge layer is grown to about 3000 Å. Here, the shutter of the K cell for Ge is closed, and a hydrogen radical beam is irradiated after 10 seconds.

The flow rate of hydrogen was 10 sccm and 13.5.
Hydrogen plasma is generated by a high frequency power supply of MHz and 400 W, hydrogen radicals are generated in a radical beam cell, and the Ge substrate is irradiated with the hydrogen radicals.

After irradiating the hydrogen radical beam for 10 seconds (τ2), the shutter of the K cell for As is opened while irradiating the hydrogen radical beam, and after 10 seconds, the shutter of the K cell for Ga is opened. After growing in this state for 25 seconds (τ3), the hydrogen radical beam is stopped.

Since GaAs is supplied after the hydrogen atom layer is formed on the Ge layer, GaAs is buried under the hydrogen atom layer and epitaxially grows on the Ge layer.
At this time, since the Ge atom has a weaker segregation force than the hydrogen atom, Ge segregation hardly occurs. Since the hydrogen atom has a strong segregating force, even if the GaAs crystal growth starts, the hydrogen atom always moves to the surface by the segregation.

Therefore, the segregation of Ge is prevented, and FIG.
A depth-direction profile of each atomic concentration as shown in (B) is obtained. We started the crystal growth of GaAs from As,
You can also start with Ga. In this case, after irradiating the hydrogen radical beam for 10 seconds, the shutter of the K cell for Ga is opened while irradiating the hydrogen radical beam to grow the Ga1 atomic layer, that is, after 1.3 seconds, the shutter of As is opened. Good.

Next, an example of growing a Ge layer on the GaAs layer will be described. GaAs is a group III atom, G
The vapor pressure of a is relatively low, but the vapor pressure of As, which is a group V element, is high.

Therefore, especially when the growth is stopped in the As layer, it is desirable to prevent the atomic layer from being disturbed by the re-evaporation of As. For this purpose, τ1 can be used.

First, the GaAs substrate is introduced into the growth chamber, evacuated to an ultrahigh vacuum, and then heated to 600 ° C. while irradiating with As to remove the natural oxide film. The substrate temperature is kept at 600 ° C. and the GaAs layer is grown to about 3000 Å. At the end of the growth, the shutter of the K cell for Ga is first closed, and at the same time, the irradiation of the hydrogen radical beam is started. A after 10 seconds (τ1)
The shutter of the K cell for s is closed.

After continuously irradiating the hydrogen radical beam for 10 seconds (τ2), the shutter of the K cell for Ge is opened while continuing the irradiation of the hydrogen radical beam. 25 seconds (τ3)
After growing in this state for a while, the hydrogen radical beam is stopped. For example, a Ge layer having a thickness of about 2000Å is grown.

Although the growth of the GaAs layer is terminated on the As plane, it may be terminated on the Ga plane. In this case, the shutter of the K cell for As is closed 10 seconds after the irradiation with the hydrogen radical beam, and the shutter of the K cell for Ga is closed after the Ga1 atomic layer is grown, that is, 1.3 seconds later.

At the end of the crystal growth of the GaAs layer,
Since its surface is covered by the hydrogen atomic layer, Ge
Segregation of GaAs is prevented when growing the layer. Ge diffuses under the hydrogen atom layer and grows on the GaAs layer. Therefore, a depth-direction profile of atomic concentration as shown in FIG. 5C is obtained.

Although the case where the Ge layer and the GaAs layer are continuously grown has been described above, other semiconductors can be grown as the first semiconductor and the second semiconductor. It is also possible to grow a layer having the same base material but different impurities or different impurity concentrations.

When it is desired to grow a layer having a low impurity concentration on a layer having a high impurity concentration, if segregation occurs and a region having a low impurity concentration has a high impurity concentration, desired semiconductor device performance is You won't get it.

A case where a region doped with Sn is embedded in the GaAs layer will be described. Undoped GaAs layer,
First, it grows up to 3000Å, and 1 × 10 ¹⁸ cm ^-3 S is grown on it.
An n-doped GaAs layer is grown to 100 Å, and a non-doped GaAs layer is further grown to about 2000 Å on its surface.

When each of these layers is grown as it is by the conventional technique, as shown in FIG.
The n concentration has a profile with a tail on the surface side. In this embodiment, a GaAs substrate is introduced into the growth chamber and
The natural oxide film is removed by heating to 600 ° C. while irradiating. After that, the substrate temperature is set to 550 ° C. and a non-doped GaAs layer is grown by about 3000 Å.

In order to form the Sn-doped layer, the hydrogen radical beam is irradiated for 10 seconds (τ1) before Sn irradiation. While irradiating a hydrogen radical beam (τ2), Sn
To form a Sn-doped region of 100 Å and stop the Sn irradiation.

After the Sn irradiation, the hydrogen radical beam is irradiated and the growth of the GaAs layer is continued. 25 after stopping the Sn irradiation
After 2 seconds (τ3), the hydrogen radical beam is stopped. In this embodiment, the outermost surface of the growth layer is covered with the hydrogen atom layer while the Sn-doped GaAs layer is grown. Therefore, Se is prevented from segregating, and a sharp impurity distribution is obtained.

Thus, by covering the crystal growth surface with the hydrogen atom layer during crystal growth, it is possible to prevent segregation and form an interface in which the atomic concentration profile changes abruptly. The interface includes a case of a heterojunction and a case of an interface of an impurity-doped region.

The semiconductor stack having such a steep interface can be used for various semiconductor devices. For example, it can be used for a high electron mobility transistor, especially a complementary high electron mobility transistor, a hetero bipolar transistor, or the like.

The present invention has been described above with reference to the embodiments.
The present invention is not limited to these. For example,
It will be apparent to those skilled in the art that various changes, improvements, combinations and the like can be made.

[0064]

As described above, according to the present invention,
Since crystal growth that prevents segregation can be performed, it is possible to obtain a laminated structure in which the atomic concentration distribution changes sharply. Therefore, it is possible to form a heterojunction or a doping region having good characteristics.

[Brief description of drawings]

FIG. 1 is a conceptual diagram for explaining an embodiment of the present invention.

FIG. 2 is a conceptual diagram for explaining diffusion and segregation.

FIG. 3 is a schematic sectional view of a molecular beam epitaxy apparatus used in an example of the present invention.

FIG. 4 is a timing chart showing a raw material supply timing in the embodiment of the present invention.

FIG. 5 is a graph showing a composition distribution according to an example of the present invention.

FIG. 6 is a graph showing a composition (impurity) distribution according to an example of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 1st semiconductor layer 2 2nd semiconductor layer 3 Hydrogen atom layer 6 1st semiconductor layer 7 2nd semiconductor layer 8 Interface 11 Ultra high vacuum chamber 12 Substrate 13 Substrate holder 14 Substrate heating heater 15 Shutter 16 K cell 17 Liquid nitrogen shroud 19 Hydrogen cylinder 20 Flow controller 21 Radical beam cell 22 Turbo molecular pump

Claims (2)

[Claims] 1. A method for growing a semiconductor crystal, comprising: growing a second semiconductor layer (2) having a partially different composition on the first semiconductor layer (1). A step of covering the surface with a substantially uniform hydrogen atom layer (3); a raw material for the second semiconductor layer is supplied while the hydrogen atom layer (3) is held, and the first semiconductor layer ( 1) A method of growing a semiconductor crystal, which comprises a step of growing a second semiconductor layer (2) having approximately two atomic layers on the surface. 2. A step of growing the first semiconductor layer (1), and hydrogen is supplied before the growth of the first semiconductor layer is finished, and the surface of the first semiconductor layer is almost completely left when the growth of the first semiconductor layer is finished. The method for growing a semiconductor crystal according to claim 1, further comprising the step of covering with a uniform hydrogen atom layer (3).