JPH0661150A - Compound semiconductor device and plasma doping - Google Patents

Abstract

PURPOSE:To obtain a compound semiconductor device having a sheet carrier concentration of a saturation point higher than a conventional saturation point by a method wherein second gas containing a group V element, which is different from that of a compound semiconductor layer, as its main component is supplied on the compound semiconductor layer consisting of a III-V compound semiconductor. CONSTITUTION:First, TMA and TMI, which are the raw material for a group III element, are supplied on a semi-insulative InP substrate 10A heated in a reactor simultaneously with the raw material for a group V element, such as an arsine, to grow an undoped AlInAs layer 16A. Then, the supply of the TMA and the TMI is stopped to interrupt the growth fo the layer 16A and in this state, disilane, for example, is introduced in the reactor for forming an Si planar doped layer 18 consisting of Si. At this time, the arsine is continued to introduce in the reactor. Moreover, second gas, such as phosphine gas, containing a group V element, which is different from that of a compound semiconductor layer, as its main component is supplied. After this, the growth of the layer 16A is again started to form an undoped AlInAs layer 20A.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a compound semiconductor device and a planar doping method, and more particularly to a compound semiconductor device having a high sheet carrier concentration and a planar doping method.

[0002]

2. Description of the Related Art A planar doping method (also called a delta doping method or an atomic layer doping method) is used for selective doping of a hetero interface field effect transistor (HIFET) represented by a high electron mobility transistor (HEMT). Has been applied to. MOCVD
A planar doping layer is formed by performing planar doping on the compound semiconductor layer using a vapor deposition method such as
Further, the sheet carrier concentration of the three-layer structure layer obtained by forming the compound semiconductor layer on the planar doping layer becomes saturated when the doping gas supply amount increases.
It is considered that this saturation value depends on the impurity element used for planar doping and the element forming the compound semiconductor layer. For example, III-V made of GaAs
When the group compound semiconductor layer and the planar doping layer made of Si are combined, the sheet carrier concentration is about 1.
Saturates at 5 × 10 ¹³ cm ^-2 (for example, refer to the document "DONOR SU
BBANDS FOR THE δ-DOPING LAYER IN GaAs-CENTRAL CEL
L AND VALLEY-ORBIT EFFECTIN TWO DIMENSIONS ", A. Zr
enner, et al., 18th International Conference onthe
Physics of Semiconductors (World Scientific, Sing
apore, 1987), pp 1523, or the document "Efficient
Si Planar Doping in GaAs by Flow-Rate Modulation E
pitaxy ", N. Kobayashi, et al., Japanese Journal of
Applied Physics, 25, L746, 1986).

The present applicant has previously proposed that a compound semiconductor layer made of AlInAs / a planar doping layer made of Si /
When a three-layer structure layer of a compound semiconductor layer made of AlInAs is formed, a method of growing a compound semiconductor is proposed which utilizes a phenomenon in which the sheet carrier concentration is saturated when the supply amount of a doping gas is set to a certain value or more. Kaihei 3-21
No. 8008 (Japanese Patent Application No. 2-13125). According to this method, a doping gas having a uniform carrier concentration can be obtained by supplying a doping gas at a supply flow rate at which the sheet carrier concentration is saturated to perform planar doping. Can prevent non-uniformity of the threshold voltage in the layer.

The conventional planar doping method using the MOCVD method will be described below. The compound semiconductor layer is made of Al
It is made of InAs and the planar doping layer is made of Si. More specifically, as shown in FIG. 1, undoped A
A planar doping layer 18 made of Si is sandwiched between the lInAs layers 16A and 20B. The undoped AlInAs layer 16A is formed on the semi-insulating InP substrate 10A.

In order to obtain such a layer structure, first, a semi-insulating substrate 10A which is placed in a reactor and heated is
A group III element material such as trimethylaluminum (TMA) and trimethylindium (TMI) are supplied simultaneously with a group V element material such as arsine (AsH ₃ ). Thereby, the undoped AlInAs layer 16A can be grown on the semi-insulating substrate 10A.

Next, for example, T, which is a raw material of the group III element,
The supply of MA and TMI to the reactor is stopped to interrupt the growth of the undoped AlInAs layer 16A. And
In this state, for example, disilane (Si ₂ H ₆ ) is introduced into the reactor in order to form the planar doping layer 18 made of Si. At this time, a gas containing As which is the group V element of the compound semiconductor layer as a main component, such as AsH _3, is continuously introduced into the reactor. The reason is that the compound semiconductor layer (Al
This is to prevent As having a high vapor pressure from being thermally dissociated from the InAs layer 16A). When the formation of the planar doping layer 18 is completed, the supply of the raw material of the planar doping layer, for example, Si ₂ H ₆ is stopped and the growth of the undoped AlInAs layer 20A is started again. As a result, the structure shown in FIG. 1 can be obtained.

FIG. 2 shows a supply state of various gases in this case. The supply state of various gases is generally called a growth sequence. “ON” in FIG. 2 indicates a state where the raw material gas is introduced into the reactor. In the case of FIG. 2, t1 and t
Reference numeral 5 corresponds to the growth period of the AlInAs layers 16A and 20A, and t3 corresponds to the formation period of the planar doping layer 18. t2
And t4 are purge periods. At t2, the raw materials of the group III element, such as TMA and TMI, are purged, and at t4, the raw materials of the planar doping layer, such as Si ₂ H _6, are purged. Purge period t2, t4
May be omitted.

The sheet carrier concentration obtained by the planar doping method is the flow rate ratio (flow rate: sccm) of the source gas of the planar doping layer made of Si such as Si ₂ H ₆ and the supply time of the Si ₂ H ₆ (planar). It is a function of the product of the doping layer formation period t3). Assuming that the value of this product is the total supply amount (total mass flow), as shown in FIG. 3, when the value of this product exceeds a certain value, the value of the sheet carrier concentration is saturated to a certain value.
In the case of the compound semiconductor layer made of AlInAs and the planar doping layer made of Si, the sheet carrier concentration is saturated at about 1.5 × 10 ¹³ cm ^-2 , as shown in FIG.

[0009]

As described above, AlI
Compound semiconductor layer made of nAs or GaAs and Si
In the case of a planar doping layer consisting of, the sheet carrier concentration saturates at approximately 1.5 × 10 ¹³ cm ^-2 .

In HIFET and the like, it is desirable to make the sheet carrier concentration as high as possible. The reason for this will be described below with the HEMT whose schematic cross-sectional view is shown in FIG.
The HEMT includes a channel layer 102 formed on the semi-insulating substrate 100, a spacer layer 106 formed on the channel layer 102, and a barrier layer 108 formed on the spacer layer 106. The two-dimensional electron channel 104 is formed in the channel layer 102 near the interface with the spacer layer 106.

The relationship between the doping concentration (Nd) of the barrier layer 108 and the maximum value of the two-dimensional electron gas density that can be accumulated in the two-dimensional electron channel layer 104 in the HEMT made of the following materials is shown in FIG. The thickness of the layer 106 is shown as a parameter in FIGS. 5A and 5B. 5A; Barrier layer (108): n-type AlGaAs spacer layer (106): Undoped AlGaAs channel layer (102): Undoped GaAs Semi-insulating substrate (100): GaAs FIG. 5B; Barrier Layer (108): n-type AlInAs spacer layer (106): undoped AlInAs channel layer (102): undoped GaInAs semi-insulating substrate (100): InP

The upper limit of the two-dimensional electron gas density is the barrier layer 1.
This is defined by the generation of carriers at 08 (so-called parallel conduction). The donor ionized in the barrier layer 108 functions to lift the energy of the conduction band of the barrier layer away from the Fermi energy according to the Poisson equation due to the space charge. Therefore, the higher the doping concentration of the barrier layer, the more the spacer layer 10
The thinner 6 is, the harder it is for parallel conduction to appear,
The maximum value of the two-dimensional electron gas density can be increased.

However, if the spacer layer is made thin, inconvenience occurs. Based on the sample shown in FIG. 5B, the relationship between the maximum value of the two-dimensional electron gas density and the electron mobility at 300 ° K is shown in FIG. 6 with the thickness of the spacer layer 106 as a parameter. The electron mobility is obtained by Hall measurement. The electron mobility has a convex curve. The increase in electron mobility on the side of the lower maximum of the two-dimensional electron gas density is considered to be due to the effect of screening or the decrease in Coulomb scattering due to the increase in Fermi energy. The decrease in electron mobility on the side where the maximum value of the two-dimensional electron gas density is high is considered to be due to parallel conduction in the barrier layer.

Now, considering only the side where the maximum value of the two-dimensional electron gas density without parallel conduction is low, the thicker the spacer layer, the higher the electron mobility tends to be. It is considered that this is because the thicker the spacer layer, the larger the distance between the ionized donor and the two-dimensional electron gas in the barrier layer and the smaller Coulomb scattering.

As described above, in order to generate a larger amount of two-dimensional electron gas, it is necessary to reduce the thickness of the spacer layer or increase the doping concentration of the barrier layer, but from the viewpoint of electron mobility. Therefore, it is not desirable to reduce the thickness of the spacer layer. Therefore, if a barrier layer having a high doping concentration can be formed and the spacer layer can be made thicker, a larger amount of two-dimensional electron gas can be generated at a higher speed, and a HEMT having high characteristics can be manufactured.

However, AlInAs or GaAs
When a compound semiconductor layer made of is used, as described above,
The sheet carrier concentration is saturated at about 1.5 × 10 ¹³ cm ^-2 , and it is hard to say that the maximum value of the sheet carrier concentration is sufficiently high. For example, the design condition of HIFET is limited 1
It is one of the factors.

Therefore, an object of the present invention is to provide a compound semiconductor having a sheet carrier concentration higher than the conventional saturation value in a three-layer structure layer of compound semiconductor layer / planar doping layer / compound semiconductor layer obtained by a planar doping method. To provide a device. Further, the object of the present invention is to
It is an object of the present invention to provide a planar doping method suitable for manufacturing such a compound semiconductor device.

[0018]

The above-mentioned objects are III-
A compound semiconductor layer made of a group V compound semiconductor, a doping gas formed in the compound semiconductor layer, a first gas containing a group V element of the compound semiconductor layer as a main component, and a group V element of the compound semiconductor layer What is claimed is: 1. A compound semiconductor device comprising: a planar doping layer made of a main component element of a doping gas, which is formed on the basis of a second gas whose main component is a different V group element. The bonding energy of the group V element of the gas of the above-mentioned V constitutes the compound semiconductor layer with the main element of the doping gas.
This can be achieved by the compound semiconductor device of the present invention, which is equal to or larger than the bonding energy of the group element.

In a preferred embodiment of the compound semiconductor device of the present invention, the main component element of the doping gas is Si, the V group element of the second gas is P, and the V group element constituting the compound semiconductor layer is As, or a combination of As and P.

The above object is further obtained by adding a doping gas on a compound semiconductor layer made of a III-V group compound semiconductor.
A first gas containing a group V element of the compound semiconductor layer as a main component,
And supplying a second gas containing a group V element different from the group V element of the compound semiconductor layer as a main component to form a planar doping layer made of the main component element of the doping gas on the compound semiconductor layer. Can be achieved by the planar doping method of the present invention.

The bonding energy of the main element of the doping gas and the group V element of the second gas is equal to or larger than the bonding energy of the main element of the doping gas and the group V element of the compound semiconductor layer. A satisfactory first gas and second gas are selected. In a preferred embodiment of the planar doping method of the present invention, the main element of the doping gas is Si, the group V element of the first gas is As, or a combination of As and P, and the second gas is Group V element is P.

The compound semiconductor layer made of a III-V group compound semiconductor is, for example, AlInAs, GaAs, AlG.
aAs, GaInAs, AlGaInAs, InAs,
It can be composed of GaInAsP and InAsP.

[0023]

According to the conventional planar doping method, a doping gas and a gas containing a group V element of the compound semiconductor layer as a main component are supplied onto the III-V group compound semiconductor layer to form the main component element of the doping gas. A planar doping layer is formed on the compound semiconductor layer. The reason why the gas containing the group V element of the compound semiconductor layer as the main component is supplied is that the group V element (eg, As) having a high vapor pressure is applied from the compound semiconductor layer (eg, AlInAs layer) on which the planar doping layer is to be formed. This is for avoiding thermal dissociation of. On the other hand, the planar doping method of the present invention is further characterized by supplying a second gas containing a V group element different from the V group element of the compound semiconductor layer as a main component.

It is assumed that the main component element of the doping gas is Si, the group V element of the first gas is As, and the group V element of the second gas is P. The compound semiconductor layer is assumed to be made of AlInAs. The bond energies of each element are as follows (Reference "Im
purity effect on grown-in dislocation density of I
nP and GaAs crystals ", Y. Seki, et al., Journal of
Applied Physics, 49 (1978), pp 822).

As vacancy exists in the compound semiconductor layer made of AlInAs. When planar doping is applied to this compound semiconductor layer, As which is a group V element of the first gas
And P, which is the group V element of the second gas, fills the As vacancy. When Si, which is the main element of the doping gas, is adsorbed on the surface of the compound semiconductor layer, P exists on the surface of the compound semiconductor layer. The bond energy of Si-P (2.06 keV) is higher than the bond energy of Si-As (1.65 keV). Therefore, when P is present on the surface of the compound semiconductor layer, the number of Si atoms adsorbed on the compound semiconductor surface is increased as compared with the case where only As is present on the surface of the compound semiconductor layer. It is considered that a high sheet carrier concentration can be obtained.

As described above, in the ordinary planar doping method, in order to suppress the dissociation of the group V element from the compound semiconductor layer, the gas containing the group V element of the compound semiconductor as the main component is supplied onto the compound semiconductor layer. However, in the present invention, in addition to that, a second gas containing a group V element, which is considered to have a strong bond with the atoms constituting the planar doping layer, as a main component is further supplied, By suppressing the desorption action of the group V element of the compound semiconductor layer as much as possible and controlling the supply flow rate ratio of each gas, it is possible to obtain a sheet carrier concentration higher than the saturation value of the sheet carrier concentration in the conventional planar doping method.

[0027]

EXAMPLES The planar doping method of the present invention will be described below. The compound semiconductor layer made of a III-V group compound semiconductor is made of AlInAs, and the planar doping layer is made of Si. More specifically, as shown in FIG. 1, undoped AlInAs layers 16A and 20A.
(Compound semiconductor layer made of III-V group compound semiconductor)
A planar doping layer 18 made of Si is sandwiched between them.

In order to obtain such a layer, first, as in the conventional case, first, a semi-insulating substrate 10A placed in a reactor and heated, a raw material of a group III element such as trimethyl aluminum (TMA) and Trimethylindium (TMI) is a raw material for Group V elements, such as arsine (A
sH ₃ ) and supplied at the same time. As a result, on the semi-insulating substrate 10A, for example, 200 nm thick undoped Al is formed.
The InAs layer 16A can be grown.

Next, for example, T, which is a raw material of the group III element,
The supply of MA and TMI to the reactor is stopped to interrupt the growth of the undoped AlInAs layer 16A. And
In this state, in order to form the planar doping layer 18 made of Si, for example, disilane (Si ₂ H ₆ 50 ppm /
H ₂ ) is introduced into the reactor. At this time, a gas whose main component is As which is a group V element of the compound semiconductor layer, for example, A
Continue to introduce sH ₃ (10% / H ₂ ) into the reactor. The reason is to avoid thermal dissociation of As having a high vapor pressure from the compound semiconductor layer (AlInAs layer) in which the planar doping layer is formed.

In addition to this, V of the compound semiconductor layer is further added.
A second gas (for example, phosphine, PH ₃ 20% / H ₂ ) containing a group V element different from the group element as a main component is supplied.

When the formation of the planar doping layer 18 is completed, the source material of the planar doping layer is, for example, S.
The supply of i ₂ H ₆ is stopped and the growth of the undoped AlInAs layer 20A is started again to form the undoped AlInAs layer 20A having a thickness of 200 nm, for example. As a result, the structure shown in FIG. 1 can be obtained.

The growth sequence in this case is shown in FIG.
t1 is the formation period of the first compound semiconductor layer made of the III-V group compound semiconductor, and t5 is the formation period of the second compound semiconductor layer made of the III-V group compound semiconductor. Further, t2 and t4 are purge periods. t2, t
4 may be omitted. In FIG. 7, a second gas, for example, phosphine is introduced during the planar doping layer formation period of t3. Further, the first gas, for example AsH _3, has a supply flow rate that is smaller than the AsH ₃ supply amount during the growth period (t1) of the first layer made of the compound semiconductor.

The Si atom adsorbed on AlInAs is
By heating in an AsH ₃ atmosphere, AlIn
It is known to desorb from As. The effect of elimination has _third supply flow rate dependency AsH, at high AsH ₃ supplied flow rate for many Si desorbed from AlInAs, the sheet carrier concentration is reduced. Therefore, it is necessary to suppress the desorption of Si by reducing the supply flow rate of AsH ₃ during the formation of the planar doping layer.

Thus, the compound semiconductor device of the present invention can be manufactured. The compound semiconductor device manufactured by the above-mentioned embodiment is formed in (a) the compound semiconductor layer made of AlInAs and (b) the compound semiconductor layer, and the doping gas (Si ₂ H ₆ ) and V of the compound semiconductor layer are formed.
A first gas (As) containing a group element (As) as a main component
H ₃ ), and a second gas (P) mainly containing a group V element (P) different from the group V element (As) of the compound semiconductor layer.
H ₃ ), a planar doping layer made of a main component element (Si) of the doping gas, and a main component element (Si) of the doping gas and a group V element (P) of the second gas. Is larger than the bonding energies of the main component element (Si) of the doping gas and the group V element (As) forming the compound semiconductor layer.

FIG. 8 shows the values of the sheet carrier concentration of compound semiconductor layer / planar doping layer / compound semiconductor layer obtained by using the method of the present invention described above. III-
The formation of the first and second compound semiconductor layers made of a Group V compound semiconductor and the formation of the planar doping layer are performed under reduced pressure MOC.
Using the VD method, the pressure was 200 Torr and the temperature was 640 ° C. The vertical axis of FIG. 8 is the sheet carrier concentration obtained by Hall measurement, and the horizontal axis is the first gas (As
It is the supply flow rate ratio of the second gas (PH ₃ ) introduced at the same time as H ₃ ). The supply flow rate ratio of the first gas (AsH ₃ ) during the planar doping was constant. The supply flow rate ratio of Si ₂ H ₆ was set to 220 sccm, 440 sccm and 660 sccm.

As is clear from FIG. 8, the sheet carrier concentration increases as the Si ₂ H ₆ supply amount ratio increases. When the second gas is not supplied, the sheet carrier concentration is about 1.3 × 10 ¹³ cm ⁻² , whereas when the second gas is supplied, the sheet carrier concentration is about 2.4 × 10 ¹³ cm ^2.
It increased to cm ^-2 .

FIG. 9 shows the relationship between the surface density of Si of these samples obtained by the secondary ion mass spectrometry (SIMS analysis) method and the supply ratio of the second gas (PH ₃ ). As is apparent from FIG. 9, when the second gas is supplied, the surface density of Si increases, and the higher the ratio of Si ₂ H ₆ supply, the higher the surface density of Si. Therefore, it is considered that the sheet carrier concentration increased due to the increase in the surface density of Si.

A schematic sectional view of a HEMT produced by the planar doping method of the present invention is shown in FIG. In FIG. 10A, 10 is a semi-insulating GaAs substrate,
12 is a buffer layer, 14 is an undoped GaAs layer, 16
Is a first compound semiconductor layer made of undoped AlGaAs, 18 shown by a broken line is a planar doping layer made of Si, 20 is a second compound semiconductor layer made of undoped AlGaAs, and 22 is a capping layer. The barrier layer (electron supply layer) is composed of the first compound semiconductor layer 16, the planar doping layer 18, and the second compound semiconductor layer 20, and is composed of undoped GaAs near the interface between the first compound semiconductor layer 16 and the undoped GaAs layer 14. Two-dimensional electron channels are formed in layer 14.

FIG. 10B is a schematic sectional view of another HEMT manufactured by the planar doping method of the present invention.
Shown in. In FIG. 10B, 10A is a semi-insulating InP substrate, 12A is a buffer layer, and 14A is undoped GaIn.
As layer, 16A is a first layer made of undoped AlInAs
Is a planar doping layer made of Si, 20A is a second compound semiconductor layer made of undoped AlInAs, and 22A is a capping layer. The barrier layer (electron supply layer) includes the first compound semiconductor layer 16A, the planar doping layer 18 and the second compound semiconductor layer 20A, and the first compound semiconductor layer 16A.
A two-dimensional electron channel is formed in the undoped GaInAs layer 14A near the interface between the undoped GaInAs layer 14A and the undoped GaInAs layer 14A. In the HEMT shown in FIG. 10,
Although the planar doping layer is one layer, the sheet carrier concentration can be further increased by forming a plurality of planar doping layers in the compound semiconductor layer.

The basic structure of the heterojunction bipolar transistor (HBT) manufactured by the planar doping method of the present invention is shown in the schematic sectional view of FIG. The HBT requires a high concentration and thin base layer. HBTs typically have an np-type with a p-type base layer.
Although it has an n structure, a p-n-p structure having an n-type base layer is also conceivable. In the case of a GaAs-based HBT, it is possible to obtain a sheet carrier concentration higher in the p-type than in the n-type by uniform doping, so that it is easy to form the n-pn structure, but the p-n-p structure is easy. Is difficult to form.
However, according to the planar doping method of the present invention, a high sheet carrier concentration can be obtained even in the n-type, and further, the base layer can be easily thinned. For example, when a planar doping layer made of Si is formed on a compound semiconductor layer made of GaAs, 1.5 ×
It is possible to obtain an n-type sheet carrier concentration of about 10 ¹³ cm ^-2 . When the thickness of the base layer is 5 nm, the carrier concentration per cm ^{3 of} the base layer is 3.0 × 10 ¹⁹ . This carrier concentration value is a value that cannot be obtained by uniform doping.

The HBT shown in FIG. 11A is a p-type GBT.
a collector layer 32 made of p-GaAs, a base layer 34 made of n-GaAs, formed on the aAs substrate 30;
The emitter layer 38 is made of p-AlGaAs. A planar doping layer 36 made of Si is formed in the base layer 34. In this example, four planar doping layers are formed. Incidentally, various electrodes and capping layers are omitted in FIG.

The HBT shown in FIG. 11B is a p-type I.
A collector layer 32A made of p-GaInAs, a base layer 34A made of n-GaInAs, and an emitter layer 38 made of p-AlInAs are formed on the nP substrate 30A.
Composed of A. A planar doping layer 36 made of Si is formed in the base layer 34A. In this example, three planar doping layers are formed.

The HBT shown in FIG. 11C is a p-type I.
The collector layer 32B made of p-GaInAs, the base layer 34B made of n-GaInAs, and the emitter layer 38B made of p-InP are formed on the nP substrate 30B. A planar doping layer 36B made of Si is formed in the base layer 34B. In this example, four planar doping layers are formed.

Although the present invention has been described based on the preferred embodiments, the present invention is not limited to these embodiments. Instead of Si ₂ H ₆ as a doping gas,
SiH ₄ and Si ₃ H ₈ can be used. Further, Ge or Sn can be used as the main element of the doping gas. The planar doping layer and the compound semiconductor layer can be formed by the MBE method as well as the MOCVD method.

Various known gases such as triethylindium can be used for forming the compound semiconductor layer made of a III-V group compound semiconductor. Further, for example, the compound semiconductor layer is GaInAsP or InA.
In the case of sP, Group V elements (As and P) of the compound semiconductor layer
A first gas containing as a main component (for example, AsH ₃ and P
H ₃ ), and group V elements (As and P) of the compound semiconductor layer
A second gas (for example, PH ₃ ) containing a group V element (P) different from that as a main component is supplied. At this time, the supply amounts of the first gas PH ₃ and the second gas PH ₃ are changed. It should be adjusted appropriately.

[0046]

As described above, in the conventional planar doping method, only the first gas containing the group V element of the compound semiconductor layer as the main component is supplied onto the compound semiconductor layer. In the invention, the second gas containing a group V element different from the group V element of the compound semiconductor layer, which is considered to have a strong bond with the atoms constituting the planar doping layer, as a main component is further supplied to perform the desorption action. It is possible to obtain a sheet carrier concentration higher than the conventional saturation concentration by controlling the supply flow rate ratio of the hidden gas such as suppressing the supply amount of the first gas which is considered to be present as much as possible. Therefore, the doping layer can be set at a position away from the channel layer, and the spacer layer shown in FIG. 4 can be set thick. As a result, scattering of two-dimensional electrons due to impurities can be reduced, and a higher performance compound semiconductor device can be manufactured.

[Brief description of drawings]

FIG. 1 is a schematic cross-sectional view showing a partial layer structure of a HEMT.

FIG. 2 is a growth sequence including a conventional planar doping method.

FIG. 3 is a diagram showing a relationship between a sheet carrier concentration and a total supply amount (total mass flow) obtained by a conventional method.

FIG. 4 is a schematic cross-sectional view of a HEMT for explaining that it is desirable to make the sheet carrier concentration as high as possible.

FIG. 5 is a diagram showing the relationship between the doping concentration of the barrier layer and the maximum two-dimensional electron gas density that can be accumulated in the two-dimensional electron channel layer.

FIG. 6 is a diagram showing a relationship between maximum two-dimensional electron gas density and electron mobility.

FIG. 7 is a growth sequence including the planar doping method of the present invention.

FIG. 8 is a diagram showing the relationship between the sheet carrier concentration of the three-layer structure layer obtained by the planar doping method of the present invention and the supply flow rate ratio of the second gas.

FIG. 9 is a diagram showing a relationship between a surface density of Si and a supply amount ratio of a second gas.

FIG. 10 is a schematic cross-sectional view of a HEMT manufactured by the planar doping method of the present invention.

FIG. 11 is a schematic cross-sectional view of an HBT manufactured by the planar doping method of the present invention.

[Explanation of symbols]

10 semi-insulating GaAs substrate 12 buffer layer 14 undoped GaAs layer 16,20 compound semiconductor layer made of undoped AlGaAs 18 planar doping layer made of Si 22,22A capping layer 10A semi-insulating InP substrate 12A buffer layer 14A undoped GaInAs layer 16A, 20A Compound semiconductor layer made of undoped AlInAs 30, 30A, 30B Substrate 32, 32A, 32B Collector layer 34, 34A, 34B Base layer 36 Planar doping layer 38, 38A, 38B Emitter layer 100 Semi-insulating substrate 102 Channel layer 104 Two-dimensional Electron channel 106 Spacer layer 108 Barrier layer

Claims (4)

[Claims] 1. A compound semiconductor layer composed of a III-V group compound semiconductor, a doping gas formed in the compound semiconductor layer, a first gas containing a group V element of the compound semiconductor layer as a main component, and A compound semiconductor device comprising: a planar doping layer formed of a second gas containing a group V element different from the group V element of the compound semiconductor layer as a main component, the planar doping layer including a main component element of a doping gas. The bonding energy of the main component element of the doping gas and the group V element of the second gas is equal to or greater than the bonding energy of the main component element of the doping gas and the group V element forming the compound semiconductor layer. Compound semiconductor device. 2. The main component element of the doping gas is Si, the V group element of the second gas is P, and the V group element constituting the compound semiconductor layer is As or a combination of As and P. The compound semiconductor device according to claim 1, wherein the compound semiconductor device is present. 3. A doping gas and V of the compound semiconductor layer on the compound semiconductor layer made of a III-V group compound semiconductor.
A first gas containing a group element as a main component and a second gas containing a group V element different from the group V element of the compound semiconductor layer as a main component are supplied, and the planar doping of the main component element of the doping gas is performed. A planar doping method, wherein the layer is formed on a III-V compound semiconductor layer. 4. The main component element of the doping gas is Si, the group V element of the first gas is As, or a combination of As and P, and the group V element of the second gas is P. The planar doping method according to claim 3, wherein: