JPS6177368A - Field effect transistor - Google Patents

Abstract

PURPOSE:To obtain the title element whose operating layer is an InGaAs layer having the good microwave characteristic of small gate leakage current and small resistance, by providing an InGaAs layer in high electron density and an InP layer in low electron density, and the former is made as the main current path. CONSTITUTION:The electron density of the InGaAs layer 2 in lattice matching with a semi-insulation InP substrate 101 is 2X10<17>cm-<3>, and that of the InP layer 3 is 1X10<14>cm<-3>. Further, a gate electrode 4g forming the Schottky junction with the InP layer 2 is made of gold. Next, a source electrode 4s and a drain electrode 4 make ohmic contacts with the InGaAs layer 2 each, and is made of AuGe/Ni. This layer 2 itself is used as the electron transit region 5. The operation principle of this element is basically equal to that of the present GaAs MESFET except that the gate has the hetero junction of InGaAs/InP.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to the structure of a field effect transistor (hereinafter referred to as FET) having an InGaAs layer as an active layer.

[Technical background of the invention and its problems]

In I-x Ga, A lattice matched to InP substrate
s (x = 0.47, hereinafter simply referred to as InGaAs)
GaAs is currently the mainstream material for microwave devices because its electron mobility is approximately twice as high as that of GaAs at room temperature, and its contact resistance is low when forming ohmic contacts.
Schottky junction field effect transistor using As (
It is expected that the transfer conductance will be larger and the source resistance (sum of contact resistance and channel resistance) will be lower than that of GaAsM
It is attracting attention as a material for high-performance microwave devices that exceeds ES FETs.

However, InGaAs has a bandgap of 0.7
8 eV, which is smaller than GaAs, so InGaAs
Even when attempting to form a Schottky junction on the surface, a large leakage current flows in the reverse direction, making it impossible to construct a Schottky junction FET using InGaAs. To solve this problem, attempts have been made to change the gate from a Schottky junction type to an MIS (metal-insulator-semiconductor) type, or to a pn junction type. The transfer conductance is small due to the effect of the generated interface states. In addition, due to reasons such as insufficient reduction of gate leakage current in the latter,
No matter which method is used, an element with small gate leakage current and good microwave performance has not been obtained.

More recently, a technique called "r-modulation doping" has been developed, in which the mobility of electrons accumulated at the interface of the GaAlAs/GaAs heterojunction is dramatically increased compared to that of the bulk GaAs crystal, and this principle has been applied to InGaAs.
/InP systems are also being considered to improve microwave performance by increasing mobility and reduce leakage current at the gate. Below, as a conventional example, InGaAs/I
An nP-based device using modulation doping will be described.

FIG. 3 is a cross-sectional view of a modulation doped FET using an InGaAs/InP junction, and FIG. 3 shows an energy band diagram in the depth direction of the gate portion in a thermal equilibrium state. In both figures, (101) is a semi-insulating InP substrate, (10
2) is an undoped InGaAs layer (n!lXl0”C
m-'), (103) is an n''InP layer (n=5 x 1
0"cm-3), (104g) is the n"InP layer (1
03) and a gate electrode forming a Schottky junction, (104s
) and (104d) are a source electrode and a drain electrode, respectively, which make ohmic contact with the n''InP layer (103).

In the InGaAs/TnP junction, InGaAs
Since the electron affinity of InP is about 0.2 eV larger than that of InP, if the electron density of both materials is the above-mentioned value, that is, the electron density of the InP layer is sufficiently higher than that of the InGaAs layer, the fourth As shown at the bonding interface in the figure, the electron storage layer (1
05) is formed. These accumulated electrons act as ionized impurity scatterers and are spatially separated from the donor ions in InP, so they move through the electron storage layer (105).
In Figure 4, the mobility of electrons traveling in the direction perpendicular to the paper (from left to right in Figure 3) is approximately 7500~ at room temperature.
It reaches 80,000m2/v-5ec. The modulation doped FET shown in FIG. 3 also has the following problems.

(i) Even if an attempt is made to reduce the gate leakage current by forming a Schottky junction on the n+InP layer (103), since the electron density of this InP layer is high, it is not easy to reduce the leakage current, and the improvement effect is small.

(it) In the InGaAs/InP system, the degree of discontinuity at the lower end of the conductor (indicated by ΔEc in Figure 4) is Ga 1
-xAl xAslGaAs system (9 for x = 0.3
Therefore, the surface density of electrons accumulated in the storage layer (105) is at most 5 x 10"cm-2, for example, the surface density of the active layer of a recessed GaAs MES FET is 5.
This is about an order of magnitude smaller than X 10"cm-2. Therefore, the channel resistance (source resistance) of the device increases, and the microwave performance does not improve.

(City) In order to form a modulation doped structure, that is, to form a band structure as shown in FIG.

Normally, the electron density of InGaAs layer is 1×1015CI11
Must be -3 or less. However, InGaAs
In such mixed crystals, it is not easy to reduce the amount of impurities in the crystal, and there are many difficulties in crystal growth.

As mentioned above, the modulation doped FET shown in FIG.
However, the improvement effect was insufficient to achieve the originally intended purpose of reducing gate leakage current and source resistance, and therefore microwaves were
It is inferior to ES FET, and does not take advantage of the high mobility of InGaAs.

[Purpose of the invention]

The present invention has been made to solve the above-mentioned problems, and provides an FET having an InGaAs layer as an active layer, which has low gate leakage current, low source resistance, and good microwave characteristics.

[Summary of the invention]

As a result of repeated consideration of the problems of the prior art, the inventor found that I
In order to take advantage of the high mobility of nGaAs and to reduce gate leakage current, the gate part is made of InGaAs/I.
In addition to forming an In-type nP heterojunction, the inventors discovered that by making the electron density of the InGaAs layer high and the electron density of the InP layer sufficiently low, the InGaAs layer itself can be operated mainly as a current path, and the present invention is constructed. did. - That is, the FET according to the present invention is made of semi-insulating InP
a substrate, an InGaAs layer formed on the substrate and lattice matched with the substrate, an InP layer deposited on the InGaAs layer, a gate electrode provided on the InP layer and forming a Schottky junction therewith; InP layer or In
In a field effect transistor having a source electrode and a drain electrode that are in ohmic contact with the gate electrode sandwiched between them on a GaAs layer, an InGaAs layer is provided with a high electron density and an InP layer is provided with a low electron density.
It has a structural feature in which the As layer is the main current path.

[Embodiments of the invention]

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 shows a cross-sectional structure of an element according to the present invention, and FIG. 2 shows an energy band diagram in the depth direction of a gate portion in a state of thermal equilibrium. In the top view, (101) is semi-insulating I
The nP substrate (2) is an InGaAs layer lattice-matched to the above InP substrate (101), and its electron density is 2 x 10"cm.
-”, the thickness is 0.08 μm. Also, (3) is In
The P layer has an electron density of I x 10"cm-" and a thickness of 0.1 μm. Further, (4g) is a gate electrode forming a Schottky junction with the InP layer (2) and is made of gold. Next, (4s) and (4d) are a source electrode and a drain electrode, respectively, which make ohmic contact with the InGaAs layer (2) and are made of AuGe/Ni. Elements with such a structure can be easily manufactured using ordinary epitaxial growth techniques and element/electrode processing techniques.

The difference between the device of the present invention described above and the device described in the conventional example can be clearly distinguished by comparing FIGS. 2 and 4. That is, in the conventional device shown in FIG.
GaAs layer Although the electron storage layer (Fig. 4 (105)) that accumulates at the inP interface is used, in this invention
The As layer (2) itself is used as the electron transit region (5) (Fig. 1,
Figure 2).

The operating principle of the device shown in Figure 1 is based on the current GaAs MES.
Compared to FETs, they are basically the same except that the gate part is an InGaAs/InP heterojunction.

In this invention, the electron density of the InGaAs layer is high, the electron density of the InP layer is low, and the InGaAs layer is used as the main current path. The reason why is as shown in FIG. 2 is as follows.

In general, the necessary conditions for a Peter junction to have the band structure shown in Figure 2 are (a) χ 〉χ (where χ is electron affinity)
InGaAs InP (b) χInGaAs ” EgInGaAs <
χInP ” EgInP (however, Eg is the band gap) (c) φ 〈φ (however, φ is the work function) I
It is given by nGaAs InP. The values of χ and Eg for InGaAs and InP are listed in Table 1 below.

Substituting the values in the upper table of Table 1 into the conditions (a) and (b), In
These conditions are satisfied in the GaAs/InP system.

Regarding the remaining condition (c), φ changes depending on the electron density, but TnGaAs at the electron density of the above example,
Due to InP, φ is 4.59 eV and 4.6] eV, respectively, and the condition (c) is also satisfied, resulting in a band structure as shown in FIG.

〔Effect of the invention〕

When the device according to the present invention is compared with the conventional device, the following advantages are recognized.

(a) In the device according to the present invention, since the electron density of the InP layer forming the Schottky junction is low, a good junction with extremely low gate leakage current can be obtained.

Furthermore, since the capacitance of the gate portion is a series connection of the depletion layer capacitances of the InP layer and the InGaAs layer, the gate capacitance is significantly reduced and the device performance is improved.

(b) Since the electron density of the InGaAs layer, which is the current path, is high, the electron density directly under the source electrode is
X 1011012C or more, which reduces the channel resistance (source resistance). In addition, it is not necessary to make the aAs layer highly pure (low impurity concentration), so difficulties in epitaxial growth can be reduced.

In addition, in the device according to the present invention, since an InGaAs layer with a high electron density is used as an active layer, it is more affected by ionized impurity scattering than the device described in the conventional example shown in FIG. , "modulation doping"
The mobility does not decrease significantly compared to the structure, and for example, an InGaAs layer with an electron density of 2×10″c′3 can obtain a mobility of 7000cII12/v−8ec or more.

As described above, according to the present invention, InGaAs exhibits good microwave performance with low gate leakage current and low source resistance while taking advantage of the high mobility of InGaAs.
A FET using an As layer as an active layer can be provided.

In the above example, the electron density of the InGaAs layer is 2×1017CI11-3, and the electron density of the InP layer is I
Although the case where the
It can be widely applied when the nP layer has a low density,
According to the inventors' experiments, the electron density of the InGaAs layer is about I x 10"cm-" or more, and the electron density of the InP layer is about I x
When the value is 10I1015" or less, the effect of the present invention of reducing gate leakage current and source resistance can be obtained. Also.

In the example, the ohmic junction of the source and drain is I
Although it is formed on the nGaAs layer, it is not limited thereto, and even if it is formed on the InP layer, the advantages of the present invention can be applied as is.

[Brief explanation of the drawing]

FIG. 1 is a sectional view of an FET according to the present invention, and FIG.
FIG. 3 is a sectional view of a conventional FET, and FIG. 4 is an energy band diagram of the device shown in FIG. 3 in the depth direction of the gate portion.

Claims (1)

[Claims] A semi-insulating InP substrate, an InGaAs layer formed on this substrate and lattice matched to this substrate, and this InGaAs layer.
an InP layer deposited on the InP layer; a gate electrode provided on the InP layer and forming a Schottky junction therewith;
In a field effect transistor comprising a source electrode and a drain electrode that are in ohmic contact with the gate electrode sandwiched between them on a P layer or an InGaAs layer, an InGaAs
A field effect transistor characterized in that an S layer has a high electron density and an InP layer has a low electron density, and the InGaAs layer serves as a main current path.