JPH01241156A - Semiconductor device - Google Patents

Abstract

PURPOSE:To make a base layer thin, to realize a high frequency as well as low base resistance and to form an FET active layer having high breakdown strength and high performance, by making the base layer have a structure where a delta-doped layer is inserted between upper and lower regions containing an extremely small amount of impurities. CONSTITUTION:A base layer consisting of n-type GaAs layers 14' and 14 and a delta-doped n<++> type GaAs layer 15 is formed so that it faces to a p-type AlXGa1-XAs emitter layer 16. Further, p-type GaAs and p<+> type GaAs collector layers 13 and 12 are formed and then, an emitter electrode 20, base electrodes 21 and 21', and collector electrodes 22 are formed. On the other hand, the delta-doped layer 15 which is inserted between thin layers 14 and 14' having low concentrated impurities is formed even as the active layer of an FET. As far as the base layer of a bipolar transistor is concerned, it is extremely thin and its resistance becomes low in this way. Thus, this device reduces base resistance and achieves a high frequency. Moreover, it is formed as the active layer of the FET so that it may have high breakdown strength and high efficiency.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to an ultra-high speed, ultra-high frequency semiconductor device, and in particular to a semiconductor device suitable for configuring a field effect transistor and a bipolar transistor on the same substrate. Regarding.

[Prior art] Recently, a specific atom PyJ in gallium arsenide (GaAs) has been studied.
(or a layer similar to an atomic layer) contains impurities (n-type or p-type).
type) doping technique (δ doping: for example, Japa type).
nese Journal of Applied P
hysics Volume 24 Nn8 19135 Page L60
8 to L610) was developed and applied to the production of field effect transistors.

[Problem to be solved by the invention]

The present invention provides 1) forming a base layer of a bipolar transistor using such δ-doping technology, and 2) δ-doping.
- The purpose is to form a bipolar transistor and a field-effect transistor in the same substrate, by using the doped region on the one hand in the base region of the bipolar transistor and on the other hand in the active layer of the field-effect transistor.

The purpose of making the base layer of a bipolar transistor using δ-doping technology is to make the base layer as thin as possible. Further, the purpose of using it in the active layer of a FET (field effect transistor) is to improve the performance of the FET by making the active layer thinner.

[Means to solve the problem]

Means for realizing the present invention is GaAs /AQGaA
Hetero-junction Biro pnp type hetero bipolar transistor with s heterojunction
This will be explained using an example of a lar transistor (abbreviated as HBT) (cross-sectional views thereof are shown in FIGS. 1(a) and 1(b)).

FIGS. 1(a) and 1(b) show cross-sectional views of the semiconductor device.

For p-type AQxGat-xAs16 emitter layer, n-
A base layer consisting of GaAs layers 14' and 14 (\ is undoped) and a δ-doped n++ GaAs layer 15 is formed, and p-GaAs 13. A p+GaAs 12 collector layer is formed, with an emitter electrode 20 and a base electrode 21, respectively.
．． 21' Collector electrode 22 temporary δ-doped layer 15 is made of FE
It can also be formed as an active layer. That is, as shown in the cross-sectional view in FIG. 1(b), the δ-doped layer 15 can be used on one side as an active layer of an FET and on the other as a base layer of a bipolar transistor.

That is, the bipolar transistor is formed in the same manner as in FIG. 1(a). On the one hand, part 14.15 of the base layer
．． 14' can be used as an n-channel FET by forming a source ff electrode 21'', a drain electrode 21'' and a shot key gate electrode 23, as shown in the FET part of FIG. 1(b). . At this time, electrode 2
2 is a substrate bias electrode. Furthermore, if two base electrodes 21 are formed independently as shown in FIG.
tion) type n-channel FET.

The operating principle has been explained in the case of the n-channel base layer of a pnp-type HBT.
A similar distinction can be made for the p-channel base layer.

[Effect]

Since the structure is such that the upper and lower regions with very small Y impurities are sandwiched between them, 1) the base layer of a bipolar transistor can be extremely thin and have low resistance, so the base resistance can be reduced and a high fr can be achieved.

2) Since a very thin active layer can be formed as the active layer of the FET, a high-performance FET with high breakdown voltage can be formed.

3) FET and bipolar can be easily formed on the same substrate.

In this way, pnp type HT3 T and n channel GaAsM
By forming an ESFET (or an npn-type HBT and a p-channel GaAs MESFET) on the same substrate, it is possible to realize a circuit format similar to the technical idea disclosed in Japanese Patent Application No. 61-40244, for example.

〔Example〕

The present invention will be explained in more detail below through examples.

Embodiment 1 An embodiment in which the present invention is applied to a pnp type HBT and an n-channel FET will be described with reference to FIGS. 1(a) and 1(b). M
Using OCVD (organometallic pyrolysis), semi-insulating G
Undoped G on the aAs substrate 10 or the Si substrate 10
a After forming the buffer layer 11 of 3000 people,
p+ GaAs 1 containing Mg 2X101θ rho3
After forming p-GaAs 13 containing about 1xto18■-δ of Mg by 3,000 people, a high-purity GaAs layer 14 (undoped layer: impurity concentration is usually less than 1011015a) was formed by 50 people. Furthermore, δ-doping method or AL
Using the E method, Se is formed to form an I
150 people formed 4'.

Next, p+A containing 3 x 1018exa-” Mg
QxGa+-, A s(x~0.3) M2S 200
Furthermore, 3000 people formed a p+GaAs layer 17 containing 2×]019m''3 of Mg.

In this embodiment, the semiconductor layers 13 and 14 are formed separately, but this is not necessarily necessary.

That is, it is sufficient if the low concentration collector layers 13 and 14 are formed in the usual manner. Further, the δ-doped layer may be extended to about 50 people.

A collector electrode 22 was formed (FIG. 1(a)).

If used as a drain, it can also be used as a junction type FET.

Further, an example in which FET/HBT are formed on the same substrate (FIG. 1(b)) will be described next. A bipolar transistor can be formed in the same manner as shown in FIG. 1(a).

FET is p-type GaAs 17 , AlGaA
s16 is selectively removed and the source (21'
) Drain (21″′) electrode Schottky gate electrode 2
3 were each formed in a conventional manner.

The n-type region 12 under the FET region is usually not made floating, but a control electrode 22 is formed in order to fix the potential. Further, the collector layer 12 may be selectively formed in a semi-insulating substrate, and then the semiconductor layer 13 or more layers may be formed (for example, Japanese Patent Laid-Open No. 6O-134479). Adjustment of the threshold voltage Vth as a FET is performed using the G a A 8 layer 14'
This can be easily achieved by controlling the film thickness (etching or changing the film thickness) or by changing the sheet concentration of the δ-doped layer.

For example, as a layer similar to the δ-doped layer in the sense of overlapping the δ-doped layer, the Si impurity concentration is 5×10”a1.
A ``''♂ n+GaAs layer may be formed by 200 PA steps in place of the δ-doped layer of this embodiment.

Example 2 The case of an npn type HBT will be explained.

In FIGS. 1(a) and 1(b), n-type and n-type are interchanged. That is, as 12, Si or Se doped (~5
0”am-8) GaAs layer, Si as 13~101
The δ-doped layer 15 is added to the layer containing approximately BQll'″δ.
The emitter layer 16 is made of an n-type AlGaAs layer containing Si dopants of 4X1017cn''8. Form a As layer.

The emitter, base, collector, gate, source/drain electrodes are also selected in the usual way, with the P-type and N-type electrodes being switched.

δ-doped structures such as
InGaAs, TnP, InAQAs, In
It goes without saying that this can also be realized with AQGaAs, Si, etc.

〔Effect of the invention〕

According to the present invention, the base layer of the HBT is formed as a δ-doped layer, and the active layer of the FET is simultaneously formed as a δ-doped layer, so 1) the base layer can be made thinner and have a higher fr (~80 GHz ductance); Can be done.

3) Since FETs and HBTs can be easily formed on the same substrate, there are effects such as the ability to realize highly integrated, ultra-high-speed LSIs (Large Scale Integrated Circuits) with low power consumption.

[Brief explanation of the drawing]

FIGS. 1(n) and 1(b) are cross-sectional structural diagrams of the semiconductor device of the present invention. 12...High concentration collector layer, 13...Low concentration collector layer, 14'...High purity layer, 15...δ doped layer (
base), 16... wide gap semiconductor layer (emitter), 20... emitter electrode, 21.21'...
x-xTl! Pole, 21', 21~...source/drain z46.23...gate electrode. 7. ′

Claims (1)

[Claims] 1. Doping impurity atoms in a single plane or in an atomic layer similar to a single plane (several atomic layers to several tens of atomic layers), at least one above and below the atomic layer (I); A semiconductor device characterized in that it operates as a bipolar transistor using a multilayer film having a high purity layer not intentionally doped with impurities as a base layer. 2. In the semiconductor device according to claim 1,
The base layer (I) is also used as an active layer of a field effect transistor, and one or more devices each having a field effect transistor operation and a bipolar type operation are formed in the same substrate. semiconductor device. 3. A semiconductor device according to claim 2, wherein the gate electrode of the field effect transistor is arranged to form a Schottky junction.