JPS63193546A - Composite semiconductor device - Google Patents

Abstract

PURPOSE:To obtain sufficient current density, by growing a first semiconductor device on a substrate, growing first and second buffer layers, growing a second semiconductor device, and forming the low-resistance connection between the semiconductor devices. CONSTITUTION:A first semiconductor device alpha is provided on an n-type GaAs substrate 9. A first buffer layer 8 (p-type InxGa1-xAs: x=0 1), in which n-type impurities or p-type impurities are added, is laminated on the first semiconductor device alpha. In said impurities, the composition is changed so that the forbidden band width continuously decreases in the direction away from the device alpha. A second buffer layer 7 (n-type InyGa1-yAs: y=1 0) is laminated on the buffer layer 8. Impurities, which have a reverse conductivity type to that of the buffer layer 8, are added to the layer 7. The composition of the impurities is changed so that the forbidden band width continuously becomes large in the direction away from the device alpha. A second semiconductor device beta is provided on the buffer layer 7. Thus the low-resistance connection between the semiconductor devices is formed, and sufficient current density is obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a structure for electrically connecting semiconductor devices in a composite semiconductor device having a plurality of semiconductor devices.

[Conventional technology]

Conventionally, a semiconductor device formed on a semiconductor substrate (hereinafter referred to as a first semiconductor device) and another semiconductor device formed on the same substrate (hereinafter referred to as a second semiconductor device) are used. When connecting through layers of different conductivity types, a method has been used in which the p-type head region and the n-type region are connected using a metal electrode. This method is illustrated in FIG. In fig.

22 is a semiconductor substrate, 1 is a p-type GaAs layer, 2 is an n-type Ga
3.4 is an ohmic metal electrode; 5 is a first semiconductor device; 6 is a second semiconductor device; 23 is a wiring metal. However, as shown in FIG. 3, this method requires a large area for connection, which limits the ability to increase the density of elements.

On the other hand, in order to solve this problem, a method of connecting by a pn junction doped with impurities at a high concentration, a so-called tunnel junction, has also been used. This principle is illustrated using an energy band diagram as shown in FIG. 4, and semiconductor devices are connected to each other by tunneling current flowing as shown by the arrows in FIG. 1. In the figure, 1 is a p-type GaAs layer, 2 is an n-type GaA layer
It is the s layer.

By using this method, a plurality of semiconductor devices can be stacked vertically as shown in FIG. 5, and unnecessary areas can be reduced. In the figure, 5 is the first semiconductor device and 6 is the second semiconductor device.
24 is a tunnel junction.

However, in semiconductor devices that use semiconductors with a large forbidden band width, such as GaAs, there is a limit to reducing the thickness of the tunnel barrier by increasing the dose. However, a connection with sufficient current density could not be obtained.

[Problem that the invention seeks to solve]

As described above, in the prior art, it has not been possible to connect a plurality of semiconductor devices without requiring a large area, with sufficiently low resistance, and with sufficient current density.

The purpose of the present invention is to provide a technology for connecting semiconductor devices with low resistance and high current density without increasing the area, which is essential for increasing the speed and performance of various composite semiconductor devices. It's about doing.

[Means for solving problems]

In the case of connecting semiconductor devices using a pn junction (tunnel junction), the present invention provides a method for connecting semiconductor devices using a narrow band via an intermediate layer (buffer N) whose forbidden band width is gradually changed by gradually changing the composition. The gist is to use a semiconductor pn junction having a forbidden band width. That is, the composite semiconductor device of the present invention has a composition change on a first semiconductor device provided on a semiconductor substrate such that the forbidden band width decreases continuously in the direction away from the first semiconductor device. A first buffer layer doped with n-type impurities or doped with n-type impurities is laminated, and a forbidden band width is continuous in a direction away from the first semiconductor device on the first buffer layer. A second buffer layer is laminated with impurities having a conductivity type opposite to that of the first buffer layer, and a second semiconductor device is provided on the second buffer layer. It is characterized by

[Effect]

The energy band diagram of the configuration of the present invention is shown in FIG. In other words, since the tunnel current in a tunnel junction increases exponentially as the forbidden band width decreases, by using the present invention, it is possible to form connections between semiconductor devices with extremely low resistance compared to conventional methods. ,
Sufficient current density can be obtained. Furthermore, since the connection method according to the present invention stacks semiconductor devices vertically, there is no increase in area.

〔Example〕

Example 1 FIG. 1 is a schematic cross-sectional view of a first example of the present invention in which a GaAs solar cell and an M (1,3 Gao, 7 As solar cell) were fabricated on a GaAs substrate.

In the figure, 9 is an n-type GaAs substrate, and 2 is an n-type GaAs substrate.
Layer 1 is P-type GaAsJl, and 2 and 1 constitute the first semiconductor device, that is, a GaAs solar cell 6
8 is the first buffer layer, p-type InxGat-xA
It is composed of an S layer (X is changed from X=0 to 1 from the interface with the p-type GaAs layer 1 to the interface with the second buffer layer 7). 7 is a second buffer layer, n-type I nyGa, -yAsM (n-type An 0 , 3 of the second semiconductor device from the interface with the first buffer layer 8);
G a O,? A y towards the interface with 8 layers 1o
; Change from 1 to 0. ). 10 is n-type AQ. ,, Ga0. t As layer, 11 is p-type Al
l. ,,Ga0. , As layers 10 and 11 constitute a second semiconductor device, that is, an An Ga As solar cell. 13 is an AuGe/Ni electrode, 12
is an AuZn/Ni electrode.

First, a QaAspn junction (first semiconductor device) is grown on an n-type GaAs substrate 9 using an epitaxial growth method, and then a first buffer layer 8 and a second buffer layer 7 are grown.
, and then An. ,.

Ga0. , an As p n junction (second semiconductor device) is grown. Finally, AuGe/Ni electrode 13 and Au
A Zn/Ni electrode 12 is formed to complete the composite solar cell.

Table 1 below shows the dopant type, doping amount, and film thickness of each layer of the composite semiconductor device of this example shown in FIG.

Table 1 Table 2 below shows that in the device of this embodiment, the bonding surface between the first buffer layer 8 and the second buffer layer 7 is Ga
This is a table showing that the use of InAs according to the present example improves the three points of forbidden band width, current density, and resistance value compared to the case of As.

The margin table below shows that the above resistance value corresponds to the p-n junction area.
This is the case.

In this example, as described above, a GaAs solar cell and an Al
l. , , Ga, , As solar cells, the forbidden band width is the first
a P-type first buffer layer whose composition changes continuously in a direction away from the first semiconductor device; and a composition whose forbidden band width continuously increases in a direction away from the first semiconductor device. The energy band diagram of the connected portion is as follows because the connection is made with the n-type second buffer M intervening.

As shown in Figure 6, the tunnel current in the tunnel junction increases exponentially as the forbidden band width decreases, so multiple solar cells can be connected with low resistance and high current density, resulting in conversion efficiency. It is possible to manufacture a broadside GaAs/GaAs solar cell that has a high current density and can obtain a sufficiently high current density.

Embodiment 2 FIG. 2 is a schematic sectional view of a second embodiment of the present invention in which an InP/InGaAsP semiconductor laser and an Ink/InGaAsP HBT (peterojunction bipolar transistor) are fabricated on an InP substrate.

In the figure, 14 is an n-type InP substrate, 15 is InP/I
nGaAsP buried stripe laser (first semiconductor device), 16 is a first buffer layer, p-type I n
G a 1- x A sy Pl-v layer (x = 140.
53,y;O-) Change as in 1. ). 17 is the second buffer layer, which is n-type InuG
a1-uAsvPl-v layer (u=0.53-+1.v=1→o in the direction away from the first semiconductor device)
Change it like this. ). 18 is n-type In
A P buffer layer, 19 an n-type InP collector layer, 20 a p-type InGaAsP base layer, 21 an n-type InP emitter layer, and 18.19° and 20.21 form the second semiconductor device, that is, InP/InGaAsP HB.
T is constructed. 13 is AuGe/Ni electrode, 12 is A
It is a uZn/Ni electrode.

First, a normal InP/InGa film is placed on the n-type 1nP substrate 14.
The first step is to fabricate an AsP embedded stripe laser.
n, , Ga0. ,, p-type first buffer layer 16 gradually changed to As, I no, , Ga0
, 4°. An n-type second buffer layer 17 whose composition is gradually changed from As to InP is grown, and each layer of the HBT, which is a second semiconductor device, is grown on this second buffer layer 17. Finally, A
A uGe/Ni electrode 13 and an AuZn/Ni electrode 12 are formed to complete a laser diode/HBT composite semiconductor device.

Table 3 below shows the dopant species and doping amounts in each layer of the composite semiconductor device of this example shown in FIG.

It is a table showing film thickness.

Margin below Table 3 Note that Table 4 below is for the apparatus of this example.

Compared to the case where the bonding surface between the first buffer layer 16 and the second buffer layer 17 is InP, Ina, 53 according to the present invention
This is a table showing that the three points of forbidden band width, current density, and resistance value are improved by using Gao and ntA8.

Table 4 However, for the above resistance value, the pn junction area is IIIfi12
This is the case.

The methods for epitaxially growing each layer of the semiconductor structure shown in the first and second embodiments include halogen transport vapor phase epitaxy, liquid phase epitaxy, molecular beam epitaxial growth, organometallic pyrolysis, etc. You can also use the method
The most suitable method can be selected from these methods depending on the type and properties of the semiconductor to be grown.

Further, in the first and second embodiments described above, the same effect can be obtained even if the first and second semiconductor devices and the first and second buffer layers are not lattice matched. Furthermore, it goes without saying that each semiconductor layer including the semiconductor substrate may contain a necessary dopant. Furthermore, it goes without saying that the thickness of each layer, the type of electrode metal, the structure of each device, etc. can be changed in various ways depending on the overall structure and purpose.

〔Effect of the invention〕

As explained above, according to the present invention, AnGaAs/Ga has high conversion efficiency and a sufficiently high current density.
Solar cells such as As can be manufactured. Furthermore, even in a so-called 0EIC (optoelectronic integrated circuit) that combines a laser and a transistor, connection resistance can be reduced and high current density can be obtained. Furthermore, p-type semiconductor and n
Since metal electrodes are not required to connect the semiconductor regions, the overall area of the device can be reduced.

As described above, according to the present invention, multiple semiconductor devices can be connected with extremely low resistance, and therefore, they can be integrated without deteriorating the characteristics of each element, making it possible to realize a multifunctional semiconductor device. Become.

[Brief explanation of the drawing]

FIG. 1 is a cross-sectional view of a composite semiconductor device according to a first embodiment of the present invention, FIG. 2 is a cross-sectional view of a composite semiconductor device according to a second embodiment of the present invention, and FIG. 3 is a conventional semiconductor device connection. 4 is an energy band diagram of a GaAspn junction, FIG. 5 is a sectional view of a pn junction type semiconductor device connection, and FIG. 6 is an Inx
FIG. 2 is an energy band diagram of a Ga1-xAs p n junction. 1...p-type GaAs layer 2-n-type GaAs layer 3.4...ohmic metal electrode 5...first semiconductor device 6...second semiconductor device 7...n-type InyGa, -yAs layer 8-p type nX Gai -X A 8 layer 9...n
Type GaAs substrate 10・=n type A11, 10a,... As layer 11・p type An6, 30 ao, 7 A 8 layer 12--・Au
Zn/Ni electrode 13...AuGe/'Ni electrode 14...n-type InP substrate 15...InP/InGaAsP embedded stripe laser 16-P-type InXGa, -, AsyPl-y layer 17
・=n-type I nuGal-uAsv P 1-v layer 18
...N-type InP buffer layer 19...N-type InP collector layer 20-P-type InGaAsP base layer 21...N-type InP emitter layer 22...Semiconductor substrate 23...Wiring metal 24... Tunnel junction patent applicant Junnosuke Nakamura, patent attorney representing Nippon Telegraph and Telephone Corporation

Claims (1)

[Claims] 1. On a first semiconductor device provided on a semiconductor substrate, a composition change is made in which the forbidden band width becomes smaller continuously in a direction away from the first semiconductor device. A first buffer layer doped with an n-type impurity or a p-type impurity is stacked on the first buffer layer, and a forbidden band width is continuously increased in a direction away from the first semiconductor device. A second buffer layer is laminated with an impurity having a conductivity type opposite to that of the first buffer layer, and a second semiconductor device is provided on the second buffer layer. A composite semiconductor device characterized by: 2. The first buffer layer is In_xGa_1_-_x
As, x is continuously changed from 0 to 1 in the direction away from the first semiconductor device, and
The buffer layer is made of In_yGa_1_-_yAs, and has a composition in which y is continuously changed from 1 to 0 in the direction away from the first semiconductor device. Composite semiconductor device. 3. The first buffer layer is In_xGa_1_-_x
As_yP_1_-_y, the composition is such that x is continuously changed from 1 to 0.53 and y is continuously changed from 0 to 1 in the direction away from the first semiconductor device, and The second buffer layer is In_uGa_1
____uAs_vP_1_-_v is a composition in which u is continuously changed from 0.53 to 1 and ri is continuously changed from 1 to 0 in the direction away from the first semiconductor device. A composite semiconductor device according to claim 1 characterized by: