JPS6134269B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a - group compound semiconductor light emitting device and a method of manufacturing the same, and more particularly to a - group compound semiconductor light emitting device with an improved electrode structure and a method of manufacturing the same.

Gallium phosphide (GaP), a typical compound semiconductor, is often used as a light-emitting element (light-emitting diode). GaP light emitting diodes are attracting attention because they can freely emit red to green light by adding impurities. For example, this configuration is as shown in FIG. That is, n
A p-type GaP layer 12 is provided on a p-type GaP substrate 11.
Electrodes 11 are placed on both sides of the pellet forming the n-junction 13.
a, 12a are provided and mounted on a header 14 with the n-type GaP substrate 11 side facing down via a conductive paste 15. Note that terminals 11c and 12c for taking out electrodes are taken out from the header 14, and a lead wire 12b is connected from the electrode 12a of the p-type GaP layer.
It is attached to the terminal 12c by the terminal 12c. Naturally, the electrode 11a of the n-type GaP substrate and the terminal 12c are electrically connected. Further, the upper part of the header 14 is covered with an epoxy resin 16.

By the way, in the configuration shown in Figure 1, electrode 1 to the p-type GaP layer
2a is in ohmic contact with the p-type GaP layer,
The lower the contact resistance, the better. Lead wire 12b
can be easily bonded. Luminous efficiency must not be reduced. p- to improve luminous efficiency
It must withstand the strong acid etching solution used in the treatment of the 13 n-junction surfaces. Performance such as the ability to perform microfabrication and etching of electrodes is required.

No electrode has been found that satisfies all of these conditions, and the reality is that electrodes are manufactured that are unsatisfactory with one or two of the above conditions. In other words, as an electrode that satisfies
Alternatively, a structure in which a Zn alloy layer is provided and an Au layer is provided thereon is known. However, with such an electrode, it is possible to satisfy , but not to satisfy , depending on the conditions. To explain the reason for this,
This will be explained with reference to FIGS. 2a to 2e.

First, an n-type GaP layer 22 and a p-type GaP layer 23 are formed on an n-type GaP substrate 21.
Electrode 24 made of Au-Si or Au-Ge alloy layer on the surface
The above-mentioned Au-Be or Au-
Electrode 2 consisting of Zn alloy layer 25a and Au layer 25b
5 are formed by a vacuum evaporation method (FIG. 2a).

Next, using photoresist (not shown), the double-sided electrodes 24 and 25 are finely processed by chemical etching. Then, the photoresist is removed and heated at 500°C for about 10 minutes in an inert gas atmosphere. With such processing, each electrode 24,
25 makes ohmic contact with the p-type GaP layer 23 and the n-type GaP substrate 21 (FIG. 2b). Then, it is processed and separated into predetermined dimensions by dicing, scribing, etc. (FIG. 2c). At this time, due to machining, the side surface of the p-n junction, which becomes the light emitting part, is crushed and the light emitting efficiency deteriorates. To restore this, the fractured layer must be removed by etching. Also, in the etching, it is desirable that the entire exposed GaP surface other than the electrodes be finished in a rough surface with irregularities rather than a smooth surface.
An etching solution that satisfies both of these requirements is a heated solution of hydrochloric acid and nitric acid. Through this process, a light emitting device pellet as shown in FIG. 2d is completed. After this, as mentioned above, apply the conductive paste 27 to the TO header 26.
The electrode 24 on the n-type GaP substrate side is fixed with adhesive, and the p-type
A gold wire 28 is bonded to the GaP layer side electrode 25 (FIG. 2e).

Although the above is a normal process, the following drawbacks occur in this case. That is, Au-Be or Au-
The Zn alloy layer 25a becomes in ohmic contact with the heat treatment, but on the contrary, the heat treatment causes Ga and P ions (particularly Ga ions) to form on the surface of the Au layer 25b.
The Be or Zn element of the Au-Be or Au-Zn alloy is also deposited through the Be or Au-Zn alloy layer 25a, and the deposited ions or elements are oxidized to form a complex oxide film such as Ga. An oxide film such as -P-Be-O or Ga-P-Zn-O is formed. As a result, the adhesion of bonding becomes extremely poor. There are times when pellets that cannot be bonded are generated, but normally bonding is not successful after one time, and bonding becomes possible only after 4 or 5 bonding operations. For example, it is possible to etch the oxide film formed on the surface and then perform bonding.
Etching is difficult because the oxide film formed on the surface is a complex oxide film. In addition, as mentioned above
Since the Ga element diffuses into the Au layer 25b (because the diffusion coefficient of Ga is large), a large amount is deposited on the electrode surface. Along with this, the amount of Ga in the GaP crystal itself is drastically reduced, the crystallinity is impaired, and the luminous efficiency is lower than the theoretical value.

Note that the alloy film mainly composed of Au is resistant to the strong acid used to provide the frost effect.

SUMMARY OF THE INVENTION In view of the above-mentioned problems, the present invention provides a - group compound semiconductor light emitting device in which the electrode structure of the p-type layer is particularly improved, and a method for manufacturing the same.

That is, the present invention is characterized in that the electrode of the p-type compound semiconductor layer has a three-layer structure, and a tantalum layer containing carbon is provided in the middle layer.

An embodiment of the present invention will be described below with reference to FIGS. 3a to 3f. FIGS. 3a to 3f show an example in which GaP crystal is used as the - group compound semiconductor, and is an example of a manufacturing method. First, as in the conventional case, the n-type GaP substrate 31
On top is an n-type GaP layer 32 and zinc (Zn) and oxygen (O).
A p-type GaP layer 33 containing the above is formed by, for example, a liquid phase epitaxial growth method. This formed n-type GaP
The donor concentration (N _D ) of layer 32 is between 2 and 10×10 ¹⁷ /cm ³
acceptor concentration (N _A ) of the p-type GaP layer 33
is about 1 to 5×10 ¹⁷ /cm ³ . After this, n-type GaP
An electrode 34 made of, for example, an Au-Ge alloy layer is formed on the substrate 31 side by vapor deposition (FIG. 3a). Then P
Zn with a thickness of about 0.05 to 0.1 μm is deposited on the GaP layer 33.
Au-Zn alloy layer 35a containing % by weight, thickness 0.2-0.4
Ta layer 39 containing μm of carbon (C) and thickness 0.2~
An electrode 35 made of an Au layer 35b having a thickness of about 1 μm is formed. Here, Ta containing carbon, which is a feature of the present invention,
The layer 39 is formed by vacuum evaporation, and to explain it specifically, first, the wafer with the Au-Zn alloy layer 35a deposited on it is placed in a evaporation apparatus, and the vacuum level is
Evacuate until the temperature reaches about 10 ^-7 to ^{10 -8} Torr, then evacuate methane gas, acetylene gas, or carbon dioxide gas to the vacuum level.
Insert until 10 ^-4 to ^{10 -6} Torr is reached, and in this state Ta
is deposited to form a Ta layer containing C. In addition
The formation of the Au-Zn alloy layer 35a and the Au layer 35b is as follows:
Perform normal vacuum deposition (vacuum level 10 ^-6 to 10 ^-8 Torr). Thereafter, heat treatment is performed at a temperature of, for example, 500° C. for about 10 minutes so that the respective electrodes 34 and 35 make ohmic contact with the n-type GaP substrate 31 and the p-type GaP layer 33 (FIG. 3b). Next, Au layer 35b, Ta containing C
The layer 39 and the Au--Zn alloy layer 35a are sequentially selectively etched to form the result as shown in FIG. 3c. When performing selective etching in this manner, the etching of the Au layer 35b is performed by selectively forming a resist (not shown), for example.
Using this as a mask, a mixed solution of iodine (I ₂ ) and potassium iodide (KI) is used to etch the Ta layer 39 containing C. Using the Au layer 35b as a mask, etching is performed using an alkaline solution such as caustic soda (NaOH) and potassium iodide (KOH). ) with a 9:1 mixture,
Moreover, the etching of the Au-Zn alloy layer 35a is
Etching is performed using the same etching solution as the one used for the Au layer, using the Ta layer 39 containing the above as a mask. Further, since this etching step is performed after the heat treatment step, there is no problem that the C-containing Ta layer 39 and the Au--Zn alloy layer 35a separate during the etching step. The subsequent steps are similar to the conventional steps, and the wafer shown in FIG. 3c is processed and separated into predetermined dimensions by dicing, scribing, etc. (FIG. 3d). At this time, the side surface of the p-n junction, which becomes the light emitting part, is crushed due to machining, and the light emitting efficiency deteriorates. Therefore, as before, p
The side surface of the -n junction is etched with a heated mixture of hydrochloric acid and nitric acid to form a rough surface with irregularities as shown in FIG. 3e. If the side surface of the p-n junction is roughened in this way,
The light emitted at the pn junction surface becomes easier to emit, and the luminous efficiency inevitably improves. After this,
n-type GaP on TO header 36 with conductive paste 37
The electrode 34 on the substrate side is fixed with adhesive, and the gold wire 3
8 to the electrode 35 on the p-type GaP layer side,
A GaP red light emitting device as shown in FIG. 3f is obtained.

According to the method of the embodiment described above, p-type GaP
Bonding of the gold wire 38 to the layer-side electrode 35 becomes easy, and a luminous efficiency close to the theoretical value can be obtained. This is electrode 3 on the p-type GaP layer side.
This is because 5 has a three-layer structure, and the Ta layer 39 containing C is provided in the middle layer.

That is, the Ta layer 39 containing C contains Au to prevent the movement of ions (or elements) such as Zn, Ga, and P during heat treatment to obtain ohmic contact.
These ions are prevented from being deposited on the surface of the layer 35b. Therefore, no complicated oxide film or the like is formed on the surface of the Au layer 35b, and the Au wire 38 can be easily bonded onto the Au layer 35b. Note that the Ta layer 39 containing C is Ta
Since this layer has a greater effect of inhibiting the movement of ions such as Zn, Ga, P, etc. during heat treatment than a single layer, the above-mentioned effects appear even more prominently. The reason for this is that when Ta contains C, it becomes a denser layer than a single Ta layer, and fine ions (or elements)
This seems to be because it no longer reaches the surface of the Au layer. When the Au layer 35b was actually analyzed using an ion microanalyzer (IMA), Zn, Ga, and P
There were almost no such things. In addition, when a layer containing Ta and C was similarly analyzed by IMA and the content of C (atomic percent = at%) was measured, it was found that the amount was relatively small (0.5 to 0.2 at%). .

Furthermore, since the Ta layer 39 containing C prevents the movement of Ga or P (particularly Ga) ions as described above, Ga deficiency, that is, deterioration of crystallinity, is less likely to occur on the surface of the p-type GaP layer. Therefore, a luminous efficiency close to the theoretical value can be obtained. At the same time, even if the heat treatment temperature for obtaining ohmic contact varies somewhat from lot to lot, the variation in luminous efficiency of light emitting element pellets from lot to lot is reduced, resulting in improved yield.

Although the above embodiments have been described with reference to a GaP light emitting device, it goes without saying that the present invention is not limited to this GaP light emitting device, but can be applied to - group compound semiconductor light emitting devices such as a GaAs light emitting device.

[Brief explanation of the drawing]

Fig. 1 is a cross-sectional view of the configuration of a commonly used GaP light emitting diode, Fig. 2 a to e are process cross-sectional views showing the manufacturing process of a conventional GaP light emitting element, and Fig. 3 a to f are an embodiment of the present invention. FIG. 3 is a process cross-sectional view showing the manufacturing process of the GaP light emitting device. 31... n-type GaP substrate, 32... n-type GaP layer,
34...N-type GaP substrate side electrode, 35...P-type
Electrode on the GaP layer side, 35a...Au-Zn alloy layer, 3
5b...Au layer, 39...Ta layer containing C.

Claims (1)

[Scope of Claims] 1. In a - group compound semiconductor light emitting device having a p-n junction, the electrode of the p-type layer constituting the p-n junction may be made of a beryllium or zinc alloy layer mainly composed of gold, a carbon 1. A - group compound semiconductor light-emitting device, characterized in that it is constructed by sequentially laminating a tantalum layer containing a metal layer containing gold or aluminum, and a metal layer consisting of gold or aluminum. 2 p-type on the n-type compound semiconductor substrate
a step of forming a p-n junction by forming a p-type group compound semiconductor layer; a step of forming a beryllium or zinc alloy layer containing gold as a main component on the p-type group compound semiconductor layer by vapor deposition; Vacuum degree on the layer
The method comprises the steps of forming a tantalum layer containing carbon by vapor deposition at a 10 ^-4 to 10 ^-6 Torr level, and forming a gold layer or an aluminum layer on the tantalum layer containing carbon by vapor deposition. Features-
A method for manufacturing a group compound semiconductor light emitting device.