JPS593872B2 - Method for manufacturing semiconductor light emitting device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for manufacturing a semiconductor light emitting device having a double heterojunction stripe structure.

Optical fibers used for optical communications etc. exhibit low transmission loss characteristics in wavelength bands of 0.85 μm, 1.05 μm, 1.3 μm, 1.6 μm, etc. Therefore, a semiconductor light emitting device having a double heterojunction stripe structure using GaAlAs, 3 parts nGaAs, GaAsSb, 3 parts nGaAsP, etc., has been proposed as a light source for generating an optical signal. Such stripe structures include a stripe electrode type, a hetero-isolation type, a planar stripe type by diffusion of Zn, etc., a mesa stripe type by mesa etching, an internal stripe type with a current confinement formed inside, etc. In both cases, the current injection region is the light emitting region. However, it is not easy to stabilize the transverse mode into a single fundamental mode, and the linearity of the 15 relationship between drive current and optical output is not sufficient. An object of the present invention is to provide a method for unifying and stabilizing the transverse mode of a semiconductor light emitting device having a double heterojunction stripe structure, improving the light output characteristics, and making it possible to easily manufacture the device through a process. There are 0 things to do.

Examples will be described in detail below. FIG. 1 is an explanatory diagram of a loss waveguide type semiconductor light emitting device, in which 1 is an n-InP substrate, 2 is an n-InP or n-In substrate with a thickness of 1 μm or more.
A cladding layer of GaAsP, 3 is an active layer of InGaAsP with a thickness of 0.1 to 0.3 μm, 254 is a cladding layer of p-InP or p-InGaAsP with a thickness of 0.3 μm, 5
is a lossy layer of InGaAsP with a thickness similar to O, 6n, region 5a is p-type, region 5b is n-type or high resistance current blocking region, and 6 is a waveguide layer of p-InP or p-part nGaAsP.
7|8 is an electrode. 30 Since the cladding layer 4 on the active layer 3 is thin, light loss occurs due to the loss layer 5, but since the waveguide layer 6 is provided, the light incident on the waveguide layer 6 does not suffer loss. propagate.

Therefore, a light emitting region with a width w is generated directly under the waveguide layer 6, and since a part of the region 355a of the loss layer 5 is of the p-type, a current flows in a range of the width s. Therefore, hole burning occurs in the injected carrier distribution at the center due to stimulated emission, and as a result, the equivalent refractive index at the center increases, making it possible to utilize the focusing effect due to the carrier distribution and increasing the luminous efficiency. can.

Furthermore, the waveguide layer 6 allows the transverse mode to be made into a stable single fundamental mode. 2A to 2D are process explanatory diagrams of an embodiment of the present invention, and as shown in FIG. 2A, an n-1nP substrate 1
A cladding layer 2 of n-1nP or n-1nGaAsP, an active layer 3 of NGaAsP, a cladding layer 4 of p-1nP or p-1nGaAsP, and a loss layer 5 of p-1nGaAsP are formed thereon by liquid phase epitaxial growth or the like. .

The thickness of each layer 2 to 5 is 1 .mu.m or more for the cladding layer 2, 0.1 to 0.3 .mu.m for the active layer 3, 0.3 .mu.m for the cladding layer 4, and 0.6 .mu.m for the loss layer. Next, as shown in FIG. 5B, a groove with a width w is formed in the loss layer 5 by selective etching.

Next, as shown in FIG. 3c, a waveguide layer 6 is formed on the entire surface.
Next, as shown in FIG. 4D, a mask is provided in a portion having a width S, and plot implantation is performed.

The depth is set near the boundary between the loss layer 5 and the cladding layer 4, so that no current flows between the electrodes 7 and 8 in a region 9 shaded with dotted lines. Then, electrodes 7 and 8 are formed to complete the process. Note that the electrode 8 can also be formed on the substrate 1 before the growth of each layer 2 to 5. The current between the electrodes 7 and 8 flows within a range of width S, but since the waveguide layer 6 has a width w and is formed on the cladding layer 4, the width of the light emitting region is w, which is smaller than the width S of the current injection region. It becomes narrow.

Note that when the waveguide layer 6 is made of p-InP, in order to obtain ohmic contact with the electrode 7, p-1nG is applied on the waveguide layer 6.
It is preferred to provide the electrode 7 after forming the layer of aAsP. In addition, the composition of each layer is such that the relationship between the band gap energy Eg of each layer, where the code of each layer is shown in (d), is Eg (3K
Eg(2), Eg(3KEg(4), Eg(4)>Eg
(5), Eg(5)〉Eg(6), Eg(4)
6). Next, a specific example will be explained.

n-1nP(
The cladding layer 2 of Snl-L7') was 8 μM. InO.
76GaO. 24ASO. 5. P.O. 45 active layer 3 with a thickness of 0.2 μTn. A p-1nP (Zn-doped) cladding layer 4 was formed with a thickness of 0.3μWL. p-1n0.76Ga0.24'F AsO. 55 pO. 45 (Zn doped) loss layer 5
The loss layer 5 was sequentially formed to a thickness of .mu.m, and then the loss layer 5 was selectively etched to form a groove having a width W of 7 .mu.m.

Next, the liquid phase epitaxial growth method is used to grow at a growth temperature of 580~
A p-1nP (Zn-doped) waveguide layer 6 is grown on the cladding layer 4 to a thickness of 1.2 μm at 600°C, and then a p-N
O. 76GaO. 24AsO. 55 pO. A 0.45 (Zn-doped) intact layer was grown to a thickness of 0.5 μm. Next, a proton implantation was performed using an Au mask having a width S=10 μMf. This proton injection is performed at an acceleration voltage of 200
KeV, and the dose was 2×10 15 tons. Next, Au
-An electrode 7 of Zn was formed. As shown in FIG. 3, such a loss waveguide type semiconductor light emitting device has a threshold current of 80 m
A, the linearity was good up to 1.5 times that, and the transverse mode was a single fundamental mode.

Note that the emission wavelength was 1.27 μm. As explained above, in the process of manufacturing a semiconductor light-emitting device with a double heterojunction stripe structure, the present invention forms a groove having a width corresponding to the width of the light-emitting region in the loss layer 5, and then conducts the entire surface. wave layer 6
By forming a mask with a width corresponding to the width of the current injection region and performing proton implantation, it is possible to form a light emitting region with a width narrower than the width of the current injection region. Since the cladding layer 4 on the optical waveguide 3 is made thin, the transverse mode can be unified and stabilized by the loss layer 5 and the waveguide layer 6.

Also, if the width of the light emitting region is made larger than the width of the current injection region,
Although it is effective in reducing the threshold current, it has the disadvantage that it tends to cause curved optical output characteristics, so-called kick. There is an advantage that the mode is stabilized and the linearity of the optical output characteristics is improved.

[Brief explanation of drawings]

FIG. 1 is an explanatory diagram of a loss-guided semiconductor light emitting device, FIGS. 2 a to d are process explanatory diagrams of an embodiment of the present invention, and FIG. 3 is a semiconductor light emitting device manufactured according to an embodiment of the present invention. FIG. 3 is a current-light output characteristic curve diagram of FIG. 1 is a substrate, 2 and 4 are cladding layers, 3 is an active layer, 5 is a loss layer, 6 is a waveguide layer, 7 and 8 are electrodes, and 9 is a proton injection region.

Claims (1)

[Claims] 1. In a method for manufacturing a semiconductor light emitting device with a double heterojunction stripe structure, a cladding layer, an active layer, a cladding layer thinner than the cladding layer, and a loss layer are formed on a substrate, and the loss layer has a width of a light emitting region. After forming a groove with a corresponding width and forming a waveguide layer on the entire surface, a mask with a width corresponding to the width of the current injection region is applied and proton injection is performed to form a light emitting region with a width narrower than the width of the current injection region. 1. A method for manufacturing a semiconductor light emitting device, the method comprising the step of forming a semiconductor light emitting device.