JPS6237829B2 - - Google Patents

Description

[Detailed description of the invention] The present invention relates to semiconductor lasers.

A semiconductor laser used as a light source for optical information processing has a double heterostructure shown in FIG.
That is, on an N-type GaAs semiconductor substrate 1, an N-type Ga ₁ - _x Al _x As layer 2 as a first confinement layer, a P-type GaAs layer 3 as an active layer, and a P-type GaAs layer 3 as a second confinement layer are formed on an N-type GaAs semiconductor substrate 1.
A layer 4 of _{Ga 1} - _x Al Laser light 6 is emitted.
In this case, the state of the end face 7, that is, the cleavage plane, has a great influence on the laser characteristics, but in the conventional example shown in FIG. 1, the entire active layer 3 is flat and exposed as it is at the end face 7. Therefore, the laser beam is easily absorbed by the GaAs on the end face 7 of the active layer 3, and as a result, heat is generated, resulting in a broken line on the end face, and high output operation is not possible.

In order to eliminate these defects, for example, as shown in Figure 2, the end faces on both sides are removed by etching.
A method has been proposed in which a material having a larger bandgap than the active layer 3, for example, a GaAlAs layer 8, is newly grown in this missing portion. But with this structure
In order to form the GaAlAu layer 8, it is necessary to carry out liquid phase epitaxial growth, which is disadvantageous in terms of workability. Further, as shown in FIG. 3, the active layer 3 is made N-type with a high concentration (>10 ¹⁸ cm ^-3 ) in advance, and Zn is diffused at a high concentration in regions other than the end face portions to form a P ⁺ type region 9. Some have formed. In this case, the band gap is largest in the active layer 3 and in the P ⁺ type region 9.
is the smallest, so the absorption of laser light at the end face 7 can be reduced to some extent. However, in addition to requiring precisely controlled Zn diffusion, doping into the active layer 3 is limited.
Furthermore, the difference in band gap between the end portion of the active layer 3 and the other portions is not very large.

The present invention has been made in order to eliminate the above-mentioned drawbacks, and includes a semiconductor substrate having a stripe-shaped groove on one main surface and having different widths at the center and both ends; A first confinement layer, an active layer, and a second confinement layer are formed sequentially on the semiconductor substrate, and the position of the lower part of the active layer is different between the center and both ends. This relates to a semiconductor laser. With this configuration, not only high output operation is possible, but also a laser can be created easily and with good workability.

Examples of the present invention will be described below with reference to FIGS. 4 to 16.

4 to 8 show the first embodiment.

As shown in FIG. 4, the semiconductor laser according to this embodiment has an N-type GaAs semiconductor substrate 11, an N-type Ga ₁ - _x Al _x As layer 12 as a first confinement layer, and a P-type GaAs layer as an active layer. layer 13, P-type Ga ₁ - _x Al _x As layer 14 as a second confinement layer, P-type GaAs layer 15
are sequentially laminated, and this laminated structure itself is the same as the conventional one. 18 is an ohmic electrode. What is important here is that the end portion 13a of the active layer 13 at the end face 17 or cleavage plane portion is located at a lower position than the light emitting portion 13b. That is, the end portions 13a of the active layer 13 are inclined downward from both ends of the light emitting portion 13b, and the second confinement layer 14 is directly above this inclined region. It is better for the end portion 13a to be as steeply sloped as possible, and it is sufficient that the end position to the end face 17 is Δh lower than the position of the light emitting portion 13b, for example, about 0.2 μ (=Δh) lower.

The slope of the end portion 13a is formed as follows. First, as shown in Figures 5 and 6, one N
Grooves 19 having a depth of 1 μm are formed in stripes on one main surface of the GaAs substrate 11 . This groove consists of a narrow linear groove 19a with a width w ₁ (eg 4μ) and a wide rectangular groove 19b with a width w ₂ (eg 20μ).The groove 19a therefore extends linearly between the grooves 19b. For example, the groove 19b is provided with a length _w3 of 200μ, and the length of the groove 19b in this length direction is
For example, w ₄ is 50μ. Then, the substrate 11 is cut into a plurality of pieces at the positions shown by the imaginary lines in FIG. 6 to obtain individual substrates 11 as shown in FIG. Therefore, on one principal surface of the substrate 11 in FIG. 5, there are grooves 19b half the size of those in FIG.
9a is present.

Before cutting (cleaving) at the position of the imaginary line in FIG. 6, each layer 12, 13, 14, 15 is grown in a liquid phase on the substrate 11 as shown in FIGS. 4 and 5, for example, as described below. The electrode 18 may be deposited on the entire surface by selective proton injection so that a current flows only in the groove portion, and the substrate 11 may be cut together with each of these layers.

According to this embodiment, each layer 12, 13, 14, 1
5, especially when growing 12, 13, and 14,
Due to the presence of the grooves 19, the active layer 15 grows so as to be inclined downward at the end portions as shown in FIG. That is, as clearly shown in FIGS. 7 and 8,
On the narrow groove 19a, the first confinement layer 1
2 grows relatively thick to h ₁ in the groove 19a, and the active layer 13 grows thereon in an almost flat state, whereas in the wide groove 19b, the active layer 13 grows in the groove as shown in the figure. It grows by entering into 19b. This is based on the principle of liquid phase growth, and as the first confinement layer 12 grows to a relatively thin thickness of h ₂ near the center of the wide groove 19b, it grows over that area. The active layer 13 grows in the groove 19b as if hanging down, as shown in FIG. For example, on the groove 19a, the first confinement layer 12 grows to a thickness of about 1.4μ, and the active layer 13 grows to a thickness of about 0.15μ, but on the groove 19b, the first confinement layer 12 grows to a thickness of about 0.15μ.
The active layer 13 grows to a thickness of about 0.4μ, and the active layer 13 grows to a thickness of about 0.25μ.
Therefore, since the widths of the grooves 19a and 19b are w ₁ <w ₂ , the first confinement layer 1 on the grooves 19a and 19b
It will be understood that the thickness of 2 is h ₁ > h ₂ .

In this way, the active layer 13 grows in the wide groove 19b so as to be inclined from above the groove 19a to a lower position, and as shown in FIG. 4, the active layer 13 is inclined downward on the end surface 17 side. At this time, when the active layer 13 is formed by liquid phase growth, it grows so that the groove or step part is relaxed, so the end 13a of the active layer 13 is
tends to be thicker than the light-emitting portion 13b, but even in this case, considering the central portion of the active layer 13 in the thickness direction, it can be considered that the active layer 13 is inclined downward on the end surface 17 side. For this reason, when operated by applying a voltage, the laser light from the light emitting part 13b of the active layer 13 is guided through the second confinement layer 14 on the end part 13a as shown by the arrow 20. become. As mentioned above, the second confinement layer 14
The band gap of Ga ₁ - _x Al _x As (~1.7 eV) is the same as the band gap of GaAs in the active layer 13 (~
1.425eV), the absorption of light 20 passing through the confinement layer 14 is significantly reduced compared to the case of GaAs, and therefore, the reduction in heat generation prevents edge destruction and enables high-power operation. It is.
The band gap at the end face 17 is the second confinement layer 1
4, and can vary widely.

Furthermore, as described above, after forming the groove 19, it is only necessary to perform normal liquid phase growth, so in order to obtain a high-power laser, it is necessary to precisely diffuse the P-type impurity as in the conventional method, or to perform another process. There is no need to perform epitaxial growth.

Note that in FIGS. 7 and 8, 21 is an insulating layer, which is, for example, about 200 KeV and a dose of 5×
It can be formed by selective ion implantation of H ⁺ at 10 ¹⁵ cm ^-2 . The insulator layer 21 has a groove 19
Since it is provided so that it does not exist on the width w ₁ of a, the current from the electrode 18 flows between the insulator layers 21 to the substrate 11 in the groove 19, and has the function of limiting the current path, so to speak. ing.

9 and 10 show another embodiment.

In this example, unlike the first embodiment described above,
The depth of the groove 19 is 4μ, the width of the groove 19b is narrow, for example, 1μ, on the end face 17 side (FIG. 10), and the wide groove 19a, for example, 4μ, is provided between the end faces with a length of 200μ. Therefore, in this case, the active layer 13 enters the groove 19a at an angle, but the active layer 13 enters the groove 19a at an angle.
Since it is flat on the surface b, the active layer 13 is formed so as to be lifted upward on the end surface 17 side, as shown by the imaginary line in FIG. Therefore, the laser beam 20 is transmitted to the first confinement layer 1 below the active layer 13.
Since the light is guided to the outside through Ga ₁ - _x Al _x As of No. 2, the band cap at the end face 17 becomes large, as described above, and light absorption can be reduced.

11 to 16 show methods applicable to each of the above embodiments.

First, as shown in FIG.
After growing layers 2, 13, 14, and 15, a fifth P-type Ga ₁ - _x Al _x As layer 30 is further grown. Next, this P-type layer 30 is selectively etched as shown in FIG. 12, and the remaining P-type layer 30 is used as a mask to remove the first
As shown in FIG. 3, an H ⁺ beam 31 is irradiated to form an insulating layer 21 in the epitaxial layer by ion implantation. Next, as shown in FIG. 14, the P-type layer 30 is removed and an ohmic electrode 18 is deposited to complete the laser.

In this example, as a mask for ion implantation.
_{Ga 1} ― _{x Al} - _x Al _x As) and does not cause distortion to the wafer (GaAs
The lattice constant is almost the same as that of Ga ₁ - _x Al _x As). The obtained semiconductor laser actually has a threshold current of 40-80mA and an output of
It oscillated in a single transverse mode up to 10-15 mW.

Advantageously, this ion implantation is performed as shown in FIGS. 15 and 16. That is, first, as shown in FIG. 15, when the ion beam 31 is implanted diagonally downward to the left, the beam is regulated by the upper left end and lower right end of the mask 30, and the opposing side 22a extends diagonally downward to the left as shown in the figure. , 22b.
is formed. Next, as shown in FIG. 16, when the beam implantation direction is changed to the diagonal direction downward to the right, the beam is implanted in a pattern opposite to that described above, and the left side 22a of the insulating layer 21 is aligned with the mask 30 in the same way as the side 22b. The ions are implanted so that they penetrate into the underside of the surface. Therefore, the opposing sides 22a of the finally obtained insulating layer 21,
The spacing between 22b becomes gradually narrower in the direction of the substrate 11 or in the depth direction, and becomes the minimum at the spacing l.
As a result, when the electrodes are formed and operated as shown in FIG. 14, the current path is narrowed by the narrow interval l, and the current cannot be effectively concentrated in the PN junction of the active layer 13. can. In this way, the current density becomes larger than in the case described above, and the rise characteristics during laser operation can be improved. Note that as the beam 31, in addition to H ⁺ , it is also possible to use O ⁺ , Ar ⁺ , Zn ⁺ , S ⁺ , Se ^{+ ,} etc., depending on the characteristics of the liquid phase growth layer.

Although the present invention has been described above with reference to embodiments, this embodiment can be further modified based on the technical idea of the present invention. For example, the shape and position of the groove 19 may be changed. In the above embodiment, the groove 19a or 19b is formed wider than the other grooves, but the width of this wide groove may be made even larger.For example, in FIG. It may be provided so as to penetrate up to the point. Furthermore, the conductivity types of each epitaxial layer and the substrate can be changed, and the constituent materials of each layer can also be changed.

As described above, in the present invention, the depth positions of the light emitting part and the end part of the active layer are made to be different from each other by the groove provided on one main surface of the base, so that the laser light from the light emitting part is transmitted to the end face. Alternatively, it is emitted through a layer with a large bandcap above or below the active layer on the cleavage plane side, thus reducing optical absorption, eliminating edge destruction, and enabling high-output operation.
Moreover, these effects can be achieved simply by growing each semiconductor layer on the uneven surface of the substrate using the usual method, so there is no need for the conventional precisely controlled diffusion or a separate liquid phase growth process after each layer is grown. It is completely unnecessary, and a laser with excellent characteristics can be easily created with good workability.

Furthermore, in the semiconductor laser of the present invention, by changing the width of the groove formed in the substrate, the thickness of the first confinement layer is substantially changed at the center of the active layer, which is the light emitting part. Not only gain-guided lasers but also refractive index-guided lasers can be easily created. Furthermore, even in the case of a gain-guided laser, by forming a current limiting layer on the substrate, current confinement can be achieved by, for example, changing the groove width.

[Brief explanation of the drawing]

Figures 1 to 3 show conventional examples, and
FIG. 1 is a sectional view of a conventional semiconductor laser, FIG. 2 is a sectional view of another semiconductor laser, and FIG. 3 is a sectional view of yet another semiconductor laser. Figure 4 ~ 1st
FIG. 6 shows an embodiment of the present invention.
The figure shows a cross-sectional view of the semiconductor laser according to the first embodiment, FIG. 5 is a perspective view of this laser seen from the cleavage plane side, FIG. 6 is a plan view of the semiconductor substrate before it is divided into individual semiconductor lasers, and FIG. 7 is a cross-sectional view taken along the line -- in FIG. 5, FIG. 8 is a cross-sectional view taken along the line -- in FIG. 5, FIG. 9 is a cross-sectional view similar to FIG. 7 of the semiconductor laser according to the second embodiment, and FIG. is a cross-sectional view similar to that shown in FIG. 8 of the same laser, FIGS. 11 to 14 are cross-sectional views showing the method of forming an insulator layer on a semiconductor laser in the order of steps by ion implantation, and FIGS. FIG. 3 is a cross-sectional view showing the injection method in the order of steps. In addition, in the symbols used in the drawings, 11
... Semiconductor substrate, 12... First confinement layer, 13... Active layer, 14... Second confinement layer, 17... End face or cleavage plane, 19... Groove, 21... Insulator layer, 13a... End part, 13b ...Light emitting part, 19a...Straight groove, 19b
...It is a rectangular groove.

Claims (1)

[Claims] 1. A semiconductor substrate having a stripe-shaped groove on one main surface and having different widths at the center and both ends thereof, and a first confinement layer, an active layer, and a first confinement layer, which are sequentially formed on the semiconductor substrate. 2. A semiconductor laser comprising two confinement layers, wherein the position of the lower part of the active layer is different between the central part and both ends.