JP2024004295A - Semiconductor laser element and method of manufacturing the same - Google Patents

Abstract

To improve the crystal quality in a semiconductor laser fabricated using IFVD or impurity diffusion.SOLUTION: A laminated growth layer 120 includes m (m≥1) laser resonators 200 formed therein. In each of the laser resonators 200, the degree of mixed crystallization in an active layer 134 is greater than the degree of mixed crystallization between a P-type clad layer 124 and a P-type contact layer 126. The m resonators include n (2≤n≤m) laser resonators 200 having oscillation wavelengths different from each other, and the degrees of mixed crystallization in the active layer 134 of the n laser resonators are different from each other. The degrees of mixed crystallization between the P-type clad layer 124 and the P-type contact layer 126 of the n laser resonators are substantially equal.SELECTED DRAWING: Figure 1

Description

The present disclosure relates to a multi-beam semiconductor laser device.

Semiconductor lasers (LDs) are small and highly efficient, so they are used as light sources for AR (Augmented Reality)/VR (Virtual Reality) devices such as head-mounted displays and smart glasses.

A multi-beam semiconductor laser in which a plurality of laser resonators (emitters) are monolithically integrated has been proposed as a high-power edge-emitting laser. In these applications, transverse single mode semiconductor lasers are used from the viewpoint of optical systems, but because the wavelength spectrum is narrow and coherency is high, if the wavelengths of multiple beams are the same, interference between the beams will cause a There is a problem that an unfavorable intensity distribution (interference fringes) appears in the field pattern (FFP) and the beam quality deteriorates. To solve the beam quality problem, it is necessary to intentionally shift the oscillation wavelengths of multiple channels.

IFVD (Impurity Free Vacancy Disordering) can be used as a technology for producing multi-wavelength/multi-beam lasers. In this method, point defects that serve as diffusion sources are introduced into the active layer, and heat treatment is performed to cause atoms in the quantum well layer and the guide layer (barrier layer) to interdiffuse (QWI: Quantum Well Intermixing). As a result, the well layer and the guide layer become disordered, and the energy level in the well layer can be controlled (Patent Document 1).

In IFVD, generally by forming a SiO ₂ film on a GaAs layer, group III atoms are extracted from the GaAs layer to the SiO ₂ film, and point defects of the group III vacancy type, which serve as diffusion sources, are generated in the crystal. let Then, heat treatment is performed to diffuse these into the active layer and promote QWI. When semiconductor lasers with different wavelengths are made monolithic, the amount of diffused source introduced into the crystal may be changed for each emitter. For example, in Non-Patent Document 1, the SiO ₂ film is patterned, thereby changing the proportion of the SiO ₂ film in contact with the GaAs layer. By doing this, the amount of diffused sources generated varies from region to region, making it possible to achieve spatial control of the emission wavelength.

The diffusion source in IFVD is not limited to point defects, and even if an impurity such as Zn is used, the energy level can be changed in the same manner as described above. For example, in Patent Document 2, wavelength control in the end face window structure is realized by doping the P layer with C as an impurity and thereby promoting mixed crystal formation.

Patent No. 4711623 Patent No. 5128604 JP2008-235790A

Boon Siew Ooi et al., IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 33, NO. 10, pp. 1784-1793, 1997

As explained in Patent Document 3, conventionally commonly used IFVD and impurity diffusion require diffusion on a scale of several μm from the epitaxial surface (mainly the P-side contact layer) to the active layer. In order to create a sufficient difference in emission wavelength between a plurality of emitters, it is necessary to introduce these diffused sources at a fairly high density. However, if point defects or impurities, which are diffusion sources, exist in the crystal, they form defect levels or impurity levels, which become nonradiative recombination centers or light absorption sources. Particularly in semiconductor lasers, the presence of these elements directly causes negative effects such as a reduction in internal quantum efficiency and an increase in internal loss, making it difficult to achieve both wavelength control and crystal quality.

Furthermore, in multi-wavelength/multi-beam lasers, in order to fabricate emitters with different wavelengths for each emitter, it is necessary to change the degree of QWI for each emitter, which inevitably affects the amount of diffused source incorporated into the crystal. There will be a difference. Therefore, differences in crystal quality occur between emitters, which causes variations in characteristics between emitters and variations in device reliability.

Wavelength control by IFVD or impurity diffusion can also be used to control the oscillation wavelength in a single beam semiconductor laser.

Certain embodiments of the present disclosure have been made in view of such problems, and one exemplary purpose thereof is to improve the crystal quality and/or increase the crystal quality in a semiconductor laser manufactured using IFVD or impurity diffusion. The goal is to eliminate variations in device characteristics among emitters in wavelength/multiple beam semiconductor lasers.

An embodiment of the present disclosure is an edge-emitting type semiconductor laser device, which includes an N-type semiconductor substrate, an N-type cladding layer, an active layer, a P-type cladding layer, and a P-type contact layer formed on the N-type semiconductor substrate. A laminated growth layer containing. m laser resonators (m≧1) are formed in the laminated growth layer, and in each laser resonator, the degree of mixed crystallization in the active layer varies from the P-type cladding layer to the P-type contact layer. greater than the degree of change.

Another aspect of the present disclosure also relates to an edge-emitting semiconductor laser device. This semiconductor laser device includes an N-type semiconductor substrate and a laminated growth layer including an N-type cladding layer, an active layer, a P-type cladding layer, and a P-type contact layer formed on the N-type semiconductor substrate. A plurality of laser resonators are formed in the laminated growth layer, and the degree of mixed crystallization in the active layer differs between the plurality of laser resonators, and the P-type cladding layer changes from the P-type cladding layer to the P-type laser resonator. The degree of mixed crystallization between the mold contact layers is substantially equal.

Yet another aspect of the present disclosure is also a semiconductor laser device. This semiconductor laser device includes an N-type semiconductor substrate and a laminated growth layer including an N-type cladding layer, an active layer, a P-type cladding layer, and a P-type contact layer formed on the N-type semiconductor substrate. A plurality of laser resonators with different oscillation wavelengths are formed in the laminated growth layer, and the degree of mixed crystallization in the active layer differs among the plurality of laser resonators, and P-type The film thickness, composition ratio, and impurity concentration of each layer between the cladding layer and the P-type contact layer are substantially the same.

Yet another aspect of the present disclosure is a method of manufacturing an edge-emitting semiconductor laser device. The manufacturing method includes a first step of forming a first growth layer including at least an N-type cladding layer and an active layer on an N-type semiconductor substrate, forming a dielectric film on the first growth layer, and performing heat treatment. The method includes a second step of forming a mixed crystal by applying a mixed crystal, and a third step of removing the dielectric film and forming a second growth layer including a P-type cladding layer by regrowth.

Note that arbitrary combinations of the above components, and mutual substitution of the components and expressions of the present disclosure among methods, devices, systems, etc., are also effective as aspects of the present disclosure.

According to an aspect of the present disclosure, it is possible to improve crystal quality in a semiconductor laser manufactured using IFVD or impurity diffusion, and/or eliminate device characteristic variations for each emitter in a multi-wavelength, multi-beam semiconductor laser. can.

1 is a cross-sectional view of a multi-beam semiconductor laser device according to an embodiment. It is a figure explaining the manufacturing method concerning an embodiment. It is a figure explaining the manufacturing method concerning an embodiment. It is a figure explaining the manufacturing method concerning an embodiment. It is a figure explaining the manufacturing method concerning an embodiment. It is a figure explaining the manufacturing method concerning an embodiment. It is a figure explaining the manufacturing method concerning an embodiment. 3 is a diagram illustrating a wavelength control process according to Example 1. FIG. FIG. 7 is a diagram illustrating a wavelength control process according to Example 2. FIG. 7 is a diagram illustrating a wavelength control process according to Example 3. FIG. 7 is a diagram illustrating a wavelength control process according to Example 5. FIG. 7 is a diagram illustrating a wavelength control process according to Example 6. FIG. 7 is a diagram illustrating a wavelength control process according to Example 7. FIG. 3 is a cross-sectional view of two samples created in a verification experiment. FIG. 3 is a diagram showing the results of PL (photoluminescence) measurement of four divided samples of the first sample after heat treatment. FIG. 6 is a diagram showing the results of PL (photoluminescence) measurement of four divided samples of the second sample after heat treatment. FIG. 2 is a diagram showing high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images of a cross section of the active layer of the first sample and the second sample, respectively. FIG. 3 is a diagram showing a normalized contrast profile F(y). It is a figure showing the result of PL measurement of a 1st sample and a 2nd sample. FIG. 3 is a diagram showing the relationship between the theoretically calculated PL wavelength difference and the composition transition length.

(Summary of embodiment)
1 provides an overview of some example embodiments of the present disclosure. This summary is provided as a prelude to the more detailed description that is presented later or to provide a basic understanding of the embodiments. This Summary is a simplified description of some concepts of one or more embodiments and is not intended to limit the breadth of the invention or disclosure. Moreover, this summary is not an exhaustive overview of all possible embodiments, and is not limited to essential components of the embodiments. For convenience, "one embodiment" may be used to refer to one embodiment (example or modification) or multiple embodiments (examples or modifications) disclosed in this specification.

An edge-emitting type semiconductor laser device according to one embodiment has a stacked-layer structure including an N-type semiconductor substrate, an N-type cladding layer, an active layer, a P-type cladding layer, and a P-type contact layer formed on the N-type semiconductor substrate. comprising a layer. m laser resonators (also called emitters) are formed in the laminated growth layer, and in each laser resonator, the degree of mixed crystallization in the active layer varies from the P-type cladding layer to the P-type contact layer. This is greater than the degree of mixed crystallization between layers.

In one embodiment, m≧2. The m laser resonators may include n laser resonators having different oscillation wavelengths (2≦n≦m). The degree of mixed crystallization in the active layer may be different among the n laser resonators, and the degree of mixed crystallization between the P-type cladding layer and the P-type contact layer may be substantially equal.

By varying the degree of mixed crystallization only in the active layer, it is possible to shift the wavelength of each laser resonator while maintaining crystal quality. Further, since the degree of mixed crystal formation from the P-type cladding layer to the P-type contact layer is uniform, variations in characteristics of the n laser resonators can be suppressed.

An edge-emitting type semiconductor laser device according to one embodiment has a stacked-layer structure including an N-type semiconductor substrate, an N-type cladding layer, an active layer, a P-type cladding layer, and a P-type cladding layer formed on the N-type semiconductor substrate. comprising a layer. A plurality of laser resonators are formed in the laminated growth layer, and the degree of mixed crystallization in the active layer differs between the plurality of laser resonators, and the P-type cladding layer changes from the P-type cladding layer to the P-type laser resonator. The degree of mixed crystallization between the mold contact layers is substantially equal.

An edge-emitting type semiconductor laser device according to one embodiment has a stacked-layer structure including an N-type semiconductor substrate, an N-type cladding layer, an active layer, a P-type cladding layer, and a P-type contact layer formed on the N-type semiconductor substrate. comprising a layer. A plurality of laser resonators with different oscillation wavelengths are formed in the laminated growth layer, and the degree of mixed crystallization in the active layer differs among the plurality of laser resonators, and P-type The film thickness, composition ratio, and impurity concentration of each layer between the cladding layer and the P-type contact layer are substantially the same.

In one embodiment, the difference in oscillation wavelength between the two laser cavities may be 2 nm or more.

In one embodiment, the active layer may include one of the group consisting of AlGaInP, AlGaAs, InGaAsP, AlGaInAs, AlGaInAsP, GaInNAs.

In one embodiment, different degrees of mixed crystallization can be defined by compositional transition lengths, which will be described in detail later, but may also mean that compositional transition lengths confirmed from a TEM image are separated by 0.16 nm or more. .

In one embodiment, in the active layer, the concentrations of C and Zn in optical gain regions corresponding to laser resonators with different oscillation wavelengths are higher in the laser resonators with shorter oscillation wavelengths.

A method for manufacturing a semiconductor laser device according to an embodiment includes a first step of forming a first growth layer including at least an N-type cladding layer and an active layer on an N-type semiconductor substrate; a second step in which a dielectric film is formed and mixed crystallized by heat treatment; and a third step in which the dielectric film is removed and a second growth layer including a P-type cladding layer is formed by regrowth. , is provided.

In this method, before forming the P-type cladding layer, the first growth layer including the active layer is subjected to QWI treatment such as IFVD or impurity diffusion. Therefore, a diffusion source is not directly introduced into the P-type cladding layer. This can improve crystal quality.

In one embodiment, the second step may include a step of varying at least one of the film thickness, film type, and film configuration of the dielectric film in a plurality of regions. The manufacturing method may further include a fourth step of forming a laser resonator in each of the plurality of regions. Thereby, the density of the diffused source can be made different for each region, and the oscillation wavelength can be controlled.

In one embodiment, the second step may include a step of varying the contact ratio between the first growth layer and the dielectric film in a plurality of regions. The manufacturing method may further include a fourth step of forming a laser resonator in each of the plurality of regions. Thereby, the density of the diffused source can be made different for each region, and the oscillation wavelength can be controlled.

In one embodiment, the second step includes etching the first growth layer at different depths in a plurality of regions, and forming a dielectric film on the first growth layer after etching. May include. The manufacturing method may further include a fourth step of forming a laser resonator in each of the plurality of regions. Thereby, the density of the diffused source can be made different for each region, and the oscillation wavelength can be controlled.

(Embodiment)
Hereinafter, the present disclosure will be described based on preferred embodiments with reference to the drawings. Identical or equivalent components, members, and processes shown in each drawing are designated by the same reference numerals, and redundant explanations will be omitted as appropriate. Further, the embodiments are illustrative rather than limiting the disclosure, and all features and combinations thereof described in the embodiments are not necessarily essential to the disclosure.

The dimensions (thickness, length, width, etc.) of each member shown in the drawings may be scaled up or down as appropriate for ease of understanding. Furthermore, the dimensions of multiple members do not necessarily represent their size relationship, and even if a member A is drawn thicker than another member B on a drawing, member A may be drawn thicker than member B. It may be thinner than that.

FIG. 1 is a cross-sectional view of a semiconductor laser device 100 according to an embodiment. The semiconductor laser device 100 is an edge-emitting type semiconductor laser device, and is a multi-beam semiconductor laser having a plurality of m (m≧2) emitters 102_1 to 102_3. In this embodiment, the number of emitters is three.

The semiconductor laser device 100 includes an N-type semiconductor substrate 110 and a stacked growth layer 120 formed on the N-type semiconductor substrate 110. In this embodiment, semiconductor laser device 100 is a red laser, and N-type semiconductor substrate 110 is a GaAs substrate.

The laminated growth layer 120 includes an N-type cladding layer 122, a light emitting layer 130, a P-type cladding layer 124, and a P-type contact layer 126, and is laminated in this order on the N-type semiconductor substrate 110. The light emitting layer 130 includes an N-type guide layer (lower guide layer) 132, an active layer 134 made of a quantum well layer, and a P-type guide layer (upper guide layer) 136. Active layer 134 includes one of the group consisting of AlGaInP, AlGaAs, InGaAsP, AlGaInAs, AlGaInAsP, and GaInNAs.

In the semiconductor laser device 100, m laser resonators 200 are formed adjacent to each other in the first direction (x direction) within the plane of the chip. Each laser resonator 200 has a striped structure extending in a second direction (z direction, depth direction in the paper) perpendicular to the first direction (x direction). For example, each laser resonator 200 is a transverse single mode laser, and the ridge width of the ridge portion where the P-type contact layer 126 is formed is about 2 μm. Further, the pitch between the beams can be about 30 μm.

A waveguide structure for confining light is formed in the laminated growth layer 120, and cleavage planes at both ends of this waveguide structure serve as mirrors to form a laser resonator 200. In this example, three laser resonators 200_1 to 200_3 are formed, and a beam is emitted from the emitter 102 in the z direction.

For example, a ridge structure can be used as the waveguide structure. The ridge structure is formed by partially removing the P-type cladding layer 124. The ridge structure is also simply referred to as a ridge or a ridge stripe structure. A bank may be formed between adjacent laser resonators 200. The waveguide structure can also be a buried ridge waveguide.

Alternatively, the waveguide structure may be a CSP (Channeled Substrate Planer) structure in which a groove is formed in the N-type semiconductor substrate 110 along the waveguide, and the thickness of the N-type cladding layer 122 in the groove portion is relatively thick. You can.

Although the ridge structure and the CPS structure are waveguide structures that utilize refractive index distribution, the present disclosure is not limited thereto, and a gain waveguide structure that utilizes gain distribution may be used. These structures can be understood as not only optical confinement structures but also current confinement structures.

Among the m laser resonators 200, n (2≦n≦m) have different oscillation wavelengths. In this example, m=n, and the oscillation wavelengths λ ₁ to λ ₃ of the three laser resonators 200_1 to 200_3 are different.

In order to control the oscillation wavelength of the laser resonator 200, the optical gain region of the light emitting layer 130 is formed between the active layer 134 and the N-type guide layer 132 and between the active layer 134 and the P-type guide layer 136 by IFVD or impurity diffusion. Mixed crystallization is performed. The difference in C and Zn impurity concentrations between the P-type cladding layer 124 and the N-type cladding layer 122 in each emitter may be within ±5%.

Comparing the three laser resonators 200, the degree of mixed crystallization is larger in the order of laser resonators 200_1, 200_2, and 200_3, and the oscillation wavelengths are shorter in this order (λ1<λ2<λ3). The difference in oscillation wavelength between the two laser resonators 200 is 2 nm or more.

Furthermore, when focusing on each laser resonator 200, the degree of mixed crystallization in the active layer 134 is greater than the degree of mixed crystallization between the P-type cladding layer 124 and the P-type contact layer 126.

Furthermore, the degree of mixed crystal formation between the P-type cladding layer 124 and the P-type contact layer 126 is substantially equal among the n laser resonators 200_1 to 200_3. In other words, among the n laser resonators 200, the film thickness, composition ratio, and impurity concentration of each layer between the P-type cladding layer 124 and the P-type contact layer 126 are substantially equal, and the P-type cladding No IFVD or impurity diffusion is performed between the layer 124 and the P-type contact layer 126.

As an example, when comparing the plurality of laser resonators 200_1 to 200_3, the difference in the thickness of the P-type cladding layer 124 may be uniform within ±5%. Furthermore, when comparing the plurality of laser resonators 200_1 to 200_3, the difference in the composition ratios of In, Ga, and Al in the P-type cladding layer 124 may be within ±5%. Furthermore, when comparing the plurality of laser resonators 200_1 to 200_3, the difference in concentration of impurities C, O, Mg, Zn, and Si between the P-type cladding layer 124 and the P-type contact layer 126 is within ±5%. Good too.

The above is the configuration of the semiconductor laser device 100.

Next, a method for manufacturing the semiconductor laser device 100 will be explained.

2A to 2F are diagrams illustrating the manufacturing method according to the embodiment. As shown in FIG. 2A, an N-type cladding layer 122, an N-type guide layer 132, an active layer 134, a P-type guide layer 136, and a cap layer 160 are formed on a GaAs substrate, which is an N-type semiconductor substrate 110, by metal organic vapor phase growth. Epitaxial growth is performed by MOCVD (Metal Organic Chemical Vapor Deposition) method.

N-type cladding layer 122 may be Al _0.5 In _0.5 P with a thickness of approximately 2 μm. In order to fabricate a waveguide, the compositions of the N-type guide layer 132 and the P-type guide layer 136 need to be adjusted so that they have a higher refractive index than the N-type cladding layer 122 and the P-type cladding layer 124. In this embodiment, x=0.7 and y=0.5 in (Al _x Ga _1-x ) _1-y In _y P, and the thickness may be 100 nm. The active layer 134 may be two layers of Ga _0.5 In _0.5 P with a thickness of approximately 5 nm. The guide layer may be made of Al _0.35 Ga _1.5 In _0.5 P similar to the guide layer. The cap layer 160 was made of GaAs grown to a thickness of 50 nm. The layer formed in FIG. 2A is referred to as a first growth layer 162.

Subsequently, as shown in FIG. 2B, a dielectric film 164 is formed on the surface of the first growth layer 162. The process in FIG. 2B and the subsequent process in FIG. 2C are referred to as a wavelength shift process or a wavelength control process. This dielectric film 164 is formed for the purpose of extracting Ga, which is a group III element, from the GaAs of the cap layer 160 and introducing Ga vacancies into the crystal. Note that the combination of the cap layer 160 and the dielectric film 164 may be selected from the viewpoint of generation of group III vacancy defects due to extraction of group III atoms. Although not limited thereto, for example, a typical material for the dielectric film 164 suitable for extracting Ga from the cap layer 160 is SiO ₂ .

In IFVD, the degree of mixed crystallization between the active layer 134 and the N-type guide layer 132 and between the active layer 134 and the P-type guide layer 136 changes depending on the vacancy density of the dielectric film 164. The amount of change in wavelength is determined by The amount of Ga vacancies can be controlled by the following conditions (i) to (viii).
(i) Film thickness of dielectric film 164 (ii) Film thickness of cap layer 160 (iii) Epi film thickness from wafer surface to active layer 134 (iv) Film type (material) of dielectric film 164
(v) Film quality of the dielectric film 164 (vi) Structure of the dielectric film 164 (vii) Proportion of the dielectric film 164 in contact with the cap layer 160 (viii) Can be controlled by thermal diffusion treatment conditions, etc.

The semiconductor laser device 100 is divided into three regions A1 to A3 in which three emitters (laser resonators) are formed. By applying a process in which at least one of conditions (i) to (viii) is different for each region, the degree of mixed crystallization of the light emitting layer 130 can be changed for each emitter.

Next, the Ga vacancies are diffused by thermal diffusion treatment to form a mixed crystal between the active layer 134 and the N-type guide layer 132 and between the active layer 134 and the P-type guide layer 136. In this embodiment, heat treatment may be performed at 950° C. for 4 minutes in an RTA (Rapid Thermal Anneal) furnace.

Subsequently, as shown in FIG. 2C, after heat treatment, wet etching with hydrofluoric acid is performed to remove the dielectric film 164 formed on the wafer surface, and further, the cap layer 160 is removed by ion milling or the like in the processing chamber. The light emitting layer 130 was exposed. In the present disclosure, since the distance between the crystal layer into which the diffused source is introduced and the active layer 134 is on the order of sub-μm scale, the wavelength can be controlled without applying excessive heat treatment.

Subsequently, as shown in FIG. 2D, regrowth is performed using the MOCVD method to form a second growth layer 166 including the P-type cladding layer 124 and the P-type contact layer 126. Here, the P-type cladding layer 124 is made of Al _0.5 In _0.5 P with a thickness of 2 μm, and may be doped with Mg as a P-type dopant.

Subsequently, an interface layer was formed from the P-type cladding layer 124 to the P-type contact layer 126 in order to continuously change the band edge, and finally the P-type contact layer 126 was formed. Note that an AlInP layer, a GaInP layer, and an AlGaAs layer may be formed in this order as the interface layer, and AlInP and GaInP may be doped with Mg, and AlGaAs may be doped with Zn. Further, the P-type contact layer 126 may be doped with Zn.

As described above, in the present disclosure, since the P-type cladding layer 124 (and the P-type contact layer 126) in all emitters are formed under the same crystal conditions, it is possible to reduce variations in crystal quality among emitters. be.

Subsequently, as shown in FIG. 2E, after the multilayer growth process of the second growth layer 166, a ridge structure 168 may be formed using a conventional method. For example, the ridge structure 168 is formed by dry etching so that the ridge width is 2 μm. The pitch interval of each emitter is, for example, 30 μm. Note that an isolation groove having a function of blocking current and guided light may be formed between the emitters, or a bank may be formed.

Subsequently, as shown in FIG. 2F, after forming an insulating layer 140 and opening the upper part of the ridge, a P electrode 150 may be formed, and an N electrode 152 may be formed on the back surface of the N type semiconductor substrate 110. Furthermore, the semiconductor laser device 100 is completed through cleavage and chipping.

Through the above steps, it is possible to realize a device with improved crystal quality compared to multi-wavelength, multi-beam semiconductor lasers fabricated using conventional IFVD technology, and which can reduce differences in crystal quality between emitters and variations in device characteristics. can.

Next, some examples will be described regarding the wavelength control process shown in FIG. 2B.

(Example 1)
FIG. 3 is a diagram illustrating a wavelength control process according to the first embodiment. In the first embodiment, the condition (i), that is, the thickness t of the dielectric film 164A formed on the first growth layer 162 is made different for each region, so that the thickness of the first emitter 102_1 to the third emitter 102_3 is In order, the wavelength becomes shorter. For example, SiO ₂ is used as the material of the dielectric film 164A, and dielectric films 164_1, 164_2, and 164_3 having different thicknesses t ₁ to t ₃ are formed in three regions A1 to A3 by CVD and sputtering. Ru.

As a result, SiO 2 with a first thickness t ₁ (for example, 500 nm) is formed in the first region A1, and SiO ₂ with a second thickness t ₂ (for example, 300 nm) is formed in the second region A2 _. , SiO ₂ having a third thickness t ₃ (for example, 100 nm) is formed in the third region A3.

The number of pores in the GaAs cap layer 160, which serves as a diffusion source, increases as the SiO ₂ film thickness t increases. In other words, the amount of hole defects introduced into the cap layer 160 made of GaAs is larger in the order of the first region A1, the second region A2, and the third region A3. As a result, the oscillation wavelength λ1 of the first region A1 becomes the shortest, the oscillation wavelength λ2 of the second region A2 becomes the next shortest, and the oscillation wavelength λ3 of the third region A3 becomes the longest.

(Example 2)
FIG. 4 is a diagram illustrating a wavelength control process according to the second embodiment. In Example 2, as in Example 1, the first emitter is The wavelength is made shorter in the order of the third emitter 102_1 to the third emitter 102_3. SiO ₂ is used as the material of the dielectric film 164, and the dielectric film 164_1 with different thicknesses t ₁ to t 2 is formed in two regions A1 and A2 of the three regions A1 to _A3 by CVD and sputtering. , 164_2 are formed.

In region A3, the film thickness of SiO ₂ is zero, and SiN is formed as the dielectric film 164_3 instead.

The configuration of FIG. 3 uses SiO ₂ and SiN as materials for the dielectric film 164, and forms the films of SiO ₂ and SiN by CVD and sputtering, and removes the dielectric film 164 by photolithography and wet etching. This can be achieved through repetition.

As a result, SiO 2 with a first thickness t ₁ (for example, 500 nm) is formed as a dielectric film 164_1 in the first region A1, and SiO ₂ with a second thickness t 1 is formed as a dielectric film 164_2 in the second region A2. ₂ (for example, 100 nm) is formed, and SiN with _a third thickness t ₃ (for example, 100 nm) is formed as the dielectric film 164_3 in the third region A3.

When a single-layer SiN film is formed on the GaAs cap layer 160, almost no Ga element is extracted into the dielectric film 164_3. In other words, by forming SiN, QWI can be suppressed compared to _SiO2 . Here, if neither SiN nor SiO ₂ is formed in the third region A3, the group V atoms in the surface layer will be eliminated in the subsequent heat treatment. On the other hand, by forming SiN in the region where the SiO ₂ film thickness is 0, it is possible to prevent V group atoms from being desorbed. According to the second embodiment, the amount of hole defects introduced into the cap layer 160 made of GaAs increases in the order of the first region A1, the second region A2, and the third region A3, and satisfies λ1<λ2<λ3. can do.

(Example 3)
FIG. 5 is a diagram illustrating a wavelength control process according to the third embodiment. In Example 3, the condition (iv), that is, the film type (material) of the dielectric film 164C, differs from region to region.

For example, the dielectric films 164_1 to 164_3 formed in the first region A1 to the third region A3 may be made of SiO ₂ (100 nm), Al ₂ O ₃ (100 nm), and SiN (100 nm), respectively.

In IFVD, the amount of group III atoms extracted can be changed depending on the type of dielectric film 164C. Note that the film type of the dielectric film 164C used for the wavelength control process is not limited to the material used in this embodiment, and for example, TiO ₂ , Ta ₂ O ₅ , ZrO, AlN, etc. can be used.

(Example 4)
In Example 4, wavelength control is performed by making the film quality of condition (v) different for each region. Even if the dielectric film 164 is made of the same material and has the same thickness, the film quality of the dielectric film 164 can be changed from region to region depending on the film formation method and film formation conditions. For example, the same film type having different film stresses may be formed in the regions A1 to A3 by a CVD (Chemical Vapor Deposition) method, an ALD (Atomic Layer Deposition) method, a sputtering method, or the like.

(Example 5)
FIG. 6 is a diagram illustrating a wavelength control process according to the fifth embodiment. In Example 5, the condition (vi), that is, the configuration of the dielectric film 164D, differs from region to region.

The dielectric film 164D has a multilayer structure in which films of two or more materials having different degrees of extraction of group III atoms in IFVD are laminated. Here, the dielectric film 164D has a two-layer structure including a lower layer film 164L in contact with the cap layer 160 and an upper layer film 164U, and the combination of the film thicknesses of the two-layer structure is different for each region A1 to A3. SiN and SiO ₂ can be used as a combination of the lower layer film 164L and the upper layer film 164U.

For example, the film thicknesses of SiN/SiO ₂ formed on the first region A1, the second region A2, and the third region A3 are (25 nm/125 nm), (50 nm/100 nm), and (75 nm/75 nm), respectively. In such a structure, the film 164L in contact with the cap layer 160 can suppress the action of layer III atom extraction by the dielectric film 164U formed above. Moreover, this degree can be controlled by the thickness of each dielectric film 164L, 164U.

(Example 6)
FIG. 7 is a diagram illustrating a wavelength control process according to the sixth embodiment. In Example 6, the condition (vii), that is, the proportion of the dielectric film 164E in contact with the cap layer 160 is different for each of the regions A1 to A3. For example, SiO ₂ can be used for the dielectric film 164E.

For example, by patterning the dielectric film 164E, the ratio of the area where the dielectric film 164E contacts the cap layer 160 to the area where the cap layer 160 (the GaAs crystal part) is exposed may be made different for each region A1 to A3. good. For example, in the first region A1, the second region A2, and the third region A3, the contact ratio between the dielectric film 164E and the cap layer 160 may be 90%, 70%, and 50%, respectively.

As the contact ratio between the dielectric film 164E and the cap layer 160 increases, the amount of group III atoms extracted by SiO ₂ increases, and the amount of group III vacancies introduced into the crystal portion increases.

(Example 7)
FIG. 8 is a diagram illustrating a wavelength control process according to the seventh embodiment. In Example 7, condition (ii), that is, the thickness of the cap layer 160 is different for each region.

In Example 7, after the growth of the first growth layer 162 is completed, the cap layer 160 is etched, and the thickness h ₁ , h ₂ , h ₃ of the cap layer 160 is changed for each region A1 to A3. Note that the dielectric film 164F may be formed of SiO ₂ with the same thickness in all regions A1 to A3. The thickness of the cap layer 160 satisfies h ₁ <h ₂ <h ₃ , and as the film thickness increases, the amount of group III vacancies introduced into the crystal portion tends to increase. Therefore, the degree of QWI can be controlled depending on the thickness of the cap layer 160F.

In this embodiment, etching may be performed so that the thickness of the GaAs cap layer 160 becomes 200 nm, 100 nm, and 50 nm in the order of the first region A1, the second region A2, and the third region A3. Subsequently, a 300 nm thick SiO ₂ film may be formed as the dielectric film 164F by CVD.

(Example 8)
In Example 8, an impurity diffusion method is used in place of IFVD in the wavelength control step. For example, the dielectric film 164 formed on the cap layer 160 may be a two-layer film of ZnO and SiO ₂ , and the thickness of the ZnO film may be changed for each region. The thickness of ZnO in each region A1, A2, and A3 may be 300 nm, 100 nm, and 50 nm, and the thickness of SiO ₂ may be 50 nm.

(Verification experiment 1)
The results of experiments that verified the effects of the semiconductor laser device 100 or the manufacturing method thereof according to the embodiment will be described. FIG. 9 is a cross-sectional view of two samples created in a verification experiment.

The first sample S1 simulates the cross-sectional structure in the wavelength control step in the embodiment. The second sample S2 imitates the cross-sectional structure in the wavelength control process in the prior art. In the conventional technology, a GaAs layer (cap layer) is formed after forming a P-type cladding layer and a P-type contact layer, and the AlInP layer of the second sample S2 has a P-type cladding layer 124 and a P-type contact layer 126. It is a model of

The first sample S1 and the second sample S2 have different distances from the active layer 134 to the outermost crystal surface. The first sample S1 has a structure in which the distance from the active layer 134 to the outermost layer is smaller than that of the second sample S2, and the effect of the present disclosure is relatively large.

A SiO ₂ film having a thickness of 500 nm may be formed on these samples S1 and S2. Thereafter, each of Sample 1 and Sample 2 was divided into four parts, and heat treatment at 950° C. was performed for each divided sample at different times of 0 minutes, 2 minutes, 6 minutes, and 1 minute.

FIG. 10 is a diagram showing the results of PL (photoluminescence) measurement of four divided samples of the first sample S1 after heat treatment. FIG. 11 is a diagram showing the results of PL (photoluminescence) measurement of four divided samples of the second sample S2 after heat treatment.

Looking at the results of PL measurement conducted after heat treatment, it is found that in the first sample S1, the PL wavelength changes significantly as the heat treatment time increases, whereas in the second sample S2, there is no significant change in the PL wavelength. I know you can't see it. This indicates that even under the same heat treatment conditions, mixed crystal formation occurred in the active layer 134 only in the first sample S1. In other words, even if the thermal energy introduced into the crystal is smaller than in the past, it is possible to create a desired wavelength difference, and it is possible to suppress the introduction of damage caused by heat treatment. This was supported by verification experiments.

(Verification experiment 2)
A sample is prepared by crystal-growing the first growth layer 162 on the N-type semiconductor substrate 110, and the sample is divided into two parts.Only one side (the second sample) is subjected to the wavelength control process, and the other side (the first sample) is subjected to the wavelength control process. ) was not subjected to a wavelength control process. Two samples were measured by High-Angle Annular Dark Field Scanning Transmission Electron Microscopy (HAADF-STEM). FIG. 12 is a diagram showing HAADF-STEM images of the cross sections of the active layers of the first sample and the second sample.

In the first sample shown in Fig. 12, the interface between the active layer (quantum well layer) and the guide layer is steep, whereas in the second sample shown in Fig. 12, the steepness of the interface is lost and mixed crystal formation occurs. I can confirm that there is. The degree of mixed crystallization can be evaluated, for example, by observation using a TEM. Further, the degree of mixed crystallization can be expressed quantitatively using a TEM contrast profile.

FIG. 13 is a diagram showing the normalized contrast profile F(y). Specifically, the contrast profile averaged in the x direction (horizontal direction of the paper) within the broken line area of the TEM images of the first sample and the second sample in FIG. 12 is f(y), the contrast in the active layer is Cw, and the contrast in the guide layer is When the contrast is Cb, the normalized contrast profile F(y) is expressed by the following formula.
F(y)=(f(y)-Cb)/(Cw-Cb)

Here, the vertical axis in FIG. 13 indicates the normalized contrast intensity, which is expressed as composition contrast. The distance at which the composition contrast changes from 0.8 to 0.2 is defined as the composition transition length. In the first sample, the composition transition length is 1.1 nm, and in the second sample, the composition transition length is 2.3 nm.

FIG. 14 is a diagram showing the results of PL measurement of the first sample and the second sample. The compositional transition lengths of the first sample and the second sample are 1.1 nm and 2.3 nm, and the difference in peak wavelength is 11 nm.

FIG. 15 is a diagram showing the theoretically calculated relationship between the PL wavelength difference and the composition transition length. In the theoretical calculation, the wavelength difference between the compositional transition lengths of 1.1 nm and 2.3 nm is 13 nm, which generally agrees with the experimental results shown in FIG.

By summarizing these verification results, the following findings can be obtained.

The degree of mixed crystallization can be quantitatively evaluated as the composition transition length confirmed from a TEM image. In this embodiment, when the difference in oscillation wavelength for each emitter is 2 nm, the difference in compositional transition length of the active layer in the two laser resonators is 0.16 nm at the minimum and 0.7 nm at the maximum.

In summary, "the degree of mixed crystallization differs" between two laser cavities means the range in which the difference in the compositional transition length of the active layer is the minimum with respect to the oscillation wavelength difference Δλ in FIG. is larger than the range in which it changes linearly. For example, when Δλ is 2 nm, the compositional transition length is 0.16 or more.

(Modified example)
The embodiments and examples described above are merely illustrative, and those skilled in the art will understand that various modifications can be made to the combinations of their constituent elements and processing processes. Hereinafter, such modified examples will be explained.

In the embodiment, the case where m=3 has been described, but the number m of beams is not limited to 3, and may be m=1 or m=2, or m may be 4 or more. Even in the semiconductor laser device 100 where m=1, that is, a single beam, the technology according to the present disclosure is effective when wavelength control is performed by QWI.

Furthermore, in the embodiment, a case has been described where the number n of laser resonators having different oscillation wavelengths is equal to m, that is, a case where all the laser resonators have different oscillation wavelengths, but this is not the case.

The embodiments merely illustrate the principles and applications of the present invention, and the embodiments may include many modifications and changes in arrangement without departing from the spirit of the present invention as defined in the claims. is recognized.

100 Semiconductor laser element 110 N-type semiconductor substrate 120 Laminated growth layer 122 N-type cladding layer 124 P-type cladding layer 126 P-type contact layer 130 Light-emitting layer 132 N-type guide layer 134 Active layer 136 P-type guide layer 140 Insulating layer 150 P-electrode 152 N electrode 160 Cap layer 162 First growth layer 164 Dielectric film 166 Second growth layer 168 Ridge structure 200 Laser resonator

Claims (12)

An edge-emitting semiconductor laser device,
an N-type semiconductor substrate;
a laminated growth layer including an N-type cladding layer, an active layer, a P-type cladding layer, and a P-type contact layer formed on the N-type semiconductor substrate;
Equipped with
m laser resonators (m≧1) are formed in the laminated growth layer,
A semiconductor laser device characterized in that, in each laser resonator, a degree of mixed crystallization in the active layer is greater than a degree of mixed crystallization between the P-type cladding layer and the P-type contact layer. m≧2, and the m laser resonators include n laser resonators having different oscillation wavelengths (2≦n≦m),
The degree of mixed crystallization in the active layer is different among the n laser resonators,
2. The semiconductor laser device according to claim 1, wherein the degree of mixed crystallization from the P-type cladding layer to the P-type contact layer is substantially equal among the n laser resonators. An edge-emitting semiconductor laser device,
an N-type semiconductor substrate;
a laminated growth layer including an N-type cladding layer, an active layer, a P-type cladding layer, and a P-type contact layer formed on the N-type semiconductor substrate;
Equipped with
A plurality of laser resonators are formed in the laminated growth layer,
The degree of mixed crystallization in the active layer is different among the plurality of laser resonators,
A semiconductor laser device characterized in that the degree of mixed crystallization from the P-type cladding layer to the P-type contact layer is substantially equal among the plurality of laser resonators. An edge-emitting semiconductor laser device,
an N-type semiconductor substrate;
a laminated growth layer including an N-type cladding layer, an active layer, a P-type cladding layer, and a P-type contact layer formed on the N-type semiconductor substrate;
Equipped with
A plurality of laser resonators having different oscillation wavelengths are formed in the laminated growth layer,
The degree of mixed crystallization in the active layer is different among the plurality of laser resonators,
A semiconductor laser device characterized in that the film thickness, composition ratio, and impurity concentration of the P-type cladding layer are substantially the same among the plurality of laser resonators. 5. The semiconductor laser device according to claim 2, wherein the difference in oscillation wavelength between the two laser resonators is 2 nm or more. 5. The semiconductor laser device according to claim 1, wherein the active layer includes one of the group consisting of AlGaInP, AlGaAs, InGaAsP, AlGaInAs, AlGaInAsP, and GaInNAs. 5. The semiconductor laser device according to claim 1, wherein the different degrees of mixed crystallization mean that compositional transition lengths confirmed from a TEM image are separated by 0.16 nm or more. . 5. In the active layer, C and Zn impurity concentrations in optical gain regions corresponding to laser resonators having different oscillation wavelengths are higher in the laser resonators having shorter oscillation wavelengths. A semiconductor laser device according to claim 1. A method for manufacturing an edge-emitting semiconductor laser device, the method comprising:
A first step of forming a first growth layer including at least an N-type cladding layer and an active layer on the N-type semiconductor substrate;
a second step of forming a dielectric film on the first growth layer and performing a heat treatment to form a mixed crystal;
a third step of removing the dielectric film and forming a second growth layer including a P-type cladding layer by regrowth;
A manufacturing method characterized by comprising: The second step includes a step of making at least one of the film thickness, film type, and film configuration of the dielectric film different in a plurality of regions,
10. The manufacturing method according to claim 9, further comprising a fourth step of forming a laser resonator in each of the plurality of regions. The second step includes a step of varying the contact ratio between the first growth layer and the dielectric film in a plurality of regions,
10. The manufacturing method according to claim 9, further comprising a fourth step of forming a laser resonator in each of the plurality of regions. The second step is
etching the first growth layer at different depths in a plurality of regions;
forming the dielectric film on the first growth layer after etching;
including;
10. The manufacturing method according to claim 9, further comprising a fourth step of forming a laser resonator in each of the plurality of regions.

Images