Patent Application
20080013581
The above and other objects, advantages and features of the present invention will be more apparent from the following description of certain preferred embodiments taken in conjunction with the accompanying drawings, in which:
The invention will be now described herein with reference to illustrative embodiments. Those skilled in the art will recognize that many alternative embodiments can be accomplished using the teachings of the present invention and that the invention is not limited to the embodiments illustrated for explanatory purposes.
Referring to the accompanying drawings, self-pulsating semiconductor lasers of following embodiments are used, for example, as a light source of an optical disk device such as a DVD device.
In
Furthermore, a block layer BLK is formed on both sides of the mesa structure of the second upper clad layer 109. That is, the block layer BLK is formed to cover the side faces of the mesa structure MS and the etching stop layer 108 in a region where the mesa structure MS is not formed. As will be described later, the block layer BLK functions to constrict the injected electric current injected into the active layer 105 to the mesa structure. Further, the block layer BLK also functions to provide an optical waveguide (horizontal transverse mode) in the X-direction. In the present embodiment, the block layer BLK includes a layer which has a larger band gap than the active layer 105 and has a smaller refractive index than the second upper clad layer 109.
Further, the top face of the second upper clad layer 109 (mesa structure MS) is covered by a second conductivity type cap layer 110. A second conductivity type contact layer 113 is formed on the cap layer 110 and the block layer BLK.
The semiconductor laser element according to an embodiment of the present invention is constituted with a semiconductor laminated structure described above. The emission wavelength is, for example, about 650 nm. Examples of each layer that constitutes such semiconductor laser element will be described below. In the examples below, the first conductivity type is an n-type, and the second conductivity type is a p-type. Naturally, the n-type and the p-type may be reversed. Further, there are cases where (AlxGa1-x)0.5In0.5P is simply written as AlGaInP. In that case, Al composition ratio x is written within parentheses.
Semiconductor substrate 101: n-type GaAs
Buffer layer 102: n-type GaAs; thickness=650 nm; impurity concentration=5×1017 cm−3
Lower clad layer 103: n-type AlGaInP (x=0.7); thickness=1200 nm; impurity concentration=5×1017 cm−3
Lower guide layer 104: AlGaInP (x=0.45); thickness=30 nm
The band gap of the active layer 105 is smaller than those of the guide layer and the clad layer provided in surrounding region. The refractive index of the active layer 105 is larger than those of the guide layer and the clad layer provided in surrounding region. In the present embodiment, the active layer 105 has a multi-quantum well structure in which a plurality of quantum wells are laminated. Each of the wells is isolated by a respective barrier layer. Each well is formed with GaInP, and the thickness thereof is 5.0 nm. Each barrier layer is formed with AlGaInP (x=0.45), and the thickness thereof is 5.0 nm. The compression distortion applied to the wells is adjusted to achieve a desired oscillation wavelength at about 650 nm.
Upper guide layer 106: AlGaInP (x=0.45); thickness=30 nm
First upper clad layer 107: p-type AlGaInP (x=0.7); thickness d=300 nm; impurity concentration=
Etching stop layer 108: p-type AlGaInP (x=0.2); thickness=10 nm; impurity concentration 6×1017 cm−3
Second upper clad layer 109: p-type AlGaInP (x=0.7); thickness=1000 nm; impurity concentration=6×1017 cm−3; width W of the bottom of the mesa structure MS=4.0 μm
Cap layer 110: p-type GaAs; thickness=300 nm; impurity concentration=1.5×1018 cm−3
Contact layer 113: p-type GaAs; thickness=3000 nm; impurity concentration=2×1018 cm−3
The block layer BLK includes an n-type or undoped (AlxGa1-x)0.5In0.5P layer 111 and an n-type GaAs layer 112 formed thereon. The Al composition x may also be 1. In this case, the block layer BLK includes the n-type or undoped AlInP layer 111. The thickness of the AlInP layer 111 (or AlGaInP layer 111) is, for example, 150 nm. The thickness of the n-type GaAs layer 112 is 850 nm, for example, and the impurity density thereof is 3×1018 cm−3, for example. The band gap of the AlInP layer 111 (or AlGaInP layer 111) is larger than that of a light-emitting portion of the active layer 105, and the refractive index thereof is smaller than that of the second upper clad layer 109 (mesa structure MS). That is, the block layer BLK with a small optical absorption coefficient is formed.
In the structure above described, there is a difference generated in the refractive indexes in the X-direction due to the block layer BLK that is formed on the second upper clad layer 109 and on both sides thereof. That is, there is a difference generated in the refractive indexes between the portion that corresponds to the stripe-shaped mesa structure MS and the portion other than the mesa structure MS. This difference is called an effective index difference Δn. The effective refractive index difference Δn related to the optical waveguide in the X-direction depends also on the thickness d of the first upper clad layer 107. In the structure presented in the above-described example, the effective refractive index difference Δn related to the optical waveguide in the X-direction is about 2.0×10−3.
Referring to
The mesa structure is formed so that an edge thereof is placed at the vicinity of the active layer 105, and the width where the electric current is injected in the active layer 105 nearly corresponds to the width W of the bottom face of the mesa structure MS. As a result, in the active layer 105, there is a gain (inverted distribution) generated only in the region that corresponds to the mesa structure MS. In
Further, regarding the optical waveguide, optical confinement (vertical transverse mode) in the Y-direction is achieved by the double heterostructure described above. Meanwhile, optical confinement (horizontal transverse mode) in the X-direction is achieved by the effective refractive index difference Δn described above. More specifically, the light is not confined inside the active layer 105, but it slightly leaks to the clad layer provided in the circumstance thereof due to the tunnel effect. The leaked light senses the block layer BLK with a relatively low refractive index, which is formed in the vicinity of the active layer 105. As a result, the above-described effective refractive index difference Δn is generated in the X-direction, and the light is confined. In this sense, the block layer BLK is also considered to be functioning as an “X-direction optical waveguide mechanism”.
In the active layer 105, the width of the optical waveguide region is larger than the width of the gain region 114. The difference between the optical waveguide region and the gain region 114, i.e. the optical waveguide region on the outer side of the gain region 114, functions as a “saturable absorber region 115”. The self-pulsation is achieved by the saturable absorber region 115. However, the extent of the self-pulsation depends on the volume of the saturable absorber region 115. The volume of the saturable absorber region 115 is determined depending on the size of the optical waveguide region and the size of the gain region 114. The size of the optical waveguide region is determined depending almost on the above-described effective refractive index difference Δn. Meanwhile, the size of the gain region 114 corresponds to the distribution width of the injected electric current that is injected into the active layer 105, and the distribution width of the injected electric current depends not only on the width W of the bottom of the mesa structure MS but also on the temperature.
Under a high-temperature condition, the distribution width of the injected electric current becomes relatively large, because the spread of the hole carrier in the X-direction (hereinafter, referred to as “lateral spread” in some cases) becomes large in the first upper clad layer 107 and the active layer 105 right under the mesa structure MS. Therefore, the gain region 114 becomes relatively large, so that a loss region that can function as the saturable absorber region 115 becomes relatively small. Inversely, under a low-temperature condition, the lateral spread of the injected electric current becomes small, and the gain region 114 becomes relatively small as well. Accordingly, the loss region that can function as the saturable absorber region 115 becomes relatively large.
As described, the volume of the loss region that can function as the saturable absorber region 115 changes depending on the temperature. Therefore, the self-pulsation also exhibits a temperature dependency. For example, when the volume of the loss region that can function as the saturable absorber region 115 becomes too large under a low-temperature condition, the self-pulsation is weakened due to the excessive loss. However, when the gain of the active layer itself is too large with respect to the loss, the gain becomes excessive inversely. In such a case, the self-pulsation may be weakened. In the meantime, under a high-temperature condition, when the volume of the loss region that can function as the saturable absorber region 115 becomes too small, the self-pulsation is weakened due to the excessive gain. However, when the gain of the active layer itself is too small with respect to the loss, the loss becomes excessive inversely. In such a case, the self-pulsation may be weakened. When the self-pulsation is weakened, the optical feedback noise becomes prominent. In order to improve the operation reliability of a semiconductor laser, it is important to design the laser by considering the temperature dependency, so that a stable self-pulsation can be maintained over a wide temperature range (at least −10° C. to 75° C.).
In order to achieve the stable self-pulsation, it is necessary to generate the oscillation itself stably. The oscillation is generated when the gain by an induced radiation exceeds a loss (transmission, absorption, dispersion, etc). Thus, it is preferable to reduce the loss as much as possible. The block layer BLK according to the present embodiment includes the AlInP layer 111 (or AlGaInP layer 111) whose an optical absorption coefficient in the oscillation wavelength region is small. Unlike the related techniques mentioned before, the block layer BLK is formed not only with the GaAs layer that has a characteristic of absorbing the light. As a result, the waveguide path loss is decreased, and the oscillation is enhanced. That is, the threshold current (a current value with which the oscillation is started) is decreased. Though the threshold current tends to increase in accordance with an increase in the temperature, the absolute value of the threshold current is small in this embodiment, so that it is possible to suppress weakening of the oscillation due to an insufficient gain, even under the high-temperature condition.
Further, since the waveguide path loss and the threshold current can be decreased, it is possible to obtain a desired optical output power with a still smaller operating current.
Next, the inventor of the present application investigated the temperature dependency of the self-pulsation. Specifically, the inventor of the present application studied a primary interference index γ and a temperature dependency of the relative intensity noise RIN for the optical feedback noise. When the strong self-pulsation is being obtained, the wavelength chirping of the longitudinal mode becomes large, and the values of γ and RIN are small. Inversely, when the self-pulsation is weakened, the values of γ and RIN are large. It is therefore possible to check whether or not a stable self-pulsation is being obtained by measuring the γ and RIN. As the references indicating the stable self-pulsation, it is required for γ to be 60% or less and RIN to be −110 dB/Hz or less.
Furthermore, the inventor of the present application conducted an experiment, in which each sample having respective thickness d of the first upper clad layer is tested, and obtained the same kind or results for the samples. The thickness d affects the above-described effective refractive index difference Δn that determines the size of the optical waveguide region. When the thickness d becomes small, the effective refractive index difference Δn becomes large, thereby decreasing the size of the optical waveguide region. Meanwhile, when the thickness d becomes large, the effective refractive index difference Δn becomes small, thereby increasing the size of the optical waveguide region.
However, in a case where the thickness d is 180 nm, the values of the γ and RIN are increased again at a high-temperature condition (75° C.). This means that the self-pulsation becomes weakened and the optical feedback noise is increased. In the case where the thickness d is 180 nm, the size of the optical waveguide region is decreased compared to the other cases. In addition, the lateral spread of the injected electric current becomes large under a high-temperature condition, and the gain region 114 becomes relatively large. This results in an excessive gain. Therefore, the volume of the saturable absorber region 115 becomes too small, so that the self-pulsation becomes weakened or stopped.
Further, in a case where the thickness d is 480 nm, the values of the γ and RIN are increased at a low-temperature condition (−10° C.) and at a high-temperature condition (75° C.). This also means that the self-pulsation becomes weakened and the optical feedback noise is increased. In the case where the thickness d is 480 nm, the size of the optical waveguide region is increased compared to the other cases. In addition, the lateral spread of the injected electric current becomes small under the low-temperature condition, and the gain region 114 becomes relatively small. This results in an excessive loss. Therefore, the volume of the saturable absorber region 115 becomes too small, so that the self-pulsation becomes weakened or stopped. Furthermore, at the high-temperature condition, the gain of the active layer 105 itself is decreased due to an influence of the carrier overflow. Therefore, the self-pulsation is also weakened or stopped because of the excessive loss.
As described above, from the view point of the temperature dependency of the volume of the saturable absorber region 115, the thickness d of the first upper clad layer 107 is preferable to be set as a value within the range from 220 nm to 450 nm.
Furthermore, the inventor of the present application conducted experiments by changing the other parameters of samples variously, and obtained same kind of results.
Thereafter, the influence of the carrier concentration (p-concentration) of the first upper clad layer 107 was investigated.
Then, investigated was the width W of the bottom of the mesa structure in the X-direction, which is one of the parameters that determine the distribution of the injected electric current.
Next, an example of a method for manufacturing the above-described semiconductor laser element will be described.
First, as shown in
The lower clad layer 103 (n-type AlGaInP (x=0.7), thickness=1200 nm, the impurity concentration=5×1017 cm−2) is formed on the semiconductor substrate 101 (n-type GaAs) via the buffer layer 102 (n-type GaAs, thickness=650 nm, the impurity concentration=5×1017 cm−3). The multi-quantum well active layer 105 (well layer; GaInP, thickness=5.0 nm; barrier layer: AlGaInP (x=0.45), thickness=5.0 nm u) is formed on the lower clad layer 103 via the lower guide layer 104 (AlGaInP (x=0.45), thickness=30 nm). The compression distortion applied to the well layer is adjusted to have a desired oscillation wavelength at about 650 nm.
Furthermore, the first upper clad layer 107 (p-type AlGaInP (x=0.7)) is formed on the multi-quantum well active layer 105 via the upper guide layer 106 (AlGaInP (x=0.45), thickness=30 nm). The thickness d of the first upper clad layer 107 is within the range from 220 nm, to 450 nm, and the carrier concentration is within the range from 5×1017 cm−3 to 2×1018 cm−3. On the first upper clad layer 107, the second upper clad layer 109 (p-type AlGaInP (x=0.7), thickness=1000 nm, the impurity concentration=6×1017 cm−3) is formed via the etching stop layer 108 (p-type AlGaInP (x=0.2), thickness=10 nm, the impurity concentration=1.5×1017 cm−3). Further, the cap layer 110 (p-type CaAs, thickness=300 nm, is the impurity concentration 1.5×1018 cm−3) is formed on the second upper clad layer 109.
Then, an SiO2 mask 200 is formed in a prescribed region on the cap layer 110 through thermal CVD, photolithography, and etching by hydrofluoric acid. Subsequently, as shown in
Then, as shown in
Next, after removing the SiO2 mask 200, a contact layer 113 (p-type GaAs, thickness=3000 nm, the impurity concentration=2×1018 cm−3) is formed by an epitaxial growth as shown in
In this manner described above, the semiconductor laser element according to the present embodiment can be manufactured. Through the above-described steps, the surface of the multi-quantum well active layer 105 is not to be exposed to the atmosphere. As a result, formation of non-luminescence center (dark defect) on a surface of the multi-quantum well active layer 105 can be prevented. Thus, it is possible to suppress weakening of the self-pulsation due to the insufficient gain of the multi-quantum well active layer 105 itself. Furthermore, the life of the element can be extended since the operating current can be decreased.
In a case that the block layer is formed only with an optical absorbing GaAs layer, the waveguide path loss becomes significant so that it is difficult to have a laser oscillation. As a result, the threshold current is increased, and the operating current becomes high as well. In particular, such problem is significant in the case of a semiconductor laser that includes an active layer made of a GaInP/AlGaInP type material, due to an influence of the carrier overflow that becomes prominent at high temperature.
The block layer BLK according to the present embodiment includes the AlGaInP layer that has a larger band gap than that of the active layer 105. That is, there is formed a block layer BLK whose optical absorption coefficient in the oscillation wavelength region is small. Because of such block layer BLK, the waveguide path loss is decreased, and oscillation can be easily generated. As a result, the threshold current is decreased, so that the slope efficiency is improved and the operating current can be decreased. The threshold current tends to increase in accordance with an increase in the temperature. However, the absolute value of the threshold current is decreased, so that it is possible to suppress weakening of the self-pulsation due to an insufficient gain even under a high-temperature condition. Further, since the threshold current can be decreased and the slope efficiency is improved, it is possible to obtain a desired optical output power with a still smaller operating current (see
Furthermore, in the present embodiment, the thickness d of the first upper clad layer 107 is designed to have a value within the range from 220 nm to 450 nm. In this case, the interference index γ and the relative intensity noise RIN can be suppressed to sufficiently low values over a wide temperature range (−10° C. to 75° C.) as shown in
With the above-described structure, “the loss by the saturable absorber layer” and “the gain of the active layer itself” which may vary depending on temperature can be balanced properly over a wide temperature range. Thus, a gain characteristic suited for the self-pulsation can be achieved over a wide operating temperature range. Such oscillation characteristic can be achieved with a low threshold current and high slope efficiency, so that an element with an excellent long-term reliability can be obtained. Furthermore, it is possible with the above-described structure to decrease the in-plane variations in the temperature dependency of the self-pulsation intensity and to obtain a high reproducibility. This enables a manufacture yield to be maintained high and stable, thereby improving the productivity.
The block layer BLK according to the present embodiment is constituted only with a first-conductive (AlxGa1-x)0.5In0.5P layer 120 without an GaAs layer. For example, the block layer BLK includes an n-type AlInP layer 120 (x=1). The thickness of the n-type AlInP layer 120 is 1000 nm, for example, and the impurity density is 3×1018 cm−3, for example. With such structure, it is possible to obtain the same effects as those of the first embodiment. The manufacturing method of the semiconductor laser element according to the present embodiment is the same as that of the first embodiment.
In the case presented in the first embodiment, the semiconductor laminated structure on the semiconductor substrate 101 is formed with a GaInP/AlGaInP type material, and the emission wavelength is about 650 nm. The present invention is also effective for a self-pulsating semiconductor laser in which the semiconductor laminated structure is formed with a GaAs/AlGaAs type material, and the emission wavelength is about 780 nm.
Semiconductor substrate 301: n-type GaAs
Buffer layer 302: n-type GaAs; thickness=650 nm; impurity concentration=5×1017 cm−3
Lower clad layer 303: n-type AlGaAs (x=0.5); thickness=1200 nm impurity concentration=1×1018 cm−3
Lower guide layer 304: AlGaAs (x=0.34); thickness=80 nm
Multi-quantum well active layer 305: well layer (AlGaAs (x=0.05), thickness=4.8 nm); barrier layer (AlGaAs (x=0.34), thickness=5.0 nm)
Upper guide layer 306: AlGaAs (x=0.34); thickness=80 nm
First upper clad layer 307: p-type AlGaAs (x=0.5); thickness d=250 nm; impurity concentration=5×17 cm−3
Etching stop layer 308: p-type AlGaAs (x=0.2); thickness=10 nm; impurity concentration=5×1017 cm−3
Second upper clad layer 309: p-type AlGaAs (x=0.5); thickness=1000 nm; impurity concentration=5×1017 cm−3; width W of the bottom of the mesa structure MS=4.5 μm
Cap layer 310: p-type GaAs; thickness=300 nm; impurity concentration=1.5×1018 cm−3
Contact layer 313: p-type GaAs; thickness 3000 nm n; impurity concentration=2×1018 cm−3
Like the first embodiment, the block layer BLK includes an n-type or undoped AlInP layer 311 (or AlGaInP layer 311) and an n-type GaAs layer 312 formed thereon. The thickness of the AlInP layer 311 (or AlGaInP layer 311) is 150 nm, for example. The thickness of the n-type GaAs layer 312 is 850 nm, for example, and the impurity density thereof is 3×1018 cm−3, for example. Alternatively, the block layer BLK may be constituted only with the n-type AlGaInP layer as in the second embodiment. With the above-described structures, the effective refractive index difference Δn in the X-direction becomes about 2.5×10−3.
With the structure according to the present embodiment, it is also possible to obtain same effects as those of the first embodiment. That is, the stable self-pulsation can be maintained over a wide temperature range through setting the parameter d appropriately. Further, the semiconductor laser element according to the present embodiment can be manufactured by the method same as that of the first embodiment.
In the self-pulsating semiconductor laser according to the present invention, a plurality of light sources having different emission wavelengths may be integrated monolithically. For example, a first light source which is presented in the first embodiment (whose emission wavelength is about 650 nm) and a second light source which is presented in the third embodiment (whose emission wavelength is about 280 nm) may be formed monolithically on a semiconductor substrate. In this case, the structure shown in
First, as shown in
Then, after removing the SiO2 mask 401, each of the above-described semiconductor layers 302-310 is formed in order as shown in
Then, as shown in
Then, as shown in
Thereafter, the block layer BLK is formed on both sides of each mesa structure MS1 and mesa structure MS2 in each of the first and second regions. The method for forming each block layer BLK is the same as that of the above-described embodiments. In this manner described above, the first light source and the second light source having different emission wavelengths are formed monolithically on the semiconductor substrate 101. The effective refractive index difference Δn is about 2.0×10−3 for the first light source (emission wavelength=650 nm) and about 2.5×10−3 for the second light source (emission wavelength=780 nm).
With the structure according to the present embodiment, it is also possible to obtain the same effects as those of the above-described embodiments. That is, the stable self-pulsation can be maintained over the wide temperature range for each of the first and second light sources. Further, the operating current can be reduced for each of the first and second light sources.
In the fourth embodiment, material of the second upper clad layer 109 is a p-type AlGaInP (x=0.7), and that of the second upper clad layer 309 is a p-type AlGaAs (x=0.5). Even though the different materials are used for those layers, the mesa structures MS1 and MS2 can be formed at once by a proper etching process. In cases where such proper etching process cannot be employed, it is preferable to form the second upper clad layers with a same material for all light sources.
For example, the p-type AlGaInP (x=0.7) that is the same material as that of the first region may be used for the second upper clad layer in the second region. A cross sectional view corresponding to a part of manufacturing steps in this case is shown in
Buffer layer 502: n-type GaAs; thickness=650 nm; impurity concentration=5×1017 cm−3
Lower clad layer 503: n-type AlGaAs (x=0.65); thickness=1200 nm; impurity concentration=1×1018 cm−3
Lower guide layer 504: AlGaAs (x=0.4); thickness=5 nm
Multi-quantum well active layer 505: well layer (AlGaAs (x=0.04), thickness=4.5 nm); barrier layer (AlGaAs (x=0.4), thickness=5.0 nm)
Upper guide layer 506; AlGaAs (x=0.4); thickness=5 nm
First upper clad layer 507: p-type AlGaAs (x=0.65); thickness d=250 nm; impurity concentration=5×1017 cm−3
Etching stop layer 508: p-type AlGaAs (x=0.2); thickness=10 nm; impurity concentration 6×1017 cm−3
Second upper clad layer 509: p-type AlGaInP (x=0.7); thickness=1000 nm, impurity concentration=6×1017 cm−3;
Cap layer 510: p-type GaAs; thickness=300 nm; impurity concentration=1.5×1018 cm−3
As described above, the p-type AlGaInP (x=0.7) that is the same material as that of the first region is used for the second upper clad layer 509 in the second region. As a result, the mesa structures MS1 and MS2 can be formed by etching the first and second regions at the same time. Other manufacturing steps are the same as those described in the fourth embodiment. It is also possible with the present embodiment to obtain the same effects as those of the fourth embodiment.
As described above, the temperature dependency of the self-pulsation in a self-pulsating semiconductor laser can be fully considered in embodiments of the present invention. As a result, the stable self-pulsation can be maintained over a wide temperature range. Since the optical feedback noise can be suppressed finely over the entire range of the required temperatures, the operation reliability can be improved. Further, the operating current can be decreased, so that the long-term reliability can be improved. The present invention can be applied not only to the self-pulsating semiconductor lasers of GaInP/AlGaInP, GaAs/AlGaAs types but also to the other types of self-pulsating semiconductor laser such as InGaAsP/InP, GaN, ZnSe type.
It is apparent that the present invention is not limited to the above embodiments, but may be modified and changed without departing from the scope and spirit of the invention.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2006-190211 | Jul 2006 | JP | national |
