JP2010219476A - Light emitting device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a light emitting device having a forbidden band width of an active layer whose temperature variation is reduced. <P>SOLUTION: The light emitting deice includes an active layer provided with a quantum dot formed of a first material whose forbidden band width becomes wider with temperature rise, and a barrier layer formed of a second material whose forbidden band width becomes narrower with temperature rise and having the quantum dot embedded therein. <P>COPYRIGHT: (C)2010,JPO&amp;INPIT

Description

  The present invention relates to a light emitting device.

  In optical communication, a semiconductor laser that emits light in a wavelength region of 1.3 μm to 1.6 μm where the loss of an optical fiber that is a transmission path is small is used as a light source. Such a semiconductor laser is formed of InGaAsP and has good characteristics including reliability.

  On the other hand, as the amount of information to be transmitted becomes enormous, wavelength division multiplexing (WDM) using a large number of wavelengths has been used instead of the conventional optical communication using a single wavelength.

  Semiconductor lasers used in wavelength division multiplexing (WDM) include a Fabry-Perot (FP) laser and a distributed feedback (DFB) laser.

  The FP laser is used for CWDM (coarse wavelength division multiplexing) with a wide wavelength interval. On the other hand, the DFB laser is used in both DWDM (dense wavelength division multiplexing) and CWDM where the wavelength interval is narrow.

JP-A-8-316578 JP-A-9-219561

  For WDM, it is not preferable that the wavelength of the light source is not constant and changes with time. However, the emission wavelength of the semiconductor laser (the wavelength of light emitted from the light emitting element) varies depending on the variation of the element temperature.

  For example, the emission wavelength of the FP laser changes when the forbidden band width of the active layer fluctuates due to changes in element temperature. For example, the emission wavelength of the FP laser changes at a large rate of about 0.5 nm / K near the emission wavelength of 1.3 μm.

  Such temperature change of the emission wavelength can be avoided by mounting the FP laser on the Peltier element and controlling the temperature. However, providing a Peltier element and its control unit in an optical transmission device that uses an FP laser as a light source has a problem of complicating the configuration of the device.

  On the other hand, the emission wavelength of the DFB laser changes when the refractive index of the light guide layer provided with the diffraction grating fluctuates due to a change in element temperature. However, the temperature change of the refractive index of the light guide layer is much smaller than the temperature change of the forbidden band width of the active layer. For this reason, the temperature change of the emission wavelength of the DFB laser is much smaller than the temperature change of the emission wavelength of the FP laser. However, the DFB laser has a problem that its manufacturing process is complicated.

  When the element temperature changes, the gain band of the DFB laser tends to dissociate from the emission wavelength determined by the period of the diffraction grating. For this reason, it is preferable that the gain band of the DFB laser is wide. However, when the gain band of the DFB laser is widened, there is a problem that the light emission efficiency of the DFB laser is lowered.

  As described above, various problems occur when an attempt is made to reduce the temperature change of the emission wavelength of the semiconductor laser.

  However, if the active layer can be formed of a material having a small band gap temperature change, the temperature change of the emission wavelength of the semiconductor laser can be reduced without facing these problems.

  Accordingly, an object of the present invention is to provide a semiconductor light emitting device in which an active layer (light emitting layer) is formed by a material layer having a temperature change of the forbidden band width smaller than that of a material such as InGaAsP conventionally used for forming an active layer. It is to be.

It has been proposed to form an active layer of a semiconductor laser using Hg _0.4 Cd _0.6 Te or GaInTlP whose band gap does not change with temperature (Patent Documents 1 and 2).

  However, these semiconductor lasers have a problem that the substrate used for the formation thereof is special and that the manufacture itself is difficult. For this reason, these semiconductor lasers have not yet been put into practical use.

  In order to achieve the above object, the first aspect of the light emitting device includes a quantum dot formed of a first material whose width of the forbidden band is widened when the temperature is increased, and a width of the forbidden band when the temperature is increased. Is a light emitting device including an active layer formed of a second material that becomes narrow and having a barrier layer in which the quantum dots are embedded.

The second side surface of the present light emitting device is a light emitting device including an active layer having quantum dots formed of Hg _1-y Cd _y Te (0.4 <y <0.6).

  According to this light emitting device, the temperature change of the forbidden band width of the light emitting layer can be made smaller than that of the conventional light emitting device.

FIG. 3 is an enlarged cross-sectional view of a part of an active layer that forms the semiconductor laser according to the first embodiment. FIG. 5 is an enlarged cross-sectional view of a part of an active layer forming a semiconductor laser according to a second embodiment. FIG. 6 is an enlarged cross-sectional view of a part of an active layer forming a semiconductor laser according to a third embodiment. 1 is a plan view illustrating a configuration of a semiconductor laser according to Example 1. FIG. It is the figure which looked at the cross section in the vv line | wire of FIG. 4 from the direction of the arrow. 6 is a diagram illustrating a cross section of an active layer of a semiconductor laser according to Example 1. FIG. It is a figure explaining the temperature change of the emission wavelength of the semiconductor laser regarding Example 1. FIG. (Example 1) which is a figure explaining the relationship between the height of a quantum dot, and the light emission wavelength, and the relationship between the height of a quantum dot, and the temperature coefficient of a light emission wavelength. The emission wavelength and its temperature coefficient when the quantum dots formed in _{Hg 1-y Cd y Te (} 0 ≦ y ≦ 0.4), is a diagram for explaining a function of the height of the quantum dot (Example 1) . It is a top view explaining the model of a quantum dot used for the simulation of the temperature change of light emission wavelength. It is the figure which looked at the cross section in the xi-xi line of FIG. 10 from the direction of the arrow. It is a figure explaining distribution of potential used for simulation. FIG. 6 is a process cross-sectional view illustrating the method for manufacturing the active layer in Example 1. 6 is a plan view illustrating a cross section of a quantum dot model used for obtaining a temperature change of an emission wavelength of Example 2. FIG. FIG. 6 is a diagram for explaining a potential distribution of a model related to Example 2. It is a figure explaining the temperature change of the emission wavelength of the semiconductor laser regarding Example 2. FIG. It is a figure explaining the relationship between the thickness of a HgTe layer, and a light emission wavelength, and the relationship between the thickness of a HgTe layer, and the temperature coefficient of a light emission wavelength. It is a figure explaining the relationship between the thickness of an InAs layer, and a light emission wavelength, and the relationship between the thickness of an InAs layer, and the temperature coefficient of a light emission wavelength. FIG. 9 is a process cross-sectional view illustrating the manufacturing method of the semiconductor laser of Example 2 (No. 1). It is process sectional drawing explaining the manufacturing method of the semiconductor laser of Example 2 (the 2). FIG. 6 is a diagram for explaining a temperature change in the emission wavelength of a semiconductor laser relating to Example 3. (Example 3) which is a figure explaining the relationship between the quantum dot height and the light emission wavelength, and the relationship between the quantum dot height and the temperature coefficient of the light emission wavelength.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, the technical scope of the present invention is not limited to these embodiments, but extends to the matters described in the claims and equivalents thereof. In addition, the same code | symbol is attached | subjected to the corresponding part even if drawings differ, The description is abbreviate | omitted.

(Embodiment 1)
(1) Configuration FIG. 1 is an enlarged cross-sectional view of a part of an active layer forming the semiconductor laser of the present embodiment.

  The active layer of this semiconductor laser is formed by stacking quantum dots 2 and quantum dot layers 5 (see FIG. 1) having barrier layers (barrier layers) 4 in which the quantum dots 2 are embedded. Here, the lower (substrate side) barrier layer 4a and the upper (opposite side substrate) barrier layer 4b of the quantum dots 2 may be formed of the same material. Alternatively, both barrier layers 4a and 4b may be formed of different materials. The barrier layers 4a and 4b may be formed of a single material. Alternatively, the barrier layers 4a and 4b may be formed by combining a plurality of portions formed of different materials. The barrier layer is also called an intermediate layer or a buried layer. The upper barrier layer is sometimes called a cap layer.

  Here, the quantum dots 2 are formed of a first material (for example, HgTe) whose width of the forbidden band becomes wider as the temperature rises. On the other hand, the barrier layer 4 is formed of a second material (for example, GaAs) whose width of the forbidden band is narrowed when the temperature rises.

  Here, the first material may be a semiconductor or a semimetal. On the other hand, a semiconductor is preferable as the second material.

  Note that the forbidden band width of the material is the energy difference between the bottom of the conduction band and the top of the valence band.

  By the way, the quantum dots 2 shown in FIG. 1 are self-organized quantum dots formed by the Stranski-Krastanov mode. Therefore, the quantum dots 2 are formed on the wetting layer 6.

  However, the quantum dots 2 may be quantum dots other than such self-organized quantum dots. For example, the quantum dot 2 may be a quantum dot formed by combining an electron beam exposure method and etching.

(2) Operation
When current is supplied to the semiconductor laser, carriers (electrons and holes) are injected into the active layer and move to the quantum dots 2.

  The electrons that have moved to the quantum dot 2 transition to a quantum level formed in the conduction band of the quantum dot 2. Similarly, holes that have moved to the quantum dot 2 transition to a quantum level formed in the valence band of the quantum dot 2.

  Then, the electrons that have transitioned to the quantum level of the conduction band recombine with the holes that have transitioned to the quantum level of the valence band by stimulated emission to generate laser light. Note that laser light generated in such a process is generated by the transition of electrons from the ground level of the conduction band to the ground level of the valence band.

  Therefore, the energy difference between the ground level formed in the conduction band of the quantum dot 2 and the ground level formed in the valence band of the quantum dot 2 (hereinafter referred to as the energy difference between the quantum levels) is activated. This is called the forbidden bandwidth of the layer.

  By the way, a part of the wave function of electrons (and holes) staying at the quantum level of the quantum dot 2 leaks into the barrier layer 4.

  For this reason, the quantum level formed in the quantum dot 2 is influenced not only by the physical properties of the material forming the quantum dot 2 (for example, HgTe) but also by the physical properties of the material forming the barrier layer 4 (for example, GaAs). receive.

  For example, if the forbidden band width of the material forming the quantum dots 2 increases, the energy difference between the quantum levels formed in the quantum dots 2 also increases. Moreover, if the forbidden band width of the material forming the barrier layer 4 (for example, GaAs) is widened, the energy difference between the quantum levels formed in the barrier layer 4 is also widened.

  Conversely, if the forbidden band width of the material forming the quantum dots 2 is narrowed, the energy difference between the quantum levels is also narrowed. Further, if the forbidden band width of the material forming the barrier layer 4 is narrowed, the energy difference between the quantum levels is also narrowed.

  By the way, as described above, the forbidden band width of the material forming the quantum dots 2 becomes wider as the temperature rises. On the other hand, the forbidden band width of the material layer forming the barrier layer 4 becomes narrower as the temperature rises.

  For this reason, in this Embodiment, the material which forms the quantum dot 2 tries to widen the energy difference between quantum levels, when temperature rises. On the other hand, the material forming the barrier layer 4 tends to narrow the energy difference between the quantum levels when the temperature rises. Therefore, the effect of both materials on the energy difference between the quantum levels with respect to changes in temperature is offset. Therefore, the temperature change of the forbidden band width of the active layer is smaller than the temperature change of the forbidden band width of each material forming the active layer.

  On the other hand, the active layer of a conventional semiconductor laser is usually formed of only a material such as GaInAsP or GaAs, whose forbidden band width becomes narrower as the temperature rises. For this reason, the forbidden band width of the conventional active layer changes in temperature in the same manner as the forbidden band width of the material forming these active layers.

  Therefore, according to the present embodiment, the active layer of the light emitting device is formed by a material layer (quantum dot layer) whose temperature change (absolute value) of the forbidden band is smaller than that of a conventional material (such as InGaAsP). can do.

  Note that “the temperature change of the forbidden bandwidth” means the magnitude of the absolute value of the increment of the forbidden bandwidth when the temperature changes. Accordingly, the “temperature change of the forbidden bandwidth” when the increment of the forbidden bandwidth due to the temperature change is −2 meV is more than the “temperature change of the forbidden bandwidth” when the increment of the forbidden bandwidth due to the temperature change is +1 meV. large. The magnitude of “temperature change in emission wavelength” also means the magnitude of the absolute value of the increase in emission wavelength when the temperature changes.

  By the way, as described above, the quantum dots 2 are formed of a material whose rate of change of the forbidden band with respect to temperature (hereinafter referred to as a temperature coefficient of the forbidden band) is positive. However, the temperature coefficient of most materials (InGaAsP and GaAs) forming a semiconductor laser is negative.

The few materials with positive temperature coefficient include HgTe, Hg _1-y Cd _y Te (0 ≦ y ≦ 0.4), HgSe, and Hg _1-y Cd _y Se (0 ≦ y <0.8). There is. However, the lattice constants of these materials are greatly different from the lattice constants of GaAs and InP widely used as semiconductor substrates. For example, the lattice constant of HgTe is 0.6429 nm. On the other hand, the lattice constant of GaAs is 0.5653 nm.

  Such lattice mismatching is generally undesirable because it causes crystal defects in the active layer and degrades the characteristics of the semiconductor laser.

  However, quantum dots are formed using such lattice mismatch. In addition, crystal defects do not occur in the active layer formed of quantum dots. Therefore, even if this semiconductor laser is manufactured using a GaAs substrate or an InP substrate, characteristics such as light emission efficiency do not deteriorate.

(Embodiment 2)
FIG. 2 is an enlarged cross-sectional view of a part of the active layer forming the semiconductor laser of the present embodiment.

  The active layer of the semiconductor laser is also formed by stacking quantum dots 8 and quantum dot layers 9 each having a barrier layer (barrier layer) 4 in which the quantum dots 8 are embedded.

  However, the quantum dot 8 has a first portion 10 and a second portion 12.

  The first portion 10 forming the present quantum dot 8 is formed of a material (for example, InAs) whose width of the forbidden band is narrowed when the temperature rises.

Further, the second portion 12 forming the present quantum dot 8 is formed of a material (for example, Hg _1-y Cd _y Te (0 ≦ y ≦ 0.4)) whose width of the forbidden band becomes wider as the temperature rises. Has been.

  The barrier layer 4 in which the quantum dots 8 are embedded is formed of a material (for example, GaAs) whose width of the forbidden band becomes narrower as the temperature rises, as in the first embodiment.

  Therefore, in the present quantum dot 8, the negative temperature coefficient of the first portion 10 and the negative temperature coefficient of the barrier layer 4 cancel each other out of the positive temperature coefficient of the second portion 10.

  For this reason, the temperature change of the energy difference between the quantum levels generated in the quantum dots 8 is smaller than the temperature change of the forbidden band width of the active layer formed only of the material having a positive temperature coefficient (for example, GaInAsP). .

  Therefore, according to the present embodiment, the active layer of the light emitting device is formed by a material layer (quantum dot layer) whose temperature change (absolute value) of the forbidden band is smaller than that of a conventional material (such as InGaAsP). can do.

  2 is formed on the semiconductor layer in which the first wetting layer 14 and the second wetting layer 16 are laminated.

  Here, the first wetting layer 14 is formed of a material that forms the first portion 10. Further, the second wetting layer 16 is formed of a material that forms the second portion 12.

  In the example shown in FIG. 2, the second portion is stacked on the first portion. However, the first part may be stacked on the second part.

(Embodiment 3)
FIG. 3 is an enlarged cross-sectional view of a part of the active layer forming the semiconductor laser of the present embodiment.

  As shown in FIG. 3, the active layer of this semiconductor laser is also formed by stacking quantum dots 18 and quantum dot layers 21 each having a barrier layer (barrier layer) 20 in which the quantum dots 18 are embedded.

Here, the quantum dot 18 is formed of Hg _1-y Cd _y Te (0.4 <y <0.6).

The temperature coefficient of Hg _1-y Cd _y Te becomes substantially zero when the composition y of Cd is between 0.4 and 0.6.

  For this reason, the temperature change of the energy difference between the quantum levels generated in the quantum dots 20 is smaller than the temperature change of the forbidden band width of the active layer formed only of the material having a positive temperature coefficient (for example, GaInAsP). .

  Therefore, according to the present embodiment, the active layer of the light emitting device is formed by a material layer (quantum dot layer) whose temperature change (absolute value) of the forbidden band is smaller than that of a conventional material (such as InGaAsP). can do.

_{Incidentally,} Hg _1-y Cd _y Te to form a quantum well structure or a bulk single layer by (0.4 <y <0.6), it can be an active layer of a semiconductor laser.

  However, such an active layer has a large amount of crystal defects due to lattice mismatch with the substrate. Therefore, even if a semiconductor laser is formed using such an active layer, good characteristics cannot be obtained.

(1) Configuration (i) Overall Configuration FIG. 4 is a plan view for explaining the configuration of the semiconductor laser of this example. FIG. 5 is a view of the cross section taken along the line vv of FIG. 4 as viewed from the direction of the arrow.

  The semiconductor laser 24 has an n-type GaAs substrate 26 and an n-type AlGaAs lower clad layer 28 formed on the n-type GaAs substrate 26 (see FIG. 5).

  The semiconductor laser 24 also has an active layer 30 formed on the lower cladding layer 28. Here, the active layer 30 is even formed by the quantum dots 2 described with reference to FIG.

  The semiconductor laser 24 also has an upper clad layer 32 made of p-type AlGaAs formed on the active layer 30.

  The semiconductor laser 24 has a p-type GaAs electrode layer 36 formed on the upper cladding layer 32.

  The semiconductor laser 24 has a protective film 38 made of SiN that is formed on a band-like mesa 34 including the upper cladding layer 32 and the electrode layer 36 and opens on the electrode layer 36.

  The semiconductor laser 24 has an upper electrode 40 formed on the electrode layer 36 and a lower electrode 42 formed on the back surface of the n-type GaAs substrate 26.

  The semiconductor laser 24 has end faces 44 and 46 formed by cleavage. These end faces 44 and 46 function as a reflecting mirror and form an optical resonator.

  In the semiconductor laser 24, an active layer 30 is disposed inside the optical resonator, and an FP laser is formed.

(Ii) Active Layer FIG. 6 is a diagram for explaining a cross section of the active layer 30 of the semiconductor laser 24.

  The active layer 30 is formed by stacking a plurality of quantum dot layers 48.

  The configuration of the quantum dot layer 48 is substantially the same as the configuration of the quantum dot layer 5 of the first embodiment (see FIG. 1).

  That is, the quantum dot layer 48 includes the quantum dots 2 formed of HgTe (or HgCdTe) and the GaAs barrier layer (barrier layer) 4. The quantum dots 2 are formed on the wetting layer 6.

Here, the temperature coefficient of the forbidden band width of HgTe is 5.6 × 10 ⁻⁴ eV / K. On the other hand, the temperature coefficient of the forbidden band width of GaAs is −5.0 × 10 ⁻⁴ eV / K.

  Thus, the quantum dots 2 are formed of a material (HgTe or HgCdTe) whose width of the forbidden band becomes wider as the temperature rises. On the other hand, the barrier layer 4 is made of a material (for example, GaAs) whose width of the forbidden band is narrowed when the temperature rises.

  For this reason, the effect of each material on the quantum level is offset, and the temperature change (absolute value) of the emission wavelength of the semiconductor laser 24 is reduced.

(2) Operation The semiconductor laser 24 is driven by connecting a power source (not shown) between the upper electrode 40 and the lower electrode 42. A current supplied from the power source to the semiconductor laser 24 is injected into the active layer 30. As a result, an optical gain is generated. Then, the spontaneous emission light amplified by this optical gain is fed back by the optical resonator formed by the end faces 44 and 46 to become laser light.

(I) Temperature change of emission wavelength (HgTe quantum dot active layer)
FIG. 7 is a diagram for explaining the temperature change of the emission wavelength of the semiconductor laser 24. The horizontal axis is temperature. The vertical axis represents the emission wavelength. The temperature change of the emission wavelength shown in FIG. 7 is a result obtained by simulation. The simulation method will be described in “(iii) Simulation” below.

  Points indicated by “x” are simulation results when the height and diameter of the quantum dots 2 are 1.8 nm and 18 nm, respectively. In addition, the height of the quantum dot 2 is a distance from the bottom of the wetting layer 6 to the top of the quantum dot 2.

  As shown in FIG. 7, the emission wavelength of the semiconductor laser 24 is approximately 1.3 μm regardless of the temperature. This emission wavelength is suitable for optical fiber communication.

  Here, the rate of change of the emission wavelength with respect to the temperature (hereinafter referred to as the temperature coefficient of the emission wavelength) is −0.04 nm / K. This value is much smaller (absolute value) than the temperature coefficient of the emission wavelength of the conventional FP laser in which the active layer is made of InGaAsP + 0.5 nm / K.

  The reason why the temperature coefficient of the emission wavelength decreases in this manner is that the temperature change of the forbidden band width of HgTe forming the quantum dots 2 and the temperature change of the forbidden band width of GaAs forming the barrier layer 4 are offset. .

  FIG. 8 is a diagram illustrating the relationship between the height of the quantum dot 2 and the emission wavelength, and the relationship between the height of the quantum dot 2 and the temperature coefficient of the emission wavelength. The horizontal axis represents the height of the quantum dot 2. The left vertical axis represents the emission wavelength of the semiconductor laser 24. The right vertical axis represents the temperature coefficient of the emission wavelength of the semiconductor laser 24.

  The dots indicated by “●” indicate the emission wavelength. The point indicated by “◯” indicates the temperature coefficient of the emission wavelength.

  As is clear from the data indicated by “◯”, the temperature coefficient of the emission wavelength decreases as the quantum dot 2 increases, and eventually becomes a negative value.

  When the quantum dot 2 becomes high, the ratio of the wave function distributed in the quantum dot increases. As a result, the influence of HgTe having a positive temperature coefficient of the forbidden bandwidth is increased. For this reason, as the quantum dot increases, the temperature coefficient of the emission wavelength decreases and eventually becomes a negative value.

  FIG. 8 shows that when the height of the quantum dot 2 is 1.5 nm or more and 2.0 nm or less, the temperature coefficient of the emission wavelength becomes a small value of −0.1 nm / K to 0.1 nm / K. Yes.

  As is clear from the data indicated by “●”, when the height of the quantum dot 2 is 1.5 nm or more and 2.0 nm or less, the emission wavelength is suitable for optical fiber communication of 1.2 μm to 1.4 μm. Value.

  These results indicate that the height of the quantum dots 2 formed of HgTe is preferably 1.5 nm or more and 2.0 nm or less.

  Moreover, if the height of the quantum dot 2 is 1.6 nm or more and 1.9 nm or less, the temperature change (absolute value) of the emission wavelength becomes smaller, and such a wavelength range is more preferable.

  The data shown in FIG. 8 is a simulation result obtained when the aspect ratio (the ratio of the height to the diameter) of the quantum dots 2 is 10%. All the data described below are results obtained when the aspect ratio is 10%.

(Ii) Temperature change of emission wavelength (HgCdTe quantum dot active layer)
FIG. 9 is a diagram for explaining the emission wavelength and the temperature coefficient thereof as a function of the height of the quantum dot 2 when the quantum dot 2 is formed of Hg _1-y Cd _y Te (0 ≦ y ≦ 0.4). is there. It is assumed that the barrier layer 4 is made of GaAs.

  The horizontal axis represents the height of the quantum dot 2. The left vertical axis represents the emission wavelength of the semiconductor laser 24. The right vertical axis represents the temperature coefficient of the emission wavelength.

  The point indicated by “●” indicates the emission wavelength when the quantum dots 2 are formed of HgTe. The point indicated by “◯” indicates the temperature coefficient of the emission wavelength when the quantum dots 2 are formed of HgTe. These data are the same as the data shown in FIG.

On the other hand, the point indicated by “▲” indicates the emission wavelength when the quantum dot 2 is formed of Hg _0.6 Cd _0.4 Te. The point indicated by “Δ” indicates the temperature coefficient of the emission wavelength when the quantum dot 2 is formed of Hg _0.6 Cd _0.4 Te.

  FIG. 9 shows that when the Cd composition ratio y is 0.0 or more and 0.4 or less, the temperature coefficient of the emission wavelength is adjusted to −0.1 nm / K or more and 0.1 nm / by adjusting the height of the quantum dot 2. It shows that it can be made K or less. At this time, the height of the quantum dots 2 is adjusted between 1.5 nm and 3.0 nm according to the Cd composition ratio y.

  In this case, the emission wavelength of the semiconductor laser becomes a value suitable for optical fiber communication from 1.2 μm to 1.6 μm (see FIG. 9).

The above results indicate that the height of the quantum dots 2 formed of Hg _1-y Cd _y Te (0 ≦ y ≦ 0.4) is preferably 1.5 nm or more and 3.0 nm or less. Furthermore, the height of the quantum dots 2 is preferably 2.0 nm or more and 2.5 nm or less.

(Iii) Simulation-Model and Solution-
FIG. 10 is a plan view for explaining a quantum dot model used for the simulation of the temperature change of the emission wavelength. FIG. 11 is a cross-sectional view taken along line xi-xi in FIG. 10 as viewed from the direction of the arrow.

  As shown in FIGS. 10 and 11, the quantum dot model used in this simulation is a disk 50 having a diameter D and a height H.

  Here, the diameter D of the disk 50 is the diameter of the quantum dot 2. Further, the height H of the disk 50 is the height of the quantum dots 2.

  When simulation is performed on electrons in the conduction band, a potential corresponding to the bottom of the conduction band of the material (for example, GaAs) forming the barrier layer 4 is assigned to the outside of the disk 50. On the other hand, a potential corresponding to the bottom of the conduction band of the material forming the quantum dots 2 (for example, HgTe) is assigned to the inside of the disk 50.

  Similarly, when a simulation is performed on holes in the valence band, a potential corresponding to the top of the valence band of the material forming the barrier layer 4 (for example, GaAs) is present outside the disk 50. Assigned. On the other hand, a potential corresponding to the top of the valence band of the material forming the quantum dots 2 (for example, HgTe) is assigned to the inside of the disk 50.

  Note that the actual quantum dots 2 often grow on the wetting layer 6 (see FIG. 1). However, the quantum level generated in the quantum dots 2 is hardly affected by the wetting layer 6. Therefore, in this model, the wetting layer 6 is not considered.

  For such a model, a wave equation approximating the Schroedinger equation by effective mass approximation is solved by numerical analysis (for example, Euler method). Then, quantum levels (including ground levels) formed inside the disk 50 are obtained for each of electrons in the conduction band and holes in the valence band.

  The emission wavelength of the present semiconductor laser is obtained from the energy difference between the ground levels obtained in this way.

  Here, in this simulation, the temperature change of the emission wavelength is obtained on the assumption that the forbidden band width of the material forming the quantum dots 2 and the material forming the barrier layer 4 changes according to the temperature.

  In addition, the cross section of an actual quantum dot is not a rectangle as shown in FIG. However, the quantum level obtained on the assumption that the cross section is rectangular and the actual quantum level substantially coincide.

  Further, the ground quantum level formed in the quantum dot 2 (disk 50) is substantially determined by the height H of the quantum dot 2. Therefore, even if the aspect ratio (= H / D) is changed, the obtained result is hardly changed. Therefore, in this simulation, the aspect ratio is fixed at 10%.

-Potential distribution-
FIG. 12 is a diagram for explaining the potential distribution used in this simulation.

  The horizontal axis is the position coordinate z in the direction perpendicular to the substrate (direction perpendicular to the bottom surface of the disk 50). The vertical axis represents the electron potential. Therefore, the vertical axis becomes the hole potential by changing the sign.

  FIG. 12 shows a potential when the quantum dot 2 having a height of 1.8 nm is formed of HgTe and the barrier layer 4 is formed of GaAs.

  FIG. 12 shows a potential distribution 52 formed by the bottom of the conduction band and a potential distribution 54 formed by the top of the valence band. Here, since HgTe is a semimetal, the bottom 52 of the conduction band and the top of the valence band 54 are reversed in the quantum dot 2.

  FIG. 12 also shows the conduction band ground level 56 and its wave function 58. FIG. 12 also shows the valence band ground level 60 and its wave function 62.

  As shown in FIG. 12, the wave functions 58 and 62 leak into the barrier layer 4. This leakage reduces the temperature change of the emission wavelength.

-Parameter-
The parameter values used in this simulation are as follows.

1. HgTe Parameters 1.1 Forbidden Bandwidth The HgTe forbidden band width Eg at 300K is −0.15 eV, respectively.

  However, two-dimensional strain ε (biaxial stress) in a direction parallel to the substrate surface is generated in HgTe forming the quantum dots. This distortion changes the forbidden bandwidth of the semiconductor. The change ΔEg in the forbidden bandwidth that occurs at this time is expressed by the following equation.

ΔEg / ε = 2a (1−C ₁₂ / C ₁₁ ) −b (1 + 2C ₁₂ / C ₁₁ )
(1)
Here, C ₁₁ and C ₁₂ are elastic constants. a and b are deformation potentials.

C ₁₁ and C ₁₂ of HgTe are 5.4 × 10 ¹⁰ Pa and 3.8 × 10 ¹⁰ Pa, respectively. Moreover, a and b of HgTe are −3.69 eV and −0.8 eV, respectively.

  Here, the two-dimensional strain ε (biaxial stress) is obtained from the difference between the lattice constant of HgTe and the lattice constant of the substrate.

  When the change in the forbidden band width due to the two-dimensional strain ε obtained from the equation (1) is added to the forbidden band width of HgTe in the absence of strain−0.15 eV, the strain HgTe at 300 K ) Forbidden bandwidth.

The forbidden band width of the strain HgTe at a temperature other than 300K can be obtained based on the forbidden band width of the strain HgTe at 300K and the temperature coefficient of the forbidden band width of HgTe + 5 × 10 ⁻⁴ eV / K.

1.2 Effective mass The effective mass of HgTe electrons is 0.029 m ₀ . On the other hand, the effective mass of holes in HgTe, respectively, is 0.3 m _0.

2. HgCdTe Parameter The parameter relating to Hg _1-y Cd _y Te is obtained from the parameter relating to HgTe and the parameter relating to CdTe, assuming that the parameter changes in proportion to the composition ratio y.

  Each parameter regarding CdTe is as follows.

First, the forbidden bandwidth of CdTe at 300K is +1.50 eV. The temperature coefficient of the forbidden bandwidth of CdTe is −4.1 × 10 ⁻⁴ eV / K.

Next, C ₁₁ and C ₁₂ of CdTe are 5.35 × 10 ¹⁰ Pa and 3.68 × 10 ¹⁰ Pa, respectively. Further, a and b of CdTe are −3.42 eV and −0.8 eV, respectively.

Next, the effective mass of the CdTe electrons and holes, respectively, is 0.11 m ₀ and 0.35 m _0.

1.3 Band Offset The band offset at the interface between the Hg _1-y Cd _y Te quantum dot and the GaAs barrier layer can be determined based on the following parameters.

  The first parameter used to calculate the band offset is the electron affinity of CdTe and the electron affinity of GaAs. Here, the electron affinity of CdTe and GaAs is 4.28 eV and 4.07 eV, respectively.

  The second parameter used for calculating the band offset is the energy difference between the top of the CdTe valence band and the top of the HgTe valence band. This energy difference is +0.46 eV.

  Since the band offset is considered to be proportional to the composition ratio y, the band offset can be obtained using these parameters.

  The values of the parameters described above are described in any of the following documents.

Reference 1: V. Latussek, CR Becker, G. Landwehr, R. Bini and L. Ulivi, Phys. Rev. B 71, 125305 (2005). "
Reference 2: AM de Paula, CRM de Oliveira, GE Marque, AM Cohen, RD Feldman and RF Austin, MN Islam, and CL Cesar, Phys. Rev. B 59, 10158 (1999).
Reference 3; J. -H. Song, J.-S. Kim, K.-U. Jung, and S. -H. Suh, Proc. SPIE. 3436, 34 (1998).
Reference 4: Koichi Tachioka et al., Shizuoka University Graduate School of Electronic Science Research Report 14, 166-169 (1992).

  The parameters relating to GaAs are well known and will not be described in particular. For example, the temperature coefficient of the forbidden band of GaAs is described in Reference 4 below.

Reference 5; Appendix II, in Optical Processes in Semiconductors, by JI Pankove, New York 1975.
(3) Manufacturing Method The manufacturing method of the semiconductor laser 24 is the same as the manufacturing method of a normal FP laser.

  However, the manufacturing method of the active layer 30 is different from the manufacturing method of the active layer of a normal FP laser. Therefore, a method for manufacturing the active layer 30 will be mainly described.

  Growth of each layer including the quantum dot layer 48 is performed by a molecular beam epitaxy (MBE) method or a metal organic chemical vapor deposition (MOCVD) method. Here, a case where the quantum dot layer 48 made of HgCdTe is grown by the MBE method will be described.

  FIG. 13 is a process cross-sectional view illustrating a method for manufacturing the active layer 30.

(I) Lower clad layer forming step and lower barrier layer forming step (see FIG. 13A)
First, n-type AlGaAs to be the lower cladding layer 28 is grown on the n-type GaAs substrate 26.

  Next, GaAs to be the lower barrier layer 4a is grown.

(Ii) Quantum dot formation step (see FIG. 13B)
Next, the substrate temperature is lowered to 200 to 300 ° C. suitable for the growth of Hg _1-y Cd _y Te.

  Next, the Ga molecular beam and As molecular beam used for the growth of the lower barrier layer 4a are blocked, and the Te molecular beam is irradiated to the substrate. At this time, molecular beams (for example, Si molecular beams) used for doping the lower cladding layer 28 and the like are also blocked.

Next, Hg molecular beam and Cd molecular beam are supplied for Hg _1-y Cd _y Te several atomic layers.

Then, the quantum dots 2 made of Hg _1-y Cd _y Te grow in the Stranski-Krastanov mode. At this time, the wetting layer 6 also grows.

In addition, Hg _1-y Cd _y Te has the same senzincite structure as GaAs. Therefore, the growth of the quantum dots 2 is easy.

(Iii) Upper barrier layer (cap layer) formation step (see FIG. 13C)
Next, the Hg molecular beam, Cd molecular beam, and Te molecular beam are blocked.

  Next, Ga molecular beam and As molecular beam are irradiated to the substrate 26 to grow GaAs to be the upper barrier layer 4b.

  Through the above steps, the quantum dot layer 48 is further grown.

(Iv) Repeated Growth of Quantum Dot Layer Thereafter, the above steps (ii) to (iii) for forming the quantum dot layer 48 are repeated as many times as necessary to form the active layer 30. Here, the upper barrier layer 4b (cap layer) of each quantum dot 2 becomes the lower barrier layer 4a of the next quantum dot layer. That is, one barrier layer 4 also serves as the upper barrier layer (cap layer) and the lower barrier layer of the quantum dots formed one above.

(V) Upper Cladding Layer and Electrode Layer Formation Step Thereafter, the substrate temperature is raised to a temperature suitable for GaAs growth (for example, 600 ° C.) to form the upper cladding layer 32 and the electrode layer 36.

(1) Configuration The configuration of this semiconductor laser is substantially the same as that of the semiconductor laser of Example 1. However, the configuration of the quantum dot layer forming the active layer is different from that of the semiconductor laser of Example 1.

  The configuration of this quantum dot layer is substantially the same as that of the quantum dot layer 9 of Embodiment 2 described with reference to FIG. That is, the present quantum dot layer 9 has quantum dots 8 and a barrier layer (barrier layer) 4 in which the quantum dots 8 are embedded.

  And this quantum dot 8 has the 1st part 10 and the 2nd part 12, as shown in FIG. The first portion 10 is made of InAs. On the other hand, the second portion 12 is made of HgTe.

  Here, InAs is a material in which the width of the forbidden band becomes narrow as the temperature rises. On the other hand, HgTe is a material whose width of the forbidden band becomes wider as the temperature rises.

  On the other hand, the barrier layer 4 is made of GaAs. Here, GaAs is a material in which the width of the forbidden band becomes narrow as the temperature rises.

  Therefore, in the present quantum dot 8, the negative temperature coefficient of the first portion 10 and the barrier layer 4 cancels out the positive temperature coefficient of the second portion 10.

  Therefore, the temperature change of the emission wavelength of this semiconductor laser is smaller than that of a conventional FP laser in which an active layer is formed only from a material (InGaAsP or the like) having a negative forbidden band temperature coefficient.

  The material forming the first portion 10 is not limited to InAs. For example, InGaAs may be used.

(2) Characteristics (i) Quantum dot model and potential distribution The temperature change of the emission wavelength of the semiconductor laser can also be obtained by simulation. The quantum dot model used for this simulation is also disk-like, similar to the quantum dot model described with reference to FIGS.

  The shape of the model used for obtaining the temperature change of the emission wavelength is the same as the shape of the model of Example 1 described with reference to FIGS. That is, the external shape of this model is a disk shape.

  FIG. 14 is a plan view for explaining a cross section of the quantum dot model of the present embodiment.

  As shown in FIG. 14, the model is divided into a lower part having a thickness h1 and an upper part having a thickness h2.

  The lower part is a region corresponding to InAs. The thickness h1 of this portion is the thickness of the InAs layer at the center of the actual quantum dot 8.

  On the other hand, the upper part corresponds to HgTe. The thickness h2 of this portion is the thickness of the HgTe layer at the center of the actual quantum dot 8.

  FIG. 15 is a diagram for explaining the potential distribution of the model. The horizontal axis is the position coordinate z in the direction perpendicular to the substrate (direction perpendicular to the bottom surface of the disk 50). The vertical axis represents the electron potential.

  FIG. 15 shows a potential when a quantum dot layer is formed of a quantum dot 8 having an InAs layer 10 having a thickness of 1.1 nm, a HgTe layer 12 having a thickness of 1.3 nm, and a barrier layer 4 made of GaAs. Is described.

  FIG. 15 even shows a potential distribution 52 formed by the bottom of the conduction band and a potential distribution 54 formed by the top of the valence band. Since HgTe is a semimetal, the bottom 52 of the conduction band and the top 54 of the valence band are reversed in the HgTe layer 12.

(Ii) Temperature change of emission wavelength FIG. 15 shows not only the potential distribution but also the ground level 56 of the conduction band and its wave function 58. FIG. 15 also shows the valence band ground level 60 and its wave function 62.

  As shown in FIG. 15, the wave functions 58 and 62 include an HgTe layer (second portion 12) in which the temperature coefficient of the forbidden band is positive and an InAs layer (first first) in which the temperature coefficient of the forbidden band is negative. 10) and the GaAs layer (barrier layer 4).

  As described above, since the wave function spreads in a region where the temperature coefficient is positive and a region where the temperature coefficient is negative, each temperature coefficient is canceled out and the temperature change of the emission wavelength becomes small.

  In the present quantum dot 8, the center of the electron wave function 58 does not coincide with the center of the hole wave function 62. However, since the misalignment between the centers is not large, the decrease in the light emission efficiency of the semiconductor laser is slight.

  FIG. 16 is a diagram for explaining the temperature change of the emission wavelength of the semiconductor laser. The horizontal axis is temperature. The vertical axis represents the emission wavelength.

  The temperature change of the emission wavelength shown in FIG. 16 is a result obtained by simulation for the above model. The simulation method is the same as the method described in the first embodiment.

  The point indicated by “◯” is a simulation result when the thickness of the InAs layer 10 is 3 nm and the thickness of the HgTe layer 12 is 1.3 nm. The diameter of the quantum dot 8 is 43 nm.

  The point indicated by “x” is a simulation result when the thickness of the InAs layer 10 is 1.1 nm and the thickness of the HgTe layer 12 is 1.3 nm. The diameter of the quantum dot 8 is 43 nm.

  That is, the data represented by “◯” and “×” are both simulation results when the thickness of the HgTe layer 12 is 1.3 nm. However, the thickness of the InAs layer 10 is different.

  As shown in FIG. 16, the emission wavelength when the thickness of the InAs layer 10 indicated by “◯” is 3 nm is approximately 1.55 μm regardless of the temperature. On the other hand, the emission wavelength when the thickness of the InAs layer 10 indicated by “x” is 1.1 nm is approximately 1.3 μm regardless of the temperature. These emission wavelengths are suitable for optical fiber communication.

  Here, the rate of change of the emission wavelength with respect to temperature is +0.01 nm / K when the thickness of the InAs layer 10 is 3 nm. On the other hand, when the thickness of the InAs layer 10 is 1.1 nm, it is +0.05 nm / K.

  These values are much smaller than the temperature coefficient of the emission wavelength of the conventional FP laser in which the active layer is formed of InGaAsP + 0.5 nm / K.

  FIG. 17 is a diagram for explaining the relationship between the thickness of the HgTe layer 12 and the emission wavelength, and the relationship between the thickness of the HgTe layer 12 and the temperature coefficient of the emission wavelength. The horizontal axis represents the thickness of the HgTe layer 12. The left vertical axis represents the emission wavelength of the semiconductor laser. The right vertical axis represents the temperature coefficient of the emission wavelength of the semiconductor laser.

  Note that all the data shown in FIG. 17 is data when the thickness of the InAs layer 10 is 2 nm.

  The dots indicated by “●” indicate the emission wavelength. The point indicated by “◯” indicates the temperature coefficient of the emission wavelength.

  As is clear from the data indicated by “◯”, the temperature coefficient of the emission wavelength decreases as the HgTe layer 12 becomes thicker and eventually becomes a negative value.

  As the HgTe layer 12 becomes thicker, the proportion of the wave function distributed in the HgTe layer 12 increases. As a result, the influence of HgTe having a positive temperature coefficient of the forbidden band width increases, and the temperature coefficient of the emission wavelength decreases.

  Here, as is clear from the data indicated by “◯”, when the thickness of the HgTe layer 12 is 1.1 nm or more and 1.5 nm or less, the temperature coefficient of the emission wavelength is −0.1 nm / K to 0.00. It becomes a small value (absolute value) of 1 nm / K.

  As is clear from the data indicated by “●”, when the thickness of the HgTe layer 12 is 1.1 nm or more and 1.5 nm or less, the emission wavelength is suitable for optical fiber communication of 1.33 μm to 1.55 μm. Value.

  That is, the thickness of the HgTe layer 12 (thickness on the lower side of the top of the quantum dots 8) is preferably 1.1 nm or more and 1.5 nm or less. Further, as is clear from FIG. 17, the thickness of the HgTe layer 12 is more preferably 1.2 nm or more and 1.4 nm or less.

  FIG. 18 is a diagram illustrating the relationship between the thickness of the InAs layer 10 and the emission wavelength, and the relationship between the thickness of the InAs layer 10 and the temperature coefficient of the emission wavelength. The horizontal axis represents the thickness of the InAs layer 10. The left vertical axis represents the emission wavelength of the semiconductor laser. The right vertical axis represents the temperature coefficient of the emission wavelength of the semiconductor laser.

  Note that all the data shown in FIG. 18 is data when the thickness of the HgTe layer 12 is 1.3 nm.

  The point indicated by “▲” indicates the emission wavelength. The point indicated by “Δ” indicates the temperature coefficient of the emission wavelength.

  As is clear from the data indicated by “Δ”, the temperature coefficient of the emission wavelength decreases once as the InAs layer 10 becomes thicker, and then increases.

  As the InAs layer 10 becomes thicker, the ratio of the wave function leaking to the GaAs barrier layer 4 decreases and the ratio of the wave function distributed in the HgTe layer 12 increases. For this reason, the influence of HgTe having a positive temperature coefficient of the forbidden bandwidth is increased, and the temperature coefficient of the emission wavelength is decreased.

The absolute value of the temperature coefficient of the forbidden band width of InAs (−3.3 × 10 ⁻⁴ eV / K) is the same as the temperature coefficient of the forbidden band width of GaAs (−5.0 × 10 ⁻⁴ eV / K). Less than absolute value.

  However, when the InAs layer 10 is further thickened, the proportion of the wave function distributed in the InAs layer 10 having a negative forbidden band temperature coefficient increases, so that the temperature coefficient of the emission wavelength starts to increase.

  Here, as is apparent from the data indicated by “Δ”, when the thickness of the InAs layer 10 is 1 nm or more and 4 nm or less, the temperature coefficient of the emission wavelength is −0.1 nm / K to 0.1 nm / K. (Absolute value) becomes a small value.

  As is clear from the data indicated by “▲”, when the thickness of the InAs layer 10 is 1 nm or more and 4 nm or less, the emission wavelength is also a value suitable for optical fiber communication from 1.3 μm to 1.6 μm. .

  That is, the thickness of the InAs layer 10 (thickness on the lower side of the top of the quantum dots 8) is preferably 1 nm or more and 4 nm or less. As is clear from FIG. 18, the thickness of the InAs layer 10 is more preferably 2 nm or more and 3 nm or less.

  The forbidden bandwidth of InAs used in this simulation is obtained based on the following equations (2) and (3).

The following formula (2) gives a forbidden band width Eg (x, T) of In _x Ga _1-x As without distortion as shown below.

Eg (x, T) = (1.519-5.408 × 10 ⁻⁴ T ² /(T+204))(1-x)+(0.417-2.76×10 ⁻⁴ T ² / (93 + T) ) X-0.475x (1-x) (2) where x is the composition ratio of In. T is an absolute temperature.

Equation (3), on the other hand, gives the forbidden bandwidth at 300 K of strained In _x Ga _1-x As formed on GaAs as shown below.

Eg (x) = 1.43-1.11x + 0.45x ² (3)
Here, x is a composition ratio of In.

  From these expressions (2) and (3), the ratio of the change in the forbidden band width ΔEg generated in InAs due to strain to the strain ε (= ΔEg / ε) is obtained at 300K.

  Next, ε at each temperature at which the simulation is performed is obtained from the difference between the thermal expansion coefficients of InAs and GaAs.

  Next, ΔEg at each temperature is obtained by multiplying ΔEg / ε at 300 K by the strain ε at each temperature.

  Finally, the forbidden band width of the strain InAs at each temperature is obtained based on this ΔEg and the equation (2).

  The band offset between InAs and GaAs is obtained as follows.

  First, the difference in the forbidden bandwidth between InAs and GaAs is obtained.

  Then, 65% is allocated to the conduction band side and 35% is allocated to the valence band side, and the band offset is obtained (ΔEc: ΔEv = 65: 35).

  Note that the band offset between InGaAs and GaAs can also be obtained by assigning the difference in the forbidden bandwidth in the same ratio.

(3) Manufacturing Method FIGS. 19 and 20 are process cross-sectional views illustrating the method of manufacturing the semiconductor laser.

  The manufacturing method of this semiconductor laser is substantially the same as the manufacturing method of the first embodiment.

  However, the quantum dots 8 grown by this manufacturing method are formed by laminating an InAs layer and an HgTe layer.

(I) Lower clad layer forming step and lower barrier layer forming step (see FIG. 19A)
First, n-type AlGaAs to be the lower cladding layer 28 is grown on the n-type GaAs substrate 26.

  Next, GaAs to be the lower barrier layer 4a is grown.

(Ii) Quantum dot formation step (see FIGS. 19B and 20A)
Next, the Ga molecular beam used for the growth of the lower cladding layer 28 and the like is cut off, the substrate temperature is lowered to 400 to 520 ^° C., and the In molecular beam is irradiated onto the substrate 26 for several atomic layers.

  At this time, the irradiation of the As molecular beam used for the growth of the lower cladding layer 28 is continued. On the other hand, molecular beams (for example, Si molecular beams) used for doping the lower cladding layer 28 and the like are blocked.

  Then, a dot-like InAs layer (first portion 10) grows in the Stranski-Krastanov mode. At this time, the first wetting layer 14 also grows (see FIG. 19B).

Next, the In molecular beam and the As molecular beam are blocked, the substrate temperature is lowered to 200 to 300 ^° C., and the Te molecular beam is irradiated to the substrate 26.

  Next, Hg molecular beams are supplied for several atomic layers.

  Then, a dot-like HgTe layer (second portion 12) grows in the Stranski-Krastanov mode. At the same time, the second wetting layer 16 also grows (see FIG. 20A).

  By the way, the relaxation of strain proceeds more in the InAs layer (first portion 10) than in the InAs wetting layer (first wetting layer 14). For this reason, HgTe grows thicker on the dot-like InAs layer (first portion 10) to become the HgTe layer (second portion 12).

(Iii) Upper barrier layer forming step (see FIG. 20B)
Next, the Hg molecular beam and the Te molecular beam are blocked.

  Next, Ga molecular beam and As molecular beam are irradiated to the substrate 26 to grow GaAs to be the upper barrier layer 4b.

  Through the above steps, the quantum dot layer 64 is further grown.

  Thereafter, the steps (ii) to (iii) for forming the quantum dot layer 64 are repeated to form an active layer.

(Iv) Upper Cladding Layer and Electrode Layer Formation Step Thereafter, the substrate temperature is raised to a temperature suitable for GaAs growth (for example, 600 ° C.) to form the upper cladding layer and the electrode layer.

  By the way, when heat treatment is performed after the growth of the InAs layer 10, the InAs layer 10 is deformed into a ring shape. When HgTe is grown on the substrate on which such ring-shaped InAs are formed, quantum dots filled with HgTe in the ring are formed.

  You may manufacture this semiconductor laser using such a quantum dot.

(1) Configuration The configuration of this semiconductor laser is substantially the same as that of the semiconductor laser of Example 1. However, the configuration of the quantum dot layer forming the active layer is different from that of the semiconductor laser of Example 1.

The quantum dot layer of this example is substantially the same as the quantum dot layer 21 of the third embodiment (see FIG. 3). That is, the quantum dot 18 is formed of Hg _1-y Cd _y Te (0.4 <y <0.6). On the other hand, the barrier layer 20 is made of Al _0.1 Ga _0.9 As. In place of Al _0.1 Ga _0.9 As, the barrier layer is made of another semiconductor (including AlGaAs) whose forbidden band is wider than GaAs and whose top of the valence band is lower than the top of the valence band of GaAs. 20 may be formed.

  The upper clad layer 28 and the lower clad 32 are made of AlGaAs having an Al composition greater than 0.1.

The temperature coefficient of the forbidden band of Hg _1-y Cd _y Te becomes substantially zero when the composition y of Cd is between 0.4 and 0.6. For this reason, the temperature change of the emission wavelength of this semiconductor laser is smaller than that of a conventional FP laser in which the entire active layer is formed of a semiconductor having a negative forbidden band temperature coefficient.

By the way, the band offset between Hg _1-y Cd _y Te (0.4 <y <0.6) and GaAs is slight. For this reason, in this embodiment, the barrier layer 20 is formed of Al _0.1 Ga _0.9 As so that carriers (particularly, holes) are not sufficiently confined in the quantum dots. The Al composition is preferably 0.15 or less.

(2) Temperature change of emission wavelength FIG. 21 is a diagram for explaining a temperature change of the emission wavelength of the semiconductor laser. The horizontal axis is temperature. The vertical axis represents the emission wavelength. The temperature change of the emission wavelength shown in FIG. 21 is a result obtained by simulation. The simulation method is the same as the method described in the first embodiment.

  Points indicated by “◯” are the simulation results when the height and diameter of the quantum dots 18 are 4.5 nm and 45 nm, respectively.

  As shown in FIG. 21, the emission wavelength of this semiconductor laser is about 1.55 μm regardless of the temperature. This emission wavelength is suitable for optical fiber communication.

  FIG. 22 is a diagram for explaining the relationship between the height of the quantum dot 18 and the emission wavelength, and the relationship between the height of the quantum dot 18 and the temperature coefficient of the emission wavelength. The horizontal axis represents the height of the quantum dot 18. The left vertical axis represents the emission wavelength of the semiconductor laser. The right vertical axis represents the temperature coefficient of the emission wavelength of the semiconductor laser.

  The point indicated by “●” indicates the emission wavelength when the composition y of Cd is 0.5. The point indicated by “◯” indicates the temperature coefficient of the emission wavelength when the composition y of Cd is also 0.5.

  The point indicated by “▲” indicates the emission wavelength when the composition y of Cd is 0.6. The point indicated by “Δ” indicates the temperature coefficient of the emission wavelength when the composition y of Cd is 0.6.

  As is clear from the data indicated by “◯” and “Δ”, when the height of the quantum dot 18 is 2 nm or more and 8 nm or less, the temperature coefficient of the emission wavelength is −0.05 nm / K to 0.1 nm / K ( Small value (absolute value). However, the composition y of Cd is any value between 0.5 and 0.6.

  As is clear from the data indicated by “●” and “▲”, when the height of the quantum dots 18 is 2.0 nm or more and 8 nm or less, the emission wavelength is 1.2 μm to 1.6 μm in optical fiber communication. It becomes a suitable value.

  Considering these results and the result of y = 0.4 described with reference to FIG. 9, the height of the quantum dots 18 formed of HgCdTe (0.4 <y <0.6) It can be seen that 2.0 nm or more and 8.0 nm or less is preferable. The height of the quantum dot 18 is preferably 2.0 nm or more and 6.0 nm or less.

(Modification)
Examples 1 to 3 described above relate to the FP laser. However, the light emitting device to which the first to third embodiments can be applied is not limited to the FP laser. For example, the first to third embodiments can be applied to, for example, a DFB laser, an optical amplifier, and a light emitting diode.

  In Examples 1 to 3, the quantum dots are formed of HgTe or HgCdTe. However, the material that can be used for forming the quantum dots of the light emitting device is not limited to HgTe or HgCdTe. For example, the quantum dots may be formed of HgSe or HgCdSe. The temperature coefficient of the forbidden band width of these materials is also positive, like the temperature coefficient of HgTe or HgCdTe. In addition, Pb salt (for example, PbS, PbSe, or PbTe) is also available as a material that has a positive temperature coefficient of the forbidden band width.

  In the first to third embodiments, the barrier layer is made of GaAs or AlGaAs. However, the material that can be used for forming the barrier layer is not limited to GaAs or AlGaAs. For example, the quantum dots may be formed of InP, InGaAsP, InAlGaAs, or the like.

  In Examples 1 to 3, a GaAs semiconductor substrate is used. However, other semiconductor substrates (for example, InP substrates) may be used.

  In Examples 1 to 3, the upper cladding layer and the lower cladding layer are made of AlGaAs. However, the upper cladding layer and the lower cladding layer may be formed of other semiconductor materials (for example, InP or AlGaAsP).

2 ... Quantum dot 4 of the first embodiment 4 ... Barrier layer 5 of the first embodiment ... Quantum dot layer 6 of the first embodiment ... of the first embodiment ) Wetting layer 8... (Dot 2) quantum dot 9... (Dot 2) quantum dot layer 10... First portion 12... Second portion 14. First wetting layer 16 ... second wetting layer 18 ... quantum dots (in the third embodiment)
20 ... (Embodiment 3) Barrier layer 21 ... (Embodiment 3) Quantum dot layer 22 ... (Embodiment 2) Wetting layer 24 ... (Example 1) Semiconductor laser 26 ... n-type GaAs substrate 28 ... lower clad layer 30 ... active layer 32 ... upper clad layer 34 ... band-like mesa 36 ... electrode layer 38 ... protective film 40 ... Upper electrode 42 ... Lower electrode 44, 46 ... End face 48 ... Quantum dot layer 50 ... Disc 52 ... Potential distribution 54 formed by the bottom of the conduction band ... Valence electrons Potential distribution formed by the top of the band 56... Ground level of conduction band 58... Wave function 60... Ground level of valence band 62. Of) quantum dot layer

Claims (5)

A quantum dot formed of a first material that increases the width of the forbidden band as the temperature rises;
An active layer formed of a second material whose width of the forbidden band is narrowed when the temperature rises, and having a barrier layer in which the quantum dots are embedded;
A light emitting device provided. The light-emitting device according to claim 1.
The first material is Hg _1-y Cd _y Te (0 ≦ y ≦ 0.4),
A light emitting device characterized. The light-emitting device according to claim 1.
A part of the quantum dots is formed of a third material that narrows the width of the forbidden band as the temperature rises.
A light emitting device characterized. The light-emitting device according to claim 3.
The first material is Hg _1-y Cd _y Te (0 ≦ y ≦ 0.4),
A light emitting device characterized. An active layer having quantum dots formed of Hg _1-y Cd _y Te (0.4 <y <0.6),
A light emitting device provided.

Images