JP2010092898A - Semiconductor light emitting element - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a semiconductor light emitting element having a structure capable of suppressing relaxation oscillation. <P>SOLUTION: The semiconductor light emitting element has a laminate structure including n-type semiconductor layers 102 and 103, p-type semiconductor layers 110 and 111, and an active layer 105 interposed between those n-type semiconductor layers 102 and 103 and p-type semiconductor layers 110 and 111. The active layer 105 includes a function layer such that a carrier diffusion length in a direction perpendicular to the laminate direction is larger than a carrier diffusion length in the laminate direction. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a semiconductor light emitting device such as a semiconductor laser or a semiconductor optical amplifier.

  2. Description of the Related Art Generally, a semiconductor laser diode (LD) for an optical disk is driven by using a pulse current in order to reduce relative intensity noise (RIN) induced by return light from the optical disk during information reproduction. The That is, RIN is generated by interference between the light emitted from the laser diode and the return light reflected by the optical disk. Therefore, in order to weaken the coherence of the laser light, the DC bias current for driving the laser diode is increased from about 50 MHz to several. A driving method in which a high frequency current in a range of about GHz is superimposed is adopted. Such a driving method is called a high frequency superposition method.

  The high-frequency superimposition method is used for existing optical disks such as AlGaAs (Aluminum Gallium Arsenide) semiconductor lasers that are light sources for CDs (Compact Discs) and AlGaInP (Aluminum Gallium Indium Phosphide) semiconductor lasers that are light sources for DVDs (Digital Versatile Discs). It can be used not only as a semiconductor laser but also as an effective means in a GaN (Gallium Nitride) -based semiconductor laser that is a light source for next-generation optical disks. Prior art relating to the high-frequency superposition method is disclosed in, for example, Japanese Patent Application Laid-Open No. 2008-53600.

As disclosed in Patent Document 1, when a pulsed drive current is supplied to a laser diode, the density of injected carriers (electrons and holes) in the laser diode varies periodically. According to this variation, A phenomenon (relaxation oscillation) occurs in which the intensity of the laser light attenuates while periodically oscillating and converges to a certain intensity. Due to this relaxation vibration, the peak level of the light emission output becomes excessive at the rising edge of the pulse, and there is a risk that the recorded contents of the optical disc will be erased by mistake. Non-Patent Document 1 (K. Furuya et al., Appl. Opt. 17, 1949-1952 (1978)) can suppress relaxation oscillation if the stripe width and carrier diffusion length of the semiconductor laser diode are made the same level. It is reported.
JP 2008-53600 A K. Furuya, Y. Suematsu, and T. Hong, "Reduction of resonancelike peak in direct modulation due to carrier diffusion in injection laser," Appl. Opt. 17, 1949-1952 (1978). A. Kaneta, T. Mutoh, Y. Kawakami, S. Fujita, G. Marutsuki, Y. Narukawa, and T. Mukai, "Discrimination of local radiative and nonradiative recombination processes in an InGaN / GaN single-quantum-well structure by a time-resolved multimode scanning near-field optical microscopy, "Appl. Phys. Lett., 83, 3462-3464 (2003).

  When a high frequency current is superimposed on a DC bias current to reduce RIN, there is a problem that RIN varies irregularly when operating conditions change. For example, with respect to changes in operating conditions such as the optical output of the semiconductor laser diode, the operating environment temperature, the frequency and amplitude of the high frequency current, the intensity of RIN increases and decreases unstablely without monotonous changes. There is a problem that it is difficult to read the optical signal stably. This tendency is particularly noticeable in GaN semiconductor lasers.

  As a result of diligent research on the relationship between the relaxation oscillation and RIN, focusing on this problem, the present inventors have found that the larger the relaxation oscillation amplitude, the greater the variation in RIN with respect to changes in the operating conditions of the semiconductor laser. I found out that The present inventors have found that suppression of relaxation oscillation is effective in reducing fluctuations in RIN.

  In view of the above, the present invention provides a semiconductor light emitting device having a structure capable of suppressing relaxation oscillation.

  According to the present invention, there is provided a semiconductor light emitting device having a stacked structure including an n-type semiconductor layer, a p-type semiconductor layer, and an active layer interposed between the n-type semiconductor layer and the p-type semiconductor layer. Is done. In this semiconductor light emitting device, the active layer includes a functional layer in which the carrier diffusion length in the direction perpendicular to the stacking direction is larger than the carrier diffusion length in the stacking direction.

  As described above, suppression of relaxation oscillation is effective for reducing fluctuations in RIN. Since the active layer of the semiconductor light emitting device according to the present invention includes a functional layer in which the carrier diffusion length in the direction perpendicular to the stacking direction is larger than the carrier diffusion length in the stacking direction, when a high frequency current is superimposed, the carrier density in the light emitting region Fluctuations can be suppressed. As a result, relaxation oscillation is suppressed, so that irregular fluctuations in RIN can be reduced.

  Embodiments according to the present invention will be described below with reference to the drawings. In all the drawings, the same components are denoted by the same reference numerals, and detailed description thereof is appropriately omitted so as not to overlap.

  FIG. 1 is a diagram schematically showing a cross-sectional structure of a semiconductor laser diode (semiconductor light emitting device) according to an embodiment of the present invention. This semiconductor laser diode includes an n-type buffer layer 102, an n-type cladding layer 103, an n-side optical confinement layer 104, an active layer 105, a cap layer 106, a p-side optical confinement layer 107 and a current on an n-type compound semiconductor substrate 101. The constriction layer 108 has a structure in which these layers are stacked in this order. The current confinement layer 108 has a stripe-shaped opening 109 that constitutes a current flow path, and a p-type superlattice cladding layer 110 is formed in the opening 109 and on the current confinement layer 108. A p-type contact layer 111 is formed on the p-type superlattice cladding layer 110. A p-side electrode 112 is provided on the p-type contact layer 111, and an n-side electrode 113 is provided on the back surface of the n-type compound semiconductor substrate 101.

  This semiconductor laser diode can be composed of a group III nitride compound semiconductor. For example, the n-type compound semiconductor substrate 101 may be an n-type GaN substrate, or a template formed by crystal growth of a group III nitride semiconductor layer such as a GaN layer on a base substrate made of another material such as sapphire or SiC. It may be a substrate. A GaN layer having a thickness of about 1 μm as the n-type buffer layer 102, an AlGaN layer having a thickness of about 2 μm as the n-type cladding layer 103, and a GaN layer having a thickness of about 0.1 μm as the n-side optical confinement layer 104, An AlGaN layer having a thickness of about 10 nm is used as 106, a GaN layer having a thickness of about 0.1 μm is used as the p-side optical confinement layer 107, and an AlN layer having a thickness of about 0.1 μm is used as the current confinement layer 108. it can. Since the current confinement layer 108 generates a refractive index difference in the horizontal direction (direction perpendicular to the stacking direction), it also has a light distribution control function in addition to the current confinement function.

  The p-type superlattice cladding layer 110 may have a superlattice structure in which, for example, a GaN layer having a thickness of about 2.5 nm and an AlGaN layer having a thickness of about 2.5 nm are alternately stacked for 130 periods. . As the p-type contact layer 111, for example, a GaN layer having a thickness of about 0.1 μm may be used. In such a stacked structure, the n-type impurity introduced into the n-type semiconductor layer is, for example, silicon (Si), and the p-type impurity introduced into the p-type semiconductor layer is, for example, magnesium (Mg).

The active layer 105 includes a functional layer in which the carrier diffusion length in the direction perpendicular to the stacking direction is larger than the carrier diffusion length in the stacking direction. The active layer 105 has a quantum well structure in which quantum well layers and barrier layers are alternately stacked, and the barrier layer has a wider band gap than the quantum well layer. In this case, a functional layer can be formed in one or both of the quantum well layer and the barrier layer. The functional layer has a superlattice structure in which first semiconductor layers (superlattice well layers) and second semiconductor layers (superlattice barrier layers) having a wider band gap than the first semiconductor are alternately stacked. Can be realized. The first semiconductor layer and the second semiconductor layer are made of materials having different compositions (for example, Al _x In _y Ga _1-xy N; 0 ≦ x ≦ 1, 0 ≦ y ≦ 1, 0 ≦ x + y ≦ 1). When the functional layer is formed only in the barrier layer, the quantum well layer is configured as a single layer made of, for example, In _z Ga _1-z N (0 ≦ z ≦ 1), and the barrier layer is formed of the first semiconductor layer. A superlattice structure composed of the second semiconductor layer can be used. On the other hand, when the functional layer is formed only in the quantum well layer, the barrier layer is formed of, for example, a single GaN layer, and the quantum well layer is a superlattice composed of the first semiconductor layer and the second semiconductor layer. Can be configured with a structure.

  As described above, in the semiconductor laser diode of FIG. 1, the active layer 105 is interposed between the n-type semiconductor layers 102 and 103 and the p-type semiconductor layers 110 and 111. The active layer 105 has a carrier diffusion length in the direction perpendicular to the stacking direction (hereinafter referred to as “first carrier diffusion length”) called the carrier diffusion length in the stacking direction (hereinafter referred to as “second carrier diffusion length”). .) Including a larger functional layer. In general, the carrier diffusion length L is proportional to the square root of the diffusion coefficient D, and the carrier mobility μ is proportional to the diffusion coefficient D. Therefore, in this functional layer, the carrier mobility in the direction perpendicular to the stacking direction (hereinafter referred to as “first carrier mobility”) is the carrier mobility in the stacking direction (hereinafter referred to as “second carrier mobility”). Larger than).

  Note that the first carrier diffusion length is L1, the second carrier diffusion length is L2, the first carrier mobility is μ1, and the second carrier mobility is μ2. At this time, the following expressions (1a) to (1c) and (2a) to (2c) are established.

L1 = (τ · D1) ^(1/2) (1a)
D1 = k · T / (e · μ1) (1b)
μ1 = 1 / (n · e · ρ1) (1c)
L2 = (τ · D2) ^(1/2) (2a)
D2 = k · T / (e · μ2) (2b)
μ2 = 1 / (n · e · ρ2) (2c)

  Here, D1, D2: diffusion coefficient, k: Boltzmann constant, T: temperature, e: elementary electric charge, n: carrier concentration, ρ1: electrical resistivity in the direction perpendicular to the stacking direction, ρ2: electrical resistance in the stacking direction Rate. Therefore, if the above equations (1a) to (1c), (2a) to (2c) and the measured values of electrical resistivity ρ1 and ρ2 are used, carrier mobility μ1 and μ2 and carrier diffusion lengths L1 and L2 are obtained. Can be calculated.

  As described above, the functional layer can be formed in at least one of the quantum well layer and the barrier layer constituting the active layer 105. FIG. 2 is a diagram schematically showing a conduction band of the active layer 105 when a functional layer is formed in the barrier layer. In this case, the quantum well layer 201 is made of, for example, InGaN having a thickness of 6 nm. The barrier layer 202 is composed of, for example, a functional layer having a thickness of 14 nm, and this functional layer has a superlattice structure in which, for example, a GaN layer having a thickness of 2 nm and an AlGaN layer having a thickness of 2 nm are alternately stacked. On the other hand, FIG. 3 is a diagram schematically showing a conduction band of a normal active layer not including a functional layer. In the active layer of FIG. 3, the quantum well layers 301 and the barrier layers 302 are merely stacked alternately. As described above, the barrier layer 202 is a superlattice in which a superlattice well layer having a relatively narrow band gap and a superlattice barrier layer having a wider bandgap than the superlattice well layer are alternately and periodically stacked. It has a structure. In the barrier layer 202, an energy band called a miniband having a superlattice structure may be formed. If the thickness of the superlattice barrier layer and the thickness of the superlattice well layer are sufficiently thin, a spatial overlap occurs between the wave functions of different superlattice well layers, and the degenerate quantum levels are split and small energy is generated. A band or miniband is formed.

  FIG. 4 is a graph showing the measurement result of the optical output of the semiconductor laser diode (Example) having the band structure of FIG. 2 and the measurement result of the optical output of the semiconductor laser diode (Reference Example) having the band structure of FIG. It is. The vertical axis of the graph indicates the optical output in arbitrary units (a.u.). The light output waveforms of the example and the reference example are adjusted so that the values obtained by integrating the respective light outputs over time are substantially the same. Further, the reference example has the same structure as the example except for the functional layer. As shown in FIG. 4, it is shown that the structure of the example suppresses relaxation oscillation compared to the reference example. Therefore, the structure of the example reduces the variation of RIN.

  The superlattice structure of the present embodiment has a structure in which the thickness of each layer constituting the superlattice structure is thin enough to form a miniband. Therefore, a layer structure such as a general quantum well structure or a hetero structure is different from the superlattice structure of the present embodiment. The thickness of each layer constituting the superlattice structure depends on the constituent material and composition, but is preferably about 5 nm or less, particularly about 3 nm or less.

  The “superlattice structure” in this specification can be applied to the Kronig-Penny model in which the thickness of each layer constituting the superlattice structure approximates the potential in the crystal with a well-type periodic potential. Thick to the extent. A structure in which layers having a thickness of an atomic layer level are stacked is regarded as a pseudo mixed crystal in which the composition of these layers is averaged, and thus is different from the superlattice structure in this specification. The thickness of each layer constituting the superlattice structure depends on the material and composition, but is preferably about 1 nm or more, for example.

  Note that the periodicity of the superlattice does not have to be strictly established in the functional layer, and the periodicity of the superlattice may be destroyed as long as a certain amount of carrier diffusion length L2 and a certain amount of carrier mobility μ2 can be secured. . For example, even when the thickness of one or both of the superlattice well layer and the superlattice barrier layer gradually changes in the direction of the cap layer 106 or the direction of the n-side optical confinement layer 104, a certain amount of carrier diffusion length is obtained. If L2 and a certain amount of carrier mobility μ2 can be secured, the same effect as in the superlattice structure of the present embodiment can be obtained.

  Next, FIGS. 5A and 5B are diagrams schematically showing the conduction band of the active layer 105 when a functional layer is formed in the quantum well layer, respectively. 5A is a structure in which the quantum well layers 401 and the barrier layers 402 are alternately stacked, and the quantum well layer 401 is a functional layer. The structure in FIG. And barrier layers 502 are alternately stacked, and the quantum well layer 501 is a functional layer. Examples of the functional layer in FIGS. 5A and 5B include a superlattice structure containing InGaN / GaN, InGaN / InGaN, or InGaN / AlGaN.

  Next, an example of a method for manufacturing the semiconductor laser diode of the present embodiment will be described.

  For manufacturing the element structure, a reduced pressure MOVPE apparatus that enables a growth pressure of 300 hPa is used. A mixed gas of hydrogen and nitrogen is used as the carrier gas, and trimethylgallium, trimethylaluminum, and trimethylindium are used as the Ga source, Al source, and In source, respectively. Silane is used as the n-type impurity, and biscyclopentadienyl magnesium is used as the p-type impurity.

  After the n-type GaN substrate 101 is put into the reduced pressure MOVPE apparatus, the substrate 101 is heated while supplying ammonia into the reduced pressure MOVPE apparatus, and the growth is started when the temperature reaches the growth temperature. In the first growth, the n-type GaN buffer layer 102, the n-type AlGaN cladding layer 103, the n-side GaN light confinement layer 104, the active layer 105, the AlGaN cap layer 106, the p-side GaN light confinement layer 107, and the AlN current confinement layer 108. Grow sequentially. The active layer 105 has a multiple quantum well structure in which InGaN quantum well layers 201 and AlGaN / GaN superlattice barrier layers 202 are alternately stacked. The growth temperature may be, for example, 200 to 800 ° C. when the AlN current confinement layer 108 is grown, 800 ° C. when the active layer 105 is grown, and 1100 ° C. when the other layers are grown. Since the AlN current confinement layer 108 grows at a relatively low temperature, it is in an amorphous state at the end of the first growth.

Thereafter, a SiO ₂ film is deposited on the entire device, and an opening is formed in the SiO ₂ film using a photolithography technique to form a SiO ₂ mask. Next, a desired opening 109 is formed in the AlN current confinement layer 108 using a mixed solution of phosphoric acid and sulfuric acid maintained at a temperature of 50 to 200 ° C. as an etchant. At this time, amorphous AlN is easily etched, and single crystal GaN is difficult to etch. Therefore, etching with high selectivity and good controllability is performed.

  Next, after the element is put into the reduced pressure MOVPE apparatus again, the substrate is heated while ammonia is supplied into the apparatus, and the second growth is started when the temperature reaches the growth temperature. At this time, the AlN current confinement layer 108 undergoes single crystallization in the process of raising the temperature of the substrate. Next, a p-type AlGaN / GaN superlattice cladding layer 110 and a p-type GaN contact layer 111 are grown sequentially. Thereafter, the p-side electrode 112 is formed on the upper surface of the device, and the n-side electrode 113 is formed on the back surface of the device, whereby the semiconductor laser diode of FIG. 1 is manufactured.

  The effects of the semiconductor laser diode of this embodiment are as follows.

  As described above, the active layer 105 includes a functional layer in which the carrier diffusion length in the direction perpendicular to the stacking direction is larger than the carrier diffusion length in the stacking direction. The carrier diffusion effect in the direction perpendicular to the stacking direction is enhanced. Thereby, when the semiconductor laser diode is driven using the drive current on which the high-frequency current is superimposed, the fluctuation of the carrier density in the light emitting region can be suppressed. Therefore, since relaxation oscillation is suppressed, irregular fluctuations in RIN can be reduced. The reason will be described in detail below.

  FIG. 6 is a diagram cited from Non-Patent Document 1. The upper diagram of FIG. 6 shows a typical structure of a semiconductor laser diode having a stripe-shaped active layer, and the lower diagram of FIG. 6 shows the relationship between the light intensity distribution in the horizontal direction and the carrier density distribution. It is a graph. As shown in the graph of FIG. 6, the carrier density distribution is not uniform in the horizontal direction (X-axis direction), and the vicinity of the center of the stripe is recessed. This phenomenon is a phenomenon called spatial hole burning in which many carriers are consumed near the center where the light intensity is high. Therefore, the carrier density varies greatly especially near the center of the stripe. According to Non-Patent Document 1, when the stripe width W and the carrier diffusion length L are approximately the same, carriers are quickly supplied from the peripheral region of the stripe to the central region, so that the dent in the carrier density distribution is reduced. On the other hand, when W> L as in a conventional GaN-based semiconductor laser, carriers are not sufficiently supplied from the peripheral region of the stripe to the central region, so that the dent in the carrier density distribution is not reduced.

  The structure of the active layer 105 of the present embodiment enhances the horizontal carrier diffusion effect. Therefore, by increasing the carrier supply amount from the peripheral region of the stripe to the central portion (the opening 109 and its vicinity), the carrier density distribution The dent can be reduced. Therefore, it is possible to suppress fluctuations in carrier density.

  As described above, in order to suppress the relaxation oscillation, it is effective to make the stripe width W and the carrier diffusion length L approximately the same. However, the stripe width W is a single transverse mode. And various characteristics such as a radiation angle are set to be desired characteristics. On the other hand, since the carrier diffusion length L is a value unique to the semiconductor material, the stripe width W and the carrier diffusion length L are not necessarily the same level. In particular, in the GaN-based semiconductor laser, the stripe width W is set in the range of 1 μm to 2 μm, while the carrier diffusion length L is reported to be a value of 1 μm or less, for example, a value of about several hundred nm (non- Patent Document 2: A. Kaneta et al., Appl. Phys. Lett., 83, 3462-3464 (2003)). In other words, GaN semiconductor lasers are unique to GaN-based semiconductor lasers, where L is significantly smaller than W due to the relationship between various characteristics required as a light source for next-generation optical disks and physical properties inherent in nitride semiconductors. There is a situation.

  In the above embodiment, in order to improve the above effect, it is preferable to provide a functional layer in the vicinity of a layer having a high light intensity and a large carrier density fluctuation. Therefore, for example, when a functional layer is provided in the barrier layer in the active layer 105, it is preferable to provide the functional layer in at least the barrier layer sandwiched between the quantum well layers.

  In addition, when the impurity is added to the functional layer in the active layer 105 and the functional layer is configured by alternately stacking a layer having a small band gap and a layer having a large band gap, the band gap is large. It is preferable to add impurities only to the layer. The reason is that horizontal diffusion of carriers mainly occurs in a layer having a small band gap. Therefore, if impurities are also added to a layer having a small band gap, carrier scattering due to the impurities occurs, which weakens the horizontal carrier diffusion effect. Because it will end up.

  As mentioned above, although embodiment of this invention was described with reference to drawings, these are the illustrations of this invention, Various structures other than the above are also employable. For example, the above-described manufacturing method is an exemplification, and it can be understood by those skilled in the art that various modifications are possible.

  The semiconductor laser diode of FIG. 1 has a Fabry-Perot optical resonator structure, but is not limited to this. The structure of the active layer 105 may be applied to a semiconductor photodiode or a semiconductor optical amplifier that does not have an optical resonator structure.

  The semiconductor laser diode of FIG. 1 has an inner stripe waveguide type structure in which the current confinement layer 108 is embedded in a laminated structure, but the present invention is not limited to this. The structure of the active layer 105 may be applied to a ridge waveguide type semiconductor light emitting device having a p-type cladding layer and a p-type contact layer processed into a ridge stripe shape.

  Further, not all of the barrier layers of the active layer 105 may be functional layers, and some of the barrier layers may be functional layers. Similarly, a functional layer may be provided on all or part of the well layer of the active layer 105.

It is a figure which shows roughly the cross-sectional structure of the semiconductor laser diode (semiconductor light emitting element) of one Embodiment which concerns on this invention. It is a figure which shows roughly the conduction band of an active layer when a functional layer is formed in a barrier layer. It is a figure which shows roughly the conduction band of the normal active layer which does not have a functional layer. It is a graph which shows the measurement result of the optical output of a semiconductor laser diode. (A), (B) is a figure which shows roughly the conduction band of an active layer when a functional layer is formed in a quantum well layer, respectively. It is the figure quoted from the nonpatent literature 1.

Explanation of symbols

101 n-type compound semiconductor substrate 102 n-type buffer layer 103 n-type cladding layer 104 n-side optical confinement layer 105 active layer 106 cap layer 107 p-side optical confinement layer 108 current confinement layer 109 opening 110 p-type superlattice cladding layer 111 p Type contact layer 112 p-side electrode 113 n-side electrode 201, 301, 401, 501 quantum well layer 202, 302, 402, 502 barrier layer

Claims (7)

an n-type semiconductor layer;
a p-type semiconductor layer;
An active layer interposed between the n-type semiconductor layer and the p-type semiconductor layer;
Having a laminated structure including
The active layer includes a functional layer in which a carrier diffusion length in a direction perpendicular to the stacking direction is larger than a carrier diffusion length in the stacking direction.   2. The semiconductor light emitting device according to claim 1, wherein in the functional layer, the carrier mobility in the vertical direction is larger than the carrier mobility in the stacking direction.   3. The semiconductor light emitting element according to claim 1, wherein the functional layer is formed by alternately stacking first semiconductor layers and second semiconductor layers having a wider band gap than the first semiconductor layers. Semiconductor light emitting device having a superlattice structure.   4. The semiconductor light emitting device according to claim 3, wherein an impurity is added to the second semiconductor layer and no impurity is added to the first semiconductor layer. A semiconductor light emitting device according to any one of claims 1 to 4,
The active layer has a quantum well structure in which quantum well layers and barrier layers having a wider band gap than the quantum well layers are alternately stacked,
The number of quantum well layers is two or more,
The semiconductor light emitting device, wherein the barrier layer includes the functional layer. A semiconductor light emitting device according to any one of claims 1 to 4,
The active layer has a quantum well structure in which quantum well layers and barrier layers having a wider band gap than the quantum well layers are alternately stacked,
The semiconductor light emitting device, wherein the quantum well layer includes the functional layer.   7. The semiconductor light emitting device according to claim 1, wherein each of the n-type semiconductor layer, the p-type semiconductor layer, and the active layer is made of a group III nitride compound semiconductor. A semiconductor light emitting device.

Images