JP4320654B2 - Semiconductor light emitting device - Google Patents

Description

  The present invention relates to a semiconductor light emitting device, and more particularly to a high brightness semiconductor light emitting device using a transparent conductive film as a current spreading layer.

  Conventionally, light-emitting diodes (hereinafter abbreviated as LEDs), which are semiconductor light-emitting elements, have recently been able to grow GaN-based and AlGaInP-based high-quality crystals by the MOVPE method, so blue, green, orange, yellow, red High brightness LED can be manufactured.

  However, in order to obtain high brightness, it is necessary to improve current dispersion characteristics so that current is uniformly injected into the LED chip surface. For example, in an AlGaInP-based LED element, the thickness of the current dispersion layer is 5 μm. It was necessary to increase the thickness to about 10 μm. For this reason, the raw material cost required for the growth of the current spreading layer is increased, which inevitably increases the manufacturing cost of the LED element, which hinders low cost manufacturing.

  Therefore, ITO (Indium Tin Oxide) or ZnO (Zinc Oxide) is used as a current-dispersing film that has sufficient translucency and good electric current dispersion characteristics. A method used for the layer has been proposed (see Patent Document 1). A method of directly forming an ITO film on a p-type cladding layer has also been proposed (see Patent Document 2).

Thus, if the ITO film can be used as the current spreading layer, the conventional method of thickening the semiconductor layer to about 5 μm to 10 μm is not required as the current spreading layer, and the corresponding epitaxial layer becomes unnecessary. Thus, it becomes possible to manufacture LED elements having high brightness and epitaxial wafers for LED elements at low cost.
JP-A-8-83927 US Reissue Patent No. 35665 Specification

  However, when an ITO film is used for the window layer, a contact resistance is generated between the semiconductor layer and the ITO film which is a metal oxide, and there is a problem that the forward operation voltage becomes high. That is, the ITO film as the transparent conductive film (transparent electrode) is an n-type semiconductor, while the upper cladding layer in contact with the ITO film is a p-type semiconductor. Therefore, when a forward operating voltage is applied to the LED, a reverse bias state is established between the transparent conductive film (transparent electrode) and the p-type cladding layer, so that a current flows unless a large voltage is applied. Absent.

  As a solution to this, there is a method in which a thin high carrier concentration layer (contact layer) is provided so as to be in contact with the ITO film, and the LED is driven at a low voltage by a tunnel junction (for example, Patent Document 2).

  However, this contact layer is intended to be a tunnel junction and acts as an absorption layer for the light emitted from the active layer. Therefore, it is necessary to form the contact layer in a thin film with a high carrier concentration layer. Since dopant diffusion easily occurs, the following two problems are caused.

  First, when the dopant diffuses in the depth direction of the LED element and diffuses to the active layer of the LED element, it becomes a non-radiative recombination component in the active layer, resulting in a decrease in output.

  Second, the diffusion of the p-type dopant reduces the substantial carrier concentration of the contact layer, which is a thin high carrier concentration layer, making it difficult to achieve the tunnel junction described above and increasing the tunnel voltage. As a result, the drive voltage of the LED element increases.

Although Zn is widely used as a p-type dopant in AlGaInP-based compound semiconductors, it is known that the diffusion constant is relatively large and adverse effects due to thermal processes and the like occur. Therefore, when Zn is used as a dopant and the carrier concentration of the p-type cladding layer is increased, Zn diffuses into the active layer and the characteristics of the light emitting element deteriorate. Furthermore, Mg, which has a smaller diffusion constant than Zn, is used as a p-type impurity to increase the carrier concentration of the p-type cladding layer, and the p-type impurity of the p-type contact layer has a carrier concentration of 1 × 10 ¹⁹ / cm ³ or more. When using Zn which can be obtained relatively easily and can obtain a sufficiently small contact resistance, the p-type dopants Zn and Mg have an interdiffusion action, so that the adverse effects due to the diffusion of the dopants are remarkable.

  Here, there are three ways of adding the dopant: single Mg, single Zn, a combination of Zn and Mg, but the diffusion amount is large in this order. That is, the magnitude relationship of the diffusion amounts is (Mg single) <(Zn single) <(combination of Zn and Mg). Thus, as a method for suppressing interdiffusion, single or simple dopant addition is preferable, but single dopant addition has the following advantages and disadvantages.

  First, with the addition of Mg alone, it is difficult to make the contact layer have a high carrier concentration. For example, it is extremely difficult to achieve a tunnel junction with an ITO film. However, on the other hand, Mg diffusion from the p-type cladding layer or the like to the active layer is extremely small, and there is an advantage that a stable LED element having a long element life can be obtained.

  Next, the addition of Zn alone has a completely opposite relationship compared to the case of addition of Mg alone, and a tunnel junction with the ITO film can be obtained relatively easily. That is, it is relatively easy to increase the carrier concentration of the contact layer. On the other hand, however, a relatively large amount of Zn is diffused from the p-type cladding layer or the like into the active layer, and the element life of the LED element is reduced as compared with the case of adding Mg alone. In the case of Zn, there is a limit to the range of carrier concentration that can be set because it is difficult to obtain crystals with a high carrier concentration compared to Mg for AlGaInP-based materials. There is also a problem that it is difficult to obtain.

  On the other hand, an LED element to which a dopant is added by a combination of Zn and Mg can be obtained to some extent by adopting the following configuration. First, Zn is used as a dopant for the contact layer, thereby obtaining a tunnel junction with the ITO film. Next, Mg is used for the p-type semiconductor layer other than the contact layer such as the buffer layer and the p-type cladding layer to obtain a p-type semiconductor layer having a high carrier concentration, thereby obtaining a high-luminance LED element.

  However, the only problem in the case of using the combination of Zn and Mg is the above-mentioned problem of mutual diffusion of Zn and Mg, and it is necessary to suppress the decrease in device life due to this.

  On the other hand, there is a method of providing a contact layer directly on a p-type cladding layer without providing a buffer layer and providing an ITO film thereon (for example, Patent Document 2). When such a structure is adopted, the p-type cladding is used. Since the thickness of the layer is thin, the dopant diffusion easily reaches the active layer, and the device life is likely to be reduced as described above. Furthermore, since the p-type cladding layer is thin, the element is easily broken due to damage during wire bonding.

  Accordingly, an object of the present invention is to provide a semiconductor light emitting device capable of solving the above-described problems and suppressing a decrease in light emission output with time and an increase in driving voltage in addition to high luminance and low driving voltage. It is to provide.

  In order to achieve the above object, the present invention is configured as follows.

In the semiconductor light emitting device according to the first aspect of the present invention, a light emitting portion including at least an n-type cladding layer, an active layer, and a p-type cladding layer is formed on a substrate, and a p-type dopant concentration is 1 × above the light emitting portion. A thin As-type p-type contact layer formed by adding a dopant of a material different from the dopant added to the p-type cladding layer and having a density of 10 ¹⁹ / cm ³ or more, is formed on the p-type contact layer. In a semiconductor light emitting device having a current spreading layer made of a metal oxide material formed thereon, between the p-type cladding layer and the p-type contact layer, p-type conductivity and intentionally or unavoidably C (carbon) is included and the film thickness is L = 6.872 × 10 ⁻¹⁴ × N ^0.733 [where N is the C concentration (unit: cm ⁻³ ), and L is the p-type contact layer Dopa added to DOO diffusion length (unit: [mu] m) and is] to have the buffer layer composed of a group III / V semiconductor is diffusion length L or more sought characterized.

  Here, “intentionally” includes being positively or intentionally added (doping), and “inevitable” includes such active, intentional, or intentional This refers to an unavoidable phenomenon in which C (carbon) is naturally mixed into a crystal, although it is not added (doping). Furthermore, the expressions “undoped”, “non-added”, and “not added” used in the present specification mean that active, intentional or intentional addition (doping) is not performed. It does not indicate the phenomenon that impurities such as C (carbon) and H (hydrogen) are inevitably mixed into the crystal.

  According to a second aspect of the present invention, in the semiconductor light emitting device according to the first aspect, the dopant of the p-type cladding layer is Mg, and the dopant of the p-type contact layer is Zn.

According to a third aspect of the present invention, in the semiconductor light emitting device according to the first or second aspect, the material constituting the p-type contact layer is Al _x Ga _{1 -x} As (where 0 ≦ X ≦ 0.4). It is characterized by that.

There is a limit to semiconductor materials that can stably realize a contact layer having a high carrier concentration of 1.0 × 10 ¹⁹ / cm ³ or more. As a semiconductor material, Zn-doped Al _x Ga _1-x As (0 ≦ X ≦ 0.4) is optimal. However, since the AlGaAs is not transparent to the emission wavelength, it is necessary to form it with a thin film of about 30 nm or less.

According to a fourth aspect of the present invention, in the semiconductor light emitting device according to any one of the first to third aspects, the buffer layer is Al _x Ga _1-x As (provided that 0.4 ≦ X ≦ 1). And

According to a fifth aspect of the present invention, in the semiconductor light emitting device according to any one of the first to fourth aspects, the current dispersion layer is made of ITO (tin-added indium oxide), SnO ₂ (tin oxide), ATO (antimony-added tin oxide). In ₂ O ₃ (indium oxide), ZnO (zinc oxide), GZO (gallium-doped zinc oxide), BZO (boron-doped zinc oxide), AZO (aluminum-doped zinc oxide), CdO (cadmium oxide), CTO (tin-added) It is characterized in that it is formed using at least one metal oxide material of cadmium oxide) and IZO (indium-added zinc oxide).

  According to a sixth aspect of the present invention, in the semiconductor light emitting device according to any one of the first to fifth aspects, the film thickness of the current dispersion layer is a calculation formula d = A × λp / (4 × n) [where A Is a constant (1 or 3), λp is a wavelength (unit: nm), and n is a refractive index].

According to a seventh aspect of the present invention, in the semiconductor light emitting device according to any one of the first to sixth aspects, the material constituting the light emitting portion is (Al _x Ga _{1 -x} ) _y In _{1 -y} P (where 0 ≦ X ≦ 1, 0.4 ≦ Y ≦ 0.6), and the content of Al (aluminum) contained in the p-type cladding layer and the n-type cladding layer is the content of Al contained in the active layer. It is characterized by being higher than the rate.

According to an eighth aspect of the present invention, in the semiconductor light emitting device according to any one of the first to seventh aspects, the carrier concentration of the current dispersion layer is 7 × 10 ²⁰ / cm ³ or more.

  The invention of claim 9 is the semiconductor light emitting device according to any one of claims 1 to 8, wherein the thickness of the p-type contact layer is in the range of 1 nm to 30 nm.

  A tenth aspect of the present invention is the semiconductor light emitting device according to any one of the first to ninth aspects, wherein the refractive indexes of the high refractive index material and the low refractive index material differ between the substrate and the n-type cladding layer. Two types of semiconductor layers are used as one pair, and a light reflection layer having 10 to 30 pairs is formed.

According to an eleventh aspect of the present invention, in the semiconductor light emitting device according to any one of the first to tenth aspects, the main material constituting the light reflecting layer is Al _x Ga _1-x As (provided that 0.4 ≦ X ≦ 1), or _{(Al x Ga 1-x)} y in 1-y P ( where, 0 ≦ X ≦ 1,0.4 ≦ Y ≦ 0.6), or characterized by comprising these combinations.

  The invention of claim 12 is the semiconductor light emitting device according to any one of claims 1 to 11, wherein the active layer is composed of a light emitting layer having a narrow band gap and a barrier layer having a wider band gap than the light emitting layer. And a plurality of these layers are laminated.

  A thirteenth aspect of the present invention is the semiconductor light emitting device according to the twelfth aspect of the present invention, wherein the active layer is a quantum well structure comprising a thin film having a thickness of 9 nm or less included in the active layer, or the light emission. The crystal lattice constant of the layer is a strained quantum well structure different from the lattice constant of the n-type cladding layer or the p-type cladding layer.

According to a fourteenth aspect of the present invention, in the semiconductor light emitting device according to the first to thirteenth aspects, the Mg concentration contained in the p-type cladding layer is in the range of 1 × 10 ¹⁷ / cm ³ or more and 5 × 10 ¹⁸ / cm ³ or less. And the p-type cladding layer set in the above range occupies all or at least a part of the layer.

When the Mg concentration is less than 1 × 10 ¹⁷ / cm ³ , the carrier concentration of the p-type cladding layer becomes too low, and it becomes difficult to exhibit a sufficient effect as the carrier supply layer. Further, when excessive Mg addition exceeding 5 × 10 ¹⁸ / cm ³ is performed, crystal defects substantially depending on the Mg concentration are generated in the p-type cladding layer, and the diffusion of the dopant is promoted and the internal quantum efficiency of the LED is increased. In general, the light emission output of the LED element is reduced.

  A fifteenth aspect of the present invention is the semiconductor light emitting device according to any one of the first to fifteenth aspects, wherein the substrate has a semiconductor material of any one of GaAs, Ge, and Si, or a thermal conductivity higher than a thermal conductivity of the Si. A metal material having a high rate is used.

  The invention according to claim 16 is the semiconductor light emitting device according to any one of claims 1 to 15, wherein a semiconductor layer in which no dopant is added between the active layer and the p-type cladding layer, the p-type cladding. Either a semiconductor layer having a lower p-type dopant concentration than the layer, a semiconductor layer in which a n-type dopant and a p-type dopant are added simultaneously, and a pseudo-neutral semiconductor layer, or a combination thereof Further, the diffusion preventing layer made of a semiconductor layer is provided in a range of 300 nm or less.

  The invention according to claim 17 is the semiconductor light emitting device according to any one of claims 1 to 16, wherein a semiconductor layer to which no dopant is added is provided between the active layer and the n-type cladding layer, and the n-type cladding. Either a semiconductor layer having a lower n-type dopant concentration than the layer, a semiconductor layer in which a n-type dopant and a p-type dopant are added simultaneously, and a pseudo-neutral semiconductor layer, or a combination thereof The diffusion prevention layer made of a semiconductor layer is provided in a range of 200 nm or less.

  In the present invention, the buffer layer intentionally or inevitably contains C (carbon), and the Zn diffusion length can be set by the C concentration value, so that the Zn diffusion length is within the range of the thickness of the buffer layer. Can be set to fit.

  According to the invention, it is possible to effectively suppress the interdiffusion of dopants from the p-type contact layer and the other p-type semiconductor layers, and to produce an LED element having a high output and a low operating voltage, and a light emission output over time. LED elements capable of suppressing the decrease in the voltage and the increase in the driving voltage can be manufactured.

  Hereinafter, the present invention will be described based on the illustrated embodiments.

  FIG. 1 shows a configuration of a light emitting diode according to the present embodiment. In this light emitting diode, an n-type GaAs buffer layer 2 is formed on an n-type GaAs substrate 1 which is a semiconductor substrate, and an n-type AlGaInP cladding layer (also simply referred to as an n-type cladding layer) is formed on the n-type GaAs buffer layer 2. 3. An undoped AlGaInP active layer (also simply referred to as an active layer) 4 and a p-type AlGaInP cladding layer (also simply referred to as a p-type cladding layer) 5 are sequentially crystal-grown to form a light emitting portion. Further, an As-based p-type AlGaAs contact layer (also simply referred to as a p-type contact layer) 7 to which a p-type dopant is added at a high concentration is laminated on the p-type cladding layer 5. Further, an ITO film (current dispersion layer) 8 which is a transparent conductive film is laminated on the p-type contact layer 7 as a current dispersion layer made of a metal oxide material, and a surface electrode 9 is formed on the surface side of the ITO film 8. A back electrode 10 is formed on the back side of the n-type GaAs substrate 1.

The p-type contact layer 7 is Al _x Ga _1-x As (where, 0 ≦ X ≦ 0.4) consists, thickness is at 1nm or 30nm or less, Zn as p-type dopant, the carrier concentration of 1 × It is added at a high concentration of 10 ¹⁹ / cm ³ or more.

The thickness of the ITO film 8 which is a current dispersion layer is calculated by the following formula: d = A × λp / (4 × n) [where A is a constant (1 or 3), λp is a wavelength (unit: nm), and n is The refractive index is in the range of ± 30% of d obtained by The ITO film 8 serving as the current dispersion layer is formed by vacuum deposition or sputtering, and has a carrier concentration of 7 × 10 ²⁰ / cm ³ or more immediately after the film formation.

  As a feature of this light emitting diode, a p-type buffer layer 6 made of a III / V semiconductor in which Mg is added as a p-type dopant between the p-type contact layer 7 and the p-type cladding layer 5. Form.

Specifically, the p-type buffer layer 6 is Al _x Ga _1-x As (0.4 ≦ X ≦) that is optically transparent to the emission wavelength and lattice-matched to the AlGaInP-based material constituting the light emitting portion. 1). Further, the p-type buffer layer 6 has a film thickness t equal to or greater than the diffusion length of Zn added to the contact layer 7.

  The buffer layer 6 is an AlGaAs layer having a high Al composition, is optically transparent with respect to the emission wavelength of an LED element made of an AlGaInP-based material, and is also compared with a quaternary material such as AlGaInP. Since the crystal growth is easy and the lattice matching with the AlGaInP-based material constituting the light emitting portion is almost the same, the material can reduce the operating voltage of the LED element.

The diffusion length of Zn added to the contact layer 7 has a correlation with the concentration of C (carbon) contained in the p-type buffer layer 6, as shown in FIG.
L = 6.872 × 10 ⁻¹⁴ × N ^0.733
[Where N is the C concentration (unit: cm ⁻³ ), and L is the diffusion length (unit: μm) of the dopant added to the p-type contact layer].

  Therefore, in the embodiment of the present invention, the thickness t of the p-type buffer layer 6 is such that the thickness is larger than the Zn diffusion length L obtained from the above relational expression or the curve of FIG. The C concentration in the mold buffer layer 6 is set. The setting of the C concentration is precisely controlled by parameters such as the V / III ratio or Al composition when the buffer layer 6 is grown.

As a specific example, in the case of the p-type buffer layer 6 having a film thickness of 5 μm, for example, the C concentration in the p-type buffer layer 6 is set to 1 × 10 ¹⁸ / cm ³ . By doing so, the diffusion length of Zn added to the contact layer 7, which is obtained from the above equation or the curve of FIG. 2, is about 1.1 μm, and is within the thickness of the p-type buffer layer 6. As a result, the diffusion of Zn in the p-type contact layer 7 is suppressed, and accordingly, the mutual diffusion of Zn in the p-type contact layer 7 and dopant Mg from the other p-type semiconductor layers 5 and 6 is effectively performed. Can be suppressed.

  As a result of intensive studies, the present inventors have found that the mutual diffusion reaction between Zn contained in the contact layer 7 and Mg contained in the p-type cladding layer 5 and the p-type buffer layer 6 is caused in the buffer layer 6. It has been found that the concentration is very close to the concentration of C contained, and that the phenomenon is particularly remarkable when the material is an AlGaAs layer having a high Al composition. Therefore, by controlling closely the parameters such as the C concentration of the buffer layer 6 made of such a material, the V / III ratio during the growth of the buffer layer 6, or the Al composition, the mutual diffusion distance of Zn and Mg is controlled, and the long life The film thickness of the buffer layer 6 necessary for obtaining a simple LED element can be suitably set as appropriate.

In the above embodiment, between the n-type GaAs substrate 1 and the n-type cladding layer 3, for example, between the n-type GaAs buffer layer 2 and the n-type cladding layer 3, a high refractive index material and a low refractive index material are used. Two types of semiconductor layers having different refractive indexes may be used as one pair, and a light reflecting layer having 10 to 30 pairs may be formed.
Moreover, in the said embodiment, it is from the semiconductor layer to which the dopant is not added between the active layer 4 and the p-type cladding layer 5, or between the active layer 4 and the n-type cladding layer 3, the p-type cladding layer 5. Diffusion formed by a semiconductor layer with a low dopant concentration, a semiconductor layer in which a n-type dopant and a p-type dopant are added simultaneously, and a pseudo-neutral semiconductor layer, or a combination thereof A prevention layer may be provided. The thickness of the diffusion preventing layer is desirably 300 nm or less when intervening between the active layer 4 and the p-type cladding layer 5. Further, when a diffusion prevention layer is provided between the active layer 4 and the n-type cladding layer 3, the film thickness is desirably 200 nm or less.

  Next, the grounds for adopting the light emitting diode structure of the above-described embodiment will be described in detail.

  First, regarding the setting of the film thickness of the AlGaAs buffer layer 6 shown in the present invention, Zn added to the contact layer 7 and Mg added to the other p-type semiconductor layers 5 and 6 as a main premise. In order to obtain a long-life and high-reliability LED element, it is important to set the film thickness so that the above-described impurities can be prevented from being mixed into the active layer 4 due to the interdiffusion caused by the above.

The mutual diffusion action of Zn and Mg at that time is closely related to the concentration of C contained in the buffer layer 6, and the inventors of the present invention include the diffusion length (diffusion distance) of the impurities in the buffer layer 6. It has been found that it can be controlled by the C concentration. This can be known from the experimental results shown in FIG. 2, for example, showing the relationship between the C concentration in the AlGaAs buffer layer 6 and the diffusion length of Zn. Based on these results, the diffusion length of Zn added to the contact layer 7 is a relational expression of L = 6.872 × 10 ⁻¹⁴ × N ^0.733 [where N is the C concentration (unit: cm ^−3). L is the diffusion length (unit: μm) of the dopant added to the p-type contact layer 7], and in the present invention, the film has a Zn diffusion length L or more determined by the above relational expression. By providing the AlGaAs buffer layer 6 having the thickness t, it is possible to obtain an LED element having good characteristics in both initial characteristics and reliability.

  Incidentally, there is no need to particularly set an upper limit for the film thickness value t of the buffer layer 6. For example, when the film thickness is made thicker than the preferred film thickness obtained for the above reason, The following can be said.

  A sufficient effect can be expected for the current dispersion characteristics of the LED element by the current dispersion layer 8 made of a metal oxide provided on the contact layer 7, for example, an ITO or ZnO transparent conductive film. For this reason, the buffer layer 6 itself is not particularly required to have a current dispersion characteristic. Even if the buffer layer 6 having a film thickness of about 10 μm or more is provided, the current dispersion effect obtained by the current dispersion layer 8 is dominant, and a dramatic improvement in output as an LED element is achieved. I can't hope. Rather, the cost of manufacturing the LED element is increased, resulting in a demerit that the cost of the LED element is increased. Therefore, it can be said that it is preferable in production that the film thickness of the buffer layer 6 is limited to a degree satisfying the above-described preferred form.

  Secondly, the current spreading layer made of a metal oxide, for example, the ohmic contact 7 layer in contact with the ITO film 8 needs to be doped with a very high concentration, that is, has a very high carrier concentration. It is necessary to be.

Specifically, in the case of the contact layer 7 to which Zn (zinc) is added, it is preferable that the crystal material has an Al mixed crystal ratio of 0 to 0.4, that is, GaAs to Al _0.4 GaAs. The carrier concentration is preferably 1 × 10 ¹⁹ / cm ³ or more, and this is preferably as high as possible.

  The ITO film 8 basically belongs to an n-type semiconductor material, and the LED element is generally manufactured by p-side-up in many cases. For this reason, as a matter of course, the LED element in which the ITO film 8 is applied to the current spreading layer has an n / p / n junction from the substrate side. For this reason, in the LED element, a large potential barrier is generated at the interface between the ITO film 8 and the p-type semiconductor layer, and the LED usually has a very high operating voltage.

In order to solve this problem, the p-type semiconductor layer is required to have a very high carrier concentration. Further, the reason why the band gap (forbidden band width) of the contact layer 7 is narrow depends largely on the fact that a narrow band gap material is generally easier to increase the carrier. Further, in relation to the increase in the carrier of the contact layer 7, the carrier concentration of the current spreading layer in contact with the contact layer 7, for example, the ITO film 8, is also important for reducing the tunnel voltage. For the same reason as the contact layer 7, it preferably has a carrier concentration of 7 × 10 ²⁰ / cm ³ or more.

In addition, as a method for forming a transparent conductive film (current dispersion layer 8 in the present invention) having a carrier concentration of 7 × 10 ²⁰ / cm ³ or more, a vacuum deposition method or a sputtering method can be used. In particular, in the sputtering method, it is known that a transparent conductive film having an extremely high carrier concentration can be obtained when a RF sputtering is applied to a DC sputtering method. Other forming methods include a coating method using a MOD (Metal Organic Decomposition) solution and a spray pyrolysis method. However, it is difficult to obtain a transparent conductive film having a high carrier concentration by these methods, and the formation is assumed. In this case, adverse effects may occur due to the heat applied to the epitaxial wafer.

  Thirdly, the thickness of the contact layer 7 is preferably in the range of 1 nm to 30 nm. This is because the contact layer 7 absorbs the light emitted from the active layer 4 at least, so that the light emission output decreases as the film thickness increases. . The relationship between the film thickness of these contact layers 7 and the transmittance at the emission wavelength of the LED element is shown in FIG.

  According to the figure, the transmittance of visible light at the emission wavelength is lowered depending on the film thickness of the contact layer 7. That is, the upper limit of the film thickness of the contact layer 7 is preferably 30 nm from the viewpoint of obtaining a high-power LED element. Further, when the thickness of the contact layer 7 is less than 1 nm, that is, about several angstroms (Å), the tunnel junction between the ITO film 8 and the contact layer 7 becomes difficult. It becomes difficult to stabilize the operating voltage and operating voltage. Therefore, the thickness of the contact layer 7 in contact with the ITO film 8 is preferably 1 nm to 30 nm.

  Fourth, the thickness of the current dispersion layer 8 made of a metal oxide is calculated by the following formula: d = A × λp / (4 × n) [where A is a constant (1 or 3), and λp is a wavelength (unit: nm) and n is the refractive index] and is in the range of ± 30% of d obtained from In addition, the wavelength λp in the above calculation formula refers to the emission peak wavelength of the LED element.

  As a current dispersion layer formed on the LED epitaxial wafer, for example, the ITO film 8 has a refractive index approximately in the middle between the semiconductor layer and the air layer, and optically functions as an antireflection film. Therefore, in order to improve the light extraction efficiency of the LED and obtain an LED element with higher output, it is preferable to set the film thickness according to the above calculation formula. However, as the thickness of the ITO film 8 is naturally increased, the transmittance tends to deteriorate. When the transmittance of the ITO film 8 decreases, the ratio of the light emitted from the active layer 4 absorbed by the ITO film 8 increases, and as a result, the light emission output decreases. Furthermore, as the film thickness of the current spreading layer 8 increases, light interference in the layer 8 increases, and the wavelength region with high light extraction efficiency becomes narrow. For these, the ITO film 8 is appropriately formed on the GaAs substrate 1, light is incident on the sample vertically, and the result of measuring the spectrum of the reflected light at this time is shown in FIG.

  That is, for these reasons, the more preferable thickness d of the current spreading layer 8 is in the above calculation formula, and the constant A is preferably 1 or 3. In the most preferred example, the constant A is 1. Further, the thickness of the current dispersion layer formed on the LED epitaxial wafer, for example, the ITO film 8 suffices to be within a range of ± 30% of the value d obtained by the above calculation formula. This is because the wavelength band having a low optical reflectance as the antireflection film, that is, the wavelength band having a high light extraction efficiency has a certain width. For example, as an antireflection film, the allowable value of the film thickness at which the reflectance is 15% or less when light is vertically incident on the LED epitaxial wafer on which the ITO film 8 is formed is obtained from the above formula. Of ± 30%. When the film thickness exceeds the range of ± 30% of d, the effect as an antireflection film is reduced, and the light emission output of the LED element is relatively lowered.

  Fifth, the thickness of the diffusion prevention layer provided adjacent to the active layer 4 is 300 nm or less when intervening between the active layer 4 and the p-type cladding layer 5, and the active layer 4 and n When intervening with the mold cladding layer 3, it is desirable that the thickness be 200 nm or less. The basis for the upper limit of these film thickness values is as follows.

First, according to the intention of the present invention, by appropriately setting the C (carbon) concentration contained in the buffer layer 6 and the film thickness of the layer 6, interdiffusion of Zn and Mg as p-type dopants. The action is preferably suppressed, and it does not diffuse into the active layer 4 and mix in a large amount. However, when the film thickness value of the buffer layer 6 is set to have almost no safety margin, or depending on the dopant concentration error or the film thickness error in the epitaxial growth process, the dopant is mixed into the active layer 4. Can happen. In such a case, in addition to appropriately setting the C concentration contained in the buffer layer 6 and the film thickness of the buffer layer 6, a diffusion preventing layer as described above is provided between the active layer 4 and the p-type cladding layer 5. By taking measures to provide the LED element, the life and stability of the LED element can be improved. However, the diffusion preventing layer is not necessarily simply thick, and the film thickness value has the upper limit of the above value. That is, if the diffusion prevention layer is too thick, the carrier (in this case, holes) from the p-type cladding layer 5 is not efficiently injected, and the forward voltage of the LED element increases. It can be a cause of deteriorating the characteristics required for the above. Therefore, as a suitable example, it can be said that the film thickness of the diffusion prevention layer provided between the active layer 4 and the p-type cladding layer 5 is in the range of 300 nm or less, and more preferably 200 nm or less.

  Next, for the same reason as the case of the diffusion preventing layer provided on the p-type cladding layer 5 side, not only a small amount of n-type dopant added to the n-type cladding layer 3 may diffuse into the active layer 4. Even if the diffusion distance is short, the n-type dopant added to the n-type cladding layer 3 may be mixed into the active layer 4 by the so-called memory effect during the growth of the n-type cladding layer 3 and the active layer 4. . Due to the mixing of these impurities, the light emission output of the LED element decreases as in the case of diffusion of the p-type dopant. As a means for suitably solving such problems, it is desirable to provide a diffusion prevention layer between the active layer 4 and the n-type cladding layer 3. In this case, the upper limit of the diffusion preventing layer is preferably in the range of 200 nm or less, and the grounds thereof are the same as those in the case of the diffusion preventing layer provided on the p-type cladding layer 5 side. Therefore, it can be said that a more preferable film thickness is 100 nm or less.

  Sixth, it is preferable that the total number of stacked light reflection layers, that is, the so-called number of pairs is in the range of about 10 to 30 pairs. The basis for the lower limit depends on the number of pairs required to have sufficient reflectivity as the light reflecting layer, which is about 10 pairs or more. FIG. 8 shows the relationship between the number of stacked pairs of light reflecting layers and the vertical reflectance.

Next, the grounds for the above upper limit of the number of pairs are as follows. As the light reflecting layer is stacked infinitely thick, the reflectance does not increase and the light emission output of the LED element does not increase. As shown in FIG. 8, the reflectivity of the light reflecting layer almost completely shows a saturation tendency from the vicinity of more than 20 pairs, and is completely saturated in the vicinity of more than 30 pairs. Therefore, the number of pairs capable of obtaining an effective reflectance is required to be more than a certain level. Further, from the viewpoint of inexpensively and efficiently manufacturing LED elements and LED epitaxial wafers, pairs of light reflecting layers are required. There is an upper limit on the number. That is, it is 10 pairs or more and 30 pairs or less as described above, and a more preferable range is selected from 15 pairs or more and 25 pairs or less for the above reason. As the preferred material that constitutes the light-reflecting layer of _{said, Al x Ga 1-x As} ( where, 0.4 ≦ X ≦ 1), or _{(Al x Ga 1-x)} y In 1-y P (However, 0 ≦ X ≦ 1, 0.4 ≦ Y ≦ 0.6). The reason for selecting these materials is strongly dependent on being optically transparent with respect to the wavelength of the light emitted from the LED element among the materials substantially lattice-matched to the GaAs substrate 1. As is well known, the DBR that is the light reflection layer has a wider reflection wavelength band and higher reflectivity when the refractive index difference between the two types of materials is larger. Therefore, the material to be selected should be selected from the above materials.

Seventh, the AlGaAs constituting the buffer layer 6 is preferably Al _x Ga _1-x As (where 0.4 ≦ X ≦ 1). The reason for setting the range on the left is that the buffer layer 6 is located on the surface of the LED element where light is emitted, that is, on the LED element surface side, so that it is optically transparent to the light emitted from the LED. This is because it is advantageous in terms of output. Even if an AlGaAs layer deviating from the above composition formula is used, the effect of the present invention is not hindered, but is not preferable from the viewpoint of obtaining a high-power LED element.

Eighth, the concentration of Mg contained in the p-type cladding layer 5 is preferably in the range of 1 × 10 ¹⁷ / cm ³ to 5 × 10 ¹⁸ / cm ³ . The reason for determining the lower limit is that if the Mg concentration is lower than the above lower limit, the carrier concentration of the layer 5 becomes too low, and it becomes difficult to exert a sufficient effect as a carrier supply layer, and consequently the light emission output of the LED element is reduced. May be connected. On the other hand, if excessive Mg is added, crystal defects are generated in the p-type cladding layer 5 almost depending on the Mg concentration, which promotes dopant diffusion and reduces the internal quantum efficiency of the LED. This is because it causes a decrease in the light output of the LED element.

  As Example 1, a red LED epitaxial wafer having a structure as shown in FIG. An epitaxial growth method, an epitaxial layer thickness, an epitaxial structure, an electrode formation method, and an LED element manufacturing method are as follows.

An n-type (Si-doped) GaAs buffer layer 2 (thickness 200 nm, carrier concentration 1 × 10 ¹⁸ / cm ³ ), n-type (Si-doped) is formed on an n-type GaAs substrate 1 whose dopant is Si by MOVPE. (Al _0.7 Ga _0.3 ) _0.5 In _0.5 P cladding layer 3 (film thickness 400 nm, carrier concentration 1 × 10 ¹⁸ / cm ³ ), undoped (Al _0.1 Ga _0.9 ) _0.5 In _0.5 P active layer 4 (film thickness 600 nm), p Type (Mg doped) (Al _0.7 Ga _0.3 ) _0.5 In _0.5 P cladding layer 5 (film thickness 400 nm, carrier concentration 1.2 × 10 ¹⁸ / cm ³ ), p type (Mg doped) Al _0.85 Ga _0.15 As buffer layer 6 (Film thickness 5 μm, carrier concentration 2 × 10 ¹⁸ / cm ³ ), p-type (Zn doped) Al _0.1 Ga _0.9 As contact layer 7 (film thickness 3 nm, carrier concentration 7.5 × 10 ¹⁹ / cm ³ ) Stacked growth It was.

The growth temperature in the MOVPE growth was 650 ° C. from the n-type GaAs buffer layer 2 to the p-type buffer layer 6, and the p-type contact layer 7 was grown at 550 ° C. The other growth conditions were a growth pressure of about 6666 Pa (50 Torr), a growth rate of each layer of 0.3 to 1.0 nm / sec, and a V / III ratio of about 150. However, the V / III ratio of the p-type contact layer 7 was set to 10. The buffer layer 6 is AlGaAs having an Al composition of about 0.8 to 0.9, and the concentration of C (carbon) contained in the buffer layer 6 is about 1 × 10 ¹⁸ / cm ^3. Similarly, the V / III ratio at the time of growth was set (the V / III ratio at this time is about 50). Incidentally, the V / III ratio mentioned here is a ratio (quotient) when the denominator is the number of moles of a group III material such as TMGa or TMAl and the molecule is the number of moles of a group V material such as AsH ₃ or PH _3. Point to.

For example, trimethylgallium (TMGa) or triethylgallium (TEGa) in the case of Ga, trimethylaluminum (TMAl) in the case of Al, and trimethylindium (TMIn) in the case of In are used as raw materials used in the MOVPE growth. In addition, a hydride gas such as arsine (AsH ₃ ) was used as the As source and phosphine (PH ₃ ) was used as the P source. For example, disilane (Si ₂ H ₆ ) was used as an additive material for an n-type layer such as the n-type GaAs buffer layer 2. Further, biscyclopentadienylmagnesium (Cp ₂ Mg) was used as an additive material for the conductivity determining impurities of the p-type layers such as the p-type cladding layer 5 and the p-type buffer layer 6. However, diethyl zinc (DEZn) was used only for the p-type contact layer 7.

In addition, hydrogen selenide (H ₂ Se), monosilane (SiH ₄ ), diethyl tellurium (DETe), and dimethyl tellurium (DMTe) can also be used as an additive material for the conductivity determining impurity of the n-type layer. In addition, dimethyl zinc (DMZn) can also be used as a supply source of Zn.

  Next, after this LED epitaxial wafer was unloaded from the MOCVD apparatus, an ITO film 8 having a film thickness of about 290 nm was formed on the surface of the wafer, that is, on the surface side of the p-type contact layer 7 by vacuum deposition. In this structure, the ITO film 8 becomes a current dispersion layer.

At this time, the glass substrate for evaluation set in the same batch of ITO vapor deposition work was taken out and cut into a size capable of Hall measurement, and then the electrical characteristics of the ITO film alone were evaluated. As a result, the carrier concentration was 1.05 × 10 ²¹ / cm ³ , the mobility was 20.3 cm ² / Vs, and the resistivity was 2.94 × 10 ⁻⁴ Ω · cm.

  Then, a circular surface electrode 9 having a diameter of about 110 μm is vacuum-deposited in a dot matrix form on the upper surface of this epitaxial wafer using a well-known method and equipment used in a general photolithography process such as a resist and a mask aligner. Formed by the law. At this time, the lift-off method was used for forming the electrode after the deposition. The surface electrode 9 was formed by depositing Ni (nickel) and Au (gold) in the order of 20 nm and 500 nm, respectively. Further, a back electrode 10 was formed on the entire bottom surface of the epitaxial wafer by the same vacuum deposition method. The back electrode 10 is formed by depositing AuGe (gold / germanium alloy, germanium content 7.4%), Ni (nickel), and Au (gold) in the order of 60 nm, 10 nm, and 500 nm, respectively, and then alloying the electrodes. The alloy process was performed by heating to 440 ° C. in a nitrogen gas atmosphere and heat treating for 5 minutes.

  Thereafter, the LED epitaxial wafer with an electrode configured as described above was cut using a dicing apparatus so that the circular surface electrode 9 was at the center, and an LED bare chip having a chip size of 300 μm square was produced. Furthermore, the LED bare chip was mounted (die bonding) on the TO-18 stem via Ag paste, and then the LED bare chip mounted was further wire-bonded to produce an LED element.

  As a result of evaluating the initial characteristics of the LED element thus fabricated, an LED element having initial characteristics of a light emission output of 0.99 mW and a working voltage of 1.842 V when energized with 20 mA (at the time of evaluation) was obtained.

  Furthermore, the LED element was driven at 50 mA in an environment of normal temperature (about 23 ° C.) and normal humidity (about 40%), and a continuous energization test was performed for 168 hours (one week) as it was. As a result, the relative comparison value before the test is 102.1% output (the light emission output before energization is 100%, hereinafter abbreviated as relative output), and the operating voltage is 1.843V (about 0.1% increase). It was.

  Moreover, the SIMS analysis of the LED element of the state immediately after LED element preparation and the state after performing an electricity supply test on said conditions after LED element preparation was performed. The SIMS analysis result after the current test at this time is shown in FIG. 12 (Note that the sample of the LED element shown in FIG. 12 is mechanically polished to remove several μm from the surface in order to improve the measurement resolution of SIMS analysis. ing).

As a result of SIMS analysis, the concentration of C (carbon) contained in the buffer layer 6 before and after the energization test is approximately 1.0 × 10 ¹⁸ / cm ³ , and in the LED element of this example after the energization test, No diffusion of Zn was confirmed.

  Here, based on the above results, FIG. 13 shows the results of changing the V / III ratio when producing the AlGaAs buffer layer 6 as needed and measuring the C concentration in the layer 6 by SIMS analysis at that time. According to the figure, it can be seen that the C concentration contained in the buffer layer 6 has a good correlation with the V / III ratio when the AlGaAs buffer layer 6 is produced. However, the C concentration of the layer 6 is not simply determined by the V / III ratio, but also changes depending on the growth temperature at the time of fabrication, the Al composition of the layer 6 or the like. It is not limited to the III ratio.

Moreover, the diffusion length of Zn added to the contact layer 7 in the LED structure described in Example 1 was measured using a sample prepared by changing the V / III ratio shown in FIG. The relationship between the C concentration in the AlGaAs buffer layer and the Zn diffusion length is shown in FIG. According to the figure, it can be seen that the diffusion length of Zn strongly depends on the C concentration in the AlGaAs buffer layer. These results can be expressed as
L = 6.872 × 10 ⁻¹⁴ × N ^0.733
[Wherein N is a C (carbon) concentration (unit: cm ⁻³ ), L is a diffusion length of a dopant added to the p-type contact layer (unit: μm)], and the above relationship In the case of an AlGaAs buffer layer having a film thickness equal to or greater than the diffusion length L determined by the equation, Zn added to the contact layer 7 does not diffuse and reach the active layer 4, and both initial characteristics and reliability are achieved. A very good LED element can be obtained. That is, the reason why the characteristics of the LED element described in Example 1 were favorable was as described above. In the LED elements shown in Example 1, no element destruction occurred in almost all the elements.

  As Example 2, an epitaxial wafer for red LED having a structure shown in FIG. The epitaxial growth method, the epitaxial layer thickness, the epitaxial layer structure, and the LED device manufacturing method are basically the same as those in the first embodiment (FIG. 1). The points different from the first embodiment will be listed below, and detailed explanations will be given accordingly.

  In Example 2, a structure in which a so-called undoped layer that is not actively added between the active layer 4 and the p-type cladding layer 5 is provided as the diffusion preventing layer 11. The diffusion prevention layer 11 is a layer for preventing the p-type dopant diffusing from the p-type semiconductor layer above the p-type cladding layer 5 from being mixed into the active layer 4. The composition of the layer 11 was the same as that of the p-type cladding layer 5 and the film thickness was 100 nm.

  In addition, as a reference example of Example 2, an LED epitaxial wafer in which the above-described diffusion prevention layer 11 was inserted under the same conditions into an LED element described in Comparative Example 1 (FIG. 9) described later was also produced.

  Next, the LED epitaxial wafer produced as described above is made into an element, and the process is the same as in the first embodiment.

  As a result of evaluating the initial characteristics of the LED element thus manufactured, even in the case of the reference example of Example 2, the LED characteristics at the time of energization of 20 mA (during evaluation) were a light emission output of 0.96 mW and an operating voltage of 1.854 V. Yes, an LED element having excellent initial characteristics could be obtained.

  Furthermore, when an energization test was performed under the same conditions as in Example 1, the relative output of the LED element in the reference example of Example 2 was 76.1% and the operating voltage was 1.904 V (an increase of about 2.6%). The relative output was higher than that in Comparative Example 1 below.

  Next, the LED element of this Example 2, that is, the LED element in which the diffusion preventing layer 11 was added to the LED structure described in Example 1 as shown in FIG. 3 was evaluated. As a result, in the case of Example 2, the initial characteristics of the light emission output of 0.97 mW and the operating voltage of 1.843 V were obtained.

  In addition, when the LED element of Example 2 was subjected to an energization test in the same manner as in Example 1, the relative output of the LED element was 101.3% and the operating voltage was 1.843 V (no change). .

As described above, the diffusion prevention layer 11 shown in the present embodiment 2 does not inhibit the fundamental diffusion of the p-type dopant , but the active layer 4 is provided by providing the layer 11 when the diffusion occurs. It is possible to suppress the diffusion amount of the p-type dopant that is diffused and mixed. As a result, it is possible to obtain an LED element that is superior in terms of relative output as compared with the LED element shown in Comparative Example 1 as described below. Further, the layer 11 is not limited to the structure in which the diffusion of the p-type dopant occurs, and even when the present invention is implemented, the layer 11 does not have an adverse effect, and the effect of the diffusion preventing layer 11 can be handled in the same manner.

  As Example 3, a red LED epitaxial wafer having a structure as shown in FIG. The epitaxial growth method, the epitaxial layer thickness, the epitaxial layer structure, and the LED device manufacturing method are basically the same as those in the first embodiment.

However, 20 n-type AlInP layers and 20 n-type Al _0.4 Ga _0.6 As layers are alternately provided between the n-type GaAs buffer layer 2 and the n-type cladding layer 3 for a total of 20 pairs of DBRs ( A light reflecting layer 12 made of a distributed Bragg reflector (distributed Bragg reflector) was provided.

The film thickness constituting the light reflecting layer 12 is an equation of “λp / 4 × n” [where λp is the emission peak wavelength (unit: nm) of the LED element, and n is the semiconductor material constituting the light reflecting layer 12. Refractive index]. The carrier concentration of the light reflecting layer 12 was uniformly set to about 1 × 10 ¹⁸ / cm ³ .

  As a result of evaluating the initial characteristics of the LED element thus fabricated, an LED element having excellent initial characteristics with a light emission output of 1.53 mW and an operating voltage of 1.855 V when the 20 mA current is applied (evaluation) is obtained. I was able to. Furthermore, when an energization test was performed under the same conditions as in Example 1, the relative output was 101.6% and the operating voltage was 1.856 V (an increase of about 0.1%).

  As described above, by adopting the structure of the LED element described in the third embodiment, that is, the light reflecting layer 12 made of a semiconductor multilayer film is provided between the n-type buffer layer 2 and the n-type cladding layer 3. By providing, in addition to the effect intended by the present invention shown in Example 1, the LED element having higher output than the LED element shown in Example 1 could be obtained. This is because the effective light extraction efficiency is improved by the light reflection layer 12.

  As Example 4, a red LED epitaxial wafer having a structure shown in FIG. The epitaxial growth method, the epitaxial layer thickness, the epitaxial layer structure, and the LED element manufacturing method were basically the same as those in Example 1.

However, the difference is that an MQW active layer 13 having a multiple quantum well (MQW) structure as the structure of the active layer 4 is used. In the multiple quantum well, the barrier layer was made of (Al _0.5 Ga _0.5 ) _0.5 In _0.5 P, and the film thickness was about 7.5 nm. The well layer as the light emitting layer was made of Ga _0.5 In _0.5 P, and the film thickness was set to about 5.0 nm. And the combination of those barrier layers and well layers was made into one pair, and 40.5 pairs were formed in total.

  As a modification of Example 4, the strained multiple quantum well structure in which the composition ratio of Ga and In constituting the well layer is changed and stress in the compression or tensile direction is applied to the starting substrate GaAs is an active layer. The LED element used for the above was also manufactured.

  Incidentally, the well layer of the LED element employing the strained multiple quantum well structure manufactured in Example 4 has a configuration in which the Ga composition is decreased and the difference in In composition is increased, and the lattice constant of the well layer is increased. However, unlike the n-type cladding layer 3 which is the base of the active layer, for example, it is characterized by receiving a compressive strain due to the degree of lattice mismatch.

  As a result of evaluating the initial characteristics of the two types of LED elements produced in this way (Example 4 and modified example of Example 4), the LED characteristics at the time of 20 mA energization (evaluation) were 1.14 mW in terms of light emission output, respectively. The operating voltage was 1.844 V, the light emission output was 1.23 mW, and the operating voltage was 1.843 V, and it was possible to obtain an LED element having excellent initial characteristics.

  Furthermore, when an energization test was performed under the same conditions as in Example 1, the relative outputs of the two types of LED elements were 102.5% and 102.3%, respectively.

  As described above, the structure used for the active layer 4 of Example 1 is a multiple quantum well structure or a strained quantum well structure. The light output is increasing. In other words, by adopting the above-described quantum well structure, the internal quantum efficiency of the LED element is improved, and the characteristics of the LED element are improved as a whole. Further, the policy intended by the present invention shown in the first embodiment is used. It is shown that it is fully applicable even when is taken.

  In the above-described embodiment of the present invention, only a red LED element having an emission wavelength of 630 nm is used as an example of manufacture. However, other LED elements manufactured using the same AlGaInP-based material, for example, an emission wavelength of 560 nm to 660 nm. This LED element has a wavelength band different from that of the present embodiment by appropriately setting the measure intended by the present invention, that is, the C (carbon) concentration in the buffer layer and the film thickness of the layer. However, the effect of the present invention can be obtained.

  Moreover, in the said Example, it was set as the LED element structure which provided the buffer layer 2 between the GaAs substrate 1 and the n-type clad layer 3. FIG. However, even if the structure in which the n-type cladding layer 3 is directly laminated on the GaAs substrate 1 is adopted, the intended effect of the present invention can be obtained.

  Moreover, in the said Example, although the shape which made the shape of the surface electrode 9 formed on the outermost surface of an LED element always circular was taken, other shapes, for example, a square, a rhombus, a polygon, etc., were taken. Furthermore, it is possible to adopt a form having electrodes that are branched into a wing shape or a branch shape, and even the electrodes of those forms can achieve the intended effect of the present invention.

  In the above embodiment, only an example in which n-type GaAs is used for the semiconductor substrate has been described. However, for example, an epitaxial wafer for LED having Ge (germanium) as a starting substrate, or a starting substrate with GaAs or The effect intended by the present invention can be obtained even in an epitaxial wafer for LED in which Ge is removed later and Si (silicon) or a metal having a thermal conductivity equal to or higher than Si is used as an alternative free-standing substrate for a permanent substrate. Can do.

  In the above embodiment, AlGaAs having an Al composition in the range of about 0.8 to 0.9 is used as the material of the buffer layer 6. However, the present invention is not particularly limited to this range, and if the measure intended by the present application, that is, the C (carbon) concentration in the buffer layer and the film thickness of the layer are appropriately set, for example, Al in the AlGaAs layer is used. There is no problem even if the composition is in a range other than the above, and the V / III ratio is a value other than the values shown in the above examples.

  For example, in the case of an LED element having an emission wavelength of about 650 nm, the AlGaAs layer used for the buffer layer hardly absorbs light emitted from the active layer even if the Al composition is in the range of about 0.6 to 0.7. And a high-power LED element can be obtained. At the same time, since the C (carbon) concentration in the buffer layer is directly related to lowering the Al composition, the V / III ratio to be set is the Al composition of 0.8 to 0 in the above example. Compared with the case where the buffer layer 6 of .9 is used, it can be set low.

  Conversely, in the case of an LED element having an emission wavelength of, for example, about 570 nm, the Al composition of the buffer layer is set to around 0.9 in order to eliminate the effect of light absorption by the buffer layer, and the V / III ratio at that time This time, measures should be taken so that it is set higher than before. In other words, as described above, even if the manufacturing parameters such as the composition of the material constituting the buffer layer, the V / III ratio during epitaxial growth, and the growth temperature change, even the measures described in the present invention are taken. Thus, the intended effect of the present invention can be obtained.

  Moreover, in the said Example, only ITO was used as a material of the electric current dispersion layer 8. FIG. However, instead of the above-described ITO, a current transparent layer 8 having a generally known visible light transmittance, such as ZnO or CTO, and a low electric resistance, that is, a general transparent conductive film may be used. It can be applied as an alternative material. However, more important factors as a material used for the current spreading layer are not only the above two points but also the carrier concentration. The importance of the carrier concentration is as described above, and considering the reduction of the operating voltage of the LED element, the material applicable to the current spreading layer is limited to some extent and is correctly selected from the material group. Should.

[Comparative Example 1]
As Comparative Example 1, an epitaxial wafer for red LED having a structure as shown in FIG. The epitaxial growth method, the epitaxial layer thickness, the epitaxial layer structure, and the LED element manufacturing method were basically the same as those in Example 1. The points different from the first embodiment will be listed below, and detailed explanations will be given accordingly.

In Comparative Example 1, the V / III ratio during growth of the p-type buffer layer 16 was 10, and the concentration of C contained in the p-type buffer layer 16 was 1.8 × 10 ¹⁹ / cm ³ .

  Next, the LED epitaxial wafer produced as described above is made into an element, and the process is the same as in the first embodiment.

  As a result of evaluating the initial characteristics of the LED element fabricated in this way, an LED element having initial characteristics of a light emission output of 0.92 mW and a working voltage of 1.855 V when 20 mA was energized (during evaluation) was obtained. Furthermore, when an energization test was performed under the same conditions as in Example 1, the relative output was 52% and the operating voltage was 1.915 V (an increase of about 3%).

  Moreover, the SIMS analysis of the LED element of the state immediately after LED element preparation and the state after performing an electricity supply test on said conditions after LED element preparation was performed. The SIMS analysis result after the current test at this time is shown in FIG. 11 (note that the sample of the LED element shown in FIG. 11 is mechanically polished to remove several μm from the surface in order to improve the measurement resolution of SIMS analysis. ing).

As a result of the SIMS analysis, the concentration of C (carbon) contained in the buffer layer 16 before and after the energization test is 1.8 × 10 ¹⁹ / cm ³ , and in the LED element of Comparative Example 1 after the energization test, the active layer 4 It was confirmed that Zn as the dopant of the p-type contact layer 7 was diffused and mixed therein. That is, the element life of the LED element shown in Comparative Example 1, that is, the cause of the decrease in reliability is due to this dopant diffusion.

[Comparative Example 2]
As Comparative Example 2, an epitaxial wafer for red LED having a structure as shown in FIG. The epitaxial growth method, the epitaxial layer thickness, the epitaxial layer structure, and the LED device manufacturing method were basically the same as those in Comparative Example 1. The points different from the comparative example 1 are listed below, and detailed explanations will be given accordingly.

  In Comparative Example 2, the p-type buffer layer 16 was not provided. If the film thickness of the p-type cladding layer 5 is about 400 nm, it is sufficient for the carrier confinement effect and the carrier (hole) supply layer. That is, the p-type cladding layer 5 has a film thickness of about 400 nm and sufficiently fulfills the role as a cladding layer. That is, the LED element described in Comparative Example 2 is the same as Comparative Example 1 except that the AlGaAs buffer layer 16 is not provided.

  Next, the LED epitaxial wafer produced as described above is made into an element, and the process is the same as in Comparative Example 1.

  As a result of evaluating the initial characteristics of the LED element thus fabricated, an LED element having initial characteristics of a light emission output of 0.89 mW and an operating voltage of 1.840 V when 20 mA was energized (during evaluation) could be obtained.

  However, at the stage of evaluating the initial characteristics, there were about 20 to 30% of elements already destroyed such as no light emission. For this reason, although the above-mentioned characteristic was obtained with the element which was not destroyed, evaluation was not performed at all with other 20 to 30% elements. This is expected to be element destruction due to wire bonding damage in the wire bonding process before element evaluation. When an energization test under the same conditions as in Comparative Example 1 was performed with an element that was not destroyed, the relative output was 79% and the operating voltage was 1.850 V (an increase of about 0.5%).

  As described above, in the structure in which the buffer layer is not provided, there is a problem in the yield of manufacturing the LED elements, and it is difficult to say that the light emission output and the reliability are the best. That is, only an effect that the relative output is slightly improved as compared with Comparative Example 1 was obtained, and conversely, the yield was lowered (In Comparative Example 1, there was almost no problem with destruction of the LED elements).

1 is a cross-sectional structure diagram of an AlGaInP red LED according to an embodiment of the present invention and Example 1. FIG. It is the figure which showed the relationship between C density | concentration of the buffer layer concerning Example 1 thru | or 4 of this invention, and the diffusion distance of Zn. It is a cross-section figure of AlGaInP type red LED concerning Example 2 of this invention. It is a cross-section figure of AlGaInP type red LED concerning Example 3 of this invention. It is a cross-section figure of AlGaInP type red LED concerning Example 4 of this invention. It shows the film thickness of the contact layer and the transmittance at the emission wavelength of the LED. It is the figure which showed the reflectance spectrum of the ITO film | membrane formed on the GaAs substrate. It is the figure which showed the relationship between the lamination | stacking pair number of a light reflection layer, and a vertical reflectance. 6 is a cross-sectional structure diagram of an AlGaInP red LED according to Comparative Example 1. FIG. 6 is a cross-sectional structure diagram of an AlGaInP red LED according to Comparative Example 2. FIG. It is the figure which showed the SIMS analysis result in the comparative example 1. FIG. 3 is a diagram showing SIMS analysis results in Example 1. It is the figure which showed V / III ratio at the time of buffer layer growth, and C density | concentration in a buffer layer.

Explanation of symbols

1 n-type GaAs substrate 2 n-type GaAs buffer layer 3 n-type AlGaInP cladding layer (n-type cladding layer)
4 Undoped AlGaInP active layer (active layer)
5 p-type AlGaInP cladding layer (p-type cladding layer)
6 p-type buffer layer 7 p-type AlGaAs contact layer (p-type contact layer)
8 ITO film (current dispersion layer)
9 Front electrode 10 Back electrode 11 Diffusion prevention layer 12 Light reflection layer 13 MQW active layer 16 p-type buffer layer

Claims (9)

A layer that is epitaxially grown on a substrate by a vapor deposition method, and a light-emitting portion including at least an n-type cladding layer, an active layer, and a p-type cladding layer is formed, and a p-type dopant concentration is 1 × above the light-emitting portion. A p-type contact layer is formed in which the main component of the group V element of the thin film is 10 ¹⁹ / cm ³ or more and is formed by adding a dopant of a material different from the dopant added to the p-type cladding layer. In the semiconductor light emitting device in which a current spreading layer made of a metal oxide material is formed on the p-type contact layer,
Between the p-type cladding layer and the p-type contact layer, it is composed of a III / V group semiconductor which is p-type conductive and contains H (hydrogen) and C (carbon) inevitably mixed during growth. Having a buffered layer,
The thickness of the p-type contact layer is in the range of 1 nm to 30 nm, and the dopant of the p-type contact layer is Zn;
In the p-type cladding layer, the main component of the group V element is P (phosphorus),
The dopant of the buffer layer is Mg, the material constituting the buffer layer is Al _x Ga _1-x As (where 0.4 ≦ X ≦ 1), and the buffer layer has an Al composition ratio of p-type. Higher than the cladding layer,
^Further , the relationship between L = 6.872 × 10 ⁻¹⁴ × N ^0.733 is defined by setting the C concentration of H (hydrogen) and C (carbon) at least by setting the V / III ratio during growth. When the diffusion length is determined from the formula [where N is the C concentration (unit: cm ⁻³ ) and L is the diffusion length (unit: μm) of the dopant added to the p-type contact layer], the buffer The semiconductor light emitting element , wherein the thickness of the layer is not less than the diffusion length L. The semiconductor light emitting device according to claim 1,
A semiconductor light emitting device, wherein the dopant of the p-type cladding layer is Mg . The semiconductor light emitting device according to claim 1 or 2,
A semiconductor light-emitting element, wherein a material constituting the p-type contact layer is Al _x Ga _1-x As (where 0 ≦ X ≦ 0.4). In the semiconductor light-emitting device according to any one of claims 1 to 3 ,
The current spreading layer is made of ITO (tin-added indium oxide), SnO ₂ (tin oxide), ATO (antimony-added tin oxide), In ₂ O ₃ (indium oxide), ZnO (zinc oxide), GZO (gallium-added zinc oxide). , BZO (boron-added zinc oxide), AZO (aluminum-added zinc oxide), CdO (cadmium oxide), CTO (tin-added cadmium oxide), IZO (indium-added zinc oxide), at least one metal oxide material. A semiconductor light emitting element formed by using one or more kinds. The semiconductor light emitting device according to any one of claims 1-4,
The material constituting the light emitting portion is (Al _x Ga _1-x ) _y In _1-y P (where 0 ≦ X ≦ 1, 0.4 ≦ Y ≦ 0.6), and the p-type cladding layer And the content rate of Al (aluminum) contained in the said n-type clad layer is higher than the content rate of Al contained in the said active layer, The semiconductor light-emitting device characterized by the above-mentioned. The semiconductor light emitting device according to any one of claims 1 to 5
A semiconductor light-emitting element, wherein the current distribution layer has a carrier concentration of 7 × 10 ²⁰ / cm ³ or more. In the semiconductor light-emitting device according to any one of claims 2 to 6 ,
The Mg concentration contained in the p-type cladding layer is in the range of 1 × 10 ¹⁷ / cm ³ or more and 5 × 10 ¹⁸ / cm ³ or less, and the p-type cladding layer set in the above range includes all of the layers Alternatively, a semiconductor light-emitting element occupying at least a part. In the semiconductor light-emitting device according to any one of claims 1 to 7 ,
A semiconductor layer to which no dopant is added between the active layer and the p-type cladding layer, a semiconductor layer having a lower p-type dopant concentration than the p-type cladding layer, or an n-type dopant and p A diffusion preventing layer made of a semiconductor layer in which any of the semiconductor layers in a pseudo-neutral state to which a dopant of a type is added at the same time, or a combination thereof, is provided in a range of 300 nm or less. A semiconductor light emitting device. In the semiconductor light-emitting device according to any one of claims 1 to 7 ,
A semiconductor layer to which no dopant is added between the active layer and the n-type cladding layer, a semiconductor layer having a lower n-type dopant concentration than the n-type cladding layer, or an n-type dopant and p A diffusion preventing layer made of a semiconductor layer in which any of the semiconductor layers in a pseudo-neutral state to which a dopant of a type is added at the same time, or a combination thereof, is provided in a range of 200 nm or less. A semiconductor light emitting device.

Images