LED and LED manufacturing method

Abstract

Provided is an LED having a quantum well structure and comprising: a quantum well layer having a composition AlaGa1-aN (aluminum gallium nitride) (0 < a < 1); and a high-level layer that has a composition AlbGa1-bN (a < b ≤ 1), is adjacent to the quantum well layer, and has an energy level higher than that of the quantum well layer. ΔX, which is the difference between a and b, and the width D [nm] of the quantum well layer satisfy [relational expression 1] D = k / ΔX (where, in the relational expression 1, D > 3, 0 < ΔX < 0.15, and 0.45 ≤ k ≤ 1.25 (unit of the coefficient k is [nm])).

Description

LED and LED manufacturing method

[0001] The present invention relates to LEDs and methods for manufacturing LEDs.

[0002] In Patent Document 1, it is stated that "the quantum well structure 35 is a stack structure in which a plurality of well layers having a thickness of about 5 nm to about 50 nm are combined with a barrier layer having a band gap energy larger than that of the well layers" (paragraph 0034), and "the fabricated LED has a homoepitaxial AlN layer (about 200 nm thick) and an n-AlN layer for the n-type electrical contact layer 30. ０．７５ Ga ０．２５ N layer and an AlGaN / AlGaN multi-quantum well active region 35 (four i-Al ０．６ Ga ０．４ N layer) and p-Al ０．９０ Ga ０．１０ "The semiconductor device includes a layered film of an N-electron blocking layer (not shown), a p-type AlGaN cladding layer 42 (about 10 nm thick), and a p-type GaN electrical contact layer 45 (about 200 nm thick)." (Paragraph 0048) Non-Patent Document 1 states, "We obtained efficient PL emission of 234 and 245 nm from AlN / Al ０．１８ Ga ０．８２ N and Al ０．８ Ga ０．２ N / Al ０．１８ Ga ０．８２ N MQWs, respectively, at 77 K. The optimum value of well thickness was approximately 1.5 nm. (Machine translation: AlN / Al ０．１８ Ga ０．８２ N MQW and Al ０．８ Ga ０．２ N / Al ０．１８ Ga ０．８２Efficient PL emissions of 234 nm and 245 nm were obtained from the N MQWs at 77 K. The optimal thickness of the wells was approximately 1.5 nm. [Prior art documents] [Patent documents] [Patent documents 1] Japanese translation of PCT publication No. 2016-511938 [Non-patent documents 1] "[Optical Properties of AlGaN Quantum Well Structures]", Hideki Hirayama, Yasushi Enomoto, Atsuhiro Kinoshita, Akira Hirata&Yoshinobu Aoyagi, MRS Internet Journal of Nitride Semiconductor Research volume 5, pages 696-702 (2000)” General disclosure

[0003] In a first aspect of the present invention, there is provided an LED having a quantum well structure. ａ Ga １－ａ A quantum well layer having a composition of N (aluminum gallium nitride) (0<a<1), and Al ｂ Ga １－ｂ The quantum well layer has a composition of N (a<b≦1), is adjacent to the quantum well layer, and has a higher energy level than the quantum well layer, and ΔX, which is the difference between a and b, and a width D [nm] of the quantum well layer, satisfy the following relational expression 1: [Relational expression 1] D=k / ΔX (where, in relational expression 1, D>3, 0<ΔX<0.15, and 0.45≦k≦1.25 (the unit of coefficient k is [nm])).

[0004] In the above LED, D may be the smallest value within the range of D such that the energy difference between the valence band potential or the conduction band potential of the polarization electric field present in the quantum well layer and the quasi-Fermi level of the carriers present in the quantum well layer becomes zero.

[0005] In any of the above LEDs, D may be the smallest value within the range of D in which the energy difference between the valence band potential and the quasi-Fermi level of holes present in the quantum well layer becomes zero.

[0006] In a second aspect of the present invention, there is provided an LED having a quantum well structure. １－ｃ Ga ｃ N (indium gallium nitride) (0<c<1) quantum well layer, In １－ｄ Ga ｄ The quantum well layer has a composition of N (c<d≦1), is adjacent to the quantum well layer, and has a higher energy level than the quantum well layer, and ΔX, which is a difference between c and d, and a width D [nm] of the quantum well layer, satisfy the following relational expression 2: [Relational expression 2] D=k / ΔX (where, in relational expression 2, D≧2, 0<ΔX≦0.15, and 0.3≦k≦0.6 (the unit of coefficient k is [nm])).

[0007] In the above LED, D may be the smallest value within the range of D such that the energy difference between the valence band potential or the conduction band potential of the polarization electric field present in the quantum well layer and the quasi-Fermi level of the carriers present in the quantum well layer becomes zero.

[0008] In any of the above LEDs, D may be the smallest value within the range of D in which the energy difference between the valence band potential and the quasi-Fermi level of holes present in the quantum well layer becomes zero.

[0009] In a third aspect of the present invention, there is provided a method for manufacturing an LED having a quantum well structure. ａ Ga １－ａ forming a quantum well layer having a composition of N (aluminum gallium nitride) (0<a<1); ｂ Ga １－ｂ and forming a high-level layer adjacent to the quantum well layer and having an energy level higher than that of the quantum well layer, the high-level layer having a composition of a<b<1), wherein ΔX, which is the difference between a and b, and a width D [nm] of the quantum well layer satisfy the following relational expression 1: [Relational expression 1] D=k / ΔX (where, in Relational expression 1, D>3, 0<ΔX<0.15, and 0.45≦k≦1.25 (the unit of coefficient k is [nm])).

[0010] In a fourth aspect of the present invention, there is provided a method for manufacturing an LED having a quantum well structure. １－ｃ Ga ｃ forming a quantum well layer having a composition of InN (indium gallium nitride) (0<c<1); １－ｄ Ga ｄ and forming a high-level layer adjacent to the quantum well layer and having an energy level higher than that of the quantum well layer, wherein ΔX, which is a difference between c and d, and a width D [nm] of the quantum well layer satisfy the following relational expression 2: [Relational expression 2] D=k / ΔX (where, in Relational expression 2, D≧2, 0<ΔX≦0.15, and 0.3≦k≦0.6 (the unit of coefficient k is [nm])).

[0011] The above summary of the invention does not list all of the features of the present invention, and subcombinations of these features may also be inventions.

[0012] 1 is a schematic diagram showing an example of the structure of the LED 100 according to the first embodiment; FIG. 2 is a diagram showing an example of the Z-axis position and Al composition X of each layer of the LED 100 according to the first embodiment; FIG. 3 is a table showing a first simulation example of the Al composition X, film thickness, etc. of each layer of the LED 100 according to the first embodiment; FIG. 4 is a graph showing an example of the relationship between the Al composition of the high-level layer of the LED 100 according to the first embodiment and the emission wavelength; FIG. 5 is a graph showing a simulation result of the relationship between the width (D) of the quantum well layer 110 and the luminous efficiency (IQE) based on the first simulation example of FIG. 3; FIG. 6 is a graph showing the relationship between the Al composition difference (ΔX) between the high-level layer and the quantum well layer 110 and the width (D) of the quantum well layer 110 at which the luminous efficiency is maximized for each Al composition difference (ΔX); FIG. 7 is a table showing a second simulation example of the Al composition X, film thickness, etc. of each layer of the LED 100 according to the first embodiment; FIG. 8 is a graph showing an example of the relationship between the Al composition X (a) of the quantum well layer 110 of the LED 100 according to the first embodiment and the emission wavelength. 10 is a graph showing a simulation result of the relationship between the width (D) of the quantum well layer 110 and the luminous efficiency (IQE) in the second simulation example of FIG. 7 . 11 is a graph showing the relationship between the Al composition difference (ΔX) in FIG. 9 and the width (D) of the quantum well layer 110 at which the luminous efficiency is maximized for each Al composition difference (ΔX). 12 is a diagram showing an example of an energy band graph and graphs of the wave functions of electrons and holes in the LED 100 according to the first embodiment when the width D of the quantum well layer 110 is 3 nm. 13 is a diagram showing an example of an energy band graph and graphs of the wave functions of electrons and holes in the LED 100 according to the first embodiment when the width D of the quantum well layer 110 is 30 nm. 14 is a diagram showing an example of a conduction band potential graph, a graph of the electron wave function, and a graph of the electron quasi-Fermi level in the LED 100 according to the first embodiment. 15 is a diagram showing an example of a valence band potential graph, a graph of the hole wave function, and a graph of the hole quasi-Fermi level in the LED 100 according to the first embodiment. 1 is a table showing an example of the magnitude of the polarization electric field of the quantum well layer 110 in the LED 100 according to the first embodiment. FIG. 2 is a graph showing an example of the relationship between the width D of the quantum well layer 110 and the valence band potential in the LED 100 according to the first embodiment.19 is a graph showing an example of the energy difference G between the energy level of a hole and the quasi-Fermi level of a hole in an LED according to a comparative example, when the Al composition difference ΔX between the high-level layer and the quantum well layer is 0.10 and the width D of the quantum well layer is 3 nm. 20 is a graph showing an example of the energy difference G between the energy level of a hole and the quasi-Fermi level of a hole in an LED 100 according to the first embodiment, when the Al composition difference ΔX between the high-level layer and the quantum well layer 110 is 0.10 and the width D of the quantum well layer 110 is 5 nm. 21 is a table showing an example of a simulation of the Al composition and film thickness when the Al composition of each layer is uniformly changed by +α in the LED 100 according to the first embodiment. 22 is a graph showing the simulation results of the transition of the graph of the quasi-Fermi level of a hole and the graph of the valence band potential when the Al composition of each layer is uniformly changed by +α, according to the simulation example of FIG. 20. 23 is a graph showing the relationship between each α of FIG. 20 and the energy difference ΔE. 26 is a graph showing a simulation example of the relationship between the Al composition difference (ΔX) between the high-level layer and the quantum well layer 110 and the width (D) of the quantum well layer 110, which changes depending on the influence of the interface between the high-level layer and the quantum well layer 110 in the LED 100 according to the first embodiment.

[0041] FIG. 27 is a schematic diagram showing an example of the structure of an LED 200 according to a second embodiment.

[0042] FIG. 28 is a diagram showing an example of the Z-axis position and Ga composition X of each layer of the LED 200 according to the second embodiment.

[0043] FIG. 29 is a table showing a simulation example of the Ga composition X, film thickness, etc. of each layer of the LED 200 according to the second embodiment.

[0044] FIG. 29 is a graph showing a simulation result of the relationship between the width (D) of the quantum well layer 210 and the luminous efficiency (IQE) in the simulation example of FIG. 26.

[0045] FIG. 29 is a graph showing the relationship between the Ga composition difference (ΔX) in FIG. 26 and the width (D) of the quantum well layer 210 at which the luminous efficiency is maximized for each Ga composition difference (ΔX).

[0046] FIG. 30 is a table showing an example of the magnitude of the polarization electric field of the quantum well layer 210 in the LED 200 according to the second embodiment. 10 is a graph showing a simulation example of the relationship between the Al composition difference (ΔX) between the high-level layer and the quantum well layer 210 and the width (D) of the quantum well layer 210 when the strain state of the quantum well layer 210 is set to coherent strain in the LED 200 according to the second embodiment.32 is a graph showing a simulation example of the relationship between the Al composition difference (ΔX) between the high-level layer and the quantum well layer 110 and the width (D) of the quantum well layer 110 when the quantum well layer 110 of the LED 100 according to the first embodiment is a three-layer multiple quantum well layer (MQW layer) in which the strain state is coherent.

[0041] FIG. 33 is a diagram showing another example of the Z-axis position and Al composition X of each layer of the LED 100 according to the first embodiment.

[0042] FIG. 34 is a table showing an experimental example of the Al composition X, film thickness, etc. of each layer of the LED 100 according to the first embodiment.

[0043] FIG. 35 is a graph showing experimental results of the relationship between the width (D) of the quantum well layer 110 and the emission wavelength and intensity, based on the experimental example of FIG. 32.

[0013] The present invention will be described below through embodiments of the invention, but the following embodiments do not limit the scope of the invention as claimed. Furthermore, not all of the combinations of features described in the embodiments are necessarily essential to the solution of the invention.

[0014] 1 is a schematic diagram showing an example of the structure of an LED 100 according to a first embodiment. The LED described in this application has a quantum well structure that enhances internal quantum efficiency (IQE, luminous efficiency), and is formed by stacking multiple layers on a substrate. More specifically, the LED 100 according to the first embodiment includes a quantum well layer 110, an N-type contact layer 120, an electron blocking layer 130, a P-type cap layer 140, a P-type contact layer 150, an N-layer electrode 160, a P-layer electrode 170, and a substrate 180.

[0015] The LED 100 may also be referred to as an ultraviolet LED, an ultraviolet light-emitting diode, a deep ultraviolet LED, a deep ultraviolet light-emitting diode, or the like. The LED 100 is, for example, a back-emitting LED that emits light from the back surface of the substrate 180 at the bottom. As shown in FIG. 1 , the stacking direction of the LED 100 is defined as the Z-axis direction, and the position of the surface of the substrate 180, i.e., the position of one end face of the N-type contact layer 120 facing the substrate 180, is defined as Z = 0 [nm], and the direction from the one end face of the N-type contact layer 120 toward the other end face is defined as the positive Z-axis direction. In the following description, the length of any layer included in the LED 100 in the Z-axis direction may be referred to as the width or film thickness. This also applies to the following embodiments, and redundant description will be omitted.

[0016] The quantum well layer 110 may also be referred to as a light-emitting layer, a QW (Quantum Well) layer, an MQW (Multiple QW) layer, an active layer, or the like, and is a layer in which electrons and holes recombine to generate light. The quantum well layer 110 recombines electrons and holes by spatially confining them. As an example, the quantum well layer 110 is made of Al. ａ Ga １－ａ The quantum well layer 110 has a composition of aluminum gallium nitride (N) (0<a<1). The quantum well layer 110 is, for example, a single quantum well layer (QW layer). In the following description, subscripts indicating the content, such as "a" and "b," may be omitted, and expressions such as "AlGaN," "InGaN," and "AlInGaN" may be used as abbreviations or general terms.

[0017] The N-type contact layer 120 is a layer adjacent to the negative side of the quantum well layer 110 on the Z axis. The N-type contact layer 120 may also be referred to as an N-type semiconductor layer or an N layer, and is a layer that supplies electrons to the quantum well layer 110. As an example, the N-type contact layer 120 is an n-Al doped with at least one kind of impurity, for example, Si. ｂ１ Ga １－ｂ１ N (a<b1≦1). The N-type contact layer 120 has a higher energy level than the quantum well layer 110. The N-type contact layer 120 is an example of a high-level layer, which will be described later. The LED 100 may additionally include a barrier layer between the N-type contact layer 120 and the quantum well layer 110, the barrier layer having a higher energy level than the N-type contact layer 120. In this case, the barrier layer is an example of a high-level layer. The barrier layer and the quantum well layer 110 may be collectively referred to as a light-emitting layer.

[0018] The electron blocking layer 130 is a layer adjacent to the quantum well layer 110 on the positive side of the Z axis. The electron blocking layer 130 may also be called an EBL (Electron Blocking Layer) or an electron blocking layer, and is a layer that prevents electrons from leaking (overflowing) from the quantum well layer 110. The electron blocking layer 130 is, for example, made of Al ｂ２ Ga １－ｂ２The electron blocking layer 130 has a composition of N (a<b2≦1). For example, the electron blocking layer 130 may be doped with one or more impurities such as Mg. The electron blocking layer 130 has a higher energy level than the quantum well layer 110. The electron blocking layer 130 is an example of a high-level layer.

[0019] The P-type cap layer 140 is a layer adjacent to the electron blocking layer 130 on the positive side of the Z axis. The P-type cap layer 140 may also be referred to as a P-type semiconductor layer or a P layer, and is a layer that supplies holes to the quantum well layer 110. As an example, the P-type cap layer 140 is made of p-Al doped with at least one kind of impurity, for example, Mg. ｂ３ Ga １－ｂ３ N (a<b3≦1). Note that LED 100 may not include electron blocking layer 130. In this case, P-type cap layer 140 is an example of a high-level layer that is adjacent to quantum well layer 110 and has a higher energy level than quantum well layer 110.

[0020] The P-type contact layer 150 is a layer adjacent to the positive side of the P-type cap layer 140 along the Z axis. The P-type contact layer 150 may also be referred to as a P-type semiconductor layer or a P layer, and, like the P-type cap layer 140, is a layer that supplies holes to the quantum well layer 110. As an example, the P-type contact layer 150 has a composition of p-GaN doped with at least one type of impurity, for example, Mg.

[0021] The N-layer electrode 160 is provided on the positive side of the N-type contact layer 120 along the Z axis and is electrically connected to the N-type contact layer 120. The N-layer electrode 160 is formed of an ohmic metal material, such as Ni or Au. The P-layer electrode 170 is provided on the positive side of the P-type contact layer 150 along the Z axis and is electrically connected to the P-type contact layer 150. Like the N-layer electrode 160, the P-layer electrode 170 is formed of an ohmic metal material, such as Ni or Au. Note that the LED 100 does not necessarily have to include the P-type contact layer 150. In this case, the P-layer electrode 170 is provided on the positive side of the P-type cap layer 140 along the Z axis and is electrically connected to the P-type cap layer 140. The substrate 180 is, for example, a sapphire substrate.

[0022] The LED 100 according to the first embodiment has an object to set the width D of at least the quantum well layer 110 so as to maximize the IQE of the LED 100, when at least one of the emission wavelength of the LED 100, the Al composition of the quantum well layer 110 and the high-level layer, and the difference in Al composition therebetween are predetermined, while the width D of at least the quantum well layer 110 can be freely set.

[0023] 2 is a diagram showing an example of the Z-axis position and Al composition X of each layer of the LED 100 according to the first embodiment. FIG. 3 is a table showing a first simulation example of the Al composition X, film thickness, etc. of each layer of the LED 100 according to the first embodiment.

[0024] The horizontal axis in Figure 2 indicates the position Z [nm] in the Z-axis direction, and the vertical axis indicates the Al composition X, which is defined in the range of 0 to 1 (0% to 100%). In this embodiment, "X" collectively refers to the above-mentioned "a," "b1," "b2," and "b3," as well as the "b" described below. Figure 2 shows the Al composition X in each layer of the LED 100 as a continuous thick line graph. Figure 2 also shows the layers present at each position Z in a bar graph above the graph.

[0025] The first line of the table in FIG. 3 lists, from the left, X (Al composition), film thickness [nm], and doping concentration [cm －３ 3 show the Al composition X, film thickness, etc. of the P-type contact layer 150 (p-GaN), the P-type cap layer 140 (p-AlGaN), the electron block layer 130 (EBL), the quantum well layer 110 (QW), and the N-type contact layer 120 (n-AlGaN), in that order.

[0026] In the first simulation example shown in FIG. 3, the Al composition X, film thickness, and doping concentration of the P-type contact layer 150 are set to 0, 100 [nm], 1×10 １９ [cm －３ The Al composition X(b3), film thickness, and doping concentration of the P-type cap layer 140 are set to 0.70, 70 [nm], 1×10 １９ [cm －３ The Al composition X(b2) and film thickness of the electron blocking layer 130 are set to 0.80 and 9 nm, respectively.

[0027] 3, the Al composition X(b1) of the N-type contact layer 120 is changed from 0.45 to 0.70 in increments of 0.025 or 0.05. The thickness and doping concentration of the N-type contact layer 120 are 350 nm and 1×10 １９ [cm －３ ]. In the first simulation example, the film thickness, i.e., width, of the quantum well layer 110 is changed in the range of 1-30 nm for each Al composition X(b1) of the N-type contact layer 120. The Al composition X(a) of the quantum well layer 110 is set to 0.40. When the Al composition X(a) of the quantum well layer 110 is 0.40, the emission wavelength of the LED 100 is approximately 280-300 nm. It is known that if the Al composition X(b1) of the N-type contact layer 120 is too high, for example, higher than 0.9, the electrical resistance of the N-type contact layer 120 increases, making it difficult for electrons to flow. On the other hand, if the Al composition X(b1) of the N-type contact layer 120 is too low, for example, set to the same as the Al composition X(a) of the quantum well layer 110, the electron confinement in the quantum well layer 110 becomes shallow, resulting in a decrease in light emission efficiency.

[0028] On the graph of FIG. 2 , the manner in which the Al composition X (b1) of the N-type contact layer 120 is varied is indicated by a filled-in arrow, corresponding to the first simulation example of FIG. 3 . The width of the quantum well layer 110 is also indicated by D, and the manner in which the width D is varied, i.e., the position Z of the surface of the quantum well layer 110 is varied, corresponding to the first simulation example of FIG. 3 . The graph also shows, with a filled-in arrow, the manner in which the position Z of the surface of the electron blocking layer 130 and the P-type cap layer 140 varies with the variation of the width D. The graph also shows, with a filled-in arrow, the difference in Al composition between the N-type contact layer 120, which is an example of a high-level layer, and the quantum well layer 110, as ΔX (=b1−a). The definitions above regarding FIG. 2 apply to subsequent figures similar to FIG. 2 , and redundant explanations will be omitted.

[0029] 4 is a graph showing an example of the relationship between the Al composition of the high-level layer and the emission wavelength of the LED 100 according to the first embodiment. The horizontal axis of Fig. 4 indicates the Al composition X of the high-level layer, and the vertical axis indicates the emission wavelength [nm] of the LED 100.

[0030] As mentioned above, the high level layer is Al ｂ Ga １－ｂ The high-level layer has a composition of N (a<b≦1), is adjacent to the quantum well layer 110, and has a higher energy level than the quantum well layer 110. In the configuration of the LED 100 shown in FIG. 1 , the high-level layer may refer to both the N-type contact layer 120 and the electron blocking layer 130, or to either one of them. Furthermore, in the first embodiment, the expression "Al composition b of the high-level layer" may refer to the average of the Al composition b1 of the N-type contact layer 120 and the Al composition b2 of the electron blocking layer 130, or to either one of them. ΔX refers to the Al composition difference (b−a) between the Al composition b of the high-level layer and the Al composition a of the quantum well layer 110. In the following descriptions of several embodiments, only the N-type contact layer 120 will be referred to as the high-level layer simply for clarity.

[0031] As shown in FIG. 4, it can be seen that even if the Al composition X(b1) of the N-type contact layer 120, which is a high-level layer, is changed, the emission wavelength of the LED 100 hardly changes.

[0032] Fig. 5 is a graph showing the simulation results of the relationship between the width (D) of the quantum well layer 110 and the luminous efficiency (IQE) in the first simulation example of Fig. 3. The horizontal axis of Fig. 5 indicates the width D [nm] of the quantum well layer 110, and the vertical axis indicates the luminous efficiency (IQE) [%] of the LED 100.

[0033] In this simulation, the LED 100 is irradiated with 10 [A / cm ２ ] is passed through the LED 100. In this simulation, the change in IQE of the LED 100 with the change in the width D of the quantum well layer 110 was confirmed when the Al composition X(b1) of the N-type contact layer 120 was set to 0.45, 0.50, 0.60, and 0.70. FIG. 5 shows a graph of the change when the Al composition X(b1) of the N-type contact layer 120 was set to 0.45, 0.50, 0.60, and 0.70. Calculated values ​​of the width D and IQE are plotted on each graph. As described in FIG. 3, the width D was changed in the range of 1-30 nm.

[0034] As shown in FIG. 5, it can be seen that there exists a width D of the quantum well layer 110 that gives the LED 100 a maximum value of IQE, regardless of the Al composition X(b1) of the N-type contact layer 120.

[0035] 6 is a graph showing the relationship between the Al composition difference (ΔX) between the high-level layer and the quantum well layer 110 and the width (D) of the quantum well layer 110 at which the luminous efficiency is maximized for each Al composition difference (ΔX). The horizontal axis of Fig. 6 indicates the Al composition difference ΔX (= b1 - a) between the high-level layer, i.e., the N-type contact layer 120, and the quantum well layer 110, and the vertical axis indicates the width D [nm] of the quantum well layer 110.

[0036] 6 , it can be seen that the smaller the Al composition difference ΔX (= b−a) between the Al composition a in the quantum well layer 110 and the Al composition b in the high-level layer, the larger the width D of the quantum well layer 110 that provides the maximum IQE value for the LED 100. In other words, it can be seen that there is a negative correlation between ΔX (= b−a), which is the difference between the Al composition a in the quantum well layer 110 and the Al composition b in the high-level layer, and the width D [nm] of the quantum well layer 110 that provides the maximum IQE value. Specifically, the smaller ΔX is, the larger the D that provides the maximum IQE.

[0037] 5 and 6 , it can be seen that when the Al composition difference ΔX between the Al composition of the quantum well layer 110 and the Al composition of the high-level layer is small, i.e., when the confinement of electrons and the like in the quantum well layer 110 is shallow, the rate at which electrons and the like that are desired to emit light in the quantum well layer 110 leak out increases, and the maximum IQE value of the LED 100 is lower than when the Al composition difference ΔX is large. However, according to the LED 100 of this embodiment, even under any constraints other than the width D of the quantum well layer 110, for example, even if the emission wavelength of the LED 100 is predetermined, i.e., the Al composition of the quantum well layer 110 is predetermined and the Al composition difference ΔX is also predetermined, the width D can be set to maximize the IQE. Furthermore, according to the LED 100, if it is possible to increase the Al composition of the high-level layer, i.e., if the Al composition difference ΔX can be increased, the IQE can be further increased by increasing the Al composition difference ΔX and setting the width D to maximize the IQE.

[0038] 7 is a table showing a second simulation example of the Al composition X, film thickness, etc. of each layer of the LED 100 according to the first embodiment. As in the table of FIG. 3, the first row of the table of FIG. 7 lists, from the left, X (Al composition), film thickness [nm], doping concentration [cm －３ 3, the second to fifth rows of the table in FIG. 7 indicate, in order, X, film thickness, etc. of the P-type contact layer 150 (p-GaN), the P-type cap layer 140 (p-AlGaN), the electron blocking layer 130 (EBL), the quantum well layer 110 (QW), and the N-type contact layer 120 (n-AlGaN).

[0039] In the second simulation example shown in FIG. 7, the Al composition X, film thickness, and doping concentration of the P-type contact layer 150 are set to 0, 100 [nm], 1×10 １９ [cm －３ The Al composition X(b3), film thickness, and doping concentration of the P-type cap layer 140 are set to 0.95, 70 [nm], 1×10 １９ [cm －３The Al composition X(b2) and thickness of the electron blocking layer 130 are set to 1 and 9 [nm]. The Al composition X(b1), thickness and doping concentration of the N-type contact layer 120 are set to 0.88, 350 [nm] and 1×10 １９ [cm －３ ]

[0040] 7, the Al composition X(a) of the quantum well layer 110 is varied from 0.70 to 0.86, more specifically, from 0.74 to 0.86 in increments of 0.04. Also in the second simulation example, for each Al composition X(a) of the quantum well layer 110, the film thickness, i.e., width, of the quantum well layer 110 is varied within a range of 1-30 nm. When the Al composition X(a) of the quantum well layer 110 is 0.70-0.86, the emission wavelength of the LED 100 is approximately 220-240 nm.

[0041] 8 is a graph showing an example of the relationship between the Al composition X(a) of the quantum well layer 110 of the LED 100 according to the first embodiment and the emission wavelength. The horizontal axis of FIG. 8 indicates the Al composition X(a) of the quantum well layer 110, and the vertical axis indicates the emission wavelength [nm] of the LED 100. As shown in FIG. 8, it can be seen that the emission wavelength of the LED 100 changes when the Al composition X(a) of the quantum well layer 110 is varied.

[0042] Fig. 9 is a graph showing the simulation results of the relationship between the width (D) of the quantum well layer 110 and the luminous efficiency (IQE) in the second simulation example of Fig. 7. The horizontal axis of Fig. 9 indicates the width D [nm] of the quantum well layer 110, and the vertical axis indicates the luminous efficiency (IQE) [%] of the LED 100.

[0043] In this simulation, as in the first simulation, the LED 100 is supplied with 10 [A / cm ２ ] is passed through the LED 100. Figure 9 shows graphs of the IQE of the LED 100 as the width D of the quantum well layer 110 changes when the Al composition X(a) of the quantum well layer 110 is set to 0.74, 0.78, 0.82, and 0.86. Calculated values ​​of the width D and IQE are plotted on each graph. As in the first simulation, the width D was varied within the range of 1-30 nm.

[0044] 9, it can be seen that there exists a width D of the quantum well layer 110 that gives a maximum value of IQE of the LED 100, regardless of the Al composition X(a) of the quantum well layer 110. Note that, as a result of the second simulation example, it has been confirmed that when the Al composition X(a) of the quantum well layer 110 is 0.86, the IQE reaches a maximum when the width D of the quantum well layer 110 is 40 nm, and that even if the width D is made larger than 40 nm, the IQE hardly decreases.

[0045] 10 is a graph showing the relationship between the Al composition difference (ΔX) in FIG. 9 and the width (D) of the quantum well layer 110 at which the luminous efficiency is maximized for each Al composition difference (ΔX). The horizontal axis of FIG. 10 indicates the Al composition difference ΔX (=b1−a) between the high-level layer, i.e., the N-type contact layer 120 and the quantum well layer 110, and the vertical axis indicates the width D [nm] of the quantum well layer 110.

[0046] The second simulation example also yielded results similar to those of the first simulation example. Specifically, as shown in FIG. 10 , it can be seen that the smaller the Al composition difference ΔX (= b−a) between the Al composition a of the quantum well layer 110 and the Al composition b of the high-level layer, the larger the width D of the quantum well layer 110 that provides the maximum IQE value of the LED 100. In other words, it can be seen that there is a negative correlation between ΔX (= b−a), which is the difference between the Al composition a of the quantum well layer 110 and the Al composition b of the high-level layer, and the width D [nm] of the quantum well layer 110 that provides the maximum IQE value. Specifically, the smaller ΔX is, the larger the D that provides the maximum IQE.

[0047] 8 to 10 , it can be seen that in order to shorten the emission wavelength of LED 100, it is necessary to increase the Al composition of quantum well layer 110, and that increasing the Al composition of quantum well layer 110 reduces the Al composition difference ΔX with the high-level layer, thereby weakening the confinement of electrons and holes in quantum well layer 110 and reducing the maximum value of IQE of LED 100. Therefore, if it is desired to shorten the emission wavelength of LED 100, width D of quantum well layer 110 may be set to a width D that gives the maximum value of IQE of LED 100, for example, based on the results of FIG.

[0048] As a result of extensive research into a number of simulations according to the first embodiment described with reference to FIGS. 3 to 10, the inventors of the present application have found the following.

[0049] The radiative recombination in the quantum well layer 110 is proportional to the product (sum) of the electron concentration and the hole concentration in the quantum well layer 110. In other words, the luminous efficiency IQE of the LED 100 corresponds to the probability that electrons and holes meet in the quantum well layer 110, and is therefore determined by the product of the electron and hole concentrations in the quantum well layer 110. The spatial distribution of the carrier (electron and hole) concentration in the quantum well layer 110 is determined by three factors: (1) the shape of the wave function (degree of spatial overlap) at the energy levels of the carriers involved in light emission, (2) the number of energy levels of the carriers involved in light emission, and (3) the energy difference between the energy level of the carriers involved in light emission and the quasi-Fermi level.

[0050] Reducing the width D of the quantum well layer 110 improves the spatial overlap of (1) and increases the probability of electrons and holes meeting, but on the other hand, the number of (2) decreases and the energy difference of (3) increases, resulting in a lower carrier concentration and a lower probability of electrons and holes meeting. Increasing the width D of the quantum well layer 110 reduces the spatial overlap of (1) and decreases the probability of electrons and holes meeting, but on the other hand, the number of (2) increases and the energy difference of (3) decreases, resulting in a higher carrier concentration and a higher probability of electrons and holes meeting. However, if the width D of the quantum well layer 110 is made larger than a predetermined size, the increase in carrier concentration due to (2) and (3) is eliminated, and the probability of electrons and holes meeting due to (1) continues to decrease. In other words, by setting the width D of the quantum well layer 110 to the predetermined size, the IQE of the LED 100 can be maximized.

[0051] The energy difference (3) depends not only on the width D of the quantum well layer 110 but also on the magnitude [eV / nm] of the polarization electric field present in the quantum well layer 110, and the magnitude [eV / nm] of the polarization electric field depends on the magnitude of ΔX, which is the difference between the Al composition in the quantum well layer 110 and the Al composition in the high-level layer. Therefore, the width D of the quantum well layer 110 that provides the maximum IQE value of the LED 100 can be determined based on ΔX. For example, if the Al compositions of the quantum well layer 110 and the high-level layer are predetermined, i.e., ΔX is predetermined, the width D of the quantum well layer 110 that provides the maximum IQE value of the LED 100 can be determined based on the magnitude of ΔX.

[0052] Here, the above (1) and (2) will be explained using Fig. 11 and Fig. 12. Fig. 11 is a diagram showing an example of an energy band graph and graphs of the wave functions of electrons and holes when the width D of the quantum well layer 110 in the LED 100 according to the first embodiment is 3 nm. Fig. 12 is a diagram showing an example of an energy band graph and graphs of the wave functions of electrons and holes when the width D of the quantum well layer 110 in the LED 100 according to the first embodiment is 30 nm. The horizontal axis in Fig. 11 and Fig. 12 indicates the position Z [nm], and the vertical axis indicates the energy [eV].

[0053] A polarization field exists in semiconductors such as the quantum well layer 110 of the LED 100. The polarization field is divided into a conduction band where electrons exist and a valence band where holes exist, with a band gap between them where no carriers exist. Figures 11 and 12 show graphs of the energy level at the low-energy edge of the conduction band potential and the high-energy edge of the valence band potential, respectively. The same applies to similar drawings described below, and redundant explanations will be omitted.

[0054] Electrons present in the conduction band of the quantum well layer 110 tend to move toward lower energy levels, and therefore tend to accumulate on the positive Z-axis side of the quantum well layer 110, where the conduction band potential is lowest. On the other hand, holes present in the valence band of the quantum well layer 110 tend to move toward higher energy levels, and therefore tend to accumulate on the negative Z-axis side of the quantum well layer 110, where the valence band potential is highest. Therefore, the shape of the wave function at the energy level of the carriers (electrons and holes) involved in light emission, i.e., the shape of the graph of the carrier wave function, is, for example, as shown by the bold line in FIG. 11 when the width D of the quantum well layer 110 is very small, such as about 3 nm, and as shown by the bold line in FIG. 12 when the width D of the quantum well layer 110 is very large, such as about 30 nm.

[0055] 11 and 12 , it can be seen that the smaller the width D of the quantum well layer 110, the better the spatial overlap between the wave function at the energy level of the electrons involved in light emission and the wave function at the energy level of the holes involved in light emission (there are more areas where the distributions of both wave functions match, the peak positions Z in the graphs of both wave functions are closer, etc.). However, it can also be seen that the smaller the width D of the quantum well layer 110, the shallower the carrier confinement due to the smaller width D of the quantum well layer 110, and the fewer the number of energy levels of the carriers involved in light emission in (2) above (for example, carriers leak from the quantum well layer 110 to the N-type contact layer 120 or the electron blocking layer 130).

[0056] Next, the energy level and quasi-Fermi level of carriers involved in light emission in (3) above will be described with reference to FIGS. 13 and 14 . FIG. 13 shows an example of a graph of the conduction band potential, a graph of the electron wave function, and a graph of the electron quasi-Fermi level in the LED 100 according to the first embodiment. FIG. 14 shows an example of a graph of the valence band potential, a graph of the hole wave function, and a graph of the hole quasi-Fermi level in the LED 100 according to the first embodiment. The horizontal axis in FIGS. 13 and 14 indicates the position Z [nm], and the vertical axis indicates the energy [eV]. In FIGS. 13 and 14 , only the graph of the carrier quasi-Fermi level is shown by a curved dashed line, and this also applies to the subsequent figures.

[0057] According to the graphs of electron and hole wave functions shown in Figures 13 and 14, for example, there are two electron energy levels involved in light emission, and three hole energy levels involved in light emission. Correspondingly, Figure 13 shows two graphs of electron wave functions, and Figure 14 shows three graphs of hole wave functions. The product of these energy levels results in six cross terms. The overall light emission can be evaluated by evaluating the ease with which each of these six terms emits light from the perspective of (3) above and adding them all together. In the following explanation, simply for the sake of clarity, the number of carrier energy levels involved in light emission, i.e., the number of carrier wave function graphs, is assumed to be only one.

[0058] The quasi-Fermi level of the carriers may be an energy level at which the carrier concentration is 50% when the quantum well layer 110 of the LED 100 is in a non-thermal equilibrium state, for example, when an operating current (J) is flowing through the LED 100. In the above (3), a small energy difference between the energy level of the carriers involved in light emission and the quasi-Fermi level may mean that the concentration of carriers that contribute to light emission among the carriers confined in the quantum well layer 110 in a non-thermal equilibrium state is high.

[0059] In general, it is easy to make an N-type semiconductor and difficult to make a P-type semiconductor in nitride semiconductors, that is, it is easy to supply electrons but difficult to supply holes. ａ Ga１－ａ In the quantum well layer 110 having a composition of N (0<a<1), a high electron concentration is easily obtained, and the position of the electron quasi-Fermi level is not a significant issue. In other words, the electron quasi-Fermi level is located at a position where the energy difference (distance) with the conduction band potential is not a significant issue. On the other hand, Al ａ Ga １－ａ In the quantum well layer 110 having a composition of N (0<a<1), it is difficult to obtain a high hole concentration, and the quasi-Fermi level of holes is located at a position where the energy difference with the valence band potential can become a problem. Therefore, holes are a more dominant factor than electrons in the above (3). Therefore, in the following explanation of the above (1) to (3), holes will be considered.

[0060] FIG. 15 is a table showing an example of the magnitude of the polarization electric field of the quantum well layer 110 in the LED 100 according to the first embodiment. The first row of the table in FIG. 15 shows 1.1 as the magnitude of the polarization electric field of the quantum well layer having a GaN / AlN composition. In the quantum well layer having a GaN / AlN composition, the difference in Al composition between the Al composition of GaN and the Al composition of AlN can be considered to be 1.0. Therefore, in the LED 100 according to this embodiment, the magnitude of the polarization electric field of the quantum well layer 110 (eV / nm) may be estimated based on the difference in Al composition between the quantum well layer 110 and the high-level layer. The second row of the table in FIG. 15 shows 0.11 as the magnitude [eV / nm] of the polarization electric field of the quantum well layer 110 estimated when the Al composition difference is 0.10, the third row shows 0.22 as the magnitude [eV / nm] of the polarization electric field of the quantum well layer 110 estimated when the Al composition difference is 0.20, and the fourth row shows 0.33 as the magnitude [eV / nm] of the polarization electric field of the quantum well layer 110 estimated when the Al composition difference is 0.30.

[0061] 16 is a graph showing an example of the relationship between the width D of the quantum well layer 110 and the valence band potential in the LED 100 according to the first embodiment. The horizontal axis of FIG. 16 indicates the position Z [nm], and the vertical axis indicates the energy [eV].

[0062] 16 shows a graph of the valence band potential in three LEDs 100, each having the same Al composition difference between the quantum well layer 110 and the high-level N-type contact layer 120, the same quasi-Fermi level of holes, and different widths D of the quantum well layer 110. The LED 100 with the smallest width D of the quantum well layer 110 among these three LEDs 100 has the same Al composition difference between the quantum well layer 110 and the high-level N-type contact layer 120, the same quasi-Fermi level of holes, and different widths D of the quantum well layer 110. Ｉ The solid line shows the graph of the valence band potential I when the width of the quantum well layer 110 is the next smallest, D ＩＩ The dashed line indicates the graph of the valence band potential II when the width of the quantum well layer 110 is the largest, D ＩＩＩ The dashed line also shows a graph of the valence band potential III when

[0063] First, comparing the graph of valence band potential I and the graph of valence band potential II shown in FIG. 16, the magnitude of the polarization electric field is the same, but the width of the quantum well layer 110 is smaller, D Ｉ The graph of the valence band potential I when the width of the quantum well layer 110 is larger is far from the graph of the hole quasi-Fermi level. ＩＩ The graph of the valence band potential II is tangent to the graph of the hole quasi-Fermi level when the width of the quantum well layer 110 is small. That is, there is an energy difference between the valence band potential I and the hole quasi-Fermi level when the width of the quantum well layer 110 is small, while the energy difference between the valence band potential II and the hole quasi-Fermi level when the width of the quantum well layer 110 is large is zero.

[0064] 16, both graphs of valence band potential II and valence band potential III are in contact with the graph of the hole quasi-Fermi level, but the magnitude of the polarization electric field is different. That is, the energy difference between the hole quasi-Fermi level and both valence band potential II and valence band potential III is 0, while the width D of the quantum well layer 110 corresponding to valence band potential III is 0. ＩＩＩ is the width D of the quantum well layer 110 corresponding to the valence band potential II ＩＩ is greater than.

[0065] As described above, increasing the width D of the quantum well layer 110 reduces the spatial overlap of the above (1) and reduces the probability of electrons and holes meeting, but on the other hand, the number of the above (2) increases and the energy difference of the above (3) decreases, so the carrier concentration increases and the probability of electrons and holes meeting increases. More specifically, as can be seen from the graph in Figure 16, as the width D of the quantum well layer 110 increases, the IQE increases due to the effect of the increase in hole concentration caused by the above (2) and (3) until the energy difference between the valence band potential and the quasi-Fermi level of the holes becomes zero.

[0066] However, as described above, if the width D of the quantum well layer 110 is made larger than a predetermined size, the increase in carrier concentration due to (2) and (3) will disappear, and the probability of encounter due to (1) will simply continue to decrease. More specifically, as can be seen from the graph in Fig. 16, if the width D of the quantum well layer 110 is still increased even after the energy difference between the valence band potential and the quasi-Fermi level of the holes becomes 0, the increase in hole concentration due to (2) and (3) will not occur, and the IQE will decrease due to the effect of (1).

[0067] As described above, it can be seen from the graph in Fig. 16 that there exists a width D of the quantum well layer 110 that gives a maximum value of IQE of the LED 100. From the above, in the LED 100, the width D of the quantum well layer 110 may be the smallest value within the range of D in which the energy difference between the valence band potential or conduction band potential of the polarization electric field present in the quantum well layer 110 and the quasi-Fermi level of the carriers present in the quantum well layer 110 becomes zero. For example, the width D may be the smallest value within the range of D in which the energy difference between the valence band potential and the quasi-Fermi level of the holes present in the quantum well layer 110 becomes zero. For example, the width D of the quantum well layer 110 may be the smallest value within the range of D in which the energy difference between the valence band potential and the quasi-Fermi level of the holes present in the quantum well layer 110 becomes zero among the three widths D shown in Fig. 16 . ＩＩ may be.

[0068] As a more specific example, assuming the carrier concentration during operation of the LED 100, the energy difference between the hole quasi-Fermi level and the valence band potential may be approximately 0.4 eV. In this case, for example, if the Al composition difference ΔX is 0.10, i.e., if the magnitude of the polarization electric field is 0.11 eV / nm (based on the table of FIG. 15 ), the width D of the quantum well layer 110 may be approximately 3.6 nm, and the energy difference between the hole quasi-Fermi level and the valence band potential may become 0. Therefore, when the Al composition difference ΔX is 0.10, the width D of the quantum well layer 110 may be set to approximately 3.6 nm to maximize the IQE of the LED 100.

[0069] In the example of Figure 16, the three LEDs 100 with different widths D of the quantum well layer 110 are described as having the same quasi-Fermi level of holes, but the quasi-Fermi level of carriers can change depending on the carrier concentration, and the quasi-Fermi level can increase as the carrier concentration increases.

[0070] Next, the energy difference between the energy level of carriers involved in light emission and the quasi-Fermi level in (3) above will be described with reference to FIGS. 17 and 18 . FIG. 17 shows an example of the energy difference G between the energy level of holes and the quasi-Fermi level of holes in an LED according to a comparative example, where the Al composition difference ΔX between the high-level layer and the quantum well layer is 0.10 and the width D of the quantum well layer is 3 nm. FIG. 18 shows an example of the energy difference G between the energy level of holes and the quasi-Fermi level of holes in the LED 100 according to the first embodiment, where the Al composition difference ΔX between the high-level layer and the quantum well layer 110 is 0.10 and the width D of the quantum well layer 110 is 5 nm. In FIGS. 17 and 18 , the range of the quantum well layer is indicated by a shaded area. The horizontal axis in FIGS. 17 and 18 indicates position Z [nm], and the vertical axis indicates energy [eV].

[0071] In the comparative LED shown in Figure 17, the graph of the hole quasi-Fermi level and the graph of the valence band potential do not touch. As shown in Figure 17, the energy difference between the hole quasi-Fermi level and the valence band potential is approximately 0.45 eV. Therefore, if the Al composition difference ΔX is 0.10, i.e., if the magnitude of the polarization electric field is 0.11 eV / nm (based on the table in Figure 15), the quantum well layer width D is approximately 4.1 nm, and the energy difference between the hole quasi-Fermi level and the valence band potential can be zero. However, in the comparative LED, the quantum well layer width D is 3 nm, so there is an energy difference between the hole quasi-Fermi level and the valence band potential, and the energy difference G between the hole energy level and the hole quasi-Fermi level is relatively large. Therefore, the concentration of carriers contributing to light emission among the carriers confined in the quantum well layer in a non-thermal equilibrium state is relatively low. 17, the graph of the hole wave function is included in a relatively low proportion of the quantum well layer, i.e., the number of energy levels of the carriers involved in the light emission described above in (2) is relatively small. Taking all of the above into consideration, it can be seen that the IQE of the LED according to the comparative example does not reach the maximum value.

[0072] On the other hand, in the LED 100 according to the first embodiment shown in FIG. 18 , the graph of the hole quasi-Fermi level and the graph of the valence band potential are tangent. As shown in FIG. 18 , the energy difference between the hole quasi-Fermi level and the valence band potential is approximately 0.52 eV. Therefore, if the Al composition difference ΔX is 0.10, i.e., if the magnitude of the polarization electric field is 0.11 eV / nm (based on the table in FIG. 15 ), the quantum well layer width D is approximately 4.7 nm, and the energy difference between the hole quasi-Fermi level and the valence band potential can be zero. In the LED 100 according to the first embodiment, the quantum well layer width D is 5 nm, so the energy difference between the hole quasi-Fermi level and the valence band potential is zero, and the energy difference G between the hole energy level and the hole quasi-Fermi level is relatively small. Therefore, the concentration of carriers contributing to light emission among the carriers confined in the quantum well layer in a non-thermal equilibrium state is relatively high. 18, the graph of the hole wave function is included in the quantum well layer 110 to a relatively high extent, i.e., the number of energy levels of the carriers involved in the light emission described in (2) above is relatively large. Taking all of the above into consideration, it can be seen that the IQE of the LED 100 according to the first embodiment is approximately the maximum value.

[0073] As explained above, the energy difference between the quasi-Fermi level of holes and the valence band potential is determined by the Al composition difference ΔX between the quantum well layer 110 and the high-level layer, and the width D of the quantum well layer 110. To support this, simulation results will be explained using Figures 19 to 21, showing that the energy difference between the quasi-Fermi level of holes and the valence band potential remains approximately constant even when the Al composition of each layer of the LED 100 is changed uniformly.

[0074] Fig. 19 is a table showing a simulation example of the Al composition and film thickness when the Al composition of each layer in the LED 100 according to the first embodiment is uniformly changed by +α. The first row of the table in Fig. 19 shows, from left to right, X (Al composition) and film thickness [nm]. The second to fifth rows of the table in Fig. 19 show X and film thickness of the P-type contact layer 150 (p-GaN), P-type cap layer 140 (p-AlGaN), electron blocking layer 130 (EBL), quantum well layer 110 (QW), and N-type contact layer 120 (n-AlGaN), respectively.

[0075] 19 , the Al composition X and thickness of the P-type contact layer 150 are set to 0 and 100 nm. The Al composition X (b3) and thickness of the P-type cap layer 140 are set to 0.5+α and 70 nm. The Al composition X (b2) and thickness of the electron block layer 130 are set to 0.6+α and 9 nm. The Al composition X (a) and thickness of the quantum well layer 110 are set to 0.2+α and 2 nm. The Al composition X (b1) and thickness of the N-type contact layer 120 are set to 0.4+α and 350 nm. In this simulation example, the Al compositions of the P-type cap layer 140, electron block layer 130, quantum well layer 110, and N-type contact layer 120 are uniformly changed by +α (α=0, 0.1, 0.2, 0.3, 0.4).

[0076] Fig. 20 is a graph showing the simulation results of the transition of the graph of the hole quasi-Fermi level and the graph of the valence band potential when the Al composition of each layer is uniformly changed by +α in the simulation example of Fig. 19. The horizontal axis of Fig. 20 indicates the position Z [nm], and the vertical axis indicates the energy [eV].

[0077] 20 shows graphs of the hole quasi-Fermi level and the valence band potential for five patterns of α = 0, 0.1, 0.2, 0.3, and 0.4. Fig. 20 also shows the energy difference ΔE between the hole quasi-Fermi level and the valence band potential.

[0078] Fig. 21 is a graph showing the relationship between each α in Fig. 20 and the energy difference ΔE. The horizontal axis of Fig. 21 indicates α, and the vertical axis indicates the energy difference ΔE [eV]. Fig. 20 and Fig. 21 show that the energy difference between the quasi-Fermi level of holes and the valence band potential remains approximately constant even if the Al composition of each layer of LED 100 is uniformly changed.

[0079] In the first and second simulation examples described above, the strain state of the quantum well layer 110 was set to coherent strain (R = 0%). Here, simulation results for different strain states of the quantum well layer 110 are shown. Fig. 22 is a graph showing a simulation example of the relationship between the Al composition difference (ΔX) between the high-level layer and the quantum well layer 110 and the width (D) of the quantum well layer 110, which changes depending on the influence of the interface between the quantum well layer 110 and a layer adjacent to the quantum well layer 110, in the LED 100 according to the first embodiment. The horizontal axis of Fig. 22 indicates the Al composition difference ΔX between the high-level layer and the quantum well layer 110, and the vertical axis indicates the width D [nm] of the quantum well layer 110.

[0080] 22 plots calculated values ​​for the quantum well layer 110 when the strain state is coherent (R=0%) and fully relaxed (R=100%), with the AlGaN of the N-type contact layer 120, which is a high-level layer, as a reference. Furthermore, as another indicator of the strain state of the quantum well layer 110, FIG. 22 plots calculated values ​​when the lattice constant of the reference N-type contact layer is a=3.112 [Å], which is the lattice constant of AlN, and calculated values ​​when the lattice constant of GaN is a=3.189 [Å], which is the lattice constant of GaN. The strain states of the four graphs shown in FIG. 22 are shown in Table 1 below. a=3.112 [Å] represents the state in which the AlGaN-based LED 100 is subjected to the strongest compressive strain, and a=3.189 [Å] represents the state in which the AlGaN-based LED 100 is subjected to the strongest tensile strain. 22 , in the LED 100 according to the first embodiment, the difference ΔX between the Al composition in the quantum well layer 110 and the Al composition in the high-level layer, i.e., the difference between a and b, and the width D [nm] of the quantum well layer 110 satisfy the following relational expression 1: [Relational Expression 1] D=k / ΔX (where, in Relational Expression 1, D>3, 0<ΔX<0.15, and 0.45≦k≦1.25 (the coefficient k is in nm)).

[0081] The configurations and compositions of the LED 100 according to the first embodiment described above are not limited to the case where the LED 100 is an AlGaN-based LED, but may be extended to an InGaN (indium gallium nitride)-based LED, an AlInGaN (aluminum indium gallium nitride)-based LED, or the like.

[0082] For example, in an InGaN-based LED, the quantum well layer 110 is made of In １－ｃ Ga ｃ The high-level layer may have a composition of InN (0<c<1). １－ｄ Ga ｄ In this case, the difference between the Ga composition in the quantum well layer 110 and the Ga composition in the high-level layer, i.e., the difference between the above-mentioned c and d, ΔX (=d−c), and the width D [nm] of the quantum well layer 110 satisfy the following relational expression 2: [Relational expression 2] D=k / ΔX (where, in relational expression 2, D≧2, 0<ΔX≦0.15, and 0.3≦k≦0.6 (the unit of the coefficient k is [nm])).

[0083] An InGaN-based LED will be described in more detail with reference to Fig. 23 to Fig. 28. Fig. 23 is a schematic diagram showing an example of the structure of an LED 200 according to the second embodiment. The LED 200 according to the second embodiment includes a quantum well layer 210, an N-type contact layer 220, an electron blocking layer 230, a P-type contact layer 250, an N-layer electrode 260, a P-layer electrode 270, and a substrate 280.

[0084] 23 , the structure of the LED 200 according to the second embodiment differs from the structure of the LED 100 according to the first embodiment in that it does not include the P-type cap layer 140, and a P-type contact layer 250 is located on the positive side of the Z axis of the quantum well layer 210. The functional configuration of each layer of the LED 200 according to the second embodiment is similar to the functional configuration of the corresponding layer of the LED 100 according to the first embodiment, and therefore the same reference numbers as those of the corresponding layers of the LED 100 according to the first embodiment are used and redundant explanations will be omitted.

[0085] The quantum well layer 210 is, for example, In １－ｃ Ga ｃ The N-type contact layer 220 has a composition of N (0<c<1). For example, the N-type contact layer 220 is n-In doped with at least one kind of impurity, for example, Si. １－ｄ１ Ga ｄ１ The electron blocking layer 230 has a composition of AlN (c<d1≦1). ｊ In １－ｄ２ Ga ｄ２ In the LED 200 according to the second embodiment, the N-type contact layer 220 and the electron blocking layer 230 have a composition of In １－ｄ Ga ｄ 23 , the high-level layer may refer to both the N-type contact layer 220 and the electron blocking layer 230, or to either one of them. Furthermore, in the second embodiment, the expression "Ga composition d of the high-level layer" may refer to the average of the Ga composition d1 of the N-type contact layer 220 and the Ga composition d2 of the electron blocking layer 230, or to the Ga composition of either one of them. ΔX refers to the Ga composition difference (d−c) between the Ga composition d of the high-level layer and the Ga composition c of the quantum well layer 210.

[0086] Fig. 24 is a diagram showing an example of the Z-axis position and Ga composition X of each layer of the LED 200 according to the second embodiment. Fig. 25 is a table showing a simulation example of the Ga composition X, film thickness, etc. of each layer of the LED 200 according to the second embodiment.

[0087] The horizontal axis in FIG. 24 indicates the position Z [nm] in the Z-axis direction, and the vertical axis indicates the Ga composition X, which is defined in the range of 0 to 1 (0% to 100%). In this embodiment, "X" is a collective term for the above-mentioned "c," "d," "d1," and "d2." However, the vertical axis exceptionally shows the range up to the Ga composition X of 1.2, which corresponds to the case where an Al composition of 0.2 is added to a Ga composition X of 1. In FIG. 24, the Ga composition X in each layer of the LED 200 is shown as a continuous thick line graph. In addition, FIG. 24 shows the layers present at each position Z in a bar graph above the graph.

[0088] The first line of the table in FIG. 25 shows, from the left, X (Ga composition), film thickness [nm], and doping concentration [cm －３ The second to fifth rows of the table in Fig. 25 show the Ga composition X, film thickness, etc. of the P-type contact layer 250 (p-GaN), the electron blocking layer 230 (EBL), the quantum well layer 210 (QW), and the N-type contact layer 220 (n-InGaN), in that order.

[0089] In the simulation example shown in FIG. 25, the Ga composition X, film thickness, and doping concentration of the P-type contact layer 250 are set to 0, 100 [nm], 1×10 １９ [cm －３ The Ga composition X (d2), film thickness, and doping concentration of the electron blocking layer 230 are set to 1.2 (Ga: 1.0, Al: 0.20), 9 [nm], and 1×10 １９ [cm －３ ]

[0090] 25, the Ga composition X(d1) of the N-type contact layer 220 was set to 0.85, 0.90, 0.95, and 0.98, respectively. The thickness and doping concentration of the N-type contact layer 220 were set to 350 nm and 1×10 １９ [cm －３ In this simulation example, the film thickness, i.e., width, of the quantum well layer 210 is varied within the range of 1-12 nm for each Ga composition X(d1) of the N-type contact layer 220. The Ga composition X(c) of the quantum well layer 210 is set to 0.80. When the Ga composition X(c) of the quantum well layer 210 is 0.80, the emission wavelength of the LED 200 is approximately 400 nm.

[0091] Fig. 26 is a graph showing the simulation results of the relationship between the width (D) of the quantum well layer 210 and the luminous efficiency (IQE) based on the simulation example of Fig. 25. The horizontal axis of Fig. 26 indicates the width D [nm] of the quantum well layer 210, and the vertical axis indicates the luminous efficiency (IQE) [%] of the LED 200.

[0092] In this simulation, the LED 200 is irradiated with 10 [A / cm ２ ] is passed through the LED 200. Figure 26 shows graphs of the IQE of the LED 200 as the width D of the quantum well layer 210 changes when the Ga composition X (d1) of the N-type contact layer 220 is set to 0.85, 0.90, 0.95, and 0.98. Calculated values ​​of the width D and IQE are plotted on each graph. As described in Figure 25, the width D was changed within the range of 1-12 nm.

[0093] As shown in FIG. 26, it can be seen that there exists a width D of the quantum well layer 210 that gives the LED 200 a maximum value of IQE, regardless of the Ga composition X(d1) of the N-type contact layer 220.

[0094] Fig. 27 is a graph showing the relationship between the Ga composition difference (ΔX) in Fig. 26 and the width (D) of the quantum well layer 210 at which the luminous efficiency is maximized for each Ga composition difference (ΔX). The horizontal axis of Fig. 27 indicates the Ga composition difference ΔX (= d1 - c) between the high-level layer, i.e., the N-type contact layer 220 and the quantum well layer 210, and the vertical axis indicates the width D [nm] of the quantum well layer 210.

[0095] 27 , it can be seen that in the range of 0<ΔX≦0.15, the smaller the Ga composition difference ΔX (=d−c) between the Ga composition c in the quantum well layer 210 and the Ga composition d in the high-level layer, the larger the width D of the quantum well layer 210 that provides the maximum IQE value for the LED 200. In other words, it can be seen that in this range, there is a negative correlation between ΔX (=d−c), which is the difference between the Ga composition c in the quantum well layer 210 and the Ga composition d in the high-level layer, and the width D [nm] of the quantum well layer 210 that provides the maximum IQE value. Specifically, the smaller ΔX is, the larger the D that provides the maximum IQE.

[0096] FIG. 28 is a table showing an example of the magnitude of the polarization electric field of the quantum well layer 210 in the LED 200 according to the second embodiment. The first row of the table in FIG. 28 shows 1.6 as the magnitude of the polarization electric field of the quantum well layer having an InN / GaN composition. In the quantum well layer having an InN / GaN composition, the Ga composition difference between the Ga composition of InN and the Ga composition of GaN can be regarded as 1.0. Therefore, in the LED 200 according to the second embodiment, as in the LED 100 according to the first embodiment, the magnitude of the polarization electric field of the quantum well layer 210 (eV / nm) may be estimated based on the Al composition difference between the quantum well layer 210 and the high-level layer. The second row of the table in FIG. 28 shows 0.16 as the magnitude [eV / nm] of the polarization electric field of the quantum well layer 210 estimated when the Ga composition difference is 0.10, the third row shows 0.32 as the magnitude [eV / nm] of the polarization electric field of the quantum well layer 210 estimated when the Ga composition difference is 0.20, and the fourth row shows 0.48 as the magnitude [eV / nm] of the polarization electric field of the quantum well layer 210 estimated when the Ga composition difference is 0.30.

[0097] 29 is a graph showing a simulation example of the relationship between the Al composition difference (ΔX) between the high-level layer and the quantum well layer 210 and the width (D) of the quantum well layer 210 when the strain state of the quantum well layer 210 is set to coherent strain in the LED 200 according to the second embodiment. The horizontal axis of Fig. 29 indicates the Al composition difference ΔX between the high-level layer and the quantum well layer 210, and the vertical axis indicates the width D [nm] of the quantum well layer 210.

[0098] FIG. 29 shows a quantum well layer 210 made of In ０．２ Ga ０．８ The calculated values ​​are for the case where the quantum well layer 210 has a composition of In. ０．３ Ga ０．７29 plots calculated values ​​for the case where the quantum well layer 210 has a composition of N. The simulation results in Fig. 29 reveal that in the LED 200 according to the second embodiment, the difference between the Al composition in the quantum well layer 210 and the Al composition in the high-level layer, i.e., the difference between the above-mentioned c and d, ΔX (= d - c), and the width D [nm] of the quantum well layer 110, satisfy the following relational expression 2: [Relational expression 2] D = k / ΔX (where, in relational expression 2, D ≥ 2, 0 < ΔX ≤ 0.15, and 0.3 ≤ k ≤ 0.6 (the coefficient k is in nm)).

[0099] In the above embodiments, the quantum well layers 110 and 210 are, for example, single quantum well layers (QW layers). Alternatively, the quantum well layers 110 and 210 may be multiple quantum well layers (MQW layers). When the quantum well layer 110 according to the first embodiment is an MQW layer, at least one of the layers for confining carriers in the MQW layer is Al. ａ Ga １－ａ When the quantum well layer 210 according to the second embodiment is an MQW layer, at least one of the layers for confining carriers in the MQW layer is made of In. １－ｃ Ga ｃ It has a composition of N (0<c<1).

[0100] 30 is a graph showing a simulation example of the relationship between the Al composition difference (ΔX) between the high-level layer and the quantum well layer 110 and the width (D) of the quantum well layer 110 when the quantum well layer 110 of the LED 100 according to the first embodiment is a three-layer multiple quantum well (MQW) layer with a coherent strain state. As with the graph of FIG. 22 , the horizontal axis of FIG. 30 indicates the Al composition difference ΔX between the high-level layer and the quantum well layer 110, and the vertical axis indicates the width D [nm] of the quantum well layer 110. The simulation results of FIG. 30 show that even when the quantum well layer 110 is, for example, a three-layer MQW layer, the difference between the Al composition in the quantum well layer 110 and the Al composition in the high-level layer, i.e., the difference ΔX between a and b, and the width D [nm] of the quantum well layer 110 satisfy the following relational expression 1: [Relationship 1] D = k / ΔX (where, in Relationship 1, D > 3, 0 < ΔX < 0.15, and 0.45 ≦ k ≦ 1.25 (the coefficient k is in nm)).

[0101] Fig. 31 is a diagram showing another example of the Z-axis position and Al composition X of each layer of the LED 100 according to the first embodiment. Fig. 32 is a table showing experimental examples of the Al composition X, film thickness, etc. of each layer of the LED 100 according to the first embodiment.

[0102] The horizontal axis in Fig. 31 indicates the position Z [nm] in the Z-axis direction, and the vertical axis indicates the Al composition X, which is defined in the range of 0 to 1 (0% to 100%). In Fig. 31, the Al composition X in each layer of the LED 100 is shown as a continuous thick line graph. In Fig. 31, the layers present at each position Z are also shown in a bar graph above the graph.

[0103] The first row of the table in Fig. 32 shows, from left to right, X (Al composition) and film thickness [nm]. The second to fifth rows of the table in Fig. 32 show, in order, the Al composition X and film thickness of the P-type contact layer 150 (p-GaN), the P-type cap layer 140 (p-AlGaN), the electron blocking layer 130 (EBL), the quantum well layer 110 (QW), and the N-type contact layer 120 (n-AlGaN).

[0104] 32, five LEDs 100 were prepared, each differing only in the thickness of the quantum well layer 110, ranging from 3 to 20 nm. For each of the five LEDs 100, the Al composition X(a) of the quantum well layer 110 was 0.82, the Al composition X and thickness of the P-type contact layer 150 were 0 and 50 nm, and the Al composition X(b3) and thickness of the P-type cap layer 140 were 0.94 and 90 nm. The Al composition X(b2) and thickness of the electron blocking layer 130 were 1 and 10 nm, and the Al composition X(b1) and thickness of the N-type contact layer 120 were 0.87 and 1000 nm.

[0105] Fig. 33 is a graph showing experimental results of the relationship between the width (D) of the quantum well layer 110, the emission wavelength, and the intensity, based on the experimental example of Fig. 32. The horizontal axis of Fig. 33 indicates the width D [nm] of the quantum well layer 110, the vertical axis on the right indicates the emission wavelength [nm] of the LED 100, and the vertical axis on the left indicates the emission intensity [a.u.] of the LED 100.

[0106] In this experimental example, a current of 20 mA was passed through each LED 100, and the emission wavelength and emission intensity of each LED 100 were measured. As shown in the graph in Figure 33, the experimental results showed that the emission wavelength [nm] remained almost constant regardless of the film thickness, even when the width D of the quantum well layer 110 of the LED 100, i.e., the film thickness, varied within a range of 3-20 nm. Furthermore, the experimental results showed that the emission intensity increased almost monotonically as the film thickness of the quantum well layer 110 of the LED 100 increased within a range of 3-20 nm. The experimental results shown in Figure 33 are close to the behavior of the simulation results when the Al composition difference between the N-type contact layer 120 and the quantum well layer 110 was set to 0.05, as in this experimental example, and it can be seen that the calculations in the simulation and the experimental results are consistent.

[0107] In the above embodiments, the LEDs 100 and 200 may include a light-emitting layer composed of well layers and barrier layers instead of the quantum well layers 110 and 210, and the Al composition of the well layers may be lower than that of the barrier layers. The light-emitting layer may have a multiple quantum well structure including multiple pairs of well layers and barrier layers, and in this case, the Al composition of each well layer and barrier layer may be different for each pair. That is, the Al composition of the well layers may be different for each well, and the Al composition of the barrier layers may be different for each barrier. In any of these cases, the light-emitting layer closest to the N-type contact layer 120 or 220 may be a well layer or a barrier layer. Similarly, the light-emitting layer closest to the electron blocking layer 130 or 230 may be a well layer or a barrier layer.

[0108] Although the present invention has been described above using embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments. It will be apparent to those skilled in the art that various modifications and improvements can be made to the above embodiments. It is clear from the claims that such modifications and improvements can also be included within the technical scope of the present invention.

[0109] It should be noted that the order of execution of each process, such as operations, procedures, steps, and stages, in the devices, systems, programs, and methods shown in the claims, specifications, and drawings is not specifically stated as "before," "prior to," etc., and that the processes can be performed in any order unless the output of a previous process is used in a subsequent process. Even if the operational flow in the claims, specifications, and drawings is described using "first," "next," etc. for convenience, this does not mean that the processes must be performed in this order.

[0110] 100 LED 110 Quantum well layer 120 N-type contact layer 130 Electron blocking layer 140 P-type cap layer 150 P-type contact layer 160 N-layer electrode 170 P-layer electrode 180 Substrate 200 LED 210 Quantum well layer 220 N-type contact layer 230 Electron blocking layer 250 P-type contact layer 260 N-layer electrode 270 P-layer electrode 280 Substrate

Claims

1. An LED having a quantum well structure, Al ａ Ga １－ａ A quantum well layer having a composition of N (aluminum gallium nitride) (0 < a < 1), Al ｂ Ga １－ｂ A high-level layer having composition N (a < b ≤ 1), adjacent to the quantum well layer, and having a higher energy level than the quantum well layer, Equipped with, The difference between a and b, ΔX, and the width D [nm] of the quantum well layer satisfy the following relational expression 1. LED. [Relationship 1] D = k / ΔX (However, in relation 1, D > 3, 0 < ΔX < 0.15, and 0.45 ≤ k ≤ 1.25 (the unit of the coefficient k is [nm]).)

2. The high-level layer is an N-type contact layer that supplies electrons to the quantum well layer. The LED according to claim 1.

3. The a satisfies 0.7 ≤ a < 1. The LED according to claim 1.

4. The high-level layer is an N-type contact layer that supplies electrons to the quantum well layer, The above a satisfies 0.7 ≤ a < 1. The LED according to claim 1.

5. The high-level layer is an N-type contact layer that supplies electrons to the quantum well layer, The above a satisfies 0.7 ≤ a < 1, The above D satisfies D ≥ 5. The LED according to claim 1.

6. The value D is the smallest value within the range of D at which the energy difference between the valence band potential or conduction band potential of the polarization electric field present in the quantum well layer and the pseudo-Fermi level of the carriers present in the quantum well layer becomes zero. The LED according to claim 1.

7. The value D is the smallest value within the range of D at which the energy difference between the valence band potential and the pseudo-Fermi level of holes present in the quantum well layer becomes zero. The LED according to claim 6.

8. An LED having a quantum well structure, In １－ｃ Ga ｃ A quantum well layer having the composition N (indium gallium nitride) (0 < c < 1), In １－ｄ Ga ｄ A high-level layer having a composition of N (c < d ≤ 1), adjacent to the quantum well layer, and having an energy level higher than the quantum well layer. Equipped with, The difference between c and d, ΔX, and the width D [nm] of the quantum well layer satisfy the following relational equation 2. LED. [Relationship 2] D = k / ΔX (However, in relation 2, D ≥ 2, 0 < ΔX ≤ 0.15, and 0.3 ≤ k ≤ 0.6 (the unit of the coefficient k is [nm]).)

9. The quantum well layer is a single quantum well layer, The aforementioned high-level layer is an N-type contact layer that supplies electrons to the quantum well layer. The LED according to claim 8.

10. The value D is the smallest value within the range of D at which the energy difference between the valence band potential or conduction band potential of the polarization electric field present in the quantum well layer and the pseudo-Fermi level of the carriers present in the quantum well layer becomes zero. The LED according to claim 8.

11. The value D is the smallest value within the range of D at which the energy difference between the valence band potential and the pseudo-Fermi level of holes present in the quantum well layer becomes zero. The LED according to claim 10.

12. A method for manufacturing an LED having a quantum well structure, Al ａ Ga １－ａ forming a quantum well layer having a composition of N (aluminum gallium nitride) (0 < a < 1), Al ｂ Ga １－ｂ Having a composition of N (a < b ≤ 1), adjacent to the quantum well layer, and forming a high-level layer with a higher energy level than the quantum well layer. Equipped with, The difference between a and b, ΔX, and the width D [nm] of the quantum well layer satisfy the following relational expression 1. LED manufacturing method. [Relationship 1] D = k / ΔX (However, in relation 1, D > 3, 0 < ΔX < 0.15, and 0.45 ≤ k ≤ 1.25 (the unit of the coefficient k is [nm]).)

13. A method for manufacturing an LED having a quantum well structure, In １－ｃ Ga ｃ Forming a quantum well layer having the composition N (indium gallium nitride) (0 < c < 1), In １－ｄ Ga ｄ Having a composition of N (c < d ≤ 1), adjacent to the quantum well layer, and forming a high-level layer with a higher energy level than the quantum well layer. Equipped with, The difference between c and d, ΔX, and the width D [nm] of the quantum well layer satisfy the following relational equation 2. LED manufacturing method. [Relationship 2] D = k / ΔX (However, in relation 2, D ≥ 2, 0 < ΔX ≤ 0.15, and 0.3 ≤ k ≤ 0.6 (the unit of the coefficient k is [nm]).)