JP2008311532A - White light emitting device, and forming method of white light emitting device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a white light emitting device which cause damage to the retina to a small extent, and to provide a forming method of the white light emitting device. <P>SOLUTION: An LED 20 is disposed in a recessed portion of a package 21, and the recessed portion is filled with a phosphor 22 and a sealing resin to a certain height. The phosphor 22 receives light from the LED 20 and emits light having a wavelength longer than the light emission wavelength thereof. The light emitted through the phosphor serves as a white light source having light emission spectrum components within a wavelength range of 350 to 700 nm. A filter 23 is provided to remove light emission spectrum components in part of a wavelength range of 380 to 480 nm in emissive light from the white light source. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a white light emitting device including a blue wavelength to a red wavelength and a configuration method thereof.

  Visible light emitting diodes or the like using GaN-based III-V group compound semiconductors such as GaN and AlGaN have been put into practical use as illumination light sources from display light sources. And as a light emitting diode for illumination, white LED with high luminous efficiency has been developed.

  By the way, a human can feel each color of R, G, and B by three types of cone cells, and recognizes the color by mixing these colors. White light is formed by mixing three kinds of colors of red (R), blue (B), and green (G). One of the models that model color recognition and give rules for color handling is color quantification and chromaticity diagram by CIE1931 color matching function.

  The color matching function is shown in FIG. In the figure, the vertical axis represents relative sensitivity or specific visibility, and the horizontal axis represents the wavelength of light. On the other hand, light incident on the human cornea reaches the fundus without being substantially attenuated in the visible light region.

  Even weak light (LED) that does not cause thermal action causes the molecules in the cell to be excited by light, causing a chemical reaction and causing tissue damage. Specifically, the photochemical reaction is strongest in the vicinity of 435 to 440 nm of blue light, and the fundus is in a wavelength region that is easily damaged. FIG. 21 shows a retinal damage action spectrum (retinal damage) showing the susceptibility to retinal damage with respect to the wavelength of light in a graph. As can be seen from the figure, the sensitivity to feel blue and the action spectrum of retinal damage are very close. That is, it can be said that the blue color causes retinal damage.

  However, in order to create a white light source, it must contain a blue component, which can also be a spectrum that damages the retina. So far, this retinal damage has not been taken into account, and the risk of retinal damage has been increased by prolonged irradiation.

  The present invention was created to solve the above-described problems, and an object thereof is to provide a white light emitting device with a low degree of retinal damage and a method for forming a white light emitting device based on a quantitative evaluation method. It is said.

  In order to achieve the above object, the invention according to claim 1 is a white light source having an emission spectrum component within a wavelength range of 350 nm to 700 nm, and a range of wavelengths from 380 nm to less than 480 nm among the emitted light from the white light source. A white light emitting device comprising spectral component removing means for removing an emission spectral component in at least a part of the wavelength region.

  The invention according to claim 2 is the white light emitting device according to claim 1, wherein the removing means is constituted by a filter.

  Further, in the invention according to claim 3, the white light source absorbs a part of the light emitted from the semiconductor light emitting diode having an emission spectrum peak in a wavelength range of 350 nm to 420 nm and the semiconductor light emitting diode. The white light-emitting device according to claim 1, wherein the white light-emitting device comprises a phosphor that emits light having a wavelength longer than that of the emitted light.

  According to a fourth aspect of the present invention, there is provided a semiconductor light emitting device in which a white light source having an emission spectrum component in a wavelength range of 350 nm to 700 nm has an emission spectrum peak in each of a wavelength range of 480 nm to 520 nm and a wavelength range of 630 nm to 670 nm. A white light-emitting device including a diode.

  The invention described in claim 5 is characterized in that the semiconductor light emitting diode having an emission spectrum peak in the wavelength range of 480 nm to 520 nm is made of a nitride semiconductor having an m-plane as a main surface for crystal growth. 4. The white light emitting device according to 4.

  According to a sixth aspect of the present invention, there is provided a semiconductor laser wherein a white light source having an emission spectrum component in a wavelength range of 350 nm to 700 nm has an emission spectrum peak in each of a wavelength range of 480 nm to 520 nm and a wavelength range of 630 nm to 670 nm. A white light-emitting device including a diode.

  According to a seventh aspect of the invention, the semiconductor laser diode having an emission spectrum peak in the wavelength range of 480 nm to 520 nm is made of a nitride semiconductor having an m-plane as a main surface for crystal growth. 6. The white light emitting device according to 6.

  In the invention according to claim 8, the white light source has an emission spectrum peak in each of a wavelength range of 380 nm to 420 nm, a wavelength of 480 nm to 500 nm, and a wavelength of 520 nm to 670 nm, and the emission spectrum in a wavelength range of 420 nm to 480 nm. 3. The white light emitting device according to claim 1, wherein no component is contained.

  The invention according to claim 9 is an emission source having an emission spectrum peak in the wavelength range of 380 nm to 420 nm, an emission source having an emission spectrum peak in the wavelength range of 480 nm to 500 nm, and a wavelength range of 520 nm to 670 nm. 9. The white light emitting device according to claim 8, wherein the light emission source having a peak of the emission spectrum is composed of a semiconductor light emitting diode.

  The invention described in claim 10 is characterized in that the semiconductor light emitting diode having the emission spectrum peak in the wavelength range of 380 nm to 420 nm and wavelength of 480 nm to 500 nm is made of a nitride semiconductor. A light emitting device.

  Further, in the invention according to claim 11, when the spectrum of the emitted light from the white light emitting device is evaluated by a color matching function of CIE1931, CIEx is within a range of 0.3 to 0.5 in the chromaticity diagram. And CIEy is contained in the range of 0.3-0.44, The white light-emitting device of any one of Claims 1-10 characterized by the above-mentioned.

  Further, in the invention according to claim 12, among the contribution component amount to the CIE1931 blue color matching function in the spectral component of the emitted light from the light emitting device, the wavelength 450 nm to the first contribution component amount in the wavelength range of 400 nm to 450 nm. The white light emitting device according to any one of claims 1 to 11, wherein the amount of the second contribution component in the 500 nm range is larger.

  The invention according to claim 13 is the white light emitting device according to claim 12, wherein the amount of the first contribution component is not more than half of the amount of the second contribution component.

The invention according to claim 14 is a white light emitting device comprising: a white light source having an emission spectrum component in a wavelength range of 350 nm to 700 nm; and a spectral component removing means for removing the emission spectrum component in a part of the wavelength region. The spectrum of radiated light from the white light emitting device is quantified using a retinal damage action spectrum and a retinal sensitivity spectrum, and the degree of retinal damage is reduced based on the quantified numerical value. It is a method for forming a white light emitting device, which constitutes a spectral component removing means.

  The invention according to claim 15 is the method for forming a white light emitting device according to claim 14, wherein the white light source is configured based on the quantified numerical value.

  Further, in the invention of claim 16, the quantified numerical value is D × I, where D is the retinal damage action spectrum, S is the retinal sensitivity spectrum, and I is the spectrum of the emitted light. 16. The method of forming a white light emitting device according to claim 14, wherein the ratio is an integral value and an S × I integral value.

  The invention according to claim 17 determines the white color classification of the white light source using chromaticity coordinates defined by one or more types of color matching functions and the retinal damage action spectrum. The method for forming a white light emitting device according to any one of claims 14 to 16, which constitutes the spectral component removing means.

  The invention according to claim 18 is the white light-emitting device according to any one of claims 14 to 17, wherein the quantified numerical value is determined according to an emission direction of light from the white light source. It is a forming method.

  According to the present invention, of the white light emitted from the white light emitting device, the spectral component in a part of the wavelength region that affects the human retina, such as damage, or the spectral component in the whole wavelength region is weakened. Or since the removal means which removes completely is provided, the risk of retinal damage can be reduced. In addition, since the retinal damage spectrum and the retinal sensitivity spectrum are used to configure the white light source and the spectral component removal means, it is possible to prevent a decrease in the amount of radiated light and fluctuations in chromaticity due to the spectral component removal means, Desired white light can be formed while reducing the risk of retinal damage. It is a big effect also about the point which can quantify these.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 shows a cross-sectional structure of a first white light emitting device of the present invention.

  In the white light emitting device of the present invention, the LED 20 is disposed in the recess of the package 21 made of resin or the like, and is filled with the phosphor 22 and the sealing resin up to a certain height of the recess. Further, a white light source is formed with a configuration excluding the filter 23. The phosphor 22 receives light emitted from the LED 20 and emits light having a wavelength longer than the emission wavelength.

  For example, if the LED 20 is configured as a semiconductor light emitting diode that generates near-ultraviolet light or violet light, the phosphor 22 emits three types of phosphors, that is, red (R), blue (G), and green (B). It is composed of a mixture of phosphors, or composed of OYBG (orange yellow, blue, green) or the like. Then, each phosphor is excited by the light from the LED 20 to fluoresce, and these lights are mixed to become white. If the LED 20 is configured as a semiconductor light emitting diode that emits blue light, the phosphor 22 is composed of a YAG phosphor such as YAG: Ce, and the YAG phosphor emits yellow fluorescence that is a complementary color of blue. Therefore, the blue light of the LED 20 and the yellow light of the YAG phosphor 22 are mixed to become white. In this embodiment, the LED 20 is formed as a semiconductor light emitting diode that generates near-ultraviolet light or violet light, and the phosphor 22 is a mixture of three types of R, G, and B phosphors, or OYBG (orange). (Yellow, Blue, Green) etc.

  The LED 20 is formed to have a peak emission wavelength in a range of 350 nm to 420 nm corresponding to near ultraviolet light or violet light. Metal wirings 10 and 12 are formed on the concave surface of the package 21, and the LED 20 is bonded to the bottom surface of the concave portion of the package 21. The metal wiring 12 is connected to the p-electrode of the blue LED 20 by the lead wire 13, while the metal wiring 10 is connected to the n-electrode of the blue LED 20 by the lead wire 11.

  A filter 23 is provided so as to cover the surface of the phosphor filled in the recess of the package 21. The filter 23 corresponds to a spectral component removing unit, and is adapted to weaken or not transmit at least a part of the light emitted from the LED 20 within a wavelength range of 380 nm to less than 480 nm. Further, the intensities of all wavelengths within the range of 380 nm to less than 480 nm may be weakened or cut so as not to be transmitted. Therefore, in the light emitted from the white light emitting device of FIG. 1, the emission spectrum component in at least a part of the wavelength region within the wavelength range of 380 nm to 480 nm is reduced in intensity or removed. In order to achieve the above function, a dielectric multilayer film, colored glass, or the like is used for the filter 23. Further, the resin or phosphor 22 and the filter 23 may be arranged apart from each other.

Next, a specific configuration of the LED 20 is shown in FIG. The LED 20 is composed of a GaN-based semiconductor element made of a GaN (gallium nitride) -based semiconductor. The GaN-based semiconductor element uses a III-V group GaN-based semiconductor, which is a hexagonal compound semiconductor, and the III-V group GaN-based semiconductor is a quaternary mixed crystal Al _x Ga _y In _z N (x + y + z). = 1, 0 ≦ x ≦ 1, 0 ≦ y ≦ 1, 0 ≦ z ≦ 1).

  The LED 20 is formed by sequentially growing a GaN-based semiconductor layer on the GaN single crystal substrate 1. The GaN single crystal substrate 1 is bonded to the bottom surface of the recess of the package 21 serving as a support. As described above, the metal wirings 10 and 12 are formed in the package 21. The n-side electrode 9 and the metal wiring 12 are connected by a bonding wire 13, and the p-side electrode 8 and the metal wiring 10 are connected by a bonding wire 11.

The n-type contact layer formed on the GaN single crystal substrate 1 is composed of an n-type GaN layer 2 to which silicon (Si) is added as an n-type dopant. The layer thickness is preferably 3 μm or more. The doping concentration of silicon is, for example, 10 ¹⁸ cm ⁻³ . The InGaN / GaN active layer 3 stacked on the n-type GaN layer 2 uses an InGaN layer doped with silicon (for example, 3 nm thick) as a well layer, and a GaN layer (for example, 9 nm thick) as a barrier layer. It is composed of a multiple quantum well structure (MQW) that is alternately stacked with a predetermined period (for example, five periods). A GaN final barrier layer 4 (for example, 40 nm thick) is laminated between the InGaN / GaN active layer 3 and the p-type AlGaN layer 5. Further, by changing the In composition ratio of the InGaN well layer of the InGaN / GaN active layer 3, the emission wavelength of the LED 20 can be changed. For example, if the In composition ratio is changed to about 10% or more, the emission wavelength peak can be 350 nm to 420 nm.

The p-type AlGaN layer 5 serves as an electron blocking layer and is composed of an AlGaN layer to which magnesium (Mg) as a p-type dopant is added. The layer thickness is, for example, 28 nm. The magnesium doping concentration is, for example, 3 × 10 ¹⁹ cm ⁻³ .

The p-type contact layer formed on the p-type AlGaN layer 5 is composed of a p-type GaN layer 6 to which magnesium as a p-type dopant is added at a high concentration. The layer thickness is, for example, 70 nm. The doping concentration of magnesium is, for example, 10 ²⁰ cm ⁻³ . The surface of the p-type GaN layer 6 is a mirror surface. More specifically, the unevenness of the surface of the p-type GaN layer 6 is 100 nm or less. This surface is a light extraction side surface from which light generated in the InGaN / GaN active layer 3 is extracted.

  The transparent electrode 7 is composed of a transparent thin metal layer (for example, 200 mm or less) composed of Ni (refractive index 1.8) and Au (refractive index 1.6).

  The n-side electrode 9 is a film composed of a Ti and Al layer. The GaN single crystal substrate 1 is a substrate made of a GaN single crystal having an m-plane as a main surface. More specifically, an m-plane which is a nonpolar plane (nonpolar plane) is a main plane, and this main plane is a plane having an off angle within ± 1 ° from the plane orientation of the nonpolar plane.

  The GaN single crystal substrate 1 having the m-plane as the main surface can be produced by cutting from a GaN single crystal having the c-plane as the main surface, for example. The m-plane of the cut substrate is polished by, for example, a chemical mechanical polishing process, and the orientation error with respect to both the (0001) direction and the (11-20) direction is within ± 1 ° (preferably ± 0.3). (Within °). In this way, a GaN single crystal substrate having the m-plane as the main surface and free from crystal defects such as dislocations and stacking faults can be obtained. There is only an atomic level step on the surface of such a GaN single crystal substrate. A light emitting diode (LED) structure is grown on the GaN single crystal substrate thus obtained by a known MOCVD method.

  The GaN single crystal substrate 1 has an m-plane as a main surface, and a GaN-based semiconductor layer that is a group III nitride semiconductor is stacked by crystal growth on the main surface. Therefore, not only the GaN-based semiconductor layer that is crystal-grown on the m-plane of the GaN single crystal substrate 1 but also the main growth surface of all the GaN-based semiconductor layers up to the uppermost p-type GaN layer 6 is the m-plane. .

  By the way, the crystal structure of a GaN-based semiconductor can be approximated by a hexagonal system, and the plane (the top surface of the hexagonal column) whose normal is the c-axis along the axial direction of the hexagonal column is the c-plane (0001). . In the GaN-based semiconductor, the polarization direction is along the c-axis. For this reason, the c-plane shows different properties on the + c-axis side and the −c-axis side, and is therefore called a polar plane. On the other hand, the side surfaces (column surfaces) of the hexagonal columns are m-planes (10-10), respectively, and the surfaces passing through a pair of ridge lines that are not adjacent to each other are a-planes (11-20). Since these are crystal planes perpendicular to the c-plane and orthogonal to the polarization direction, they are called non-polar planes, that is, nonpolar planes.

As described above, since the crystal growth main surface is a non-polar m-plane, an electric field due to natural polarization or a piezo electric field based on lattice distortion does not occur at the interface of the heterojunction, and a decrease in luminous efficiency is suppressed. Is done. Therefore, in an LED used in second and third white light emitting devices described later, in order to generate a light emission wavelength in a blue-green region or a light emission wavelength in a green region, the InGa well layer In of the InGaN / GaN active layer 3 Even if the composition ratio is increased, the polarization electric field does not become strong and can be used as a good ultraviolet to green light source. Depending on the emission wavelength, the main surface of the GaN single crystal substrate 1 may be the c-plane. Further, as the growth substrate, a dissimilar substrate such as sapphire, SiC, ZnO may be used instead of the GaN single crystal substrate.

  Next, FIG. 3 shows a retinal damage action spectrum D indicating the susceptibility to damage to the human retina and a retinal sensitivity spectrum indicating the sensitivity of the human retina. Here, the vertical axis represents relative sensitivity, the horizontal axis represents wavelength, and B represents a blue color matching function of CIE 1931, which is a kind of retinal sensitivity spectrum. As can be seen from this figure, the wavelength region of 435 nm to 440 nm can be broadly divided into a wavelength region where the contribution of retinal damage is large and a wavelength region where the contribution of feeling blue is large. Based on these two wavelength regions, a wavelength of about 440 nm or less that greatly contributes to retinal damage is cut, and a white light emitting device having desired chromaticity after removing this wavelength is obtained. A quantitative construction method is described below.

  FIG. 4 shows the emission spectrum I of the white light emitting device, the CIE 1931 color matching functions R (red), G (green), B (blue), and the retinal damage action spectrum D. The thick line graph in the figure is the emission spectrum I of the white light emitting device before improvement. As can be seen from the comparison with the spectrum D, the intensity of the emission spectrum component in the vicinity of the peak wavelength of D is also high, which increases the risk of retinal damage. Therefore, in order to reduce the risk of retinal damage and form desired white light, the spectrum of the emitted light from the white light emitting device is quantified using the retinal damage action spectrum D and the retinal sensitivity spectrum, and this quantification is performed. The device is configured based on numerical values.

First, the emission spectrum I in FIG. 4 is expressed as a function of the wavelength λ and is defined as I (λ). D is expressed as a function of D (λ) and wavelength. Of I (λ), D (λ ) to the components that contribute _{I D} first performs a multiplication with ID = I (λ) × D (λ), is determined by integrating the ID. Therefore, I _D = ∫ (ID). Here, ∫ represents an integral symbol.

Next, numerical values called tristimulus values that are also used for the chromaticity coordinates of the chromaticity diagram are obtained. The spectral sensitivity characteristics of three types of color receptors (color sensors) in the human eye are defined as a function of wavelength λ, x (λ), y (λ), and z (λ). These are called CIE1931 color matching functions. Compared with the blue color matching function B, the green color matching function G, and the red color matching function R in FIG. 4, x (λ) = B (λ), y (λ ) = G (λ), z (λ) = R (λ). The amount of light received by the three sensors in the human eye is called tristimulus values X, Y, and Z. The spectral distribution I of the light source is multiplied by the color matching functions x (λ), y (λ), and z (λ). It is obtained by integrating. Therefore, if the symbols in FIG. 4 are used, the tristimulus values X, Y, and Z are expressed by the following equations.
X = ∫I (λ) B (λ), Y = ∫I (λ) G (λ), Z = ∫I (λ) R (λ)
Here, the integration interval is a wavelength of 380 nm to 780 nm.

Next, relative damage R is defined as a quantified evaluation index. Relative damage R means damage to the sensitivity to the eye sensor and is R = I _D / (X + Y + Z). If the relative damage R quantified as described above is small, it can be said that the retinal damage is relatively small compared to the retinal sensitivity. Conversely, if the relative damage is large, it can be said that the retinal damage is relatively large.

FIG. 19 shows a chromaticity diagram based on the CIE 1931 color matching function, and FIG. 20 shows the white rank in the color locus of black body radiation in FIG. By the way, the chromaticity coordinates (CIE _X , CIE _Y ) of the chromaticity diagram are expressed as follows by the tristimulus values X, Y, and Z described above.
CIE _X = X / (X + Y + Z), CIE _Y = Y / (X + Y + Z)
When the emission spectrum I in FIG. 4 is represented by the chromaticity coordinates (CIE _X , CIE _Y ) in FIG. 19, the chromaticity range is (0.31, 0.34), and the relative damage R is 0.21. It was.

  Next, as shown in FIG. 5, the spectral component removing means, that is, the filter 23 in the case of the configuration of FIG. 1, is used to cut all emission spectral components in the wavelength region of about 440 nm or less to obtain the shape of the emission spectrum I1. To do. In this state, as described above, when the relative damage R was obtained, it was 0.16, and the value was decreased, thereby reducing the risk of retinal damage. However, the chromaticity range is (0.33, 0.37), and the chromaticity range is worse than the state of FIG. In order to return this to the state shown in FIG. 4 (0.31, 0.34), the emission spectrum distribution as shown in FIG. 6 is obtained by increasing the emission intensity of the LED 20 and increasing the height of the spectral component. I2 can be obtained.

  The emission spectrum of FIG. 6 shows that the emission spectrum intensity increases from 450 nm to 500 nm as compared with FIG. In addition, since the wavelength of about 440 nm or less is all cut by the filter 23, the risk of retinal damage is not increased as much as in FIG. The chromaticity range in this case was (0.31, 0.34), and the relative damage R was 0.19.

  In this way, the wavelength that contributes to retinal damage is cut, while the emission spectrum component that is lost because the wavelength that contributes to retinal damage is cut in advance, taking into account the blue color matching function, how much human In order to calculate whether or not the sensitivity is lost by the eye and to compensate for this, the white light emitting device has an optimal chromaticity range and a low risk of retinal damage by increasing the light emission intensity of the white light source (LED 20 in the above example). Can be produced.

As described above, the contribution component amount I _B2 at 450 to 500 nm is made larger than the contribution component amount I _{B1 at} 400 to 450 nm so as to optimize the white chromaticity. In order to reduce the risk of retinal damage while optimizing the chromaticity range, the contribution component amount I _B1 at 400 to 450 nm of the emission spectrum component based on the blue color matching function is changed to the contribution component amount I at 450 to 500 nm. It is desirable that it is 50% or less of _B2 . If this method is performed according to the emission angle of the light from the white light source, the quantification can be performed for each emission direction.

  As described above, it is possible to appropriately form the white light source and the spectrum removing means by determining the sensitivity amount of brightness / retinal damage and then calculating the spectrum of the emitted light from the white light emitting device. However, when evaluated with the CIE 1931 color matching function, the CIEx is within the range of 0.3 to 0.5 and the CIEy is within the range of 0.3 to 0.44 in the chromaticity diagram of FIG. It becomes possible to design.

  Next, the spectrum of the emitted light from the white light emitting device is quantified using the retinal damage action spectrum D and the retinal sensitivity spectrum of a different type, and a spectrum removing unit is configured based on the quantified numerical value. A method for forming the spectrum so as to be a white spectrum as designed on the chromaticity diagram will be described below.

  FIG. 13 shows a retinal damage action spectrum D indicating the ease of retinal damage due to the above-described photoexcitation reaction, and a relative visibility spectrum S which is a kind of retinal sensitivity spectrum. The specific visibility spectrum S indicates a wavelength region and relative sensitivity that the human eye feels bright. S is substantially equivalent to a green color matching function, and D and S have almost no overlapping wavelength region, so that a component included in a wavelength region with high S sensitivity is avoided while avoiding a wavelength region with high D sensitivity. To make it larger.

  Optical information coming from the outside passes through the cornea, pupil, lens, and vitreous in order of the human eye and is received by the retina. Visual cells that sense light include rod cells that recognize light and darkness (Rods) and cone cells that recognize color (Cones). However, when lighting is considered, it can be designed in consideration of light adaptation. That is, only the pyramidal cells are considered, and the above-described relative luminous sensitivity spectrum S, which is bright adaptation visual sensitivity, may be used.

For example, assume that there is light emission of an emission spectrum as shown in FIG. Further, D is expressed as a function of D (λ), S is expressed as a function of S (λ) and wavelength. As shown in FIG. 15 (a), the out of I (lambda), component contributing _{I D} to D (lambda) first performs a multiplication with ID = I (λ) × D (λ), integrates the ID Is required. Therefore, it becomes ∫ (ID). Here, ∫ represents an integral symbol. On the other hand, as shown in FIG. 15B, when the component I _S that contributes to S (λ) of I (λ) is obtained, ∫ (IS) is obtained. Here, IS = I (λ) × S (λ). Then, I _S / I _D , that is, the ratio of brightness sensitivity / retinal damage is obtained. If this ratio is large, it is felt bright and retinal damage is relatively small. Conversely, if the ratio value is small, it can be said that the brightness is not felt so much and the retinal damage is relatively large.

As described above, the white light is formed so as to increase the ratio of _IS / _ID . For example, the light emission spectrum of a light bulb is known to be close to the blackbody radiation spectrum, but is as shown in FIG. In this case, ID and IS are as shown in the figure, the component contributing to retinal damage is very small, and the component contributing to brightness sensitivity is large. The ratio of I _S / I _D is 5.5. On the other hand, the emission spectrum of sunlight, which is known to contain a lot of ultraviolet rays, is as shown in FIG. Here, ID and IS are as shown in the figure, and the components contributing to retinal damage are large, and the components contributing to brightness sensitivity are also large. In this case, I _S / I _D is 1.5.

Next, an emission spectrum of a device that uses a blue LED often used as a white LED and a phosphor such as YAG to mix blue light and yellow light into white is shown by I in FIG. Here, ID and IS are as shown in the figure, and components that contribute to retinal damage are not so small, and components that contribute to brightness sensitivity are also normal. In this case, I _S / I _D is 1.9. Since there is still a considerable amount of retinal damage component, the wavelength cut filter 23 is provided as in the present invention to cut the component in at least a part of the wavelength range of 380 to 480 nm. _{The S} / _ID value is improved. On the other hand, by providing the wavelength cut filter 23, it is considered that a part of the emission spectrum component is lost and the chromaticity range is deteriorated. In this case, the emission intensity of the blue LED is increased as described above. It ’s fine.

  The white light emitting device formation method described above can be applied to all white light emitting devices of the present invention, such as a second white light emitting device and a third white light emitting device described later.

  Next, the configuration of white light generated from the white light emitting device will be considered. The first white light emitting device is configured to be excited by light emission in the near-ultraviolet or violet region and mix the three colors of light from the excited phosphor to make white. The white light emitting device 2 generates white light by mixing light emission in the blue-green region and light emission in the red region. FIG. 12 is a relative sensitivity curve with respect to the wavelengths of D, B, G, and R already described. The intersection C1 (about 500 nm) between the B curve and the G curve is a blue color sensor corresponding to a human eye color sensor. Since the sensitivity of the container and the green receptor is the same, a blue-green light source is formed so as to form a spectral distribution in which a peak wavelength exists in the vicinity of this wavelength, and the intersection C2 (about about R2) at the same sensitivity as C1 The red light source is configured so as to form a spectral distribution having a peak wavelength in the vicinity of the wavelength of 650 nm.

  Here, the wavelength near C1 is 500 nm ± 20 nm (480 nm to 520 nm), and the wavelength near C2 is 650 nm ± 20 nm (630 nm to 670 nm). As described above, when a blue-green light source and a red light source are produced and the light outputs of these two light sources are made the same, and simultaneously emitted and mixed, blue and green are received with the same sensitivity at the emission wavelength near C1. In addition, it is possible to sense red with the same sensitivity as blue and green at the emission wavelength near C2, so that the three primary colors of blue, green and red can be felt with the same intensity, and ideal white light can be obtained. Can be formed. In addition, since components that damage the retina having a wavelength of less than 480 nm are weakened, the risk of retinal damage can be suppressed.

  In order to specifically configure the white light emitting device as described above, in the case of the second white light emitting device in which the blue-green and red light sources are configured by light emitting diodes (LEDs), and in the case of the second configured by a laser diode (LD). The case of the white light emitting device 3 will be described.

  First, in the case of the 2nd white light-emitting device comprised with a light emitting diode, the same thing as the structure of FIG. 2 is used independently as blue-green LED. The difference is the In composition ratio in the InGaN well layer of the InGaN / GaN active layer 3. The In composition ratio is made higher than that in the case of blue light emission, and the wavelength is longer than the blue light emission wavelength, for example, 20% or more.

  On the other hand, the red LED is composed of a quaternary mixed crystal AlInGaP-based semiconductor as shown in FIG. The crystal growth of each semiconductor layer is performed by a known metal organic chemical vapor deposition method (MOVPE). On the inclined n-type GaAs substrate 51, an n-type GaAs buffer layer 52, an n-type AlGaInP clad layer 53, an AlGaInP active layer 54, a p-type AlGaInP clad layer 55, and a transparent electrode 56 made of ITO or the like are laminated. An n electrode 58 is formed on the back side of 51, and a p electrode 57 is formed on the transparent electrode 56. As the n-type GaAs substrate 51, a substrate whose crystal orientation is inclined by 10 to 15 degrees from (001) is used.

The AlGaInP active layer 54 is formed of, for example, (Al _0.5 Ga _0.5 ) _0.5 In _0.5 P. The n-type AlGaInP clad layer 53 contains n-type impurity Si and the p-type AlGaInP clad layer 55. Is doped with a p-type impurity Zn. The p-electrode 57 uses a multilayer metal film of Ti and Au, and the n-electrode 58 uses an alloy layer of Au, Ge, and Ni and a multilayer metal film of Ti and Au. Note that by changing the composition ratio of the AlGaInP active layer 54, it is possible to emit light within a wavelength range of 650 nm ± 20 nm (630 nm to 670 nm) in the vicinity of C2.

  By arranging the above two LEDs side by side on a support substrate and causing them to emit light simultaneously, color mixing occurs and white light is obtained.

  Next, the case of a third white light emitting device constituted by a laser diode (LD) will be described. 8 shows a blue-green laser diode configuration, and FIG. 10 shows a perspective view of the blue-green LD of FIG. The blue-green LD is composed of a III-V group GaN-based semiconductor.

  On the GaN single crystal substrate 61, an n-type GaN contact layer 62 (for example, a film thickness of 2 μm), an n-type AlGaN cladding layer 63 (a film thickness of 1.5 μm or less, for example, 1.0 μm thickness), an n-type GaN light guide layer 64 ( For example, an InGaN active layer (light emitting layer) 65 is sequentially formed. Next, as a p-type semiconductor layer, on the active layer 65, a p-type AlGaN electron blocking layer 66 (for example, a film thickness of 20 nm), a p-type GaN light guide layer 67 (for example, a film thickness of 0.1 μm), a p-type AlGaN cladding. A layer 68 (film thickness of 1.5 μm or less, for example, 0.4 μm thickness) and a p-type GaN contact layer 69 (for example, film thickness of 0.05 μm) are sequentially stacked.

  Here, the GaN single crystal substrate 61 has an m-plane as a main surface, and a GaN-based semiconductor layer that is a group III nitride semiconductor is stacked by crystal growth on the main surface. Therefore, not only the GaN-based semiconductor layer that is crystal-grown on the m-plane of the GaN single crystal substrate 61 but also the growth principal surface of all GaN-based semiconductor layers up to the uppermost p-type GaN contact layer 69 is the m-plane. Become.

  As described above, since the crystal growth main surface is a non-polar m-plane, an electric field due to natural polarization or a piezo electric field based on lattice distortion does not occur at the interface of the heterojunction, and a decrease in luminous efficiency is suppressed. Is done. Therefore, even if the In composition ratio in the InGaN active layer 65 is increased in order to generate an oscillation wavelength in the blue-green region, the polarization electric field does not become strong, and it can be used as a good blue-green light source. Of course, depending on the wavelength, conventional c-plane growth can also be used.

The n-type GaN contact layer 62 and the p-type GaN contact layer 69 are low resistance layers for making ohmic contact with the n electrode 72 and the p electrode 71, respectively. The n-type GaN contact layer 62 is a semiconductor layer in which GaN is doped with, for example, 3 × 10 ¹⁸ cm ^{−3 of} n-type dopant Si, and the p-type GaN contact layer 69 is 3% of p-type dopant Mg, for example. It is a semiconductor layer doped with × 10 ¹⁹ cm ⁻³ .

The n-type AlGaN cladding layer 63 and the p-type AlGaN cladding layer 68 produce a light confinement effect that confines light from the active layer 65 between them. The n-type AlGaN cladding layer 63 is a semiconductor layer in which AlGaN is doped with, for example, 3 × 10 ¹⁸ cm ^{−3 of} n-type dopant Si, and the p-type AlGaN cladding layer 68 is, for example, 3 p-type dopant Mg in AlGaN. It is a semiconductor layer doped with × 10 ¹⁹ cm ⁻³ . The AlGaN cladding layers 63 and 68 are produced with an Al composition ratio of 7% or less. The n-type AlGaN cladding layer 63 has a wider band gap than the n-type GaN optical guide layer 64, and the p-type AlGaN cladding layer 68 has a wider band gap than the p-type GaN optical guide layer 67. Thereby, good confinement can be performed, and a low threshold and high efficiency semiconductor laser diode can be realized.

The n-type GaN light guide layer 64 and the p-type GaN light guide layer 67 are semiconductor layers that generate a carrier confinement effect for confining carriers (electrons and holes) in the active layer 65. Thereby, the efficiency of recombination of electrons and holes in the light emitting layer 65 is increased. The n-type GaN light guide layer 64 is a semiconductor layer obtained by doping GaN with, for example, 3 × 10 ¹⁸ cm ^{−3 of} n-type dopant Si, and the p-type GaN light guide layer 67 is GaN with, for example, 5 × 10 p-type dopant Mg. It is a semiconductor layer doped with ¹⁸ cm ⁻³ . The light guide layers 64 and 67 may be made of InGaN containing several percent or less of In.

The p-type AlGaN electron blocking layer 66 doped with, for example, 5 × 10 ¹⁸ cm ^{−3 of the} p-type dopant Mg prevents the outflow of electrons from the active layer 65 and increases the recombination efficiency of electrons and holes. . The p-type AlGaN electron block layer 66 is formed with an Al composition ratio of 5 to 30%.

  The active layer 65 has, for example, an MQW (multipule-quantum well) structure containing InGaN, and light is generated by recombination of electrons and holes, and the generated light is This is a layer for amplification. Specifically, the active layer 65 is configured by repeatedly laminating InGaN well layers (for example, 3 nm thickness) and GaN barrier layers (for example, 9 nm thickness) alternately for about 2 to 7 periods. In this case, when the InGaN well layer has an In composition ratio of 5% or more, the band gap becomes relatively small and a quantum well layer is formed. On the other hand, the GaN barrier layer functions as a barrier layer having a relatively large band gap.

  The emission wavelength is adjusted so as to increase the In composition ratio in the InGaN well layer, so that an oscillation wavelength of 480 nm to 520 nm in the blue-green region can be obtained. For example, it is desirable that the In composition of the InGaN well layer is 20% or more and the InGaN well layer is about 30%. In the MQW structure, the number of quantum wells containing In is preferably 3 or less.

  A part of the p-type semiconductor stacked body from the p-type AlGaN electron blocking layer 66 to the p-type GaN contact layer 69 is removed by mesa etching to form a ridge stripe A. More specifically, the p-type contact layer 69, the p-type AlGaN cladding layer 68, and the p-type GaN light guide layer 67 are partially removed by etching to form a mesa-shaped ridge stripe A. As shown in FIG. 10, the ridge stripe A is formed along the c-axis direction. A pair of end surfaces formed by cleaving at both ends in the longitudinal direction (c-axis direction) of the ridge stripe A are parallel to each other, and both are perpendicular to the c-axis and constitute a c-plane and a −c-plane. A resonator is formed between these end faces, and light generated in the active layer 65 is amplified by stimulated emission while reciprocating between the end faces of the resonator. In order to extract a part of the amplified light from the end face of the resonator as a laser beam to the outside of the device, the following configuration was adopted.

The insulating film 81 formed so as to cover the resonator end face which is the + c plane is made of, for example, a single film of ZrO ₂ . On the other hand, the insulating film 80 formed on the end face of the resonator that is the −c plane is composed of a multiple reflection film in which, for example, SiO ₂ films and ZrO ₂ films are alternately and repeatedly stacked. The ZrO ₂ single film constituting the insulating film 81 has a thickness of λ / 2n ₁ (where λ is the emission wavelength of the active layer 65 and n ₁ is the refractive index of ZrO ₂ ). On the other hand, multiple reflection film constituting the insulating film 80, a thin film lambda / 4n ₂ (where _{n 2} is the refractive index of SiO ₂₎ laminating a SiO ₂ film, a ZrO ₂ film with a thickness of lambda / 4n ₁ alternately It has a structure.

  With such a structure, the reflectance at the end face of the + c plane is small, and the reflectance at the end face of the −c plane is large. More specifically, for example, the reflectance of the end face of the + c plane is about 20%, and the reflectance of the end face of the −c plane is about 99.5% (almost 100%). Therefore, a larger laser output is emitted from the end face of the + c plane. In addition, the laser has a spectral width on the order of nanometers and becomes sharp, so there is no component that greatly contributes to retinal damage, so the risk of retinal damage can be suppressed.

The n electrode 72 is made of, for example, Al metal, and the p electrode 71 is made of, for example, Al metal or a Pd / Au alloy, and is ohmically connected to the p-type contact layer 69 and the GaN single crystal substrate 61, respectively. The exposed surfaces of the p-type GaN light guide layer 67 and the p-type AlGaN cladding layer 68 are covered so that the p-electrode 71 contacts only the p-type GaN contact layer 69 on the top surface (stripe-shaped contact region) of the ridge stripe A. An insulating layer 70 is provided. As a result, current can be concentrated on the ridge stripe A, so that efficient laser oscillation is possible. The insulating layer 70 can be made of an insulating material having a refractive index greater than 1, for example, SiO ₂ or ZrO ₂ .

  Further, the top surface of the ridge stripe A is an m-plane, and a p-electrode 71 is formed on the m-plane. The back surface of the GaN single crystal substrate 61 on which the n-electrode 72 is formed is also m-plane. As described above, since both the p-electrode 71 and the n-electrode 72 are formed on the m-plane, it is possible to realize reliability that can sufficiently withstand high-power laser and high-temperature operation.

  When the semiconductor laser diode of FIG. 8 is manufactured, first, the GaN single crystal substrate 61 having the m-plane as the main surface can be manufactured by cutting from a GaN single crystal having the c-plane as the main surface. The m-plane of the cut substrate is polished by, for example, a chemical mechanical polishing process, and an orientation error with respect to both the (0001) direction and the (11-20) direction is within ± 1 ° (preferably ± 0.3). (Within °). In this way, a GaN single crystal substrate 61 having the m-plane as a main surface and free from crystal defects such as dislocations and stacking faults is obtained.

  Next, a specific structure of the red LD is shown in FIG. The red LD is composed of an AlInGaP-based semiconductor. The crystal growth of each semiconductor layer is performed by a known metal organic chemical vapor deposition method (MOVPE). On the inclined n-type GaAs substrate 32, an n-type AlGaInP cladding layer 33, an AlGaInP light guide layer 34, an MQW active layer 35, an AlGaInP light guide layer 36, a p-type AlGaInP first cladding layer 37, an AlGaInP etching stop layer 38, an n-type An AlGaInP block layer 41, a p-type AlGaAs second cladding layer 39, a p-type GaAs contact layer 40, and a p-electrode 42 are stacked, and an n-electrode 31 is formed on the back side of the n-type GaAs substrate 32. As the n-type GaAs substrate 32, a substrate whose crystal orientation is inclined by 10 to 15 degrees from (001) is used.

The MQW active layer 35 is formed of three GaInP well layers and two undoped (Al _0.5 Ga _0.5 ) _0.5 In _0.5 P barrier layers. The n-type AlGaInP cladding layer 33 is n-type impurity Si-doped (Al _0.7 Ga _0.3 ) _0.5 In _0.5 P, the AlGaInP light guide layer 34 and the AlGaInP light guide layer 36 are undoped (Al _{0. 5} Ga _0.5 ) _0.5 In _0.5 P, p-type AlGaInP first cladding layer 37 is made of p-type impurity Zn-doped (Al _0.7 Ga _0.3 ) _0.5 In _0.5 P, AlGaInP The etching stop layer 38 includes three layers of _unstrained (Al _0.1 Ga _0.9 ) _0.5 In _0.5 P doped with p-type impurity Zn and (Al _0.4 Ga _{0. 6) 0.5} in _0.5 P with two layers layers are alternately stacked, a p-type AlGaAs second cladding layer 39 is p-type impurity Zn-doped _Al 0.5 GaAs, p-type GaAs contactor Layer 40 is p-type impurity Zn-doped GaAs, the n-type AlGaInP block layer 41 is composed of n-type impurity Si-doped _{(Al 0.8 Ga 0.2) 0.5 In} 0.5 P. The p electrode 42 uses a multilayer metal film of Ti and Au, and the n electrode 31 uses an alloy layer of Au, Ge, Ni, and a multilayer metal film of Ti and Au.

  The MQW active layer 35 is sandwiched between AlGaInP light guide layers 34 and 36 from both sides. These light guide layers are formed to confine light in the vertical direction, and the vertical spread angle can be controlled by the composition and thickness of the light guide layer. When this vertical light confinement is weakened, the light emission spot expands in the vertical direction, and the vertical spread angle of the outgoing beam (the size in the FFP stacking direction) is reduced.

  In the high-power red semiconductor laser diode shown in FIG. 9, the p-type AlGaAs second cladding layer 39 and the p-type GaAs contact layer 40 form a striped ridge portion B, and both sides of the ridge portion B are n-type AlGaInP. It has a buried ridge structure covered with a block layer 41. The current does not flow in the n-type AlGaInP block layer 41 that is reverse-biased and the lower portion thereof, but flows in the striped ridge portion B.

  As described above, assuming that the constructed blue-green LD is LD1, and the red LD is LD2, LD1 and LD2 are arranged on the support substrate 90 as shown in FIG. If a light coupler or the like is provided on the surface side (not shown) and laser light is mixed, a white light emitting device with reduced risk of retinal damage can be formed as can be seen from FIG.

  As another example, by using an LED composed of a GaN-based semiconductor and changing the type of phosphor or filter, a white light emitting device, for example, a light source having an emission spectrum peak in the wavelength range of 380 nm to 420 nm, Consists of three types of light sources, having a light emission spectrum peak in the wavelength range of 480 nm to 500 nm, and a light source having a light emission spectrum peak in the wavelength range of 520 nm to 670 nm. It can also be made to remove emission spectral components in the range.

For example, two LEDs composed of the nitride semiconductor of FIG. 2 are used, and the first LED has a wavelength range of 380 nm to 420 nm by changing the In composition ratio in the InGaN well layer of the InGaN / GaN active layer 3. In the second LED, the In composition ratio in the InGaN well layer of the InGaN / GaN active layer 3 is set to be higher than that of the first LED, and the wavelength ranges from 480 nm to 500 nm. Can be formed so as to have an emission spectrum peak. For a light source having an emission spectrum peak in the wavelength range of 520 nm to 670 nm, a structure using a semiconductor light emitting diode and a phosphor as shown in FIG. 1 is used so that light in the wavelength range of 520 nm to 670 nm is emitted. A filter may be provided for the three light sources. The light source having an emission spectrum peak in the wavelength range of 520 nm to 670 nm may be a light emitting diode that outputs red light made of an AlGaInP-based semiconductor.

It is a figure which shows the cross-section of the 1st white light-emitting device of this invention. It is a figure which shows the cross-section of LED used for a 1st white light-emitting device. It is a figure which shows a retinal damage action spectrum and a blue color matching function. It is a figure which shows the white emission spectrum before improvement. It is a figure which shows the white emission spectrum after removing the spectrum component regarding retinal damage. FIG. 6 is a diagram showing a white emission spectrum after increasing the emission intensity from FIG. 5. It is a figure which shows the cross-section of the semiconductor light-emitting diode used for a white light-emitting device. It is a figure which shows the cross-section of the semiconductor laser diode used for a white light-emitting device. It is a figure which shows the cross-section of the semiconductor laser diode used for a white light-emitting device. It is a perspective view of the semiconductor laser diode of FIG. It is a figure which shows the white light-emitting device at the time of using two types of semiconductor laser diodes. It is a figure which shows the wavelength range from which the relative sensitivity of each color matching function of R, G, B is the same. It is a figure which shows a retina action spectrum and a specific luminous efficiency spectrum. It is a figure which shows an example of a white emission spectrum. It is a figure which shows the part which contributes to a retinal action spectrum in the white light emission spectrum component of FIG. 13, and the part which contributes to a relative luminous sensitivity spectrum. It is a figure which shows the part which contributes to a retinal action spectrum with respect to the light emission spectrum of a light bulb, and the part which contributes to a specific luminous efficiency spectrum. It is a figure which shows the part which contributes to a retinal action spectrum with respect to the light emission spectrum of sunlight, and the part which contributes to a specific luminous efficiency spectrum. It is a figure which shows the part which contributes to a retinal action spectrum with respect to the emission spectrum of white LED, and the part which contributes to a specific luminous efficiency spectrum. It is a figure which shows a chromaticity diagram. It is a figure which shows the white classification | category in the chromaticity diagram of FIG. It is a figure which shows a color matching function.

Explanation of symbols

1 GaN single crystal substrate 2 n-type GaN layer 3 InGaN / GaN active layer 4 GaN final barrier layer 5 p-type AlGaN layer 6 p-type GaN layer 7 Transparent electrode 8 p-side electrode 9 n-side electrode 10 Metal wiring 11 Bonding wire 12 Metal Wiring 13 Bonding wire 20 LED
21 Package 22 Phosphor 23 Filter

Claims (18)

A white light source having an emission spectrum component in a wavelength range of 350 nm to 700 nm;
A white light emitting device comprising: spectral component removing means for removing an emission spectral component in at least a part of a wavelength region within a wavelength range of 380 nm to less than 480 nm of the emitted light from the white light source.   The white light emitting device according to claim 1, wherein the spectral component removing unit includes a filter.   The white light source emits light having a longer wavelength than the emitted light by absorbing a part of the emitted light from the semiconductor light emitting diode having an emission spectrum peak in a wavelength range of 350 nm to 420 nm and the semiconductor light emitting diode. The white light-emitting device according to claim 1, wherein the white light-emitting device is a phosphor.   A white light source having an emission spectrum component in a wavelength range of 350 nm to 700 nm includes a semiconductor light emitting diode having emission spectrum peaks in a wavelength range of 480 nm to 520 nm and a wavelength range of 630 nm to 670 nm, respectively. White light emitting device.   5. The white light emitting device according to claim 4, wherein the semiconductor light emitting diode having an emission spectrum peak in the wavelength range of 480 nm to 520 nm is made of a nitride semiconductor having an m-plane as a main surface for crystal growth.   A white light source having an emission spectrum component in a wavelength range of 350 nm to 700 nm includes a semiconductor laser diode having emission spectrum peaks in a wavelength range of 480 nm to 520 nm and a wavelength range of 630 nm to 670 nm. White light emitting device.   7. The white light emitting device according to claim 6, wherein the semiconductor laser diode having an emission spectrum peak in the wavelength range of 480 nm to 520 nm is made of a nitride semiconductor having an m-plane as a main surface for crystal growth.   The white light source has an emission spectrum peak in each of a wavelength range of 380 nm to 420 nm, a wavelength of 480 nm to 500 nm, and a wavelength of 520 nm to 670 nm, and does not include an emission spectrum component in the wavelength range of 420 nm to 480 nm. The white light emitting device according to any one of claims 1 and 2.   An emission source having an emission spectrum peak in the wavelength range of 380 nm to 420 nm, an emission source having an emission spectrum peak in the wavelength range of 480 nm to 500 nm, and an emission source having an emission spectrum peak in the wavelength range of 520 nm to 670 nm, 9. The white light emitting device according to claim 8, comprising a semiconductor light emitting diode.   10. The white light emitting device according to claim 9, wherein the semiconductor light emitting diode having an emission spectrum peak in a wavelength range of 380 nm to 420 nm and a wavelength range of 480 nm to 500 nm is made of a nitride semiconductor.   When the spectrum of the emitted light from the white light emitting device is evaluated by the color matching function of CIE1931, CIEx is in the range of 0.3 to 0.5 in the chromaticity diagram, and CIEy is 0.3 to 0.44. The white light-emitting device according to claim 1, wherein the white light-emitting device is included in the range of 1 to 10.   Of the contribution component amount to the CIE1931 blue color matching function in the spectral component of the emitted light from the light emitting device, the second contribution component amount in the wavelength range of 450 nm to 500 nm is more than the first contribution component amount in the wavelength range of 400 nm to 450 nm. The white light emitting device according to claim 1, wherein the white light emitting device is large.   The white light emitting device according to claim 12, wherein the first contribution component amount is half or less of the second contribution component amount. A white light source having an emission spectrum component in a wavelength range of 350 nm to 700 nm;
A method for forming a white light emitting device comprising a spectral component removing means for removing an emission spectral component in a partial wavelength region,
The spectrum component removing means is configured to quantify the spectrum of the emitted light from the white light emitting device using a retinal damage action spectrum and a retinal sensitivity spectrum, and to reduce the degree of retinal damage based on the quantified numerical value. A method for forming a white light emitting device.   The method for forming a white light emitting device according to claim 14, wherein the white light source is configured based on the quantified numerical value.   The quantified numerical value is a ratio of an integral value of D × I and an integral value of S × I, where D is the retinal damage action spectrum, S is the retinal sensitivity spectrum, and I is the spectrum of the emitted light. The method for forming a white light emitting device according to claim 14, wherein the white light emitting device is formed.   A chromaticity coordinate defined by one or a plurality of types of color matching functions and the retinal damage action spectrum are used to determine a white color classification of the white light source and to constitute the spectral component removal means. The method for forming a white light emitting device according to any one of claims 14 to 16.   The method for forming a white light emitting device according to any one of claims 14 to 17, wherein the quantified numerical value is determined according to an emission direction of light from the white light source.

Images