JP2010016292A - Lighting device and method of manufacturing lighting device - Google Patents

Abstract

An illumination device that is excellent in stability against color variation and in which color stability against a change with time is ensured, and a manufacturing method thereof.
A reflector 31b made of a white resin and containing a semiconductor light emitting element 3 having an opening in an upper surface 31a for emitting light from the semiconductor light emitting element 3, and a reflector 31 is provided. The phosphor-containing resin layer 2 made of a transparent resin and a phosphor that fills the recess 31b and covers the semiconductor light-emitting element 3 and absorbs light emitted from the semiconductor light-emitting element 3 and emits light having a longer wavelength. with weight average particle diameter with the true density of the phosphor is set to 3 g / cm ³ or more 4.7 g / cm ³ or less in the range is the 15μm following range of 7 [mu] m, the phosphor-containing resin layer 2, the upper surface An illumination device is adopted in which 2c is recessed from the upper surface 31a and the amount d of the recess is set in the range of −20 μm to −100 μm from the upper surface 31a.
[Selection] Figure 1

Description

  The present invention relates to a lighting device and a method for manufacturing the same, and more particularly to a lighting device including a semiconductor light emitting element and a method for manufacturing the same.

  In recent years, white LEDs have been adopted as backlights for liquid crystal displays of mobile phones, and the production of white LEDs has dramatically increased. In the future, it seems to be certain that white LEDs will be used for the backlight of liquid crystal displays in applications other than mobile phones.

  By the way, when a white LED is used as a backlight of a liquid crystal display, there is a strict standard for color variation in the entire display surface of the display. If it is defined as a range of color coordinates, it is required to suppress variation of about ± 0.008 for each of X and Y. In order to keep the color variation on the entire display surface of the display within the above-mentioned range, the white LED package also needs to be within the range. Here, the white LED package refers to an illumination device including a semiconductor light emitting element, and includes a lead frame on which a semiconductor light emitting element having a recess made of a reflective surface is placed, and a semiconductor light emitting element disposed on the lead frame. And a phosphor-containing sealing resin formed so as to cover the semiconductor light emitting element.

  If you find a package that deviates from the required range by measuring it with a handler tester, you must remove the package. Thus, the cost of the package will increase by that amount unless all packages are within the required color gamut. In other words, in order to popularize the backlight using white LEDs, it is necessary to reduce the cost of the white LED package, and for that purpose, it is required to reduce the color variation as much as possible in the manufacturing process of the package. .

  There are two types of color variation, and there are cases where the color constellation differs for each package, and color constellation varies depending on the direction of one package. Various proposals have been made to suppress these two types of color variations (Patent Documents 1 to 4).

  Furthermore, if the color changes with time, it will deviate from the standard. Therefore, it is necessary to suppress changes in luminance and color with time as much as possible. There are two main causes of changes over time. These are the discoloration of the plating surface of the resin material, phosphor-containing sealing resin or lead frame constituting the reflector, the decrease in the light emission efficiency of the semiconductor light emitting element itself, and the wavelength shift of the semiconductor light emitting element.

  In Patent Documents 3 and 4, there is the following description in order to prevent the color constellation from being different depending on the direction in one package. It is essential that the thickness of the resin layer containing the phosphor existing from the light emitting element to the light extraction surface of the lighting device is constant. In order to achieve this, a specially shaped package is used, or the sealing resin is divided into two layers and the phosphor-containing layer is used to cover the light emitting element with a uniform thickness. It is necessary to apply a body-containing sealing resin. Therefore, Patent Document 2 proposes the following method mainly for the purpose of preventing the color constellation from varying depending on the direction. That is, when the sealing resin is injected into the white package container, the phosphor partially sinks, and the distance through which the sealing resin light passes from the light emitting chip cannot be made constant. It has been proposed that the layer containing the phosphor is made by another method such as electrophoresis in which particles do not settle, and the thickness of the layer containing the phosphor is made uniform.

  Further, Patent Document 1 proposes preventing emission colors from varying depending on directions in one package in a direction different from the above. That is, by uniformly dispersing the phosphor in the sealing resin, it is possible to prevent the color constellation from varying depending on the direction in one package. Also in Patent Documents 2, 3, and 4, uniform dispersion of the phosphor is assumed in the resin layer containing the phosphor, and the difference from that is not necessarily clear.

Moreover, if the thickness of the phosphor-containing sealing resin layer covering one light emitting element varies, the emission color varies depending on the direction of one package. Furthermore, if the thickness of the layer containing the phosphor differs for each package, the color constellation varies for each package. Although there is no clear description of how to prevent the color constellation from being different for each package, it is a precondition that the thickness of the phosphor-containing encapsulating resin layer covering the light emitting element is the same as much as possible.
JP 2005-277441 A JP 2003-110153 A US Pat. No. 6,576,930 US Pat. No. 7,126,162

  Illumination using LEDs has been developed mainly as a demand for use as a backlight for liquid crystal displays. However, at present, small displays for mobile phones are mainly used, and they are not adopted for large area liquid crystal displays such as monitors for personal computers and large TVs. This is because, as described above, stability against color variation and stability over time are not sufficiently ensured.

  In a liquid crystal display such as a cellular phone, even if the color reproduction range is about 70% in terms of NTCS ratio, there is no practical problem. Therefore, it is common to develop a white color by combining a yellow phosphor and a blue LED. However, in the case of a large liquid crystal television, a color reproducibility of at least 82% is required. In order to achieve this, it is difficult to use a single yellow phosphor, and it is necessary to use a mixture of a green phosphor emitting green light with blue light and a red phosphor emitting red light. As described above, when two types of phosphors are used, unless the phosphors are evenly dispersed, the color coordinates vary depending on the illumination device and the color coordinates vary depending on the direction in the same illumination device. The problem becomes more and more serious. However, since the density and particle size distribution are generally different between the green phosphor and the red phosphor, there is a difference in sedimentation speed, and the color coordinates of the emission color change. Manufacturing technology that minimizes variations in color gamut depending on the direction of a single lighting device and color gamut for each lighting device, and allows for more than 95% to be within ± 0.008 for both X and Y LCD backlight standards. is necessary.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide an illuminating device that is excellent in stability against color variation and in which color stability against temporal changes is ensured, and a method for manufacturing the same. To do.

In order to achieve the above object, the present invention employs the following configuration.
[1] A semiconductor light emitting device having a main light emission peak in a wavelength region of 430 nm or more and 500 nm or less, and a reflector made of a white resin, and a concave portion that houses the semiconductor light emitting device is opened on an upper surface, and the semiconductor light emitting device And a phosphor and a transparent resin which are formed so as to cover the semiconductor light emitting element and fill the recess of the reflector and absorb light emitted from the semiconductor light emitting element to emit longer wavelength light from consisting a phosphor-containing resin layer, the true density of the phosphor is the phosphor mass average particle size of less 15μm or more 7μm in conjunction are 3 g / cm ³ or more 4.7 g / cm ³ or less in the range The upper surface of the phosphor-containing resin layer is recessed from the upper surface, and the amount of the depression is in the range of −20 μm to −100 μm from the upper surface. Lighting apparatus characterized by being constant.
[2] The semiconductor light emitting device includes at least a substrate, a seed layer made of AlN formed on the substrate, and a laminated semiconductor layer mainly composed of GaN formed on the seed layer, The laminated semiconductor layer is formed by laminating an n-type semiconductor layer, a light emitting layer, and a p-type semiconductor layer in this order from the substrate side, and an X-ray rocking curve of the (0002) plane of p-type GaN constituting the p-type semiconductor layer. The half-value width of the X-ray rocking curve of the (10-10) plane is 250 arcsec. Or less, and the half-value width of the illumination device according to [1].
[3] The half width of the X-ray rocking curve of the (0002) plane of the seed layer is 100 arcsec. Or less, and the half width of the X-ray rocking curve of the (10-10) plane is 1.7 degrees or less. [2] The lighting device according to [2].
[4] The lighting device according to any one of [1] to [3], wherein the phosphor-containing resin layer is a silicone resin.
[5] A semiconductor light emitting device having a main light emission peak in a wavelength region of 430 nm or more and 500 nm or less, and a reflector made of white resin, and a recess for housing the semiconductor light emitting device is opened on an upper surface, and the semiconductor light emitting device A reflector that emits light of a predetermined wavelength, a phosphor that fills the recess of the reflector and covers the semiconductor light emitting element, absorbs light emitted from the semiconductor light emitting element, and emits light having a longer wavelength, and transparent a method of manufacturing a lighting device and a phosphor-containing resin layer comprising a resin, the mass average particle diameter of 7μm or more 15μm or less with a true density in the range of 3 g / cm ³ or more 4.7 g / cm ³ or less In addition to preparing a mixed resin paste by mixing a phosphor in the range of uncured transparent resin, the viscosity of the mixed resin paste is 4000 adjusting the range of cP to 15000 cP, covering the semiconductor light emitting element with the mixed resin paste, and denting the liquid surface of the mixed resin paste from the upper surface of the reflector, and reducing the amount of depression from the upper surface − The manufacturing method of the illuminating device characterized by comprising the process set to the range of 20 micrometers--100 micrometers, and the process of hardening the said mixed resin paste and forming a fluorescent substance containing resin layer.
[6] The semiconductor light emitting device includes at least a substrate, a seed layer made of AlN formed on the substrate, and a laminated semiconductor layer mainly made of GaN formed on the seed layer, A laminated semiconductor layer is formed by laminating an n-type semiconductor layer, a light emitting layer, and a p-type semiconductor layer in this order from the substrate side, and an X-ray rocking curve of the (0002) plane of p-type GaN constituting the p-type semiconductor layer The illumination device according to [5], wherein a semiconductor light emitting element having a half width of not more than 60 arcsec. And a half width of an X-ray rocking curve of the (10-10) plane being not more than 250 arcsec. Production method.
[7] The method for manufacturing a lighting device according to any one of [5] or [6], wherein the transparent resin is a silicone resin.

  ADVANTAGE OF THE INVENTION According to this invention, it is excellent in stability with respect to the dispersion | variation in a color, and can provide the illuminating device with which the stability of the color with respect to a temporal change was ensured, and its manufacturing method.

  When the phosphor is uniformly dispersed in the sealing resin and it is housed in a reflector made of a white resin having a reflectance of 90% or more, the light entering the white resin is reflected and returns to the phosphor-containing layer. The light that came out of the light-emitting element finally passes through the phosphor-containing layer and escapes to the light-emitting surface after a complicated movement that goes back and forth as it enters the white resin again. Get out. Therefore, it has been found that the phenomenon that the color constellation varies depending on the direction in one lighting device is independent of the thickness of the phosphor-containing layer in each direction. Furthermore, the color constellation for each lighting device does not depend on the size of the lighting device or the amount of the phosphor-containing sealing resin, and it has been found that the color gamut can be controlled with good reproducibility by controlling only the shape of the light emitting surface. .

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a schematic cross-sectional view illustrating an example of a lighting device according to the present embodiment. 2 is a schematic cross-sectional view showing the semiconductor light-emitting element provided in the lighting device shown in FIG. 1, FIG. 3 is a schematic plan view of the semiconductor light-emitting element, and FIG. 4 shows the configuration of the semiconductor light-emitting element. It is a cross-sectional schematic diagram which shows the laminated semiconductor layer to perform.

  In the present specification, in the notation of (10-10) indicating the orientation plane of the crystal, “−1” means “−1” where a bar should be originally added above “1”. It is a thing.

  The illuminating device 1 shown in FIG. 1 is roughly composed of a lead frame 32 made of a copper alloy, a semiconductor light emitting element 3 placed on the lead frame 32, and a phosphor-containing resin layer 2. The surface of the lead frame 32 is silver-plated for the purpose of improving wire bonding and improving reflection characteristics. A reflector 31 made of white resin is attached to the lead frame 32, and a recess 31 b is opened on the upper surface 31 a of the reflector 31. A part of the lead frame 32 is exposed at the bottom of the recess 31b.

  The lead frame 32 is exposed to the outside of the reflector 31 and serves as a terminal for applying a current to the semiconductor light emitting element 3. When surface mounting is assumed, it is desirable to bend the lead frame 32 and dispose its tip at the bottom of the reflector 31 as shown in FIG.

  On the lead frame 32 exposed in the concave portion 31b of the reflector 31, the semiconductor light emitting element 3 is bonded with a die bond agent made of silicone resin or epoxy resin. The lead frame 32 is divided into two inside the reflector 31, and the semiconductor light emitting element 3 is connected to both via wire bonding.

  The semiconductor light emitting element 3 has an n-type electrode and a p-type electrode, and the n-type electrode and the p-type electrode and the lead frame 32 are connected via bonding wires 33 and 34. The phosphor-containing resin layer 2 is filled in the concave portion 31 b of the reflector 31. The semiconductor light emitting element 3 is covered with the phosphor-containing resin layer 2 and the bonding wires 33 and 24 are protected.

The semiconductor light emitting element 3 has a main light emission peak in a wavelength region of 430 nm or more and 500 nm or less, and as shown in FIG. 2, a substrate 101 and a seed layer 102 made of AlN formed on the substrate 101, It includes at least a laminated semiconductor layer 20 mainly composed of GaN formed on the seed layer 102.
The laminated semiconductor layer 20 is formed by laminating the n-type semiconductor layer 104, the light emitting layer 105, and the p-type semiconductor layer 106 in this order from the substrate 101 side, and the (0002) plane of the p-type GaN constituting the p-type semiconductor layer 106. The half width of the X-ray rocking curve is set to 60 arcsec. Or less, and the half width of the (10-10) plane X-ray rocking curve is set to 250 arcsec.

  Next, in the illumination device 1 shown in FIG. 1, a reflector 31 made of a white resin containing a white pigment such as titania fine particles is attached to the lead frame 32. In the white resin, the white pigment content, particle size, and the like are adjusted so that the reflectance exceeds 90%. In other words, the light absorption rate of the reflector 31 is set to 10% or less. It is preferable to use titania as the white pigment. Titania (titanium oxide) has a higher refractive index than other white pigments and has a low light absorption rate, so that it can be suitably used for the reflector 31 of the present embodiment. Other white pigments, such as aluminum oxide, are not preferred because the brightness is greatly reduced compared to titania. Moreover, since titania generally has a photocatalytic action, the resin is altered. Therefore, it is necessary to perform surface treatment with aluminum hydroxide or the like. If the surface treatment is not complete, the resin constituting the reflector is denatured and absorbs a specific wavelength, so that the color locus changes with time. Therefore, the selection and addition amount of titania as a white pigment is extremely important. In order to increase the reflectivity, it is better to add a large amount of titania, but in order to make a reflector shape by injection molding or the like, the fluidity becomes poor and there is a limit.

  In addition, since there are a plurality of processes that require temperature such as solder reflow in the manufacturing process, a material that sufficiently considers heat resistance is selected for the white resin. The most common resin is PPA (polyphthalamide), but it may be a liquid crystal polymer, an epoxy resin, polystyrene, or the like. The bottom surface shape of the recess 31b of the reflector 31 may be any of a circle, a rectangle, an ellipse, and a polygon. The opening shape on the upper surface side of the recess 31b may be any of a circle, a rectangle, an ellipse, and a polygon, and may be the same as or different from the bottom shape.

  Furthermore, the cross-sectional shape of the inner wall surface of the recess 31a may be an arc shape, an elliptical arc shape, or a parabolic shape. The cross-sectional shape should be selected to have the highest luminous efficiency, and the white resin reflectivity, the type of silver plating on the lead frame, the orientation characteristics of light emission of the semiconductor light-emitting element 3 and so on are complicated. It may be selected as appropriate. The important point of the present invention is that even though complicated factors are involved as described above, the phosphor can be uniformly dispersed, and the phosphor-containing resin layer 2 is injected into the concave portion 31b of the reflector 31 made of white resin. In this case, there is no color constellation variation for each direction in one lighting device, and the color constellation variation for each lighting device is a finding that the shape of the light emitting surface may be managed by the following method.

  Next, the phosphor-containing resin layer 2 filling the concave portion 31b of the reflector 31 includes a phosphor (hereinafter also referred to as phosphor powder) 2a that absorbs light emitted from the semiconductor light emitting element 3 and emits light having a longer wavelength. And a transparent resin 2b containing the phosphor powder 2a in a uniformly dispersed state. The phosphor powder 2a includes a green phosphor that emits green light by absorbing light emitted from the semiconductor light emitting device 2, and a red phosphor that emits red light by absorbing light emitted from the semiconductor light emitting device 2. Yes. Since the red phosphor is also excited by the light of the green phosphor, the movement of the color constellation due to the ratio of the green and red phosphors changes very complexly. The color locus is adjusted depending on the emission wavelength of the semiconductor light emitting element, the ratio of the green and red phosphors, the phosphor concentration, and the shape of the upper surface of the phosphor-containing resin layer 2.

  In the illuminating device 1 of this embodiment, since the three primary colors of blue, green, and red are aligned with the blue light emitted from the semiconductor light emitting element 2, the green light emitted from the green phosphor, and the red light emitted from the red phosphor, White light is emitted from the injected phosphor-containing resin exit surface 2c. When this white color is used as the backlight of the liquid crystal display, the wavelength and half width of the light emission peak of the semiconductor light emitting device 3 and the wavelength and half width of the light emission peak of the phosphor powder 2a are determined. The conditions of the present invention are necessary in consideration of dispersibility after satisfying the excitation characteristics and emission characteristics.

The green phosphor suitably used in the lighting device 1 of the present embodiment is preferably a silicate phosphor (BaSiO ₄ : Eu ²⁺ ), and the red phosphor is preferably a nitride phosphor (CaAlSiN ₃ : Eu ²⁺ ). These phosphors have a relatively low true density of 3.5 to 4.7 g / cm ³ and can produce a powder having an average particle diameter of about 10 μm in mass average, which is suitable for the purpose of the present invention. Yes.

Since the green phosphor (BaSiO ₄ : Eu ²⁺ ) has an excitation wavelength of 380 to 440 nm and an emission wavelength of 508 nm, both the excitation wavelength and the emission wavelength are too short to be used for the purpose of the present invention. By replacing with other alkaline earth elements such as Mg, the excitation wavelength and the emission wavelength can be shifted to the longer wavelength side. In the present invention, the excitation wavelength is preferably adjusted to 455 nm and the emission wavelength is set to 528 nm.

The red phosphor (CaAlSiN ₃ : Eu ²⁺ ) has an excitation wavelength of 400 to 500 nm and an emission wavelength of 640 nm. The full width at half maximum of the peak wavelength varies depending on the purpose of use. For illumination, the peak wavelength waveform should be as broad as possible, but for LCD TV backlights, the waveform should be as sharp as possible. In order to increase the color rendering (the degree that looks the same as the color seen under sunlight) for lighting, it is necessary to have the same wavelength distribution as that of sunlight, and all wavelengths are included to the same extent. There is a need. When used for the backlight of a liquid crystal television, the color reproduction range needs to be as wide as possible on the blue, green and red color constellations. That is, it is important that the color purity is as high as possible. For this purpose, it is preferable to make the wavelength distribution of the three primary colors as sharp as possible.

  The phosphor-containing resin layer 2 is formed by pouring into the recess 31b with a dispenser and then curing by heat treatment. If the same volume as the recess 31b is injected, the upper surface becomes flat. Actually, it does not become completely flat but becomes convex or concave. If it becomes convex, the lighting devices will adhere to each other through the phosphor-containing resin layer 2 when sent by a parts feeder or the like. Therefore, it is necessary to make the upper surface concave, but the color locus moves depending on the degree of the depression. Therefore, it is desirable to define the degree of depression.

  The phosphor-containing resin layer 2 is provided with an emission surface 2c that emits white light. In FIG. 1, an emission surface 2 c is formed on the upper surface 31 b side of the reflector 31.

In the illuminating device 1 of this embodiment, as shown in FIG. 1, the output surface 2c is dented from the upper surface 31a, and the amount of the dent is set in the range of −20 μm to −100 μm from the upper surface. As shown in FIG. 1, the dent amount d is the difference between the height of the upper surface 31a and the minimum height of the exit surface 2c. In this specification, when the height of the upper surface 31a is 0 μm, the semiconductor light emitting element 3 side is minus (−). Accordingly, the indentation d is in the range of −20 μm to −100 μm from the upper surface when the height of the upper surface 31a is 0 μm and the minimum height of the emission surface 2c is in the range of 20 to 100 μm from the upper surface. It is located on the 2 side.
In the case of FIG. 1, the dent d is set virtually by setting an extension surface L of the upper surface 31 a of the reflector 31, and the center point of the phosphor-containing resin layer 2 injected from the extension surface L into the recess 31 b is A. It is obtained from the distance between A and the extended surface L. Actual measurement can be performed with a laser displacement meter (for example, ZSHLD2 manufactured by OMRON). What is necessary is just to measure the distance from the midpoint of the straight line connecting the end portions B and B of the phosphor-containing resin layer 2 to the center point A of the phosphor-containing resin layer 2.

The phosphor powder 2a contained in the phosphor-containing resin layer 2, along with its true density is in the range of 3 g / cm ³ or more 4.7 g / cm ³ or less, 15 [mu] m or less the weight-average particle diameter of 7μm or more It is considered as a range. The transparent resin 2b is preferably a silicone resin from the viewpoint of heat resistance.

  Hereinafter, the reason for limiting the amount d of the indented phosphor-containing resin layer, the true density of the phosphor powder 2a, and the mass average particle diameter will be described.

  When a plurality of illumination devices according to the present embodiment are used as a backlight of a liquid crystal display device, the color coordinates (X, It is necessary to keep the variation of Y) within a range of ± 0.008. In addition, if the illuminating device continues to be lit, if the luminescent color changes over time, the color coordinates of the luminescent color will vary during use, which must be suppressed at the same time.

  In the illumination device 1 of the present embodiment, it has been found that the dispersion from the target color coordinates can be suppressed by uniformly dispersing the phosphor powder in the transparent resin. The factors that determine the white color chart are as follows. That is, three types are considered: the concentration of the phosphor powder in the phosphor-containing resin layer, the ratio of the green phosphor and the red phosphor, and the shape of the upper surface of the injected phosphor-containing resin. Therefore, the cause of the variation in the luminescent color of the lighting device 1 is that the concentration differs depending on the location, the ratio of the green and red phosphors is shifted depending on the location, or the shape of the upper surface of the phosphor-containing resin injected by the package varies. The color chart will move. As long as the phosphor sinks after injection, concentration variation and ratio variation will occur. Therefore, the inventors have found out that it is most important to thoroughly pursue uniform dispersion of the phosphor powder in order to stabilize the color locus. In addition, when uniform dispersion is achieved, the color locus moves greatly depending on the shape of the upper surface of the phosphor-containing resin, and if this shape management cannot be achieved at the same time, dispersion may be increased due to uniform dispersion. It is possible.

  If the phosphor powder 2a sinks before the transparent resin 2b is cured, it is impossible to control all the sinking methods for each phosphor-containing resin layer 2 in the same manner. Therefore, in order for the illuminating device 1 to show the emission color of the same color coordinate, it is desirable that the green phosphor and the red phosphor are dispersed as uniformly as possible in the transparent resin 2b.

The uncured transparent resin 2b containing the phosphor powder 2a can be stirred until it is potted into the recess 31b of the reflector 31. However, after potting, the uncured transparent resin 2b is heated and heated. Generally, it takes about 1 hour to cure. During this time, the phosphor powder 2a may settle. By the way, the sedimentation velocity vs of the phosphor powder 2a in the uncured transparent resin 2b is in accordance with the Stokes equation represented by the following equation (1). In the formula (1), Dp is the mass average particle size of the phosphor powder 2a, the [rho _p There the true density of the phosphor powder 2a, the [rho _f is the density of the transparent resin. g is the acceleration of gravity, and η is the viscosity of the uncured transparent resin 2b.

^{_{vs = D p 2 (ρ p}} -ρ f) g / 18η ... (1)

  Here, in order to reduce the sedimentation rate so that the phosphor powder 2a does not settle, it is necessary to increase the viscosity of the transparent resin 2b before curing, but if the viscosity is too high, it is possible to perform potting quantitatively. Disappear.

  Moreover, if the mass average particle diameter of the phosphor powder 2a is reduced, the sedimentation speed is lowered. However, if the mass average particle diameter is too small, it becomes difficult to maintain the light emission characteristics of the phosphor powder 2a.

Furthermore, if the true density of the phosphor powder 2a is the same as the true density of the transparent resin, the phosphor powder 2a does not settle, but the true density of the uncured transparent resin 2b is 1.4 to 1.8 g / cm. ^On the other hand, since most phosphor powders 2a contain heavy metals, the true density is much higher than 2.0 g / cm ³ .

  As described above, only one of the three characteristic values of the mass average particle diameter of the phosphor powder 2a, the true density of the phosphor powder 2a, and the viscosity of the uncured transparent resin 2b prevents sedimentation. The conclusion is that it is difficult. In other words, a transparent resin with the maximum viscosity is used as long as it can be injected quantitatively by potting, and the smallest particle size is selected as long as the light emission characteristics can be maintained at the highest level. It is necessary to select a phosphor with a low density.

  When the thickness of the phosphor-containing resin layer 2 of the lighting device 1 shown in FIG. 1 is about 0.5 mm, the phosphor powder 2a sinks 0.5 mm during one hour when the uncured transparent resin 2b is cured. All of the phosphor powder 2a settles on the bottom of the recess 31b. In order to suppress the variation in color constellation, it is necessary to set the settling distance in 1 hour to 0.05 mm or less in order to suppress the settling distance within a negligible range. In order to enter the settling distance, it is necessary to put the mass average particle diameter of the phosphor, the true density of the phosphor, and the viscosity of the uncured sealing resin within the above ranges.

  In addition, the amount of the phosphor-containing resin layer 2 is a cause of causing variation in the color coordinates of the emission color. In the case of a so-called top view package as shown in FIG. 1, the shape of the upper surface of the injected phosphor-containing resin moves depending on the amount of the phosphor-containing resin layer 2 as shown in FIG. There are cases where the upper surface 31a of the reflector 31 swells in a convex shape or a concave shape. The amount of phosphor powder that works more effectively when the injected phosphor-containing resin swells in a convex manner increases the amount of phosphor powder. Therefore, the same action as when the concentration of the phosphor powder is increased, and the color coordinates of the emission color are X , Y shift to the larger one. However, if it bulges convexly, it will stick to each other when transported by a parts feeder or the like, making handling extremely difficult. Therefore, it must be avoided to become convex. Therefore, as shown in FIG. 1, it is necessary to align the injected phosphor-containing resin into a concave shape. Even in that case, the color coordinate moves depending on the degree. When the inventors measured the correlation, it was found that the Y value moved by 0.01 when the pull-in amount moved by 100 μm. In the case of a backlight for a liquid crystal television, since X and Y are required to be within ± 0.008 respectively, the dent amount d of the injected phosphor-containing resin layer must be controlled to −20 to −100 μm. I must.

From the above, the true density of the phosphor powder 2a contained in the phosphor-containing resin layer 2 with a 3 g / cm ³ or more 4.7 g / cm ³ or less of the range, 7 [mu] m or more 15μm or less a weight average particle diameter It is preferable to set it as the range. Moreover, it is preferable that the recessed amount d of the injected phosphor-containing resin layer 2 is set to −20 to −100 μm.

Next, a semiconductor light emitting element 3 suitable for use in the illumination device 1 of the present embodiment will be described.
2 to 4, the semiconductor light emitting device 3 according to this embodiment includes a substrate 101, a laminated semiconductor layer 20 including a light emitting layer 105 laminated on the substrate 101, and an upper surface of the laminated semiconductor layer 20. A laminated translucent electrode 109 and a p-type bonding pad electrode 107 laminated on the translucent electrode 109 are provided. The semiconductor light emitting device 3 of the present embodiment is a face-up mount type light emitting device that extracts light from the light emitting layer 105 from the side on which the p-type bonding pad electrode 107 is formed.

As shown in FIG. 2, the stacked semiconductor layer 20 is configured by stacking a plurality of semiconductor layers. More specifically, the laminated semiconductor layer 20 is configured by laminating an n-type semiconductor layer 104, a light emitting layer 105, and a p-type semiconductor layer 106 in this order from the substrate side. Part of the p-type semiconductor layer 106 and the light emitting layer 105 is removed by means such as etching, and a part of the n-type semiconductor layer is exposed from the removed part. An n-type bonding pad electrode 108 (hereinafter referred to as n-type electrode 108) is stacked on the exposed surface 104c of the n-type semiconductor layer.
A translucent electrode 109 and a p-type bonding pad electrode 107 are stacked on the upper surface 106 a of the p-type semiconductor layer 106. The translucent electrode 109 and the p-type bonding pad electrode 107 constitute a p-type electrode 111.

  In the semiconductor light emitting device 3 of the present embodiment, a main light emission peak is present in the wavelength region of 430 nm to 500 nm from the light emitting layer 105 by passing a current between the p-type bonding pad electrode 107 and the n-type electrode 108. Almost blue light can be emitted.

  The n-type semiconductor layer 104, the light emitting layer 105, and the p-type semiconductor layer 106 are preferably composed mainly of a compound semiconductor, preferably composed mainly of a group III nitride semiconductor, and more preferably composed mainly of GaN.

  The translucent electrode 109 stacked on the p-type semiconductor layer 106 preferably has a small contact resistance with the p-type semiconductor layer 106. In addition, since the light from the light emitting layer 105 is taken out to the side where the p-type bonding pad electrode 107 is formed, the light-transmitting electrode 109 is preferably excellent in light transmittance. Further, in order to uniformly diffuse the current over the entire surface of the p-type semiconductor layer 106, the translucent electrode 109 preferably has excellent conductivity.

From the above, as the translucent electrode 109, ITO (indium tin oxide (In ₂ O ₃ —SnO ₂ )), IZO (indium zinc oxide (In ₂ O ₃ —ZnO)), AZO (aluminum zinc oxide ( ZnO—Al ₂ O ₃ )), GZO (gallium zinc oxide (ZnO—Ga ₂ O ₃ )), fluorine-doped tin oxide, titanium oxide and the like are preferable. By providing these materials by conventional means well known in this technical field, the translucent electrode 109 can be formed.

  Moreover, the structure of the translucent electrode 109 can be used without any limitation including any conventionally known structure. The translucent electrode 109 may be formed so as to cover almost the entire upper surface 106a of the p-type semiconductor layer 106, or may be formed in a lattice shape or a tree shape with a gap. After forming the translucent electrode 109, thermal annealing may be performed for the purpose of alloying or transparency, but it may not be performed.

  Next, the p-type bonding pad electrode 107 is preferably excellent in both adhesion to the transparent electrode and adhesion to the bonding wire. Therefore, the adhesiveness with the transparent electrode requires a metal such as Ti or Cr in which the oxide of the element is stable, and the adhesiveness with the gold wire is preferably Au with the highest possible purity. Since the transparent electrode and Au do not form a bond at all, they peel off when Au diffuses and reaches the transparent electrode / pad interface. When Ti or Cr diffuses to the Au side, Au becomes hard and wire bonding is difficult. Therefore, a diffusion preventing layer such as Pt or Rh is generally sandwiched between about 1800 mm. Therefore, a typical pad configuration is Ti / Pt / Au.

  The n-type electrode 108 also serves as a bonding pad, and is in contact with the n-contact layer of the n-type semiconductor layer 104 constituting the laminated semiconductor layer 20. Therefore, as shown in FIGS. 2 and 3, the n-type electrode 108 is exposed by removing a part of the p-type semiconductor layer 106, the light emitting layer 105, and the n-type semiconductor layer 104 to expose the n-contact layer. The surface 104c is formed in a substantially circular shape. There are cases where it is directly attached to GaN and cases where it is attached on a transparent conductive positive electrode such as ITO, IZO, etc. However, since Ti, Cr, etc. are in ohmic contact with each other, the material of the n-type electrode 108 is the p-type bonding pad electrode 107. Same as Different pads may be formed by forming pads separately.

Next, the substrate and the laminated semiconductor layer 20 constituting the semiconductor light emitting element 3 will be described.
(substrate)
Although it does not specifically limit as the board | substrate 101, It is preferable to use the sapphire board | substrate which uses c surface as a main surface.

(Seed layer)
The group III nitride semiconductor crystal forming the seed layer 102 preferably has a single crystal structure. A single crystal is a crystal having no grain boundary, and is a crystal having the same crystal orientation in all parts. However, as long as it is not a perfect crystal, some defect exists, and the crystal orientation slightly changes in the crystal depending on the arrangement of the defect. The change in crystal orientation can be grasped by the diffraction peak width of the (0002) plane. The fact that the diffraction peaks are sufficiently sharp means that the planes with no gaps are arranged at a constant plane interval. Next, a measure of whether or not the seed layer 102 faces in the same direction is the rocking curve sharpness (FWMH). If this is disturbed, it may grow in an arbitrary direction, and a smooth surface cannot be secured. In addition, the (10-10) plane of the group III nitride semiconductor crystal shows a rocking curve peak as an indicator of how many locations are partially rotated when viewed from a direction perpendicular to the substrate surface. The full width at half maximum. If this worsens, a defect penetrating in the C-axis direction will be created, and this will affect the light emission characteristics of the LED.

  From the above, in the seed layer 102 according to the present embodiment, the half-value width of the (0002) plane X-ray rocking curve is 100 arcsec. Or less, and the half-width of the (10-10) plane X-ray rocking curve is 1. It is preferable that it is 7 degrees or less. When these values serving as an index of crystallinity are out of the above range, the crystallinity of the GaN single crystal grown on the seed layer 102 is lowered, and the crystallinity of the stacked semiconductor layer 20 including the light emitting layer 105 is decreased. Is lowered, the light emission characteristics are deteriorated, the luminance of the semiconductor light emitting element 3 itself is deteriorated, and the wavelength is shifted.

(Underlayer)
Next, as the underlayer 103, Al _x Ga _y In _z N (0 ≦ x ≦ 1, 0 ≦ y ≦ 1, 0 ≦ z ≦ 1, x + y + z = 1) can be given. This is preferable because a good base layer 103 can be formed.
The film thickness of the base layer 103 is preferably in the range of 0.5 μm to 20 μm. If the thickness is 0.5 μm or more, a GaN layer with good crystallinity is easily obtained. Moreover, even if the film thickness exceeds 20 μm, there is no change in the function, and the processing time is simply extended unnecessarily. A preferable film thickness of the underlayer 103 is in the range of 1 μm to 15 μm.

  In order to improve the crystallinity of the base layer 103, the base layer 103 is preferably not doped with impurities. However, when p-type or n-type conductivity is required, acceptor impurities or donor impurities can be added.

(N-type semiconductor layer)
The n-type semiconductor layer 104 is generally preferably composed of an n-contact layer 104a and an n-cladding layer 104b. The n contact layer 104a can also serve as the n clad layer 104b. Further, the above-described base layer 103 may be included in the n-type semiconductor layer 104.

The n contact layer 104a is a layer for providing an n-type electrode. The n contact layer 104a is preferably composed of a GaN layer. The n contact layer 104a is preferably doped with an n-type impurity, and the n-type impurity is preferably 1 × 10 ¹⁷ to 1 × 10 ²⁰ / cm ³ , preferably 1 × 10 ¹⁸ to 1 × 10 ¹⁹ / cm. A concentration of ³ is preferable in terms of maintaining good ohmic contact with the n-type electrode, suppressing crack generation, and maintaining good crystallinity. Although it does not specifically limit as an n-type impurity, For example, Si, Ge, Sn, etc. are mentioned, Preferably Si and Ge are mentioned.

  The thickness of the n contact layer 104a is preferably 0.5 to 5 μm, and more preferably set to a range of 1 to 3 μm. When the film thickness of the n-contact layer 104a is in the above range, the semiconductor crystallinity is maintained well.

  An n-clad layer 104b is preferably provided between the n-contact layer 104a and the light-emitting layer 105. The n-clad layer 104b is a layer that injects carriers into the light emitting layer 105 and confines carriers. The n-clad layer 104b can be formed of AlGaN, GaN, GaInN, or the like. Alternatively, a heterojunction of these structures or a superlattice structure in which a plurality of layers are stacked may be used. Needless to say, when the n-cladding layer 104b is formed of GaInN, it is desirable to make it larger than the band gap of GaInN of the light emitting layer 105.

The film thickness of the n-clad layer 104b is not particularly limited, but is preferably 0.005 to 0.5 μm, and more preferably 0.005 to 0.1 μm. The n-type doping concentration of the n-clad layer 104b is preferably 1 × 10 ¹⁷ to 1 × 10 ²⁰ / cm ³ , more preferably 1 × 10 ¹⁸ to 1 × 10 ¹⁹ / cm ³ . A doping concentration within this range is preferable in terms of maintaining good crystallinity and reducing the operating voltage of the device.

By providing the n-clad layer 104b, it is possible to provide effects such as electron supply to the active layer and relaxation of the lattice constant difference. In the present embodiment, the n-clad layer 104b is, for example, In _0.1 Ga _0.9 N with a thickness of 20 nm having a doping concentration of 1 × 10 ¹⁸ / cm ³ . Alternatively, a heterojunction of these structures or a superlattice structure in which a plurality of layers are stacked may be used. In the case where the n-type cladding layer 104b is made of GaInN, it is desirable to make it lower than the In concentration of GaInN in the well layer of the light emitting layer 105.

(Light emitting layer)
As the light emitting layer 105 stacked on the n-type semiconductor layer 104, there is a light emitting layer 105 having a single quantum well structure or a multiple quantum well structure. As the light emitting layer 105 having the quantum well structure, as shown in FIG. 4, a barrier layer 105a made of a group III nitride compound semiconductor and a well layer 105b made of a group III nitride compound semiconductor containing indium are alternately repeated. A barrier layer 105a is disposed on the n-type semiconductor layer 104 side and the p-type semiconductor layer 106 side. In the example shown in FIG. 4, the light emitting layer 105 includes six barrier layers 105 a and five well layers 105 b that are alternately stacked, and the barrier layers 105 a are arranged on the uppermost layer and the lowermost layer of the light emitting layer 105. The multi-quantum well configuration is such that a well layer 105b is disposed between the barrier layers 105a.

As the well layer 105b, a group III nitride semiconductor layer made of Ga _1-y In _y N (0 <y <0.4) is usually used. The film thickness of the well layer 105b can be set to a film thickness that provides a quantum effect, for example, 1 to 10 nm, and preferably 2 to 6 nm, from the viewpoint of light emission output.
The barrier layer 105a needs to be thick enough to allow electrons to pass through the tunnel effect, and a layer made of GaN can be used. A thickness of about 4 to 30 nm is appropriate. The well layer 105b and the barrier layer 105a may or may not be doped with impurities by design.

(P-type semiconductor layer)
The p-type semiconductor layer 106 is generally composed of a p-cladding layer 106a and a p-contact layer 106b. The p contact layer 106b can also serve as the p clad layer 106a.

The p-cladding layer 106a is a layer for confining carriers in the light emitting layer 105 and injecting carriers. The p-cladding layer 106a is not particularly limited as long as it has a composition larger than the band gap energy of the light-emitting layer 105 and can confine carriers in the light-emitting layer 105, but is preferably Al _x Ga _1-x N. (0 <x ≦ 0.4). When the p-clad layer 106a is made of such AlGaN, it is preferable in terms of confining carriers in the light-emitting layer. The thickness of the p-clad layer 106a is not particularly limited, but is preferably 1 to 400 nm, more preferably 5 to 100 nm. The p-type doping concentration of the p-clad layer 106a is preferably 1 × 10 ¹⁸ to 1 × 10 ²¹ / cm ³ , more preferably 1 × 10 ¹⁹ to 1 × 10 ²⁰ / cm ³ . When the p-type dope concentration is in the above range, a good p-type crystal can be obtained without reducing the crystallinity.
The p-clad layer 106a may have a superlattice structure in which a plurality of layers are stacked.

The p contact layer 106b is a layer for providing a positive electrode. The p contact layer 106b is preferably Al _x Ga _1-x N (0 ≦ x ≦ 0.4). When the Al composition is in the above range, it is preferable in terms of maintaining good crystallinity and good ohmic contact with the p ohmic electrode. When a p-type impurity (dopant) is contained at a concentration of 1 × 10 ¹⁸ to 1 × 10 ²¹ / cm ³ , preferably 5 × 10 ^{19 to} 5 × 10 ²⁰ / cm ³ , good ohmic contact can be obtained. It is preferable in terms of maintenance, prevention of crack generation, and good crystallinity. Although it does not specifically limit as a p-type impurity, For example, Preferably Mg is mentioned. The thickness of the p contact layer 106b is not particularly limited, but is preferably 0.01 to 0.5 μm, and more preferably 0.05 to 0.2 μm. When the film thickness of the p-contact layer 106b is within this range, it is preferable in terms of light emission output.

  The p-contact layer 106b has a half-value width of the X-ray rocking curve of the (0002) plane of the p-type GaN constituting the p-contact layer 106b of 60 arcsec. Or less, and the X-ray rocking of the (10-10) plane. The half width of the curve is preferably 250 arcsec. Or less. The crystallinity of the p contact layer 106b, which is the uppermost layer of the laminated semiconductor layer 20, is affected by the crystallinity of the seed layer 102 and the underlayer 103, and when the crystallinity of the entire laminated semiconductor layer 20 is improved, the semiconductor light emitting device 3 itself Therefore, the wave wavelength shift is less likely to occur.

Next, a method for manufacturing the semiconductor light emitting element 3 will be described.
In order to manufacture the semiconductor light emitting device 3, first, a substrate 101 such as a sapphire substrate is prepared, and then a seed layer 102 is laminated on the upper surface of the substrate 101.

The seed layer 102 is preferably formed after pretreatment of the substrate 101.
The pretreatment can be performed, for example, by placing the substrate 101 in a chamber of a sputtering apparatus and generating a plasma with power applied between the substrate and the chamber before forming the seed layer 102. Specifically, pretreatment for cleaning the upper surface may be performed by exposing the substrate 101 to Ar or N ₂ plasma in the chamber.

Then, a seed layer 102 is formed on the substrate 101 by sputtering. When the seed layer 102 having a single crystal structure is formed by sputtering, the ratio of the nitrogen flow rate to the nitrogen source flow rate in the chamber and the flow rate of the inert gas is 20% to 90%, preferably 75%. It is desirable to do so. When the flow rate ratio is less than this, the sputtered metal adheres as it is, while when the flow rate ratio is higher than this, the amount of inert gas is small and the sputtering rate is reduced. As the nitrogen raw material, a generally known compound can be used without any problem, but in particular, when nitrogen is used as the raw material, an apparatus is simple and it is difficult to obtain a high reaction rate. However, N ₂ is used in this embodiment. Even with N ₂ , a usable film forming speed can be obtained, and it is the most suitable nitrogen source in view of the balance with the apparatus cost.

  In the present embodiment, among the sputtering methods, it is preferable to employ an RF (radio frequency) sputtering method in which the target surface is unlikely to be charged up and the deposition rate is stable. The substrate temperature during the formation of the seed layer 102 by sputtering is preferably set to 300 to 800 ° C. Further, Al is preferably used as a sputtering target. Furthermore, the pressure in the chamber may be 0.3 Pa or higher. At pressures lower than this, it is difficult to stabilize the discharge. The upper limit of the pressure is not particularly defined as long as plasma can stably exist.

Introducing an inert gas such as Ar and N _{2 using} Al as a target, adjusting the pressure to 0.1 to 10 Pa, and applying a voltage between the target and the chamber from the RF power source causes discharge, and Al atoms are generated. Be knocked out of the target. When Al and N atoms reach the surface of the sapphire substrate, they get a very large kinetic energy from the plasma and can move a considerable range. Since AlN has a very large grain boundary energy, a single crystal having no grain boundary is the most stable structure. Therefore, as long as atoms can move, it will become a single crystal without grain boundaries.

  Since the GaN structure is different from the sapphire structure, a GaN single crystal cannot be grown on the sapphire substrate as it is. On the other hand, a GaN single crystal substrate is generally obtained by inserting a GaN layer called a low temperature buffer. However, since the crystallinity of the low-temperature buffer itself is not good, the GaN crystallinity when using the low-temperature buffer is naturally limited, and the half-value width of the rocking curve on the (0002) plane is 100 arcsec., And on the (100) plane is 300 arcsec. The degree is the limit.

The reason for using AlN for the seed layer 102 is different from the low-temperature buffer layer. It is very natural that AlN is sandwiched between aluminum oxide (Al ₂ O ₃ ), which is sapphire, and GaN for the following reason.
If aluminum and oxygen are replaced with gallium and nitrogen at the same time, the change is too great. First, oxygen is replaced with nitrogen to make AlN, and Al in AlN is changed to Ga to make GaN. Therefore, taking an Al ₂ O ₃ / AlN / GaN structure means that AlN is used as an adhesive agent between Al ₂ O ₃ and GaN, and there is no further intermediate layer.

  As described above, AlN is optimal as a seed layer for connecting sapphire and GaN, and sputtering is optimal as a method for forming the AlN as a single crystal film having high crystallinity. Therefore, a GaN layer having very high crystallinity can be formed by depositing AlN by sputtering. As a result, it is possible to obtain a p-type semiconductor layer having a rocking curve half width of 60 arcsec. Or less on the (0002) plane and 250 arcsec. Or less on the (100) plane. By using a light-emitting element having a p-type semiconductor layer with high crystallinity, there is no luminance degradation and no wavelength shift, and almost no change with time occurs. Thus, by configuring the lighting device 1 in which the phosphors are uniformly dispersed, it is possible to control the variation in the color coordinates of the emission color, including the initial and in-use.

Next, a single crystal base layer 103 is formed on the seed layer 102. The base layer 103 is preferably formed using a sputtering method. When the sputtering method is used, the apparatus can have a simple configuration as compared with the MOCVD method, the MBE method, or the like. When forming the underlayer 103 by sputtering, it is preferable to use a reactive sputtering method in which a group V material such as nitrogen is circulated in the reactor.
In general, in the sputtering method, the higher the purity of the target material, the better the film quality such as crystallinity of the thin film after film formation. When the underlayer 103 is formed by sputtering, it is possible to use a group III nitride semiconductor as a target material as a raw material and perform sputtering by plasma of an inert gas such as Ar gas. The group III metal alone and the mixture thereof used as the target material in can be highly purified as compared with the group III nitride semiconductor. For this reason, in the reactive sputtering method, the crystallinity of the underlying layer 103 to be formed can be further improved.

  The temperature of the substrate 101 when the base layer 103 is formed, that is, the growth temperature of the base layer 103 is preferably 800 ° C. or higher, more preferably 900 ° C. or higher, and 1000 ° C. or higher. Most preferably. This is because atom migration is likely to occur by increasing the temperature of the substrate 101 when the base layer 103 is formed, and dislocations can be driven out of the crystal. In addition, the temperature of the substrate 101 when the base layer 103 is formed needs to be lower than the temperature at which the crystal is decomposed, and is preferably less than 1200 ° C. If the temperature of the substrate 101 when forming the base layer 103 is within the above temperature range, the base layer 103 with good crystallinity can be obtained.

  After forming the base layer 103, the n-type semiconductor layer 104 is formed by stacking the n-contact layer 104a and the n-cladding layer 104b. The n contact layer 104a and the n clad layer 104b may be formed by sputtering or MOCVD.

The light emitting layer 105 can be formed by either sputtering or MOCVD, but MOCVD is particularly preferable. Specifically, the barrier layers 105a and the well layers 105b are alternately and repeatedly stacked, and the barrier layers 105a may be stacked in the order in which the barrier layers 105a are disposed on the n-type semiconductor layer 104 side and the p-type semiconductor layer 106 side. .
Further, the p-type semiconductor layer 106 may be formed by either sputtering or MOCVD. Specifically, the p-cladding layer 106a and the p-contact layer 106b may be sequentially stacked.

  In the case where the p-type semiconductor layer 106 is formed by the MOCVD method, it is preferable that the p-type semiconductor layer 106 be formed at a temperature higher than the deposition temperature of the light-emitting layer 105 in terms of improving the crystallinity of the p-type semiconductor layer 106. In the semiconductor light emitting device 3 of the present embodiment, since the seed layer 102 having a single crystal structure is formed, the crystallinity of the light emitting layer 105 is improved, whereby the p-type semiconductor layer 106 is made higher than the deposition temperature of the light emitting layer 105. It becomes possible to form at high temperature, and the crystallinity of the p-type semiconductor layer 106 is improved.

Thereafter, a translucent electrode is stacked on the p-type semiconductor layer 106, and the translucent electrode other than a predetermined region is removed by, for example, a generally known photolithography technique. Subsequently, similarly, patterning is performed by, for example, photolithography, and a part of the laminated semiconductor layer in a predetermined region is etched to expose a part of the n contact layer 104a. After a protective film such as SiO ₂ is formed on the entire surface, only the pad portion is extracted by photolithography, and the p-type electrode 107 and the n-type electrode 108 are formed on the exposed surface of the translucent electrode and the n-contact layer 104a.
In this way, the semiconductor light emitting device 3 shown in FIGS. 2 to 4 is manufactured.

Next, the manufacturing method of the illuminating device 1 is demonstrated.
First, as shown in FIG. 5, a reflector 31 is formed by injection molding a white resin on the lead 32. Next, the semiconductor light emitting element 3 is placed on the lead 32 in the recess 31 b of the reflector 31, and the lead 32 is connected to the p-type electrode and the n-type electrode of the semiconductor light emitting element 3 by bonding wires 33 and 34.

Next, as shown in FIG. 6, the concave portion 31b is filled with a mixed resin paste P containing phosphor powder and an uncured transparent resin. Covers the semiconductor light-emitting element 3 by mixing resin paste P, and the liquid level _{P 1} of the mixed resin paste P is recessed from the upper surface 31a, it is set in a range of -20μm~-100μm its recessed amount d from the upper surface 31a. In Figure 6, the difference between the height of the end portion _{B 1} of the center height and the liquid level _{P 1} of _{A 1} of the liquid level _{P 1} in the range of 20 to 100 [mu] m.

Here, dented d becomes the difference of the height of the upper surface 31a, the minimum height of the liquid level P _1. In the present specification, when the height of the upper surface 31a is 0 μm, the semiconductor light emitting element 3 side is minus (−). Therefore, the recessed amount d in the range of -20μm~-100μm from the top, when the height of the upper surface 31a and 0 .mu.m, the semiconductor light-emitting in the range of 20~100μm than the lowest height of the liquid level _{P 1} is the upper surface It means that it is located on the element 2 side.

  The filling of the mixed resin paste P is preferably performed by a podding method using a paste discharge device. This discharge device includes a discharge nozzle N that discharges a mixed resin paste and a control unit.

Mixed resin paste, the phosphor powder mass average particle size with a true density in the range of 3 g / cm ³ or more 4.7 g / cm ³ or less is 15μm or less in the range of 7μm is uncured transparent resin The viscosity is adjusted to the range of 4000 cP to 15000 cP.

Here, when the empty volume of the recess 31b in the state where the semiconductor light emitting element 3 is mounted is 6.5 μL, and the repetition accuracy of the injection amount of the mixed resin paste is, for example, ± 0.2 μL, the injection amount is 0. When it varies by 2 μL, the height of the central portion of the liquid surface P ₁ of the mixed resin paste P after injection varies by about 25 μm. When the height of the liquid surface moves by 100 μm, the color coordinate Y of the emission color moves by 0.01, and therefore the movement of the color coordinate Y that moves due to the variation in the injection amount becomes ± 0.0025.

The injection amount is controlled by the injection pressure. If the injection amount is moved by the minimum adjustment amount, the injection amount varies by about 0.3 μL. This corresponds to moving the height of the liquid level P ₁ by 38 μm. If the control range of the color coordinates Y of the luminescent color was ± 0.008, injection volume by controlling the injection pressure in a situation such as the liquid level P ₁ after the change implantation, etc. in the viscosity of the mixed resin paste will move it is varied, thereby it is necessary to control the level of the liquid surface P ₁ to ± 40 [mu] m.

Therefore, it is necessary to accurately measure the height of the liquid level P _1. The height of the liquid level can be measured with a laser displacement meter (for example, ZSHLD2 manufactured by OMRON). Measuring the distance from the connecting's midpoint the end of the recess 31b to the liquid surface P ₁ after injection. The height of the liquid surface is measured before the curing heat treatment after the injection, and the injection pressure may be controlled so that the position falls within the range of −20 to −100 μm from the midpoint connecting both ends of the recess 31b.

  Next, the mixed resin paste P is cured to form the phosphor-containing resin layer 2. The curing process may be performed by heating, for example. In this way, the lighting device 1 shown in FIG. 1 is manufactured.

  According to the illuminating device 1 described above, the phosphor-containing resin layer 2 covering the semiconductor light emitting element 3 is configured by mixing phosphor powder having a predetermined range of true density and mass average particle diameter in a transparent resin. Therefore, the phosphor powder can be uniformly dispersed in the transparent resin. Moreover, since the emission surface 2c of the phosphor-containing resin layer 2 is recessed from the upper surface 31a, and the amount of the depression is set in a range of −20 μm to −100 μm from the upper surface 31a, the injected phosphor-containing resin emission surface From 2c, it is possible to emit white light in which the width of the color coordinate Y is within a range of ± 0.008.

  Moreover, according to said illuminating device 1, while the half value width of the X-ray rocking curve of the (0002) plane of the p-type semiconductor layer of the semiconductor light emitting element 3 is 60 arcsec. Or less, the X-ray of the (10-10) plane Since the full width at half maximum of the rocking curve is 250 arcsec or less, it is possible to configure the lighting device 1 with little luminance deterioration with time and little shift of emission wavelength with time.

  Moreover, according to the manufacturing method of said illuminating device 1, the phosphor powder which has a true density and a mass average particle diameter of a predetermined range is mixed with uncured transparent resin, and mixed resin paste of predetermined viscosity Since the phosphor-containing resin layer 2 is formed by filling and curing the semiconductor light-emitting element 3 so as to cover the semiconductor light emitting element 3, the phosphor powder can be uniformly dispersed in the transparent resin. Moreover, the illuminating device 1 in which white light is radiate | emitted from the output surface 2c can be comprised by denting the output surface 2c rather than the upper surface 31a, and setting the amount of the dent to the range of -20 micrometers--100 micrometers from the upper surface 31a.

(Manufacture of semiconductor light emitting devices)
A c-plane sapphire substrate was introduced into a sputtering apparatus, the substrate was heated to 500 ° C. in a chamber, and nitrogen gas was introduced. Thereafter, the pressure in the chamber was maintained at 2.0 Pa, a high frequency bias was applied to the substrate side, and the substrate surface was cleaned by exposure to nitrogen plasma for 15 seconds.

  Subsequently, argon and nitrogen gas are introduced into the chamber, the substrate temperature is set to 500 ° C., high-frequency power of a predetermined output is applied to the target side, the pressure in the furnace is maintained at 1.0 Pa, argon gas is 10 sccm, nitrogen Formation of an AlN layer having a thickness of 21 nm to 40 nm as a seed layer was started on the c-plane of the sapphire substrate under the condition that the gas was circulated at 30 sccm (ratio of nitrogen to the whole gas was 75%). Then, after depositing AlN for a predetermined time, the plasma was stopped to lower the substrate temperature.

  Next, the substrate was taken out from the sputtering apparatus and introduced into the MOCVD furnace, and an n-type semiconductor layer (GaN layer) was formed by the following method.

  First, the substrate was placed on the susceptor in the MOCVD furnace, and after flowing nitrogen gas through the MOCVD furnace, the substrate temperature was raised to 1150 ° C. The amount of ammonia was adjusted so that the Group V element / Group III element ratio was 6000. Subsequently, hydrogen containing trimethylgallium (TMG) vapor was supplied into the MOCVD furnace, and deposition of a GaN layer on the substrate was started. After the undoped GaN layer (underlayer) having a thickness of 6 μm was grown, the supply of the raw material to the MOCVD furnace was terminated and the growth was stopped. Thereafter, power supply to the heater was stopped, and the temperature of the substrate was lowered to room temperature. The taken-out substrate exhibited a colorless and transparent mirror shape.

  Returning to the MOCVD furnace again, an n-type semiconductor layer comprising an n-type contact layer and an n-type cladding layer, a light emitting layer comprising a barrier layer and a well layer, and a p-type semiconductor layer comprising a p-type cladding layer and a p-contact layer. Were formed by MOCVD (metal organic chemical vapor deposition) method.

On the p-contact layer of the laminated semiconductor layer thus obtained, a light-transmitting positive electrode was formed using a photolithography method, and further dry-etched to partially expose the n-contact layer. A protective film of SiO ₂ was formed on the entire surface, and only the pad position was exposed by photolithography. Thereafter, a p-pad and an n-pad were formed on the exposed surface, thereby forming a semiconductor light emitting device.

  Therefore, for the semiconductor light emitting devices having seed layers with different thicknesses, the crystal plane of the underlayer of each device was measured by the rocking curve method. First, sample A1, sample A2, sample A3, and sample A4 were prepared. Samples B1 and B2 were also produced. Samples B1 and B2 deviate from the preferable film thickness conditions of the seed layer. Details of samples A1 and B2 are shown in Table 1.

  About these samples, the half width of the X-ray rocking curve of the (0002) plane and the (10-10) plane of the underlayer was measured. As the X-ray source, CuKα ray was used, and incident light having a divergence angle of 0.01 ° was used and measured using a Spectralic PANalytical X'pert Pro MRD apparatus. In consideration of errors due to the way the substrate is attached to the apparatus and the orientation direction with respect to the substrate differ depending on the sample to be measured, the rocking curve measurement of the (0002) plane finds a peak corresponding to the (0002) plane, The correction was performed by optimizing 2θ and ω, and then adjusting Psi to perform rocking curve measurement in the direction in which the peak intensity becomes maximum.

  Moreover, the rocking curve measurement of the (10-10) plane was performed using X-rays transmitted through the surface under the condition that X-rays were totally reflected. Specifically, when X-rays that diverge in the vertical direction with respect to a horizontally placed substrate are incident from the horizontal direction, a part of the X-rays are totally reflected. The detector was fixed at a 2θ position corresponding to the (10-10) plane, and φ scan was performed. Then, a six-fold symmetric peak was measured, and after fixing the optical system at the peak position showing the maximum intensity, 2θ and ω were optimized, and rocking curve measurement was performed. The results are shown in FIG.

  As shown in FIG. 7, the half-value width of the X-ray rocking curve of the (0002) plane of the underlayer is as large as about 180 arcsec in the sample B1 where the film thickness of the seed layer is 20 nm or less. On the other hand, Sample A1, Sample A2, Sample A3, and Sample A4 included in the range where the film thickness of the seed layer is 21 nm or more and 40 nm or less has a small value of about 50 arcsec or less. In Sample B2 where the seed layer thickness was 41 nm or more, it was about 50 arcsec or less.

  On the other hand, as shown in FIG. 7, the half width of the X-ray rocking curve of the (10-10) plane of the underlayer is as large as about 270 arcsec in the sample B1 in which the seed layer thickness is 20 nm or less. On the other hand, Sample A1, Sample A2, Sample A3, and Sample A4 in which the seed layer thickness is in the range of 21 nm to 40 nm are stable in the range of 200 to 225 arcsec. And in sample B2 in which the film thickness of the seed layer is 41 nm or more, it is about 260 arcsec, and tends to be large in the region where the film thickness exceeds 41 nm.

  Considering the above results, in the region where the thickness of the seed layer is less than 21 nm, the crystallinity of the seed layer is poor, and the orientation of the (0002) plane and (10-10) plane of the underlying layer formed thereon It is considered that the X-ray rocking curve half-width was increased. On the other hand, when the seed layer is thicker than 21 nm, the seed layer is crystallized and the crystal planes are aligned, so that the orientation of the (0002) plane and the (10-10) plane of the underlayer is improved. It is thought that the rocking curve half-width has decreased. However, since the crystallinity of the (10-10) plane is obtained from the substrate, if the seed layer thickness exceeds 40 nm, the underlying layer formed thereon becomes difficult to obtain information from the substrate. It is considered that the degree of orientation deteriorated and the X-ray rocking curve half width increased.

  As described above, by optimizing the film formation conditions of the seed layer formed on the substrate 11, the crystallinity of the stacked semiconductor layer stacked thereon is improved. Therefore, a homogeneous semiconductor light emitting device can be obtained from the laminated semiconductor layer having good crystallinity.

  Next, a semiconductor light emitting device of Sample A2 in FIG. 7 and Table 1 was prepared. In addition, three types of Enomoto package lead frames were prepared. This package lead frame is formed by injection-molding 144 or 40 reflectors made of white resin. Each of the two types of white reflectors whose outer dimensions are vertical and horizontal dimensions of 3.8 mm × 0.8 mm and 3.5 mm × 2.8 mm are formed on one lead frame. In addition, one type of white reflector having an outer dimension of 5.0 mm × 5.5 mm in a vertical and horizontal dimension is formed on one lead frame. In addition, all the depths of the recesses are 0.6 mm.

  Each package lead frame is die-bonded to the semiconductor light-emitting element of sample A2 using a die attach paste, heated at 150 ° C. for 2 hours to cure the die attach paste, and then made of Tanaka Kikinzoku with a diameter of 25 μm. Wire bonding was performed using a 4N gold wire. In the case of 3.8 mm × 0.8 mm and 3.5 mm × 2.8 mm, one light emitting element was mounted on one reflector. In the case of 5.0 mm × 5.5 mm, three light emitting elements were mounted in parallel on one reflector.

There silicone sealing resin viscosity during injection density at 9000cP 1.7g / cm ^3, an excitation wavelength of 450nm at a density 4.5 g / cm ^3, a silicate phosphor having a weight average particle size of 12μm at the emission wavelength 530nm A nitride phosphor having a density of 4.2 g / cm ³ , an excitation wavelength of 460 nm, an emission wavelength of 640 nm, and a mass average particle diameter of 9 μm was prepared. Regarding phosphors, those with an emission wavelength of 530 nm were prepared by classification, and those with an emission wavelength of 640 nm were also prepared with 7 μm and 15 μm. Silicone resins having a viscosity at the time of injection of 3000 cP and 20000 cP were further prepared. In this example, the green phosphor was 5.8% by mass, and the red phosphor was 2.1% by mass.

And silicone sealing resin was inject | poured with the automatic resin injector. 1.2 μL was injected for an outer diameter of 3.8 mm × 0.8 mm, 3 μL was injected for an outer diameter of 3.5 mm × 2.8 mm, and 6.8 μL was injected for a 5.0 mm × 5.5 mm.
The reflectors having dimensions of 3.8 mm × 0.8 mm and 3.5 mm × 2.8 mm have 144 reflectors in one frame. For these, 100 frames were prepared, and a total of 14400 lighting devices with the number of reflectors were prototyped.
In the case of 5.0 mm × 5.5 mm, there are 40 reflectors in one frame. For this, 400 frames were prepared, and a total of 16000 lighting devices were manufactured as the number of reflectors. It took about 1.5 seconds to inject the resin into one reflector.

About the obtained illuminating device, the shape of the upper surface of a reflector was measured with the laser displacement meter. Before starting, the dummy was poured and the injection amount was finely adjusted so that the average dent amount of 10 pieces became −60 μm ± 10 μm. When the average dent amount of one frame was shifted by 20 μm or more, the injection pressure was adjusted so as to fall within −60 μm ± 10 μm. The variation in one frame was about 16 μm for σ in the case of 3.8 mm × 0.8 mm, about 13 μm in the case of 3.5 mm × 2.8 mm, and 11 μm in the case of 5.0 mm × 5.5 mm.
The smaller the injection amount, the larger the variation in the injection amount. However, when one frame was averaged, the movement of the dent amount moved almost the same with time. As a result, it was possible to put all of the reflectors into a dent amount of -20 to -100 μm. However, sample no. In No. 8, the viscosity of the phosphor-containing sealing resin was too high, so that the injection amount could not be accurately controlled, and could not be achieved. After the injection, curing treatment was performed at 150 ° C. for 4 hours, and lighting devices of sample Nos. 1 to 10 as shown in Table 2 were manufactured.

  The output device (mW), total luminous flux (lm), luminous efficiency (lm / W), and color coordinates (x, y) of the obtained illuminating device were measured with an integrating sphere manufactured by Labsphere. A measurement result shows the average value of 20 illuminating devices. Further, the orientation dependency of the light emission of the lighting device was measured using a spectral radiance meter (Konica Minolta CS1000S). The direction perpendicular to the light emitting surface was 180 degrees, and the parallel direction was rotated 360 degrees when the angle perpendicular direction was tilted 60 degrees from the perpendicular. The degree of variation in color gamut is indicated by Δmax-min between X and Y. Further, measurements were performed on 1000 lighting devices, and the variation of color coordinates, which is the standard deviation between X and Y, was evaluated. In addition, 20 lighting devices were turned on by flowing 20 mA current (60 mA for 5.0 mm × 5.5) at 75 ° C., and 20 changes in total luminous flux (lm) and 20 color coordinates were measured after 1000 hours. And compared with average values. The results are shown in Table 3.

  First, sample no. Other than 7, it can be seen that the variation due to the direction of one illumination device is extremely uniform from the result of the spectral radiance meter. Sample No. In FIG. 8, since the injection amount could not be controlled, the variation in color locus when 1000 pieces were measured was very large at σ = 0.084. Next, sample no. In No. 7, since the phosphor had sunk, the degree of sag was slightly different depending on each lighting device, and as a result, the variation in color constellation when 1000 pieces were measured was large. The other samples were further separated by color with a handler tester. Since the center values were different, each sample was sorted with a center value of ± 0.008, and a yield of 95% or more was obtained.

FIG. 1 is a schematic cross-sectional view showing an example of a lighting device according to an embodiment of the present invention. FIG. 2 is a schematic cross-sectional view showing a semiconductor light emitting element provided in the illumination device shown in FIG. FIG. 3 is a schematic plan view showing the semiconductor light emitting element provided in the illumination device shown in FIG. FIG. 4 is a schematic cross-sectional view showing a laminated semiconductor layer constituting the semiconductor light emitting device shown in FIGS. FIG. 5 is a process diagram illustrating a method for manufacturing a lighting device according to an embodiment of the present invention. FIG. 6 is a process diagram illustrating a method for manufacturing a lighting device according to an embodiment of the present invention. FIG. 7 is a graph showing the relationship between the film thickness of the seed layer and the full width at half maximum of the rocking curve of the underlayer.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Illuminating device, 2 ... Phosphor containing resin layer, 2a ... Phosphor powder, 2b ... Transparent resin, 2c ... Outgoing surface (upper surface), 3 ... Semiconductor light emitting element, 20 ... Multilayer semiconductor layer, 101 ... Substrate, 102 ... seed layer, 104 ... n-type semiconductor layer, 105 ... light-emitting layer, 106 ... p-type semiconductor layer, d ... dent amount, P ... mixed resin paste, P ₁ ... liquid surface of the mixed resin paste.

Claims (7)

A semiconductor light emitting device having a main emission peak in a wavelength region of 430 nm or more and 500 nm or less;
A reflector made of a white resin, wherein a concave portion for accommodating the semiconductor light emitting element is opened on an upper surface, and a reflector for emitting light from the semiconductor light emitting element;
A phosphor-containing resin layer made of a phosphor and a transparent resin, which is formed so as to cover the semiconductor light-emitting element by being filled in the concave portion of the reflector, and absorbs light emitted from the semiconductor light-emitting element and emits light having a longer wavelength; With
The true density of the phosphor is between the phosphor ranges mass average particle diameter of 7μm or 15μm following with are 3 g / cm ³ or more 4.7 g / cm ³ or less in the range,
The phosphor-containing resin layer has an upper surface recessed from the upper surface, and the amount of the depression is set in a range of −20 μm to −100 μm from the upper surface. The semiconductor light emitting device includes at least a substrate, a seed layer made of AlN formed on the substrate, and a laminated semiconductor layer mainly composed of GaN formed on the seed layer,
The stacked semiconductor layer is formed by stacking an n-type semiconductor layer, a light emitting layer, and a p-type semiconductor layer in this order from the substrate side, and X-ray rocking of the (0002) plane of p-type GaN constituting the p-type semiconductor layer. The lighting device according to claim 1, wherein the half-value width of the curve is 60 arcsec. Or less and the half-value width of the X-ray rocking curve of the (10-10) plane is 250 arcsec. Or less.   The half width of the X-ray rocking curve of the (0002) plane of the seed layer is 100 arcsec. Or less, and the half width of the X-ray rocking curve of the (10-10) plane is 1.7 degrees or less. The lighting device according to claim 2.   The lighting device according to claim 1, wherein the phosphor-containing resin layer is a silicone resin. A semiconductor light emitting device having a main emission peak in a wavelength region of 430 nm or more and 500 nm or less;
A reflector made of a white resin, wherein a concave portion for accommodating the semiconductor light emitting element is opened on an upper surface, and a reflector for emitting light from the semiconductor light emitting element;
A phosphor-containing resin layer comprising a phosphor and a transparent resin, which is formed so as to fill the concave portion of the reflector so as to cover the semiconductor light-emitting element, absorbs light emitted from the semiconductor light-emitting element, and emits light having a longer wavelength. A method of manufacturing a lighting device comprising:
True density is mixed with 3 g / cm ³ or more 4.7 g / cm ³ or less of the phosphor transparent resin uncured mass average particle diameter of 15μm or less in the range of 7μm with a range, the mixed resin paste And adjusting the viscosity of the mixed resin paste to a range of 4000 cP to 15000 cP;
Covering the semiconductor light emitting element with the mixed resin paste, making the liquid surface of the mixed resin paste recessed from the upper surface of the reflector, and setting the amount of the depression in a range of −20 μm to −100 μm from the upper surface; ,
Curing the mixed resin paste to form a phosphor-containing resin layer;
The manufacturing method of the illuminating device characterized by comprising. The semiconductor light emitting device includes at least a substrate, a seed layer made of AlN formed on the substrate, and a laminated semiconductor layer mainly composed of GaN formed on the seed layer,
The laminated semiconductor layer is formed by laminating an n-type semiconductor layer, a light emitting layer, and a p-type semiconductor layer in this order from the substrate side, and X-ray locking of the (0002) plane of p-type GaN constituting the p-type semiconductor layer. 6. The illumination device according to claim 5, wherein a semiconductor light emitting device having a half width of the curve of 60 arcsec. Or less and a half width of the X-ray rocking curve of the (10-10) plane of 250 arcsec. Or less is used. Manufacturing method.   The said transparent resin is a silicone resin, The manufacturing method of the illuminating device as described in any one of Claim 5 or Claim 6 characterized by the above-mentioned.

Images