JP3130292B2 - Semiconductor light emitting device and method of manufacturing the same - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a flip-chip type semiconductor light-emitting device using a gallium nitride compound used for an optical device such as a blue light-emitting diode, and more particularly to a method for efficiently reflecting light from a P-side electrode. The present invention relates to a method for manufacturing a semiconductor light emitting device that is well collected and emits light from a light extraction surface.

[0002]

2. Description of the Related Art GaN, GaAlN, InGaN and I
GaN-based compound semiconductors such as nAlGaN have been widely used as semiconductor materials for visible light emitting devices and high-temperature operating electronic devices, and have been developed in the fields of blue and green light emitting diodes.

[0003] In the production of a semiconductor of this GaN compound,
Generally, insulating sapphire is used as a crystal substrate for growing a semiconductor film on the surface.
In the case of using an insulating crystal substrate such as sapphire, electrodes cannot be exposed from the crystal substrate side.
The p and n electrodes provided on the semiconductor layer are formed on one surface facing the crystal substrate.

FIG. 8 is a schematic perspective view of a conventional GaN-based semiconductor light emitting device.

The GaN-based semiconductor light emitting device 50 uses a sapphire substrate 50a as an insulating substrate, forms an n-type layer 51 and a p-type layer 52 thereon, and partially etches the p-type layer 52 to form an n-type layer. The mold layer 51 is exposed. And
On the n-type layer 51, an n-side electrode 51a for bonding is formed, and on the p-type layer 52, which serves as a light emitting region, a transparent electrode 52a is formed on almost the entire upper surface thereof, and a p-type electrode The basic configuration is to provide the side electrode 52b.

Here, the transparent electrode 52a is a laminated film of Ni and Au or a laminated film of Co and Au, and the p-side electrode 52b is mostly formed of a laminated film of the same combination. The n-side electrode 51a is generally used as a laminated film of Ti and Au or a laminated film of V and Al.

The material of each of the transparent electrode 52a, the p-side electrode 52b, and the n-side electrode 51a is selected on the premise that the condition for ohmic connection to the GaN system is satisfied. That is, Ni, Co, Ti, and V are electrode materials suitable for ohmic connection to the element side, and Au is hardly oxidized, and is used because the bonding property can be improved.

In such a semiconductor light emitting device 50, the active layer is InGaN laminated between the pn junction region between the n-type layer 51 and the p-type layer 52 or between them, and the surface of the p-type layer 52 is mainly used. It is mounted on a lead frame or the like as a light extraction surface. Then, by bonding an Au wire (not shown) to each of the n-side electrode 51a and the p-side electrode 52b to conduct the light to the lead frame side, light emission from the main light extraction surface is obtained.

The sapphire used as the substrate 50a is optically transparent and has n-side and p-side electrodes 51.
Since a and 52b are included on the same side surface, a flip-chip type assembly is possible. In this method, bump electrodes are formed on each of the n-side and p-side electrodes 51a and 52b, and these are bonded to the mount-side electrodes by an ultrasonic pressure bonding method or the like, thereby forming a wireless bonding assembly. When the flip-chip type is assembled, the upper surface of the substrate 50a when the light emitting element shown in FIG. 8 is turned upside down is the main light extraction surface.

On the other hand, in such a light emitting device 50 in which a GaN-based compound semiconductor layer is laminated on an insulating sapphire substrate 50a, static electricity is generated due to physical constants such as a dielectric constant ε of the device material and the device structure. It is known to be very weak. For example, in the case of an LED lamp in which the light emitting element 50 is mounted on a mount portion of a lead frame and sealed with an epoxy resin or the like, the LED lamp and a capacitor charged with static electricity are opposed to each other to cause discharge between the two. In some cases, breakdown occurs at a static voltage of about 100 V in the forward direction and at about 30 V in the reverse direction.

On the other hand, in order to prevent the light emitting element 50 from being destroyed by an overcurrent such as static electricity, it is effective to provide a Si diode as an electrostatic protection element. This electrostatic protection element is proposed by the applicant of the present application in advance, and is disclosed in Japanese Patent Application No. Hei.
A structure in which an Si diode having an n-type silicon substrate as a base material is connected to a light-emitting element while conducting so as to have a reverse polarity relationship, which can be applied to those described in the specification and drawings already filed as US Pat. It is what it was.

FIG. 9 shows an example in which the light emitting device 50 of FIG. 8 is mounted on a Si diode 53 for electrostatic protection to form a composite device. FIG. 9A is a plan view, and FIG. It is a longitudinal cross-sectional view by the CC line arrow of FIG.

The Si diode 53 is an n-type silicon substrate 5
9A, a p-type semiconductor region 53b is partially formed by injecting and diffusing impurity ions from the upper surface at a position deviated rightward in FIG. 9A. Then, a portion corresponding to the n-type semiconductor region 53b and an n-side electrode 54 and a portion
Side electrodes 55 are respectively formed, and further, on the lower surface, an n-electrode 56 for electrically conducting with a lead frame or the like is provided. Here, the n-side electrode 54 of the Si diode 53 and n
The resistance between the electrodes 56 functions as a protection resistance.

The n-side electrode 54 of the Si diode 53 is connected to the p-side electrode 52b of the light emitting element 50 via the micro bump 57, and the p-side electrode 55 is connected to the n-side electrode 51a via the micro bump 58 to emit light. The element 50 and the Si diode 53 are connected with opposite polarities. A part of the p-side electrode 55 is a bonding area of a wire connected to a lead frame or the like.

By the connection of the reverse polarity, when an overcurrent due to a high voltage is applied, the reverse voltage applied to the light emitting element 50 is close to the forward voltage of the Si diode 53, that is, the bypass is opened at 0.9V. By
As for the forward voltage applied to the light emitting element 50, an overcurrent is caused to flow by opening the bypass near the voltage drop due to the resistance component of the Si diode 53 and near the zener voltage Vz (for example, 10V). Therefore, destruction of the light emitting element 50 due to static electricity can be reliably prevented.

Here, in the flip-chip type semiconductor light emitting device, since a light emitting layer is formed above the transparent electrode 52a, light from this light emitting layer passes through the sapphire substrate 50a and emits light using the upper surface as a light extraction surface. There are two types: one that goes to the transparent electrode 52a side. For this reason, if the light traveling toward the transparent electrode 52a is reflected by a thick-film electrode having a high reflectance instead of the transparent electrode, the luminous efficiency from the light extraction surface can be increased. In this case, the thick film electrode becomes a p-side electrode including the bonding pad portion. Also in this case, the condition for the selection of the electrode material is that the ohmic connection to the GaN-based semiconductor laminated film is still possible, and the selection of the electrode material is in addition to the condition for the ohmic connection, Optimization is achieved by including a material that reflects light efficiently.

[0017]

When a thick film electrode is used instead of the transparent electrode 52a and a p-side electrode including a bonding pad is used, this p-side electrode is made of the same material as that shown in FIG. 8, that is, Ni and Au. Or a laminated film of Co and Au, it is possible to obtain a film which can be ohmic-connected to a p-type semiconductor laminated film and has good bonding properties of the microbumps 57.

However, in the laminated film of Ni and Au, the layer of Ni for making ohmic contact is generally very thin, and Au is formed thicker than this. Co and Au
Similarly, Co is thin and Au is thick. Such a relationship of the thickness comes from the condition for the optimum ohmic property, that is, the condition for minimizing the contact resistance between the electrode and the semiconductor (p-type GaN layer).

If Au is thicker than Ni or Co as described above, it is alloyed after the heat treatment, and the p-side electrode becomes gold or yellowish gold. A gold layer having such a gold or yellowish color at the interface of the p-side electrode is formed.

Here, it is known that the reflectivity when vertically projecting light of various wavelengths onto a fresh surface formed by vacuum-depositing a metal changes depending on the wavelength. this is,
For example, it is shown as a table of spectral reflectance of a metal surface on page 519 of "Science Chronology" of the 1996 edition. According to this table, when light is applied to the Au deposition surface, the reflectance is 8 if the wavelength is 0.550 μm (corresponding to green) or more.
When the wavelength is 0.500 μm, the reflectance sharply drops to 50% or less, while the reflectance is 0% or more. And
40% if the wavelength is 0.450 μm (corresponding to blue)
It can be seen that the reflectance decreases greatly as the wavelength becomes shorter.

Therefore, if there is a gold layer or a yellowish yellow layer at the interface between the p-type semiconductor laminated film and the p-side electrode,
The amount of light absorbed from the light emitting layer is larger, and the efficiency of reflection collection on the light extraction surface side is greatly reduced.

As described above, when Au is used as the electrode material, even if the light reflected from the p-side electrode is collected on the light extraction surface side to increase the luminous efficiency as in the flip chip type, the effect is sufficient. is not.

Furthermore, the above-mentioned "science chronology" of the 1996 edition
According to the table of the spectral reflectance of the metal surface described on page 519,
In the visible light region, the metal material having the highest reflectance is A
g. However, in the field of semiconductor devices, when Ag is used as an electrode material, it is necessary to prevent the occurrence of migration. This migration is caused by the ionization of Ag under certain conditions with respect to moisture, electric field, and the like, and causes a short circuit in the circuit.

In the case of this light emitting element as well, if Ag is used for the p-side electrode, in an LED lamp sealed with an epoxy resin, Ag migration is prevented by an electric field applied to the light emitting element and moisture permeating the epoxy resin. This causes a problem that a leak failure between the p-side electrode and the n-side electrode of the light emitting element occurs by short-time energization.

It is an object of the present invention to provide a method of manufacturing a flip-chip type semiconductor light emitting device in which Ag having the highest reflectance can be used for a p-side electrode without causing migration.

[0026]

According to the present invention, an n-type layer and a p-type layer of a GaN-based compound semiconductor are laminated on an insulating and light-transmitting substrate, and n-type layers are formed on the surface of the n-type layer. Forming a side electrode, forming a thin-film transparent electrode on almost the entire surface of the p-type layer, and forming a p-side electrode in a region on the transparent electrode and occupying a part of the surface of the p-type layer. A flip-chip type semiconductor light emitting device, a first main surface having two electrodes formed at positions corresponding to the p-side and n-side electrodes of the flip-chip type semiconductor light emitting device, and a second surface having a whole surface electrode formed thereon. And a sub-mount element having two main surfaces, and the flip-chip type semiconductor light-emitting element is made to face and conduct through two bumps on the first main surface of the sub-mount element via micro-bumps. A method of manufacturing a device, comprising: After the body light emitting elements are electrically connected to each other via a microbump with the electrodes of both elements facing each other on a wafer on which the submount elements are formed in a matrix, electrolytic plating is performed by dissolving Ag together with the light emitting elements. The electrode of the wafer is connected to a negative electrode of a power supply for electrolysis, and the Ag is deposited on the surface of the transparent electrode as a reflective layer by electrolytic plating.

According to the present invention, even if the p-side electrode is a light-emitting element with a transparent electrode, the sub-mount element is mounted on a wafer formed in a matrix, and then the electro-plating method is used. Ag compound layer can be easily attached and formed, and as a composite device with excellent external quantum efficiency,
Further, by using a Si diode as the submount element, it is possible to provide a composite element having excellent electrostatic withstand voltage.

[0028]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows G in the semiconductor light emitting device of the present invention.
FIG. 1A is a schematic view of an aN-based compound semiconductor light emitting device, in which FIG. 1A is a plan view, and FIG. 1B is a longitudinal sectional view taken along line AA of FIG.

1A and 1B, in the light emitting device 1, a plurality of semiconductor thin film layers are formed on the surface of an insulating transparent sapphire substrate 1a by a conventionally known metal organic chemical vapor deposition method. Things. For example, a GaN buffer layer 1b and an n-type GaN layer 1
c, an InGaN active layer 1d, a p-type AlGaN layer 1e, and a p-type GaN layer 1f, which have a double hetero structure or a quantum well structure.

The upper surface of one corner of the n-type GaN layer 1c is removed in a step shape by etching, and an n-side electrode 2 is formed on the removed portion by a vapor deposition method.
A p-side electrode 3 is similarly formed on the upper surface of the uppermost p-type GaN layer 1f excluding a portion cut off by etching by a vapor deposition method. Micro bumps 4 and 5 are formed on the n-side electrode 2 and the p-side electrode 3, respectively. However, the micro bumps 4 and 5 may be formed on the electrode on the submount element side in some cases.

FIG. 2 is a schematic view of a semiconductor light emitting device having the light emitting element 1.

A mount portion 6a formed at the upper end of the lead frame 6 has a Si diode element 7 for electrostatic protection.
Is mounted as a submount element, and
It is bonded and fixed with paste. And this Si
On the upper surface of the diode element 7, the light emitting element 1 shown in FIG.

The micro-bumps 4 and 5 of the light emitting element 1 are electrically connected to the p-side electrode 7b and the n-side electrode 7a of the Si diode element 7 and are sealed with the epoxy resin 8, respectively. The upper surface of 1a is a light extraction surface.

When the light-emitting element 1 is energized, the InGaN active layer 1d in the semiconductor laminated film becomes a light-emitting layer, and light from this light-emitting layer emits light from the light-extraction surface of the sapphire substrate 1a to the light-emitting side as shown in the conventional example. It goes to the p-side electrode 3. Then, the light directed toward the p-side electrode 3 may be efficiently reflected. However, if the electrode material of the p-side electrode 3 includes Au, the reflectance of blue light is reduced. .

On the other hand, in the present invention, in order to increase the reflection efficiency from the p-side electrode 3, the material is changed to the p-type GaN layer 1f.
Ni, Co or S as a contact layer 3a that contacts
b is formed to a thickness of 50 nm, and a reflective layer 3b is formed thereon by stacking a 1.5 μm-thick silver-white metal Al or Zn having a high reflectance with respect to blue and green light.
Here, the reason for thinning the contact layer 3a is that p-type Ga
In order to make a good ohmic contact with the N layer and to make the metal used for it thin enough to transmit most of the light because the reflectivity for blue and green light is not always good. Thus, light is reflected by the reflective layer 3b having a high reflectance laminated thereon. Further, it is necessary that the thickness of the reflective layer 3b is sufficient to prevent light from passing therethrough. Thus, light traveling toward the p-side electrode 3 can be efficiently reflected. When the reflective layer 3b is made of Zn, an Au layer is formed on the reflective layer 3b in order to improve the bonding property with the micro bumps 4 and 5.

When Ag having the highest reflectance for blue or green light is used for the reflective layer 3b, a migration phenomenon is accompanied. In addition, in the case of the flip-chip type semiconductor light emitting device, the p-side electrode 3 and the n-side electrode 2 are adjacent to the same surface, so that a migration phenomenon is likely to occur. There is a disadvantage that mode failure occurs.

On the other hand, in the present invention, A
When g is used, the reflection layer 3 is formed as an alloy of Pd or Pt.
b is formed. When Pd or Pt is mixed with Ag, the reflectance slightly decreases, but the migration of Ag is suppressed. The appropriate value of the mixing is about 30 wt% in the case of Pd, and about 10 wt% in the case of Pt, and the film thickness of the reflective layer 3 b is 1.5 μm.

FIG. 3 is an outline of a semiconductor light emitting device according to the present invention, which is a composite element comprising a light emitting element and a submount element. (A) of the same figure is a schematic plan view thereof, and (b) of the same figure is a longitudinal sectional view taken along line BB of the same figure (a).

The GaN-based compound semiconductor light-emitting element used here is a conventional light-emitting element 50 having a transparent electrode 52a, which is bonded on a submount element (Si diode) 7 via a microbump to form a composite element. The reflective layer 10 is formed by forming a silver-white metallic material, Ag or Zn, having a high reflectance to blue and green light on the transparent electrode by 1.5 μm.
It is formed by electrolytic plating with a film thickness of. In this case, since the reflection layer 10 is formed after forming the composite element,
Its cross-sectional shape is different from that of FIG. In other words, in the case of FIG. 2, the reflective layer 3b is formed only on the p-side electrode 3 of the light emitting element 1.
In the case of FIG. 3, the reflective layer 10 is formed on the entire surface of the transparent electrode 52a, the p-side electrode 52b, the microbump 5, and the n-side electrode 7a of the Si diode.
Is formed, but it is the portion formed on the transparent electrode 52a that functions as a reflective layer. This reflection layer 10
The configuration is the same as that of FIG. 2 except for the light emitting element 50 and the p-side electrode portion.

Also, in this embodiment, when Ag is used as the reflective layer 10, a migration countermeasure is required. In this case, as shown in FIG. 4A, a protective film 11 made of a Ni plating layer is formed on the surface of the Ag plating layer that is the reflection layer 10. Thereby, the contact between water and Ag that have penetrated into the resin can be cut off to suppress migration. The reason why Ni is used as the protective film 11 is that a uniform plating layer can be easily formed. Due to the formation of the protective film, leak mode defects between electrodes due to Ag migration are dramatically reduced.

The effect of the protective film 11 is the same in the case of a light emitting element in which the reflective layer 3b of the thick p-side electrode shown in FIG. 4B is formed using Ag.

As described above, Ag is applied to the surface of the transparent electrode 52a.
By providing the reflective layer 10 on which a metal such as is adhered by a plating method, light from the active layer is reflected by the reflective layer 10 to the main light extraction surface side. Therefore, as compared with the conventional structure in which the light that has exited the light emitting element 50 is passed again and collected, the attenuation of the light emission energy can be suppressed to be small, and the light emission luminance can be improved.

The temperature of the transparent electrode 52a, which is a light-emitting region, tends to be higher than that of other portions. Since the transparent electrode 52a is extremely thin and has a small thermal conductivity, the influence of light emission luminance on heat cannot be ignored. On the other hand, the transparent electrode 52a
Since the reflective layer 10 formed by Ag plating is laminated and thickened, the heat dissipation increases due to the high thermal conductivity of Ag. Therefore, not only the emission luminance by Ag can be improved, but also the resistance can be improved by preventing overheating. The reflective layer 10 is formed not only on the surface of the transparent electrode 52a but also on the bump electrode 5a.
Also, since it is formed on the surface of the n-side electrode 7a of the Si diode 7, the heat dissipation can be further improved by utilizing the expansion of the surface area and the heat conduction by Ag.

Further, by further covering the protective film 11, the surface of the transparent electrode 52a can be further thickened, so that the heat capacity and the heat dissipation are further improved, and the durability of the light emitting element 50 is improved.

FIG. 5 shows the reflection layer 1 of the semiconductor light emitting device of the present invention.
FIG. 2 is a schematic view illustrating steps of a method for manufacturing a protective film 11 and a protective film 11.

In the drawing, a wafer 12 made of n-type silicon and having a pattern of Si diodes 7 formed thereon is prepared as a material. This wafer 12 is shown in FIG.
In addition to the formation of the p-type semiconductor region 7e of the Si diode element shown in FIG. 1, the patterns of the n-side and p-side electrodes 7a and 7b are formed on one side, and the n-side electrode 7c is formed on the other side.

On the other hand, the n-side and p-side electrodes 51a, 52b
The light emitting elements 50 singly diced from a wafer having micro bumps 4 and 5 formed by plating or stud are arranged on an expanded sheet. Then, the light emitting element 50 is picked up by a collet, and the micro bumps 5 and 4 are brought into contact with each other in accordance with the patterns of the n-side and p-side electrodes 7a and 7b of the wafer 12, and the wafer is melt-bonded by weighting, heat and ultrasonic waves. 12 on. In this case, the micro bumps are
It may be formed on the side of the electrodes 7a and 7b on the p-side and the p-side.

Next, the wafer 12 on which the light emitting elements 50 are mounted is immersed in the electrolytic plating solution 14 in the electrolytic plating tank 13. In the electrolytic plating tank 13, a positive electrode 13a for plating is arranged at a position facing the wafer 12, and an n-electrode 7c on the back surface of the wafer 12 is provided as a negative electrode for plating. By connecting a power source between the plating solution and the plating solution, the plating solution in the electrolytic plating tank 13 is electrolyzed. Accordingly, first, in the case of Ag plating of the reflective layer, the n-side electrode 7a, the microbump 5, and the p-side electrode 52b and the transparent electrode 52a of the light emitting element 50, in which Ag ions in the plating solution are conducted to the n-electrode 7c of the wafer 12.
Is deposited and adhered to the surface of the substrate to form an Ag reflective layer 10 as shown in FIG.

Thereafter, the wafer 12 is taken out and washed sufficiently, and then diced by a dicer. Thereby, the light emitting element 50 in which the Ag reflective layer 10 is formed on the portion including the transparent electrode 52a of the light emitting element 50 and the Si diode 7 are used. A composite device is obtained.

In the case of the example in which the protective film 11 is provided as shown in FIG. 4A, after forming the reflective layer 10 of Ag,
The protective film 11 may be immersed in a metal plating solution for the protective film 11 in another electrolytic plating tank, and the protective film 11 may be formed by an electrolytic plating method by the same operation. Also, as shown in FIG. 4B, in the light emitting element 1 having the thick p-side electrode 3,
When Ag is used for the side electrode 3, after bonding to the wafer 12, the protective film 11 may be immersed in a metal plating solution for the protective film 11 in the electrolytic plating tank, and the protective film 11 may be formed by the electrolytic plating method by the same operation. it can.

FIG. 6 shows an n-side electrode 7a and a p-side electrode 7b formed on a wafer 12 for manufacturing the Si diode 7.
FIG. 5 is a diagram showing a pattern and a conductive structure with an n-electrode 7c.

As shown in FIG. 6A, the plating surface is the bump 5 having the same potential as the n-side electrode 7a, the p-side electrode 52b of the light emitting element 50, and the transparent electrode 52a. It is preferable to use the n-side electrode 7a instead of the n-electrode 7c. That is, the n-side electrode 7c on the back surface and the n-side electrode 7 on the front surface
Since a resistance R of several ohms per unit area exists between the transparent electrode 52a and the transparent electrode 52a, it is preferable to use a negative electrode for plating on the n-side electrode 7a in order to facilitate plating on the transparent electrode 52a. Further, as shown in FIG. 6B, it is preferable that all adjacent n-side electrodes 7a are made conductive by the connection electrodes 9 in order to make them the same potential.

However, the electric current flowing between the plating electrodes is 0.1.
Since the current is as small as about 03 mA / mm ^2,
No inconvenience occurs even if c is used as a negative electrode for plating.

As described above, Ag is formed on the entire surface of the transparent electrode 52a.
Since the reflection layer 10 is formed by plating, as described above, the light that passes through the transparent electrode 52a from the active layer is reflected by the reflection layer 10, so that the light emission luminance is improved.
Ag plating is performed on the p-side electrode 52b, the micro bump 5
In addition, since it also adheres to the n-side electrode 7a of the Si diode 7, heat radiation of each part is promoted, and the influence of heat on light emission can be suppressed.

In the illustrated example, the combination with the Si diode 7 for protecting the static electricity is used.
A substrate or the like which has electrodes conducting to the p-side and p-side electrodes 51a and 52b and can provide a conducting structure for electrolytic plating may be used.

FIG. 7 shows a semiconductor light emitting device according to the present invention having the requirements of claim 1, wherein a composite element having a protective film 11 formed on the surface of a reflective layer 10 is mounted and joined to a lead frame, and a silicone resin and an epoxy resin are attached. It is a longitudinal cross-sectional view of the LED lamp molded with.

The structure will be described in detail. The reflection layer 10 shown in FIG.
After the composite element having the protective film 11 formed on the surface thereof is conductively connected, the bonding gap between the flip-chip type semiconductor light emitting element 50 and the Si diode element 7 is filled with the silicone resin 15. Then, the lead frame 6 including this mounting portion
Is molded with epoxy resin 8. However, the light extraction surface of the flip-chip type semiconductor light emitting element 50 is not covered with the silicone resin. The reason is,
This is because the light transmittance of the silicone resin is lower than that of the epoxy resin for molding, and the light extraction efficiency is lower because the refractive index is smaller than that of the epoxy resin.

This embodiment shows a method that can more reliably prevent Ag migration that occurs when Ag is used for the reflective layer. That is, in the second embodiment, when the surface of the Ag reflective layer 10 is covered with the Ni protective film 11, the migration of the resin is blocked, so that the Ag migration can be suppressed. In the formation process of No. 11, there may be a case where a portion that cannot cover the Ag surface, such as a pinhole, occurs. In that case, moisture invades from the pinhole, and heat and an electric field are applied to ionize Ag and melt, and a leak mode defect due to Ag migration occurs. Even if such a pinhole is present, if the silicone resin is filled in the bonding gap of the composite element shown in FIG. 7, migration of Ag is reliably suppressed and no leak mode defect occurs.

Table 1 summarizes the results of the Ag migration acceleration reliability test in the following three cases.

[0061]

[Table 1]

In the case of, the reflective layer 10 is made of Ag, without a protective film, and without a silicone resin. In the case of, the reflective layer 10 is made of Ag, with a protective film of Ni, and without a silicone resin. In the case of, the reflective layer is made of Ag, with a protective film of Ni, and a silicone resin. Is filled. The conditions of the acceleration test are as follows: temperature is 85 ° C, humidity is 85%,
Vf = 3.3 V and If = 12 mA.

As shown in Table 1, 100% of leak mode failures occurred in 24 hours, but 1/3
It can be seen that no leak mode failure occurred for the sample of 240 hours or more, and no leak mode failure occurred even for 500 hours or more.

[0064]

According to the present invention, a Ga having a conventional transparent electrode is used.
After mounting and bonding an N-based compound semiconductor light emitting device on a wafer with submount elements formed in a matrix, a reflective layer with high reflectivity can be easily formed on the transparent electrode of the light emitting device by electrolytic plating. It is possible to improve the luminance by using this chip. Further, even when Ag is used for the reflective layer, a protective film can be easily formed thereon, so that Ag migration can be suppressed and a light emitting device with high luminance can be obtained. In addition, since a thick plating layer is formed on the transparent electrode and in the vicinity thereof, heat radiation on the light emitting surface is promoted, and resistance can be improved.

[Brief description of the drawings]

FIG. 1 is a flip-chip type semiconductor light emitting device according to an embodiment of the present invention, wherein (a) is a plan view and (b) is a schematic longitudinal sectional view taken along line AA of FIG.

FIG. 2 is a schematic diagram of an LED lamp including the light emitting device of FIG.

3A and 3B are schematic views of a composite element portion of a semiconductor light emitting device according to an embodiment of the present invention, in which FIG. 3A is a plan view, and FIG. 3B is a longitudinal section taken along line BB in FIG. Figure

4A and 4B are longitudinal sectional views showing an example in which a protective layer is formed on the surface of a reflective layer, wherein FIG. 4A is a longitudinal sectional view showing a case of a reflective layer formed by an electrolytic plating method, and FIG. Longitudinal sectional view showing the case of the reflective layer of the p-side electrode

FIG. 5 is a schematic view of the steps in the manufacturing method of the present invention.

6A and 6B are diagrams illustrating a plating process of the present invention, wherein FIG. 6A illustrates a conductive structure of electrolytic plating, and FIG. 6B illustrates a pattern of an n-side electrode and a p-side electrode formed on a Si diode wafer. Top view of

FIG. 7 is a longitudinal sectional view showing an example in which the composite element is mounted on a lead frame and joined and molded.

FIG. 8 is a perspective view showing an example of a GaN-based compound semiconductor light emitting device.

9A and 9B are examples in which the light emitting device of FIG. 8 is mounted on a Si diode to form a composite device, wherein FIG. 9A is a plan view, FIG. 9B is a longitudinal sectional view taken along line CC of FIG.

[Explanation of symbols]

 Reference Signs List 1 light emitting element 1a crystal substrate 1b GaN buffer layer 1c n-type GaN layer 1d InGaN active layer 1ep p-type AlGaN layer 1f p-type GaN layer 2 n-side electrode 3 p-side electrode 3a ohmic junction layer 3b reflection layer 4,5 microbump 6 Lead frame 6a Mount part 7 Si diode element 7a n-side electrode 7b p-side electrode 7c n-electrode 7d n-type silicon substrate 7e p-type semiconductor region 8 epoxy resin 9 connection electrode 10 reflective layer 11 protective film 12 wafer 13 electrolytic tank 13a positive electrode 14 Electroplating solution 15 Silicone resin

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-57-106087 (JP, A) JP-A-6-314825 (JP, A) JP-A-6-69546 (JP, A) JP-A-61-1986 220344 (JP, A) JP-A-9-232632 (JP, A) JP-A-8-64871 (JP, A) JP-A-5-159618 (JP, A) (58) Fields investigated (Int. ⁷ , DB name) H01L 33/00 H01S 5/00-5/50 H01L 21/18 H01L 21/44 H01L 29/40

Claims (1)

(57) [Claims] 1. A method in which G is placed on an insulating and light-transmitting substrate.
An n-type layer and a p-type layer of an aN-based compound semiconductor are laminated, an n-side electrode is formed on the surface of the n-type layer, and a thin-film transparent electrode is formed on almost the entire surface of the p-type layer. A flip-chip type semiconductor light emitting device in which a p-side electrode is formed in a region on the electrode and occupying a part of the surface of the p-type layer; and a p-side and n-side electrode of the flip-chip type semiconductor light emitting device. A submount element having a first main surface on which two electrodes are formed at corresponding positions and a second main surface on which a full-surface electrode is formed;
A method of manufacturing a semiconductor light-emitting device in which the flip-chip type semiconductor light-emitting elements are opposed to each other via micro-bumps on two main surfaces of the semiconductor light-emitting element to conduct the semiconductor light-emitting elements, wherein the sub-mount elements are arranged in a matrix. After the electrodes of both elements are opposed to each other on the wafer formed on the wafer and electrically connected via micro-bumps, the wafer is immersed together with the light emitting elements in an electrolytic plating solution in which Ag is dissolved, and the electrodes of the wafer are electrolyzed. A method for manufacturing a semiconductor light emitting device, comprising: connecting to a negative electrode of a power source for use; and forming the Ag as a reflective layer on the surface of the transparent electrode by electrolytic plating.