DE19820777A1 - Electrode for light emitting semiconductor device - Google Patents

Abstract

An electrode, for a light emitting semiconductor device, comprises (a) a transparent electrode in contact with a p-conducting GaN-based compound semiconductor surface; and (b) a wire bonding electrode which is in electrical contact with the transparent electrode and which includes a semiconductor-contacting region having a higher contact resistance per unit area wrt. the semiconductor than that of a transparent electrode region in contact with the semiconductor. Also claimed are similar electrodes in which the transparent electrode consists of a first layer of Au, Pt and/or Pd and an overlying second transparent metal oxide layer selected from oxides of Ni, Ti, Sn, Cr, Co, Zn, Cu, Mg and In. Further claimed are processes for producing the above electrode.

Description

The present invention relates to an electrode for a light-emitting Semiconductor device on a surface of a p-type composite GaN-based semiconductor is formed, and a manufacturing method the electrode.

In recent years, compound semiconductor materials have been based on GaN as a semiconductor material for light emitting devices that emit light emit with a short wavelength, attracting attention. Of the GaN-based compound semiconductors are made on various oxide substrates, such as sapphire crystal or a III-V group composite substrate the metal organic chemical vapor deposition process (MOCVD method), a molecular beam epitaxy method (MBE method) or a other such process is formed.

A compound semiconductor based on GaN is a compound semiconductor of the III-V group, which are generally represented by Al _x Ga _y In _1-xy N, (O ≦ x ≦ 1, O ≦ y ≦ 1, O ≦ x + y ≦ 1).

In the case of a light-emitting device that is coated or Laminating layers of this compound semiconductor based on GaN det, which is a substrate made of an electrically insulating material, such as a sapphire substrate, can be used on the back surface of the substrate  no electrode is provided, as opposed to when a half conductor substrate, such as a GaAs or GaP substrate is used, the elek is tric leading. Accordingly, two positive and negative Electrodes on the same surface of the light emitting device educated. When electricity passes through the two electrodes to To produce a light emission, the light is emitted from the surface tiert on which the electrodes are provided because the sapphire or other such substrate material is an insulator. The light is after emitted above.

A property of the GaN-based compound semiconductor material is that the current diffusion in the transverse direction is small. Even if electrodes are formed and light is emitted by electricity between them due to this property, the main part of the current takes place flow directly under the electrodes, consequently the light emis sion limited to the area just below the electrodes and diffuses not easily to the peripheral area of the electrodes. Therefore in the In the case of conventional opaque electrodes, light emission interrupted by the electrode itself and cannot be from the upper side of the electrode. As a result, the intention not achieved improvement in light emission intensity.

In order to overcome this disadvantage, JP-A-6-314 822 discloses one Technology related to device construction which allows light to pass through sige electrode, which comprises a very thin metal, as a p-type Electrode is used and covers almost the entire front upper surface of the device is formed to thereby allow that emitted light passes through the translucent electrode and is emitted from the upper side to the outside. With this revelation  For example, Au, Ni, Pt, In, Cr or Ti are used as the electrode material used and the metal film that is formed by vapor deposition is heat-treated at a temperature of 500 ° C or higher to ne sublimation of the metal so that the thickness on between 0.001 and 1 µm is reduced, thereby ver light transmission lend. The term "translucent" as used herein with respect to the Electrode used refers to an electrode through which one Light emission generated under the electrode can be observed can. To allow observation through the electrode takes place, the electrode must have a light transmission of at least Have 10%.

However, such a thin metal film has low strength, which makes it impossible to bond wires directly to the thin film or to inject to inject electrical power from an external source adorn. For this reason, electrodes are used for light emitting semiconductor devices generally have a structure which includes that in addition to the translucent electrode Wire bonding or wire bonding electrode is formed, the one has electrical contact with the translucent electrode, and that this wire bonding electrode is used to start the wire close, which is used to supply electricity to the translucent electrode to transport.

In general, when a transparent electrode is formed using a thin metal film as shown in Fig. 23, the structure includes that the wire bonding electrode 8 is formed on the transparent electrode 7 . However, it is difficult with this structure, adhesion between the front surface of the transparent elec trode 7 and the lower surface of the Drahtbondingelektrode 8 determine safe, which sometimes causes the Drahtbondingelektrode 8 peels off during the electrode production process.

To overcome this, JP-A-7-94782 discloses a technique for improving the bonding properties, which is illustrated in Fig. 24. With this arrangement, a window is formed in the translucent electrode 7 , above which the surface of the semiconductor 9 is exposed, the wire bonding electrode is formed on the window 70 to cause direct contact between the wire bonding electrode 8 and the surface of the semiconductor 9 .

In most cases, a thick film with a thickness of about 1 micron used for the wire bonding electrode as a way to influence of the wire bonder (the contacting device). Because he is so thick, the wire bonding electrode can not let light through be awarded. This means that the light emission is direct occurs under the wire bonding electrode by the wire bonding electrode de is interrupted, and thus not be emitted to the outside can. In order to achieve a higher emission brightness, there is an upside construction required, through the current not directly into the semiconductor section is injected below the wire bonding electrode, but instead flows to the translucent electrode.

JP-A-8-250 768 discloses a technique by which there is no current to it Area under the wire bonding electrode flows. This is accomplished by the semiconductor layers under the wire bonding electrode by various process with a high resistance area such as by forming a silicon oxide layer in  an area is left free of no treatment for p-wire off education undergoes by heat treatment or ion implant tation etc. is used. The area with high resistance prevails that current flows under the wire bonding electrode, the current instead is passed directly to the translucent electrode to thereby using the electricity efficiently.

However, in the disclosure of JP-A-8-250 768, the structure of the the high resistance area under the wire bonding electrode sees the formation of silicon oxide layers and steps to counter them to increase the size of the semiconductor. This makes the process complicated and it takes a long time to manufacture. For example, the To form silicon oxide layers, it is necessary to order photolithography to effect patterning, or plasma CVD processes and the like to use. Similarly, photolithography, ion implantation, Heat treatment and other such processes are used to to form a semiconductor region with high resistance. All of these processes They are complex and time consuming.

If the above-described arrangement having a high resistance area is to be applied to the configuration of JP-A-7-94782 above, in which the wire bonding electrode 8 is provided on the window 70 ( Fig. 24), the area becomes also high resistance is formed in the semiconductor 9 below the wire bonding electrode 8 . This creates an arrangement in which the current from the peripheral portion 8 a of the wire bonding electrode 8 must flow through the translucent electrode 7 in the semiconductor 9 , wherein a light emission in the injection area 91 is generated. Because the peripheral portion 8 a acts as a barrier to the light generated, the light emission can not be removed upwards. The light emission is therefore wasted, reducing the emission efficiency.

The aim of the present invention is to provide an electrode for light emitting rende semiconductor devices to provide a simple structure used, which is able to flow a current under the wire bonding to safely block the electrode and reduce the light emission efficiency can improve.

The present invention achieves the above object by using an electrode for a semiconductor light emitting device is provided on the Surface of a p-type compound semiconductor based on GaN is formed comprising: a translucent electrode formed so that it comes into contact with the surface of the semiconductor, and a Wire bonding electrode that is in electrical contact with the translucent Sigen electrode is and is formed such that it is in partial contact reaches the surface of the semiconductor, at least one area, which is in contact with the semiconductor has a higher contact resistance per unit area with respect to the semiconductor has as a range of translucent electrode that is in contact with the semiconductor.

The wire bonding electrode can have a multilayer structure in which the top layer is made of Al or Au.

The translucent electrode may include a first layer that is formed such that it is in contact with the surface of the semiconductor arrives, and which comprises at least one component, which from the group pe is selected, which consists of Au, Pt and Pd, and it can be a second Include layer formed on the first layer and a light  permeable metal oxide comprising an oxide of at least one Contains metal selected from the group consisting of Ni, Ti, Sn, Cr, Co, Zn, Cu, Mg and In exist.

The second layer has an oxygen content that gradually decreases from that second layer towards the first layer in the area near he interface between the second layer and the first layer decreases.

The first layer may contain a metal element which is a main element is part of the metal oxide that forms the second layer.

The translucent electrode can be formed such that it on egg a portion above the top surface of the wire bonding electrode lies on which the wire bonding electrode is arranged.

The transparent electrode can be formed such that it over the circumference of the top surface of the wire bonding electrode.

The translucent electrode can be formed such that it ge entire upper surface of the wire bonding electrode is covered.

A portion of the second layer of the translucent electrode, the overlying the wire bonding electrode can be removed so that the first Layer is exposed.

The electrode for a light-emitting semiconductor according to the invention direction also includes an electrode placed on the surface of a p-lei tendency compound semiconductor is formed based on GaN, which is also a light  permeable electrode formed to be in contact with the surface of the semiconductor, and the one wire bonding includes electrode that is in electrical contact with the translucent Electrode stands and is formed such that a lower surface in part wise contact with the surface of the semiconductor and the light permeable electrode lies over an upper surface.

The transparent electrode can be formed such that it over the circumference of the top surface of the wire bonding electrode.

The translucent electrode can be formed such that it ge entire upper surface of the wire bonding electrode is covered.

The present invention also provides a method of making a Electrode for a semiconductor light-emitting device ready on a surface of a p-type compound semiconductor based on GaN comprising: a first step that a wire bond elec trode is formed on a portion of the surface of the semiconductor, ei NEN second step that a first layer on the surface of the half conductor is formed, wherein the first layer at least one component which is selected from the group consisting of Au, Pt and Pd stands, and is formed such that it is at a portion above the upper Surface of the wire bonding electrode lies on which the wire bonding electrode is arranged, a third step that on the first layer forming a second layer comprising at least one metal which is selected from the group consisting of Ni, Ti, Sn, Cr, Co, Zn, Cu, Mg and In, and a fourth step is that a translucent electrode is formed by placing the first and second layers in an oxygen containing atmosphere can be heat treated.  

The method of manufacturing an electrode for a light emitting Semiconductor device according to the present invention can instead comprise: a first step that a wire bonding electrode on a Section of the surface of the semiconductor is formed, a second Step that an alloy layer on the surface of the semiconductor ge is formed, wherein the alloy layer comprises an alloy which min contains at least one metal selected from the group consisting of Au, Pt and Pd, and which contains at least one metal that consists of the Group is selected from Ni, Ti, Sn, Cr, Co, Zn, Cu, Mg and In be stands, and which is formed such that it is at a portion above the top surface of the wire bonding electrode, on which the wire receipt thing electrode is arranged, and a third step that a light permeable electrode is formed by the alloy layer in a Oxygen containing atmosphere is heat treated to on the Semiconductor side to form a first layer made of metal or an Le alloy, and to form a second layer consisting of a light permeable metal oxide formed on the first layer.

As described above, the area is the wire bonding electrode, which is in contact with the semiconductor, formed in such a way that he has a higher contact resistance per unit area with respect to has the semiconductor as the region of the transparent electrode, which is in contact with the semiconductor, which makes it possible to prevent current from flowing under the wire bonding electrode, so that the total current from around the wire bonding electrode into the light permeable electrode is injected, from which it enters the laminate occurs and contributes to the light emission function. That means it won't Light emission generated under the wire bonding electrode, so that by  the light is not obstructed by the wire bonding electrode, we sentlich all the light generated by the translucent electro de can be emitted upwards. This allows the electricity to be effective exploited and the light emission efficiency can be improved.

This electrode design, which is a wire bonding electrode and a having translucent electrode can be formed by thin films using a method such as a vapor deposition ver driving, growing up. The process is very simple by only includes the vapor deposition of the metal material so that the formation that films can be made quickly. That is, a current flow under The wire bonding electrode can be securely attached using a simple structure blocked, which can be easily formed without complex processes to have to make.

The invention will be exemplified below with reference to the drawing wrote in this shows:

Fig. 1 is an overall view of the structure of the electrode of the invention,

Fig. 2 is a graph (Span-current voltage characteristics) of each metal or any alloy with respect to a p-type GaN semiconductor shows the contact resistance properties,

Fig. 3 shows the two-layer structure of Drahtbondingelektrode of the present invention,

Fig. 4 shows the entirety of the upper surface of the Drahtbondingelek trode of Fig. 3, on which the transparent electrode is located,

Fig. 5 shows the peripheral portion of the upper surface of the Drahtbon ding electrode of Fig. 3, on which the transparent electrode is located,

Fig. 6 shows the two-layer structure of the transparent electrode of the present invention,

FIG. 7 shows a second method for producing the two-layer transparent electrode of FIG. 6, FIG. 7 (a) showing the first stage and FIG. 7 (b) showing the second stage,

Fig. 8, the whole of the upper surface of the Drahtbondingelek trode of FIG. 3, over which a two-layer electrode is lichtdurchlässi ge,

Has been removed in the over the entirety of the upper surface of the upper Drahtbondingelektrode of FIG. 3 placed a zweischichti ge transparent electrode and the second layer Fig. 9 an arrangement

Fig. 10 shows the peripheral portion of the upper surface of the Drahtbon ding electrode of Fig. 3, on which the transparent electrode is two-layered light,

Fig. 11 is a plan view showing the arrangement of the approximate shape electrode for a light-emitting device according to a first exporting the present invention,

Fig. 12 is a sectional view taken along line 12-12 of Fig. 11,

Fig. 13 is a view showing the depth profile of the respective elements of the transparent electrode of the first embodiment obtained by Auger electron spectroscopy was measured

Fig. 14 is a thin-film XRD pattern of the second layer of the transparent electrode of the first embodiment,

Fig. 15 is a cross-sectional view according to the guide form the structure of the electrode for a light-emitting device of a second one of the present invention, wherein FIG 15 (a). The first stage, Fig. 15 (b) shows the second stage, and FIG. 15 (c) shows the final state,

Fig. 16 is a view showing the depth profile of the respective elements of the transparent electrode of the second embodiment obtained by Auger electron spectroscopy was measured

Fig. 17 is a plan view showing the arrangement of the approximate shape electrode for a light-emitting device according to a third exporting the present invention,

Fig. 18 is a cross-sectional view taken along line 18-18 of Fig. 17,

Fig. 19 is a plan view showing the arrangement of the approximate shape electrode for a light-emitting device according to a fourth exporting the present invention,

Fig. 20 is a sectional view taken along line 20-20 of Fig. 19,

Fig. 21 is a plan view showing the arrangement of the approximate shape electrode for a light-emitting device according to a fifth embodiment of the present invention;

Fig. 22 is a cross sectional view taken along line 22-22 of Fig. 21,

Fig. 23 is a cross-sectional view of a conventional p-type electrode, and

Fig. 24 is a sectional view of a conventional p-type electrode, wherein a window is provided with a Drahtbondingelektrode.

Embodiments of the present invention will now be described under Be described on the drawing.

Fig. 1 is an overall view showing the structure of the electrode for a semiconductor light-emitting device of the invention. In the drawing, the electrode 1 for a light-emitting semiconductor device includes a light-transmissive electrode 10 which is arranged on the upper layer side of a laminated body 3 which forms a light-emitting semiconductor device where it is formed so as to be in contact with the surface of a p conductive semiconductor based on GaN 30 , and a wire bonding electrode 20 which is in electrical contact with the light-permeable electrode 10 and is formed such that it comes into partial contact with the surface of the semiconductor 30 . The wire bonding electrode 20 is formed such that at least one area 20 a, which is in contact with the semiconductor 30 , has a higher contact resistance per unit area with respect to the semiconductor 30 than an area 10 a of the translucent electrode 10 , which is in contact with the semiconductor 30 .

In the embodiment shown in FIG. 1, electrode 1 has two positive and negative electrodes are formed on the side of the semiconductor layer 30 of the body 3. However, the negative electrode is omitted from FIG. 1.

The area 20 a of the wire bonding electrode 20 , which is in contact with the semiconductor 30 , is formed from a metal or an alloy of two or more metals, which are selected from the group consisting of Tl, In, Mn, Ti, Al, Ag, Sn, AuBe, AuZn, AuMg, AlSi, TiSi and TiBe. Subsequent heat treatment can increase the firm adhesion between the electrode 1 and the semiconductor 30 .

Of these metals and alloys, the selection of Ti, TiSi, TiBe, Al, AlSi or AuBe or the like enables stronger adhesion to the surface of the semiconductor 30 to be preserved.

In contrast, the area 10 a of the transparent electrode 10 is formed from a metal that is selected from a group of metals, each having a lower contact resistance relative to the semiconductor 30 , such as Au, Pd, Pt, Ni and Cr. The metal can be given light transmittance by forming the metal as a thin film having a thickness of 1 nm to 1000 nm.

Fig. 2 shows the contact resistance properties (voltage-current properties) of each metal or alloy with respect to a p-type GaN semiconductor. As shown in FIG. 2, Ni and Au have a low contact resistance with respect to a p-type GaN semiconductor, while the AuBe alloy has an extremely high contact resistance, which makes it suitable as a material for the wire bonding electrode region 20 a shows that is in contact with the semiconductor 30 .

Because the area 20 a, which is in contact with the semiconductor 30, is formed such that it has a higher contact resistance with respect to the semiconductor 30 than the translucent electrode area 10 a, which is in contact with the semiconductor 30 , as described above it is possible to safely prevent current from flowing under the wire bonding electrode 20 , thereby ensuring that all of the current is injected from around the wire bonding electrode 20 into the translucent electrode 10 , from where it enters the laminate 3 and contributes to the light emission effect. That is, no light emission is generated under the wire bonding electrode 20 , so that by not obstructing the light by the wire bonding electrode 20 , substantially all of the generated light can be emitted outward from the translucent electrode 10 (upward with respect to FIG . 1). This enables the power to be effectively used and the light emission efficiency is improved.

The electrode structure consisting of the wire bonding electrode 20 and the transparent electrode 10 can be formed by growing thin films using a vacuum deposition method or the like. The process is very simple, involving only the selection and vapor deposition of the metal materials so that the films can be grown in a short period of time. That is, the current flow under the wire bonding electrode 20 can be securely blocked by a simple structure that can be easily formed without having to carry out complex processes.

In most cases, gold wire is used to connect the power supply to the wire bonding electrode 20 . More specifically, small gold bonding balls are used to effect the connection between the gold wire and the wire bonding electrode 20 using ultrasonic waves, and thus to heat and fuse the bonding balls with the electrode material. The electrode materials that fuse with the bonding balls are limited, with Au and Al being known materials. If a metal or alloy is used which is suitable for making contact with the semiconductor 30 but does not fuse well with the bonding balls, a multilayer structure for the wire bonding electrode 20 can be used to provide the surface of the wire bonding electrode 20 provided, which is formed from a metal that melts well with the bonding balls ver.

Fig. 3 shows a Drahtbondingelektrode which has a two-layer structure. In the drawing, the wire bonding electrode 21 is made of a lower layer 21 a formed of AuBe which has a high contact resistance to the semiconductor 30 , while the upper layer 21 b is formed of Au which has good fusion properties with respect to the Has bonding balls.

In cases where there is poor adhesion between the metal used for the upper layer 21b and the metal used for the lower layer 21b , a three-layer structure or a structure with more than three layers can be used to be used. The use of such a multilayer structure makes it possible to realize a wire bond that has both a high contact resistance to the semiconductor 30 and a good fusibility with Bondingku gels.

If a metal such as Al is used, which has a high contact resistance relative to the semiconductor 30 and a high bonding ball-fusibility, the entire wire bonding electrode 20 can be formed as a single layer of Al.

In the arrangements shown in Figs. 1 and 3, the translucent electrode 10 is shown as it is formed around the outer peripheral surface of the wire bonding electrode 20 or 21 and is in contact therewith. However, the translucent electrode 10 may instead be formed so that it overlies the wire bonding electrode 20 or 21 .

Fig. 4 shows an arrangement in which the translucent electrode lies over the entire upper surface of the wire bonding electrode. By forming the translucent electrode 10 to cover the entire upper surface of the wire bonding electrode 20 as shown in the drawing, the translucent electrode 10 , which previously only contacted the side peripheral surface of the wire bonding electrode 20 , now contacts the entire exterior of the wire bonding electrode 20 except for the lower surface, thereby producing a major improvement in the adhesion between the light-transmitting electrode 10 and the wire-bonding electrode 20 . As a result, even if there is little adhesion between the material used to form the translucent electrode 10 and the material used to form the wire bonding electrode 20 , peeling or peeling of the translucent electrode may result 10 can be prevented by the wire bonding electrode 20 .

Even if there is some deviation of the pattern during mask alignment, the contact between the translucent electrode 10 and the wire bonding electrode 20 is not affected because the translucent electrode covers the entire wire bonding electrode except for its lower surface.

In addition, because the transparent electrode 10 trode the Drahtbondingelek 20 downwards in the direction of the semiconductor 30 presses, is the of adhesion between the Drahtbondingelektrode 20 and the semiconductor 30 increases, which also prevents the Drahtbondingelektrode 20 separates from the semiconductor 30th

If the translucent electrode 10 is to be formed such that it covers the entire upper surface of the wire bonding electrode 20 and if the translucent electrode 10 is formed of a material which has good bonding ball fusibility, ie good bonding properties, such as Au or Al, can the Drahtbondingelektrode 20 and the portion of the transparent electrode 10 on the upper side of the Drahtbondingelektrode 20 as an integrated Drahtbondingelek trode 20 are considered, and its upper surface can be used for the receipt, as described above.

On the other hand, when the translucent electrode 10 is formed of a material having poor bonding properties, an arrangement such as that shown in Fig. 5 can be used in which a portion of the translucent electrode 10 that cuts the middle portion of the wire bonding electrode 20 is removed, leaving only the portion of the translucent electrode 10 around the circumference of the upper surface of the wire bonding electrode 20 , thereby exposing the wire bonding electrode 20 where the central portion of the wire bonding electrode 20 has been removed. Bonding to the wire bondingelektrode 20 is effected at the exposed portion, or the exposed portion may also be coated properties with a material having good Bondingei, such as Au or Al, with a coated From section 20 c is formed of the integral as a part the wire bonding electrode 20 can be viewed, which allows the upper surface of the coated portion 20 c to be used for bonding.

The use of such an arrangement makes it possible to ensure good ticket properties even if the translucent electrode 10 is formed of a material with poor bonding properties. Even if the portion of the translucent electrode 10 is removed, the translucent electrode 10 still remains in contact around the upper surface of the wire bonding electrode 20 , so that when the translucent electrode 10 is placed over the wire bonding electrode 20 , preventing loosening, the reduction in the impact of pattern misalignment and other such effects can be preserved.

The use of a two-layer design for the transparent electrode 10 will now be explained.

Fig. 6 shows a transparent electrode 10 having a two-layer structure. In Fig. 6, a translucent electrode 11 consists of a first layer 11 a made of a translucent metal, which is formed on the surface of the semiconductor 30 , and a second layer 11 b, which contains a translucent metal oxide and on the first layer 11 a ge is formed.

The first layer 11 a, which contacts the p-type semiconductor 30 , can be formed from a metal that, when heat-treated, provides good ohmic contact, the metal being selected from a first group consisting of Au, Pt and Pd exists. The first layer 11 a can also be formed from an alloy of at least two of the metals from the first group.

An alloy can be used to achieve good ohmic contact can be used, which is obtained by adding to that described above Metal a small amount of at least one metal, such as Zn, Ge, Sn, Be or Mg, as an impurity is added.

The metal oxide contained in the second layer 11 b is an oxide which has a relatively good light transmittance and a superior adhesive property to the first layer 11 a, and it can be used an oxide of at least one metal made of a second group is selected, which consists of Ni, Ti, Sn, Cr, Co, Zn, Cu, Mg and In. Of these, it is widely known that NiO, TiO ₂ , SnO, Cr ₂ O ₃ , CoO, ZnO, Cu ₂ O, MgO, In ₂ O _{3 are} translucent, and those which comprise an oxide are mainly useful, in which the above-described metal oxide and another metal element together are present. It is also preferred to use an oxide that has good adhesion to the backing of the first layer 11 a.

The term "metal oxide" as used in the present invention det refers to a mixture of oxides in which the Oxidation number of the metals differs, and includes the case in which a non-oxidized metal is included. A metal that is not oxidized can be included among them. The second layer is translucent casualness, and accordingly it is of course an advantage that the oxides, the composition of which differs, that most translucent material as the main ingredient is used.

This is described below, taking Ni as an example. Known oxides of Ni include NiO, Ni ₂ O ₃ , NiO ₂ and Ni ₃ O ₄ . Each be contained by the sen or a mixture thereof may be used as the composition of the material forming the second layer 11b, or it may even Ni, which is a non-oxidized metal. However, of these several types of oxide, it is known that NiO is most effective in promoting light transmittance, and therefore a second layer comprising NiO as a main component is advantageous.

With conventional translucent electrodes, which consist of a very thin individual metal film layer are formed, causes a heat  act to make ohmic contact with the backing (semiconductor) realize a phenomenon called "ball up", which the metal coagulates into a ball due to the fact that the upper surface tension of the metal is greater than its adherence to the backing is. This spherical formation phenomenon occasionally creates gaps and Cracks in the thin metal film, resulting in a loss of electrical Connection and a loss of function as a translucent Leads electrode.

It is conceivable that as a means to the ball formation phenomenon prevent the thickness of the metal electrode from being enlarged. However, leads this is a reduction in transmission, resulting in that Electrode will lose its transmission properties.

Therefore, one of the objects of the present invention is to transmit light through casual electrode for a semiconductor light emitting device provide transmission properties and a structure that in is able to effectively prevent the ball formation phenomenon, and a To provide a method for manufacturing the translucent electrode len.

Therefore, in the arrangement of this embodiment, the light-transmissive electrode 11 is formed as a layer arrangement consisting of the second layer 11 b, which is formed of a metal oxide, which is coated on the first layer 11 a of metal, which is on the semiconductor 30 ge is formed. This makes it possible to prevent the ball formation phenomenon that occurs when a conventional single-layer structure is used. This makes it possible to achieve a major improvement in the ohmic properties between the transparent electrode 11 and the semiconductor 30 , and it is also possible to achieve a substantial increase in the two-way bonding strength.

Because the second layer 11 b formed of metal, also good permeability can be imparted to see the entirety of the light transmitting electrode 11 with a superior permeability to ver.

The first layer 11 a, which comprises a metal, and the second layer 11 b, which comprises a translucent metal oxide, each preferably have good adhesion properties. For this purpose, the electrode preferably has a structure so that the oxygen content gradually decreases from the second layer 11 b in the direction of the first layer 11 a in the region near the interface between the second layer 11 b and the first layer 11 a, so that the composition undergoes a continuous change from the composition containing a metal oxide to the composition comprising a metal.

In order to achieve strong adhesion between the first layer 11 a and the second layer 11 b, the first layer 11 a preferably contains a metal component of the metal oxide which is contained in the second layer 11 b. The constituent of the second layer 11 b can be contained in the first layer 11 a in a constant concentration in the entire first layer 11 a, or the concentration can have a gradient, so that the concentration along the direction of the interface 11 c with the second layer 11 b in the direction of the surface of the semiconductor 30 becomes smaller. The component of the second layer 11 b can be contained in the entire first layer 11 a, or it can be contained in only part of the side of the interface 11 c with the second layer 11 b.

The first layer 11 a is preferably formed such that it has a thickness of 1 nm to 500 nm in order to obtain a light transmittance. It is preferable to adjust the layer thickness so that light transmission is obtained which is calculated from the absorption coefficient as a value of a physical property inherent in a metal from 10% to 90%.

The second layer 11 b preferably has a thickness of 1 nm to 1000 nm, whereby a light transmittance is realized, an excellent spherical protective effect is achieved and good light transmittance is achieved. It is preferred to ensure that the translucent electrode 11 , which consists of the first layer 11 a and the second layer 11 b, has a transmission of at least 10%, and particularly preferably at least 30%.

The above two-layer translucent electrode 11 can be manufactured by one of two methods. The first method involves the use of conventional resistive heat deposition, electron beam heat deposition, sputtering, or other such method to form the lower layer of metal from the first group and the upper layer of metal from the second group. At this stage, the thin film comprising each of the two layers has a dark color with a metallic sheen.

Next, heat treatment in an oxygen-containing atmosphere is used to oxidize the upper layer, which is made of a metal from the second group. An oxygen-containing atmosphere means an atmosphere containing oxygen gas (O ₂ ) or steam (H ₂ O) or the like. This heat treatment oxidizes the metal from the second group from which the upper layer is formed, making it a translucent metal oxide.

By means of the above method, starting from the side of the semiconductor 30 , a translucent first layer 11 a with good ohmic contact with the semiconductor 30 and a second layer 11 b ge, which consists of metal oxide with a high light transmittance, thereby forming the To form transparent electrode 11 , which has a two-layer structure.

In the first manufacturing method, the heat treatment diffuses the metal from the second group, which is used to the metal oxide of the second layer 11 to form b, effectively, resulting in the first layer 11 a to a two-layer structure with a good adhesiveness.

If metals from the first and second group are selected, which can readily be alloyed by heat treatment, the Wär can mebehandlung in an atmosphere containing oxygen at the same time as the heat treatment thereof, which is used to oxidation of the second layer 11b to a metal oxide to cause, and as the heat treatment, which is carried out to diffuse the metal component from the second group, which comprises the second layer 11 b, into the first layer 11 a. Since a small amount of the metal from the second group, which is diffused into the first layer 11 a, the interface 11 d between the semiconductor 30 and the first layer 11 a it reaches and reacts on the surface of the semiconductor 30 with the oxide layer , this can be given the function of destroying the oxide layer.

This can be used to effectively remove the oxide layer, which deteriorates the properties of the contact between the semiconductor 30 and the first layer 11 a and prevents the flow of current. As a result, the heat treatment used to oxidize the metal from the second group can also serve as the heat treatment to improve the properties of the contact between the translucent electrode 11 and the semiconductor 30 .

The second method of manufacturing a transparent electrode having a two-layer structure will now be described with reference to FIG. 7. First, a thin film 11 m is formed on the surface of the semiconductor 30 , the thin film 11 m being an alloy comprising a metal from the first group which has low reactivity with oxygen and a metal from the second group which has reacted with oxygen to form the translucent metal oxide ( Fig. 7 (a)). The thin film 11 m can be formed by ordinary resistance heating deposition, electron beam heating deposition, sputtering or another such method. At this stage, the thin film 11 m has a dark color with a metallic sheen.

Next, as shown in Fig. 7 (b), the thin film 11 m is subjected to heat treatment in an oxygen-containing atmosphere to initiate oxidation of the metal from the second group and a metal oxide film on the surface of the thin film 11 m to build. As in the first manufacturing method, an oxygen-containing atmosphere means an atmosphere containing oxygen gas (O ₂ ) or steam (H ₂ O) or the like.

This method produces the two-layer translucent electrode 11 by separating the thin film 11 m into a translucent first layer 11 a, which comprises a metal from the first group, which is in contact with the semiconductor 30 and produces a good ohmic contact, and in a translucent second layer 11 b, which comprises an oxide of the metal from the second group, which covers the surface of the first layer 11 a be.

When the second manufacturing method is used, the alloy thin film 11 m formed on the surface of the semiconductor 30 contains , as a component, the metal from the second group, which is highly reactive with oxygen. Thus, in the heat treatment process, the oxide layer on the surface of the semiconductor 30 is destroyed, which leads to good electrical contact properties between the metal from the first group and the semiconductor 30 . At the same time, the metal from the second group reacts with vapor phase oxygen which has diffused extensively onto the surface of the thin film 11 m, whereby it becomes a translucent metal oxide which is fixed on the surface of the thin film 11 m so that the second layer 11 b is formed. Thereby, the heat treatment used to oxidize the metal from the second group also serves as a heat treatment to improve the properties of the contact between the transparent electrode 11 and the semiconductor 30 .

When using the second manufacturing method, it is desirable to Select metals from the first and from the second group, too  can be used together as an alloy to make a thin to form film.

The term "alloy" as used herein does not mean only metals that are connected at the atomic level, but also a mixture or a mixture of fine crystal grains. Thereby For example, a sputtering target that consists of a mixture of two Metals are used to make a mixture of crystals two metals to adhere to a substrate, even if the metals are are not an alloy when they are fused together the. The word "alloy" as used herein also includes this type of a mixture of fine crystals.

The mixing ratio of the metals from the first and second groups in the alloy 11 m, which includes the thin film 11 m, can be determined by calculating back from the thickness ratios of the layers which have been formed after the heat treatment.

In the above first and second manufacturing methods, the concentration of oxygen in the atmosphere in which the heat treatment is carried out must be determined based on the properties of the metal from the second group to be oxidized. Based on various studies, it was found that whichever molecules were used to introduce the oxygen atoms, the translucent electrode 11 could not consistently show the translucency when the oxygen concentration was less than 1 ppm. Therefore, it has been demonstrated that it is necessary that the heat treatment atmosphere contain at least 1 ppm oxygen.

The oxygen content of the heat treatment should be particularly preferred atmosphere must be less than 25%. More than 25% oxygen can be too damage to the GaN-based compound semiconductor during the Heat treatment process.

The temperature and the time of the heat treatment must be selected from the second group according to the metal to be oxidized. According to studies by the inventors, metals from the second group could not be completely and evenly oxidized, however long the heat treatment lasted if the heat treatment temperature was below 300 ° C. On the other hand, the metals could be oxidized uniformly in less than an hour at a heat treatment temperature of 450 ° C or higher. Since the metal can be continuously oxidized at higher heat treatment temperatures, any temperature above 300 ° C can be used. However, a temperature should of course be used which does not cause the semiconductor 30 to decompose.

Regardless of how high the temperature is within the above range is set, he can not complete and uniform oxidation be sufficient if the heat treatment time is less than 1 minute. Accordingly, the heat treatment must be for 1 minute or longer be performed.

The thin films can be placed in an oven under normal atmospheric conditions Pressure conditions or heat treated at a lower pressure will. However, a pressure of at least 1 torr is desirable. If the pressure is less than 1 torr, it is difficult during the heat  treatment maintains a high oxygen concentration in the furnace to get what makes it impossible to constantly transmit light to reach.

The "lift-off method" can be used as the patterning method used to form the shape of the light-transmissive electrode 11 , or a method that includes forming a thin metal film over the entire surface that a resist is used to form a negative of the pattern on the thin metal film, and then an etchant is used to etch away the exposed portions of the thin metal film.

Forming a layer of Ni on the semiconductor surface followed by a layer of Au on the Ni layer, and then the implementation heat treatment to reverse the inversion of the element distribution effecting depth is an example of a known conventional one Process used to build a two-layer structure the. This two-layer structure helps to make ohmic contact with the Increase semiconductor surface, reduce the resistance and the Increase bonding strength.

However, the heat treatment must be carried out at a high temperature be led to a hike up the Au to the surface of the Electrode layer and to cause diffusion over a wide range ken, and to cause an inversion of the element distribution. As one As a result, it is difficult to control the diffusion reaction constantly and egg a stable quality for the ohmic properties and light transmission to achieve casualness required for the translucent electrode are. In contrast, the two manufacturing methods above can  of this invention with only a small degree of diffusion what makes it possible for the diffusion reaction to be still low to control their temperatures, and can be used to translucent electrodes with a constant quality over one to reach yet another range of heat treatment temperatures.

As an example, the first manufacturing method was used to manufacture a transparent electrode 11 using Au for the first layer 11 a and Ni for the second layer 11 b, and the transparent electrode 11 was compared with an electrode formed by the above conventional methods had been prepared. The results showed that in both electrodes the transmission was 10% immediately after the deposit and rose to 50% after the heat treatment at 550 ° C. However, when the heat treatment temperature was reduced to 450 ° C, the transmission increased to 30% in the case of the conventional electrode after the heat treatment, while in the case of the electrode made by the first method, the transmission after the heat treatment rose to 50%.

The second manufacturing method was used to manufacture a translucent electrode 11 using a Ni-Au alloy thin film 11 m, likewise for comparison with an electrode manufactured by the same conventional method. Again, the results showed that the transmission in both electrodes was 10% immediately after the deposition and rose to 50% after the heat treatment at 550 ° C. However, when the heat treatment temperature was reduced to 400 ° C, in the case of the conventional electrode, the transmission after the heat treatment only increased to 15%, while in the case of the electrode made by the second method, the transmission increased the heat treatment rose to 50%.

Thus, the use of the first and second manufacturing methods according to the present invention makes it possible to manufacture good quality translucent electrodes 11 over a wide range of heat treatment temperatures.

The arrangement comprising the light-transmissive electrode 11 formed to overlie the wire-bonding electrode 20 ( 21 ) will now be explained.

Fig. 8 shows a two-layer transparent electrode 11, wherein the transparent electrode is located over the entire top surface of the Drahtbondingelektrode 20th The forming of the transparent elec trode 11 so that the entire upper surface of Drahtbondingelek trode 20 covers, makes it possible the adhesion between the first layer 11 a and the Drahtbondingelektrode 20 to improve. This makes it possible to prevent the translucent electrode 11 from peeling off or peeling off from the wire bonding electrode 20 even if there is little adhesiveness between the material of the first layer 11 a and the material of the wire bonding electrode 20 .

As with the arrangement shown in Fig. 4, any pattern deviation that might occur during the mask alignment procedure will not affect the contact between the translucent electrode 11 and the wire bonding electrode 20 because the translucent electrode 11 will cover the entire wire bonding electrode 20 except its bottom surface covered.

In addition, since the Drahtbondingelektrode 20 is pushed from the transparent electrode 11 down toward the semiconductor 30, the adhesion is increased between the Drahtbondingelektrode 20 and the semiconductor 30, which also helps to prevent that the wire bonding to electrode 20 of the semiconductor 30 peels off.

Covering the wire bonding electrode 20 with the two-layer translucent electrode 11 enables the effects explained below to come to light. From the first layer 11 a and the second layer 11 b, which form the translucent electrode 11 , the second layer 11 b is a metal oxide with a low conductivity, wherein in the arrangement shown in FIG. 6, in which the translucent electrode 11 around and around the outer peripheral surface of the wire bonding electrode 20 is formed, the current that enters the transparent electrode 11 from around the wire bonding electrode 20 flows only through the thickness portion of the first layer 11 a, so that the conductivity is halved is. In contrast, because the translucent electrode 11 covers the entire wire bonding electrode 20 except for its lower surface, the area of the electrical contact with the first layer 11 a can be greatly expanded. This greatly improves the conductivity of the current from the wire bonding electrode 20 into the translucent electrode 11 with a corresponding reduction in resistance, thereby allowing wasteful energy consumption to be reduced.

When the transparent electrode 11 is formed so as to be 20, and if the second layer 11 b of the transparent electrode 11 is formed of a metal having good bonding properties which ge entire upper surface of the Drahtbondingelektrode covered wire can bondingelektrode 20 and the portion of the light-transmissive electrode 11 are considered on top of the Drahtbondingelektrode 20 as an integrated Drahtbondingelektrode 20, and the upper surface of its second layer 11 b can be used for bonding as described above.

In most cases, the metal comprising the second layer 11 b, poor bonding properties, which make it difficult to bond to the layer. If this is the case, an arrangement such as that shown in FIG. 9 can be used, in which a portion of the second layer 11 b, which corresponds to the central portion of the wire bonding electrode 20 , is removed, only that Part of the second layer 11 b remains around the circumference of the upper surface of the wire bonding electrode 20 , whereby the first layer 11 a is exposed at this section.

If the first layer 11 a is formed from a material with good bonding properties, this section of the first layer 11 a can be regarded as an integral part of the wire bonding electrode 20 , thereby permitting the exposed upper surface of the first layer 11 a to Bonding is used. If the electrode height is not sufficient, Al or Au, which have good bonding properties, can be used as a layer arrangement on the first layer 11 a, and the coated section 20 c can be regarded as an integral part of the wire bonding electrode 20 , and the upper surface of the coated section 20 c can be used for bonding.

If both the first layer 11 are a and the second layer 11b formed of a material having poor bonding properties, the first layer 11 are removed a same time b to the second layer 11, wherein only the part of the transparent electrode 11 is left, of the wire bonding to the periphery of the top surface electrode 20 is, as shown in Fig. 10, the upper surface of the Drahtbondingelektrode 20 is partially exposed. The exposed portion of the wire bonding electrode 20 can be used for bonding, or the exposed portion can be coated with a material such as Au or Al that has good bonding properties, and the coated portion 20 c can then be an integral part of the wire Bonding electrode 20 can be considered, and the upper surface of the coated portion 20 c can be used for bonding.

The use of this arrangement makes it possible to ensure good bonding properties, even if the second layer 11 b or both the second layer 11 b and the first layer 11 a are formed from a material with poor bonding properties.

Even if only the second layer 11 b is removed, the first layer 11 of the Drahtbon remains a yet in contact to the upper surface ding electrode 20 around, so that when the transparent electrode is 11 through the Drahtbondingelektrode 20, moreover, the loosening prevention, the reduction in the impact of pattern misalignment and other such effects can be preserved. Even if both the second layer 11 b and the first layer 11a are removed, but the transparent electrode 11 remains in contact around the periphery of Drahtbondingelektrode 20 around, so that when the transparent electrode is 11 on the Drahtbondingelektrode 20, and the Prevention of loosening, reducing the impact of pattern misalignment, and other such effects can be maintained.

Examples of the light emitting electrode will now be described semiconductor devices according to the present invention fen. However, it is to be understood that the invention is not limited to the examples is limited.

example 1

Fig. 11 is a plan view showing the arrangement of the electrode for a semiconductor light emitting device which is a first example of the present invention, and Fig. 12 is a cross sectional view taken along the line 12-12 of Fig. 11. In the drawing an electrode 1 A for a light-emitting semiconductor device (hereinafter also referred to simply as a "device electrode") on a laminated body 3 A is arranged. The device electrode 1 A and the laminate 3 A form ei ne light emitting device. The device electrode 1 A comprises a p-type electrode 101 and an n-type electrode 106th

The laminate 3 A comprises a sapphire substrate on which an AlN buffer layer, an n-type GaN layer, an InGaN layer, a p-type AlGaN layer and a p-type GaN layer 301 in this order are coated. The p-type electrode 101 was formed by the following method on the laminated body 3 A.

First, a known photolithography technology was used to form an AuBe layer 211 a of a wire bonding electrode 211 on the p-type GaN layer 301 .

In order to form the wire bonding electrode 211 (AuBe layer 211 a), the laminate 3 A was first placed in a vacuum deposition device (not shown), in the outer layer, which initially contained 1% by weight of Be, to a thickness of 500 nm was deposited over the entire surface of the p-type GaN layer 301 at a pressure of 3 × 10 ^-6 Torr, thereby forming an AuBe thin film. The layer body 3 A, was deposited on the AuBe thin film was then kuumablagerungsvorrichtung from the Va removed, and it was a normal Photo used lithography technique, to form a positive pattern on the resist based on a Drahtbondingelektrode which comprised a resist. Then, the laminated body was immersed in a 3 A Au etchant to the exposed portions of the AuBe thin film to be removed, whereby the AuBe layer was formed 211 a.

According to the following method, a transparent electrode 111 having a two-layer structure was formed by forming a first layer 111 a of Au on the p-type GaN layer 301 and forming a second layer 111 b of NiO on the first layer 111 a has been.

Then photolithography was used to prepare a negative resist pattern on the basis of the light-transmissive electrode 111 to form on the layer body 3 A, on which the Drahtbondingelektrode was completed 211th Next, the laminated body 3 A was placed in the vacuum deposition apparatus in which Au was first formed with a thickness of 20 nm on the p-type GaN layer 301 under a pressure of 3 × 10 ^-6 Torr and the formation of deposits of 10 nm of Ni in the same device followed. The layer body 3 A, on which the Au and Ni were deposited was then removed from the Vakuumablagerungsvor direction behan punched to form a two-layer thin film made of Au and Ni in a Ge desired shape removed and replaced with a common lift-off method.

The laminated body 3 A was then placed in a heat treatment furnace at a temperature of 550 ° C for 10 minutes in an atmosphere of flowing argon containing 1% of oxygen gas, heat-treated. Was removed as the layer body 3 A from the oven, the two thin films 111 and 111 were a b of Au and Ni on the layer A body 3 of a dark bluish gray and showed light transmittance. This was how the two-layer light-transmissive electrode 111 was formed. This heat treatment also served as a heat treatment to maintain ohmic contact between the light-transmissive electrode 111 and the p-type GaN layer 301 and as a heat treatment to prevent the adhesion between the wire bonding electrode 211 (AuBe layer 211 a). and to improve the p-type GaN layer 301 .

The translucent electrode 111 thus produced by the above method showed a transmission of 45% in the case of light with a wavelength of 450 nm. This transmission was measured on a sample whose structure was identical to that of the translucent electrode 111 produced, but which was of a size suitable for measurement.

Auger electron spectroscopy (AES) was used to analyze components in the direction of the depth of the translucent electrode 111 . This showed that there was no great difference in the thickness of the transparent electrode 111 between before and after the heat treatment, but the AES showed that a large amount of oxygen was taken up by the second layer 111b , which caused the oxidation of the Ni . Fig. 13 is a profile in the direction of the depth of the respective elements of the electrode as measured by the AES.

The profile of the electrode composition in the direction of depth shown in Fig. 13 shows that the second layer 111 b is made of an oxide of Ni containing Ni and oxygen, and that the first layer 111 a is made of Au with a low Ni Content exists, and that there is a gradient region of the composition R1 in the area near the interface between the first layer 111a and a second layer 111b , in which the oxygen concentration gradually decreases as one goes in the direction of the substrate .

The second layer 111 b was evaluated using thin film X-ray diffraction (XRD) and was found to have the spectrum shown in FIG. 14. From the peak positions, it is known that the peaks P1, P2, P3 and P4 correspond to the diffraction of the ( 111 ), ( 200 ), ( 220 ) and ( 311 ) faces of the NiO, respectively, which shows that the second layer 111 b consists of randomly oriented crystals of NiO. In the spectrum, a weak diffraction peak P6 was also detected from the ( 111 ) face of the Ni. Diffraction peaks P5 and P7 were also found from the Au ( 111 ) and ( 220 ) surfaces which form the first layer 111 a. This is considered to indicate that there is an accumulation of NiO crystal grains in which a small amount of Ni crystal grains are mixed. This confirmed that the second layer 111 b was made of NiO and a small amount of Ni.

Next, a known photolithography technology was used to form a pattern resist and thus expose a portion of the wire bonding electrode 211 (AuBe layer 211 a). The layer body 3 A was then immersed ge in concentrated hydrochloric acid to the exposed portion of the NiO layer to be removed. In this way, the NiO of the second layer 111 b was removed in a region of the AuBe layer 211 a, thereby exposing the Au layer which formed the first layer 111 a.

The layer body 3 A was then placed in a vacuum deposition apparatus, and it was vacuum deposition used in the same manner as in the deposition of the AuBe layer 211 a to a to form Au film nm with a thickness of 500th This vapor deposition process created a fusion with the first layer 111 a, which formed the backing, the Au, which was then deposited by steam, was integrated with the Au of the first layer 111 a. The laminated body 3 A was removed from the device and treated by a lift-off procedure, whereby the wire-bonding electrode 211 was completed, which had a structure that, starting from the side of the laminated body 3 A, from the AuBe layer 211 a and the Au layer 211 b.

As a result, the p-type electrode 101 was formed, which includes the transparent electrode 111 and the wire bonding electrode 211 . The Au that was used for the first layer 111 a is a metal that provides good ohmic contact with the p-type GaN layer 301 . The presence of the NiO used to form the second layer 111 b could prevent the spheroidization phenomenon from occurring. The AuBe layer 211 a is an alloy that forms a contact with high resistance with the p-type GaN layer 301 .

Then dry etching was used to partially expose the n-type GaN layer of the laminate 3 A and thus form an n-type electrode de 106 , and then the n-type electrode 106 was formed from Al on the exposed area and heat treated, to cause ohmic contact of the n-type electrode 106 .

The wafer, which was provided with device electrodes 1 A, the p-type electrode 101 and n-type electrodes 106 having, was then cut into 400-micron square chips, which were then mounted on a lead frame and connected to the lines, to thereby form light emitting diodes. The light emitting diode showed an emission output of 80 µW at 20 mA and a forward voltage of 2.8 V. There was no peeling off of the wire bonding electrode 211 during the bonding process. 16,000 chips were obtained from the wafer, which was 2 inches in diameter. Chips with an emission intensity of less than 76 µW were removed, which resulted in a yield of 98%.

Example 2

Fig. 15 is a cross-sectional view trode the structure of the Vorrichtungselek shows, which is a second example of the present invention, wherein Fig. 15 (a) shows the first stage, Fig. 15 (b) the second stage, and FIG. 15 (c) shows the final state. The difference between the first and second examples is that the two-layer device electrode 112 of the second example is formed as a single-layer alloy thin film layer 112 m.

The device electrode 1 B of this example is arranged on a laminated body 3 B, which has the same structure as the laminated body 3 A of the first example. The device electrode 1 B and the layer body 3 B form a light-emitting device. The device electrode 1 B comprises a p-type electrode 102 and an n-type electrode 107 .

In order to form the p-side electrode 102 on the p-type GaN layer 302 that is disposed on the layer body 3 B The following procedure is used.

First, a known photolithography technique was used to form a 500 nm thick AuBe layer 212 a from a wire bonding electrode 212 on the p-type GaN layer 302 .

The alloy thin film layer 112 m, which was made of an alloy of Au and Ni, was formed as follows just at the area where the light-transmitting electrode 112 was formed on the p-type GaN layer 302 ( Fig. 15 (a)) .

First, the laminate 3 B was placed in a vacuum deposition apparatus in which an alloy of Au and Ni was deposited on the p-type GaN layer 302 under a pressure of 3 × 10 ^-6 Torr. For this process, pieces of Au and Ni were placed on a resistance heating tungsten boat in the volume ratio of Au: Ni = 2: 1. After the boat was heated by the passage of a current, it was confirmed that the metal was melted, and was after lan ge maintained enough to ensure that the metals were mixed from reaching the seal between the Schichtkör per 3 B was and opened the boat to begin vapor deposition and form a 30 nm thick AuNi alloy thin film layer 112 m.

The laminate 3 B on which the alloy thin film layer 112 m was formed was then removed from the vacuum deposition device and treated according to an ordinary lift-off procedure to form the alloy thin film 112 m in a desired shape, where in this manner and how a single-layer alloy thin film layer 112 m was formed on the p-type GaN layer 302 . Thin film X-ray diffraction was used to confirm that the alloy thin film layer 112 m was formed from an AuNi alloy. The 112 m alloy thin film layer was dark gray and had a metallic sheen. Almost no light transmission was observed.

The laminate 3 B was heat-treated in a heat treatment furnace at a temperature of 500 ° C. for 10 minutes in an atmosphere of streaming argon containing 20% oxygen gas. When the laminate 3 B was removed from the furnace, the alloy thin film layer was 112 m of a dark bluish gray and showed a light transmittance after being the transparent electrode 112 ( Fig. 15 (b)).

The translucent electrode 112 thus produced by the above method showed a transmission of 45% with respect to light with a wavelength of 450 nm. A measurement of the transmission and the thin film X-ray diffraction measurement described below were carried out using translucent electrodes 112 which had been made to a size suitable for measurement applications.

AES was used to measure the profile of each element toward the depth of the translucent electrode 112 . The results are shown in Fig. 16. It was found that the light-transmissive electrode 112 after the heat treatment in a first layer 112 a, which consisted of substantially pure Au, which was in contact with the p-type GaN layer 302 , and in the second surface layer 112 b which was composed of an oxide of Ni. There is a gradient region of the composition R2 in the region near the interface between the first layer 112 a and the second layer 112 b, in which the oxygen concentration gradually decreases as one goes in the direction of the substrate. It was also found that the first layer 112 a, which consisted of Au and was in contact with the p-type GaN layer 302 , contained essentially no Ni.

Trace amounts of Ga were also detected in the first layer 112 a, indicating that during the initial phase of the heat treatment, the Ni destroyed a Ga oxide layer that was present on the surface of the p-type GaN layer 302 .

When evaluating the second layer 112 b after the heat treatment using the thin film X-ray diffraction, it was found that the second layer 112 b was made of NiO and a small amount of Ni.

Next, as in the first example, uses a known Photoli thographietechnologie to expose a portion of the second layer 112 b on the Drahtbondingelektrode 212 (AuBe layer 212 a) to entfer NEN. The layer body 3 B was then set direction in a Vakuumablagerungsvor, and it was used vacuum deposition to a form Au layer nm with a thickness of 500, and then Ver application of a normal lift-off procedure treated, to thereby form a Drahtbondingelektrode To complete 212 , which had a multilayer structure, which, starting from the side of the laminate 3 B, consisted of an AuBe layer 212 a and an Au layer 212 b, whereby a p-side electrode 102 was formed, which consists of the translucent electrode 112 and wire bonding electrode 212 .

Then, an n-type Al electrode 107 was formed on the exposed surface and heat-treated to cause ohmic contact of the n-type electrode 107 .

The wafer, which was provided with device electrodes 1 B, the p-guide de electrodes 102 and n-type electrodes 107 having, was then cut into 400-micron square chips, which then mounted men at a Leitungsrah and connected to the lines to thereby form light emitting diodes. The light emitting diodes each had a lighting output of 80 µW at 20 mA and a forward voltage of 2.9 V. There was no peeling of the wire bonding electrode 212 during the bonding process. 16,000 chips were obtained from the wafer, which was 2 inches in diameter. Chips with an emission intensity of less than 76 µW were removed, which resulted in a yield of 98%.

Example 3

Fig. 17 is a plan view showing the arrangement of a third example of the device electrode of the present invention, and Fig. 18 is a cross sectional view taken along the line 18-18 of Fig. 17. The difference between this third example and the first example that the wire bonding electrode 213 has a three-layer structure, and that the layers 113 a and 113 b, which form the translucent electrode 113 , are each formed from Pd or SnO.

The device electrode 1 C in this example is arranged on a layer body 3 C, which has the same structure as the device electrode 1 A in the first example. The device electrode 1 C and the laminate 3 C together form a light-emitting device. The device electrode 1 C comprises a p-type electrode 103 and an n-type electrode 108 .

The following procedure was used to form p-type to p-type electrode 103 on the GaN layer 303 was disposed on the layer body 3 C.

First, a known photolithography technique was used to deposit Ti on the p-type GaN layer 303 and thus form a Ti layer 213 a of the wire bonding electrode 213 . The Ti layer 213 a was formed by the same method used in the first example.

The following method was used to form a multilayer light-transmissive electrode 113, which comprised a first layer 113 a of Pd formed on the p-type GaN layer 303 and a second layer 113 b of SnO on the first layer 113 a was formed.

Then, photolithography was used to form a negative resist-based pattern of the transparent electrode 113 on the laminate 3 C on which the formation of the wire bonding electrode 213 (Ti layer 213 a) had been completed. Next, the layer body 3 C was placed in the vacuum deposition device in which a first 5 nm thick first layer 113 a of Pd was formed on the p-type GaN layer 303 under a pressure of 3 × 10 ^-6 Torr, the a deposit formation of a 10 nm thick second layer 113 b of Sn in the same device followed. The laminate 3 C on which the Pd and Sn had been deposited was then removed from the vacuum deposition device and treated by an ordinary lift-off method to form a thin film of Pd and Sn in a desired shape.

The laminate 3 C was then heat treated at 500 ° C for 60 minutes in an oxygen gas atmosphere in a heat treatment furnace. The heat treatment had a dual purpose, namely to oxidize the Sn of the second layer 113 b to form transparent SnO and to form an ohmic contact between the first layer 113 a and the p-type GaN layer 303 . In this heat treatment, the translucent electrode 113 was formed, which comprised the first layer 113 a made of Pd and the second layer 113 b made of SnO.

To remove the portion of the SnO from the translucent electrode 113 that overlapped the wire bonding electrode 213 (Ti layer 213 a), photolithography was used to form a pattern resist for exposing a portion of the translucent electrode 113 that is the Ti layer 213 a overlapped, followed by immersion in acid. An acid was used which only dissolved the SnO without dissolving the Pd. In this way, the second layer 113 b of SnO was completely removed in a part of the transparent electrode 113 , which overlapped the Ti layer 213 a, the first layer 113 a of Pd being exposed.

After the area had been placed by the removal of the second layer 113 b free, a Re was formed sistmuster next to another area. The laminate 3 C was placed in a vacuum deposition apparatus, 36941 00070 552 001000280000000200012000285913683000040 0002019820777 00004 36822, and vacuum deposition was used to form an Au layer by the same method used to form the Ti layer 213 a. the laminate 3 C was removed from the device and heat-treated by a lift-off process. This left only the Au that was deposited on the first layer 113a when the backing was exposed. This completed the wire bonding electrode 213 , which had a multilayer structure in the form that, starting from the side of the laminate 3 C, had the Ti layer 213 a, the Pd layer 213 b and the Au layer 213 c.

Thereby, the p-type electrode 103 was formed, which comprised the transparent electrode 113 and the wire bonding electrode 213 . The SnO used for the second layer 113b is known as a conductive oxide, however, it has a high resistivity compared to metal, which makes it difficult to achieve good continuity even when it is Metal is contacted. That is, at the portion of the wire bonding electrode 213, the second layer 113 b is removed to improve the bonding properties. The Pd that is used for the first layer 113 a is a metal that provides good ohmic contact with the p-type GaN layer 303 . The Ti, which is used for the lower layer 213 a of the wire bonding electrode 213 , is a metal which forms a contact with high resistance with the p-type GaN layer 303 .

Dry etching was used to partially expose the n-type GaN layer of the laminate 3 C and thus form an n-type electrode de 108 , and a Ti layer 108 a was formed on the exposed portion, and it became one Al layer 108 b formed on the Ti layer 108 a, and a heat treatment was used to bring about an ohmic contact with the n-type GaN layer and to form through the two-layer n-type electrode 108 .

The wafer, which was formed in this way with the device electrodes 1 C, a p-type electrode 103 and an n-type electrode 108 had, was then cut into 400-micron square chips mounted on lead frames and connected to the lines were connected to form light emitting diodes. These light emitting diodes showed a lighting output of 80 µW at 20 mA and a forward voltage of 2.8 V. There was no peeling of the wire bonding electrode 213 during the bonding process. 16,000 chips were obtained from the wafer, which was 2 inches in diameter. Chips with an emission intensity of less than 76 µW were removed, which resulted in a yield of 98%.

Example 4

Fig. 19 is a plan view showing the arrangement of a fourth example of the device electrode according to the present invention, and Fig. 20 is a cross-sectional view taken along the line 20-20 of Fig. 19. The wire bonding electrode 214 of this fourth example has a three-layer structure and the transparent electrode 114 is of a two-layer type made of alloy thin film monolayers of Pt and TiO ₂ .

The device electrode 1 D in this example is arranged on a layer body 3 D which has the same structure as the device electrode 1 A in the first example. The device electrode 1 D and the laminate 3 D together form a light-emitting device. The device electrode 1 D includes a p-type electrode 104 and an n-type electrode 109 .

The following procedure was used to form the p-type electrode 104 on the p-type GaN layer 304, which was placed on the laminated body 3 D.

First, a known photolithography technique was used to form a pattern of an AuBe layer 214 a in the form of a Drahtbondingelek trode 214 on the form 304 p-type GaN layer.

Then, photolithography and lift-off techniques were used to form a thin film of an alloy of Pt and Ti only on the area of the p-type GaN layer 304 where a transparent electrode 114 was formed.

Regarding the formation of the alloy thin film layer, the laminate 3 D was placed in the vacuum deposition device, and a 20 nm thick alloy thin film layer was formed under pressure using a PtTi alloy target in which the Pt and the Ti were contained in an equal volumetric ratio formed by 3 × 10 ^-6 torr. The laminate 3 D on which the PtTi alloy thin film was deposited was then removed from the vacuum deposition apparatus and treated by an ordinary lift-off method to form the alloy thin film in a desired shape. In this way, a PtTi alloy thin film was formed on the p-type GaN layer 304 . The thin film was silver in color with a metallic sheen and showed essentially no light transmission.

The layer body 3 D was then heat treated in a mebehandlungsofen Wär at 450 ° C for one hour in a nitrogen gas atmosphere containing 10% oxygen gas. When the alloy thin film was observed through an optical microscope after the heat treatment, it was found to have turned yellow and lost its metallic luster, and it showed light transmittance by becoming the light-transmissive electrode 114 .

The translucent electrode 114 thus manufactured by the above method was not changed in thickness by the heat treatment and showed a transmission of 30% with respect to light with a wavelength of 450 nm. Measurements using AES and thin-film X-ray diffraction showed that this Ti had been oxidized in the PtTi alloy, with TiO _{2 being} formed and being deposited on the thin film surface. That is, following the heat treatment, the translucent electrode 114 became a thin film with a layer structure consisting of a first layer 114 a of Pt and a second layer 114 b of Ti oxide.

To remove the portion of the TiO _{2 of} the translucent electrode 114 that overlapped the wire bonding electrode 214 , photolithography was used to form a pattern resist for exposing the portion of the translucent electrode 114 that overlapped the wire bonding electrode 214 and the laminate 3 D was then immersed in acid. An acid was used that only dissolved the TiO ₂ without dissolving the Pt. In this way, the second layer 114 b of TiO _{2 was} completely removed from a part of the light-transmitting electrode 114 which overlapped the wire bonding electrode 214 , the first layer 114 a of Pt being exposed.

Next, after the area was exposed by removing the TiO ₂ , a resist pattern was formed on another area. The layer body 3 D was set ge in a vacuum deposition apparatus, and it was used vacuum deposition to form a to form an Au layer by the same method that was used to form the Drahtbondingelektrode 214 (AuBe layer 214 a). The layer body 3 D was removed from the device and treated by a lift-off method. This left only the Au, which on the first Pt layer 114 was deposited a, exposed as the backing and completed the Drahtbondingelektrode 214, which had a multilayer structure, 3 D, starting from the side of the laminated body, an AuBe layer 214 a, a Pt layer 214 b and an Au layer 214 c. Thus, the p-type electrode 104 was formed, which includes the transparent electrode 114 and the wire bonding electrode 214 .

As in the first example, an n-side Al electrode 109 was formed and subjected to heat treatment to make ohmic contact cause.

The wafer, which was formed in this way with the device electrodes 1 D, a p-type electrode 104 and an n-type electrode 109 had was cut into 400-micron square chips, which are then attached to egg nem lead frame and with the leads were connected to form light emitting diodes. These light emitting diodes showed an illumination output of 80 µW at 20 mA and a forward voltage of 2.9 V. There was no peeling of the wire bonding electrode 214 during the bonding process. 16,000 chips were obtained from the wafer, which was 2 inches in diameter. Chips with an emission intensity of less than 76 µW were removed, which resulted in a yield of 96%.

Example 5

Fig. 21 is a plan view showing the arrangement of the device electrode according to a fifth example of the present invention, and Fig. 22 is a cross-sectional view taken along line 22-22 of Fig. 21. In this fifth example, the wire bonding electrode 215 is made of Al ge forms, and the layers 115 a and 115 b, which form the transparent electrode 115 , are formed from Au and Cr ₂ O ₃ , respectively.

The device electrode 1 E of this example is provided on a layer body 3 E which comprises a sapphire substrate on which an Al _0.8 Ga _0.2 N buffer layer, an n-type GaN layer, an InGaN layer, the is doped with Zn and Si, a p-type AlGaN layer and a p-type GaN layer 305 are coated in this order. The device electrode 1 E and the laminate 3 E together form a light-emitting device. The device electrode 1 E also includes a p-type electrode 105 and an n-type electrode 110 .

The following procedure was used to form p-type to p-type electrode 105 on the GaN layer 305 was deposited on the layer body 3 E.

The preprocessed laminate 3 E was placed in a vacuum deposition device which was evacuated to a pressure of 3 × 10 ^-6 Torr, and the furnace of the electron beam radiant heater of the device was filled with Al as the source metal.

After confirming the vacuum, the Al became melting is heating. After confirming that the Al is reflected by the electron beam radiation had melted, the shutter between the stove and of the semiconductor substrate chuck opened to the Al deposit process to begin. There became a quartz oscillator type thickness gauge used to monitor the thickness of the film. As the film thickness 1 µm the deposition process was stopped.

The laminate 3 E was removed from the device, and photolithography was used to form a wire-bonding electrode photoresist pattern on the laminate 3 E. The laminated body 3 E was then immersed in an Al etchant to exposed portions that were not covered by the photoresist, remove, and thereby form an Al-Drahtbondingelektrode 215 in the desired shape, after the photoresist had been removed.

Next, photolithography was used to edge a Photoresistmu a transparent electrode layer on the body 3 E to the bil. The laminated body 3 E was set in a vacuum deposition apparatus, which was set at a pressure of 3 × 10 ^-6 Torr, and it uses WUR de The same procedure as described above to form a 10 nm thick layer of Au for the first Form layer 115 a and a 20 nm thick layer of Cr for the second layer 115 b. The laminated body 3 E was then removed from the device and treated by an ordinary lift-off method, whereby a two-layer thin film of Au and Cr was formed in a desired shape.

Then, the laminated body 3 E was heat-treated in a heat treatment furnace at 650 ° C. for 10 minutes in a nitrogen gas atmosphere containing 500 ppm of oxygen. The reaction with the oxygen oxidized the Cr of the second layer 115 b, which led to a translucent electrode 115 with a planar structure which, starting from the side of the layer body 3 E, had a first layer 115 a made of Au and a second layer 115 b Cr ₂ O ₃ existed. To remove the portion of the Cr ₂ O _{3 of} the translucent electrode 115 that overlapped the wire bonding electrode 215 , photolithography was used to form a pattern resist for exposing the portion of the translucent electrode 115 that overlapped the wire bonding electrode 215 and the laminate 3 E was then immersed in hydrochloric acid to dissolve only the Cr ₂ O ₃ . In this way, the second layer 115 b made of Cr ₂ O _{3 was} completely removed from a part of the light-transmitting electrode 115 that overlapped the wire-bonding electrode 215 , the first layer 115 a made of Au being exposed. Thereby, the p-type electrode 105 was formed, which consisted of the transparent electrode 115 and the wire bonding electrode 215 . Then the n-type electrode was formed, and ohmic contact was caused to the n-type electrode. In this way, the p-type electrode de 105 was formed, which comprised the translucent electrode 115 and the wire bonding electrode 215 . As in the first example, an n-type Al electrode 110 was formed and subjected to heat treatment to cause ohmic contact.

The wafer, which was formed in this way with the device electrodes 1 E, having a p-type electrode 105 and an n-type electrode 110 was cut to 400 micron square chips that have been mounted frame on line, and Electrodes were connected to the leads using an ultrasonic wire bonder. During the wire bonding process, essentially no peeling of the wire bonding electrode 215 was observed.

The chips were then sealed in resin packages to light them up tive diodes that have a lighting output of 80 µW at 20 mA and a forward voltage of 2.9 V showed. There were Obtained 16,000 chips from the wafer, which was 2 inches in diameter. Chips with an emission intensity of less than 78 µW were removed distant, which led to a yield of 98%.  

Conventional device electrodes were made, and the pre Directional electrodes of Examples 1 to 5 were made with the conventional ones Device electrodes compared.

Comparative Example 1

The device electrodes of the first comparative example were on egg Nem laminate provided the same structure as that of first example. First a vacuum deposition device device used to form a translucent electrode made of was composed of a single layer of Au, the thickness of which was 25 nm scam. The laminate was subjected to a heat treatment which from heating to 550 ° C for 10 minutes in an argon atmosphere existed for an ohmic contact with a p-type GaN layer to effect. After the heat treatment it turned out that the translucency of the surface of the translucent elec trode had enlarged, but the metallic sheen had disappeared was. The translucent electrode was elec- trode coated. In the case of light-emitting diodes made from this semi-lead Substrate were manufactured, the light emission was only directly below generated the wire bonding electrode, and the light emission was at the Surface of the translucent electrode not visible. An observer attention through an optical microscope showed that the Au thin film, the the translucent electrode formed spherical agglomerations and lacked continuity.  

Comparative Example 2

The device electrodes of the second comparative example were on egg Nem laminate provided the same structure as that of first example, and the same operations were used around a layer of Au on the laminate and on the Au layer a NiO layer and thereby a translucent electrode with a to form a two-layer structure. In contrast to the first comparison For example, a window section was attached to the translucent electrode a position where the wire bonding electrode is placed that should, with the semiconductor substrate exposed at this section has been. Then a lift-off technique was used to get on the window section to form a multi-layer wire bonding electrode an AuBe layer on the side of the semiconductor substrate and one Au layer on top of the AuBe layer. Dry etching was used around the n-type layer at the point of formation of the n-type Expose the electrode. Then an n-type Al electrode was attached to the exposed section formed. It was subjected to heat treatment led to form an ohmic contact for the n-type electrode the.

The wafer thus formed with p-type and n-type electrodes was then cut into 400 µm square chips attached to a lei tion frame and connected to the lines to to form light emitting diodes. These light emitting diodes show a lighting output power of 50 µW at 20 mA, the lower than any of the light emitting diodes of Examples 1-5. With The forward voltage was very high at 15.0 V. This was through the Me Talloxidschicht causes which is the top layer of the translucent  Electrode forms and as an obstacle between the translucent Electrode and the wire bonding electrode acts.

While the inventive examples with reference to light emitting rende diodes have been described, the device electrode of the Invention can also be applied to laser diodes.

The present invention provides as described above was, the following effects.

The area of the wire bonding electrode that is in contact with the semiconductor stands, has a higher contact resistance per unit area in relation on the semiconductor as a region of the translucent electrode, that is in contact with the semiconductor. This makes it possible to get under bind that current flows under the wire bonding electrode, so that the ge entire current from around the wire bonding electrode into the light through casual electrode is injected, which contributes to the light emission effect. That is, there is no light emission under the wire bonding electrode instead, and it can essentially generate all of the light emitted be emitted upwards by the translucent electrode. There through, electricity is effectively exploited and the light emission effect degree is improved.

This electrode design, which is a wire bonding electrode and a having translucent electrode can be easily formed by Thin films using a method such as Vacuum Vapor Depo sition to be grown up. The process is very simple, with only the Vapor deposition of the metal material comprises, so that the formation of Movies can be effected quickly. That is, a current flow under  the wire bonding electrode by means of a simple structure that is easy to ge can be formed, can be blocked safely, without complex processes to do this.

The invention also includes a wire bonding electrode with one more layered structure. This means that if the area of the wire receipt thing electrode, which is in contact with the semiconductor, made of a material is formed with high contact resistance, which is not a good bond egg properties, the surface of the wire bonding electrode with egg nem metal with good bonding properties, where through a wire bonding electrode with a high contact resistance the semiconductor and good bonding properties is realized.

The device electrode of the invention also provides an arrangement riding, in which the translucent electrode made of a second metal oxide layer that is on the first metal oxide layer on the semiconductor side is formed. This makes it possible to prevent bullet formation which occurs in the conventional monolayer arrangement, and verbes thereby the properties of the ohmic contact between the translucent electrode and the semiconductor surface and reinforced the connection between them. Because the second layer is made of a metal oxide is formed, good light transmittance can be given, resulting in a translucent electrode with good translucency throughput.

The second layer, which is formed from metal oxide, makes it possible to to give the second layer a good permeability and an out recorded total permeability of the translucent electrode real.  

The device electrode also includes an arrangement in which the Percentage of oxygen from the second layer towards the first Layer in the area near the interface between the second Layer and the first layer decreases so that the composition a continuous change from a composition that a me contains talloxide, undergoes a composition comprising a metal. This leads to a stronger attachment between the first and the second layer.

The device electrode also includes an arrangement in which the first Layer contains a metal element that is a major component of the metal is oxide that forms the second layer. This also leads to a stronger one Adhesion between the first and second layers.

The device electrode also includes an arrangement in which the light permeable electrode is formed such that it over an upper upper surface of the wire bonding electrode. Because the translucent electrode the entire exterior of the wire bonding electrode except the bottom covered surface, this creates a major improvement in the surface adhesion between the translucent electrode and the wire bonding electrode. Even if there is little adhesion between the material, used to form the translucent electrode, and the material used to make the wire bonding electrode can result in peeling and peeling of the translucent light The electrode can be prevented by the wire bonding electrode.

Even if there is some variation in the pattern during the masks alignment occurs, is also the contact between the translucent  Electrode and the wire bonding electrode are not affected because of the translucent electrode the entire wire bonding electrode with off covered their lower surface.

Because also the translucent electrode is the wire bonding electrode downward in the direction of the semiconductor, the adhesion is between between the wire bonding electrode and the semiconductor enlarges what also prevents the wire bonding electrode from coming off the semiconductor solves. Since the translucent electrode covers the entire wire bond electro de with the exception of its lower surface is also covered electrical contact with the first layer is greatly expanded. This improves the conductivity of the current from the wire bonding electrode in the translucent electrode with a corresponding reduction in Resistance strong, which enables wasteful energy consumption is reduced.

If the second layer of the translucent electrode consists of a Me talloxide is formed with a low conductivity, the current flowing in the translucent electrode from around the wire bonding electrode occurs in the conventional arrangement only by the thickness cut the first layer and is therefore halved. Because in contrast there to the translucent electrode be the entire wire bonding electrode is sufficient continuity with the wire bonding electrode ensured.

The device electrode also includes an arrangement in which the light permeable electrode is formed such that it extends over a circumference of the top surface of the wire bonding electrode. This allows the translucent electrode directly with an exposed central Ab  cut the wire bonding electrode to be contacted, or the exposed Section can be with a metal with good bonding properties be layered. This can ensure good bonding properties even if the second layer of translucent elec trode is formed from a material with poor bonding properties or both the second layer and the first layer of materials are formed with poor bonding properties. This arrangement It also makes it possible to avoid the effects of preventing peeling translucent electrode reducing the impact of a Mu misalignment and an improvement in the strength of the attachment between the wire bonding electrode and the semiconductor.

The device electrode also includes an arrangement in which the light permeable electrode is formed such that it covers the entire upper Surface of the wire bonding electrode covered. With this arrangement can also prevent loosening, reducing the out effect of pattern misalignment, improving the strength of the Adhesion between the wire bonding electrode and the semiconductor and other such effects are preserved.

The device electrode also includes an arrangement in which the Section of the second layer of the translucent electrode, which over the wire bonding electrode is removed, so that the first layer is free is laid. This enables the bonding properties to be improved by cutting the section with a material with good bonding properties is coated. This can have good bonding properties be ensured even if the second layer of translucent electrode made of a material with poor bonding properties  is formed or both the second layer and the first layer Materials with poor bonding properties are formed.

The device electrode also includes an arrangement in which the top Surface of the wire bonding electrode from the translucent electro de is covered, so that the contact between the two over the ge entire surface extends and only the lower surface of the wire receipt thing electrode excludes. This provides a major improvement in the on adhesion between the translucent electrode and the wire bonding electrode. This ensures good adhesion, even if the translucent electrode and the wire bonding electrode made of material s are formed with poor adherence and it prevents also peeling off the translucent electrode.

Even if there is some deviation during mask alignment of the pattern occurs, is also the contact between the translucent gene and the wire bonding electrode not affected because the translucent electrode with the entire wire bonding electrode Except for their lower surface covered.

Because also the translucent electrode is the wire bonding electrode downward in the direction of the semiconductor, the adhesion is between the wire bonding electrode and the semiconductor enlarged, whereby the wire bonding electrode is also prevented from coming off the half head triggers.

Because the translucent electrode also covers the entire wire bond The surface is covered with the exception of its lower surface electrical contact with the first layer is greatly expanded. This  improves the conductivity of the current from the wire bonding electrode in the translucent electrode with a corresponding reduction in Resistance strong, which enables wasteful energy consumption is reduced.

The device electrode also includes an arrangement in which the light permeable electrode is formed such that it extends over a circumference of the top surface of the wire bonding electrode. This allows the translucent electrode directly with an exposed central Ab cut the wire bonding electrode to be contacted, or the exposed Section can be made with a material with good bonding properties be layered. Therefore, good bonding properties can be guaranteed be made, even if the translucent electrode from a Ma material is formed with poor bonding properties. This arrangement it also makes possible the effects of preventing peeling the translucent electrode and a reduction in impact preserve pattern misalignment.

The device electrode also includes an arrangement in which the light permeable electrode is formed such that it covers the entire upper upper surface of the wire bonding electrode covered. This arrangement further increases ner the effects of preventing peeling of the translucent Electrode, reducing the impact of pattern misalignment device and the strength of the adhesion between the wire bonding electrode and the semiconductor is improved.

The invention also provides a method of making the device Electrode ready, which includes that metals are used to the to form the layers of the transparent electrode, and that a  Heat treatment in an oxygen atmosphere. It does it light, the metal of the second layer, which is the surface layer, to oxidize in order to give a translucency. Also part of the metal component of the second layer becomes the first Layer diffuses what is the oxide layer on the surface of the half head destroyed. The provision of light transmission and ohmic Contact between the first layer and the semiconductor can be through a single heat treatment can be achieved.

With the two-layer structure, which is transparent in a conventional casual electrode is used, a heat treatment at a higher temperature to be carried out after a hike above to the surface of the electrode layer and a diffusion over a wide area. As a result, diffusion is difficult to constantly control the reaction and a constant quality for the ohm properties and translucency to achieve that for the translucent electrode are required. In the present invention However, the surface layer is oxidized, so that only a small amount Degree of diffusion is required, which makes it possible to diffuse action at even lower temperatures, and so it can used to make translucent electrodes with stable qua lity over an even wider range of heat treatment temperatures reach.

The present invention also provides a method of making the Device electrode ready, which is a heat treatment of alloy layers in an oxygen atmosphere to provide a translucent to form ge electrode, which has two layers. This can be done by first forming an alloy monolayer, and  can therefore be achieved with a simple process. Metal with the oxygen in the alloy layer reacts strongly, can easily oxi dated to give light transmission, and part of the Metals can be used to destroy oxides that are on the top surface of the semiconductor are present. The provision of light transmission liquid and the ohmic contact with the semiconductor can with a single heat treatment can be achieved. With the two-layer Structure that ver. In a conventional translucent electrode it was difficult to achieve consistent qualities that were required for a translucent electrode. This invention can be done with just a small degree of diffusion it makes the diffusion reaction possible at even lower temperatures control, and can thus be used to transmit light electrodes with consistent quality over an even wider range of heat treatment temperatures.

In summary, contains an electrode for a light-emitting semiconductor device a transparent electrode which is formed such that it comes into contact with the surface of the semiconductor, and a wire bonding electrode that is in electrical contact with the translucent Electrode stands and is formed such that it is in partial contact with reaches the surface of the semiconductor, with at least one area being the is in contact with the semiconductor, a higher contact resistance per Area unit with respect to the semiconductor has as an area of translucent electrode that is in contact with the semiconductor. This device electrode is formed by a wire receipt thing electrode on a portion of the surface of a p-type ver GaN-based semiconductor is formed that on the surface of the Semiconductor a first layer is formed, which consists of at least one Be  constituent selected from the group consisting of Au, Pt and Pd exists, and which is formed such that it overlaps a section the top surface of the wire bonding electrode on which the wire bonding electrode is arranged that on the first layer a second Layer is formed, which consists of at least one metal, which consists of is selected from the group consisting of Ni, Ti, Sn, Cr, Co, Zn, Cu, Mg and In and that a transparent electrode is formed by the first and second layers in an oxygen-containing atmosphere be heat treated.

Claims (36)

1. Electrode for a semiconductor light emitting device based on a surface of a p-type compound semiconductor GaN base is formed, comprising: a translucent electrode, the is formed so that it is in contact with the surface of the half conductor arrives, and a wire bonding electrode, which is in electrical There is contact with the translucent electrode and so formed det is that they are in partial contact with the surface of the half conductor arrives, with at least one area in contact with the semiconductor stands, a higher contact resistance per area Chen unit with respect to the semiconductor has as an area of translucent electrode that is in contact with the semiconductor stands. 2. The electrode of claim 1, wherein the area of wire bonding electrode that is in contact with the semiconductor, at least one Includes component selected from a group consisting of Tl, In, Mn, Ti, Al, Ag, Sn, AuBe, AuZn, AuMg, AlSi, TiSi and TiBe be stands. 3. The electrode of claim 1, wherein the area of wire bonding electrode that is in contact with the semiconductor, at least one Includes an ingredient selected from a group consisting of Ti, Al and AuBe exist.   4. The electrode of claim 1, wherein the region of the translucent metal, which is in contact with the semiconductor which is selected from a group consisting of Au, Pd, Pt, Ni and Cr exists. 5. The electrode of claim 1, wherein the wire bonding electrode is one has a multilayer structure, in which the top layer Al or Au is formed. 6. The electrode of claim 1, wherein the translucent electrode comprises: a first layer formed to be in contact reaches the surface of the semiconductor, and the at least egg comprises an ingredient selected from a group consisting of consists of Au, Pt and Pd, and a second layer on which he most layer is formed and a translucent metal oxide summarizes, which contains an oxide of at least one metal, the egg ner group selected from Ni, Ti, Sn, Cr, Co, Zn, Cu, Mg and In consists. 7. The electrode of claim 6, wherein the first layer is made of Au and the second layer is made of an oxide of Ni. 8. The electrode of claim 6, wherein the second layer is an acid Has material content that gradually from the second layer in Direction of the first layer in an area near the Interface between the second layer and the first layer decreases.   9. The electrode of claim 6, wherein the first layer is a Metallele ment contains, which is a main component of the metal oxide that the second layer forms. 10. Electrode according to one of claims 1 to 5, wherein the light casual electrode is formed such that it is on a portion overlies an upper surface of the wire bonding electrode which the wire bonding electrode is arranged. 11. An electrode according to any one of claims 6 to 9, wherein the light casual electrode is formed such that it is on a portion overlies an upper surface of the wire bonding electrode which the wire bonding electrode is arranged. 12. The electrode of claim 10, wherein the translucent electrode is formed so that it over a circumference of the upper surface surface of the wire bonding electrode. 13. The electrode of claim 11, wherein the translucent electrode is formed so that it over a circumference of the upper surface surface of the wire bonding electrode. 14. The electrode of claim 10, wherein the translucent electrode is formed such that it covers an entire upper surface of the Wire bonding electrode covered. 15. The electrode of claim 11, wherein the translucent electrode is formed such that it covers an entire upper surface of the Wire bonding electrode covered.   16. The electrode of claim 11, wherein the first layer of light permeable electrode at the portion of the second layer free is placed, which lies over the wire bonding electrode. 17. The electrode of claim 16, wherein at least a portion of the free section of the first layer is coated with Al or Au. 18. Electrode for a semiconductor light emitting device based on a surface of a p-type compound semiconductor GaN base is formed, comprising: a translucent electrode, the consists of a first layer of Au, which is formed such that it comes into contact with the surface of the semiconductor, and the consists of a second layer of NiO on the first Layer is formed, and a wire bonding electrode so ge forms is that they are in electrical contact with the translucent Electrode and in partial contact with the surface of the half conductor stands, with the wire bonding electrode, from the semiconductor starting from the side, a coating of AuBe and Au includes where the translucent electrode is formed in such a way that it is connected to egg a portion over an upper surface of the wire bond trode is located on which the wire bonding electrode is arranged. 19. The electrode of claim 18, wherein the first layer of the translucent casual electrode exposed at the portion of the second layer overlying the wire bonding electrode.   20. The electrode of claim 19, wherein the exposed portion of the he Most layer of the translucent electrode coated with Au is. 21. Electrode for a semiconductor light emitting device based on a surface of a p-type compound semiconductor GaN base is formed, comprising: a translucent electrode, the consists of a first layer consisting of at least one Be component that is selected from a group, which consists of Au, Pt and Pd and which is formed in such a way that Contact with the surface of the semiconductor, and the egg there is a second layer formed on the first layer and a translucent metal oxide comprising an oxide of contains at least one metal selected from a group which consists of Ni, Ti, Sn, Cr, Co, Zn, Cu Mg and In, and one Wire bonding electrode, which is formed such that it in electri chemical contact with the translucent electrode and in part sem is in contact with the surface of the semiconductor. 22. The electrode of claim 21, wherein the first layer consists of Au and the second layer is made of an oxide of Ni. 23. The electrode of claim 21, wherein the second layer is an acid Has material content that gradually from the second layer in Direction of the first layer in an area near a Interface between the second layer and the first layer decreases.   24. The electrode of claim 21, wherein the first layer is a metallic element ment contains, which is a main component of the metal oxide that the second layer forms. 25. Electrode according to one of claims 21 to 24, wherein the light permeable electrode is formed such that it is connected to an Ab cut over an upper surface of the wire bonding electrode lies on which the wire bonding electrode is arranged. 26. The electrode of claim 25, wherein the translucent electrode is formed so that it over a circumference of the upper surface surface of the wire bonding electrode. 27. The electrode of claim 25, wherein the translucent electrode is formed so that it covers the entire top surface of the wire bonding electrode covered. 28. The electrode of claim 27, wherein the first layer of the translucent casual electrode exposed at the portion of the second layer overlying the wire bonding electrode. 29. The electrode of claim 28, wherein at least a portion of the free section of the first layer is coated with Al or Au. 30. Electrode for a semiconductor light emitting device based on a surface of a p-type compound semiconductor GaN base is formed, comprising: a translucent electrode, the is formed so that it is in contact with the surface of the half conductor arrives, and a wire bonding electrode, which is in electrical  There is contact with the translucent electrode and so formed det is that a lower surface in partial contact with the Surface of the semiconductor stands and the translucent electrode is above the top surface. 31. The electrode of claim 30, wherein the translucent electrode is formed so that it over a circumference of the upper surface surface of the wire bonding electrode. 32. The electrode of claim 30, wherein the translucent electrode is formed such that it covers an entire upper surface of the Wire bonding electrode covered. 33. Method of manufacturing an electrode for a light emitting Semiconductor device on a surface of a p-type A GaN-based compound semiconductor is formed, comprising: a first step that a wire bonding electrode on a section the surface of the semiconductor is formed, a second step, that a first layer is formed on the surface of the semiconductor the first layer comprising at least one component, which is selected from a group consisting of Au, Pt and Pd, and is formed so that it is at a portion above an above ren surface of the wire bonding electrode on which the wire bonding electrode is arranged, a third step that on the first layer a second layer is formed, the at least one Comprises metal selected from a group consisting of Ni, Ti, Sn, Cr, Co, Zn, Cu, Mg and In, and a fourth step, that a translucent electrode is formed by the first  and the second layer in an atmosphere containing oxygen be heat treated. 34. A method of manufacturing an electrode according to claim 33, wherein the oxygen-containing atmosphere has an oxygen concentration of not less than 1 ppm. 35. Method of manufacturing an electrode for a light emitting Semiconductor device on a surface of a p-type A GaN-based compound semiconductor is formed, comprising: a first step that a wire bonding electrode on a section the surface of the semiconductor is formed, a second step, that an alloy layer is formed on the surface of the semiconductor det, wherein the alloy layer comprises an alloy which contains at least one metal selected from a group which consists of Au, Pt and Pd, and which ent ent least one metal which is selected from a group consisting of Ni, Ti, Sn, Cr, Co, Zn, Cu, Mg and In exists, and is formed so that it on ei a portion over an upper surface of the wire bond trode is located on which the wire bonding electrode is arranged, and a third step that a translucent electrode is formed by placing the alloy layer in an oxygen-containing Atmosphere is heat treated to be on the semiconductor side to form the first layer, which consists of metal or an alloy, and to form a second layer made of a translucent Metal oxide is formed which is formed on the first layer.   36. A method of manufacturing an electrode according to claim 35, wherein the oxygen-containing atmosphere has an oxygen concentration of not less than 1 ppm.