JP2009105466A - Nitride semiconductor wafer, and method for manufacturing of nitride semiconductor element - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a nitride semiconductor element formed by being divided so as to be flat in a resonator end face. <P>SOLUTION: When an LD structure 251 is constituted on a GaN-based substrate 250, cleavage introduction grooves 252 are formed by being scribed by a diamond-edge blade from the surface of the LD structure 251. The cleavage introduction grooves 252 are formed between stripe-shaped optical waveguides 253 which are formed in parallel to the direction of [1-100]-orientation of a wafer, and the grooves 252 are formed in shapes of dashed lines in the direction of [11-20]-orientation of the wafer. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

The present invention relates to a laser diode that emits blue light, a nitride semiconductor device used for a light emitting diode, and a method for manufacturing the nitride semiconductor device, and more particularly to a nitride semiconductor device using a nitride semiconductor as a substrate and the nitride semiconductor device. The present invention relates to a method for manufacturing a physical semiconductor device.

A nitride III-V semiconductor (hereinafter referred to as “GaN-based semiconductor”) composed of a compound of a group III element such as Al, Ga, and In and a group V element N has a band structure, It has been expected to be used as a light emitting device or a power device because of its chemical stability, and its application has been attempted. For example, many attempts have been made to produce a nitride semiconductor device that emits a blue laser by laminating a GaN-based semiconductor on a sapphire substrate. By the way, in general, in an AlGaInAs-based or AlGaInP-based nitride semiconductor element, a resonator required for laser oscillation is manufactured using a cleavage plane.

However, when a GaN-based semiconductor layer is laminated on a sapphire substrate, sapphire is difficult to divide, so that the average of irregularities on the end face of the laminate becomes as large as 4 to 10 nm, and a good resonator can be obtained. difficult. Furthermore, when a GaN-based semiconductor layer is formed by laminating a GaN-based semiconductor layer on a sapphire substrate, the cleavage direction of sapphire and the cleavage direction of the GaN-based semiconductor layer stacked on the substrate are generally shifted by 30 °. is there. Therefore, it is difficult to reduce the unevenness of the end face even if the dividing direction is aligned with either the substrate side or the laminate side layer.

Therefore, as the substrate on which the GaN-based semiconductor layer is stacked, a GaN-based substrate having a cleavage property and having the same cleavage direction as that of the GaN-based semiconductor layer stacked on the surface is used, and the end face is produced by cleavage. It is attracting attention. Here, the GaN-based substrate is a substrate composed of a GaN-based semiconductor. When this GaN-based substrate is used, since the cleavage directions of the GaN-based semiconductor layer and the GaN-based substrate are the same, the end face is expected to be flat. Furthermore, when a GaN-based substrate is used, the lattice matching between the stacked GaN-based semiconductor layer and the GaN-based substrate is good, and there is no difference in the coefficient of thermal expansion, reducing strain and defects on the nitride semiconductor element. Therefore, it is expected that the lifetime of the nitride semiconductor device will be extended.

An example of a nitride semiconductor element in which a GaN-based semiconductor layer is laminated on a GaN-based substrate as described above and then a resonator end face is produced by cleavage is disclosed in, for example, Japanese Patent Application Laid-Open No. 11-4048.

Japanese Patent Laid-Open No. 11-4048

However, in the example of the nitride semiconductor device using the GaN-based substrate disclosed in Japanese Patent Application Laid-Open No. 11-4048, there is no detailed description regarding the resonator end face manufacturing method and the chip dividing method. Therefore, the present inventors made various attempts to cleave the wafer using the GaN-based substrate. In practice, it was difficult to divide the resonator length with a constant length and high yield without variation.

That is, as shown in FIG. 13, when the wafer 130 having the striped optical waveguide 131 and having the cleavage introduction groove 132 provided in the cleavage direction on the end face thereof is divided, Ideally, it is divided along the dividing line 133 in the same direction. Thus, when divided along the dividing line 133, the dividing surface of the striped waveguide 131 becomes flat, so that a nitride semiconductor device can be obtained with good yield. However, in reality, a wavy dividing line such as the line 134 or a dividing line having an inclination of 60 ° with respect to the direction to be divided as indicated by the line 135 frequently occurs.

As one of the causes of the generation of the lines 134 and 135, even when the cleavage introduction groove 132 is provided in the <11-20> direction (this direction will be described later), the hexagonal crystal. In the GaN-based substrate, the angle that forms 60 ° with the cleavage direction is also an equivalent cleavage direction, so that it can be easily broken by a line inclined by 60 ° with respect to the direction to be split. When such a crack occurs only once, it becomes a crack like a line 135 in FIG. 13, and when it occurs continuously, it becomes a crack like a line 134. As a further cause, unlike the sapphire substrate, the GaN-based substrate is very fragile, so that the inclined crack is likely to occur. In some cases, the elements are separated during the division.

An object of the present invention is to provide a nitride semiconductor device that is divided so that the end face of the resonator is flat. Another object of the present invention is to provide a method of manufacturing a nitride semiconductor device that can always be divided in a certain cleavage direction.

In order to achieve the above object, a nitride semiconductor device of the present invention includes a nitride semiconductor substrate composed of a compound containing a group III element and nitrogen, and a group III element and nitrogen deposited on the nitride semiconductor substrate. And a stripe optical waveguide formed in the nitride semiconductor layer, and the end face obtained by cleaving the nitride semiconductor layer and the stripe optical waveguide resonate with each other. The nitride semiconductor device constituting the device has a cleavage introduction groove formed from the surface side of the nitride semiconductor layer at a position other than directly above the stripe-shaped optical waveguide with respect to the end face. .

Such a nitride semiconductor device can be obtained by dividing the wafer provided with the cleavage introduction groove in the <11-20> direction. At this time, since the cleavage introduction groove is provided at a position other than directly above the striped optical waveguide, the end surface around the striped optical waveguide can be a mirror surface.

In addition, such a nitride semiconductor device may have a cleavage assist groove formed from the back surface side of the nitride semiconductor substrate. The cleavage assisting groove is also provided in the <11-20> direction in the same manner as the cleavage introducing groove, so that the nitride semiconductor element can be divided from the wafer without any trouble, and the end surface around the striped optical waveguide is a mirror surface. Can serve as an auxiliary to

The nitride semiconductor device of the present invention includes a nitride semiconductor substrate composed of a compound containing a group III element and nitrogen, and a compound containing a group III element and nitrogen deposited on the nitride semiconductor substrate. Nitride comprising a nitride semiconductor layer and a stripe-shaped optical waveguide formed in the nitride semiconductor layer, wherein the end face obtained by cleaving the nitride semiconductor layer and the stripe-shaped optical waveguide constitute a resonator The semiconductor element is characterized by having a cleave introduction groove formed from the back side of the nitride semiconductor substrate in addition to the position directly below the stripe optical waveguide with respect to the end face.

Such a nitride semiconductor device can be obtained by dividing the wafer provided with the cleavage introduction groove in the <11-20> direction. At this time, since the cleavage introduction groove is provided at a position other than directly below the stripe-shaped optical waveguide, the end surface around the stripe-shaped optical waveguide can be a mirror surface.

In the nitride semiconductor device described above, the depth d from the front surface or the back surface of the nitride semiconductor device to the deepest portion of the cleavage introduction groove may be in the range of 1 ≦ d ≦ 10 μm. When the nitride semiconductor layer is thin and can be marked up to the interface between the nitride semiconductor substrate and the nitride semiconductor layer, the nitride semiconductor substrate, the nitride semiconductor layer, The depth d from the interface to the deepest part of the cleavage introduction groove may be in the range of 1 ≦ d ≦ 10 μm.

Further, when the nitride semiconductor element is projected two-dimensionally with the nitride semiconductor substrate down, the cleavage guide groove is at least 50 μm from the stripe optical waveguide in a direction perpendicular to the stripe optical waveguide. It may be formed at a distant position. Further, when the nitride semiconductor element is projected two-dimensionally with the nitride semiconductor substrate down, the cleavage introduction groove is at least 100 μm from the stripe optical waveguide in a direction perpendicular to the stripe optical waveguide. The yield of the nitride semiconductor device can be improved by forming it at a distant position.

Furthermore, by setting the thickness of the above-described nitride semiconductor element to 80 to 160 μm, the wafer can be smoothly divided.

In the method for manufacturing a nitride semiconductor device of the present invention, a nitride semiconductor layer composed of a compound containing a group III element and nitrogen is deposited on a nitride semiconductor substrate composed of a compound containing a group III element and nitrogen. And a nitride semiconductor device manufacturing method for obtaining a nitride semiconductor device from a nitride semiconductor wafer in which a plurality of striped optical waveguides are configured at equal intervals in the nitride semiconductor layer. A step of adjusting the thickness of the semiconductor wafer to 80 to 160 μm, and a ruled line from the surface side of the nitride semiconductor layer to the nitride semiconductor wafer, whereby a plurality of cleavage introduction grooves are formed in an intermittent broken line shape. And the step of dividing the nitride semiconductor wafer along the cleavage introduction groove, and when forming the cleavage introduction groove, the cleavage introduction groove is the stripe. Characterized in that it is formed at a position other than directly above the optical waveguide.

At this time, by providing the cleavage introduction groove in the <11-20> direction, the nitride semiconductor wafer can be divided according to the cleavage direction. Therefore, the surface roughness at the end face around the striped waveguide can be reduced to be a mirror surface, and the yield of the nitride semiconductor device can be improved.

In such a manufacturing method, in the nitride semiconductor layer, when a semiconductor layer made of a material having a different cleavage direction from the nitride semiconductor is provided at the interface with the nitride semiconductor substrate, first, the surface of the nitride semiconductor layer Forming a plurality of cleavage assisting grooves in an intermittent broken line shape from the surface side of the nitride semiconductor layer to a depth half the thickness of the nitride semiconductor layer by scoring from the side, The cleaved introduction groove may be formed by scribing from the bottom surface of the groove.

In addition, before the step of dividing the nitride semiconductor wafer, a step of forming a cleavage assist groove by marking the nitride semiconductor wafer from the back surface side of the nitride semiconductor substrate is provided, and the cleavage is performed. The cleavage introducing groove and the cleavage assisting groove may be formed so that the cleavage introducing groove is located on the center axis of the auxiliary groove.

The method for producing a nitride semiconductor device according to the present invention includes a nitride semiconductor layer composed of a compound containing a group III element and nitrogen on a nitride semiconductor substrate composed of a compound containing a group III element and nitrogen. In the method of manufacturing a nitride semiconductor device for obtaining a nitride semiconductor device from a nitride semiconductor wafer comprising a plurality of stripe-shaped optical waveguides configured at equal intervals in the nitride semiconductor layer, A step of adjusting the thickness of the nitride semiconductor wafer to 80 to 160 μm, and scribing the nitride semiconductor wafer from the back surface side of the nitride semiconductor substrate, thereby forming a plurality of cleavages in an intermittent broken line shape. Forming the introduction groove, and dividing the nitride semiconductor wafer along the cleavage introduction groove. When forming the cleavage introduction groove, the cleavage introduction groove is Characterized in that it is formed at a position other than directly below type shaped optical waveguide.

At this time, by providing the cleavage introduction groove in the <11-20> direction, the nitride semiconductor wafer can be divided according to the cleavage direction. Therefore, the surface roughness at the end face around the striped waveguide can be reduced to be a mirror surface, and the yield of the nitride semiconductor device can be improved.

In the manufacturing method as described above, the depth d from the front surface side or the back surface side of the nitride semiconductor wafer to the deepest portion of the cleavage introduction groove may be in the range of 1 ≦ d ≦ 10 μm. When the nitride semiconductor layer is thin and can be marked up to the interface between the nitride semiconductor substrate and the nitride semiconductor layer, the nitride semiconductor substrate, the nitride semiconductor layer, The depth d from the interface to the deepest part of the cleavage introduction groove may be in the range of 1 ≦ d ≦ 10 μm.

In addition, by providing the cleavage introduction grooves at intervals of 1 mm or less on the same broken line, it is possible to suppress problems that occur when the nitride semiconductor wafer is divided. Furthermore, the cleavage introduction groove is provided between the stripe optical waveguides in the same broken line shape, so that problems occurring when the nitride semiconductor wafer is divided can be suppressed.

In the manufacturing method described above, the cleavage introduction groove itself may be formed in a continuous linear shape, or the cleavage introduction groove itself may be formed in an intermittent broken line shape.

According to the present invention, since the cleave introduction groove is provided in the cleaving direction near the optical waveguide, the cleavage is caused according to the cleave introduction groove, and a nitride semiconductor device can be obtained. Therefore, the roughness of the surface irregularities can be reduced at the end face near the optical waveguide of the nitride semiconductor element, and the mirror surface can be finished. Therefore, it is possible to improve the FFP and reduce the defects that occur when the yield in manufacturing the nitride semiconductor device is 90% or more.

The figure for demonstrating a hexagonal crystal structure. Sectional drawing which shows the structure of the wafer before dividing | segmenting the GaN-type semiconductor laser element of 1st Embodiment. Sectional drawing and top view for demonstrating the division | segmentation of the wafer in 1st Embodiment. 1 is an external perspective view of a GaN-based semiconductor laser device according to a first embodiment. Sectional drawing which shows the structure of the wafer before dividing | segmenting the GaN-type semiconductor laser element of 2nd Embodiment. Sectional drawing and top view for demonstrating the division | segmentation of the wafer in 2nd Embodiment. The external appearance perspective view of the GaN-type semiconductor laser element of 2nd Embodiment. Sectional drawing which shows the structure of the wafer before dividing | segmenting the GaN-type semiconductor laser element of 3rd Embodiment. Sectional drawing and top view for demonstrating the division | segmentation of the wafer in 3rd Embodiment. The external perspective view of the GaN-type semiconductor laser element of 3rd Embodiment. Sectional drawing and top view for demonstrating the division | segmentation of the wafer in 4th Embodiment. The external appearance perspective view of the GaN-type semiconductor laser element of 4th Embodiment. The top view for demonstrating the division | segmentation of the conventional wafer.

Embodiments of the present invention will be described below. First, definitions of terms used in this specification are clarified.

In this specification, “GaN-based semiconductor” refers to a nitride III-V group having a hexagonal crystal structure composed of a group III element such as Al, Ga, and In and a group V element N. In addition to substances that are compound semiconductors and represented by a composition ratio of Al _X Ga _Y In _1-XY N (where 0 ≦ X ≦ 1, 0 ≦ Y ≦ 1, and 0 ≦ X + Y ≦ 1), III It includes substances obtained by substituting a part of group elements (about 20% or less) with other elements and substances obtained by substituting part of group V elements (about 20% or less) with other elements.

In this specification, the “GaN-based substrate” is mainly composed of a nitride III-V compound semiconductor having a hexagonal crystal structure composed of a compound of a group III element and N, like a GaN-based semiconductor. In addition to a substrate represented by a composition ratio of Al _X Ga _Y In _1-XY N (where 0 ≦ X ≦ 1, 0 ≦ Y ≦ 1, and 0 ≦ X + Y ≦ 1), the group III A substrate in which a part of the element (about 20% or less) is replaced with another element or a substrate in which a part of the group V element (about 20% or less) is replaced with another element is included. A wafer in which a GaN-based semiconductor is deposited on a heterogeneous substrate having a substance other than a GaN-based semiconductor as a main component and then an LD structure to be described later is laminated, and before dividing the wafer in each of the following embodiments In addition, the removal of the heterogeneous substrate is also included in the category of the GaN-based substrate defined in this specification.

In this specification, the “LD structure” is a structure mainly composed of a GaN-based semiconductor including a light emitting portion and a waveguide structure other than a resonator inside, and is deposited or epitaxially grown on the GaN-based substrate. In this layer structure, the electrode metal and the insulator film inserted between the electrode and the GaN-based semiconductor are excluded. The LD structure may be mixed with a GaN-based semiconductor having a partially different crystal structure or a material other than the GaN-based semiconductor.

In this specification, the “striped optical waveguide” is an integral structure for confining and guiding light emitted from the light emitting portion and the light emitting portion.

In the present specification, the “average value Ra of surface irregularities” is an average value of a roughness curve measured using a step meter, with a center line as a reference.

In this specification, “RMS value” or simply “average of unevenness” means the size of surface unevenness, and a roughness curve is measured with an AFM (AtomicForce Microscope) over a length of 4 μm in a direction parallel to the growth layer surface. It is a value calculated as RMS (Root Mean Square: square root of the deviation from the center line to the roughness curve).

In this specification, “groove depth” means that when the center line of the interface between the GaN-based substrate and the LD structure is placed horizontally in the GaN-based semiconductor laser device, it is perpendicular to the edge of the cleavage-introducing groove or the auxiliary cleavage groove. Shows a value obtained by measuring the depth of the valley of the groove. And this “groove depth” has three types, the first is “depth based on the surface of a GaN-based semiconductor laminated as an LD structure”, and the second is “GaN-based substrate” The depth is based on the center line of the interface between the LD structure and the LD structure, and the third is the depth based on the back surface of the GaN-based substrate.

In the hexagonal crystal structure as shown in FIG. 1, <0001> represents the normal direction of the plane represented by all A, that is, the [0001] and [000-1] directions, and <1-100> represents all Represents the normal direction of the surface represented by B, ie, [1-100], [10-10], [01-10], [-1100], [-1010] and [0-110] directions, and <11 −20> is the normal direction of the surface represented by all C, that is, [11-20], [1-210], [-2110], [−1-120], [-12-10] and [2 −1-10] direction.

In each of the following embodiments, a GaN-based semiconductor laser element will be described as a representative nitride semiconductor element.

<First Embodiment>
A first embodiment of the present invention will be described below with reference to the drawings. FIG. 2 is a cross-sectional view showing the configuration of the wafer before dividing the GaN-based semiconductor laser device in this embodiment. FIG. 3 is a cross-sectional view and a top view for explaining the division of the wafer. FIG. 4 is an external perspective view of the divided GaN-based semiconductor laser element.

1. GaN semiconductor laser device manufacturing method
(Wafer formation)
First, wafer formation will be described with reference to FIG.

First, the n-GaN substrate 200 having a thickness of 100 to 500 μm with the (0001) plane as the surface for crystal growth is organically cleaned. In the present embodiment, the film thickness of the n-GaN-based substrate 200 is adjusted to 135 μm.

The cleaned n-GaN substrate is carried into a MOCVD (Metal Organic Chemical Vapor Deposition) apparatus and cleaned at a high temperature of about 1100 ° C. in a hydrogen (H ₂ ) atmosphere. Thereafter, the temperature is lowered, and 10 nmol / min of silane (SiH ₄ ) is introduced while flowing H ₂ as a carrier gas at 10 l / min, and ammonia (NH ₃ ) and trimethylgallium (TMG) are each 5 l / min at 600 ° C. Then, by introducing 20 mol / min, an n-GaN buffer layer 201 having a thickness of 10 nm to 10 μm (for example, 100 nm) is grown.

The buffer layer 201 may be a GaN buffer layer formed with a SiH ₄ introduction amount of 0 mol / min. Alternatively, a GaN buffer layer containing Al or In may be used. At this time, when the buffer layer contains Al, trimethylaluminum (TMA) may be introduced, and when the buffer layer contains In, trimethylindium (TMI) may be introduced in an appropriate amount. The buffer layer 201 is a layer provided for the purpose of relaxing the surface strain of the n-GaN-based substrate, improving the surface morphology and unevenness (flattening), and in the n-GaN-based substrate, the n-GaN for crystal growth When the crystallinity is excellent, the buffer layer 201 may not be provided.

Next, the temperature is raised to about 1050 ° C. while flowing nitrogen (N ₂ ) and ammonia (NH ₃ ) at 5 l / min. Thereafter, the carrier gas is changed from N ₂ to H ₂ , TMG is introduced at 100 μmol / min, SiH ₄ is introduced at 10 nmol / min, and the n-GaN contact layer 202 is grown by 0.1 to 10 μm (for example, 4 μm).

Next, the flow rate of TMG is adjusted to 50 μmol / min, a fixed amount of TMA is introduced, and an n-Al _x1 Ga _1-x1 N layer (for example, x1 = 0.2) is laminated, for example, the total film thickness A 0.8 μm n-AlGaN cladding layer 203 is formed. The n-AlGaN cladding layer 203 may be a film made of another material as long as it has a smaller refractive index and a larger band gap than an n-GaN light guide layer 204 described later. Further, several layers may be combined so that the average refractive index and the average band gap satisfy the comparison condition with the above-described n-GaN light guide layer 204.

After the formation of the n-AlGaN cladding layer 203, the supply of TMA is stopped and TMG is adjusted to 100 μmol / min so that the n-GaN light guide layer 204 has a thickness of 50 to 200 nm (for example, 100 nm). Grow. Thereafter, the supply of TMG is stopped, the carrier gas is changed from H ₂ to N ₂ again, and the temperature is lowered to 700 ° C. Then, a certain amount of TMI and 15 μmol / min of TMG are introduced to grow a barrier layer made of In _v Ga _1-v N (0 <v <1). After a predetermined time, the supply of TMI is increased to a certain amount, and a well layer made of In _W Ga _1-W N (0 <w <1) is grown.

By repeating this change in the supply amount of TMI, an InGaN multiple quantum well active layer 205 having an alternately laminated structure of InGaN barrier layers and InGaN well layers is formed. The composition ratio and film thickness of InGaN forming the barrier layer and the well layer are designed so that the emission wavelength is in the range of 370 to 430 nm, and the flow rate of TMI introduced during growth is a film having an In composition equal to the design value. Adjust to obtain.

In the InGaN multiple quantum well active layer 205, the number of well layers is preferably 2 to 6, and more preferably 3 layers. When the formation of the InGaN multiple quantum well active layer 205 is completed, the supply of TMI and TMG is stopped and the temperature is raised to 1050 ° C. again. Then, after changing the carrier gas from N ₂ to H ₂ again, 50 μmol / min of TMG, an appropriate amount of TMA, and 10 nmol / min of biscyclopentadienylmagnesium (Cp ₂ Mg) which is a p-type doping raw material are flowed. _{, p-Al z Ga 1-} z N (0 ≦ z ≦ 0.3) of 0~20nm thick growing evaporation preventing layer 206. When the growth of the p-AlGaN evaporation prevention layer 206 is completed, the supply of TMA is stopped and the supply amount of TMG is adjusted to 100 μmol / min, and p-GaN light having a thickness of 50 to 200 nm (for example, 100 nm) is obtained. A guide layer 207 is grown.

Next, by adjusting the flow rate of TMG to 50 μmol / min and introducing a certain amount of TMA, a p-type Al _x2 Ga _1-x2 N layer (for example, x2 = 0.2) is stacked, for example, A p-AlGaN cladding layer 208 having a total film thickness of 0.8 μm is formed. The p-AlGaN cladding layer 208 may be a film made of another material as long as it has a smaller refractive index and a larger band gap than the p-GaN light guide layer 207. In addition, several layers may be combined so that the average refractive index and the average band gap satisfy the conditions for the p-GaN light guide layer 207 described above.

Finally, the supply amount of TMG is adjusted to 100 μmol / min, the supply of TMA is stopped, and the p-GaN contact layer 209 having a film thickness of 0.01 to 10 μm (for example, 0.1 μm) is grown. By doing so, the growth of the LD structure on the GaN-based substrate 200 is completed. When the growth is completed, the supply of TMG and Cp ₂ Mg is stopped and the temperature is lowered, and then the wafer is taken out from the MOCVD apparatus at room temperature. When the flatness of the surface of the wafer thus formed was measured, the average value of surface irregularities was Ra = 100 mm or less.

Subsequently, the wafer is processed to form a laser element. First, in forming the p-electrode portion, etching is performed in a stripe shape along the <1-100> direction (see FIG. 3) of the GaN-based substrate 200 to form the ridge stripe portion 211. Thereafter, a SiO ₂ dielectric film 212 is vapor-deposited, and then the p-GaN contact layer 209 is exposed and vapor-deposited in the order of Pd / Mo / Au to form a p-electrode 213. As the p-electrode 213, one formed in the order of Pd / Pt / Au, one formed in the order of Pd / Au, or one formed in the order of Ni / Au may be used.

Next, by using a physical method such as polishing or a chemical method such as wet etching or dry etching, the back surface side of the n-GaN-based substrate 200 is shaved to adjust the thickness of the wafer to 80 to 160 μm. . In this way, the wafer is adjusted to a thickness that can be easily divided. That is, when the thickness of the wafer is smaller than this range, handling during element formation is hindered. Conversely, when the thickness is larger than this range, it is difficult to divide the wafer.

Next, the n-electrode 210 is formed in the order of Hf / Al from the back side of the n-GaN-based substrate 200. Thus, using Hf for the n-electrode 210 is effective because the contact resistance of the n-electrode can be lowered. Further, the n electrode 210 may be formed in the order of Ti / Al, formed in the order of Ti / Mo, or formed in the order of Hf / Au.

In forming the n-electrode 210, instead of forming the electrode from the back side of the n-GaN-based substrate 200, the n-GaN layer 202 is exposed from the front side of the wafer by using a dry etching method to form the n-electrode. It doesn't matter.

(Division of wafer)
Next, a wafer dividing method according to this embodiment will be described with reference to FIG. FIG. 3A is a cross-sectional view of a wafer in which the LD structure is formed on the GaN-based substrate 200 as described above, and FIG. 3B is a GaN-based substrate in which the LD structure is formed as described above. A top view of a wafer formed on 200 is shown respectively.

In FIG. 3, for simplicity, it is assumed that the GaN-based substrate 250 includes an n-GaN-based substrate 200, an n-GaN buffer layer 201, and an n-electrode 210. -GaN contact layer 202, n-AlGaN cladding layer 203, n-GaN light guide layer 204, InGaN multiple quantum well active layer 205, p-AlGaN evaporation prevention layer 206, p-GaN light guide layer 207, p-AlGaN cladding layer 208, a p-GaN contact layer 209, a SiO ₂ dielectric film 212, and a p-electrode 213.

As described above, when the LD structure 251 is formed on the GaN-based substrate 250 in the wafer, a stripe-shaped waveguide 253 is provided in the LD structure 251 as shown in FIG. The striped waveguide 253 is provided so as to be parallel to the <1-100> direction. Between the striped waveguides 253, cleavage introduction grooves 252 are provided to assist in dividing the wafer in the <11-20> direction to obtain a plurality of bars. A cross-sectional view of a portion where the cleavage introduction groove 252 is provided is as shown in FIG.

The cleavage introducing groove 252 is provided on the surface of the LD structure 251 by scribing with a diamond needle between the striped waveguides 253 as shown in FIG. 3B. At this time, the depth d from the surface of the LD structure 251 to the deepest part of the cleavage introduction groove 252 is at least 1 μm ≦ d ≦ 10 μm. In this way, when the wafer is obtained by dividing the wafer, the “bar yield” can be improved, which is the ratio at which an unbroken bar can be obtained.

Further, when the LD structure 251 is not so thick and the cleavage introduction groove 252 can reach the interface between the GaN-based substrate 250 and the LD structure 251, cleavage is introduced from the interface between the GaN-based substrate 250 and the LD structure 251. The depth d1 up to the deepest part of the groove 252 can be set to 1 μm ≦ d1 ≦ 10 μm. By doing so, the element obtained from the divided bars is an element with a flat resonator end face of the laser beam whose RMS value is 0.5 nm or less, and variation in the resonator length is within a certain range. It is possible to improve the “yield of elements”, which is the ratio at which the elements are obtained.

This is because a part of the LD structure 251 is not cleaved or a substance having a different cleaving direction is included and the divided pressure is dispersed, so that an undesirable roughness occurs and This is because it is possible to prevent the far field pattern (FFP) from being deteriorated and the reflectance of the resonator end face from being lowered.

The groove introduction direction of the cleavage introducing groove 252 is <11-20> of the GaN-based substrate 250 as described above. Further, by setting the start and end points of the cleavage introduction groove 252 to be 50 μm or more away from the striped waveguide 253, the bar can be divided with good yield, and more preferably, 100 μm or more. Thus, the device yield can be increased.

In the present embodiment, as shown in FIG. 3A, the depth from the interface between the GaN-based substrate 250 and the LD structure 251 to the deepest portion of the cleavage introduction groove 252 is constant at a depth of 1 μm. . The distance from the striped waveguide 253 to the start point and end point of the cleavage introduction groove 252 was set to 125 μm.

Further, when the cleavage introduction groove 252 is provided in this way, in order to divide the bar with a high yield, it is desirable to form it as long as a straight line in the <11-20> direction as much as possible within the above-mentioned range. It doesn't matter. Further, as a method for forming the cleavage introduction groove 252, dry etching such as RIE (Reactive Ion etching) or wet etching may be used in addition to the above-described scribing.

Further, the cleavage introduction grooves 252 are provided between the striped waveguides 253 as shown in FIG. 3B, but the intervals between the cleavage introduction grooves 252 are within 1 mm. In addition, if the distance between the start point and the end point of the stripe waveguide 253 satisfies the above-described conditions, it is not necessary to provide the gap between the stripe waveguides 253.

Next, the wafer thus provided with the cleavage introduction groove 252 is divided in the <11-20> direction to obtain a bar. In this wafer division, a breaking blade is applied to the position where the cleavage introduction groove 252 is located from the back side of the GaN-based substrate 250 to push the wafer. By doing in this way, in the divided | segmented bar | burr, the end surface which uses a cleaved surface can be formed in the part into which the striped waveguide 253 was divided | segmented. In addition, cleaving by impact given by hitting the blade, dividing by heating only the periphery of the crease line, or breaking by impact by sound waves or water flow, etc. You may make it.

By dividing in this way, many bars having a resonator length of 500 μm were obtained from the wafer as shown in FIG. The resonator length was within the set value of 500 μm ± 5 μm, the bar yield was 92%, and the element yield was 90%. At this time, in the divided bars, the average of the unevenness of the end face between the cleavage introducing grooves 252 was measured. As a result, the maximum RMS value was 10 nm at a distance within 50 μm from the cleavage introduction groove 252, but it had a very flat surface at a maximum of 0.5 nm at a point distant from the cleavage introduction groove by 100 μm or more. When a GaN-based semiconductor was deposited on a sapphire substrate, the average RMS value of this semiconductor portion was 3.5 nm, and thus it was confirmed that the quality was improved with the cleaved end face being even more flat. .

In this way, the bar obtained by dividing the wafer as shown in FIG. 3 is divided by marking the back surface or the front surface in the <1-100> direction between the striped waveguides 253. A GaN-based semiconductor laser device is obtained. At this time, by adjusting the stylus pressure (the load when the stylus is pressed against the wafer) at the time of scribing, the GaN-based semiconductor laser device is obtained by being divided by pushing in the <1-100> direction. Alternatively, it may be cut completely and divided to obtain a GaN-based semiconductor laser device.

2. Configuration of GaN-based semiconductor laser device The configuration of the GaN-based semiconductor laser device 1 formed by dividing the wafer as described above will be described with reference to FIG.

In FIG. 4, for simplicity, it is assumed that the GaN-based substrate 10 includes an n-GaN-based substrate 200 and an n-GaN buffer layer 201, and the LD structure 11 includes an n-GaN contact layer. 202, n-AlGaN cladding layer 203, n-GaN light guide layer 204, InGaN multiple quantum well active layer 205, p-AlGaN evaporation prevention layer 206, p-GaN light guide layer 207, p-AlGaN cladding layer 208, p- It is assumed that the GaN contact layer 209 and the SiO ₂ dielectric film 212 are included.

As described above, the GaN-based semiconductor laser device 1 obtained by dividing the wafer having the LD structure formed on the GaN-based substrate has the mirror end surface formed by cleavage into the LD structure 11 formed on the GaN-based substrate 10. 12 is formed. A stripe-shaped waveguide 13 is provided inside the LD structure 11 and plays a role of guiding laser light.

The n-electrode 210 provided on the lower surface of the GaN-based substrate 10 and the p-electrode 213 provided on the upper surface of the LD structure 11 are for supplying electric power from the outside during the operation of the GaN-based semiconductor laser device 1. . Furthermore, groove portions 14 are formed on the LD structure 11 side at portions located at the four corners on the surface of the GaN-based semiconductor laser device 1.

This grooving portion 14 corresponds to the cleavage introduction groove 252 (FIG. 3) formed in advance on the upper surface of the wafer in order to form the mirror end surface 12 when the wafer is divided into bars. The depth from the interface between the substrate 10 and the LD structure 11 to the deepest portion of the grooved portion 14 is 1 μm. Further, in this embodiment, when the GaN-based semiconductor laser device 1 is two-dimensionally projected with the GaN-based substrate 10 facing down, the grooving portion 14 is formed from a position away from the stripe-shaped waveguide 13 by 125 μm. Yes. In this embodiment, the number of grooving portions 14 for the GaN-based semiconductor laser device 1 is four. However, the number varies depending on the state of the cleavage introduction groove 252 formed in advance on the wafer upper surface. That's fine.

In this embodiment, when the wafer is divided to obtain a GaN-based semiconductor laser device, the grooved portion generated by the cleavage introduction groove portion may be cut off. This has an effect of removing dust or the like when forming the cleavage introduction groove.

<Second Embodiment>
A second embodiment of the present invention will be described below with reference to the drawings. FIG. 5 is a cross-sectional view showing the configuration of the wafer before dividing the GaN-based semiconductor laser device in this embodiment. FIG. 6 is a cross-sectional view and a top view for explaining the division of the wafer. FIG. 7 is an external perspective view of a divided GaN-based semiconductor laser device.

1. GaN semiconductor laser device manufacturing method
(Wafer formation)
First, wafer formation will be described with reference to FIG. In FIG. 5, the same portions of the wafer shown in FIG. 2 are denoted by the same reference numerals, and detailed description thereof is omitted.

In the present embodiment, unlike the first embodiment (FIG. 2), first, an electron is formed on the surface of an n-GaN-based substrate 200 having a thickness of 100 to 500 μm with the (0001) plane as a surface for crystal growth. A growth suppressing film is formed by vapor deposition with SiO ₂ using a beam method or a sputtering method. Thereafter, by using a lithography technique using a photo-curable resin, from the growth suppression film formed on the n-GaN substrate 200 along the <1-100> direction of the n-GaN substrate 200, A striped SiO ₂ mask 501 is formed.

The mask 501 has a mask width of 13 μm and a window width between the masks 501 of 7 μm. The growth suppressing film may be made of SiN _x , Al ₂ O ₃ , TiO ₂ or the like in addition to the above-described SiO ₂ . The mask may be hollow.

In this way, the n-GaN-based substrate 200 having the mask 501 formed on the surface is organically cleaned, and then carried into the MOCVD apparatus. The carrier gas H ₂ is flowed, and TMG is 100 μmol / min, SiH _4. 10 nmol / min is introduced to grow the n-GaN contact layer 502 by 25 μm.

After the n-GaN contact layer 502 is formed, as in the first embodiment, the N-AlGaN cladding layer 203 is formed by introducing TMA, and then the supply of TMA is stopped and the n− A GaN light guide layer 204 is formed. Thereafter, the carrier gas is changed from H ₂ to N ₂ again and the temperature is lowered to 700 ° C., and then TMI and TMG are introduced, whereby an InGaN multiple quantum well active layer 205 composed of an alternately laminated structure of InGaN barrier layers and InGaN well layers. Form.

Then, after stopping the supply of TMI and TMG and raising the temperature to 1050 ° C. again, the carrier gas is changed from N ₂ to H ₂ again, and TMG, TMA, Cp ₂ Mg is allowed to flow to prevent p-AlGaN evaporation. Layer 206 is grown. Thereafter, the supply of TMA is stopped, and the p-GaN optical guide layer 207 is grown. Further, when TMA is introduced again to form the p-AlGaN cladding layer 208, the supply of this TMA is stopped, and the p-GaN contact layer 209 is grown, and the LD on the GaN-based substrate 200 is grown. End the growth of the structure.

The wafer having the LD structure formed on the surface in this manner is processed to form a laser element when it is taken out from the MOCVD apparatus after the supply of TMG and Cp ₂ Mg is stopped and the temperature is lowered. First, etching is performed to form the ridge stripe portion 211, and then the SiO ₂ dielectric film 212 is deposited, then the p-GaN contact layer 209 is exposed, and deposited in the order of Pd / Mo / Au to form a p-electrode. 213 is formed. As in the first embodiment, the p-electrode 213 is either formed in the order of Pd / Pt / Au, formed in the order of Pd / Au, or formed in the order of Ni / Au. It may be used.

Next, by using a physical method such as polishing or a chemical method such as wet etching or dry etching, the back surface side of the n-GaN-based substrate 200 is shaved to adjust the thickness of the wafer to 80 to 160 μm. . Then, the n-electrode 210 is formed in the order of Hf / Al from the back side of the n-GaN-based substrate 200. Similarly to the first embodiment, the n electrode 210 may be formed in the order of Ti / Al, formed in the order of Ti / Mo, or formed in the order of Hf / Au. Anyway.

In forming the n-electrode 210, instead of forming the electrode from the back side of the n-GaN-based substrate 200, the n-GaN layer 202 is exposed from the front side of the epitaxial wafer by using a dry etching method. It may be formed.

(Division of wafer)
Next, a wafer dividing method according to this embodiment will be described with reference to FIG. FIG. 6A shows a cross-sectional view of the wafer in which the LD structure is formed on the GaN-based substrate 200 as described above, and FIG. 6B shows the LD structure having the GaN-based substrate as described above. A top view of a wafer formed on 200 is shown respectively.

In FIG. 6, for the sake of simplicity, it is assumed that the GaN-based substrate 510 includes the n-GaN-based substrate 200 and the n-electrode 210, and the LD structure 511 includes a mask 501 and an n-GaN contact layer. 502, n-AlGaN cladding layer 203, n-GaN light guide layer 204, InGaN multiple quantum well active layer 205, p-AlGaN evaporation prevention layer 206, p-GaN light guide layer 207, p-AlGaN cladding layer 208, p- It is assumed that the GaN contact layer 209, the SiO ₂ dielectric film 212, and the p-electrode 213 are included.

As described above, when the LD structure 511 is formed on the GaN-based substrate 510 in the wafer, as in the first embodiment, a stripe shape is formed in the LD structure 511 as shown in FIG. A waveguide 253 is provided. Between the striped waveguides 253, cleavage assist grooves 512 and cleavage introduction grooves 513 are provided to assist in dividing the wafer in the <11-20> direction to obtain a plurality of bars. A sectional view of a portion where the cleavage assisting groove 512 and the cleavage introducing groove 513 are provided is as shown in FIG.

First, the cleavage assist groove 512 is provided on the surface of the LD structure 511 by performing RIE between the striped waveguides 253 as shown in FIG. 6B. The cleavage assist groove 512 is formed so that its depth is half of the epitaxial layer thickness of the LD structure 511. That is, the cleavage assisting groove 512 is configured above the mask 501 configured in the LD structure 511.

When the cleavage assisting groove 512 is configured in this way, the cleavage introduction groove 513 is provided by scribing with a diamond needle on the center side of the cleavage assisting groove 512. The cleavage introduction groove 513 is provided so as to reach the mask 501 in the LD structure 511 whose cleavage direction is different from that of other portions. At this time, assuming that the depth from the surface of the LD structure 511 to the deepest part of the cleavage introduction groove 513 is d2, at least 1 μm ≦ d2 ≦ 10 μm is set on the basis of the surface of the LD structure 511. By doing in this way, the yield of a bar can be improved.

Further, when the LD structure 511 is not so thick and the cleavage introduction groove 513 can reach the interface between the GaN-based substrate 510 and the LD structure 511, cleavage is introduced from the interface between the GaN-based substrate 510 and the LD structure 511. The depth d3 to the deepest part of the groove 513 is 1 μm ≦ d3 ≦ 10 μm with reference to the center line of the interface between the GaN-based substrate 510 and the LD structure 511, so that the device yield can be improved.

This is desirable because, in particular, a part of the LD structure 511 such as the mask 501 is not cleaved or a substance having a different cleaving direction is contained and the dividing pressure is dispersed. This is because it is possible to prevent the FFP of the emitted light from deteriorating and the reflectance of the resonator end face from being lowered.

The groove insertion direction of the cleavage assisting groove 512 and the cleavage introducing groove 513 is <11-20> of the GaN-based substrate 510 as described above. Further, the start and end points of the cleavage assisting groove 512 and the cleavage introducing groove 513 are separated from the striped waveguide 253 by 50 μm or more so that the bars can be divided with good yield, and more preferably 100 μm or more. The yield of the element can be increased by using the above points.

In the present embodiment, as shown in FIG. 6A, the depth from the interface between the GaN-based substrate 510 and the LD structure 511 to the deepest part of the cleavage introduction groove 513 is constant at a depth of 1 μm. . The distance from the stripe-shaped waveguide 253 to the starting point and the ending point of the cleavage assisting groove 512 and the cleavage introducing groove 513 was set to 125 μm.

Further, when the cleavage assisting groove 512 and the cleavage introduction groove 513 are provided in this way, it is desirable to form them as long as possible in the <11-20> direction within the above range in order to divide the bars with a high yield. It does not matter as a shape. Further, as a method for forming the cleavage introducing groove 513, dry etching such as RIE or wet etching may be used in addition to the scribe described above.

Furthermore, the cleavage assisting groove 512 and the cleavage introducing groove 513 are provided between the striped waveguides 253 as shown in FIG. 6B. However, the intervals between the cleavage assisting grooves 512 and the cleavage introducing grooves 513 are different. If the distance is within 1 mm and the distance between the start and end points and the striped waveguide 253 satisfies the above-described conditions, it is necessary to provide the gap between the striped waveguides 253. Absent.

Next, the wafer thus provided with the cleavage assisting groove 512 and the cleavage introducing groove 513 is divided in the <11-20> direction to obtain a bar. In this wafer division, as in the first embodiment, the breaking blade is applied from the back side of the GaN-based substrate 510 to the position where the cleavage introduction groove 513 is located, and the wafer is pushed. By doing in this way, in the divided | segmented bar | burr, the end surface which uses a cleaved surface can be formed in the part into which the striped waveguide 253 was divided | segmented. In addition, cleaving by impact given by hitting the blade, dividing by heating only the periphery of the crease line, or breaking by impact by sound waves or water flow, etc. You may make it.

By dividing in this way, many bars having a resonator length of 500 μm were obtained from the wafer as shown in FIG. The resonator length was within the set value of 500 μm ± 5 μm, the bar yield was 92%, and the device yield was 96%. At this time, when the average of the unevenness of the end surface between the cleavage assisting grooves 512 was measured in the divided bars, the result was the same as that of the first embodiment. Therefore, compared with the case where a GaN-based semiconductor was deposited on the sapphire substrate, it was confirmed that the quality was improved because the cleaved end face was further flat.

In this manner, the bar obtained by dividing the wafer as shown in FIG. 6 has the back surface or the front surface in the <1-100> direction between the striped waveguides 253, as in the first embodiment. A GaN-based semiconductor laser device is obtained by scoring and dividing. At this time, the stylus pressure at the time of scribing may be adjusted and divided by pushing and dividing in the <1-100> direction to obtain a GaN-based semiconductor laser device, or by cutting completely. It may be divided to obtain a GaN-based semiconductor laser element.

2. Configuration of GaN-based semiconductor laser device The configuration of the GaN-based semiconductor laser device 2 formed by being divided from the wafer as described above will be described with reference to FIG.

In FIG. 7, the LD structure 21 includes a mask 501, an n-GaN contact layer 502, an n-AlGaN cladding layer 203, an n-GaN light guide layer 204, an InGaN multiple quantum well active layer 205, It includes a p-AlGaN evaporation prevention layer 206, a p-GaN light guide layer 207, a p-AlGaN cladding layer 208, a p-GaN contact layer 209, and a SiO ₂ dielectric film 212.

As described above, the GaN-based semiconductor laser device 2 obtained by dividing the wafer having the LD structure formed on the GaN-based substrate has the mirror end surface formed on the LD structure 21 formed on the GaN-based substrate 200 by cleavage. 22 is formed. A stripe-shaped waveguide 23 is provided inside the LD structure 21 and plays a role of guiding laser light.

The n-electrode 210 provided on the lower surface of the GaN-based substrate 200 and the p-electrode 213 provided on the upper surface of the LD structure 21 are for supplying electric power from the outside during the operation of the GaN-based semiconductor laser device 2. . Further, groove portions 24 are formed on the LD structure 21 side at portions located at the four corners on the surface of the GaN-based semiconductor laser device 2. The grooving portion 24 includes a grooving portion 24a whose depth reaches the center of the LD structure 21 and a grooving portion 24b whose depth reaches the GaN-based substrate 200 from the bottom surface of the grooving portion 24a. Composed.

This grooving portion 24 corresponds to a cleavage assisting groove 512 and a cleavage introducing groove 513 (FIG. 6) formed in advance on the upper surface of the wafer in order to form the mirror end face 22 when the wafer is divided into bars. The insertion portion 24a corresponds to the cleavage assisting groove 512, and the groove insertion portion 24b corresponds to the cleavage introduction groove 513. At this time, in the present embodiment, the depth from the interface between the GaN-based substrate 200 and the LD structure 21 to the deepest portion of the grooving portion 24 is 1 μm.

Further, in this embodiment, the grooved portion 24 is formed from a position 125 μm away from the stripe-shaped waveguide 23 when the GaN-based semiconductor laser device 2 is two-dimensionally projected with the GaN-based substrate 200 facing down. Yes. The number of grooved portions 24 with respect to the GaN-based semiconductor laser element 2 is four in this embodiment, but varies depending on the state of the cleavage assisting groove 512 and the cleavage introducing groove 513 formed in advance on the wafer upper surface. There may be at least one or more.

In this embodiment, when the wafer is divided to obtain a GaN-based semiconductor laser device, the grooved portion generated by the cleavage introduction groove portion may be cut off. This has an effect of removing dust or the like when forming the cleavage introduction groove.

<Third Embodiment>
A third embodiment of the present invention will be described below with reference to the drawings. FIG. 8 is a cross-sectional view showing the configuration of the wafer before dividing the GaN-based semiconductor laser device according to this embodiment. FIG. 9 is a cross-sectional view and a top view for explaining the division of the wafer. FIG. 10 is an external perspective view of a divided GaN-based semiconductor laser device.

1. GaN semiconductor laser device manufacturing method
(Wafer formation)
First, wafer formation will be described with reference to FIG. In FIG. 8, the same portions of the wafer shown in FIG. 2 are denoted by the same reference numerals, and detailed description thereof is omitted.

In the present embodiment, as in the first embodiment (FIG. 2), after first organically cleaning the n-GaN-based substrate 200 having a thickness of 100 to 500 μm with the (0001) plane as the surface for crystal growth, It is carried into the MOCVD apparatus and cleaned at a high temperature of about 1100 ° C. in an H ₂ atmosphere. Unlike the first embodiment, the temperature is lowered to about 1050 ° C. while flowing N ₂ and NH ₃ at 5 l / min, respectively, and then the carrier gas is changed from N ₂ to H ₂ , and TMG and SiH ₄ are introduced. Then, the n-GaN contact layer 202 is grown to 0.1 to 10 μm (for example, 4 μm).

The subsequent LD structure and n electrode and p electrode formation methods are the same as those in the first and second embodiments. That is, on the n-GaN contact layer 202, an n-AlGaN cladding layer 203, an n-GaN light guide layer 204, an InGaN multiple quantum well active layer 205, a p-AlGaN evaporation prevention layer 206, a p-GaN light guide layer 207, A p-AlGaN cladding layer 208 and a p-GaN contact layer 209 are sequentially formed, and the growth of the LD structure is completed. Then, the wafer on which the LD structure is formed is taken out from the MOCVD apparatus at room temperature.

In the wafer having the LD structure formed in this manner, the surface of the LD structure is etched in a stripe shape to form a ridge stripe portion 211 and a SiO ₂ dielectric film 212 is deposited, and then a p-GaN contact layer. 209 is exposed. Then, a p-electrode material is deposited on the surface to form a p-electrode 213. Further, when the back surface of the n-GaN-based substrate 200 is shaved and the thickness of the wafer is adjusted, an n-electrode material is deposited on the back surface side of the n-GaN-based substrate to form the n-electrode 210. .

(Division of wafer)
Next, a wafer dividing method according to this embodiment will be described with reference to FIG. FIG. 9A shows a cross-sectional view of the wafer having the LD structure formed on the GaN-based substrate 200 as described above, and FIG. 9B shows the LD structure having the GaN-based substrate as described above. A top view of a wafer formed on 200 is shown respectively.

In FIG. 9, for simplicity, it is assumed that the GaN-based substrate 700 includes the n-GaN-based substrate 200 and the n-electrode 210, and the LD structure 701 includes the n-GaN contact layer 202, n -AlGaN cladding layer 203, n-GaN light guide layer 204, InGaN multiple quantum well active layer 205, p-AlGaN evaporation prevention layer 206, p-GaN light guide layer 207, p-AlGaN cladding layer 208, p-GaN contact layer 209, a SiO ₂ dielectric film 212, and a p-electrode 213.

As described above, when the LD structure 701 is formed on the GaN-based substrate 700 in the wafer, as in the first embodiment, the stripe structure is formed in the LD structure 701 as shown in FIG. 9B. A waveguide 703 is provided. Between the striped waveguides 703, cleavage introduction grooves 702 are provided to assist in dividing the wafer in the <11-20> direction to obtain a plurality of bars. A cross-sectional view of a portion where the cleavage introduction groove 702 is provided is as shown in FIG.

The cleavage introduction groove 702 is provided by scribing the back surface of the GaN-based substrate 700 with a diamond needle. At this time, if the depth from the back surface of the GaN-based substrate 700 to the deepest part of the cleavage introduction groove 702 is d4, at least 1 μm ≦ d4 ≦ 10 μm is set. The groove introduction direction of the cleavage introducing groove 702 is <11-20> of the GaN-based substrate 700 as described above. Further, by setting the start and end points of the cleavage introduction groove 702 to be 50 μm or more away from the striped waveguide 703, the bar can be divided with good yield, and more preferably 100 μm or more. Thus, the device yield can be increased.

In this embodiment, as shown in FIG. 9A, the depth d4 from the back surface of the GaN-based substrate 700 to the cleavage introduction groove 702 is constant at a depth of 4 μm. The distance from the striped waveguide 703 to the start point and end point of the cleavage introduction groove 702 was set to 125 μm. Further, when the cleavage introduction groove 702 is provided in this way, in order to divide the bar with a high yield, it is desirable to make it as long as possible in the <11-20> direction within the above-mentioned range, but it may be a broken line shape. . Further, as a method of forming the cleavage assisting groove 702, dry etching such as RIE or wet etching may be used in addition to the above-described scribing.

Furthermore, the cleavage introduction grooves 702 are provided between the striped waveguides 703 as shown in FIG. 9B, but the intervals between the cleavage introduction grooves 702 are within 1 mm. In addition, if the distance between the start point and the end point of the stripe-shaped waveguide 703 satisfies the above-described conditions, it is not necessary to provide the gap between the stripe-shaped waveguides 703.

Next, the wafer thus provided with the cleavage introduction groove 702 is divided in the <11-20> direction to obtain a bar. In this wafer division, unlike the first embodiment, a breaking blade is applied from the surface side of the LD structure 701 to the position where the cleavage introduction groove 702 is located, and the wafer is pushed. By doing in this way, in the divided | segmented bar | burr, the end surface which uses a cleaved surface can be formed in the part into which the striped waveguide 703 was divided | segmented. In addition, cleaving by impact given by hitting the blade, dividing by heating only the periphery of the crease line, or breaking by impact by sound waves or water flow, etc. You may make it.

By dividing in this way, many bars having a resonator length of 500 μm were obtained from the wafer as shown in FIG. The resonator length was within the set value of 500 μm ± 5 μm, and the bar yield was 92%. At this time, when the average of the unevenness of the end surface between the cleavage introduction grooves 702 was measured in the divided bars, the result was the same as that of the first embodiment. Therefore, compared with the case where a GaN-based semiconductor was deposited on the sapphire substrate, it was confirmed that the quality was improved because the cleaved end face was further flat.

As described above, the bar obtained by dividing the wafer as shown in FIG. 9 has the back surface or the front surface in the <1-100> direction between the striped waveguides 703, as in the first embodiment. A GaN-based semiconductor laser device is obtained by scoring and dividing. At this time, the stylus pressure at the time of scribing may be adjusted and divided by pushing and dividing in the <1-100> direction to obtain a GaN-based semiconductor laser device, or by cutting completely. It may be divided to obtain a GaN-based semiconductor laser element.

2. Configuration of GaN-based semiconductor laser device The configuration of the GaN-based semiconductor laser device 3 formed by dividing the wafer as described above will be described with reference to FIG.

In FIG. 10, the LD structure 31 includes an n-GaN contact layer 202, an n-AlGaN cladding layer 203, an n-GaN light guide layer 204, an InGaN multiple quantum well active layer 205, and p-AlGaN for simple explanation. It is assumed that the evaporation prevention layer 206, the p-GaN light guide layer 207, the p-AlGaN cladding layer 208, the p-GaN contact layer 209, and the SiO ₂ dielectric film 212 are included.

As described above, the GaN-based semiconductor laser device 3 obtained by dividing the wafer having the LD structure formed on the GaN-based substrate has the mirror end surface formed on the LD structure 31 formed on the GaN-based substrate 200 by cleavage. 32 is formed. In addition, a stripe-shaped waveguide 33 is provided in the LD structure 31 and plays a role of guiding laser light.

The n-electrode 210 provided on the lower surface of the GaN-based substrate 200 and the p-electrode 213 provided on the upper surface of the LD structure 31 are for supplying electric power from the outside during the operation of the GaN-based semiconductor laser device 3. . Further, groove portions 34 are formed on the GaN substrate 200 side at portions located at the four corners on the back surface of the GaN semiconductor laser device 3.

The grooving portion 34 corresponds to the cleavage introduction groove 702 (FIG. 9) formed in advance on the upper surface of the wafer in order to form the mirror end surface 32 when the wafer is divided into bars. At this time, in this embodiment, the depth d4 from the back surface of the GaN-based substrate 200 to the deepest portion of the groove portion 34 is 1 μm ≦ d4 <10 μm.

Further, in this embodiment, the grooving portion 34 is formed from a position away from the stripe-shaped waveguide 33 by 100 μm or more when the GaN-based semiconductor laser device 3 is two-dimensionally projected with the GaN-based substrate 200 facing down. ing. In this embodiment, the number of grooving portions 34 with respect to the GaN-based semiconductor laser element 3 is four. However, the number varies depending on the state of the cleavage introduction groove 702 formed in advance on the lower surface of the wafer. That's fine.

In this embodiment, when the wafer is divided to obtain a GaN-based semiconductor laser device, the grooved portion generated by the cleavage introduction groove portion may be cut off. This has an effect of removing dust or the like when forming the cleavage introduction groove.

<Fourth Embodiment>
A fourth embodiment of the present invention will be described below with reference to the drawings. FIG. 11 is a cross-sectional view and a top view for explaining the division of the wafer. FIG. 12 is an external perspective view of a divided GaN-based semiconductor laser device.

1. GaN semiconductor laser device manufacturing method
(Wafer formation)
The wafer formed in this embodiment is assumed to be a wafer represented by a cross-sectional view as shown in FIG. 2 as in the first embodiment. Therefore, the detailed description of the formation method will be omitted by referring to the first embodiment.

That is, on the n-GaN-based substrate 200, an n-GaN buffer layer 201, an n-GaN contact layer 202, an n-AlGaN cladding layer 203, an n-GaN light guide layer 204, an InGaN multiple quantum well active layer 205, p- The AlGaN evaporation prevention layer 206, the p-GaN light guide layer 207, the p-AlGaN cladding layer 208, and the p-GaN contact layer 209 are grown in order to form an LD structure. In addition, after the ridge stripe portion 211 is formed on the surface of the LD structure and the SiO ₂ dielectric film 212 is deposited, the p-GaN contact layer 209 is exposed. Then, a p-electrode material is deposited on the surface to form a p-electrode 213. Further, when the back surface of the n-GaN-based substrate 200 is shaved and the thickness of the wafer is adjusted, an n-electrode material is deposited on the back surface side of the n-GaN-based substrate to form the n-electrode 210. .

(Division of wafer)
Next, a wafer dividing method in this embodiment will be described with reference to FIG. FIG. 11A is a cross-sectional view of a wafer in which the LD structure is formed on the GaN-based substrate 200 as described above, and FIG. 11B is a GaN-based substrate in which the LD structure is formed as described above. A top view of a wafer formed on 200 is shown respectively.

In FIG. 11, for simplicity of explanation, it is assumed that the GaN-based substrate 250 includes the n-GaN-based substrate 200, the n-GaN buffer layer 201, and the n-electrode 210, as in FIG. 3. The structure 251 includes an n-GaN contact layer 202, an n-AlGaN cladding layer 203, an n-GaN light guide layer 204, an InGaN multiple quantum well active layer 205, a p-AlGaN evaporation prevention layer 206, a p-GaN light guide layer 207, A p-AlGaN cladding layer 208, a p-GaN contact layer 209, a SiO ₂ dielectric film 212, and a p-electrode 213 are assumed to be included.

Similarly to the first embodiment, when the LD structure 251 is formed on the GaN-based substrate 250 in the wafer, the striped waveguide 253 is formed in the LD structure 251 as shown in FIG. Is provided.

Such a wafer is placed on a dicer, which is an apparatus for cutting a groove in a semiconductor wafer with a diamond blade so that the back surface of the GaN-based substrate 250 faces upward, and has a depth d5 (0 <d5 ≦ 40 μm), a line A cleavage assisting groove 254 having a width w (0 <w ≦ 30 μm) is provided linearly along the <11-20> direction as shown in FIG. Further, the front and back surfaces of the wafer are reversed, and scribe lines are formed on the surface of the LD structure 251 with diamond needles on the surface of the LD structure 251 in the same manner as in the first embodiment. ), A broken-line cleavage introduction groove 252 is provided. Therefore, the cleavage introduction groove 252 is intermittent, whereas the cleavage assist groove 254 is continuous.

When the cleavage introduction groove 252 is provided in this way, the depth d from the surface of the LD structure 251 to the deepest part of the cleavage introduction groove 252 is at least 1 μm ≦ d ≦ 10 μm. By doing in this way, the yield of a bar can be improved. At this time, if the depth d1 from the interface between the GaN-based substrate 250 and the LD structure 251 to the deepest portion of the cleavage introduction groove 252 is set to 1 μm ≦ d1 ≦ 10 μm, the device yield is improved. be able to.

As described above, the groove insertion direction of the cleavage introduction groove 252 is <11-20> of the GaN-based substrate 250 and is aligned with the vicinity of the central axis of the cleavage assist groove 254. Further, by setting the start and end points of the cleavage introduction groove 252 to be 50 μm or more away from the striped waveguide 253, the bar can be divided with good yield, and more preferably, 100 μm or more. Thus, the device yield can be increased.

In this embodiment, as shown in FIG. 11A, the depth from the interface between the GaN-based substrate 250 and the LD structure 251 to the deepest part of the cleavage introduction groove 252 is constant at a depth of 1 μm. . The distance from the striped waveguide 253 to the start point and end point of the cleavage introduction groove 252 was set to 125 μm. For the cleavage assisting grooves 254, the depth d5 is 20 μm, the line width w is 20 μm, and the pitch p for each cleavage assisting groove 254 in the <1-100> direction is 500 μm.

Further, when the cleavage introduction groove 252 is provided in this way, in order to divide the bar with a high yield, it is preferable to form it as long as possible in the <11-20> direction within the above-mentioned range, but it may be a broken line shape. . Further, as a method for forming the cleavage introduction groove 252, dry etching such as RIE or wet etching may be used in addition to the scribe described above.

Next, the wafer thus provided with the cleavage introduction groove 252 is divided in the <11-20> direction to obtain a bar. In this division of the wafer, the breaking blade is applied to the position where the cleavage introduction groove 252 is located from the cleavage assisting groove 254 on the back surface side of the GaN-based substrate 250 to divide the wafer. By doing in this way, in the divided | segmented bar | burr, the end surface which uses a cleaved surface can be formed in the part into which the striped waveguide 253 was divided | segmented. In addition, cleaving by impact given by hitting the blade, dividing by heating only the periphery of the crease line, or breaking by impact by sound waves or water flow, etc. You may make it.

By dividing in this way, many bars having a resonator length of 500 μm were obtained from the wafer as shown in FIG. The resonator length was within the set value of 500 μm ± 5 μm, and the bar yield was 96%. At this time, when the average of the unevenness of the end face between the cleavage introduction grooves 252 was measured in the divided bars, the result was the same as that of the first embodiment. Therefore, compared with the case where a GaN-based semiconductor was deposited on the sapphire substrate, it was confirmed that the quality was improved because the cleaved end face was further flat.

In this way, the bar obtained by dividing the wafer as shown in FIG. 11 is divided by marking the back surface or the front surface in the <1-100> direction between the striped waveguides 253. A GaN-based semiconductor laser device is obtained. At this time, the stylus pressure at the time of scribing may be adjusted and divided by pushing and dividing in the <1-100> direction to obtain a GaN-based semiconductor laser device, or by cutting completely. It may be divided to obtain a GaN-based semiconductor laser element.

2. Configuration of GaN-based semiconductor laser device The configuration of the GaN-based semiconductor laser device 1a formed by dividing the wafer as described above will be described with reference to FIG. In the GaN-based semiconductor laser device 1a of FIG. 12, the same parts as those of the GaN-based semiconductor laser device 1 of FIG. 4 are denoted by the same reference numerals, and detailed description thereof is omitted.

As in the first embodiment, the GaN-based semiconductor laser device 1a manufactured by the manufacturing method of the present embodiment has a cleavage introduction groove 252 (see FIG. 5) on the LD structure 11 side at the four corners on the surface. A grooving portion 14 corresponding to 11) is formed. Further, groove portions 15 are formed at two positions on the mirror end face side on the back side of the GaN-based substrate 10. This grooving portion 15 corresponds to a cleavage assist groove 254 (FIG. 11) formed in advance on the lower surface of the wafer.

At this time, the number of grooved portions 14 with respect to the GaN-based semiconductor laser device 1a is four in the present embodiment, but varies depending on the state of the cleavage introducing groove 252 formed in advance on the upper surface of the wafer, and is at least one or more. I just need it. Similarly, the number of grooving portions 15 is two in this embodiment. However, the number of grooving portions 15 varies depending on the state of the cleavage assisting groove 254 formed in advance on the lower surface of the wafer.

In this embodiment, when the wafer is divided to obtain a GaN-based semiconductor laser device, the grooved portion generated by the cleavage introduction groove portion may be cut off. This has an effect of removing dust or the like when forming the cleavage introduction groove.

In this embodiment, a GaN-based semiconductor laser device is divided by providing a cleavage introduction groove and a cleavage assist groove on the upper and lower surfaces of the wafer having the same configuration as that of the first embodiment, respectively. However, a GaN-based semiconductor laser is provided in which a cleavage introduction groove and a cleavage auxiliary groove are provided on the upper surface and the lower surface of the wafer, respectively, with respect to a wafer having the same configuration as that of the second embodiment (FIG. 5). The element may be divided.

In each of the embodiments described above, a specific surface is selected and described as the formation direction of the mirror end surface. However, the {0001} plane, the {11-20} plane, and the {1} plane, which are unique to the hexagonal GaN-based semiconductor, are described. A plane parallel to any of the −100} planes may be selected. However, among these, the {1-100} plane is preferable because of good cleaving properties. That is, it is preferable that any one of the (1-100), (10-10), (01-10), (-1100), (-1010), and (0-110) surfaces be the mirror end surface.

The optical waveguide structure of the semiconductor laser element to which the present invention is applied is not limited to the examples shown in the above embodiments. Starting with the ridge structure shown in each of the above-described embodiments, the self-aligned structure (SAS) structure, the electrode stripe structure, the buried hetero (BH) structure, the channeled substrate planar (CSP) structure, etc. The structure is not related to the essence of the present invention, and the same effect as described above can be obtained.

1, 1a, 2, 3 GaN-based semiconductor laser device 10, 200 GaN-based substrate 11, 21, 31 LD structure 12, 22, 32 Mirror end faces 13, 23, 33 Striped waveguides 14, 15, 25, 35 Grooving Part 201 n-GaN buffer layer 202 n-GaN contact layer 203 n-AlGaN cladding layer 204 n-GaN light guide layer 205 InGaN multiple quantum well active layer 206 p-AlGaN evaporation prevention layer 207 p-GaN light guide layer 208 p- AlGaN cladding layer 209 p-GaN contact layer 210 n-electrode 211 ridge stripe portion 212 SiO ₂ dielectric film 213 p-electrode

Claims (18)

Nitride semiconductor substrate composed of group III element and nitrogen-containing compound, nitride semiconductor layer composed of group III element and nitrogen-containing compound deposited on nitride semiconductor substrate, and nitride semiconductor A nitride semiconductor device having a stripe-shaped optical waveguide formed in a layer, and forming a resonator with the end surface obtained by cleaving the nitride semiconductor layer and the stripe-shaped optical waveguide; A nitride semiconductor device comprising a cleaved introduction groove formed from the surface side of the nitride semiconductor layer at a position other than directly above the stripe-shaped optical waveguide. The nitride semiconductor device according to claim 1, further comprising a cleavage assist groove formed from a back surface side of the nitride semiconductor substrate. Nitride semiconductor substrate composed of group III element and nitrogen-containing compound, nitride semiconductor layer composed of group III element and nitrogen-containing compound deposited on nitride semiconductor substrate, and nitride semiconductor In a nitride semiconductor device having a stripe-shaped optical waveguide formed in a layer, and forming a resonator with the end surface obtained by cleaving the nitride semiconductor semiconductor layer and the stripe-shaped optical waveguide, with respect to the end surface, In addition to the position directly below the stripe-shaped optical waveguide, the nitride semiconductor device has a cleave introduction groove formed from the back surface side of the nitride semiconductor substrate. 3. The nitride according to claim 1, wherein a depth d from a front surface or a back surface of the nitride semiconductor element to a deepest portion of the cleavage introduction groove is in a range of 1 ≦ d ≦ 10 μm. Semiconductor element. 2. The nitridation according to claim 1, wherein a depth d from an interface between the nitride semiconductor substrate and the nitride semiconductor layer to a deepest portion of the cleavage introduction groove is in a range of 1 ≦ d ≦ 10 μm. Semiconductor device. The cleavage introduction groove is at least 50 μm away from the stripe optical waveguide in a direction perpendicular to the stripe optical waveguide when the nitride semiconductor element is projected two-dimensionally with the nitride semiconductor substrate down. The nitride semiconductor device according to claim 1, wherein the nitride semiconductor device is formed at a position. The cleavage introduction groove is at least 100 μm away from the stripe optical waveguide in a direction perpendicular to the stripe optical waveguide when the nitride semiconductor element is projected two-dimensionally with the nitride semiconductor substrate down. The nitride semiconductor device according to claim 5, wherein the nitride semiconductor device is formed at a position. The nitride semiconductor device according to claim 1, wherein the nitride semiconductor device has a thickness of 80 to 160 μm.  A nitride semiconductor layer made of a compound containing a group III element and nitrogen is deposited on a nitride semiconductor substrate made of a compound containing a group III element and nitrogen. In a nitride semiconductor device manufacturing method for obtaining a nitride semiconductor device from a nitride semiconductor wafer having a plurality of stripe-shaped optical waveguides formed at equal intervals, the nitride semiconductor wafer is adjusted to a thickness of 80 to 160 μm. A step of forming a plurality of cleavage introduction grooves in an intermittent broken line shape by scoring from the surface side of the nitride semiconductor layer with respect to the nitride semiconductor wafer; and along the cleavage introduction grooves Dividing the nitride semiconductor wafer, and when forming the cleavage introduction groove, the cleavage introduction groove is formed at a position other than directly above the stripe optical waveguide. Method of manufacturing a nitride semiconductor device characterized. In the nitride semiconductor layer, when a semiconductor layer made of a material having a cleavage direction different from that of the nitride semiconductor is provided at the interface with the nitride semiconductor substrate, first, by scoring from the surface side of the nitride semiconductor layer Forming a plurality of cleavage assisting grooves in an intermittent broken line shape from the surface side of the nitride semiconductor layer to a depth half the thickness of the nitride semiconductor layer, and then scoring from the bottom surface of the cleavage assisting groove. The method of manufacturing a nitride semiconductor device according to claim 9, wherein the cleavage introduction groove is formed. Before the step of dividing the nitride semiconductor wafer, the method further comprises a step of forming a cleave assist groove by scoring the nitride semiconductor wafer from the back side of the nitride semiconductor substrate, and the cleaving assist 11. The method for manufacturing a nitride semiconductor device according to claim 9, wherein the cleavage introduction groove and the cleavage auxiliary groove are formed so that the cleavage introduction groove is located at a central axis of the groove. . A nitride semiconductor layer made of a compound containing a group III element and nitrogen is deposited on a nitride semiconductor substrate made of a compound containing a group III element and nitrogen. In a nitride semiconductor device manufacturing method for obtaining a nitride semiconductor device from a nitride semiconductor wafer having a plurality of stripe-shaped optical waveguides formed at equal intervals, the nitride semiconductor wafer is adjusted to a thickness of 80 to 160 μm. A step of forming a plurality of cleavage introduction grooves in an intermittent broken line shape by marking the nitride semiconductor wafer from the back side of the nitride semiconductor substrate, and along the cleavage introduction grooves. Dividing the nitride semiconductor wafer, and when forming the cleavage introduction groove, the cleavage introduction groove is formed at a position other than directly under the stripe optical waveguide. Method of manufacturing a nitride semiconductor device characterized and. 13. The depth d from the front surface side or the back surface side of the nitride semiconductor wafer to the deepest portion of the cleavage introduction groove is in the range of 1 ≦ d ≦ 10 μm. A method for producing a nitride semiconductor device according to claim 1. The depth d from the interface between the nitride semiconductor substrate and the nitride semiconductor layer to the deepest portion of the cleavage introduction groove is in the range of 1 ≦ d ≦ 10 μm. A method for producing a nitride semiconductor device according to any one of the above. The method for manufacturing a nitride semiconductor device according to any one of claims 9 to 14, wherein the cleavage introduction grooves are provided at intervals of 1 mm or less on the same broken line. The method for manufacturing a nitride semiconductor device according to any one of claims 9 to 15, wherein the cleavage introduction grooves are provided between the striped optical waveguides in the same broken line shape. The method for manufacturing a nitride semiconductor device according to any one of claims 9 to 16, wherein the cleavage introduction groove itself is formed in a continuous linear shape. The method for manufacturing a nitride semiconductor device according to any one of claims 9 to 16, wherein the cleavage introduction groove itself is formed in an intermittent broken line shape.

Images