JP5630434B2 - Manufacturing method of semiconductor device - Google Patents

Description

  The present invention relates to a method for manufacturing a semiconductor element. More specifically, the present invention relates to a method for manufacturing a semiconductor device in which an ohmic contact is improved between a p-type contact layer made of a group III nitride semiconductor and a p-electrode.

  In general, a semiconductor element has a p-electrode and a p-type contact layer in contact with the p-electrode. If the contact resistance between the p-electrode and the p-type contact layer is large, heat is likely to be generated. And there exists a possibility that lifetime may become short by heat_generation | fever. Further, when the contact resistance is large, the driving voltage of the semiconductor element becomes high.

  Particularly, in order to reduce the contact resistance between the p-type contact layer made of a group III nitride semiconductor and the p-electrode, it is necessary to increase the hole concentration of the p-type contact layer. However, the acceptor impurity level formed in the group III nitride semiconductor is deep. Therefore, it is difficult to increase the hole concentration.

Therefore, many studies have been made to increase the hole concentration. For example, in Patent Document 1, a hole concentration of about 10 ¹⁸ / cm ³ is realized by reducing the growth rate of the p-type contact layer.

JP 2003-23179 A

  However, even if a p-electrode is formed on a p-type contact layer having such a hole concentration, a good ohmic contact cannot be obtained. On the other hand, it is difficult to obtain an ohmic contact between a p-type contact layer made of a group III nitride semiconductor and a p-electrode because of the work function of both.

  The present invention has been made to solve the above-described problems of the prior art. That is, an object of the present invention is to provide a method for manufacturing a semiconductor device in which ohmic contact resistance between a p-type contact layer and a p-electrode of a nitride semiconductor is improved.

A method of manufacturing a semiconductor device according to the first aspect of the invention for solving this problem includes a light emitting layer forming step of forming a light emitting layer, and a p type cladding layer forming of forming a p type cladding layer on the light emitting layer. And a p-type contact layer forming step of forming a p-type contact layer on the p-type cladding layer, and a p-electrode forming step of forming a p-electrode on the p-type contact layer. The p-type contact layer forming step includes a first step of forming a first p-type contact layer on the p-type cladding layer, and a second p-type contact layer on the first p-type contact layer. And a second step. In the first step, a mixed gas of nitrogen and hydrogen is used as a carrier gas, and the molar ratio of nitrogen to the total number of moles of the carrier gas is within a range of 50% to 75%, and the first p-type contact The Mg concentration of the layer is in the range of 1 × 10 ¹⁹ / cm ^{3 to} 1 × 10 ²⁰ / cm ³ , and the thickness of the first p-type contact layer is in the range of 100 to 1000 μm. In the second step , hydrogen is used as a carrier gas, and the Mg concentration of the second p-type contact layer is 2 × 10 ²⁰ / cm ³ or more, which is higher than the Mg concentration of the first p-type contact layer, 9 × The thickness is within the range of 10 ²¹ / cm ³ or less, and the thickness of the second p-type contact layer is within the range of 20 to 90 cm .

In a semiconductor element manufactured by such a method for manufacturing a semiconductor element, ohmic contact between the p-type contact layer and the p-electrode is good. That is, the contact resistance is small. Therefore, power consumption can be suppressed. And the calorific value is also small. Further, in the case of this film thickness, carriers easily pass through the potential barrier due to the tunnel effect through lattice defects. In addition, when the Mg concentration is used, more lattice defects can be formed in the second p-type contact layer. For this reason, after Mg is activated, electrons move smoothly between the p-electrode and the p-type contact layer.

In the method for manufacturing a semiconductor device according to the second invention, an annealing process for reducing the resistance of the p-type contact layer is further performed. The hole concentration is set in the range of 5 × 10 ¹⁶ / cm ^{3 to} 6 × 10 ¹⁶ / cm ³ .

In the method for manufacturing a semiconductor device according to the third aspect of the present invention, an annealing process for reducing the resistance of the p-type contact layer is not performed. The hole concentration is in the range of 2 × 10 ¹⁶ / cm ^{3 to} 3 × 10 ¹⁶ / cm ³ and the electrical resistivity is in the range of 40 Ω · cm to 70 Ω · cm. By omitting the annealing process, cycle time is shortened and productivity is improved.

  According to the present invention, there is provided a method for manufacturing a semiconductor device in which the ohmic property of an ohmic contact between a nitride semiconductor p-type contact layer and a p-electrode is improved.

It is a schematic block diagram which shows the laminated structure of the semiconductor element which concerns on an Example. BRIEF DESCRIPTION OF THE DRAWINGS It is FIG. (1) for demonstrating the manufacturing method of the semiconductor element based on an Example. It is FIG. (2) for demonstrating the manufacturing method of the semiconductor element which concerns on an Example. It is a graph which shows the hole concentration of the 1st p-type contact layer in the light emitting element manufactured with the manufacturing method of the semiconductor element which concerns on an Example. It is a graph which shows the mobility of the hole of the 1st p-type contact layer in the light emitting element manufactured with the manufacturing method of the semiconductor element which concerns on an Example. It is a graph which shows the electrical resistivity of the 1st p-type contact layer in the light emitting element manufactured with the manufacturing method of the semiconductor element which concerns on an Example. It is a graph which shows the activation rate of the 1st p-type contact layer in the light emitting element manufactured with the manufacturing method of the semiconductor element which concerns on an Example. 4 is a graph comparing outputs of a light emitting device manufactured by a method for manufacturing a semiconductor device according to an example and a conventional light emitting device.

  Hereinafter, specific examples of the present invention will be described with reference to the drawings by taking light emitting elements as examples. However, the present invention is not limited to these examples. That is, the present invention can also be applied to various semiconductor devices such as transistors such as FEMT, light receiving elements, and light emitting elements such as LEDs and laser diodes. Moreover, the laminated structure of each layer of the light emitting element to be described later is an exemplification, and of course, a laminated structure different from the example may be used. And the thickness of each layer in each figure is shown conceptually and does not indicate the actual thickness.

1. Semiconductor Device A light emitting device 100 manufactured by the method of manufacturing a semiconductor device according to this example will be described with reference to FIG. The light emitting element 100 is a semiconductor element made of a group III nitride semiconductor. As shown in FIG. 1, the light emitting device 100 includes a sapphire substrate 10, a low-temperature buffer layer 20, an n-type contact layer 30, an n-type ESD layer 40, an n-type SL layer 50, and an MQW layer that is a light source. (Multiple quantum well layer) 60, p-type cladding layer 70, and p-type contact layer 80 are formed in this order. Further, the n-type contact layer 30 is formed with an n-electrode N1. A p-electrode P <b> 1 is formed on the p-type contact layer 80.

  The sapphire substrate 10 is for forming each of the above layers on one surface thereof by MOCVD. In order to improve the light extraction efficiency, the surface of the sapphire substrate 10 may be processed to be uneven. In addition to sapphire, SiC, ZnO, Si, GaN, or the like may be used as the growth substrate. The low temperature buffer layer 20 is for forming an upper layer while inheriting the crystallinity of the sapphire substrate 10. Examples of the material of the low temperature buffer layer 20 include AlN and GaN.

The n-type contact layer 30 is a layer that actually contacts the n-electrode N1. The n-type contact layer 30 is a layer made of GaN doped with Si. The Si concentration is 1 × 10 ¹⁸ / cm ³ or more. The n-type contact layer 30 may be a plurality of layers having different carrier concentrations. This is to improve the ohmic property with the n-electrode.

The n-type ESD layer 40 is an electrostatic withstand voltage improvement layer for preventing electrostatic breakdown of each semiconductor layer. The n-type ESD layer 40 has a laminated structure of non-doped GaN and Si-doped GaN. When doping Si, the carrier concentration is preferably 1 × 10 ¹⁸ / cm ³ or more.

  The n-type SL layer 50 is a layer having a superlattice structure for relaxing stress applied to the MQW layer 60. The n-type SL layer 50 is a layer in which GaN and InGaN are alternately stacked. Alternatively, in addition to these, n-GaN may be stacked. In particular, the layer in contact with the MQW layer 60 may be a layer made of n-GaN. The number of repetitions of repeatedly laminating the n-type SL layer 50 is 10 to 20 times. The total thickness of the n-type SL layer 50 is 60 nm to 80 nm.

  The MQW layer 60 is a light emitting layer that emits light when electrons and holes are recombined. Therefore, in the MQW layer 60, a well layer having a small band gap and a barrier layer having a large band gap are alternately formed. Here, InGaN can be used as the well layer and AlGaN can be used as the barrier layer. Further, GaN may be used as the well layer and AlGaN may be used as the barrier layer. Alternatively, these unit structures may be repeated by freely combining them, and using four or more layers as a unit structure.

  The p-type cladding layer 70 is for preventing electrons from diffusing into the p-type contact layer 80. The p-type cladding layer 70 is a layer formed by alternately repeating the unit structure using a layer made of p-InGaN and a layer made of p-AlGaN as a unit structure. The number of repetitions is 12 times. Moreover, it is good also considering the repetition frequency as the range of 3-50 times.

  The p-type contact layer 80 has a first p-type contact layer 81 and a second p-type contact layer 82. These are all layers made of p-GaN doped with Mg. The second p-type contact layer 82 is a layer that actually contacts the p-electrode P1. Therefore, the second p-type contact layer 82 appears on the surface of the light emitting element 100 opposite to the sapphire substrate 10. The first p-type contact layer 81 is below the second p-type contact layer 82.

Here, the Mg doping amount in the first p-type contact layer 81 is in the range of 1 × 10 ¹⁹ / cm ^{3 to} 1 × 10 ²⁰ / cm ³ . This Mg concentration range is a range in which a high hole concentration can be obtained without lowering the crystallinity. The Mg doping amount in the second p-type contact layer 82 is in the range of 1 × 10 ²⁰ / cm ^{3 to} 1 × 10 ²² / cm ³ . Thus, the Mg concentration of the second p-type contact layer 82 is high. The Mg doping amount in the second p-type contact layer 82 is larger than the Mg doping amount in the first p-type contact layer 81.

  The thickness of the second p-type contact layer 82 is in the range of 10 to 100 mm. Thus, the thickness of the second p-type contact layer 82 is sufficiently thin. Further, as will be described later, the carrier gas used for forming the second p-type contact layer 82 is only hydrogen and does not contain nitrogen. Therefore, the crystallinity of the second p-type contact layer 82 is poor. Therefore, the Schottky barrier formed between the p-electrode P1 and the second p-type contact layer 82 is extremely thin.

  For this reason, holes are likely to escape from the p-electrode P1 to the second p-type contact layer 82. That is, holes from the p-electrode P1 tunnel through the Schottky barrier and are easily injected into the second p-type contact layer 82. Therefore, a suitable ohmic contact can be obtained between the p-electrode P1 and the p-type contact layer 80.

2. Semiconductor Device Manufacturing Method In the semiconductor device manufacturing method according to this embodiment, crystals of the above-described layers are grown by metal organic chemical vapor deposition (MOCVD). The semiconductor element manufacturing method of this embodiment is a method characterized by the step of forming the p-type contact layer 80. Hereafter, each process is demonstrated using FIG. 2 and FIG.

The carrier gas used here is hydrogen (H ₂ ), nitrogen (N ₂ ), or a mixed gas of hydrogen and nitrogen (H ₂ + N ₂ ). Ammonia gas (NH ₃ ) was used as a nitrogen source. Trimethylgallium (Ga (CH ₃ ) ₃ : hereinafter referred to as “TMG”) was used as the Ga source. Trimethylindium (In (CH ₃ ) ₃ : hereinafter referred to as “TMI”) was used as the In source. Trimethylaluminum (Al (CH ₃ ) ₃ : hereinafter referred to as “TMA”) was used as the Al source. Silane (SiH ₄ ) was used as the n-type dopant gas. Cyclopentadienylmagnesium (Mg (C ₅ H ₅ ) ₂ : hereinafter referred to as “CP ₂ Mg”) was used as the p-type dopant gas.

2-1. Low temperature buffer layer formation process In this form, sapphire substrate 10 was used. Then, the sapphire substrate 10 was put in a MOCVD furnace. Next, the sapphire substrate 10 was cleaned in hydrogen gas, and the deposits adhered to the surface of the sapphire substrate 100 were removed. Then, the substrate temperature was set to 400 ° C., and the low temperature buffer layer 20 made of AlN was formed on the sapphire substrate 10.

2-2. Step of forming n-type contact layer Next, an n-type contact layer 30 was formed on the low-temperature buffer layer 20 (see FIG. 2A). Here, the carrier gas was hydrogen gas, and the substrate temperature was raised to 1100 ° C. while flowing ammonia gas. When the substrate temperature reached 1100 ° C., silane gas was supplied as TMG, ammonia gas, and impurity gas. Thereby, the n-type contact layer 30 made of n-GaN having a Si concentration of 4.5 × 10 ¹⁸ / cm ³ was formed.

2-3. Step of forming n-type ESD layer Next, an n-type ESD layer 40 was formed on the n-type contact layer 30. The substrate temperature was lowered to 900 ° C. to form a laminated structure of non-doped GaN and Si-doped n-GaN. The growth temperature at this time should just be 800-950 degreeC. The characteristic value defined by the product of the Si atom concentration (atom / cm ³ ) and the film thickness (nm) in this n-GaN is 0.9 × 10 ^{20 to} 3.6 × 10 ²⁰ (atom · nm / cm ³ ).

2-4. Step of forming n-type SL layer Next, an n-type SL layer 50 was formed on the n-type ESD layer 40. As the n-type SL layer 50, 2.5 nm thick InGaN layers and 2.5 nm thick Si-doped n-GaN layers were alternately stacked. This periodic structure was repeatedly formed for 15 periods. When forming the InGaN layer, the substrate temperature was set to 830 ° C., and silane gas, TMG, TMI, and ammonia were supplied. When forming the n-GaN layer, the substrate temperature was set to 830 ° C., and silane gas, TMG, and ammonia were supplied. As a result, the laminated structure shown in FIG. 2B was formed.

2-5. Light-Emitting Layer Formation Step Subsequently, the MQW layer 60 was formed on the n-type SL layer 50. The MQW layer 60 has a repeating structure of an InGaN layer and an AlGaN layer. The InGaN layer was grown at a temperature in the range of 750-800 ° C. Therefore, TMI, TMG, and ammonia source gases were supplied. Here, the composition ratio of In is in the range of 0.05 to 0.15%. The thickness for growing the crystal is 1 to 4 nm.

  The AlGaN layer was grown at a temperature in the range of 850 to 950 ° C. For this purpose, TMA, TMG, and ammonia source gases were supplied. The thickness for growing the crystal is 1 to 6 nm. Then, five layers of these InGaN layers and AlGaN layers were alternately laminated. The number of layers to be stacked may be about 3 to 7 layers.

2-6. Step of forming p-type cladding layer Next, a p-type cladding layer 70 was formed on the MQW layer 60. The p-type cladding layer 70 has a repeating structure of a p-InGaN layer and a p-AlGaN layer. When forming the p-InGaN layer, CP ₂ Mg, TMI, TMG, and ammonia were supplied at a substrate temperature of 855 ° C. A p-In _0.05 Ga _0.95 N layer was formed to a thickness of 1.7 nm.

When forming the p-AlGaN layer, CP ₂ Mg, TMA, TMG, and ammonia were supplied at a substrate temperature of 855 ° C. A p-AlGaN layer was formed to a thickness of 3.0 nm. As a result, the laminated structure shown in FIG. 2C was formed.

2-7. p-type contact layer forming step 2-7-1. First p-type contact layer forming step Subsequently, a first p-type contact layer 81 was formed on the p-type cladding layer 70 (see FIG. 3). A mixed gas of nitrogen and hydrogen was used as the carrier gas.

  The presence of hydrogen increases the migration of constituent atoms. Therefore, the crystal quality is improved and the surface flatness of the layer is improved. However, on the other hand, H is taken into the crystal and bonded to Mg. Thereby, activation of Mg cannot be inhibited and the hole concentration cannot be increased.

  Nitrogen gas, on the other hand, inhibits the decomposition of the crystal and prevents N from escaping from the crystal. However, on the other hand, it becomes a factor of reducing crystallinity. Therefore, there is an optimum range for the molar ratio of nitrogen to the total carrier gas.

A desirable range of the molar ratio of nitrogen to the total carrier gas, that is, the mixing ratio (N ₂ / (H ₂ + N ₂ )) was 40% or more and 80% or less . And the range of the more preferable mixing ratio was 50% or more and 75% or less. The specific numerical value will be described later. When the mixing ratio (N ₂ / (H ₂ + N ₂ )) is lower than 40%, a sufficient hole concentration cannot be obtained as will be described later. If the mixing ratio (N ₂ / (H ₂ + N ₂ )) is higher than 80%, surface roughness and pits may increase as described later.

  Here, the temperature for growing the crystals was in the range of 900 ° C. or higher and 1050 ° C. or lower. This is because if the temperature is too low, the crystal quality of GaN will deteriorate. This is because at temperatures higher than 1050 ° C., the reaction occurs before each gas reaches the sapphire substrate 10.

The doping amount of Mg was 1 × 10 ¹⁹ / cm ³ or more and 1 × 10 ²⁰ / cm ³ or less. This is because if the Mg doping amount is 1 × 10 ²⁰ / cm ³ or less, lattice defects are unlikely to enter the first p-type contact layer 81 to be formed. The film thickness of the first p-type contact layer 81 to be formed was in the range of 100 to 1000 mm.

2-7-2. Second p-type contact layer forming step Next, a second p-type contact layer 82 was formed on the first p-type contact layer 81. Only hydrogen was used as the carrier gas. This is because many lattice defects are intentionally formed in the second p-type contact layer 82. Therefore, the supply of nitrogen gas was stopped and only hydrogen gas was supplied as a carrier gas. Therefore, the carrier gas does not contain nitrogen. However, the formation process of each of these layers is continuously performed inside the MOCVD furnace. Therefore, residual nitrogen gas may be contained in the furnace atmosphere.

  The crystal growth temperature was in the range of 800 ° C. or higher and 1050 ° C. or lower. When the growth temperature is lower than 800 ° C., the crystal quality of GaN deteriorates. On the other hand, if the growth temperature is higher than 1050 ° C., ammonia, Ga, Mg or the like may react before reaching the location of the sapphire substrate 10 in the atmosphere in the furnace.

The amount of Mg doped into the second p-type contact layer 82 was in the range of 1 × 10 ²⁰ / cm ^{3 to} 1 × 10 ²² / cm ³ . This is because when the Mg doping amount is 1 × 10 ²⁰ / cm ³ or more, lattice defects are likely to enter the semiconductor layer to be formed.

  The film thickness of the second p-type contact layer 82 was in the range of 10 to 100 mm. This is because if it is thinner than 10 mm, it is difficult to form the lattice defect itself. Since the lattice constant of GaN in the c-axis direction is 5.185Å, a lattice defect is more easily formed with a thickness of two or more layers. On the other hand, the thicker the layer having lattice defects, the higher the electrical resistance. Therefore, 100 mm or less is preferable. More preferably, it is in the range of 20 to 90 cm. Moreover, it is still more preferable in it being in the range of 30 to 70 mm.

2-8. Next, the MOCVD furnace was cooled to room temperature in a nitrogen gas atmosphere. This is because the stacked body 90 shown in FIG. 3 is cooled in a nitrogen gas atmosphere to prevent the separated hydrogen from being taken into the stacked body 90 again.

2-9. Next, dry etching was performed from the surface side of the p-type contact layer 80 to form a groove reaching the middle of the n-type contact layer 30. Then, a p-electrode P1 was formed on the p-type contact layer 80. As the p-electrode P1, an Ni layer, an Au layer, and an Al layer were formed in this order on the p-type contact layer 80. Alternatively, ITO may be used in addition to these metals. Alternatively, a wiring electrode made of Ni / Au may be formed on the ITO electrode. Ag and Rh can also be used. An n electrode N1 was formed on the exposed n-type contact layer 30. As the n-electrode N1, an Ni layer and an Au layer were formed on the n-type contact layer 30 in this order. Alternatively, a Ti layer and an Al layer may be formed in order.

2-10. Next, a heat treatment (annealing process) was performed on the laminate 90 in a nitrogen atmosphere. This is to activate the doped Mg. The annealing process may be performed before the electrode formation process. It can also be performed before the cooling step. Thus, the light emitting device 100 illustrated in FIG. 1 was manufactured.

3. Manufactured Semiconductor Element In the light emitting element 100 of this embodiment, the p-type contact layer 80 has a first p-type contact layer 81 and a second p-type contact layer 82. Since there is the second p-type contact layer 82, the thickness of the Schottky barrier between the p-type contact layer 80 and the p-electrode P1 is thin. Therefore, the hole conductivity is high between the p-type contact layer 80 and the p-electrode P1.

  There are many lattice defects in the second p-type contact layer 82. Of course, there are more than the first p-type contact layer 81. Therefore, the hole conductivity between the p-electrode P1 and the second p-type contact layer 82 is higher.

  A case where the second p-type contact layer 82 is not provided will be described. When the p-electrode P1 and the first p-type contact layer 81 are brought into contact with each other, the drive voltage Vf has a larger value than when the second p-type contact layer 82 is not provided. That is, it is considered that a relatively thick Schottky barrier is formed between the first p-type contact layer 81 and the p-electrode P1.

  Therefore, by providing the second p-type contact layer 82, the Schottky barrier becomes thin. That is, holes easily move between the p-electrode P1 and the p-type contact layer 80. The second p-type contact layer 82 has more lattice defects than the first p-type contact layer 81. Therefore, the holes are more easily moved between the p electrode P1 and the p-type contact layer 80. Therefore, the electrical resistivity of the light emitting element 100 of this embodiment is small.

4). Experimental Results Here, experimental results performed on the light emitting element 100 of this embodiment will be described. Hereinafter, measured values of each physical quantity when the mixing ratio of nitrogen in the carrier gas is changed when the first p-type contact layer 81 is formed will be described.

4-1. Hole Concentration FIG. 4 is a graph showing the hole concentration of the first p-type contact layer 81 with respect to the nitrogen mixing ratio (N ₂ / (H ₂ + N ₂ )). In FIG. 4, white symbols indicate values when the annealing process is not performed. The black symbols indicate values when annealing is performed. These are the same in FIG. 5 and FIG.

As shown in FIG. 4, as a whole, the hole concentration when the annealing treatment is performed is higher than the hole concentration when the annealing treatment is not performed. That is, when annealing is performed, the hole concentration value is sufficiently high regardless of the mixing ratio of nitrogen. A hole concentration value of about 5 × 10 ¹⁶ / cm ^{3 to} 6 × 10 ¹⁶ / cm ³ was obtained.

On the other hand, when the annealing treatment was not performed, the hole concentration was higher when the nitrogen concentration was increased. When the mixing ratio of nitrogen was 44%, a hole concentration value of about 2 × 10 ¹⁶ / cm ^{3 to} 3 × 10 ¹⁶ / cm ³ was obtained without annealing. That is, the hole concentration is about ½ of that when annealing is performed. Therefore, it is considered that hydrogen H is bonded to half of the Mg atoms before the annealing treatment.

When the nitrogen mixing ratio was 66%, a hole concentration value of about 5 × 10 ¹⁶ / cm ³ was obtained without annealing. That is, the hole concentration was the same value as when annealing was performed. Therefore, it is considered that there is almost no bonding of hydrogen H to Mg atoms even before annealing. Thus, when the mixing ratio of nitrogen was 44% or 66%, the hole concentration value was sufficiently high.

4-2. FIG. 5 shows the mobility of holes in the first p-type contact layer 81 with respect to the mixing ratio of nitrogen (N ₂ / (H ₂ + N ₂ )). As shown in FIG. 5, the hole mobility is almost the same between the case where the annealing treatment is performed and the case where the annealing treatment is not performed. As the concentration of nitrogen gas is increased, the hole mobility is improved. That is, lattice defects are reduced. Therefore, it can be seen that a p-GaN layer with good crystal quality could be formed.

When the mixing ratio of nitrogen was 22%, the hole mobility was approximately 2 cm ² / V · s although the annealing treatment was not performed. The hole mobility of the annealed material was approximately 3 cm ² / V · s. When the mixing ratio of nitrogen was 44%, the hole mobility was 4 cm ² / V · s regardless of the presence or absence of the annealing treatment. When the mixing ratio of nitrogen was 66%, the hole mobility was 7 to 8 cm ² / V · s regardless of the presence or absence of the annealing treatment.

  Thus, the difference due to the presence or absence of annealing treatment is small. In particular, when the mixing ratio of nitrogen is 44% and 66%, there is almost no difference depending on the presence or absence of annealing treatment. Therefore, a p-GaN layer with good hole conductivity can be formed.

4-3. Electrical Resistivity FIG. 6 is a graph showing the electrical resistivity in the first p-type contact layer 81 with respect to the nitrogen mixing ratio (N ₂ / (H ₂ + N ₂ )). As shown in FIG. 6, overall, the greater the nitrogen mixing ratio, the lower the electrical resistivity. This decrease in electrical resistivity is thought to be due to an improvement in crystal quality and a decrease in resistance components due to lattice defects. This is understood from the fact that the hole mobility is improved.

  When the mixing ratio of nitrogen was 22%, the electrical resistivity of the first p-type contact layer 81 that was not annealed was about 110 Ω · cm. The electrical resistivity of the first p-type contact layer 81 that was annealed was about 40 to 50 Ω · cm. The electrical resistivity of the first p-type contact layer 81 that has not been annealed is about twice the electrical resistivity of the first p-type contact layer 81 that has been annealed. That is, before the annealing process, hydrogen H remains bonded to half of the Mg atoms.

  When the mixing ratio of nitrogen was 44%, the electrical resistivity of the first p-type contact layer 81 that was not annealed was about 40 to 70 Ω · cm. This is similar to the case where the annealing process is performed when the mixing ratio of nitrogen is 22%. When the nitrogen mixing ratio was 44%, the electrical resistivity of the annealed first p-type contact layer 81 was about 20 Ω · cm. The electrical resistivity of the first p-type contact layer 81 that has not been annealed is about twice the electrical resistivity of the first p-type contact layer 81 that has been annealed.

  When the mixing ratio of nitrogen was 66%, the electrical resistivity of the first p-type contact layer 81 that was not annealed was about 15 Ω · cm. When the nitrogen mixing ratio was 66%, the electrical resistivity of the first p-type contact layer 81 subjected to the annealing treatment was about 12 Ω · cm. That is, regardless of the presence or absence of annealing treatment, the electrical resistivity of both is comparable.

  The electrical resistivity (15 Ω · cm) of the first p-type contact layer 81 that is not annealed when the nitrogen mixing ratio is 66% is that the annealing treatment is performed when the nitrogen mixing ratio is 22%. It is sufficiently lower than the electrical resistivity (40-50 Ω · cm) of the applied one. Thus, the effect of reducing electrical resistance by forming the first p-type contact layer 81 using a mixed gas of nitrogen and hydrogen as a carrier gas is great.

4-4. Activation Rate FIG. 7 is a graph showing the activation rate of the first p-type contact layer 81 with respect to the nitrogen mixing ratio (N ₂ / (H ₂ + N ₂ )). However, these are values when annealing is performed.

  As shown in FIG. 7, when only hydrogen was used without mixing nitrogen into the carrier gas, the activation rate of the first p-type contact layer 81 was about 0.12%. As the mixing ratio of nitrogen is increased, the activation rate once decreases. When the nitrogen mixing ratio is 22%, the activation rate of the first p-type contact layer 81 is about 0.07%.

  However, when the mixing ratio of nitrogen is further increased, the activation rate starts to increase again. The activation rate when the mixing ratio of nitrogen is around 35% is almost the same as the activation rate when the mixing ratio of nitrogen is 0%. When the mixing ratio of nitrogen is 40% or more, the activation rate takes a sufficiently high value.

  When the nitrogen mixing ratio was 44%, the activation rate of the first p-type contact layer 81 was about 0.14%. When the nitrogen mixing ratio was 66%, the activation rate of the first p-type contact layer 81 was about 0.21%. This value is a sufficiently high value.

4-5. FIG. 8 shows a case where hydrogen gas is used in the formation of the first p-type contact layer 81 (conventional example) and a mixed gas of nitrogen gas and hydrogen gas (mixing ratio of nitrogen 66). %) Is a graph for comparison with the case (Example). In addition, the result shown in FIG. 8 is a result at the time of performing an annealing process. The vertical axis | shaft of FIG. 8 is the relative luminous intensity of the light. The light intensity of the conventional example was used as a reference. Therefore, the relative luminous intensity of the conventional example is of course 100%.

  As shown in FIG. 8, the case where the first p-type contact layer 81 is formed using a mixed gas of nitrogen gas and hydrogen gas (mixing ratio of nitrogen 66%) as a carrier gas, as compared with the conventional example. The light output is about 10% higher. This is because, as described above, a light emitting device manufactured using a mixed gas as a carrier gas has preferable values in electrical resistivity, hole concentration, hole mobility, and activation rate. .

  In the experiment, the best results were obtained when the mixing ratio of nitrogen was 66%. Even when the mixing ratio of nitrogen was 44%, good results were obtained. Therefore, it is considered that the present invention can be applied even in a region where the mixing ratio of nitrogen is high, that is, a region where the mixing ratio of nitrogen is 80% or less, within a range where the mixing ratio of nitrogen is 44% or more and 66% or less. However, when the mixing ratio of nitrogen is large, surface roughness and pits may occur. Accordingly, the mixing ratio of nitrogen is preferably in the range of 50% to 75%. More preferably, it is considered to be in the range of 55% or more and 70% or less.

5. Modified example 5-1. In the present embodiment, after the p-electrode P1 and the n-electrode N1 are formed on the stacked body 90, the annealing process (heat treatment) is performed. However, as shown in FIGS. 4 to 6, the annealing process is not necessarily performed. For each of the hole concentration (see FIG. 4), hole mobility (see FIG. 5), electrical resistivity (see FIG. 6), and activation rate (see FIG. 7), some degree of electricity can be obtained without annealing. A light-emitting element with low resistivity can be manufactured. This reduces the manufacturing process by one. That is, the productivity of the light emitting element is improved.

5-2. In the present embodiment, both the first p-type contact layer 81 and the second p-type contact layer 82 are made of p-GaN. However, these layers 81 and 82 may be p-InGaN instead of p-GaN. Since p-GaN has improved ohmic properties, naturally, a similar effect should be obtained even with p-InGaN having a band gap smaller than that of p-GaN.

  Specifically, there is a case where the first p-type contact layer 81 is p-GaN and the second p-type contact layer 82 is p-InGaN. Further, the first p-type contact layer 81 may be p-InGaN, and the second p-type contact layer 82 may be p-InGaN. In this case, the In composition in the second p-type contact layer 82 is preferably larger than the In composition in the first p-type contact layer 81.

6). Summary As described above in detail, in the method of manufacturing the light emitting device 100 according to the present embodiment, the first p-type contact layer forming step uses a mixed gas of nitrogen and hydrogen as the carrier gas in the first p-type contact layer forming step. In the second p-type contact layer forming step, the second p-type contact layer 82 is formed using hydrogen as a carrier gas.

  This improves the mobility of holes in the p-type contact layer 80 and manufactures a group III nitride semiconductor light-emitting device capable of reducing the contact resistance between the p-type contact layer 80 and the p-electrode P1. The method is realized.

  In addition, this Embodiment is only a mere illustration and does not limit this invention at all. Therefore, the present invention can be variously improved and modified without departing from the scope of the invention. In this embodiment, the laminate 90 shown in FIG. 3 is used. However, the laminated structure of the laminated body is not necessarily limited to that shown in FIG. You may select arbitrarily, such as a laminated structure and the repetition frequency of each layer. Of course, the composition of each layer other than the p-type contact layer 80 may be different from that of the present embodiment. Moreover, it is not restricted to a metal organic chemical vapor deposition method (MOCVD method). Any other method may be used as long as the crystal is grown using a carrier gas.

DESCRIPTION OF SYMBOLS 10 ... Sapphire substrate 20 ... Low temperature buffer layer 30 ... n-type contact layer 40 ... n-type ESD layer 50 ... n-type SL layer 60 ... MQW layer 70 ... p-type cladding layer 80 ... p-type contact layer 81 ... 1st p-type Contact layer 82 ... second p-type contact layer 90 ... laminate 100 ... light emitting element P1 ... p electrode N1 ... n electrode

Claims (3)

A light emitting layer forming step of forming a light emitting layer;
A p-type cladding layer forming step of forming a p-type cladding layer on the light emitting layer;
A p-type contact layer forming step of forming a p-type contact layer on the p-type cladding layer;
A p-electrode forming step of forming a p-electrode on the p-type contact layer;
A method of manufacturing a semiconductor device having
The p-type contact layer forming step includes
A first step of forming a first p-type contact layer on the p-type cladding layer;
A second step of forming a second p-type contact layer on the first p-type contact layer;
Have
In the first step,
As a carrier gas, a mixed gas of nitrogen and hydrogen is used ,
The molar ratio of nitrogen to the total number of moles of the carrier gas is in the range of 50% to 75%,
The Mg concentration of the first p-type contact layer is in the range of 1 × 10 ¹⁹ / cm ^{3 to} 1 × 10 ²⁰ / cm ³ ,
The film thickness of the first p-type contact layer is in the range of 100 to 1000 mm,
In the second step,
Hydrogen is used as the carrier gas ,
The Mg concentration of the second p-type contact layer is in a range of 2 × 10 ²⁰ / cm ^{3 to} 9 × 10 ²¹ / cm ³ higher than the Mg concentration of the first p-type contact layer ,
The method of manufacturing a semiconductor element, wherein the thickness of the second p-type contact layer is in the range of 20 to 90 mm . A method of manufacturing a semiconductor device according to claim 1,
Performing an annealing process for reducing the resistance of the p-type contact layer;
The method of manufacturing a semiconductor device characterized <br/> that the hole concentration 5 × 10 ¹⁶ / cm ^{3 or} more 6 × 10 ¹⁶ / cm ³ within the following range. A method of manufacturing a semiconductor device according to claim 1 or 2,
Without performing an annealing process for reducing the resistance of the p-type contact layer ,
The hole concentration is in the range of 2 × 10 ¹⁶ / cm ^{3 to} 3 × 10 ¹⁶ / cm ³ and the electrical resistivity is in the range of 40 Ω · cm to 70 Ω · cm. A method for manufacturing a semiconductor device.

Images