JP5927543B2 - Device isolation method of GaN layer - Google Patents

Description

  The present invention relates to an element separation method for a GaN layer, in which a GaN layer is separated into a plurality of elements by etching with a mask attached to a GaN layer disposed on a substrate.

  Conventionally, an element separation process is performed in which a GaN layer on a substrate is etched to the substrate and separated into a plurality of elements (for example, Patent Document 1).

  Japanese Patent Application Laid-Open No. H10-228561 describes performing an element separation process of a GaN layer using a dry etching method or the like.

Special table 2010-534943 gazette

  However, in recent years, it has been desired to develop a method for performing element isolation of the GaN layer more efficiently in the element isolation process of the GaN layer.

  Accordingly, an object of the present invention is to provide a device isolation method for a GaN layer, which performs device isolation of the GaN layer more efficiently.

  In order to achieve the above object, the present invention is configured as follows.

According to the first aspect of the present invention, there is provided an element isolation method for separating a GaN layer into a plurality of elements by attaching a mask to a GaN layer disposed on a substrate and etching the GaN layer, wherein BCl _{3 is used} as an etching gas. Provides a device isolation method for a GaN layer using a mixed gas in which CH ₂ F ₂ is mixed with a main gas.

According to the second aspect of the present invention, the CH ₂ F ₂ addition ratio in the etching gas is set to 7% or more and 35.1% or less to perform etching on the GaN layer, and the etching rate of the GaN layer is 300 nm / min or more. A device isolation method for a GaN layer according to the first aspect is provided.

According to the third aspect of the present invention, the element separation of the GaN layer according to the second aspect, wherein etching is performed on the GaN layer with the CH ₂ F ₂ addition ratio in the etching gas being 8% or more and 13.5% or less. Provide a method.

  According to the fourth aspect of the present invention, A (W) is the power for generating a bias applied to the substrate during etching, and B (W) is the power for generating plasma in the chamber that houses the substrate during etching. ), The GaN layer is etched while the value of A / (A + B) is 0.37 or more and 0.55 or less, and the etching rate of the GaN layer is 300 nm / min or more. A device isolation method for a GaN layer according to any one of three aspects is provided.

  According to the fifth aspect of the present invention, there is provided the element isolation method for a GaN layer according to the fourth aspect, wherein the GaN layer is etched while the value of A / (A + B) is 0.46 or more and 0.51 or less. To do.

  The element separation method of the GaN layer according to the present invention can perform the element separation of the GaN layer more efficiently.

1 is a configuration diagram of a dry etching apparatus according to an embodiment of the present invention. Schematic sectional view of a substrate handled by the dry etching apparatus according to the present embodiment Flow chart of element separation processing of GaN layer using dry etching apparatus according to this embodiment (A) Schematic sectional view of a substrate on which a mask pattern is formed, which is handled by the dry etching apparatus according to the present embodiment. (B) The GaN layer which is handled by the dry etching apparatus according to the present embodiment is etched to separate elements. (C) Schematic sectional view of the substrate from which the mask has been removed, handled by the dry etching apparatus according to the present embodiment Experiment No. Distribution chart of data obtained in 1-5 Experiment No. Distribution chart of data obtained in 6-11 Experiment No. Distribution chart of data obtained in 12 and 13 Experiment No. Distribution chart of data obtained in 14-16 Experiment No. Distribution map of data obtained in 17-21 Experiment No. Distribution chart of data obtained in 22-26 Experiment No. 27-32 data distribution map Experiment No. Shows the proportion of the added relationships bias ratio and the _CH 2 _{F 2} in the data obtained in 14-32

The inventors of the present invention have obtained the following knowledge as a result of intensive studies in order to solve the problems of the prior art.

That is, the present inventors have found that it is useful to improve both the etching rate of the GaN layer and the improvement of the selection ratio in order to perform the element separation method of the GaN layer by etching more efficiently. It was. That is, by improving the etching rate of the GaN layer, productivity is improved and by improving the selectivity, the etching shape of the GaN layer can be further stabilized. By improving the productivity and stabilizing the etching shape of the GaN layer, as a result, more efficient element separation of the GaN layer can be performed. Furthermore, the present inventors use a mixed gas of BCl ₃ and CH ₂ F ₂ as an etching gas for etching in order to achieve both improvement in the etching rate of the GaN layer and improvement in the selection ratio. I found out. Based on these findings, the inventors of the present invention have come up with the present invention.

  Embodiments according to the present invention will be described below in detail with reference to the drawings.

  FIG. 1 shows an inductively coupled plasma (ICP) type dry etching apparatus 1 as an example of an apparatus capable of performing etching for the element isolation method according to the present embodiment.

  The dry etching apparatus 1 includes a chamber (vacuum chamber) 3 that accommodates a substrate 2. The chamber 3 includes an inlet / outlet (not shown), and the substrate 2 is carried into / out of the chamber 3 through the inlet / outlet.

A gas introduction port 3 a and an exhaust port 3 b are provided on the side wall of the chamber 3. The chamber 3 is connected to an etching gas source 4 that holds an etching gas via a gas inlet 3a. Thereby, the etching gas can be supplied from the etching gas source 4 into the chamber 3. The etching gas source 4 in the present embodiment contains at least CH ₂ F ₂ (difluoromethane), BCl ₃ (boron trichloride), Cl ₂ (chlorine), and Ar (argon).

  The chamber 3 is connected to a decompression mechanism (not shown) through the exhaust port 3b. The air in the chamber 3 is exhausted to the outside through the exhaust port 3b and can be adjusted to a desired pressure by a decompression mechanism.

  The top of the chamber 3 is closed by a dielectric wall 5 made of quartz or the like. An antenna (plasma source) 6 serving as an upper electrode is disposed above the dielectric wall 5. The antenna 6 is electrically connected to the first high frequency power source 7A.

  On the other hand, a stage 8 for placing and holding the substrate 2 is arranged on the lower side in the chamber 3. The stage 8 is electrically connected to the second high frequency power supply unit 7B and functions as a lower electrode.

  Inside the stage 8, an electrostatic chucking electrode and a cooling device (both not shown) are incorporated. The electrostatic chucking electrode has a function of electrostatically chucking the substrate 2 on the stage 8. The cooling device has a function of cooling the substrate 2 on the stage 8. In the present embodiment, for example, He (helium) is used as the cooling gas for the substrate 2.

  The controller 9 comprehensively controls the operation of each component included in the dry etching apparatus 1 described above. Further, the flow rate at which a plurality of types of gases held in the etching gas source 4 are supplied into the chamber 3 and the respective mixing ratios can be adjusted by the control of the controller 9.

  Next, a substrate 2 handled by the dry etching apparatus 1 is shown in FIG. The substrate 2 in the present embodiment is made of sapphire. As shown in FIG. 2, a GaN layer 10 made of GaN (gallium nitride) is disposed on the substrate 2.

  The thickness of the substrate 2 may be, for example, 350-1300 μm, and the thickness of the GaN layer 10 may be, for example, 5-10 μm.

  An outline of the etching process using the dry etching apparatus 1 for the substrate 2 on which the GaN layer 10 is arranged will be described below.

  As shown in FIG. 1, first, the substrate 2 is carried into the chamber 3 and placed on the upper surface of the stage 8. Next, a DC voltage is applied to the electrostatic chucking electrode in the stage 8 to electrostatically attract and fix the substrate 2 to the stage 8. Subsequently, while supplying an etching gas from the etching gas source 4 and applying a high frequency voltage to the antenna 6 as the upper electrode from the first high frequency power source 7A, plasma (ICP) is generated in the chamber 3. This plasma turns the etching gas into plasma, generating ion components and radical components. On the other hand, a high frequency voltage (bias voltage) is applied from the second high frequency power supply 7B to the stage 8 as the lower electrode together with the application of the high frequency voltage to the antenna 6. As a result, a bias is applied to the stage 8 and the ion component of the plasma-ized etching gas in the chamber 3 is attracted to the substrate 2 on the stage 8 (ion incidence). Etching of the substrate 2 on the stage 8 is performed by the sucked plasma ion component and the plasma radical component. During the etching, the pressure in the chamber 3 is maintained at a predetermined pressure by using a pressure reducing mechanism, and the substrate 2 is cooled by using a cooling device.

  Although the outline of the etching process for the substrate 2 using the dry etching apparatus 1 has been described above, the dry etching apparatus 1 according to the present embodiment etches the GaN layer 10 on the substrate 2 as an etching process for the substrate 2. An element separation process for separating into a plurality of elements can be performed.

  An example of the flow of this element isolation method is shown in FIG. The element isolation method executed along the flowchart of FIG. 3 will be described in detail below using the cross-sectional view of the substrate 2 shown in FIG.

  First, a mask is applied and formed on the substrate 2 on the stage 8 of the dry etching apparatus 1 (step S1). Specifically, as shown in FIG. 4A, a mask 11 formed of an organic resist mask or the like is applied on the GaN layer 10 disposed on the substrate 2, and desired by photolithography or the like. The pattern is formed. The mask 11 partially protects the GaN layer 10. The thickness of the mask 11 may be, for example, 5-10 μm.

Next, etching is performed on the GaN layer 10 on the substrate 2 on which the mask 11 is formed (step S2). Specifically, the GaN layer 10 is etched by supplying a predetermined etching gas (for example, a mixed gas obtained by adding CH ₂ F ₂ to BCl ₃ ) from the etching gas source 4 into the chamber 3. When etching with an etching gas is performed, a portion of the GaN layer 10 that is not protected by the mask 11 is etched and removed from the substrate 2 as shown in FIG. Thereby, the GaN layer 10 is separated into a plurality of elements. Since not only the GaN layer 10 but also the mask 11 is etched at the same time, the height of the mask 11 after the etching is lower than that before the etching (FIG. 4A).

  Next, the mask 11 on the GaN layer 10 is removed (step S3). Specifically, only a mask 11 is removed from the GaN layer 10 on the substrate 2 by performing a predetermined mask removing process on the substrate 2 after etching. Thereby, as shown in FIG. 4C, the substrate 2 on which the GaN layer 10 separated into a plurality of elements is arranged is completed.

  As described above, the GaN layer 10 on the substrate 2 can be separated into a plurality of elements by performing element isolation processing by the dry etching apparatus 1 along the flowchart of FIG.

(Example)
Next, examples and comparative examples relating to the element isolation method of the GaN layer 10 using the dry etching apparatus 1 described above will be described.

  As shown in Table 1 below, Experiment No. Etching for element separation of the GaN layer 10 on the substrate 2 was performed under various etching conditions indicated by 1-5.

Experiment No. In 1-5, a mixed gas of Cl ₂ and Ar was used as an etching gas, and the GaN layer 10 on the substrate 2 was etched while changing the addition ratio of Cl ₂ in the etching gas.

  Experiment No. 1-5 corresponds to a comparative example. Experiment No. Specific etching conditions of 1-5 are as follows.

The amount of Cl ₂ added in the etching gas was determined according to Experiment No. 1 is 25 sccm, Experiment No. 1 2 is 40 sccm. 3 is 50 sccm, Experiment No. 4 is 75 sccm. 5 is 100 sccm.

  The amount of Ar added in the etching gas was determined according to Experiment No. 1 is 175 sccm, Experiment No. 2 is 160 sccm. 3 is 150 sccm, Experiment No. 4 is 125 sccm. 5 is 100 sccm.

The addition ratio of Cl ₂ in the etching gas was determined according to Experiment No. 1 is 12.5%. 2 is 20.0%, Experiment No. 3 is 25.0%, Experiment No. 4 is 37.5%, Experiment No. 5 is 50.0%.

  The pressure in the chamber 3 was measured according to Experiment No. 1-5 is 1 Pa.

  The power density (ICP power / volume of the chamber 3) of the ICP power, which is the power for generating plasma (ICP) in the chamber 3, was determined as Experiment No. 1-5 is 0.024107 (W / cm3). The ICP power here is the power (W) applied to the antenna 6 from the first high frequency power source 7A in the dry etching apparatus 1. 1-5 is 1350W. The volume of the chamber 3 is the same as that of Experiment No. 1-5 is 56000 cm3.

  The power density of the bias power (bias power / electrode area of the electrode in the stage 8), which is the power for generating a bias in the chamber 3, was determined as Experiment No. 1-5 is 1.403162 W / cm2. The bias power here is power (W) applied to the electrode in the stage 8 from the second high frequency power source 7B in the dry etching apparatus 1. 1-5 is 1350W. The electrode area of the electrode in the stage 8 is the same as that in Experiment No. 1-5 is about 962 cm 2 (the electrode is a perfect circle having a diameter of 350 mm).

  The bias ratio, which is the ratio of the bias power to the sum of the ICP power (W) and the bias power (W), 1-5 is 0.5.

  The temperature in the chamber 3 was measured according to Experiment No. 1-5 at 15 ° C.

  Experiment No. For 1-5, the etching rate (nm / min) of the GaN layer 10, the etching rate (nm / min) of the mask 11, and the selectivity thereof were measured. Each measurement result is as shown in Table 1.

  The selection ratio is the ratio of the etching rate of the GaN layer 10 to the etching rate of the mask 11. By increasing the selection ratio or increasing the etching rate of the GaN layer 10, the etching shape of the GaN layer 10 can be further stabilized. The stabilization of the etching shape here means that the GaN layer 10 is etched in a more vertical direction, or the shape of the GaN layer 10 after etching becomes more uniform.

In FIG. The distribution chart of the data obtained in 1-5 is shown. In FIG. 5, the abscissa indicates the addition ratio of Cl ₂ , and the ordinate indicates two types of etching rates (the etching rate of the GaN layer 10 and the etching rate of the mask 11) and the selection ratio (the etching rate of the GaN layer 10 / the mask 11). Etching rate).

As shown in FIG. 5, when a mixed gas of Cl ₂ and Ar is used as an etching gas, it can be seen that the selection ratio becomes the highest by setting the addition ratio of Cl ₂ to around 20%. . On the other hand, the selection ratio does not reach 2.0 or more regardless of the addition ratio of Cl ₂ . It can also be seen that the etching rate of the GaN layer 10 and the mask 11 increases as the Cl ₂ addition ratio increases.

  Next, as shown in Table 2 below, using the dry etching apparatus 1, the experiment No. Etching for element separation of the GaN layer 10 on the substrate 2 was performed under various etching conditions indicated by 6-11.

Experiment No. 6-11, the mixed gas of BCl ₃ and Cl ₂ was used as an etching gas, and the GaN layer 10 on the substrate 2 was etched while changing the addition ratio of BCl ₃ in the mixed gas.

  Experiment No. 6-11 corresponds to a comparative example. Experiment No. Specific etching conditions for 6-11 are as follows.

The amount of BCl ₃ added in the etching gas was determined according to Experiment No. 6 is 0 sccm. 7 is 24 sccm. 8 is 60 sccm. 9 is 100 sccm. 10 is 140 sccm, Experiment No. 11 is 200 sccm.

The amount of Cl ₂ added in the etching gas was determined according to Experiment No. 6 is 200 sccm, Experiment No. 7 is 176 sccm. 8 is 140 sccm. 9 is 100 sccm. 10 is 60 sccm. 11 is 0 sccm.

The addition ratio of BCl ₃ in the etching gas was determined according to Experiment No. 6 is 0.0%. 7 is 12.0%, Experiment No. 8 is 30.0%, Experiment No. 9 is 50.0%, Experiment No. 10 is 70.0%, Experiment No. 11 is 100.0%.

  Experiment No. 6-11, the pressure in the chamber 3, the power density of the ICP power, the power density of the bias power, the bias ratio and the temperature in the chamber 3 (and the electrodes of the ICP power, the bias power, the volume of the chamber 3 and the electrode in the stage 8) Area) is the above-mentioned experiment No. Same as 1-5.

  Experiment No. For 6-11, the etching rate (nm / min) of the GaN layer 10, the etching rate (nm / min) of the mask 11, and the selectivity thereof were measured. Each measurement result is as shown in Table 2.

In FIG. The distribution chart of the data obtained in 6-11 is shown. In FIG. 6, the horizontal axis represents the addition ratio of BCl ₃ , and the vertical axis represents two types of etching rates and selection ratios (etching rate of the GaN layer 10 / etching rate of the mask 11).

As shown in FIG. 6, it can be seen that when the mixed gas of BCl ₃ and Cl ₂ is used as the etching gas, the higher the BCl ₃ addition ratio, the higher the selection ratio. On the other hand, the selection ratio does not reach 2.0 or more regardless of the addition ratio of BCl ₃ . It can also be seen that the smaller the BCl ₃ addition ratio, the higher the etching rate of the GaN layer 10 and the mask 11.

  Next, as shown in Table 3 below, Experiment No. Etching for element isolation of the GaN layer 10 on the substrate 2 was performed under various etching conditions indicated by 12 and 13.

Experiment No. In 12 and 13, the substrate using mixed gas of Cl ₂ to _CH 2 _{F 2} (Experiment No.12) or mixed gas of BCl ₃ in _CH 2 _{F 2} (Experiment No.13) as the etching gas 2 The upper GaN layer 10 was etched.

  Experiment No. 12 corresponds to a comparative example. 13 corresponds to the example. Experiment No. Specific etching conditions of 12 and 13 are as follows.

The amount of BCl ₃ added in the etching gas was determined according to Experiment No. 12 is 0 sccm. 13 is 140 sccm.

The amount of Cl ₂ added in the etching gas was determined according to Experiment No. 12 is 140 sccm, Experiment No. 13 is 0 sccm.

The amount of CH ₂ F ₂ added in the etching gas was determined according to Experiment No. Both 12 and 13 are 45 sccm.

  The power density of the ICP power is determined as Experiment No. 12 and 13 are 0.028571 W / cm3. Experiment No. In 12 and 13, the ICP power is 1600W.

  The power density of the bias power was determined according to Experiment No. 12 and 13 are 0.675596 W / cm 2. Experiment No. In 12 and 13, the bias power is 650 W.

  The bias ratio was determined as Experiment No. 12 and 13 are 0.288889.

  Experiment No. 12 and 13, the pressure in the chamber 3 and the temperature in the chamber 3 (and the volume of the chamber 3 and the electrode area of the electrode in the stage 8) are the same as those in the experiment No. 1 described above. It is the same as 1-11.

  Experiment No. 12 and 13, the etching rate of the GaN layer 10 (nm / min), the etching rate of the mask 11 (nm / min), and the selectivity thereof were measured. Each measurement result is as shown in Table 3.

In FIG. 12 shows a distribution map of data obtained in 12 and 13. FIG. In FIG. 7, the horizontal axis indicates the type of gas added to CH ₂ F ₂ (Cl ₂ or BCl ₃ ), and the vertical axis indicates two types of etching rates and selection ratios (etching rate of GaN layer 10 / etching rate of mask 11. ).

As shown in FIG. 7, when a mixed gas in which Cl ₂ is mixed with CH ₂ F ₂ is used as an etching gas, a selectivity ratio of 2.0 cannot be obtained, whereas BCl ₃ is added to CH ₂ F _2. In the case where a mixed gas in which is used as an etching gas, a selection ratio of 2.0 or more could be obtained. Moreover, the etching rate of the GaN layer 10 is approximately 300 nm / min regardless of whether Cl ₂ or BCl ₃ is mixed.

When CH ₂ F ₂ is added to the etching gas, a carbon polymer film is deposited on the mask 11 that is a resist, thereby suppressing the etching of the mask 11. Thereby, it is considered that the etching rate of the mask 11 is lowered and the selectivity is improved. On the other hand, from the results of this experiment (particularly, Experiments Nos. 12 and 13), even when CH ₂ F ₂ was added to Cl ₂ , the selectivity could not be improved satisfactorily, whereas BCl _When CH ₂ F ₂ was added to ₃ , the selectivity could be improved satisfactorily.

This is considered to be related to the fact that the amount of Cl radicals generated is larger when Cl ₂ is added to the etching gas than when BCl ₃ is added.

Etching of the GaN layer 10 is facilitated by the incidence of Cl radicals and ions. When Cl ₂ is added, the generation of Cl radicals increases, so that the GaN layer 10 can be satisfactorily etched. On the other hand, the mask 11 is also etched at a high rate. Since the mask 11 as a base to which the carbon polymer film adheres moves back at a high speed, the carbon polymer film cannot be effectively deposited even if CH ₂ F ₂ is added. As a result, a sufficient etching suppression effect by the carbon polymer film cannot be obtained, and the etching rate of the mask 11 cannot be lowered sufficiently. It is presumed that the selection ratio cannot be improved satisfactorily by such a principle.

On the other hand, when BCl ₃ is added to the etching gas, the generation of Cl radicals is reduced, so that the mask 11 is not etched at a very high rate, and the etching rate of the mask 11 decreases. As a result, a carbon polymer film can be effectively deposited on the mask 11, so that the etching rate of the mask 11 is further reduced. On the other hand, although the generation of Cl radicals is less than when Cl ₂ is added, the GaN layer 10 can be satisfactorily etched due to other factors by adding BCl _3. Does not drop that much. That is, when BCl ₃ is added, the rate at which the etching rate of the mask 11 decreases is higher than the rate at which the etching rate of the GaN layer 10 decreases, compared with the case where Cl ₂ is added. It is considered that can be improved satisfactorily.

As described above, Experiment No. As a result of considering the results of 1-13, the present inventors etched the GaN layer 10 using a mixed gas in which CH ₂ F ₂ was added to BCl ₃ in the element isolation method of the GaN layer 10, thereby obtaining a GaN layer 10. It has been found that a high etching rate and a high selectivity can be achieved in the layer 10. If a high etching rate can be realized, productivity can be improved, and if a high selectivity can be realized, the etching shape of the GaN layer 10 can be further stabilized. As a result, device isolation of the GaN layer 10 can be performed more efficiently.

  If both a high etching rate and a high selection ratio can be achieved, a desired etching shape can be obtained in the etched GaN layer 10 while reducing the thickness of the mask 11 used for etching. Therefore, the cost for etching can be reduced.

Here, regarding the element isolation method of the GaN layer 10 using the mixed gas obtained by adding CH ₂ F ₂ to BCl ₃ as an etching gas, the present inventors added BCl ₃ and CH ₂ F ₂ in the etching gas. Further experiments were performed while changing the above.

  That is, as shown in Table 4 below, using the dry etching apparatus 1, the experiment No. Etching for element isolation of the GaN layer 10 was performed under various etching conditions indicated by 14-32.

  Experiment No. Specific etching conditions of 14-31 are as follows.

The amount of BCl ₃ added in the etching gas was determined according to Experiment No. 14-16, 25 and 30 are 140 sccm, Experiment No. 17-21, 23 and 28 are 170 sccm, Experiment No. 22 and 27 are 185 sccm. 24, 29 is 160 sccm, Experiment No. 26 and 31 are 120 sccm.

The amount of CH ₂ F ₂ added in the etching gas was determined according to Experiment No. 14-16, 25 and 30 are 45 sccm, Experiment No. 17-21, 23 and 28 are 15 sccm, Experiment No. Nos. 22 and 27 were 0 sccm. 24, 29 is 25 sccm, Experiment No. 26 and 31 are 65 sccm.

The addition ratio of CH ₂ F ₂ in the etching gas was determined according to Experiment No. 14-16, 25 and 30 are 24.3%, Experiment Nos. 17-21, 23 and 28 are 8.1%, and Experiment No. 22 and 27 are 0%, Experiment No. 24, 29, 13.5%, Experiment No. 26 and 31 are 35.1%.

  The power density of the ICP power was measured according to Experiment No. 14-16, 18, 22-26, 0.028571 W / cm3, Experiment No. 17 0.032143 W / cm3, Experiment No. 19, 27-31, 0.025893 W / cm3, Experiment No. 20 0.024107 W / cm3, Experiment No. 21 is 0.019643 W / cm3. Experiment No. 14-16, 18, 22-26, ICP power is 1600 W, Experiment No. 17, ICP power is 1800 W, Experiment No. 19, 27-31, ICP power is 1450 W, Experiment No. No. 20, ICP power is 1350 W, Experiment No. At 21, the ICP power is 1100W.

  The power density of the bias power was measured according to Experiment No. 14 0.476772 W / cm 2, Experiment No. 15, 0.675596 W / cm 2, Experiment No. 16, 18, 22-26, 1.195286 W / cm2, Experiment No. 17, 0.98741 W / cm 2, Experiment No. 19, 1.351193 W / cm 2, Experiment No. 20, 1.455131 W / cm 2, Experiment No. 21 is 1.714976 W / cm2. Experiment No. 14, the bias power is 450 W, the experiment No. 15, the bias power is 650 W, the experiment No. 16, 18, 22-26, the bias power is 1150 W, the experiment No. 17, the bias power is 950 W, the experiment No. 19 and 27-31, the bias power is 1300 W, the experiment No. 20, the bias power is 1400 W, the experiment No. In 21, the bias power is 1650W.

  The bias ratio was determined according to Experiment No. 14 0.220, Experiment No. 15 at 0.289, experiment no. 16, 18, 22-26, 0.418, Experiment No. 17 0.345, Experiment No. 19, 27-31, 0.473, Experiment No. 20 at 0.509, experiment no. 21 is 0.600.

  Experiment No. 14-31, the pressure in the chamber 3 and the temperature in the chamber 3 (and the volume of the chamber 3 and the electrode area of the electrode in the stage 8) are the same as those in Experiment No. 1 described above. It is the same as 1-13.

  Experiment No. For 14-31, the etching rate (nm / min) of the GaN layer 10, the etching rate (nm / min) of the mask 11 and the selectivity thereof were measured. Each measurement result is as described in Table 4.

  Experiment No. 14-16 is a distribution diagram of the data obtained in FIG. The distribution map of the data obtained in 17-21 is shown in FIG. A distribution map of the data obtained in 22-26 is shown in FIG. FIG. 11 shows a distribution map of data obtained in 27-31.

In FIG. 8, the horizontal axis represents the power density of the bias power, and the vertical axis represents the two types of etching rates and the selection ratio. In FIG. 9, the horizontal axis represents the bias ratio, and the vertical axis represents the two types of etching rates and selection ratios. In FIG. 10, the horizontal axis represents the addition ratio of CH ₂ F ₂ , and the vertical axis represents two types of etching rates and selection ratios. In FIG. 11, the horizontal axis represents the addition ratio of CH ₂ F ₂ , and the vertical axis represents two types of etching rates and selection ratios. 8-11 is the etching rate of the GaN layer 10 / the etching rate of the mask 11.

  From the results shown in FIG. 9, when the bias ratio is in the range of approximately 0.37 to 0.55 (corresponding to the region A1 in FIG. 9), the etching rate of the GaN layer 10 is 300 or more and the selection ratio is 2 or more. It turns out that it exists in a preferable range. Furthermore, if the bias ratio is approximately in the range of 0.46-0.51 (corresponding to the region A2 in FIG. 9), the etching rate of the GaN layer 10 is 350 nm / min or more and the selection ratio is 3 or more, which is more preferable. You can see that it is in range.

From the results shown in FIG. 10, when the bias ratio is 0.418, the CH ₂ F ₂ addition ratio is approximately within a range of 4.0% to 35.1% (corresponding to the area A3 in FIG. 10). If so, it can be seen that the etching rate of the GaN layer 10 is 300 or more and the selection ratio is 2 or more, which is in a preferable range. Similarly, if the CH ₂ F ₂ addition ratio is approximately in the range of 11.0% to 35.1% (corresponding to the region A4 in FIG. 10), the etching rate of the GaN layer 10 is 350 or more and the selection ratio is It can be seen that is in a more preferable range of 3 or more.

From the results shown in FIG. 11, when the bias ratio is 0.473, the CH ₂ F ₂ addition ratio is approximately in the range of 2.0% to 35.1% (corresponding to the region A5 in FIG. 11). If so, it can be seen that the etching rate of the GaN layer 10 is 300 or more and the selection ratio is 2 or more, which is in a preferable range. Similarly, if the addition ratio of CH ₂ F ₂ is in the range of approximately 8.0% to 35.1% (corresponding to the region A6 in FIG. 11), the etching rate of the GaN layer 10 is 350 or more and the selection ratio is It can be seen that is in a more preferable range of 3 or more.

As described above, as a result of considering the results shown in FIGS. 9-11, the present inventors, in the element isolation method of the GaN layer 10 using the mixed gas obtained by adding CH ₂ F ₂ to BCl ₃ as an etching gas, It has been found that the etching rate and the selectivity of the GaN layer 10 can be further improved by setting the addition ratio of CH ₂ F ₂ and the respective two values within a predetermined range.

From this experimental result, by using a material obtained by adding BCl ₃ in CH ₂ F ₂ as the etching gas, it was possible to obtain a high selectivity while maintaining a high etch rate of GaN layer 10, CH ₂ Increasing the addition ratio of F ₂ increases the selectivity, while increasing the deposition (sediment) inside the chamber 3. If the deposition increases, the deposition may peel off from the inner wall of the chamber 3 or the like, and particles may be generated, or the exhaust pipe of the dry etching apparatus 1 may be clogged. In order to prevent this, it is necessary to perform dry cleaning regularly or perform wet maintenance by opening the chamber 3 to the atmosphere. However, the operation rate of the dry etching apparatus 1 is lowered by either dry cleaning or wet maintenance.

From this, the present inventors have found a further problem of improving both maintainability and achieving both a high etching rate and a high selection ratio of the GaN layer 10, and have intensively studied to solve the problem. As a result, in order to solve the above-described problem, the bias ratio is increased within a range where the etching rate of the GaN layer 10 does not decrease, and the addition ratio of CH ₂ F _{2 is set} so that the bias ratio exceeds the required selection ratio. We found that it is effective to set more than the minimum ratio. In this case, the addition ratio of CH ₂ F ₂ may be set to a ratio with a slight margin from the minimum ratio.

Therefore, Experiment No. The relationship between the bias ratio in 14-31 and the addition ratio of CH ₂ F ₂ is shown in FIG.

In FIG. 12, the horizontal axis represents the bias ratio, and the vertical axis represents the CH ₂ F ₂ addition ratio.

  As shown in FIG. 12, the bias ratio is preferably in the range of approximately 0.37 to 0.55 (corresponding to the region A7 in FIG. 12), and is approximately in the range of 0.46 to 0.51 (region A8 in FIG. 12). Is more preferable.

As shown in FIG. 12, when the addition ratio of CH ₂ F ₂ is large, the etching rate and selectivity of the GaN layer 10 are easily improved. However, considering that CH ₂ F ₂ remains as deposition, CH ₂ The addition ratio of ₂ F ₂ is preferably not more than a predetermined upper limit value. On the other hand, considering that it is preferable to achieve both a high etching rate and a high selection ratio of the GaN layer 10, the addition ratio of CH ₂ F ₂ is preferably equal to or higher than a predetermined lower limit value.

Based on the above, as a result of various experiments, the present inventors have found that the practically preferable range of the addition ratio of CH ₂ F ₂ is approximately 7% -35.1% (corresponding to the region A9 in FIG. 12). It was. Further, a more preferable range of the CH ₂ F ₂ addition ratio can be approximately 8% to 13.5% (corresponding to the region A 10 in FIG. 12) from the result shown in FIG.

As described above, according to the element isolation method of the GaN layer 10 according to the example of the present embodiment, the addition ratio of CH ₂ F ₂ in the etching gas is set to 7% or more and 35.1% or less, and the GaN layer 10 The etching rate of the GaN layer 10 is set to 300 nm / min or more. Thereby, the etching rate and selectivity of the GaN layer 10 can be further improved.

In addition, according to the element isolation method of the GaN layer 10 according to the example of the present embodiment, the addition ratio of CH ₂ F ₂ in the etching gas is set to 8% or more and 13.5% or less, and the GaN layer 10 is etched. Do. Thereby, the etching rate and selectivity of the GaN layer 10 can be further improved.

  Further, according to the element isolation method of the GaN layer 10 according to the example of the present embodiment, the power for generating a bias applied to the substrate 2 at the time of etching is A (W), and the substrate 2 is accommodated at the time of etching. When the power for generating plasma in the chamber 3 is B (W), the GaN layer 10 is etched while the value of A / (A + B) is 0.37 or more and 0.55 or less, and the GaN layer 10 is etched. The etching rate of 10 is 300 nm / min or more. Thereby, the etching rate and selectivity of the GaN layer 10 can be further improved.

  Further, according to the element isolation method of the GaN layer 10 according to the example of the present embodiment, the GaN layer 10 is etched while the value of A / (A + B) is set to 0.46 or more and 0.51 or less. Thereby, the etching rate and selectivity of the GaN layer 10 can be further improved.

In the example of the present embodiment, the case where a mixed gas in which only BCl ₃ and CH ₂ F ₂ are mixed has been described. However, the mixed gas may contain other gases, and BCl ₃ If it is a mixed gas in which CH ₂ F ₂ is mixed with a main gas (for example, the ratio containing BCl ₃ is 80% or more), the same effects as described above can be obtained.

  In addition, this invention is not limited to the said embodiment, It can implement in another various aspect. For example, although the case where the substrate 2 is formed of sapphire has been described in the present embodiment, the present invention is not limited to such a case. For example, the substrate 2 is made of a material other than sapphire, such as Si (silicon), SiC (silicon carbide), or GaN. The substrate 2 may be formed.

  In the present embodiment, the case where the substrate 2 and the GaN layer 10 are distinguished from each other has been described. However, the present invention is not limited to this case. For example, the GaN layer 10 may be handled as a part of the substrate 2. .

  In the present embodiment, the case where the GaN layer 10 is directly disposed on the substrate 2 has been described. However, the present invention is not limited to such a case. For example, another layer is provided between the substrate 2 and the GaN layer 10. (Such as a buffer layer) may be disposed.

  It is to be noted that, by appropriately combining any of the above-described various embodiments, the effects possessed by them can be produced.

  The present invention can be applied to a device isolation method for a GaN layer.

DESCRIPTION OF SYMBOLS 1 Dry etching apparatus 2 Substrate 3 Chamber 3a Gas introduction port 3b Exhaust port 4 Etching gas source 5 Dielectric wall 6 Antenna 7A First high frequency power supply 7B Second high frequency power supply 8 Stage 9 Controller 10 GaN layer 11 Mask

Claims (5)

A device isolation method for separating a GaN layer into a plurality of devices by attaching a mask to a GaN layer disposed on a substrate and etching the GaN layer,
A device isolation method for a GaN layer using a mixed gas obtained by adding CH ₂ F ₂ to a gas mainly containing BCl ₃ as an etching gas. _2. The GaN layer according to claim 1, wherein the GaN layer is etched with an addition ratio of CH ₂ F ₂ in the etching gas of 7% or more and 35.1% or less, and the etching rate of the GaN layer is 300 nm / min or more. Element isolation method. The element isolation method for a GaN layer according to claim 2, wherein the GaN layer is etched with an addition ratio of CH ₂ F ₂ in the etching gas of 8% to 13.5%.   A / (A + B) where A (W) is a power for generating a bias applied to the substrate during etching, and B (W) is a power for generating plasma in the chamber accommodating the substrate during etching. The GaN layer according to any one of claims 1 to 3, wherein the GaN layer is etched while the value of) is 0.37 or more and 0.55 or less, and the etching rate of the GaN layer is 300 nm / min or more. Element isolation method.   The element isolation method for a GaN layer according to claim 4, wherein the GaN layer is etched while the value of A / (A + B) is 0.46 or more and 0.51 or less.

Images