Group iii nitride laminate

Abstract

To provide a group III nitride laminate in which the concentration of impurities in a buffer / channel layer of a high electron mobility transistor (HEMT) is controlled within a predetermined range.SOLUTION: The group III nitride laminate 100 includes a first layer (buffer / channel layer 30) made of gallium nitride and a second layer (barrier layer 40) formed on the first layer and made of a group III nitride having an electron affinity smaller than that of gallium nitride, the first layer includes a lower layer 31 and an upper layer 32 formed on the lower layer, a carbon concentration in the upper layer is lower than a carbon concentration in the lower layer, and a hydrogen concentration in the upper layer is lower than a hydrogen concentration in the lower layer, the concentration of C in the upper layer is less than or equal to 5 * 1016cm - 3 and the concentration of H in the upper layer is less than or equal to 1 * 1017cm - 3.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a group III nitride stack. [Background technology]

[0002] Epitaxial substrates in which a group III nitride layer is grown on a silicon carbide substrate are being developed, and such epitaxial substrates are used as materials for fabricating semiconductor devices such as high electron mobility transistors (HEMTs) (see, for example, Patent Document 1).

[0003] The buffer / channel layer of the HEMT is made of, for example, GaN. The concentration of impurities such as carbon in the buffer / channel layer affects the performance of the HEMT. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2016-48790 Summary of the Invention [Problem to be solved by the invention]

[0005] An object of the present invention is to provide a group III nitride stack in which the impurity concentration in the buffer / channel layer of a HEMT is controlled within a predetermined range. [Means for solving the problem]

[0006] According to one aspect of the present invention a first layer made of gallium nitride; a second layer formed on the first layer and made of a group III nitride having a lower electron affinity than gallium nitride; and the first layer has a lower layer and an upper layer formed on the lower layer, the carbon concentration in the upper layer is lower than the carbon concentration in the lower layer; the hydrogen concentration in the upper layer is lower than the hydrogen concentration in the lower layer; The carbon concentration in the upper layer is 5×10 16 cm -3 is as follows: The hydrogen concentration in the upper layer is 1×10 17 cm -3 A group III nitride stack comprising: is provided. [Effects of the Invention]

[0007] A group III nitride stack is provided in which the impurity concentration in the buffer / channel layer (first layer) of the HEMT is controlled within a predetermined range. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a schematic cross-sectional view showing a group III nitride stack according to one embodiment of the present invention. [Figure 2] FIG. 2 is a first example of a SIMS profile showing C concentration, H concentration, and Si concentration in the lower and upper layers of the buffer / channel layer according to an embodiment. [Figure 3] FIG. 3 is a second example of SIMS profiles showing C, H, and Si concentrations in the lower and upper layers of the buffer / channel layer according to an embodiment. [Figure 4] FIG. 4 is a timing chart that schematically illustrates the difference in growth conditions between the lower and upper layers of the buffer / channel layer according to the embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] <One embodiment> A III-nitride stack 100 (hereinafter also referred to as stack 100) according to one embodiment of the present invention will be described. FIG. 1 is a schematic cross-sectional view illustrating the stack 100. The stack 100 includes a substrate 10 and a III-nitride layer 60 (hereinafter also referred to as epitaxial layer 60) made of III nitride and formed on the substrate 10. The epitaxial layer 60 illustrated in FIG. 1 includes a nucleation layer 20, a buffer / channel layer 30, a barrier layer 40, and a cap layer 50. As will be described below, one feature of the stack 100 according to this embodiment is that the concentration of impurities such as carbon in the buffer / channel layer 30 is controlled within a predetermined range.

[0010] The stack 100 may be in the form of, for example, an epitaxial substrate 70 (hereinafter also referred to as epitaxial substrate 70) having a substrate 10 and an epitaxial layer 60. The stack 100 may also be in the form of, for example, a semiconductor device formed using the epitaxial substrate 70 as a material, more specifically, a high electron mobility transistor (HEMT) 200 formed by providing electrodes 210 (a source electrode 211, a gate electrode 212, and a drain electrode 213) on the epitaxial layer 60. FIG. 1 illustrates the stack 100 in the form of a HEMT 200. The stack 100 in the form of a HEMT 200 may be in the form of a wafer or may be in the form of chips obtained by dividing the wafer.

[0011] The substrate 10 is made of silicon carbide (SiC) and serves as a base substrate for heteroepitaxial growth of the epitaxial layer 60. As the SiC constituting the substrate 10, for example, semi-insulating SiC of polytype 4H or polytype 6H is used. Here, "semi-insulating" refers to, for example, a material having a resistivity of 10 5 The surface of the substrate 10 that serves as the base for growing the epitaxial layer 60 is, for example, the (0001) plane (c-plane silicon surface).

[0012] A nucleation layer 20 is formed on the substrate 10. The nucleation layer 20 is made of aluminum nitride (AlN) and functions as a nucleation layer that generates nuclei for crystal growth of the buffer / channel layer 30. The thickness of the nucleation layer 20 is preferably, for example, 1 nm or more and 200 nm or less.

[0013] A buffer / channel layer 30 is formed on the nucleation layer 20. The buffer / channel layer 30 is made of gallium nitride (GaN). A lower layer 31, which is the lower portion of the buffer / channel layer 30, serves as a base for growing an upper layer 32, which is the upper portion of the buffer / channel layer 30, and functions as a buffer layer for improving the crystallinity of the upper layer 32. The upper layer 32 also functions as a channel layer through which electrons travel when the HEMT 200 is in operation. The thickness of the lower layer 31 is preferably, for example, 100 nm or more so that it has suitable crystallinity as a base for growing the upper layer 32. The thickness of the upper layer 32 is preferably, for example, 50 nm or more so that the upper layer 32 functions suitably as a channel layer. Excessive thickness of the lower layer 31 and the upper layer 32 leads to unnecessary cost increases, so the thickness of the lower layer 31 is preferably, for example, 800 nm or less, and the thickness of the upper layer 32 is preferably, for example, 300 nm or less.

[0014] A barrier layer 40 is formed on the buffer / channel layer 30. The barrier layer 40 is made of a group III nitride having a lower electron affinity than the GaN constituting the buffer / channel layer 30, such as aluminum gallium nitride (AlGaN) containing aluminum (Al) and gallium (Ga) as group III elements. The barrier layer 40 generates a two-dimensional electron gas (2DEG) in the upper layer 32 of the buffer / channel layer 30 and functions as a barrier layer that spatially confines the 2DEG within the upper layer 32. The thickness of the barrier layer 40 is preferably, for example, 1 nm or more and 50 nm or less.

[0015] A cap layer 50 is formed on the barrier layer 40. The cap layer 50 is made of, for example, GaN, and is interposed between the barrier layer 40 and the electrode 210 in order to improve the device characteristics (such as the controllability of the threshold voltage) of the HEMT 200. The cap layer 50 is formed as necessary, and may be omitted.

[0016] On the epitaxial layer 60, a source electrode 211, a gate electrode 212, and a drain electrode 213 are formed as electrodes 210 of the HEMT 200. The gate electrode 212 is formed, for example, of a Ni / Au layer in which a nickel (Ni) layer and a gold (Au) layer are stacked. In this specification, when a stack is described as X / Y, it indicates that the layers are stacked in this order on the cap layer 50 (or on the barrier layer 40 if the cap layer 50 is omitted). Each of the source electrode 211 and the drain electrode 213 is formed, for example, of a Ti / Al layer in which a titanium (Ti) layer and an Al layer are stacked. Each of the source electrode 211 and the drain electrode 213 may also be formed, for example, by stacking a Ni / Au layer on a Ti / Al layer.

[0017] The impurity concentrations in the lower layer 31 and upper layer 32 of the buffer / channel layer 30 will be further described below. Figures 2 and 3 are secondary ion mass spectrometry (SIMS) profiles showing the carbon (C) concentration, hydrogen (H) concentration, and silicon (Si) concentration as impurity concentrations in the lower layer 31 and upper layer 32, respectively. As the SIMS profiles of the lower layer 31 and upper layer 32, a first example is shown in Figure 2, and a second example is shown in Figure 3. Note that Figures 2 and 3 are SIMS profiles measured on samples fabricated in a test in which the C concentration, H concentration, and Si concentration in the lower layer 31 and upper layer 32 were controlled to predetermined concentrations, and the depth values ​​in these SIMS profiles do not directly indicate the depth values ​​in the stack 100.

[0018] To illustrate the differences in C, H, and Si concentrations between the lower layer 31 and the upper layer 32 (formed directly above the lower layer 31), SIMS profiles are shown near the interface between the lower layer 31 and the upper layer 32. Portions 34 and 35 of predetermined thickness (100 nm in this example) are set below and above the boundary 33 between the lower layer 31 and the upper layer 32. The average values ​​of the C, H, and Si concentrations in a constant portion 37 of the lower predetermined thickness portion 34, where the C concentration changes little and is located outside (below) the slope portion 36 of the lower predetermined thickness portion 34, where the C concentration changes little, are defined as the C concentration, H concentration, and Si concentration in the lower layer 31. Similarly, the average values ​​of the C, H, and Si concentrations in a constant portion 38 of the upper predetermined thickness portion 35, where the C concentration changes little and is located outside (above) the slope portion 36 of the upper predetermined thickness portion 35, where the C concentration changes little, are defined as the C concentration, H concentration, and Si concentration in the upper layer 32. The boundary 33 is set, for example, at a position where the C concentration is exactly the middle (average) of the C concentration in the lower layer 31 and the C concentration in the upper layer 32.

[0019] The standard of variation in C concentration from the average value of the C concentration is, for example, ±25% or less, or, for example, ±20% or less, in each of the lower constant portion 37 and the upper constant portion 38. In the first and second examples, the lower constant portion 37 and the upper constant portion 38 are both regions over a thickness of 50 nm or more where the variation is ±20% or less.

[0020] In the first example (FIG. 2), the concentrations of the impurities are as follows: C concentration in the lower layer 31 is 9.1×10 16 cm -3 and the C concentration in the upper layer 32 is 1.1 × 10 16 cm -3 The H concentration in the lower layer 31 is 1.1×10 17 cm -3 and the H concentration in the upper layer 32 is 7.2 × 10 16 cm -3 The Si concentration in the lower layer 31 is 3.6×10 14 cm -3 The Si concentration in the upper layer 32 is 1.4×10 15 cm-3 is.

[0021] In the second example (FIG. 3), the concentrations of the impurities are as follows: C concentration in the lower layer 31 is 3.3×10 16 cm -3 and the C concentration in the upper layer 32 is 8.9 × 10 15 cm -3 The H concentration in the lower layer 31 is 7.7×10 16 cm -3 and the H concentration in the upper layer 32 is 6.5×10 16 cm -3 The Si concentration in the lower layer 31 is 2.7×10 14 cm -3 The Si concentration in the upper layer 32 is 1.2×10 15 cm -3 is.

[0022] As can be seen from the SIMS profiles of the first and second examples, the concentrations of the impurities in the lower layer 31 and the upper layer 32 have, for example, the following characteristics. This also makes it possible to obtain, for example, the following effects.

[0023] The carbon concentration in the upper layer 32 is lower than the carbon concentration in the lower layer 31. The lower carbon concentration in the upper layer 32 allows the crystallinity of the upper layer 32 to be higher than the crystallinity of the lower layer 31. The upper layer 32 is the portion where the 2DEG is formed and becomes the channel of the HEMT 200, and therefore the high crystallinity of the upper layer 32 can improve the performance of the HEMT 200. Specifically, the carbon concentration in the upper layer 32 is 5×10 16 cm -3 It is preferable that:

[0024] The H concentration in the upper layer 32 is lower than the H concentration in the lower layer 31. The low H concentration in the upper layer 32 can suppress the diffusion of H contained in the upper layer 32 into the barrier layer 40, and can suppress fluctuations in the H concentration in the upper layer 32 over time. Electrons in the 2DEG formed in the upper layer 32 are converted into hydrogen ions (H +Since the H concentration in the upper layer 32 is easily attracted to hydrogen ions, if the change in the H concentration over time in the upper layer 32 is large, the change in the 2DEG concentration over time will also be large. By suppressing the change in the H concentration over time in the upper layer 32, the change in the 2DEG concentration over time can be suppressed, and therefore the change in the performance of the HEMT 200 over time can be suppressed. Specifically, the H concentration in the upper layer 32 is 1×10 17 cm -3 It is preferable that:

[0025] The C concentration in the upper layer 32 is lower than the C concentration in the lower layer 31, in other words, the C concentration in the lower layer 31 is higher than the C concentration in the upper layer 32. The high C concentration in the lower layer 31 can make the conductivity of the lower layer 31 lower than the conductivity of the upper layer 32. This can suppress the leakage current of the HEMT 200. Specifically, the C concentration in the lower layer 31 is 2×10 lower than the C concentration in the upper layer 32. 16 cm -3 It is preferable that the value is higher than this.

[0026] However, in this embodiment, the C concentration in the lower layer 31 is suppressed. Specifically, the C concentration in the lower layer 31 is 5×10 17 cm -3 Preferably, it is 1×10 or less. 17 cm -3 It is more preferable that the C concentration in the lower layer 31 is less than or equal to the above. By suppressing the C concentration in the lower layer 31, it is possible to suppress a decrease in the crystallinity of the lower layer 31 due to the addition of C while adding C to the lower layer 31. This allows the lower layer 31 formed as a buffer layer to have high crystallinity even when it is thin.

[0027] The H concentration in the upper layer 32 is lower than the H concentration in the lower layer 31; in other words, the H concentration in the lower layer 31 is higher than the H concentration in the upper layer 32. As will be described later, by increasing the ratio of hydrogen gas in the carrier gas used during the growth period of the lower layer 31 (compared to the ratio of hydrogen gas in the carrier gas used during the growth period of the upper layer 32), it is possible to suppress a decrease in the crystallinity of the lower layer 31 when C is added to the lower layer 31. As a result, the H concentration in the lower layer 31 becomes higher than the H concentration in the upper layer 32.

[0028] However, if the H concentration in the lower layer 31 is excessively higher than that in the upper layer 32, there is a concern that the H concentration in the upper layer 32 will become high due to diffusion of H contained in the lower layer 31. For this reason, it is preferable that the ratio of the H concentration in the lower layer 31 to the H concentration in the upper layer 32 is small. The ratio of the H concentration in the lower layer 31 to the H concentration in the upper layer 32 is 1.5 times in the first example and 1.2 times in the second example. The ratio of the H concentration in the lower layer 31 to the H concentration in the upper layer 32 is, for example, preferably 5 times or less, more preferably 3 times or less, and even more preferably 2 times or less. The H concentration in the lower layer 31 is specifically 1×10 18 cm -3 Preferably, it is 5 x 10 or less. 17 cm -3 More preferably, it is:

[0029] Other features of the C concentration and H concentration observed from the SIMS profiles of the first and second examples include, for example, that the C concentration in the upper layer 32 is lower than the H concentration in the upper layer 32. Also, for example, that the C concentration in the lower layer 31 is lower than the H concentration in the lower layer 31.

[0030] The Si concentration in the upper layer 32 is higher than the Si concentration in the lower layer 31. The high Si concentration in the upper layer 32 compensates for electron traps in the channel, thereby suppressing current collapse. In addition, the high Si concentration in the upper layer 32 controls the electron distribution in the vertical direction (thickness direction) so that electrons gather on the upper layer 32 side, thereby improving the blocking characteristics. Specifically, the Si concentration in the upper layer 32 is 5×10 14 cm -3 It is preferable that the concentration is 1×10 or more. 15 cm -3 However, if the Si concentration in the upper layer 32 is excessively high, it will cause an increase in leakage current. Therefore, the Si concentration in the upper layer 32 should be 1×10 16 cm -3 It is preferable that:

[0031] The Si concentration in the upper layer 32 is higher than the Si concentration in the lower layer 31, in other words, the Si concentration in the lower layer 31 is lower than the Si concentration in the upper layer 32. The lower Si concentration in the lower layer 31 makes it possible to lower the conductivity of the lower layer 31 below the conductivity of the upper layer 32. This makes it possible to suppress the leakage current of the HEMT 200. Specifically, the Si concentration in the lower layer 31 is 5×10 14 cm -3 It is preferable that it is less than 10 ...

[0032] In the first and second examples, the following concentrations of each impurity were obtained: C concentration in the upper layer 32 was 5×10 16 cm -3 More specifically, 2×10 16 cm -3 The H concentration in the upper layer 32 is 1×10 17 cm -3 More specifically, 8 x 10 16 cm -3 The carbon concentration in the lower layer 31 is 5×10 17 cm -3 More specifically, 1×10 17cm -3 The H concentration in the lower layer 31 is 5×10 17 cm -3 More specifically, 2×10 17 cm -3 The Si concentration in the upper layer 32 is (1×10 16 cm -3 More specifically, 5 x 10 15 cm -3 Less than or equal to 5 x 10 14 cm -3 More specifically, 1×10 15 cm -3 The Si concentration in the lower layer 31 is 5×10 14 cm -3 More specifically, it is less than 4 × 10 14 cm -3 The details are as follows.

[0033] Next, a method for manufacturing the stack 100 according to the embodiment will be described. Here, a method for manufacturing the stack 100 in the form of a HEMT 200 will be exemplified. First, a method for manufacturing the epitaxial substrate 70 will be described. A SiC substrate is prepared as the substrate 10. The layers constituting the epitaxial layer 60, namely, the nucleation layer 20, the buffer / channel layer 30, the barrier layer 40, and (if necessary) the cap layer 50, are grown above the substrate 10 by metalorganic vapor phase epitaxy (MOVPE), thereby forming the epitaxial substrate 70.

[0034] Among the Group III source gases, trimethylaluminum (Al(CH3)3, TMA) gas is used as the Al source gas. Trimethylgallium (Ga(CH3)3, TMG) ​​gas is used as the Ga source gas. Ammonia (NH3) gas is used as the nitrogen (N) source gas, which is a Group V source gas. At least one of nitrogen gas (N2 gas) and hydrogen gas (H2 gas) is used as the carrier gas. The growth temperature can be selected from a range of 900°C to 1400°C, for example, and the V / III ratio, which is the flow rate ratio of the Group V source gas to the Group III source gas, can be selected from a range of 10 to 5000, for example. The ratio of the supply rates of each source gas is adjusted depending on the composition of each layer to be formed. The thickness of each layer to be formed can be controlled by the growth time, for example, by calculating the growth time corresponding to the designed thickness from the growth rate obtained in a preliminary experiment.

[0035] The following describes the process of forming the buffer / channel layer 30 in particular. Figure 4 is a timing chart that schematically illustrates the difference in growth conditions between the lower layer 31 and the upper layer 32 of the buffer / channel layer 30. The lower layer 31 and the upper layer 32 are made of GaN and are grown by supplying a Ga source gas and an N source gas together with a carrier gas. In order to control the concentrations of each impurity in the lower layer 31 and the upper layer 32 as described above, the growth conditions are different between the lower layer 31 and the upper layer 32.

[0036] 4, in this embodiment, the growth rate of the upper layer 32 during its growth period is set lower (slower) than the growth rate of the lower layer 31 during its growth period. Specifically, for example, the growth temperature and growth pressure are set higher during the growth period of the upper layer 32 than during the growth period of the lower layer 31, thereby setting the growth rate of the upper layer 32 lower than the growth rate of the lower layer 31. Furthermore, the ratio of H gas in the carrier gas used during the growth period of the upper layer 32 is set lower (less) than the ratio of H gas in the carrier gas used during the growth period of the lower layer 31.

[0037] For example, during the growth period of the lower layer 31, the growth temperature is set to a low value within a predetermined temperature range (e.g., not less than 900°C and less than 1040°C), and the growth pressure is set to a low value within a predetermined pressure range (e.g., not less than 6.7 kPa and less than 13.3 kPa). Furthermore, during the growth period of the lower layer 31, the carrier gas is a mixed gas of H2 gas and N2 gas, and the ratio of H2 gas in the mixed gas is increased (e.g., the partial pressure of N2 gas / the partial pressure of H2 gas is not less than 0.5 and less than 1.0).

[0038] For example, during the growth period of the upper layer 32, the growth temperature is set to be high within a predetermined temperature range (e.g., 1040°C or higher and 1200°C or lower), and the growth pressure is set to be high within a predetermined pressure range (e.g., 13.3 kPa or higher and 66.7 kPa or lower). Furthermore, during the growth period of the upper layer 32, the carrier gas is set to N gas only, or to a mixed gas of H gas and N gas with a low ratio of H gas in the mixed gas (e.g., N gas partial pressure / H gas partial pressure 1.0 or higher and 3.0 or lower).

[0039] By increasing the growth temperature and growth pressure during the growth of the upper layer 32 and making the growth rate of the upper layer 32 slower than that of the lower layer 31, decomposition of the Ga source material is relatively promoted during the growth of the upper layer 32 and relatively suppressed during the growth of the lower layer 31. As a result, the C concentration becomes relatively low in the upper layer 32 and relatively high in the lower layer 31.

[0040] During the growth period of the lower layer 31, the ratio of H2 gas in the carrier gas is increased, thereby improving the crystallinity of the GaN constituting the lower layer 31. This makes it possible to dope C into the lower layer 31 while suppressing a decrease in the crystallinity of the lower layer 31 due to the addition of C. On the other hand, during the growth period of the upper layer 32, the ratio of H2 gas in the carrier gas is reduced, thereby suppressing the incorporation of H into the upper layer 32. In this way, the H concentration becomes relatively low in the upper layer 32 and relatively high in the lower layer 31. Furthermore, when the H concentration in the upper layer 32 is increased to 1×10 17 cm -3 It can be as follows:

[0041] Si is supplied (mixed) into the growth layer at a constant rate, for example, from the reactor wall of the MOVPE apparatus. Therefore, the lower the growth rate, the higher the Si concentration in the growth layer. Therefore, by making the growth rate of the upper layer 32 slower than that of the lower layer 31, the Si concentration becomes relatively high in the upper layer 32 and relatively low in the lower layer 31.

[0042] In order to obtain a predetermined concentration of each impurity, the growth conditions, such as the growth temperature, growth pressure, V / III ratio, and carrier gas composition, during the growth period of the lower layer 31 and the growth period of the upper layer 32 may be appropriately adjusted using an MOVPE apparatus or the like.

[0043] After the epitaxial substrate 70 is manufactured, the electrodes 210 (source electrode 211, gate electrode 212, and drain electrode 213) are formed on the epitaxial layer 60 to manufacture the HEMT 200. When manufacturing the HEMT 200, other members such as a protective film may be formed as needed. The electrodes 210, protective film, etc. may be formed by known methods. In this manner, the stack 100 according to this embodiment is manufactured.

[0044] As described above, according to this embodiment, it is possible to obtain a stack 100 in which the impurity concentration (in the lower layer 31 and the upper layer 32) in the buffer / channel layer 30 of the HEMT 200 is controlled within a predetermined range. For example, the C concentration in the upper layer 32 may be set to be lower than the C concentration in the lower layer 31 (buffer layer), i.e., 5×10 16 cm -3 or less, the conductivity of the lower layer 31 can be made lower than that of the upper layer 32, thereby suppressing leakage current and improving the crystallinity of the upper layer 32. In addition, for example, the H concentration in the upper layer 32 can be made lower than the H concentration in the lower layer 31, for example, 1×10 17 cm -3 By setting the following, it is possible to suppress fluctuations in the performance of the HEMT 200 over time.

[0045] <Preferred embodiment of the present invention> Preferred embodiments of the present invention will be described below.

[0046] (Appendix 1) a first layer made of gallium nitride; a second layer formed on the first layer and made of a group III nitride having a lower electron affinity than gallium nitride; and the first layer has a lower layer and an upper layer formed on the lower layer, the carbon concentration in the upper layer is lower than the carbon concentration in the lower layer; the hydrogen concentration in the upper layer is lower than the hydrogen concentration in the lower layer; The carbon concentration in the upper layer is 5×10 16 cm -3 is as follows: The hydrogen concentration in the upper layer is 1×10 17 cm -3 The group III nitride stack is:

[0047] (Appendix 2) The carbon concentration in the lower layer is 2×10 16 cm -3 or more.

[0048] (Appendix 3) The carbon concentration in the lower layer is 5×10 17 cm -3 or less (more preferably 1 × 10 17 cm -3 3. The Group III nitride stack according to claim 1 or 2, wherein:

[0049] (Appendix 4) 4. The Group III nitride stack according to any one of claims 1 to 3, wherein the ratio of the hydrogen concentration in the lower layer to the hydrogen concentration in the upper layer is 5 times or less (more preferably 3 times or less, and even more preferably 2 times or less).

[0050] (Appendix 5) The hydrogen concentration in the lower layer is 1×10 18cm -3 or less (more preferably 5 × 10 17 cm -3 5. The Group III nitride stack according to any one of Appendices 1 to 4, wherein:

[0051] (Appendix 6) 6. The Group III nitride stack according to any one of claims 1 to 5, wherein the carbon concentration in the upper layer is lower than the hydrogen concentration in the upper layer.

[0052] (Appendix 7) 7. The Group III nitride stack according to any one of claims 1 to 6, wherein the carbon concentration in the lower layer is lower than the hydrogen concentration in the lower layer.

[0053] (Appendix 8) 8. The Group III nitride stack according to any one of claims 1 to 7, wherein the silicon concentration in the upper layer is higher than the silicon concentration in the lower layer.

[0054] (Appendix 9) The silicon concentration in the upper layer is (1×10 16 cm -3 Less than or equal to 5 x 10 14 cm -3 or more (more preferably 1×10 15 cm -3 9. The Group III nitride stack according to any one of Appendices 1 to 8, wherein:

[0055] (Appendix 10) The silicon concentration in the lower layer is 5×10 14 cm -3 10. The Group III nitride stack according to any one of appendices 1 to 9, wherein the thickness is less than 100 nm. [Explanation of symbols]

[0056] 10...substrate, 20...nucleation layer, 30...buffer / channel layer, 31...lower layer (of buffer / channel layer), 32...upper layer (of buffer / channel layer), 40...barrier layer, 50...cap layer, 60...group III nitride layer, 70...epitaxial substrate, 100...group III nitride stack, 200...HEMT, 210...electrode, 211...source electrode, 212...gate electrode, 213...drain electrode

Claims

1. a first layer composed of gallium nitride; a second layer formed on the first layer and made of a Group III nitride having an electron affinity smaller than that of gallium nitride; and the first layer has a lower layer and an upper layer formed on the lower layer, the carbon concentration in the upper layer is lower than the carbon concentration in the lower layer; the hydrogen concentration in the upper layer is lower than the hydrogen concentration in the lower layer; The carbon concentration in the upper layer is 5×10 １６ cm －３ is as follows: The hydrogen concentration in the upper layer is 1×10 １７ cm －３ A Group III nitride stack, which is:

2. The carbon concentration in the lower layer is 2×10 lower than the carbon concentration in the upper layer. １６ cm －３ The III-nitride stack of claim 1 , wherein the thickness of the stack is greater than or equal to 100 nm.

3. The carbon concentration in the lower layer is 5×10 １７ cm －３ 3. The Group III nitride stack of claim 1, wherein:

4. 4. The Group III nitride stack according to claim 1, wherein the ratio of the hydrogen concentration in said lower layer to the hydrogen concentration in said upper layer is 5 times or less.

5. The hydrogen concentration in the lower layer is 1×10 １８ cm －３ 5. The Group III nitride stack according to claim 1, wherein:

6. 6. The Group III nitride stack according to claim 1, wherein the carbon concentration in said upper layer is lower than the hydrogen concentration in said upper layer.

7. 7. The Group III nitride stack according to claim 1, wherein the silicon concentration in the upper layer is higher than the silicon concentration in the lower layer.

8. The silicon concentration in the upper layer is 5×10 １４ cm －３ The Group III nitride stack according to any one of claims 1 to 7, wherein:

Images