Calibration method

Abstract

To correctly perform calibration when measuring a carbon-containing semiconductor by an X-ray photoelectron spectroscopy.SOLUTION: A calibration method includes: forming a first layer of metal on a carbon-containing semiconductor; identifying first bond energy derived from a first energy level of the metal when measuring the semiconductor with the first layer formed thereon by an X-ray photoelectron spectroscopy; forming a second layer of an insulator on the semiconductor with the first layer formed thereon; identifying second bond energy derived from the first energy level when measuring the second layer by the X-ray photoelectron spectroscopy; and calculating fourth bond energy obtained by calibrating third bond energy derived from a second energy level of an element when measuring the second layer by the X-ray photoelectron spectroscopy by using only a difference between the first bond energy and the second bond energy.SELECTED DRAWING: Figure 10

Description

[Technical Field]

[0001] The present invention relates to a calibration method. [Background technology]

[0002] Carbon-based semiconductor materials such as diamond have characteristics such as a wide band gap, high thermal conductivity, high voltage resistance, chemical stability, and radiation resistance, and electronic devices that utilize these characteristics are being investigated. In particular, MIS (Metal Insulator Semiconductor) capacitors and MIS transistors (MISFETs) using diamond have excellent high-power, high-frequency, and high-temperature characteristics.

[0003] For example, Non-Patent Document 1 reports that a MISFET using diamond (H-diamond) whose surface is terminated with hydrogen has a high cutoff frequency of 70 GHz and a very high breakdown voltage of 1700 V. Non-Patent Document 2 also reports that a MISFET using boron-doped diamond (B-diamond) has a current on / off ratio of 10 even at a high temperature of 500°C. 9 It has been reported that the SiO2 has a higher SiO2 content than the SiO2 and has excellent high-temperature properties.

[0004] Furthermore, diamond can be doped with impurities to become a p-type or n-type semiconductor, allowing it to be used in electronic circuits. To fabricate MIS capacitors and MISFETs using diamond, it is necessary to understand the band structure at the interface between the insulating layer and diamond. Understanding this band structure allows us to optimize the capacitance of MIS capacitors and determine the threshold voltage, charge injection and extraction, leakage current density, and device stability of MISFETs. Thus, elucidating the band structure at the interface between diamond and an insulating layer is useful for fabricating highly reliable MIS devices using diamond.

[0005] Techniques for investigating band structures include XPS (X-ray Photoelectron Spectroscopy), capacitance-voltage measurement, current-voltage measurement, and Kelvin probe force microscopy. Among these, XPS has an advantage over other techniques in that it can measure band bending nondestructively. For example, XPS can be used to measure the band offset when various insulating layers are formed on diamonds whose surfaces are terminated with hydrogen (H-diamond) or boron-doped diamonds (B-diamond).

[0006] XPS investigates the elemental composition and chemical bonding state of a sample by measuring the kinetic energy spectrum of electrons emitted from the sample irradiated with X-rays. However, if the sample surface is charged by X-ray irradiation, the measured kinetic energy of the electrons will deviate from its original position. For example, if the sample surface becomes positively charged, the kinetic energy will be smaller than when there is no charge due to the Coulomb attraction between the emitted electrons and the sample surface. Therefore, XPS requires calibration to reduce this deviation in kinetic energy. Calibration is performed by identifying the C 1s binding energy resulting from carbon contamination on the sample surface using XPS and then shifting the energy spectrum by the difference between that binding energy and the known C 1s binding energy (approximately 284.8 eV) (Non-Patent Documents 3 and 4).

[0007] However, in carbon-containing semiconductor materials such as diamond, the C 1s bond energy derived from the carbon contained in the material itself is strong, making it impossible to identify the C 1s bond energy derived from carbon contamination on the material surface. Furthermore, because the C 1s bond energy derived from the carbon contained in the material itself varies depending on the carbon bonding state within the material, shifting the energy spectrum to match it to the known C 1s bond energy (approximately 284.8 eV) results in inaccurate calibration. As a result, it is difficult to accurately measure the band offset using XPS when an insulating layer is formed on a carbon-containing semiconductor material such as diamond. [Prior art documents] [Non-patent literature]

[0008] [Non-Patent Document 1] J.W. Liu, et al., "Band offsets of Al2O3 and HfO2 oxides deposited by atomic layer deposition technique on hydrogenated diamond," Applied Physics Letters, Vol. 101, pp. 252-108, 2012 [Non-patent document 2] A. Marechal and 6 others, "Energy-band diagram configuration of Al2O3 / oxygen-terminated p-diamond metal-oxide-semiconductor", Applied Physics Letters, Vol. 107, No. 141601, 2015 [Non-patent document 3] Fang De, et al., "Calibration of binding energy positions with C1s for XPS results," Journal of Wuhan University of Technology-Mater. Sci. Ed., Vol. 35, p. 711, 2020 [Non-patent document 4] Thomas R. Gengenbach, et al., “Practical guides for x-ray photoelectron spectroscopy (XPS): Interpreting the carbon 1s spectrum”, Vol. 39, No. 013204, 2021 Summary of the Invention [Problem to be solved by the invention]

[0009] In one aspect, the present invention aims to accurately perform calibration when measuring carbon-containing semiconductors by X-ray photoelectron spectroscopy. [Means for solving the problem]

[0010] According to one aspect, a calibration method includes forming a first layer of a metal on a carbon-containing semiconductor, determining a first binding energy resulting from a first energy level of the metal when the semiconductor on which the first layer is formed is measured by X-ray photoelectron spectroscopy, forming a second layer of an insulator on the semiconductor on which the first layer is formed, determining a second binding energy resulting from the first energy level when the second layer is measured by X-ray photoelectron spectroscopy, and calculating a fourth binding energy by calibrating a third binding energy resulting from a second energy level of an element when the second layer is measured by X-ray photoelectron spectroscopy by a difference between the first binding energy and the second binding energy.

[0011] In the above calibration method, forming the first layer may include forming an opening in the first layer, and forming the second layer may include forming the second layer on the semiconductor within the opening.

[0012] The calibration method may further include calculating a first difference between the fourth binding energy when carbon is used as the element and the fourth binding energy when a constituent element of the insulator is used as the element; calculating a second difference between a fifth binding energy resulting from the second energy level of the constituent element included in the third layer of the insulator and a valence band top (VBM) of the third layer; calculating a third difference between a sixth binding energy resulting from the second energy level of carbon included in the semiconductor on which the first layer is formed and a valence band top (VBM) of the semiconductor; and calculating a value obtained by subtracting the first difference and the second difference from the third difference as a valence band offset (VBO) between the semiconductor and the second layer.

[0013] In the above calibration method, the third layer may be thicker than the second layer.

[0014] In the above calibration method, the second layer and the third layer may be of the same phase.

[0015] In the above calibration method, the phase may be an amorphous phase.

[0016] The above calibration method may further include forming the third layer on a silicon substrate, and the fifth binding energy and the valence band top (VBM) of the third layer may be values ​​obtained by calibrating measurements obtained by X-ray photoelectron spectroscopy with the C 1s binding energy.

[0017] In the above calibration method, the valence band top (VBM) of the semiconductor may be a value in a state where the first layer is not formed.

[0018] In the above calibration method, the sixth binding energy may be a value obtained by calibrating the measurement value obtained by X-ray photoelectron spectroscopy without using any other binding energy.

[0019] In the above calibration method, the metal may be any one of Au (gold), Ag (silver), Cu (copper), Al (aluminum), Ni (nickel), Ti (titanium), and Mg (magnesium).

[0020] In the above calibration method, the semiconductor may be any one of diamond, graphite, carbon nanotubes, graphene, and silicon carbide.

[0021] In the above calibration method, the insulator may be any one of alumina, hafnium dioxide, hafnium silicon dioxide, silicon dioxide, silicon oxynitride, zirconium dioxide, lanthanum aluminum trioxide, tantalum pentoxide, titanium dioxide, magnesium oxide, aluminum nitride, and silicon nitride. [Effects of the Invention]

[0022] According to the present invention, accurate calibration can be performed when measuring a carbon-containing semiconductor by X-ray photoelectron spectroscopy. [Brief explanation of the drawings]

[0023] [Figure 1] 1(a) to 1(d) are cross-sectional views (part 1) of a first sample used in the calibration method according to this embodiment during its manufacture. [Figure 2] 2(a) to 2(c) are cross-sectional views (part 2) of the first sample used in the calibration method according to this embodiment during its manufacture. [Figure 3] 3(a) to 3(c) are cross-sectional views (part 3) of the first sample used in the calibration method according to this embodiment during its manufacture. [Figure 4] 4(a) and (b) are cross-sectional views (part 4) of the first sample used in the calibration method according to this embodiment during its manufacture. [Figure 5] 5(a) and 5(b) are plan views of a first sample used in the calibration method according to this embodiment during its manufacture. [Figure 6] FIG. 6 is a plan view showing the setup of the HRXPS in this embodiment. [Figure 7] FIG. 7(a) is an SEM image of the first sample according to this embodiment, and FIG. 7(b) is an SXI image of the first sample obtained by using secondary electrons in HRXPS. [Figure 8] 8(a) and (b) are cross-sectional views of a second sample used in the calibration method according to this embodiment during its manufacture. [Figure 9] FIG. 9(a) is a graph showing how much the binding energy spectrum differs depending on whether or not the first layer is present, and FIG. 9(b) is a graph showing how much the binding energy spectrum differs depending on whether or not the second layer made of alumina is present when the first layer made of gold is formed. [Figure 10] FIG. 10 is a flowchart showing an example of the calibration method according to this embodiment. [Figure 11] FIG. 11(a) is a binding energy spectrum showing the peak energy of C 1s, and FIG. 11(b) is a binding energy spectrum showing the peak energy of Al 2p. [Figure 12]FIG. 12 is a band diagram for explaining a method for calculating the valence band offset (VBO). [Figure 13] FIG. 13 is a diagram showing a method for identifying VBM (B-diamond) and VBM (Al2O3), which are the upper ends of the valence bands of a semiconductor and an insulator, respectively. [Figure 14] FIG. 14 is a flowchart showing a method for calculating the valence band offset (VBO) according to this embodiment. [Figure 15] FIG. 15 is a band diagram near the interface between the semiconductor layer and the second layer based on VBO obtained according to the flowchart of FIG. [Figure 16] FIG. 16 is a diagram showing the hardware configuration of a computer. DETAILED DESCRIPTION OF THE INVENTION

[0024] (Present embodiment) Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Note that like elements are given like reference numerals and their description will be omitted.

[0025] In this embodiment, a sample having a heterostructure of a carbon-containing semiconductor and an insulator is prepared as follows, and the sample is used to calibrate X-ray photoelectron spectroscopy.

[0026] 1 to 4 are cross-sectional views of a first sample used in the calibration method according to this embodiment during its manufacture, and FIGS. 5(a) and 5(b) are plan views thereof.

[0027] First, as shown in Figure 1(a), a type Ib diamond substrate 10 is prepared, and its (100) surface is cleaned with a mixed solution of sulfuric acid and nitric acid. The solution temperature during cleaning is 250°C to 310°C, for example, 300°C. The cleaning time is 2.5 hours to 5 hours, for example, 3 hours.

[0028] Next, as shown in FIG. 1(b), a boron-doped diamond layer is formed as a carbon-containing semiconductor layer 11 on the (100) surface of the diamond substrate 10 by microwave plasma CVD in a chamber (not shown) to a thickness of 50 nm to 3000 nm, for example, 825 nm. The deposition conditions for the semiconductor layer 11 are not particularly limited. For example, the microwave power is 1.2 kW to 1.5 kW, for example, 1.4 kW, and the substrate temperature is 950°C to 1100°C, for example, 1000°C. The deposition pressure is 16 kPa to 20 kPa, for example, 18.6 kPa.

[0029] The source gas for diamond is, for example, a mixed gas of H2 gas and CH4 gas. In this example, the flow rate of H2 gas is set to 47 sccm or more and 49.5 sccm or less, for example, 49 sccm, and the flow rate of CH4 gas is set to 0.5 sccm or more and 3 sccm or less, for example, 1 sccm. Note that, as the source material for boron, the boron remaining in the chamber when a boron-doped diamond film is formed using diborane (B2H6) gas before the semiconductor layer 11 is formed can be used. This allows the semiconductor layer 11 to contain 1×10 boron. 15 cm -3 More than 1×10 18 cm -3 Concentrations below, e.g., 4 x 10 15 cm -3 is doped at a concentration of

[0030] Thereafter, the surface of the semiconductor layer 11 is washed with a mixed solution of sulfuric acid and nitric acid. The solution temperature during washing is 250°C or higher and 310°C or lower, for example, 300°C. The washing time is 2.5 hours or higher and 5 hours or lower, for example, 3 hours. This allows the surface of the semiconductor layer 11, which was terminated with hydrogen immediately after film formation, to be terminated with oxygen.

[0031] Then, the surface of the oxygen-terminated semiconductor layer 11 is washed with acetone, and further washed with ethanol, and then washed with pure water.

[0032] Next, as shown in FIG. 1(c), the binding energy spectrum of the electrons in the semiconductor layer 11 is measured by high-resolution X-ray photoelectron spectroscopy (HRXPS). In the binding energy spectrum, a peak originating from the C 1s of carbon, which is the core level of the semiconductor layer 11, appears. Furthermore, from the binding energy spectrum, the top of the valence band (E VBM (B-diamond)) can also be identified.

[0033] The measurement device and measurement conditions for this HRXPS are not particularly limited. In this example, a PHI Quantera SXM manufactured by ULVAC-PHI, Inc. is used as the measurement device. Al Kα rays with an energy of 1486.6 eV are used as the X-rays irradiated onto the surface of the semiconductor layer 11, and the X-rays are irradiated onto an area with a diameter of 50 μm to 400 μm in plan view, for example, 200 μm.

[0034] The angle (take-off angle θ) between the surface of the semiconductor layer 11 and the incident X-ray is 15° or more and 90° or less, for example, 45°. The pass energy of the analyzer that measures the energy of electrons emitted from the sample surface is 40 eV or more and 80 eV or less, for example, 55 eV. The output energy of the X-ray source is 40 W or more and 80 W or less, for example, 50 W. The step width of the binding energy is 0.05 eV or more and 0.2 eV or less, for example, 0.1 eV.

[0035] Next, the process for obtaining the cross-sectional structure shown in FIG. 1(d) will be described. First, a LOR5A resist manufactured by Nippon Kayaku Co., Ltd. is applied onto the semiconductor layer 11 by spin coating at a rotation speed of 5000 rpm to 8000 rpm (e.g., 7000 rpm) for a coating time of 1 second, followed by baking at 180°C for 5 minutes. Then, an AZ5214E ​​resist manufactured by Merck & Co. is applied by spin coating at a rotation speed of 5000 rpm to 8000 rpm (e.g., 7000 rpm) for a coating time of 1 second, followed by baking at 110°C for 2 minutes. Then, the LOR5A and AZ5214E ​​resists are exposed using a laser exposure system (DL-1000 manufactured by Nano System Solutions). The LOR5A and AZ5214E ​​resists are then developed in a 2.38% TMAH (tetramethylammonium hydroxide) solution for a time of 120 seconds to 150 seconds, forming a first resist pattern 12.

[0036] 2(a), a gold layer is formed as a first metal layer 13 on each of the semiconductor layer 11 and the first resist pattern 12 by vapor deposition to a thickness of 50 nm to 200 nm, for example, 100 nm. The deposition conditions for the first layer 13 are not particularly limited. For example, the deposition pressure is 10 -7 Pa or more 5×10 -4 Pa or less, e.g., 10 -6 Pa, and the film formation rate is 0.05 nm / sec or more and 0.3 nm / sec or less, for example, 0.2 nm / sec.

[0037] Thereafter, as shown in FIG. 2(b), the first resist pattern 12 is removed to pattern the first layer 13 by a lift-off method, leaving the first layer 13 with the opening 13a formed therein on the semiconductor layer 11.

[0038] FIG. 5(a) is a plan view at the time when this step is completed, and the above-mentioned FIG. 2(b) corresponds to a cross-sectional view taken along line II in FIG. 5(a).

[0039] 5(a), a plurality of rectangular openings 13a in plan view are formed at intervals in the first layer 13. When the openings 13a are formed, cross-shaped alignment marks 13b are also formed in the first layer 13. The alignment marks 13b are used to align the diamond substrate 10 with an HRXPS device (not shown) in the HRXPS that is performed later.

[0040] Next, as shown in FIG. 2(c), the binding energy spectrum of electrons in the semiconductor layer 11 is measured by HRXPS. The measurement conditions are the same as those in FIG. 1(c), with the take-off angle set to 45° and the center of X-ray irradiation being the inside of the opening 13a. As a result, part of the X-ray passes through the first layer 13, and the Au 4f of the gold constituting the first layer 13 is irradiated. 7 / 2 A peak due to this appears in the binding energy spectrum.

[0041] In particular, in this example, the opening 13a is formed in the first layer 13, so that the X-rays graze the side surface of the opening 13a and the surrounding first layer 13, and the Au 4f 7 / 2 Peaks derived from this tend to appear in the binding energy spectrum.

[0042] That Au 4f 7 / 2 Peak energy E derived from Au4f7 / 2 (B-diamond) is an example of the first bond energy. Also, Au 4f 7 / 2 is an example of the first energy level. 7 / 2 In addition to the peak due to the C 1s core level of the semiconductor layer 11, a peak due to the C 1s core level also appears in the binding energy spectrum. C1s (B-diamond) is an example of the sixth bond energy.

[0043] 3(a), an amorphous alumina (Al2O3) layer is formed by atomic layer deposition on the first layer 13 and on the semiconductor layer 11 exposed in the opening 13a as the second insulator layer 14. The deposition conditions for the second layer 14 are not particularly limited, but in this example, a mixed gas of trimethylaluminum (Al(CH3)3) and ozone is used as the alumina precursor, and the substrate temperature is set to 100°C to 400°C, for example, 200°C.

[0044] The second layer 14 is a layer simulating an insulating layer in an MIS device, and is formed to a thickness as thin as possible, between 1 nm and 5 nm, for example, 3 nm, similar to that of an MIS device.

[0045] 3(c), a resist is applied onto the first layer 13, and then exposed and developed to form a second resist pattern 15 within the opening 13a. The conditions for forming the second resist pattern 15 are the same as those for forming the first resist pattern (see FIG. 1(d)), and therefore a description thereof will be omitted. The second layer 14 around the opening 13a is exposed and not covered by the second resist pattern 15.

[0046] 3(c), the second layer 14 in the portion not covered with the second resist pattern 15 is etched away with a TMAH solution. This etching is performed at room temperature for, for example, 12 minutes.

[0047] Thereafter, as shown in FIG. 4(a), the second resist pattern 15 is removed to expose the second layer 14 in the opening 13a.

[0048] Next, as shown in FIG. 4(b), the electron binding energy spectrum is measured by HRXPS. The measurement conditions are the same as those in FIG. 1(c), with the take-off angle set to 45° and the center of X-ray irradiation set to the inside of the opening 13a. As a result, the Au 4f of gold constituting the first layer 13 is measured, as in FIG. 2(c). 7 / 2 A peak due to this appears in the binding energy spectrum.

[0049] Furthermore, by forming an opening 13a in the first layer 13 and forming the second layer 14 inside the opening 13a, X-rays can be irradiated onto the side surface of the opening 13a and the surrounding first layer 13 while irradiating the second layer 14 with X-rays. This allows the Au 4f 7 / 2 Peaks derived from

[0050] That Au 4f 7 / 2 Peak energy E derived from Au4f7 / 2 (Al2O3_3nm) is an example of the second binding energy. Also, its binding energy spectrum includes Au 4f 7 / 2 In addition to the peaks due to Al 2p of aluminum, which is a constituent element of the second layer 14, and a peak due to C 1s of the semiconductor layer 11 also appear.

[0051] Al 2p is a core level of the second layer 14 and is an example of a second energy level. The peak energy E Al2p (Al2O3_3nm) is an example of the third bond energy. Similarly, C 1s is the core level of the semiconductor layer 11 and is an example of the second energy level. The peak energy E C1s (Al2O3_3nm) is also an example of the third bond energy.

[0052] With the above steps, the basic structure of the first sample 16 used in the calibration method according to this embodiment is completed.

[0053] The first sample 16 is a sample having a heterostructure in which a second insulating layer 14 is formed on a semiconductor layer 11, and the layers 11 and 14 constitute the semiconductor and insulator of the MIS device.

[0054] In this example, a total of three HRXPS measurements are performed in the steps of FIG. 1(c), FIG. 2(c), and FIG. 4(b).

[0055] FIG. 6 is a plan view showing the setup of each HRXPS.

[0056] As shown in Fig. 6, during measurement by HRXPS, the diamond substrate 10 is placed between a metal holder 20 and a metal jig 21, both of which are maintained at ground potential, and in this state, the diamond substrate 10 is fixed to the metal holder 20 by tightening metal screws 22. As a result, in the HRXPS of Fig. 2(c) and Fig. 4(b), the first metal layer 13 and the metal jig 21 come into contact with each other, and the first layer 13 is at ground potential.

[0057] Fig. 7(a) is a scanning electron microscope (SEM) image of the first sample 16. As shown in Fig. 7(a), a plurality of openings 13a are formed in a matrix in the first sample 16. In the HRXPS steps shown in Fig. 1(c), Fig. 2(c), and Fig. 4(b), any of the plurality of openings 13a may be irradiated with X-rays.

[0058] Figure 7(b) shows an SXI (Scanning X-ray Image) image of the first sample 16 using secondary electrons in HRXPS. The cross mark in this SXI image indicates the center of X-ray irradiation. In HRXPS, X-rays are irradiated so that this cross mark is within the aperture 13a.

[0059] Next, the second sample used in the calibration method according to this embodiment will be described.

[0060] 8(a) and (b) are cross-sectional views of a second sample used in the calibration method according to this embodiment during its manufacture.

[0061] 8(a), an alumina layer is formed on a silicon substrate 30 by atomic layer deposition to a thickness of 15 nm to 50 nm, for example, 30 nm, as a third layer 31 having an amorphous phase, which is the same phase as the second layer 14. In this example, a mixed gas of trimethylaluminum (Al(CH3)3) and ozone is used as the alumina precursor, and the third layer 31 is formed by setting the substrate temperature to 100°C to 400°C, for example, 200°C.

[0062] 8(b), the binding energy spectrum of electrons in the third layer 31 is measured by HRXPS. In the binding energy spectrum, a peak originating from Al 2p of aluminum, which is a constituent element of the third layer 31, appears.

[0063] Al 2p is the core level of the third layer 31, and the peak energy E Al2p (Al2O3_30nm) is an example of the fifth bond energy. Furthermore, from the bond energy spectrum, the top of the valence band (E VBM The measurement conditions for HRXPS are the same as those in Figure 1(c), so a detailed description is omitted.

[0064] In this embodiment, the third layer 31 is formed thicker than the second layer 14, so that secondary electrons emitted from the silicon substrate 30 in response to X-ray irradiation are less likely to pass through the third layer 31. This makes it difficult for the spectrum of secondary electrons originating from the silicon substrate 30 to overlap with the binding energy spectrum measured by HRXPS, making it easier to obtain the binding energy spectrum of secondary electrons originating only from the third layer 31.

[0065] Furthermore, the phase of the third layer 31 is the same amorphous phase as the second layer 14 that simulates the insulating layer of an MIS device, so that the binding energies of Al 2p and VBM in the insulating layer used in an MIS device can be obtained.

[0066] With the above steps, the basic structure of the second sample 33 used in the calibration method according to this embodiment is completed.

[0067] The first sample 16 (Fig. 4(b)) and the second sample 33 (Fig. 8(b)) are both used for calibration to reduce the shift in the HRXPS binding energy spectrum due to charging. The first sample 16 has a first layer 13 made of a gold film, which allows the charge on the sample surface to escape from the first layer 13 to the outside via the metal holder 20 (see Fig. 6). Therefore, the use of the first sample 16 is expected to reduce the shift in the binding energy spectrum due to charging to some extent.

[0068] Fig. 9(a) is a diagram showing the degree of deviation of the binding energy spectrum depending on the presence or absence of the first layer 13. The dotted line in Fig. 9(a) is the binding energy spectrum obtained by HRXPS in Fig. 1(c), which is the spectrum when the first layer 13 is not present. On the other hand, the solid line in Fig. 9(a) is the binding energy spectrum obtained by HRXPS in Fig. 2(c), which is the spectrum when the first layer 13 is present.

[0069] As shown in FIG. 9(a), when the first layer 13 is not present, the peak energy of C 1s is 283.23 eV. On the other hand, when the first layer 13 is present, the peak energy of C 1s is 283.23 eV. C1s (B-diamond) is 284.50 eV. The difference in peak energy corresponds to the shift in the binding energy spectrum due to charging. This shows that by forming the first layer 13, the shift in binding energy due to charging can be corrected by 1.27 eV. However, as will be explained below with reference to FIG. 9(b), it has become clear that the binding energy spectrum of the heterostructure of a carbon-containing semiconductor and an insulator cannot be sufficiently corrected by just forming the first layer 13.

[0070] FIG. 9(b) is a diagram showing how much the binding energy spectrum differs depending on whether or not the second layer 14 made of alumina is present when the first layer 13 made of gold is formed.

[0071] The solid line in Fig. 9(b) is the binding energy spectrum obtained by HRXPS in Fig. 2(c). Au4f7 / 2 (B-diamond) is Au 4f without the second layer 14 of alumina. 7 / 2 In this example, it is 83.97 eV. This value is 7 / 2 This is approximately equal to 83.96 eV, which is the measured value of the peak energy E Au4f7 / 2 (B-diamond) does not need to be calibrated.

[0072] On the other hand, the dotted line in Fig. 9(b) is the binding energy spectrum obtained by HRXPS in Fig. 4(b). Au4f7 / 2 (Al2O3_3nm) is a heterostructure of Au 4f with a second layer 14 of alumina formed on the semiconductor layer 11. 7 / 2 In this example, it is 84.15 eV. This value is 7 / 2 This is significantly shifted towards the higher energy side from the measured value of 83.96 eV.

[0073] As mentioned above, the first layer 13 made of gold prevents the charge on the sample surface from dissipating to the metal holder 20 during HRXPS measurement, thereby suppressing the deviation of the binding energy spectrum due to charging. However, as shown in Figure 9(b), the peak energy E Au4f7 / 2 (Al2O3_3nm) is the peak energy E Au4f7 / 2The binding energy spectrum is shifted by ΔE from (B-diamond). This indicates that when the second layer 14 is formed, charging is not sufficiently suppressed by simply forming the first layer 13 made of a gold layer, and the binding energy spectrum must be calibrated taking into account the shift due to charging. In this example, ΔE = 84.15 eV - 83.96 eV = 0.18 eV, so calibration is required to shift the binding energy to the lower energy side by this value. Based on this, calibration is performed in this embodiment as follows.

[0074] FIG. 10 is a flowchart showing an example of the calibration method according to this embodiment.

[0075] First, in step S1, the user determines the peak energy E in FIG. 9(b) based on the HRXPS binding energy spectrum in FIG. 2(c). Au4f7 / 2 Identify (B-diamond).

[0076] Next, in step S2, the user determines the peak energy E in FIG. 9(b) based on the HRXPS binding energy spectrum in FIG. 4(b). Au4f7 / 2 Identify (Al2O3_3nm).

[0077] Next, the process moves to step S3, in which the user selects E Au4f7 / 2 (Al2O3_3nm) and E Au4f7 / 2 Calculate the difference ΔE from (B-diamond). In this example, ΔE = 0.18 eV as mentioned above.

[0078] Next, the process moves to step S4, in which the user selects the peak energy E Al2p (Al2O3_3nm) and the calibration value E Al2p Calculate (Al2O3_3nm)_cal.

[0079] Peak Energy E Al2p (Al2O3_3nm) is the peak energy in the binding energy spectrum obtained by HRXPS in FIG. 4(b), and corresponds to the binding energy of Al 2p of aluminum, which is a constituent element of the second layer 14. In this example, EAl2p The value obtained by subtracting the difference ΔE from (Al2O3_3nm) is the calibration value E Al2p (Al2O3_3nm)_cal. In this embodiment, E Al2p (Al2O3_3nm)_cal=72.99eV.

[0080] Next, the process moves to step S5, where the user selects the peak energy E C1s (Al2O3_3nm) and the calibration value E C1s Calculate (Al2O3_3nm)_cal.

[0081] Peak Energy E C1s (Al2O3_3nm) is the peak energy in the binding energy spectrum obtained by HRXPS in FIG. 4(b), and corresponds to the binding energy of C 1s in the semiconductor layer 11. In this example, E C1s The value obtained by subtracting the difference ΔE from (Al2O3_3nm) is the calibration value E C1s (Al2O3_3nm)_cal. In this embodiment, E C1s (Al2O3_3nm)_cal=284.15eV.

[0082] In addition, the calibration value E Al2p (Al2O3_3nm)_cal and calibration value E C1s (Al2O3_3nm)_cal are both examples of the fourth bond energy.

[0083] This completes the basic steps of the calibration method according to this embodiment. According to this calibration method, the peak energy of C 1s derived from carbon contamination on the sample surface is not used, but the level of the first layer 13 made of metal (Au 4f 7 / 2 ) is used to calculate the difference ΔE in step S3. Therefore, even if the semiconductor layer 11 contains carbon, such as boron-doped diamond, and calibration using the C 1s peak energy becomes inaccurate, the difference ΔE can be used to accurately calibrate the binding energy spectrum.

[0084] Moreover, as shown in FIG. 9(b), the difference ΔE is E Au4f7 / 2 (Al2O3_3nm) and E when the second layer 14 is not formed. Au4f7 / 2 This is the difference between the bond energy of the second layer 14 and that of the semiconductor layer 11 (B-diamond). Therefore, the difference in bond energy due to the newly generated charge caused by forming the second layer 14 on the semiconductor layer 11 can be extracted as the difference ΔE. As a result, the bond energy spectrum of the heterostructure of the semiconductor layer 11 and the second layer 14 can be accurately calibrated based on the difference ΔE.

[0085] FIG. 11(a) is a binding energy spectrum showing the peak energy of C 1s. The dotted line in this figure is a spectrum obtained by calibrating the electron binding energy spectrum obtained by HRXPS in FIG. 4(b) by subtracting the difference ΔE as described above. As shown in this dotted line spectrum, when the second layer 14 is formed, the peak energy E of C 1s after calibration is 284.15 eV. C1s (Al2O3_3nm)_cal appears. In addition, a peak derived from the carbon-boron bond of the semiconductor layer 11 appears at a position of 282.5 eV.

[0086] On the other hand, the solid line in Fig. 11(a) is the binding energy spectrum obtained by HRXPS in Fig. 2(c), and is the same as the solid line spectrum in Fig. 9(a). As shown in this solid line spectrum, in the state where the second layer 14 is not formed, E C1s (B-diamond) appears at an energy of 284.50 eV.

[0087] From this result, it was found that when the second layer 14 was formed, the binding energy derived from C 1s shifted to the lower energy side by 0.35 eV.

[0088] FIG. 11(b) is a binding energy spectrum showing the peak energy of Al 2p of alumina. The dotted line is a spectrum obtained by calibrating the electron binding energy spectrum obtained by HRXPS in FIG. 4(b) by subtracting the difference ΔE as described above. As shown in FIG. 4(b), in this HRXPS, a first layer 13 made of a gold layer is formed. As shown in the dotted line spectrum, when the first layer 13 is formed, the peak energy E of Al 2p after calibration is 72.99 eV. Al2p (Al2O3_3nm)_cal appears.

[0089] On the other hand, the solid line in Figure 11(a) is a spectrum obtained by calibrating the binding energy spectrum obtained by HRXPS in Figure 8(b) based on the binding energy of C 1s. As shown in Figure 8(b), in this HRXPS, the first layer 13 made of a gold layer was not formed. As shown in the spectrum of the solid line, in this state where the first layer 13 was not formed, the peak energy E of Al 2p after calibration was 74.57 eV. Al2p (Al2O3_30nm)_cal appears. Also, a peak derived from the Al-O-OH bond on the surface of the second alumina layer 14 appears at 77.0 eV.

[0090] From this result, it was found that when the first layer 13 was formed, the binding energy derived from Al 2p shifted by 1.58 eV to the lower energy side.

[0091] Next, a method for calculating the valence band offset (VBO) using the calibration value will be described.

[0092] FIG. 12 is a band diagram for explaining a method for calculating the valence band offset (VBO).

[0093] This band diagram is a band diagram near the interface between a semiconductor and an insulator. Here, we use boron-doped diamond as the semiconductor and alumina as the insulator. The VBO of this system is calculated using the following equation (1):

[0094] VBO=ΔE3-ΔE1-ΔE2 … (1)

[0095] Here, ΔE1 is an example of the first difference, and is the difference between the binding energy of Al 2p in an insulator and the binding energy of C 1s in a semiconductor.

[0096] Furthermore, ΔE2 is an example of the second difference, and is the difference between the binding energy of the top of the valence band in an insulator (VBM (Al2O3)) and Al 2p.

[0097] ΔE3 is an example of the third difference, and is the difference between the valence band top (VBM (B-diamond)) in a semiconductor and the binding energy of C 1s.

[0098] The band gaps E of semiconductors and insulators are g (Al2O3), E g When (B-diamond) is known, the conduction band offset (CBO) can be calculated from the following equation (2):

[0099] CBO=E g (Al2O3)-E g (B-diamond)-VBO … (2)

[0100] FIG. 13 is a diagram showing a method for identifying VBM (B-diamond) and VBM (Al2O3), which are the upper ends of the valence bands of a semiconductor and an insulator, respectively.

[0101] The black circles in Figure 13 represent the binding energy spectrum measured by HRXPS in Figure 1(c), and the white circles represent the binding energy spectrum measured by HRXPS in Figure 8(b).

[0102] In this example, the E of the third layer 31 is used as the valence band upper limit (VBM (Al2O3)) in the insulator. VBM (Al2O3_30nm) calibrated based on the C 1s binding energy from carbon contamination VBM (Al2O3_30nm)_cal is adopted.

[0103] 8(a) and 8(b), the third layer 31 is formed on the silicon substrate 30, which does not contain carbon, so using the C 1s binding energy in this way does not result in significant inaccuracy in calibration. In this case, in the HRXPS of Fig. 8(b), calibration can be performed by identifying the C 1s binding energy resulting from carbon contamination on the surface of the third layer 31 and shifting the energy spectrum by the difference between that binding energy and the known C 1s binding energy (approximately 284.8 eV).

[0104] And E VBM (Al2O3_30nm)_cal is the binding energy at the intersection of the fitting line of the series of white circles where the intensity is approximately constant and the fitting line of the series of white circles where the intensity begins to rise, and in this example is 3.38 eV.

[0105] The upper valence band edge VBM (B-diamond) of the semiconductor is determined by the HRXPS of the semiconductor layer 11 shown in FIG. 1(c). VBM (B-diamond) calibrated based on the C 1s binding energy VBM (B-diamond))_cal is adopted.

[0106] Since the semiconductor layer 11 contains carbon, the C 1s binding energy changes depending on the bonding state of the carbon in the film, which can lead to inaccuracies in the calibration using the C 1s binding energy. Therefore, instead of the HRXPS shown in Figure 1(c), the binding energy spectrum of the semiconductor layer 11 is measured by HRXPS with the first layer 13 made of a gold layer formed as shown in Figure 2(c), and the measurement results are used as the Au 4f 7 / 2 However, this method may cause the VBM of the semiconductor layer 11 to change due to the first layer 13 made of a gold layer. Therefore, in this embodiment, the HRXPS measurement results shown in FIG. 1(c) are calibrated based on the C 1s binding energy, and the VBM of the semiconductor layer 11 is calibrated based on the C 1s binding energy. VBM (B-diamond))_cal is adopted.

[0107] In this case, E VBM (B-diamond)_cal is the binding energy at the intersection of the fitting line of the series of black circles where the intensity is approximately constant and the fitting line of the series of black circles where the intensity begins to rise, and in this example is 0.85 eV.

[0108] FIG. 14 is a flowchart showing a method for calculating the valence band offset (VBO).

[0109] First, in step S11, the user calculates the first difference ΔE1. ​​Here, the calibration value E calculated in steps S4 and S5 of FIG. C1s (Al2O3_3nm)_cal and E Al2p The difference between (Al2O3_3nm)_cal is calculated as the first difference ΔE1. ​​In this example, ΔE1 = E C1s (Al2O3_3nm)_cal-E Al2p (Al2O3_3nm)_cal=284.15eV-72.99eV=211.16eV.

[0110] Next, the process proceeds to step S12, where the user calculates the second difference ΔE2. Here, the user calculates the peak energy E Al2p (Al2O3_30nm) calibrated based on the C 1s binding energy from carbon contamination Al2p (Al2O3_30nm)_cal. Then, the user can calculate the E Al2p (Al2O3_30nm)_cal and E in Fig. 13 VBM The difference from (Al2O3_30nm)_cal is calculated as the second difference ΔE2.

[0111] E VBM For the same reason as (Al2O3_30nm)_cal, the third layer 31 is formed on the silicon substrate 30 that does not contain carbon, so E Al2p The use of the C 1s binding energy due to carbon contamination in the calibration of (Al2O3_30nm) does not result in significant inaccuracy in the calibration.

[0112] In this embodiment, as described above, E Al2p (Al2O3_30nm)_cal=74.57eV, E VBM Since (Al2O3_30nm)_cal=3.38eV, ΔE2=74.57eV-3.38eV=71.19eV.

[0113] Next, the process proceeds to step S13, where the user calculates the third difference ΔE3. Here, the user calculates the E measured by the HRXPS in FIG. C1s (B-diamond) and E in Figure 13 VBM The difference between the (B-diamond)_cal is calculated as a third difference ΔE3.

[0114] E C1s (B-diamond) is the value obtained from HRXPS in Fig. 2(c), and is the value obtained from C 1s and Au 4f 7 / 2 This is because, in the state where there is no insulating layer such as the second layer 14 on the semiconductor layer 11 as shown in FIG. 2(c), the sample surface is unlikely to be charged, and the deviation in binding energy due to charging is thought to be sufficiently suppressed by the first layer 13 made of a gold layer.

[0115] In this embodiment, as described above, E C1s (B-diamond) = 284.50 eV, E VBM Since (B-diamond)_cal=0.85eV, ΔE3=284.50eV-0.85eV=283.65eV.

[0116] Thereafter, the process proceeds to step S14, where the user calculates VBO by substituting the values ​​obtained in steps S11 to S13 into each term on the right side of the above-mentioned equation (1). In this embodiment, VBO = 283.65 eV - 211.16 eV - 71.19 eV = 1.30 eV.

[0117] This completes the basic process of the VBO calculation method according to this embodiment. According to the VBO calculation method described above, in step S11, the peak energies E Au4f7 / 2 (Al2O3_3nm), E Au4f7 / 2 Each calibration value E is calibrated based on (B-diamond) C1s (Al2O3_3nm)_cal, E Al2p The difference between (Al2O3_3nm)_cal is calculated as the first difference ΔE1. ​​Therefore, even if the semiconductor layer 11 contains carbon, such as boron-doped diamond, and calibration using the C 1s peak energy may be inaccurate, VBO can be accurately calculated from equation (1).

[0118] Fig. 15 is a band diagram near the interface between the semiconductor layer 11 and the second layer 14 based on VBO obtained according to the flowchart of Fig. 14. In Fig. 15, the dotted line represents the band diagram when the second layer 14 is not present, and the solid line represents the band diagram when the second layer 14 is present.

[0119] As described above, the doping concentration of boron in the semiconductor layer 11 is 4×10 15 cm -3 In the case of bulk boron-doped diamond, the Fermi energy E F The difference between the VBM and the ion beam is estimated to be about 0.36 eV.

[0120] Without the second layer 14, oxygen vacancies and defects at the surface of the semiconductor layer 11 would cause the surface to be positively charged, which would cause the VBM at the surface to bend downwards and increase the Fermi energy E F It is only 0.85 eV lower than

[0121] On the other hand, when the second alumina layer 14 is formed by atomic layer deposition using a precursor containing ozone, the VBM on the surface of the semiconductor layer 11 bends downward, but the bending is gentler. This is thought to be because the oxygen vacancies on the surface of the semiconductor layer 11 are compensated for by ozone.

[0122] In addition, the known band gaps E g (B-diamond), E g (Al2O3) are 5.47 eV and 7.2 eV, respectively. Substituting these values ​​and the VBO value (1.30 eV) into equation (2), CBO becomes 0.43 eV.

[0123] (Other embodiments) In the above, the steps in the flowcharts of FIGS. 10 and 14 are executed by a user, but these steps may also be executed by a computer.

[0124] 16 is a hardware configuration diagram of the computer. The computer 100 is an example of a calibration device, and includes a storage device 101, a memory 102, a processor 103, a communication interface 104, and a media reader 105. These components are connected to each other by a bus 106.

[0125] Of these, the storage device 101 is a non-volatile storage such as a hard disk drive (HDD) or a solid state drive (SSD), and stores a calibration program 110 according to this embodiment.

[0126] The calibration program 110 may be recorded on a computer-readable recording medium 111 and the processor 103 may read the calibration program 110 via the medium reader 105 .

[0127] Such recording media 111 include physically portable recording media such as CD-ROMs (Compact Disc - Read Only Memory), DVDs (Digital Versatile Discs), and USB (Universal Serial Bus) memories. Also, semiconductor memories such as flash memories and hard disk drives may be used as the recording media 111. These recording media 111 are not temporary media such as carrier waves that do not have a physical form.

[0128] Furthermore, the calibration program 110 may be stored in a device connected to a public line, the Internet, a LAN (Local Area Network), etc. In this case, the processor 103 may read and execute the calibration program 110.

[0129] On the other hand, the memory 102 is hardware that temporarily stores data, such as a DRAM (Dynamic Random Access Memory).

[0130] The processor 103 is hardware such as a CPU (Central Processing Unit) or GPU (Graphical Processing Unit) that controls each part of the computer 100. The processor 103 also executes the calibration program 110 in cooperation with the memory 102. This causes the processor 103 to execute each step of the flowcharts in FIGS.

[0131] Furthermore, the communication interface 104 is hardware such as a network interface card (NIC) for connecting the computer 100 to a network such as the Internet or a LAN.

[0132] The medium reading device 105 is hardware such as a CD drive, a DVD drive, or a USB interface for reading the recording medium 111 .

[0133] In this way, by having the computer 100 execute the calibration method according to this embodiment, it is possible to accurately and quickly calibrate a binding energy spectrum, such as that of a carbon-containing semiconductor, which is difficult to calibrate using the C 1s binding energy. Furthermore, the computer 100 can accurately and quickly calculate the VBO at the interface between a semiconductor and an insulator. As a result, for example, by having the computer 100 present the calculated VBO to a user as manufacturing data for an MIS-type device, a highly reliable MIS-type device can be manufactured based on the VBO.

[0134] Although each embodiment has been described in detail above, each embodiment is not limited to the above.For example, in the above, a boron-doped diamond layer is formed as the semiconductor layer 11, but the semiconductor layer 11 is not limited to this as long as it is a carbon-containing semiconductor layer.For example, any layer of diamond, graphite, carbon nanotube, graphene, or silicon carbide may be formed as the semiconductor layer 11.

[0135] Furthermore, although alumina is used as the insulator of the second layer 14 in the above, it is not limited to alumina as long as it is an insulator for MIS devices. Examples of such insulators include alumina, hafnium dioxide, hafnium silicon dioxide, silicon dioxide, silicon oxynitride, zirconium dioxide, lanthanum aluminum trioxide, tantalum pentoxide, titanium dioxide, magnesium oxide, aluminum nitride, and silicon nitride.

[0136] Furthermore, there are no particular limitations on the metal that constitutes the first layer 13. For example, the first layer 13 may be formed as a metal layer of any of Au (gold), Ag (silver), Cu (copper), Al (aluminum), Ni (nickel), Ti (titanium), and Mg (magnesium). [Explanation of symbols]

[0137] 10...diamond substrate, 11...semiconductor layer, 12...first resist pattern, 13...first layer, 13a...opening, 13b...alignment mark, 14...second layer, 15...second resist pattern, 16...first sample, 20...metal holder, 21...metal jig, 22...metal screw, 30...silicon substrate, 31...third layer, 100...computer, 101...storage device, 102...memory, 103...processor, 104...communication interface, 105...media reader, 106...bus, 111...recording medium.

Claims

1. forming a first layer of metal on a carbon-containing semiconductor; Identifying a first binding energy derived from a first energy level of the metal when the semiconductor on which the first layer is formed is measured by X-ray photoelectron spectroscopy; forming a second layer of an insulator on the semiconductor having the first layer formed thereon; determining a second binding energy derived from the first energy level when the second layer is measured by X-ray photoelectron spectroscopy; calculating a fourth binding energy by correcting a third binding energy derived from a second energy level of the element when the second layer is measured by X-ray photoelectron spectroscopy by a difference between the first binding energy and the second binding energy; Calibration methods, including:

2. forming the first layer includes forming an opening in the first layer; forming the second layer includes forming the second layer on the semiconductor in the opening; The calibration method according to claim 1 .

3. calculating a first difference between the fourth bond energy when carbon is used as the element and the fourth bond energy when a constituent element of the insulator is used as the element; calculating a second difference between a fifth binding energy derived from the second energy level of the constituent element included in the third layer of the insulator and a valence band top (VBM) of the third layer; calculating a third difference between a sixth binding energy derived from the second energy level of carbon contained in the semiconductor on which the first layer is formed and a valence band top (VBM) of the semiconductor; calculating a value obtained by subtracting the first difference and the second difference from the third difference as a valence band offset (VBO) between the semiconductor and the second layer; The calibration method according to claim 1 or claim 2, further comprising:

4. the third layer is thicker than the second layer; The calibration method according to claim 3 .

5. the second layer and the third layer are of the same phase; The calibration method according to claim 3 or 4.

6. The phase is an amorphous phase. The calibration method according to claim 5 .

7. forming the third layer on a silicon substrate; the fifth binding energy and the valence band top (VBM) of the third layer are values ​​obtained by calibrating measurements by X-ray photoelectron spectroscopy with the binding energy of C 1s; The calibration method according to any one of claims 3 to 6.

8. the valence band edge (VBM) of the semiconductor is a value in a state where the first layer is not formed; The calibration method according to any one of claims 3 to 7.

9. The sixth binding energy is a value obtained by calibrating a measurement value obtained by X-ray photoelectron spectroscopy using no other binding energy. The calibration method according to any one of claims 3 to 8.

10. The metal is any one of Au (gold), Ag (silver), Cu (copper), Al (aluminum), Ni (nickel), Ti (titanium), and Mg (magnesium). The calibration method according to any one of claims 1 to 9.

11. The semiconductor is any one of diamond, graphite, carbon nanotubes, graphene, and silicon carbide. The calibration method according to any one of claims 1 to 10.

12. the insulator is any one of alumina, hafnium dioxide, hafnium silicon dioxide, silicon dioxide, silicon oxynitride, zirconium dioxide, lanthanum aluminum trioxide, tantalum pentoxide, titanium dioxide, magnesium oxide, aluminum nitride, and silicon nitride; The calibration method according to any one of claims 1 to 11.

Images