Method for evaluating the quality of oxide films

Abstract

To provide an oxide film quality evaluation method that evaluates and measures oxygen vacancies in an oxide film deposited under the same condition as that used in an actual device.SOLUTION: A method includes preparing a sample 101 in which a conductive layer 16, a semiconductor layer 12, an oxide film to be measured 14, and a conductive film 15 made of a hydrogen catalyst metal or a hydrogen catalyst alloy are sequentially laminated, and measuring the quality of the oxide film 14 to be measured, especially the oxygen vacancies in the film by measuring the C-V between the conductive layer 16 and a conductive film 15 in a hydrogen atmosphere or measuring the capacitance at a bias of 0 V, etc.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a method for evaluating the film quality of an oxide film. Here, the evaluation of the film quality of the oxide film in the present invention means measuring the amount of oxygen deficiency in the oxide film and evaluating the degree thereof as the film quality.

Background Art

[0002] In semiconductor devices such as MOSFETs, the film quality of oxide films including the gate insulating film is extremely important because it directly affects device performance and reliability. Various methods for evaluating the film quality of oxide films have been disclosed, for example, as disclosed in Patent Documents 1 and 2. However, it is difficult to evaluate oxygen deficiency among the film qualities of oxide films. Regarding the oxygen deficiency in oxide films, as disclosed in Non-Patent Document 1, there are reports of evaluating the film quality of oxide films by optical methods such as the photoluminescence method after depositing a thick oxide film. However, in order to obtain the necessary measurement sensitivity, the oxide film has to be deposited very thickly. Therefore, the evaluation is carried out under conditions different from those of actual devices, and there are problems with the reliability of the measurement.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Non-Patent Documents

[0004]

Non-Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] The problem that this invention aims to solve is to provide a method for evaluating and measuring oxygen vacancies in oxide films deposited under the same conditions as those used in actual devices. [Means for solving the problem]

[0006] The configuration of the present invention is shown below. (Composition 1) Prepare a sample in which a conductive layer, a semiconductor layer, an oxide film to be measured, and a conductive film consisting of a hydrogen catalyst metal or hydrogen catalyst alloy are sequentially laminated, A method for evaluating the film quality of an oxide film, comprising measuring the electrical capacitance between the conductive layer and the conductive film under a hydrogen atmosphere. (Configuration 2) The method for evaluating the film quality of an oxide film according to configuration 1, wherein the volume concentration of the hydrogen atmosphere is 100 ppm or more and 100% or less. (Composition 3) The method for evaluating the film quality of an oxide film according to configuration 1 or 2, wherein the hydrogen partial pressure of the hydrogen atmosphere is 1 kPa or more and 10 atmospheres or less. (Composition 4) The method for evaluating the film quality of an oxide film according to any one of the three claims, wherein the conductive film comprises one or more metals selected from the group consisting of platinum, palladium, iridium, ruthenium, and nickel. (Composition 5) The method for evaluating the film quality of an oxide film according to any one of the configurations 1 to 4, wherein the thickness of the conductive film is 5 nm or more and 1 μm or less. (Composition 6) The method for evaluating the film quality of an oxide film according to any one of the configurations 1 to 5, wherein the semiconductor layer consists of one or more selected from the group consisting of GaN, Si, GaAlN, GaAs, ZnO, SiC, and Ga2O3. (Composition 7) A method for evaluating the film quality of an oxide film according to any one of the configurations 1 to 6, wherein the thickness of the semiconductor layer is 100 nm or more and 1 mm or less. (Composition 8) The method for evaluating the film quality of an oxide film according to any one of the items 1 to 7, wherein the thickness of the oxide film to be measured is 0.1 nm or more and 1 μm or less. (Composition 9) A method for evaluating the film quality of an oxide film according to any one of the configurations 1 to 8, wherein an electric field is applied to the oxide film to be measured. (Composition 10) A method for evaluating the film quality of an oxide film according to any one of the configurations 1 to 8, wherein a bias voltage is applied between the conductive layer and the conductive film. (Composition 11) The method for evaluating the film quality of an oxide film according to configuration 10, wherein the bias is between -20V and 10V. (Composition 12) The aforementioned capacity measurement is performed in a temperature environment of 23°C to 200°C, and is a method for evaluating the film quality of an oxide film as described in any one of items 1 to 11. [Effects of the Invention]

[0007] According to the present invention, a method is provided that can evaluate and measure oxygen vacancies in an oxide film deposited under the same conditions as those used in actual devices. [Brief explanation of the drawing]

[0008] [Figure 1] This is a cross-sectional structural diagram showing the composition of sample 101. [Figure 2] This is a characteristic curve showing the bias start voltage dependence of the CV characteristics of an HfO2-δ film. [Figure 3] This is a characteristic diagram showing the CV characteristics of sample 101. [Figure 4] This is a characteristic diagram showing the response characteristics of sample 101. [Figure 5] This is a characteristic diagram showing the dependence of the measured oxide film on capacitance at a bias of 0V. [Figure 6] This is a characteristic diagram showing the dependence of the relative charge density in the film on the oxide film being measured. [Figure 7] This is a characteristic diagram showing the recovery characteristics (initialization characteristics) of sample 101. [Modes for carrying out the invention]

[0009] The embodiments for carrying out the present invention will be described below with reference to the drawings.

[0010] <Embodiment 1> Embodiment 1 describes the configuration and measurement principle of the present invention.

[0011] <Sample structure and measurement method> In Embodiment 1, a sample 101 is used, which has a structure in which a conductive layer 16, a semiconductor substrate 11 and a semiconductor layer 12 (a semiconductor layer 13), an oxide film to be measured 14, and a conductive film 15 are sequentially stacked as shown in Figure 1. The sample 101 is subjected to CV measurement in a hydrogen gas environment, or capacitance measurement with a bias of 0V, etc., to evaluate the film quality of the oxide film to be measured, particularly the oxygen vacancy in the film. Here, the conductive layer 16 is not essential; if the semiconductor substrate 11 has low electrical resistance due to a high concentration of dopants or the like, the semiconductor substrate 11 can be used instead of the conductive layer 16. Also, the semiconductor layer 13 does not necessarily have to consist of the semiconductor substrate 11 and the semiconductor layer 12; it can be composed of the semiconductor layer 12 alone.

[0012] The oxide film 14 to be measured is not particularly limited, and its thickness is acceptable to be between 0.1 nm and 1 μm. The oxide film 14 to be measured may be an oxide film deposited under the same conditions as those used in actual devices. Furthermore, if the thickness of the oxide film 14 under test is within this range, the hydrogen atoms generated by catalytic action from hydrogen gas by the conductive film 15 (described later) are sufficiently trapped in the oxygen-deficient portions of the oxide film 14 under test. This allows for film quality evaluation to be performed with sufficient speed and sensitivity, and also allows for an acceptable recovery time when resetting to the initial state.

[0013] It is preferable that the oxide film 14 to be measured is formed in contact with the semiconductor layer 13. This is because a virtual image of the charge caused by atomic hydrogen trapped in the oxygen vacancies of the oxide film 14 to be measured, as described later, can be sufficiently formed on the semiconductor layer 13.

[0014] The conductive film 15 is a conductive film made of a hydrogen catalyst metal or hydrogen catalyst alloy, and has the function of converting ambient hydrogen gas into atomic hydrogen through catalytic action, and the function of temporarily storing the generated atomic hydrogen in the conductive film 15 and supplying atomic hydrogen to the oxide film 14 to be measured by diffusion.

[0015] Specific metals for the conductive film 15 include Pt (platinum), palladium (Pd), iridium (Ir), ruthenium (Ru), and Ni (nickel), and these metals may be used individually or in combination. Here, "combination" means one or more selected from the group consisting of using them in separate locations, using them in a laminated form, and using them as an alloy. Specific alloys for the conductive film 15 include LaNi5, ReNi5, MgNi, Mg2Ni, and Ce2Ni7. In order to efficiently supply the generated atomic hydrogen to the oxide film 14 under test, it is preferable that the conductive film 15 be in direct contact with the oxide film 14 under test. The thickness of the conductive film 15 is preferably 5 nm to 1 μm, from the viewpoint of ensuring the efficient generation, absorption, and diffusion supply of atomic hydrogen to the oxide film 14 to be measured. Methods for forming the conductive film 15 include thermal and electron beam deposition, sputtering, and resistance heating deposition.

[0016] As shown in Figure 1, the semiconductor layer 13 may consist of multiple layers, such as the semiconductor substrate 11 and the semiconductor layer 12, or it may consist of a single layer. The role of the semiconductor layer 13 is to generate a charge within the semiconductor layer that corresponds to the hydrogen-induced charge generated in the oxide film 14 under measurement, thereby enabling sensitive detection of hydrogen. As a result, the hydrogen gas detection sensitivity is significantly improved compared to when the semiconductor layer is replaced with a conductive layer such as a metal, and consequently, the measurement sensitivity for evaluating film quality is enhanced. A thickness of 100 nm to 1 mm for the semiconductor layer 13 is preferable because it easily enhances measurement sensitivity. Specific examples of the semiconductor layer 13 include one or more selected from the group consisting of GaN, Si, GaAlN, GaAs, ZnO, SiC, and Ga2O3.

[0017] The conductive layer 16 is not particularly limited as long as it functions as an electrode that can make ohmic contact with the semiconductor layer 13. For example, Pt (platinum), Au (gold), Ag (silver), W (tungsten), Pd (palladium), Cu (copper), and polysilicon can be used. When using Pt or the like, it is preferable to form a thin film of Ti between it and the semiconductor layer 13 for purposes such as improving adhesion. In order to make ohmic contact with the semiconductor layer 13, if the substrate is n-type, it is preferable to use a metal with a small work function, such as Ti or Al, as the metal in contact with the semiconductor layer 13. However, even if the substrate is n-type, most metals are acceptable if the concentration is high.

[0018] When performing CV measurements to evaluate the film quality of the oxide film 14 under test, a hydrogen gas environment is used as described above. Preferably, the volume concentration of the hydrogen atmosphere is 100 ppm or more and 100% or less, and the partial pressure of hydrogen is 1 kPa or more and 10 atmospheres or less. When the hydrogen gas environment falls within this range, a sufficient amount of atomic hydrogen is generated by the catalytic action of the conductive film 15, and this fills most of the oxygen-deficient areas of the oxide film 14 under test, making it possible to evaluate the film quality with high accuracy.

[0019] <Sample preparation method> Sample 101 can be prepared by the following process. First, a semiconductor layer 13 is prepared, and the interface of the semiconductor layer is cleaned by washing or other means before forming the oxide film 14 to be measured. In this case, it is preferable to form it under the same conditions as the actual device. Subsequently, a sample 101 is prepared by forming a conductive film 15 on the oxide film 14 to be measured, and a conductive layer 16 on the underside of the semiconductor layer 13, using methods such as thermal or electron beam deposition or sputtering.

[0020] <Measurement principle> The measurement principle of Embodiment 1 is shown below. The hydrogen gas in the measurement environment is converted to atomic hydrogen by the catalytic action of the conductive film 15. The generated atomic hydrogen is absorbed into the conductive film 15, and the absorbed atomic hydrogen diffuses into the oxide film 14 to be measured and is trapped in the oxygen-deficient portions of the oxide film 14. Here, since atomic hydrogen is trapped in most of the oxygen-deficient portions, the amount of trapped atomic hydrogen has a strong correlation with the amount of oxygen deficiency. Atomic hydrogen trapped in oxygen-depleted regions becomes positively charged as a result of interacting with the oxygen-depleted regions. This positively charged atomic hydrogen generates negative image charges within the semiconductor layer, shifting the CV characteristics to the negative voltage side. By utilizing this characteristic, the oxygen vacancy in the oxide film 14 under test can be detected and quantified by the voltage shift between the CV characteristic curve obtained under an inert gas environment (such as N2) and the CV characteristic curve measured under a hydrogen gas environment. A calibration curve can be used for this quantification.

[0021] Detailed investigations into the CV characteristics revealed that applying an electric field to the oxide film 14 under test accelerates the recovery of the sample 101 to its initial state. Since the hydrogen present in the oxide film 14 under test is positively charged, applying an appropriate electric field can speed up the desorption process from the oxide film 14. The electric field may be applied from outside the sample 101, or it may be applied by applying a bias between the conductive film 15 and the conductive layer 16.

[0022] Extensive experiments have revealed that a bias of -20V to 10V is preferable. This is because the positively charged hydrogen present in the oxide film 14 under test is in a relatively stable state, and there is a threshold for its movement by an externally applied electric field. It is preferable to apply an electric field large enough so as not to damage the oxide film 14 under test.

[0023] CV measurements are preferably performed in an environment between 23°C and 200°C. Higher temperatures increase catalytic efficiency, improving the efficiency of atomic oxygen generation and thus the measurement sensitivity. However, they also react with components other than hydrogen gas, such as water and hydrocarbons, increasing noise. Therefore, it is preferable to limit the temperature to 200°C.

[0024] Here, by monitoring the change in capacitance before and after the introduction of hydrogen gas under a bias of 0V, atomic hydrogen trapped in the oxygen-deficient region can be detected in the form of a positive charge generated in the film. As described above, the positive charge generated in the oxide film 14 under test is strongly correlated with the amount of oxygen deficiency. Therefore, by measuring the capacitance between the conductive film 15 and the conductive layer 16 under a bias of 0V before and after the introduction of hydrogen gas, it is possible to detect and measure the amount of oxygen deficiency in the oxide film 14 under test, and the film quality of the oxide film 14 under test can also be evaluated using this method. However, this method is suitable when the change in the CV curve (the slope of the curve) under a hydrogen gas environment is near a bias of 0V. If the change is far from the vicinity of a bias of 0V, the change in the measured quantity is small, so the measurement sensitivity and accuracy decrease.

[0025] As described above, this method is a novel method for evaluating the film quality of oxide films, particularly oxygen vacancies in oxide films, based on a principle that involves atomic hydrogenation of hydrogen gas, trapping of the atomic hydrogen in oxygen-vacuated areas of the oxide film under test, and quantifying the effect of the charge generated by the trapping using CV measurement or capacitance measurement with the bias voltage fixed to 0V. This method allows for the evaluation and measurement of oxygen vacancies in oxide films deposited under the same conditions as those used in actual devices. [Examples]

[0026] (Example 1) In Example 1, sample 101 was prepared and its properties were evaluated. The following explanation will be given with reference to the figures.

[0027] <Sample> As shown in FIG. 1, the sample 101 fabricated in Example 1 has a structure in which a conductor layer 16 made of Pt (platinum) / Ti (titanium), a semiconductor layer 13 made of a semiconductor substrate 11 and an n-GaN semiconductor layer 12, a measured oxide film 14, and a conductor film 15 made of Pt are sequentially laminated.

[0028] Here, the conductor layer 16 is formed by an electron beam evaporation method, and the thicknesses of Pt and Ti are 100 nm and 20 nm, respectively. Ti is a film formed for the purpose of strengthening the adhesion of the conductor layer 16 to the semiconductor substrate 11, and is formed so as to contact the semiconductor substrate 11. The conductor layer 16 has an ohmic contact with the semiconductor substrate 11.

[0029] The semiconductor substrate 11 is a self-supporting n + -GaN(0001) wafer with a thickness of 300 μm, and the dislocation density and carrier concentration are on the order of 10 6 cm -2 and 1×10 18 cm -3 respectively. The semiconductor layer 12 is a Si (silicon)-doped GaN homoepitaxial layer formed by metalorganic vapor phase epitaxy. Its thickness is 5 μm and the Si concentration is 2×10 16 cm -3 respectively.

[0030] As the measured oxide film 14, three types of Al2O3, HfO 2-δ , and Hf 0.57 Si 0.43 O 2-δ were prepared. Here, δ indicates oxygen deficiency. From the characteristic evaluation of the devices mounted with HfO 2-δ , and Hf 0.57 Si 0.43 O 2-δ , the δ of Example 1 is estimated to be about 0.02. Al2O3 is formed by the ALD method using trimethylaluminum (TMA) as a precursor and H2O as an oxidant gas, and is estimated to be a film with less oxygen deficiency for the following reasons.

[0031] HfO 2-δ As shown in Figure 2, when measuring CV by gradually increasing the voltage on the storage side, the CV curve shifts to the positive voltage side as the voltage increases. This is because of HfO 2-δ It is thought that the oxygen vacancies contained within act as electron traps, and a very large number of oxygen vacancies are expected. On the other hand, although oxygen vacancies also exist in Al2O3, when the sweep voltage range is changed during CV measurement, HfO 2-δ As shown above, no shift in the CV curve is observed, and all CV curves with different sweep voltage ranges overlap perfectly. From this, it can be concluded that Al2O3 is HfO 2-δ This membrane is considered to have less oxygen deficiency compared to others.

[0032] HfO 2-δ Hf 0.57 Si 0.43 O 2-δ These films were also formed using the Atomic Layer Deposition (ALD) method. The formation temperature was 300°C, and the thickness of both films was 15 nm. As a pretreatment for forming these films using the ALD method, the samples were washed with a mixture of sulfuric acid and hydrogen peroxide (H2SO4:H2O2=1:1) to remove organic residues, and then treated with buffered hydrofluoric acid (BHF; HF:NH4F=1:6). These films were then deposited immediately afterward. HfO 2-δ For the deposition, tetrakis(dimethylamino)hafnium and O2 plasma were used as precursors. 0.57 Si 0.43 O 2-δ HfO 2-δ The film is a laminated film consisting of a layer and an SiO2 layer, with thicknesses of 0.164 nm and 0.134 nm, respectively. The SiO2 layer was formed by ALD (Advanced Laser Development) using tris(dimethylamino)silane and O2 plasma.

[0033] Furthermore, samples were prepared from the oxide film 14 to be measured, one with and one without post-deposition annealing (PDA), and a characteristic evaluation was performed to investigate the effect of PDA. Here, the PDA conditions were 800°C for 5 minutes under an N2 atmosphere at a pressure of 1 atm. 2-δ It is in the monoclinic phase, Hf 0.57 Si 0.43 O 2-δ It has been confirmed that it is amorphous.

[0034] The conductive film 15 is a circular electrode with a thickness of 100 nm, formed by electron beam deposition via a shadow mask, and its area is 1 × 10⁻⁶. -4 cm 2 That is the case.

[0035] <Sample preparation> Sample 101 was prepared by the following process. First, a semiconductor substrate 11 with the above specifications was prepared, and an n-GaN semiconductor layer 12 was formed on it using the organometallic vapor phase epitaxy method. After the above cleaning, the oxide film to be measured 14 was formed by the ALD method. Here, the oxide film to be measured 14 was Al2O3, HfO, as described above. 2-δ , and Hf 0.57 Si 0.43 O 2-δ These were evaluated as three types. Subsequently, an electron beam deposition method was used to form a conductive film 15 made of Pt with a thickness of 100 nm on the oxide film 14 to be measured, and a conductive layer 16 made of Pt (100 nm) / Ti (20 nm) on the lower side of the semiconductor substrate 11. As a final step, some samples underwent Post-Metalization Annealing (PMA) at 300°C for 5 minutes under N2 flow. The resulting aqueous sample 101 was very compact, measuring approximately 150 μm in diameter and 300 μm in height.

[0036] <Method for evaluating electrical characteristics> Sample 101 was placed in a measurement chamber made of stainless steel, and its electrical properties were evaluated using a tungsten probe. The measurement chamber containing the sample is evacuated by a dry scroll vacuum pump, and then flows at room temperature (25°C) at a total pressure of 10.0 kPa (100 mL·min). -1 We introduced N2 and evaluated the CV electrical characteristics under an inert gas environment. Subsequently, a hydrogen-containing gas consisting of 1 vol% H2 and 99 vol% N2 was flowed at a total pressure of 10.0 kPa for 30 minutes, replacing the N2. Then, CV measurements were performed in the hydrogen-containing gas (1 vol% H2 + 99 vol% N2) environment. The flow rate of the hydrogen-containing gas was 100 mL·min. -1 That is the case. Capacitance was determined from CV measurements at 100 kHz. For the bias dependence measurement of the CV measurement, the bias voltage step was set to 0.1V, and a sweep measurement was performed with each step held for 1 second. For the measurement of capacitance at a bias of 0V, the capacitance was determined by applying an AC current at a frequency of 100 kHz without applying a bias voltage or sweeping.

[0037] <Measurement and Evaluation> The conditions of the evaluated samples are summarized in Table 1.

[0038] [Table 1]

[0039] Figure 3 shows the results of CV measurements performed on samples A2, B1, and C2. The measurement results in an environment with 1 vol% hydrogen gas added are shifted in the ΔV negative direction compared to the measurement results in a nitrogen gas environment. This is based on the mechanism described above, in which the presence of hydrogen gas causes atomic hydrogen to be generated by the catalytic action of the conductive film 15, and the generated atomic hydrogen is trapped in the oxide film 14 under test, causing the oxide film 14 to acquire an electric charge. Furthermore, even without performing CV measurements with a voltage sweep, atomic hydrogen can be detected by monitoring the change in capacity value before and after hydrogen introduction at 0V, for example, without applying a bias, and thus the amount of oxygen deficiency can be evaluated.

[0040] Next, Figure 4 shows the results of investigating the responsiveness of this method for samples A2, B1, and C2. Note that the bias voltage was fixed at 0V during this measurement. The oxide film 14 under test was Al2O3(A2) and HfO 2-δ (B1) takes approximately 3 minutes, Hf 0.57 Si 0.43 O 2-δ In the case of (C2), the capacitance saturated in about 1 minute, indicating that the measurement method of the present invention has excellent responsiveness.

[0041] Figure 5 compares the capacitance of each sample in a nitrogen gas environment and a 1 vol% hydrogen gas environment when the bias voltage is fixed at 0V. The larger the measured value in the hydrogen gas environment compared to the measured value in the nitrogen gas environment, the more atomic hydrogen is detected, and therefore the greater the oxygen deficiency. As a result, the capacitance was larger in the hydrogen gas environment for all samples, indicating that each oxide film 14 under test has an oxygen deficiency, and in particular, the Hf oxide film 14 under test that was treated with PDA at 800°C 0.57 Si 0.43 O 2-δ Sample (C2) was shown to have an extremely high oxygen deficiency.

[0042] Furthermore, as shown in Figure 5, using Al2O3 as the metal oxide film 14 for samples A1 and A2 is better than using HfO for samples B1 and B2. 2-δ When using [a different method], the change in electrical capacitance from the nitrogen gas environment is larger in the hydrogen gas environment, making it appear at first glance that the oxygen deficiency is large. However, this is an apparent effect due to the measurement being taken at 0V, and HfO 2-δ The oxygen deficiency is greater in Al2O3 than in Al2O3.

[0043] This is for the following reasons: Figure 5 shows the change in volume at 0V (from a nitrogen environment to a hydrogen-added environment) in Figure 3. As is clear from Figure 3(a), when Al2O3 is used, 0V happens to be in a region where the capacitance change is large. On the other hand, as is clear from Figure 3(b), HfO 2-δRegarding this, 0V corresponds to a region where the change in capacitance is small. This is HfO 2-δ In this case, the CV curve is shifted in the positive voltage direction compared to Al2O3, and not because the oxygen deficiency is large. In fact, when the change in capacity is converted to a voltage change (ΔV), as is clear from Table 2, the value is HfO 1-δ It is larger than Al2O3. Based on the above, in Figure 5, Al2O3 is better than HfO 2-δ The larger change is merely an apparent one; the actual voltage change (ΔV) and the amount of intra-film charge calculated from that value are HfO 2-δ The latter is larger. Therefore, the oxygen deficiency is HfO 2-δ There is more of it than Al2O3.

[0044] Table 2 summarizes the CV measurement results and the dielectric constant of the oxide film 14 under test. Table 2 also includes the relative charge density in the oxide film calculated from these results. Figure 6 illustrates this relative charge density. Samples B1, B2, C1, and C2, which are hafnium-containing oxide films, exhibit high relative charge densities, particularly those treated with 800°C PDA. 0.57 Si 0.43 O 2-δ It can be seen that sample (C2) has a particularly high relative charge density.

[0045] [Table 2]

[0046] To measure the initialization characteristics of sample 101, i.e., its recovery characteristics to its initial state, the change in capacitance over time was measured after placing it in a hydrogen gas environment and then switching to a dry air environment. The results are shown in Figure 7. Here, the bias voltage was fixed at 0V and the measurements were taken at room temperature (25°C). The samples measured were B1, B2, C1, and C2.

[0047] HfO without PDA processing 2-δSample B1, with the oxide film 14 being measured, recovered to the same capacitance as the initial state measured in a nitrogen environment after 10 minutes, indicating that it has excellent recovery characteristics. 2-δ Sample B2, which was subjected to PDA treatment at 800°C, also exhibits good recovery characteristics, though not as good as those of sample B1. Also, Hf without PDA processing 0.57 Si 0.43 O 2-δ Sample C1, with the oxide film 14 being measured, recovered to almost the same capacitance as the initial state measured in a nitrogen environment after 10 minutes, indicating that it has excellent recovery characteristics. On the other hand, Hf 0.57 Si 0.43 O 2-δ Sample C2, which was subjected to PDA treatment at 800°C, did not exhibit sufficient recovery characteristics.

[0048] [Table 3] [Industrial applicability]

[0049] Evaluating the film quality of oxide films is essential for improving the device performance, reliability, and quality control of semiconductor devices such as MOSFETs. The film quality of an oxide film depends greatly on the formation method, conditions, and environment. Therefore, it is desirable to evaluate the film quality of oxide films deposited under the same conditions as those used in the device. The present invention provides a method for evaluating the film quality of oxide films, which is difficult to measure, and allows for the evaluation of films deposited under the same conditions as those used in devices. Therefore, we believe that the method for evaluating the film quality of oxide films according to the present invention will greatly contribute to improving the device performance, reliability, and quality control of semiconductor elements. [Explanation of symbols]

[0050] 11: Semiconductor substrate (n + -GaN) 12: Semiconductor layer (n-GaN) 13: Semiconductor layer 14: Oxide film to be measured 15: Conductor film (Pt) 16: Conductor layer (Pt / Ti) 101: Sample

Claims

1. Prepare a sample in which a conductive layer, a semiconductor layer, an oxide film to be measured, and a conductive film consisting of a hydrogen catalyst metal or hydrogen catalyst alloy are sequentially laminated, (1) C-V measurement, or (2) measurement of capacitance at bias 0V, to detect and quantify oxygen deficiencies in the oxide film under test. It includes, The above (1) detection and quantification by C-V measurement involves performing C-V measurements on the sample under both an inert gas environment and a hydrogen gas environment, and detecting and quantifying the oxygen deficiency in the oxide film under measurement by detecting and quantifying the hydrogen gas trapped in the oxygen-deficient portion from the voltage shift amount between the C-V characteristic curves. The above (2) detection and quantification by measuring capacitance at a bias of 0V means measuring the capacitance of the sample at a bias of 0V in both an inert gas environment and a hydrogen gas environment, and detecting and quantifying the oxygen deficiency of the oxide film to be measured from the amount of change in the capacitance value, a method for evaluating the film quality of an oxide film.

2. The method for evaluating the film quality of an oxide film according to Claim 1, wherein the volume concentration of the hydrogen atmosphere in the hydrogen gas environment is 100 ppm or more and 100% or less.

3. The method for evaluating the film quality of an oxide film according to claim 1 or 2, wherein the partial pressure of hydrogen in the hydrogen atmosphere under the hydrogen gas environment is 1 kPa or more and 10 atmospheres or less.

4. The method for evaluating the film quality of an oxide film according to claim 1, wherein the conductive film comprises one or more metals selected from the group consisting of platinum, palladium, iridium, ruthenium, and nickel.

5. The method for evaluating the film quality of an oxide film according to claim 1, wherein the thickness of the conductive film is 5 nm or more and 1 μm or less.

6. The semiconductor layer is composed of GaN, Si, GaAlN, GaAs, ZnO, SiC, and Ga ２ O ３ A method for evaluating the film quality of an oxide film according to claim 1, comprising one or more selected from the group consisting of the following.

7. The method for evaluating the film quality of an oxide film according to claim 1, wherein the thickness of the semiconductor layer is 100 nm or more and 1 mm or less.

8. The method for evaluating the film quality of an oxide film according to claim 1, wherein the thickness of the oxide film to be measured is 0.1 nm or more and 1 μm or less.

9. The method for evaluating the film quality of an oxide film according to claim 1, further comprising, following the detection and quantification, applying an electric field to the oxide film to be measured to restore the sample to its initial state.

10. The method for evaluating the film quality of an oxide film according to claim 1, further comprising, following the detection and quantification, applying a bias voltage between the conductive layer and the conductive film to restore the sample to its initial state.

11. The method for evaluating the film quality of an oxide film according to claim 10, wherein the bias voltage is -20V or more and 10V or less.

12. The method for evaluating the film quality of an oxide film according to claim 1, wherein the detection and quantification by C-V measurement described above (1) is performed in a temperature environment of 23°C to 200°C.

Images