Electron spectrometer and analysis method
The electron spectrometer with multiple detection units and optimized energy sweeping reduces measurement time by integrating or averaging spectra, addressing the inefficiency of single-unit detection systems.
Patent Information
- Application Number
- JP2022097319
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2022-06-16
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2042-06-16
AI Technical Summary
Existing electron spectrometers require multiple measurements over a range of energies to collect a spectrum, leading to prolonged measurement times due to the use of a single detection unit.
An electron spectrometer with a detector comprising multiple detection units arranged in the energy dispersion direction, allowing for simultaneous detection of electrons with different energies, and a processing unit that sweeps energy in discrete steps to acquire and integrate or average spectra, reducing the number of measurements needed.
This approach significantly shortens the measurement time required to collect spectra by minimizing the number of measurements and reducing the influence of detection sensitivity variations, while maintaining spectral accuracy.
Smart Images

Figure 0007586856000020 
Figure 0007586856000021 
Figure 0007586856000022
Abstract
Description
[Technical field]
[0001] The present invention relates to an electron spectroscopic device and an analytical method. [Background technology]
[0002] In electron spectroscopy instruments such as X-ray photoelectron spectroscopy and Auger electron spectroscopy, electrons are separated by energy using a spectrometer, allowing the detection of only electrons with a specific energy. The measurement energy is selected by controlling the electron lens equipped in the spectrometer. An energy spectrum can be collected by repeating measurements while sweeping the measurement energy.
[0003] As a detector to be mounted on an electron spectrometer, Patent Document 1 discloses a detector in which a plurality of channeltrons are arranged in the energy dispersion direction of electrons whose energy has been dispersed by a spectrometer. In such an electron spectrometer, generally, when a spectrum is collected, the electrons dispersed by the spectrometer are detected by one specific channeltron. [Prior art documents] [Patent documents]
[0004] [Patent Document 1] JP 2001-312994 A Summary of the Invention [Problem to be solved by the invention]
[0005] As described above, when detecting electrons with one specific channeltron, in order to collect a spectrum, it is necessary to repeatedly perform measurements while sweeping the measurement energy, which takes a long time to collect the spectrum. [Means for solving the problem]
[0006] One aspect of the electron spectroscopy apparatus according to the present invention is a spectrometer for energy-spectroscopically analyzing electrons emitted from the sample; a detector including a plurality of detection units arranged side by side in an energy dispersion direction of the electrons energy-split by the spectrometer; A processing section; Including, The processing unit includes: a process of acquiring a plurality of first spectra obtained by sweeping energy in a first energy step in the spectrometer and detecting the electrons energy-spectroscopically separated by the spectrometer with the plurality of detection units; A process of complementing a measurement point in each of the plurality of first spectra; The plurality of first spectra with the measurement points supplemented By integrating or averaging generating a second spectrum having a second energy step smaller than the first energy step; I'll do it stomach , the processing unit sweeps the energy in the spectrometer at the first energy step to obtain the first spectrum at the first energy step for each of the detection units; The processing unit sets the first energy step so that the measured energies of the plurality of first spectra do not coincide with each other. .
[0007] In such an electron spectrometer, the number of measurements can be reduced compared to a case where electrons are detected by one specific detection unit, and therefore the measurement time required to collect a spectrum can be shortened.
[0008] One aspect of the analysis method according to the present invention is to a spectrometer for energy-spectroscopically analyzing electrons emitted from the sample; a detector including a plurality of detection units arranged side by side in an energy dispersion direction of the electrons energy-split by the spectrometer; An analysis method using an electron spectrometer, comprising: a step of sweeping energy in a first energy step in the spectrometer, and detecting the electrons energy-spectroscopically separated by the spectrometer with the plurality of detection units to obtain a plurality of first spectra; complementing measurement points in each of the plurality of first spectra; The plurality of first spectra with the measurement points supplemented By integrating or averaginggenerating a second spectrum having a second energy step smaller than the first energy step; Including fruit , In the step of acquiring a plurality of first spectra, Sweeping energy in the spectrometer at the first energy step to obtain the first spectrum at the first energy step for each of the detection units; The first energy step is set so that the measured energies of the plurality of first spectra do not coincide with each other. .
[0009] In this analysis method, the number of measurements can be reduced compared to, for example, a case where electrons are detected by one specific detection unit, and therefore, the measurement time required to collect a spectrum can be shortened. [Brief description of the drawings]
[0010] [Figure 1] FIG. 1 is a diagram showing the configuration of an electron spectroscopy apparatus according to an embodiment of the present invention. [Diagram 2] FIG. 2 is a diagram showing a schematic diagram of a detector. [Diagram 3] FIG. 2 is a diagram for explaining a function of a detector. [Figure 4] 1 is a graph for explaining the relationship between the energy step ΔE1 and the energy resolution. [Diagram 5] Graph showing the measured energy of seven channeltrons. [Figure 6] Graph showing the measured energy of seven channeltrons. [Figure 7] A graph showing the case where ΔE1 = f(Em), where f(Em) = αEm. [Figure 8] A graph showing the case where ΔE1 = f(Em), where f(Em) = max(αEm, β). [Figure 9] FIG. 13 is a diagram showing an example of a table of β and d. [Figure 10] FIG. 1 shows the measured energy of each channeltron of the detector. [Figure 11] A diagram showing seven spectra. [Figure 12] FIG. 13 is a diagram for explaining a process of complementing measurement points in each of seven spectra. [Figure 13] Graph showing the results of integrating seven spectra with linearly interpolated measurement points. [Figure 14] FIG. 13 is a diagram showing a case where the measured energies of the spectra do not match. [Figure 15] FIG. 13 is a diagram showing a case where the measured energies of the spectra are the same. [Figure 16] FIG. 13 is a diagram for explaining the S / N ratio of an actually measured measurement point and the S / N ratio of an interpolated measurement point. [Figure 17] 10 is a flowchart showing an example of processing by a processing unit. [Figure 18] A graph showing the calculation results of the noise level when the measurement step is M times the differential step. [Figure 19] A graph showing Dσm normalized by measurement time when M=1. [Figure 20] A graph showing Dσm normalized by measurement time when M=1. [Figure 21] A graph comparing the cases where the energy shift during differentiation is zero and (1 / 2) ΔE. [Figure 22] Graph showing the results of averaging Dσm. [Figure 23] Graph showing resolution function. [Figure 24] FIG. 13 is a graph showing a resolution function when the measurement step is changed. [Diagram 25] Graph showing the seven-point differentiation of the resolution function when the measurement step is changed. [Figure 26] Graph showing the resolution function and its derivative when the measurement start energy is shifted by +2 eV. [Figure 27] Graph showing the resolution function and its derivative when the measurement start energy is shifted by +4 eV [Figure 28] Graph showing the resolution function and its derivative when the measurement start energy is shifted by +6 eV [Figure 29] A graph showing the results of accumulating seven spectra. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0011] Preferred embodiments of the present invention will be described in detail below with reference to the drawings. Note that the embodiments described below do not unduly limit the contents of the present invention described in the claims. In addition, not all of the configurations described below are necessarily essential components of the present invention.
[0012] In the following, an Auger electron spectroscopy device will be described as an example of the electron spectroscopy device according to the present invention, but the electron spectroscopy device according to the present invention is not limited to this. For example, the electron spectroscopy device according to the present invention may be an X-ray photoelectron spectroscopy device.
[0013] 1. Electron spectrometer First, an electron spectroscopy apparatus according to an embodiment of the present invention will be described with reference to the drawings. Fig. 1 is a diagram showing the configuration of an electron spectroscopy apparatus 100 according to an embodiment of the present invention.
[0014] The electron spectrometer 100 is an apparatus that performs analysis of a sample by Auger electron spectroscopy, which is a method of performing elemental analysis by measuring the energy of Auger electrons emitted from a sample upon excitation by an electron beam or the like.
[0015] As shown in FIG. 1, the electron spectrometer 100 includes an electron beam irradiation device 10, a sample stage 20, a spectrometer 30, a detector 40, an irradiation control device 50, a spectrometer control device 52, a counting calculation device 54, and a processing unit 60.
[0016] The electron beam irradiation device 10 irradiates an electron beam onto a sample S. The electron beam irradiation device 10 includes an electron gun 12, an electron lens 14, and a deflector 16.
[0017] The electron gun 12 emits an electron beam. The electron lens 14 focuses the electron beam emitted from the electron gun 12. The deflector 16 deflects the electron beam focused by the electron lens 14. The deflector 16 allows the electron beam to be irradiated at a desired position. The deflector 16 also allows the sample S to be scanned with the electron beam.
[0018] The sample stage 20 is capable of holding the sample S and moving the sample S. The sample stage 20 may be equipped with, for example, a mechanism (goniometer) for tilting the sample S. The sample stage 20 can position the sample S.
[0019] The spectrometer 30 performs energy dispersion of Auger electrons generated from the sample S when the sample S is irradiated with an electron beam. The spectrometer 30 includes an input lens 32 and an electrostatic hemispherical analyzer .
[0020] The input lens 32 collects the incident electrons and guides them to the electrostatic hemispherical analyzer 34. The input lens 32 makes the energy resolution variable by decelerating the electrons. The input lens 32 is composed of, for example, a plurality of electrostatic lenses 33.
[0021] The electrostatic hemispherical analyzer 34 has an inner hemispherical electrode 35a and an outer hemispherical electrode 35b. In the electrostatic hemispherical analyzer 34, by applying a voltage between the inner hemispherical electrode 35a and the outer hemispherical electrode 35b, it is possible to extract electrons in an energy range according to the applied voltage. Cut.
[0022] The detector 40 detects the electrons whose energies have been separated by the spectrometer 30 .
[0023] Fig. 2 is a schematic diagram of a detector 40. The detector 40 has a plurality of channeltrons 42 (an example of a detection section). In the example shown in Fig. 2, the detector 40 has seven channeltrons 42. Here, a case will be described in which the detector 40 has seven channeltrons 42, but the number of channeltrons 42 is not particularly limited as long as it is two or more. The channeltrons 42 are detectors that detect electrons and output an amplified signal.
[0024] The seven channeltrons 42 are aligned in the energy dispersion direction A on the exit surface, i.e., the energy dispersion surface, of the electrostatic hemispherical analyzer 34. Therefore, the seven channeltrons 42 can detect electrons of different energies from each other. Therefore, the detector 40 can simultaneously detect electrons of different energies.
[0025] FIG. 3 is a diagram for explaining the function of the detector 40. As shown in FIG.
[0026] As shown in Fig. 3, the seven channeltrons 42 are assigned channel numbers from -3ch to +3ch. Specifically, the channeltron 42 that detects electrons passing through the center between the inner hemispherical electrode 35a and the outer hemispherical electrode 35b of the electrostatic hemispherical analyzer 34 is 0ch. In addition, the channeltron 42 of -1ch, the channeltron 42 of -2ch, and the channeltron 42 of -3ch are arranged in order from the channeltron 42 of 0ch toward the inner hemispherical electrode 35a. In addition, the channeltron 42 of +1ch, the channeltron 42 of +2ch, and the channeltron 42 of +3ch are arranged in order from the channeltron 42 of 0ch toward the outer hemispherical electrode 35b.
[0027] The seven Channeltrons 42 are arranged so that the measured energy difference D between adjacent channels is the same. Therefore, for example, when the measured energy of Channeltron 42 for 0ch is set to E0, the measured energy of -3ch is E0-3D, the measured energy of -2ch is E0-2D, the measured energy of -1ch is E0-D, the measured energy of +1ch is E0+D, the measured energy of +2ch is E0+2D, and the measured energy of +3ch is E0+3D.
[0028] In the above, the detector 40 is described as being composed of seven channeltrons 42, but the configuration of the detector 40 is not limited to this. For example, a microchannel plate, a multi-anode, a CMOS camera, or the like may be used as the detector 40. When using these, a desired number of channels may be provided in an artificial manner by combining multiple pixels into one pixel.
[0029] The irradiation controller 50 controls the electron beam irradiation device 10. The irradiation controller 50 controls the electron beam irradiation device 10 based on a control signal from the processing unit 60, for example, so that the electron beam is irradiated onto a predetermined position on the sample S.
[0030] The spectrometer control device 52 controls the spectrometer 30. Based on a control signal from the processing unit 60, the spectrometer control device 52 controls the deceleration rate of the electrons in the input lens 32 and the voltage applied between the inner hemisphere electrode 35a and the outer hemisphere electrode 35b.
[0031] The counting operation device 54 counts the electrons detected by each channeltron 42. The counting operation device 54 sends the counting results of the electrons detected by each channeltron 42, i.e., the detection results of the electrons, to a processing unit 60. The processing unit 60 calculates the number of electrons counted by each channeltron 42. Therefore, the processing unit 60 can obtain an energy spectrum (hereinafter, simply referred to as a "spectrum") for each channeltron 42. That is, the processing unit 60 can simultaneously obtain spectra according to the number of channeltrons 42.
[0032] The processing unit 60 performs processes such as generating an Auger electron spectroscopy spectrum based on the electron detection result in the detector 40 sent from the counting and calculation unit 54, and controlling the irradiation control unit 50 and the spectrometer control unit 52. The processes performed by the processing unit 60 will be described in detail later.
[0033] The processing unit 60 includes, for example, a processing circuit such as a CPU (Central Processing Unit) or an FPGA (Field Programmable Gate Array), and a storage circuit such as a semiconductor memory. In the processing unit 60, the processing circuit executes a program stored in the storage circuit, thereby performing various calculation processes and various control processes.
[0034] 2. Spectral collection 2.1. Specifying the energy resolution In the electron spectrometer 100, an energy spectrum can be collected by repeating measurements while sweeping the measurement energy in the spectrometer 30. A method for collecting a spectrum in the electron spectrometer 100 will be described below.
[0035] First, the user sets the measurement conditions. The user sets the measurement conditions by operating a setting unit (such as a user interface) of the electron spectrometer 100. The processing unit 60 accepts the set measurement conditions. The set measurement conditions are reflected in the processing of the processing unit 60.
[0036] The measurement conditions include the selection of the mode, the start energy, the end energy, and the energy resolution. The mode is selected from the CAE (Constant Analyzer Energy) mode and the CRR (Constant Retarding Ratio) mode. The details of the modes are described later. The start energy is the measurement energy at which the measurement starts, and the end energy is the measurement energy at which the measurement ends. That is, in the measurement, a spectrum can be obtained in the range from the start energy to the end energy. The energy resolution is the resolution of the spectrum, and is indicated, for example, by the half-width (full width at half maximum) of the peak that appears in the spectrum. For example, the energy resolution is set low for elemental analysis to know the elements that constitute the sample, and set high for chemical state analysis.
[0037] 2.2. Setting the energy step ΔE1 during measurement The processing unit 60 sets the energy step ΔE1 during measurement based on the energy resolution. The electron spectrometer 100 performs measurement by sweeping the energy with the energy step ΔE1. For example, in the example shown in FIG. 3, if the measurement energy of 0ch in the first measurement is E0, the measurement energy of 0ch in the second measurement is E0+ΔE1, and the measurement energy of 0ch in the third measurement is E0+2ΔE1. Therefore, the interval between the measurement points of the spectrum obtained for each channeltron 42 is the energy step ΔE1.
[0038] FIG. 4 is a graph for explaining the relationship between the energy step ΔE1 and the energy resolution.
[0039] The energy resolution can be expressed, for example, by the full width at half maximum (FWHM). In the graph shown in FIG. 4, the energy resolution at around 2000 eV is FWHM=12 eV. If the energy step ΔE1 is sufficiently small for the waveform with a maximum value of 1.0 shown in FIG. 4, a waveform with a maximum value of 1.0 is obtained as the measurement result. However, for example, When the loop ΔE1 is 6 eV, the measurement result does not necessarily have a waveform with a maximum value of 1.0.
[0040] For example, when the energy step ΔE1=6 eV, the following two patterns are obtained as the measurement energy.
[0041] Pattern 1: 1970, 1976, 1982, 1988, 1994, 2000, 2006, 2012, 2018, 2024, 2030 Pattern 2: 1973, 1979, 1985, 1991, 1947, 2003, 2009, 2015, 2021, 2027, 2033
[0042] For pattern 1, the measurement results are a waveform with a maximum value of 1.0 and FWHM = 12 eV. For pattern 2, the measurement results are a waveform with a maximum value of 0.8 and FWHM = 12 eV. For points that were not measured, linear interpolation was used to determine the maximum value and FWHM.
[0043] In this way, even if the energy step ΔE1 is the same, the degree of distortion of the spectrum changes depending on the relationship between the measurement energy and the peak position. As mentioned above, when the energy step ΔE1 is half the FWHM, the maximum value changes by 20%, so the energy step ΔE1 must be made small.
[0044] Here, the measured energy difference D between channels is proportional to the energy resolution. Therefore, calculating the energy step ΔE1 from the energy resolution is essentially the same as calculating the energy step ΔE1 from the measured energy difference D between channels. Therefore, here, the case of calculating the energy step ΔE1 from the measured energy difference D between channels will be described.
[0045] Figures 5 and 6 are graphs showing the measured energies of seven channeltrons. In Figures 5 and 6, the channels are arranged in the following order from low energy to high energy: -3ch, -2ch, -1ch, 0ch, +1ch, +2ch, and +3ch.
[0046] 5 shows the case where the energy step ΔE1 is equal to the measured energy difference D between the channels (ΔE1=D), and FIG 6 shows the case where the energy step ΔE1 is 1 / 2 the measured energy difference D between the channels (ΔE1=D / 2).
[0047] The energy step ΔE1 is expressed by the following equation using the measured energy difference D between the channels.
[0048] ΔE1=max(β,α·c·(E p +γ E m )) α = d (1 + 0.5 d c γ) β: Lower limit of energy step ΔE1 α: proportionality coefficient of energy step ΔE1 c (E p +γ E m ): Measured energy difference between channels D c: Ratio of the energy of the electrons passing through the spectrometer to the measured energy difference D between the channels E p : Pass energy in CAE mode (energy of electrons passing through the spectrometer) γ: Deceleration rate of the input lens in CRR mode E m :Measured energy d: Ratio of measured energy difference D between channels and energy step ΔE1
[0049] Note that γ is the deceleration rate of the input lens in CRR mode. For example, When the energy of the electrons becomes 1 / 2 after passing through the lens, γ=1 / 2. For example, when ΔE1=D as shown in FIG. 5, the ratio d becomes d=1, and when ΔE1=D / 2 as shown in FIG. 6, the ratio d becomes d=1 / 2.
[0050] The electron spectrometer 100 is capable of performing measurements in CAE mode and CRR mode.
[0051] In the CAE mode, the energy of the electrons passing through the spectrometer 30 is constant, i.e., the pass energy is constant, regardless of the energy of the electrons emitted from the sample S. In the CAE mode, the potential difference applied between the inner hemisphere electrode 35a and the outer hemisphere electrode 35b is kept at a constant value, and the voltage applied to the input lens 32 is swept. In the CAE mode, the energy resolution is constant over the entire measurement energy range.
[0052] In contrast, in the CRR mode, the electrons are decelerated at a constant ratio according to the kinetic energy of the electrons to be measured. In the CRR mode, the potential difference applied between the inner hemisphere electrode 35a and the outer hemisphere electrode 35b is swept together with the voltage applied to the input lens 32, and the electrons are dispersed at a constant deceleration ratio. In the CRR mode, the energy resolution is proportional to the measurement energy.
[0053] In the CAE mode, γ = 0. Therefore, in the CAE mode, the energy step ΔE1 is expressed by the following equation (1).
[0054] ΔE1=max(β,α·c·E p ) ···(1) α=d
[0055] In CAE mode, β is a fixed value and α c E p is also a fixed value. Therefore, the energy step ΔE1 is fixed.
[0056] In CRR mode, E p = 0. Therefore, in the CRR mode, the energy step ΔE1 is expressed by the following equation:
[0057] ΔE1=max(β,α c γ E m ) ···(2) α = d (1 + 0.5 d c γ)
[0058] In CRR mode, β is a fixed value, α c γ E m is proportional to the measured energy. Therefore, the energy step ΔE1 is expressed by the following equation (3).
[0059] ΔE1=f(E m ) ···(3) f(E m )=max(αE m ,β) (4) However, α in equation (4) is α = α c γ, and the α on the right hand side of this equation is the α in equation (2).
[0060] Figure 7 shows the relationship between ΔE1=f(E m ), where f(E m ) = αE m FIG. 8 is a graph showing the case where ΔE1=f(E m ), where f(E m )=max(αE m , β).
[0061] As shown in Figure 7, f(E m ) = αE m In other words, the energy step ΔE1 is set to α and E m When the measured energy E is m When is small, the energy step ΔE1 is also small, and the measurement energy E m When is large, the energy step ΔE1 is also large. Therefore, the measurement energy E m When is close to zero, the energy step ΔE1 becomes very small.
[0062] Therefore, as shown in Figure 8, f(E m )=max(αE m , β), and the energy step ΔE1 is α and E m The product of α and E m If the product of is smaller than β, ΔE1 = β. This allows the measured energy E m This solves the problem that the energy step ΔE1 becomes very small when is close to zero.
[0063] In CRR mode, the measured energy E m Therefore, the measurement energy E m is also proportional to the measured energy difference D between the channels. Therefore, when the ratio d is set, α is calculated using the correction term (1 + 0.5 d c γ) for the measured energy difference D between the channels.
[0064] 2.3. Table for calculating ΔE1 In the above formula for ΔE1, c, E p , γ, Em has a set value. On the other hand, values must be set for β, α, and d. Here, α can be calculated from the formula α=d·(1+0.5·d·c·γ), so the parameters to be determined in order to calculate ΔE1 are β and d.
[0065] For example, β and d are determined by preparing a table. Fig. 9 is a diagram showing an example of a table of β and d. Fig. 9 shows a table for performing qualitative analysis with γ=0.2 in CRR mode.
[0066] In the table shown in Figure 9, the values of β and d are set so that the change in the peak intensity of the spectrum is about 5%. When performing quantitative analysis, it is desirable to set the values of β and d small so that the change in peak intensity is approximately the same as or less than the S / N ratio. In this way, by changing the values of β and d according to the purpose of measurement, it is possible to perform measurements with an energy step ΔE1 that suits the purpose of measurement.
[0067] A plurality of tables for setting β and d are prepared according to the measurement purpose, energy resolution, and measurement mode (CAE mode, CRR mode). The processing unit 60 refers to the table according to the measurement purpose, energy resolution, and measurement mode, sets the values of β and d, and calculates the energy step ΔE1.
[0068] Measurement In the electron spectrometer 100, an electron beam emitted from an electron gun 12 is focused by an electron lens 14 and irradiated onto a sample S. Auger electrons, secondary electrons, etc. are emitted from the sample S irradiated with the electron beam.
[0069] The Auger electrons emitted from the sample S enter the input lens 32, are decelerated by the electrostatic lens 33, and enter the electrostatic hemispherical analyzer 34. The incident Auger electrons are energy-dispersed by the electrostatic hemispherical analyzer 34, and are dispersed in the energy dispersion direction A on the energy dispersion plane of the electrostatic hemispherical analyzer 34 according to their energy (kinetic energy).
[0070] The Auger electrons dispersed according to energy are detected by seven channeltrons 42 aligned in the energy dispersion direction A. The electrons detected by the seven channeltrons 42 are counted by a counting calculation device 54 for each channeltron 42, and the counting results are sent to a processing unit 60.
[0071] When collecting a spectrum, the energy is swept in the spectrometer 30 at an energy step ΔE1 while the dispersed electrons are detected by the seven channeltrons 42, and this process is repeated.
[0072] FIG. 10 is a diagram showing the measured energy of each channeltron 42 of the detector 40.
[0073] For example, in the first measurement, n=1, the energy E A Electrons are detected and have energy E A An electron of -D is detected and the energy E A -2D electrons are detected and -3ch has energy E A An electron in -3D is detected and the energy E A +D electron is detected, and energy E A An electron with +2D is detected and energy E A A +3D electron is detected. In the second measurement, n=2, the energy E B Electrons are detected and have energy E B An electron of -D is detected and the energy E B -2D electrons are detected and -3ch has energy E B An electron in -3D is detected and the energy E B +D electron is detected, and energy E B An electron with +2D is detected and energy E B +3D electrons are detected, where the energy E B and energy E A The difference between these is the energy step ΔE1. That is, EB =E A +ΔE1. In addition, the energy E detected at 0ch in the third measurement n=3 is C E C =E B +ΔE1=E A +Δ2E1. In addition, the energy E detected at 0ch in the fourth measurement n=4 is D E D =E C +ΔE1=E A +Δ3E1.
[0074] In this manner, the measurement is repeated while sweeping the energy in energy steps ΔE1 in the spectrometer 30, and a spectrum is collected for each channel. The electron spectrometer 100 can obtain seven spectra corresponding to the seven channels.
[0075] Figure 11 shows seven spectra, plotting the measured energy and the measured intensity for each channel.
[0076] In the electron spectrometer 100, the measurement energy is swept in energy steps ΔE1, so each spectrum is discrete data. The energy step of one spectrum, i.e., the measurement energy difference between adjacent measurement points, is ΔE1. Note that, although an example of the CAE mode has been described here, in the CRR mode, the measurement energy difference D between channels and the energy step ΔE1 depend on the measurement energy.
[0077] 2.4. Setting the energy step ΔE2 In the electron spectrometer 100, an energy step ΔE2 during spectrum generation is set separately from the energy step ΔE1 during measurement, which allows optimal energy steps to be set for each of measurement and spectrum generation.
[0078] In the electron spectrometer 100, the energy step ΔE2 is set to be smaller than the energy step ΔE1 (ΔE2<ΔE1). For example, the energy step ΔE2 is set to be approximately 1 / 10 to 1 / 2 of the energy step ΔE1. In other words, the energy step ΔE2 is set to a value that satisfies (1 / 10)×ΔE1≦ΔE2≦(1 / 2)×ΔE1. By setting the energy step ΔE2 in such a range, it is possible to balance the interpolation error and the amount of data.
[0079] Furthermore, in the CRR mode, the energy step ΔE1 changes depending on the measurement energy, but even when the measurement is performed in the CRR mode, the energy step ΔE2 can be constant regardless of the measurement energy.
[0080] Completion FIG. 12 is a diagram for explaining the process of complementing the measurement points in each of the seven spectra.
[0081] In each of the seven spectra, the measurement points between the actually measured measurement points are interpolated by calculation. In the example shown in FIG. 12, the measurement points between the actually measured measurement points are connected by straight lines for linear interpolation. The method for interpolating the measurement points is not particularly limited, and the measurement points may be interpolated using a method other than linear interpolation, such as spline interpolation. By performing linear interpolation, the energy step of each spectrum can be changed from energy step ΔE1 to energy step ΔE2. In each spectrum shown in FIG. 12, the measurement points are made continuous by linear interpolation, so the energy step ΔE2=0.
[0082] 2.6. Estimation FIG. 13 is a graph showing the results of integrating seven spectra whose measurement points have been linearly interpolated.
[0083] As shown in FIG. 13, one spectrum (multispectrum) can be generated by accumulating seven spectra whose measurement points have been linearly interpolated.
[0084] As shown in Fig. 11, the channeltron 42 has a variation in detection sensitivity. As shown in Fig. 13, the effect of the variation in detection sensitivity can be reduced by integrating seven spectra in which the measurement points are linearly interpolated. Therefore, in the electron spectrometer 100, it is not necessary to correct the variation in detection sensitivity of each channeltron 42.
[0085] In the above, a case has been described in which the measurement points are linearly complemented to make them continuous and the energy step ΔE2=0 is set, but as long as the energy step ΔE2=0 can be calculated, a multispectrum can be generated by a similar method regardless of the value of the energy step ΔE2. For example, measurement points may be extracted from the spectrum shown in Fig. 13 according to the energy step ΔE2 to generate a multispectrum with a set energy step ΔE2. Also, for example, in each of the seven spectra with complemented measurement points, measurement points may be extracted so as to have an energy step ΔE2, and then the seven spectra with extracted measurement points may be integrated to generate a multispectrum with an energy step ΔE2.
[0086] Although the case where seven spectra with complemented measurement points are integrated to generate one multispectrum has been described here, one multispectrum may also be generated by averaging the seven spectra with complemented measurement points.
[0087] 2.7. Complementary calculations and ratios When accumulating the seven spectra shown in FIG. 12, it is desirable that the measured energies of the spectra do not match.
[0088] Fig. 14 is a diagram showing a case where the measured energies of the seven spectra do not match, and Fig. 15 is a diagram showing a case where the measured energies of the seven spectra match.
[0089] As shown in FIG. 15, when the measured energies of the spectra match, the energy range R of only the interpolated measurement points, not the actually measured measurement points, is wide.
[0090] In contrast, as shown in FIG. 14, when the measured energies of the spectra do not match, the actually measured measurement points are scattered, so that the energy range R of only the interpolated measurement points can be narrowed.
[0091] Here, the SN ratio of the intensity of the interpolated measurement point is different from the SN ratio of the actually measured measurement point. FIG. 16 illustrates the SN ratio of the actually measured measurement point and the SN ratio of the interpolated measurement point. This is a diagram for explaining the method.
[0092] If the intensity I1 at the actual measurement point follows a Poisson distribution, the signal-to-noise ratio of the intensity I1 is I1 1 / 2 Similarly, the signal-to-noise ratio of intensity I2 is I2 1 / 2 Calculating I3 by linear interpolation, α=(E3-E1) / (E2-E1), I3=(1-α)I1+αI2. Therefore, the signal-to-noise ratio of the intensity I3 is expressed as follows:
[0093]
number
[0094] In the interpolation, 0≦α≦1, and the SN ratio of the intensity value of the interpolated measurement point is equal to or greater than the SN ratio of the intensity value when it follows the Poisson distribution. In this way, the SN ratio changes due to the interpolation, so the energy step ΔE1 is set so that this effect is not concentrated at a specific measurement energy.
[0095] Here, when the ratio d is changed, the energy step ΔE1 changes. Therefore, the ratio d is set so that the measured energies of the spectra do not match as much as possible. For example, the ratio d is set to a value smaller than 1. When the ratio d=1, that is, when the energy step ΔE1 and the measured energy difference D between the channels are equal as shown in FIG. 15, the measured energies of the spectra match. On the other hand, by setting the ratio d to a value smaller than 1, the number of matching measured energies can be reduced compared to when the ratio d is 1, and the energy range R of only the interpolated measurement points can be narrowed.
[0096] For example, let the number of channeltrons 42 be N C Then, the ratio d is d=N C / (N C +1). This makes it possible to make the measured energies of the spectra different from each other. The ratio d is not limited to the above, and the ratio d can be set as d=N C / (N C +N N ) etc. However, N N is N C It is a natural number that is not a multiple of any divisor other than 1.
[0097] 3. Processing FIG. 17 is a flowchart showing an example of the processing of the processing unit 60.
[0098] First, the processing unit 60 sets the measurement conditions (S100).
[0099] The processing unit 60 accepts the user's selection of a mode, a start energy, an end energy, and an energy resolution.
[0100] The processing unit 60 sets the energy step ΔE1 based on the specified energy resolution. For example, the processing unit 60 acquires the values of β and d corresponding to the specified measurement mode and energy resolution from a table of β and d prepared for each measurement mode and energy resolution. In the CAE mode, the processing unit 60 calculates the energy step ΔE1 from the above-mentioned formula (1), and in the CRR mode, the processing unit 60 calculates the energy step ΔE1 from the above-mentioned formula (3).
[0101] As described above, it is preferable that the energy step ΔE1 is set so that the measured energies of the spectra do not match.
[0102] Next, the processing unit 60 sweeps the energy in the spectrometer 30 at an energy step ΔE1. The electrons separated by energy in the spectrometer 30 are detected by the seven channeltrons 42 to obtain seven spectra (S102). The processing unit 60 causes the spectrometer 30 to sweep the energy in the energy step ΔE1 set in step S100 via the spectrometer control device 52, and obtains the counting results of the electrons counted for each channeltron 42 from the counting calculation device 54. As a result, seven spectra can be obtained, as shown in FIG.
[0103] Next, the processing unit 60 sets an energy step ΔE2 (S104). The processing unit 60 sets the energy step ΔE2 to be smaller than the energy step ΔE1 (ΔE2<ΔE1). The processing unit 60 sets the energy step ΔE2 to an energy step that is optimal for displaying the generated spectrum or analyzing the spectrum, for example.
[0104] Next, the processing unit 60 performs a complementary calculation to complement the measurement points in each of the seven spectra (S106). The processing unit 60 complements the measurement points by, for example, linear complementation. As a result, the measurement points can be complemented in each of the seven spectra as shown in FIG.
[0105] Next, the processing unit 60 integrates the seven spectra with the interpolated measurement points to generate one spectrum (multispectrum) (S108), thereby generating a multispectrum as shown in FIG.
[0106] The processing unit 60 causes the display unit to display the generated multispectrum, and ends the process. Note that the processing unit 60 may also differentiate the generated multispectrum to generate a derivative spectrum, and cause the display unit to display the generated derivative spectrum.
[0107] 4. Energy step ΔE1 The larger the energy step ΔE1 during measurement, the shorter the measurement time and the smaller the noise. However, the larger the energy step ΔE1 during measurement, the more the spectrum peak is distorted. Thus, there is a trade-off between noise and distortion of the spectrum peak.
[0108] In the following, we will explain the noise and distortion of the spectrum peak when the ratio of the energy step ΔE1 to the energy step ΔE2 is changed. Here, we will explain the number of channeltrons as 7.
[0109] (1) Relationship between energy step and noise In Auger electron spectroscopy, differentiated spectra are used to reduce background effects.
[0110] When n is a natural number, the differentiation at (2n+1) points is expressed by the following equation.
[0111]
number
[0112] where N(E) is the electron intensity at energy E, D(E) is the derivative of the electron intensity at energy E, and ΔE is the energy step. If n=3, this becomes a 7-point derivative, and the following equation is obtained:
[0113]
number
[0114] If the measurement start energy is E0, the electron intensity is as follows:
[0115]
number
[0116] This is expressed as follows:
[0117]
number
[0118] Differentiation is expressed in the same way.
[0119]
number
[0120] For any integer m, N m The seven-point derivative of is expressed as follows:
[0121]
number
[0122] Since the spectral intensity follows a Poisson distribution, N m Noise level σ m is expressed as follows:
[0123]
number
[0124] N m is the signal component S m and the noise component σ mIt is expressed as the sum of.
[0125]
number
[0126] Therefore, N m The seven-point differentiation of is as follows:
[0127]
number
[0128] However, D sm D m The signal component of D σm D m is the noise component.
[0129] In the following, S m-3 ~S m+3 Assume that the intensities of are approximately equal. σ m-3 ~σ m+3 Assuming that is a random variable, σ m-3 ~σ m+3 The variances of can be considered to be approximately equal.
[0130] In the following, the energy step ΔE1 during spectrum measurement is referred to as the measurement step, and the energy step ΔE (ΔE2) during differential calculation using the Savitzky-Golay method is referred to as the differential step.
[0131] First, the noise level when ΔE1 / ΔE2=1, that is, the measurement step and the differentiation step are equal, is calculated.
[0132] The noise components at each measured energy are independent (random variables are independent). Therefore, the intensity N m The noise component σ m would be as follows:
[0133]
number
[0134] From the above formula, we can see that the larger the energy step ΔE, the smaller the noise level. However, since the energy width for differential calculation is proportional to ΔE, there is a trade-off between the energy width during differential calculation and the noise level.
[0135] Next, the noise level when ΔE1 / ΔE2=2, that is, the measurement step is twice the differential step, is calculated.
[0136] If the measurement step is twice the derivative step, then the measurement step should be 2·ΔE and the derivative step should be ΔE.
[0137] To perform differential calculations using the Savitzky-Golay method, the spectrum of measurement step 2·ΔE is step-converted by linear interpolation.
[0138]
number
[0139] In contrast, linear interpolation is as follows:
[0140]
number
[0141] At this time, D σm is different when m is even and when m is odd.
[0142] When m is even (m modulo2=0), D σm is expressed as follows:
[0143]
number
[0144] When m is odd (m modulo2=1), D σmis expressed as follows:
[0145]
number
[0146] When ΔE1 / ΔE2=M for a natural number M, that is, when the measurement step is M times the differentiation step, only the calculation results of the noise level are shown below.
[0147] [Table 1]
[0148] Fig. 18 is a graph showing the calculation results of the noise level when the measurement step is M times the differential step. Note that in the graph shown in Fig. 18, the noise level is calculated using the D σm is normalized as 1.
[0149] In the table and graph showing the noise level calculation results above, the measurement steps are different, so the measurement times are also different. The larger the measurement step, the longer the measurement time at each measurement energy is, so that the measurement time is normalized to be the same as the measurement time at M=1.
[0150] When the measurement step is M times, the measurement time is also M times and the spectrum intensity is also M times. Therefore, the signal-to-noise ratio is M 1 / 2 2x improvement.
[0151] Figure 19 shows the D normalized by the measurement time for M=1. σm 19 is a graph showing the noise component. It can be seen from the graph in Fig. 19 that the noise component decreases when the measurement step is increased. For example, when the measurement step is increased by four times, the noise component is reduced to about half.
[0152] When the measurement step and the differentiation step are different, N(E) is interpolated and then the differentiation calculation is performed. Here, the energy at the time of measurement and the energy at the time of differentiation are shifted by 0 or more and ΔE or less.
[0153] Counting electrons at energies E0, E0+ΔE, E0+2·ΔE,... The numbers N0, N1, N2, ... are expressed as follows:
[0154]
number
[0155] If the energy during differentiation is shifted by (1 / 2) ΔE, interpolation is performed as follows.
[0156]
number
[0157] Also, let the energy be (f n / f d ) ΔE, the following formula is used for interpolation:
[0158]
number
[0159] When the measurement energy is shifted, the independent and dependent components of the noise components change at adjacent energy steps. Therefore, D σm also changes.
[0160] Figure 20 shows the D normalized by the measurement time for M=1. σm However, the energy at the time of differentiation is Shift the - by (1 / 2)·ΔE and then D σm was calculated.
[0161] Figure 21 is a graph comparing the cases where the energy shift during differentiation is zero and (1 / 2)·ΔE. Note that the left bar for each item is the case where the shift is zero, and the right bar is the case where the shift is (1 / 2)·ΔE.
[0162] Regardless of whether the energy at the time of differentiation is shifted or not, it can be seen that the noise level decreases as M is increased. Also, even when the energy at the time of differentiation is shifted before interpolation, the noise level fluctuates by only a few tens of percent, and the noise level does not change by more than a factor of two.
[0163] This characteristic is believed to be true not only when the shift is zero or when the shift is (1 / 2) · ΔE, but also for any shift between 0 and ΔE.
[0164] The spectrum acquired by the electron spectrometer is created by integrating the signals from multiple channels into one spectrum. However, because the electron energies that can be acquired simultaneously by the multi-channel detector are different, the energy of the signals from each channel is converted. The signal intensity after the energy conversion is calculated by interpolation.
[0165] Here, D σm For each M, calculate the average of m modulo M. Furthermore, D σm For each M, calculate the average for zero shift and for (1 / 2) ΔE.
[0166] Figure 22 shows the D σm 22 is a graph showing the average result. The graph shown in Fig. 22 shows an estimate of the noise level when the measurement step is an integer multiple of the differentiation step. From the graph shown in Fig. 22, it is possible to estimate the change in noise level when the measurement step is increased.
[0167] Even if the measurement step and differentiation step are different, the noise level during spectrum generation has been estimated. The estimated results show that the larger the measurement step, the smaller the noise level.
[0168] (2) Energy steps and spectral distortion Next, the relationship between the measurement step and the distortion of the spectrum peak will be described.
[0169] Fig. 23 is a diagram showing the resolution function. The resolution function shown in Fig. 23 has a peak at 2000 eV and a ΔE (full width at half maximum) of 12 eV. For convenience, the resolution function is illustrated here as a Gaussian distribution.
[0170] In order to detect half the maximum intensity of the peak at around 2000 eV, the measurement step must be 12 eV (ΔE) or less. However, since the qualitative analysis of Auger electron peaks also uses information on the peak shape, the measurement step must be much smaller than ΔE.
[0171] Fig. 24 shows the resolution function when the measurement step is changed. Interpolation is performed for energies that are not measured. Fig. 24 shows the resolution function when the measurement step is changed in 1 eV increments from ΔE=1 eV to 12 eV. Fig. 25 is a graph showing the results of seven-point differentiation of the resolution function when the measurement step is changed.
[0172] FIG. 26 shows the distribution of the resolution function shown in FIG. 24 when the measurement start energy is shifted by +2 eV. Fig. 27 is a graph showing the resolution function and its differential result when the measurement start energy of the resolution function shown in Fig. 24 is shifted by +4 eV. Fig. 28 is a graph showing the resolution function and its differential result when the measurement start energy of the resolution function shown in Fig. 24 is shifted by +6 eV.
[0173] As shown in Figures 25 to 28, it can be seen that the differential waveform is distorted when the measurement step becomes large. In particular, when the measurement step exceeds the energy width during differential calculation, exceeding 6 eV in the illustrated example, the distortion of the differential waveform becomes large. When the measurement start energy is changed, the differential waveform of the resolution function also changes.
[0174] Figure 29 shows the result of integrating seven spectra. As shown in Figure 29, by integrating seven spectra, the waveform becomes smoothly continuous. However, the larger the measurement step, the more distorted the waveform becomes.
[0175] From the results shown in FIGS. 24 to 29, it can be seen that for the resolution function shown in FIG. 23, the peak waveform is not distorted when the measurement step is 4 eV or less.
[0176] 5. Effects In the electron spectrometer 100, the processing unit 60 performs the following processes: acquiring a plurality of first spectra obtained by sweeping the energy in the spectrometer 30 at a first energy step ΔE1 (measurement step) and detecting the electrons energy-spectroscopically separated by the spectrometer 30 with a plurality of channeltrons 42; complementing the measurement points in each of the plurality of first spectra; and generating a second spectrum (multispectrum) with a second energy step ΔE2 smaller than the first energy step ΔE1 based on the plurality of first spectra with the complemented measurement points. Therefore, in the electron spectrometer 100, the number of measurements can be reduced compared to, for example, the case where electrons are detected with one specific channeltron. Therefore, in the electron spectrometer 100, the measurement time for collecting a spectrum can be shortened.
[0177] In the electron spectrometer 100, the energy step ΔE2 during spectrum generation is different from the energy step ΔE1 during measurement. Therefore, the electron spectrometer 100 can set optimal energy steps during measurement and spectrum generation. Therefore, even if the optimal energy step for measuring a spectrum differs from the optimal energy step for displaying a generated spectrum or analyzing a spectrum, the electron spectrometer 100 can set optimal energy steps for each.
[0178] In the electron spectrometer 100, the processing unit 60 generates a second spectrum by integrating or averaging a plurality of first spectra with the supplemented measurement points. In this way, by integrating or averaging the first spectra with the supplemented measurement points, the influence of variations in detection sensitivity can be reduced. Therefore, in the electron spectrometer 100, it is not necessary to correct the detection sensitivity of each channeltron 42.
[0179] In the electron spectrometer 100, the processing unit 60 performs a process of receiving a designation of the energy resolution and a process of setting the energy step ΔE1 based on the energy resolution. Therefore, in the electron spectrometer 100, the user does not need to set the energy step ΔE1.
[0180] In the electron spectrometer 100, the processing unit 60 sets a ratio d between the measurement energy difference between adjacent channeltrons 42 and the energy step ΔE1 based on the energy resolution, and sets the energy step ΔE1 based on the ratio d. By setting the energy step ΔE1 based on the ratio d, the energy step ΔE1 can be easily set.
[0181] In the electron spectrometer 100, the processing unit 60 sets the energy step ΔE1 so that the measurement energies of the multiple first spectra do not coincide with each other. This makes it possible to narrow the energy range R of only the interpolated measurement points shown in FIGS. 14 and 15.
[0182] For example, in the electron spectrometer 100, the processing unit 60 sets the ratio d to a value smaller than 1. By setting the ratio d to a value smaller than 1, it is possible to reduce the number of matching measurement energies in each of the multiple first spectra, compared to when the ratio d is 1. This makes it possible to narrow the energy range R of only the interpolated measurement points shown in FIGS. 14 and 15.
[0183] The analysis method using the electron spectrometer 100 includes the steps of sweeping the energy in the spectrometer 30 at a first energy step ΔE1, detecting the electrons energy-spectroscopically separated by the spectrometer 30 with a plurality of channeltrons 42 to obtain a plurality of first spectra, complementing the measurement points in each of the plurality of first spectra, and generating a second spectrum with a second energy step ΔE2 smaller than the energy step ΔE1 based on the plurality of first spectra with the complemented measurement points. In this analysis method, the number of measurements can be reduced compared to, for example, the case where electrons are detected with one specific channeltron. Therefore, in the analysis method using the electron spectrometer 100, the measurement time for collecting a spectrum can be shortened.
[0184] In the above-described embodiment, an Auger electron spectroscopy apparatus has been described as an electron spectroscopy apparatus, but the primary probe irradiated to the sample is not limited to electrons, and may be anything that can emit electrons from the sample. For example, the primary probe may be an X-ray. That is, the electron spectroscopy apparatus according to the present invention may be an X-ray photoelectron spectroscopy apparatus.
[0185] The present invention is not limited to the above-described embodiment, and various modifications are possible. For example, the present invention includes a configuration that is substantially the same as the configuration described in the embodiment. A substantially identical configuration is, for example, a configuration that has the same function, method, and result, or a configuration that has the same purpose and effect. The present invention also includes a configuration in which non-essential parts of the configuration described in the embodiment are replaced. The present invention also includes a configuration that has the same action and effect as the configuration described in the embodiment, or a configuration that can achieve the same purpose. The present invention also includes a configuration in which publicly known technology is added to the configuration described in the embodiment. [Explanation of symbols]
[0186] 10...electron beam irradiation device, 12...electron gun, 14...electron lens, 16...deflector, 20...sample stage, 30...spectroscope, 32...input lens, 33...electrostatic lens, 34...electrostatic hemispherical analyzer, 35a...inner hemispherical electrode, 35b...outer hemispherical electrode, 40...detector, 42...channeltron, 50...irradiation control device, 52...spectroscope control device, 54...counting calculation device, 60...processing unit, 100...electron spectrometer
Claims
1. a spectrometer for energy-spectroscopically analyzing electrons emitted from the sample; a detector including a plurality of detection units arranged side by side in an energy dispersion direction of the electrons energy-split by the spectrometer; A processing section; Including, The processing unit includes: a process of acquiring a plurality of first spectra obtained by sweeping energy in a first energy step in the spectrometer and detecting the electrons energy-spectroscopically separated by the spectrometer with the plurality of detection units; A process of complementing a measurement point in each of the plurality of first spectra; a process of integrating or averaging the plurality of first spectra with the interpolated measurement points to generate a second spectrum having a second energy step smaller than the first energy step; Do the following: the processing unit sweeps the energy in the spectrometer at the first energy step to obtain the first spectrum at the first energy step for each of the detection units; The processing unit sets the first energy step so that measurement energies of the plurality of first spectra do not coincide with each other.
2. In claim 1, The processing unit includes: A process of receiving a designation of an energy resolution; setting the first energy step based on the energy resolution; An electron spectroscopy device that performs this function.
3. In claim 2, The processing unit includes: setting a ratio between a measurement energy difference between adjacent detection units and the first energy step based on the energy resolution; The electron spectrometer sets the first energy step based on the ratio.
4. In claim 3, The processing unit sets the ratio to a value smaller than 1.
5. a spectrometer for energy-spectroscopically analyzing electrons emitted from the sample; a detector including a plurality of detection units arranged side by side in an energy dispersion direction of the electrons energy-split by the spectrometer; An analysis method using an electron spectrometer, comprising: a step of sweeping energy in a first energy step in the spectrometer, and detecting electrons energy-spectroscopically separated by the spectrometer with the plurality of detection units to obtain a plurality of first spectra; complementing measurement points in each of the plurality of first spectra; integrating or averaging the plurality of first spectra with the interpolated measurement points to generate a second spectrum having a second energy step smaller than the first energy step; Including, In the step of acquiring a plurality of first spectra, sweeping energy in the spectrometer at the first energy step to obtain the first spectrum at the first energy step for each of the detection units; An analysis method, comprising: setting the first energy step so that the measured energies of the plurality of first spectra do not coincide with each other.
Citation Information
Patent Citations
Analyzer for electronic energy
JP1995072102A
Electron spectroscope
JP2001312994A
Electron beam analyzer
JP2002323463A
Electronic spectrograph
JP2002340827A
Electron spectroscopic device and measuring method
JP2017204425A
