JP2012013502A - Method and device for ion scattering spectroscopy measurement - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an ion scattering spectroscopy measuring method which can appropriately determine irradiation quantity of probe particles.SOLUTION: The ion scattering spectroscopy measuring method comprises: a step for starting radiation of probe particles to a sample; a step for integrating the number of the probe particles scattered at the sample and detected to a first irradiation quantity and obtaining a first ion scattering spectrum; a step for integrating the number of the detected probe particles to a second irradiation quantity which is more than the first irradiation quantity and obtaining a second ion scattering spectrum; a step for calculating a first scattering ratio which is the number of the detected probe particles per unit irradiation quantity for a first energy range corresponding to a first element in the sample, based on the first ion scattering spectrum; a step for calculating a second scattering ratio for the first energy range based on the second ion scattering spectrum; and a step for determining whether the difference of the second scattering ratio from the first scattering ratio is small or not relative to the standard.

Description

  The present invention relates to an ion scattering spectroscopic measurement method and an ion scattering spectroscopic measurement apparatus.

  Ion scattering spectroscopy (also called Rutherford Backscattering Spectroscopy (RBS)) uses the scattering interaction between probe particles such as He ions and H ions and various elements in the sample to determine the elemental composition in the sample. It is a technique to analyze. In ion scattering spectroscopy, the sample may be damaged due to irradiation of the sample with probe particles. For example, in order to suppress such damage to the sample, it is preferable that the irradiation amount of the probe particles is appropriately determined.

JP 2010-010126 A

  An object of the present invention is to provide an ion scattering spectroscopic measurement method and an ion scattering spectroscopic measurement apparatus capable of appropriately determining the irradiation amount of probe particles.

  According to one aspect of the present invention, the step of starting irradiation of a sample of probe particles for ion scattering spectroscopy, the number of the probe particles scattered by the sample and detected by a detector, To obtain a first ion scattering spectroscopic spectrum indicating the distribution of the number of detected probe particles for each energy, and to detect the probe particles scattered by the sample and detected by the detector. Accumulating the number of particles to a second dose greater than the first dose of the probe particles to obtain a second ion scattering spectrum showing the distribution of the number of detected probe particles for each energy; and Based on the first ion scattering spectrum, the first energy range corresponding to the first element contained in the sample is expressed as the number of detected probe particles per unit dose of the probe particles. A second scattering which is the number of detected probe particles per unit dose of the probe particles for the first energy range based on the step of calculating the first scattering rate and the second ion scattering spectrum. An ion scattering spectrometry method comprising: calculating a rate; and determining whether a difference between the second scattering rate and the first scattering rate is small with respect to a predetermined first reference Provided.

  The difference between the second scattering rate obtained by irradiating the second irradiation amount of probe particles and the first scattering rate obtained by irradiating the first irradiation amount of probe particles is different from the reference. By determining whether or not the amount is small, it is possible to determine whether or not the measurement result of the ion scattering spectroscopy has become stable, so that the probe particle irradiation amount can be appropriately determined.

FIG. 1 is a cross-sectional view schematically showing an ion scattering spectrometer according to an embodiment of the present invention. FIG. 2 conceptually shows the true spectrum of ion scattering spectroscopy and the spectrum of a damaged sample. FIG. 3A shows a measurement spectrum at an irradiation dose of 1000 nC and an irradiation dose of 5000 nC obtained by ion scattering spectroscopy of the example, and FIG. 3B shows a measurement spectrum at an irradiation dose of 45000 nC. FIG. 4 is a flowchart showing a flow of a method for determining the total irradiation amount of the probe particles in the ion scattering spectrometry of the embodiment. FIG. 5A shows a measurement spectrum and an estimated spectrum at an irradiation amount of 4000 nC obtained by ion scattering spectroscopy of the example, FIG. 5B shows a measurement spectrum and an estimated spectrum at an irradiation amount of 5000 nC, and FIG. FIG. 5D shows the estimated spectra of the irradiation amount of 4000 nC and the irradiation amount of 5000 nC, and FIG. 5D shows the estimated spectra of the irradiation amount of 4000 nC and the irradiation amount of 5000 nC with the same intensity.

  First, the structure of an ion scattering spectrometer according to an embodiment of the present invention and the outline of an ion scattering spectrometer measurement method will be described.

  FIG. 1 is a cross-sectional view schematically showing the structure of the ion scattering spectrometer of the embodiment. The ion scattering spectrometer of the embodiment is formed including a probe particle irradiation device 1, a chamber 2, a detector 3, and an analysis device 4.

  The probe particle irradiation apparatus 1 includes an ion source 1a, an acceleration tube 1b, and a Wien filter 1c. The ion source 1a emits ions (for example, He ions) as probe particles. Probe particles emitted from the ion source 1a are accelerated to, for example, about several hundred kV by the acceleration tube 1b. The Wien filter 1c selectively passes probe particles having a desired energy among the probe particles accelerated by the acceleration tube 1b. Probe particles that have passed through the Wien filter 1 c enter the chamber 2.

  An analysis target sample S is held inside the chamber 2. The probe particles are irradiated onto the sample S, and interact with elements in the sample S to be scattered. The energy of the probe particles varies depending on the type of interacting element and the position of the element in the depth direction. Probe particles scattered by the sample S are detected by the detector 3.

  The detector 3 includes a magnetic field filter 3a and a position sensitive detector (PSD) 3b. Of the probe particles scattered by the sample S, those traveling in the direction of the exit of the chamber 2 pass through the exit of the chamber 2 and enter the magnetic field filter 3a. The magnetic field filter 3a includes a magnet and divides the trajectory of the probe particle according to the energy of the probe particle. Due to the magnetic field of the magnetic field filter 3a, the trajectory of the probe particles with lower energy is bent more greatly.

  Probe particles that have passed through the magnetic field filter 3a enter the PSD 3b. The PSD 3b has detection channels arranged in directions in which the trajectories of the probe particles are divided, and detects the probe particles for each channel.

  Probe particle detection information by the PSD 3b is sent to the analyzer 4. The analyzer 4 is formed including a personal computer, for example. The analyzer 4 integrates the number of incident probe particles for each channel of the PSD 3b. Since the position of the channel corresponds to the magnitude of energy, by integrating the number of probe particles for each channel, an ion scattering spectrum that is the distribution of the number of detected probe particles for each energy is obtained. Hereinafter, the ion scattering spectrum may be simply referred to as a spectrum.

  Such a high resolution ion scattering spectrometer using a magnetic field deflection type energy analyzer as the detector 3 can measure the scattered probe particles with a resolution of several hundred volts, for example, in the depth direction. Enables compositional analysis at the layer thickness level.

  In this embodiment, the high-resolution ion scattering spectrometer is described. However, in the conventional ion scattering spectrometer in which the scattered probe particles are directly incident on the semiconductor detector, each energy is also measured. Similarly, an ion scattering spectrum is obtained as the distribution of the number of detection probe particles.

  Next, problems in ion scattering spectroscopy will be considered. The ion scattering spectrum is obtained by integrating probe particles detected by a detector for each energy. Therefore, in order to reduce noise and obtain an accurate spectrum, it is desirable to increase the total amount of probe particles irradiated onto the sample so that a sufficiently large number of probe particles can be detected. However, if the irradiation amount of the probe particles becomes too large, there is a concern that the sample is damaged due to the irradiation of the probe particles and deteriorates, and the spectrum deviates from the true one.

  FIG. 2 conceptually shows the true spectrum SP and the spectrum SPD of the damaged sample. The spectrum SPD of the damaged sample is greatly deviated from the true spectrum SP, and correct measurement is not performed.

  Until now, determination of the total irradiation amount of probe particles has been made empirically while the measurer observes the shape of the spectrum and the temporal shape change of the spectrum. As described above, the total irradiation amount of the probe particles is large enough to suppress the influence of noise, and small enough to suppress the damage to the sample. However, it is difficult to subjectively determine an appropriate total irradiation amount based on the experience of the measurer, and there are many cases where an excessive irradiation amount is obtained in order to obtain an accurate measurement result.

  Note that the solid angle of the detection unit is smaller in the high-resolution ion scattering spectrometer than in the conventional ion scattering spectrometer. Accordingly, since the number of probe particles that can be collected per unit time is reduced, the irradiation becomes long and the irradiation amount tends to be excessive.

  The inventor of the present application proposes a technique capable of objectively and appropriately determining the total irradiation amount of probe particles as described below.

Next, the ion scattering spectroscopic measurement method according to the examples will be described more specifically based on experimental examples. He ion (He ⁺ ) was used as the probe particle, and the probe particle was accelerated at 400 keV and irradiated to the sample S. Sample S was about 2 nm thick SiO ₂ grown on a single crystal Si substrate. The measurement mode is aligned measurement in which the incident direction of the probe particles with respect to the substrate is matched with the crystal orientation of the substrate.

  The amount per probe particle irradiation was set to a charge amount of 1000 nC. Irradiation of probe particles with a charge amount of 1000 nC was repeated several times at the same location to increase the irradiation amount. A spectrum obtained by integration for N times of irradiation is represented as SPn.

By dividing the amount of probe particles expressed by the amount of charge by the amount of charge of He ⁺ , the number of probe particles can be converted. Hereinafter, in order to avoid complicated explanation, the amount of probe particles expressed by the amount of charge and the amount of probe particles expressed by the number are used without distinction. In addition, as a way of representing the amount of irradiation of the probe particles, the way of representing the number of irradiations N times and the way of representing the amount of charge N × 1000 nC are used without particular distinction.

  As will be described in detail later, in this experimental example, the determination for determining the total irradiation amount is performed by the analyzer while irradiating the probe particles, and it is determined that 5000 nC is appropriate as the total irradiation amount.

  Prior to the description of the method for determining the total irradiation amount, a specific example of the ion scattering spectrum and the scattering rate will be described. First, as a specific example of the spectrum, a spectrum at an irradiation dose of 1000 nC and a spectrum at an irradiation dose of 5000 nC will be described.

  FIG. 3A shows the spectra SP1 and SP5 with a dose of 1000 nC and a dose of 5000 nC. The horizontal axis indicates PSD channel (ch) number, and the vertical axis indicates intensity (arbitrary unit). As mentioned above, PSD channels correspond to energy. The larger the channel number, the higher the energy. It can be seen that, in the spectrum SP5, as compared with the spectrum SP1, the noise is reduced (the SN ratio is improved) due to the increase in the integrated amount of the detection probe particles.

  In general, in an ion scattering spectrum, the peak of a heavy element in a sample appears on the high energy side, and the peak of a light element appears on the low energy side. This is because the energy lost by the incident probe particles is larger in the interaction with the light element.

  Also, the lighter the element, the less likely it will interact with the probe particles (the lower the scattering intensity). Therefore, the detected number of probe particles scattered by light particles is smaller than the detected number of probe particles scattered by heavy elements. Therefore, lighter element peaks in the spectrum are more susceptible to noise than heavier element peaks.

  Even in the same element, the shallower one is detected on the higher energy side, and the deeper one is detected on the lower energy side, and a peak having a certain width is formed. This is because the deeper the scattered element, the greater the energy lost before the incident probe particle reaches.

  The probe particle interacts with higher energy as the element is at a shallower position. However, the higher the energy of the probe particles, the less likely they are scattered by the element. This indicates that an element closer to the surface in the sample is less susceptible to the probe particle irradiation, that is, the vicinity of the outermost surface of the sample is most unlikely to be damaged due to the probe particle irradiation.

  For example, as can be seen from spectrum SP5, there are peaks in the vicinity of 170 ch and 270 ch, respectively, peak Po near 170 ch indicates relatively light oxygen, and peak Psi near 270 ch indicates relatively heavy silicon. It can be seen that the oxygen peak Po has a lower intensity than the silicon peak Psi and is more susceptible to noise. A portion where the peak Psi of silicon falls to the high energy side, for example, a portion shown in the range of 302 ch to 322 ch indicates silicon near the outermost surface of the sample.

Next, the scattering rate will be described. The ratio I _{i /} I 0 of the irradiated probe particle number I _i scattered in a certain element i with respect to the probe particle number I _0, i.e., the unit dose per probe particles, is how the scattered probe particles Elemental i The ratio indicating whether it is detected will be referred to as the scattering rate. The scattering rate can be expressed by the following formula (1).

... (1)
Here, N _i is the number of elements i contained in the sample, dx is the depth of the layer in which scattering occurs, σ _i is the scattering cross section of element i and the probe particles, and ΔΩ and η are determined by the measuring device. It is a constant. From equation (1), it can be seen that the scattering rate is essentially a constant value. If a sufficient amount of probe particles are irradiated to the sample and a sufficient amount of probe particles are detected, the influence of noise is reduced and the scattering rate calculated based on the measurement converges to a constant value. Is expected to do.

  For normal samples (which are not particularly susceptible to damage), the probe particles have such a degree that the scattering rate converges almost uniformly (and the extent that the spectrum shape converges almost constant as described later). It can be said that the amount of irradiation is not so great as to cause damage to the sample.

  Accordingly, the scattering rate obtained based on the spectrum SPn obtained by N times of probe particle irradiation is obtained based on the spectrum SP (n-1) obtained by (N-1) times of probe particle irradiation. If there is almost no change from the scattering rate, it can be determined that N times of probe particle irradiation is a sufficient irradiation amount.

  Next, a method for determining the total irradiation amount of probe particles will be described with reference to the flowchart of FIG. First, in step S0, measurement is started. In step S1, probe particle irradiation is performed each time, and a spectrum at each irradiation frequency is obtained.

  Since the ion scattering spectrum reflects the distribution of elements in the sample, it is considered that the ion scattering spectrum has essentially a smooth shape like the true spectrum SP shown in FIG. However, the spectrum actually obtained by the ion scattering spectrometer has a shape including fine irregularities due to noise, as shown in the spectrum SP5 of FIG. 3A and the like.

  In step S2, a smooth spectrum is estimated by removing fine irregularities due to noise in the spectrum at each number of irradiations by statistical processing. The estimated spectrum is called an estimated spectrum. Hereinafter, the spectrum before estimation may be referred to as a measurement spectrum.

  The estimation is performed by Bayesian estimation under the condition that the prior distribution becomes maximum when the square sum of the second-order difference between three consecutive points is minimum, assuming that the noise component is Gaussian, for example. Can do. Such a Bayesian estimation method is described in “Bayesian estimation and statistical physics” (Yukito Iba, Iwanami lecture, physics world). Note that the estimation method is not limited to Bayesian estimation, and other methods such as a moving average method may be used.

  FIG. 5A shows a measured spectrum SP4 obtained by four irradiations (irradiation amount 4000 nC) and an estimated spectrum SPE4 obtained by Bayesian estimation from the measured spectrum SP4.

  FIG. 5B shows a measured spectrum SP5 obtained by five irradiations (irradiation amount of 5000 nC) and an estimated spectrum SPE5 obtained by Bayesian estimation from the measured spectrum SP5.

  FIG. 5C shows estimated spectra SPE4 and SPE5 side by side.

  It can be seen that the estimated spectra SPE4 and SPE5 are smoother than the measured spectra SP4 and SP5, respectively, by removing fine irregularities caused by noise. Returning to FIG. 4, the description will be continued.

  In step S3, the scattering rate is calculated based on the spectrum at each number of irradiations, and the scattering rates of the two most recent irradiations are compared. First, a method for calculating the scattering rate will be described.

  The denominator of the scattering rate is the total of the irradiation amount of the probe particles at the time of irradiation for each number of times of irradiation. As the numerator of the scattering rate, the integration of the number of probe particles detected in the energy range (channel range) corresponding to the element i at the time of irradiation of each number of irradiations can be used.

  It is desirable that the scattering rate be obtained with less influence of noise. For this reason, it is considered more desirable to obtain the scattering rate for the heaviest element in the sample. In this embodiment, the scattering rate is obtained for silicon.

  Further, if the sample is damaged, the scattering rate cannot be obtained accurately. Therefore, it is considered more desirable to obtain the scattering rate for the element in the portion with little damage near the outermost surface. In this embodiment, the range of PSD 302ch to 322ch is handled as silicon distributed near the outermost surface of the sample.

  However, as described above, the influence of noise is large if the measurement spectrum is used as it is. Therefore, it is considered more desirable to obtain the scattering rate from the estimated spectrum estimated in step S2.

  The procedure for calculating the scattering rate with N irradiations is summarized as follows. The area of 302ch to 322ch of the estimated spectrum SPEn with N irradiation times, that is, the number of probe particles scattered and detected by silicon distributed near the outermost surface of the sample is calculated. Then, the number of detection probe particles is divided by N × 1000 nC, which is the irradiation amount of the probe particles, to calculate the scattering rate (the number of detection probe particles per unit irradiation amount) at N irradiation times.

  If the scattering rate at the number of irradiations N is calculated, it is evaluated whether the difference with respect to the scattering rate at the number of irradiations (N-1) is small with respect to a predetermined reference. If the difference in the scattering rate is equal to or greater than the reference, it means that the irradiation amount is still insufficient, so the process returns to step S1, and the (N + 1) th irradiation is performed, and the spectrum at the number of irradiations (N + 1) is acquired. The On the other hand, if the difference in the scattering rate is smaller than the reference, the irradiation amount is sufficient from the viewpoint of the scattering rate, and the process proceeds to step S4.

  The flow of the procedure for comparing the scattering rates will be described more specifically. Measurement is started in step S0. First, in step S1, first irradiation is performed, and a measurement spectrum SP1 at an irradiation amount of 1000 nC is obtained. In step S2, an estimated spectrum SPE1 at an irradiation amount of 1000 nC is estimated. In step S3, the scattering rate at an irradiation amount of 1000 nC is calculated. The scattering rate at an irradiation dose of 1000 nC was 0.08. At this stage, since only one spectrum at an irradiation dose of 1000 nC has been acquired, the scattering rates cannot be compared.

  Next, returning to step S1, the second irradiation is performed, and a measurement spectrum SP2 at an irradiation amount of 2000 nC is obtained. In step S2, an estimated spectrum SPE2 at an irradiation amount of 2000 nC is estimated. In step S3, the scattering rate at an irradiation amount of 2000 nC is calculated. The scattering rate at an irradiation dose of 2000 nC was 0.13.

  Further, in step S3, the amount of deviation of the scattering rate at the irradiation amount of 2000 nC from the scattering rate at the irradiation amount of 1000 nC is evaluated. The amount of deviation is, for example, | An−A (n−1) | / A where A (n−1) is the scattering rate at the number of irradiations (N−1) and An is the scattering rate at the number of irradiations N. (N-1) is calculated. Then, it is determined whether the deviation amount is smaller than a reference value, for example, 0.1. Here, the amount of deviation is obtained as | 0.13-0.08 | /0.08=0.63, which is 0.1 or more, so the third time is returned to step S1 and the irradiation amount is increased to 3000 nC. Probe particle irradiation is performed. Note that the formulas and reference values for evaluating the difference in scattering rate are not limited to those in this example.

  Such a procedure is repeated until the scattering amount of the scattering rate becomes smaller than the reference value. In this experimental example, the scattering rate at an irradiation amount of 4000 nC was 0.196, and the scattering rate was 0.192 at an irradiation amount of 5000 nC. The amount of deviation of the scattering rate at the irradiation amount of 5000 nC from the scattering rate at the irradiation amount of 4000 nC is | 0.192−0.196 | /0.196=0.02, which is smaller than 0.1. From the viewpoint of rate, it is determined that 5000 nC is a sufficient dose, and the process proceeds to step S4.

  Thus, by evaluating the change in the scattering rate, an appropriate total irradiation amount of the probe particles can be estimated.

  Furthermore, in the present embodiment, as described below, in step S4, the accuracy of ion scattering spectrometry can be improved by evaluating the spectral shape change and determining the total dose.

  In step S4, the difference between the shape of the estimated spectrum SPEn at the number of irradiations N that has passed through step S3 and the shape of the estimated spectrum SPE (n-1) at the number of irradiations (N-1) is a predetermined reference. It is evaluated whether it is less for.

  Specifically, in the present experimental example, the measurement result at the number of irradiations of 5 passes through step S3, and in step S4, the difference between the shape of the estimated spectrum SPE5 and the shape of the estimated spectrum SPE4 is evaluated. A method for comparing spectral shapes will be described.

  First, the area of the estimated spectrum SPE5 surrounded by the spectrum in the range of 302ch to 322ch (that is, the number of probe particles scattered and detected by silicon distributed near the outermost surface of the sample) is divided by that of the estimated spectrum SPE4. Thus, the magnification α is calculated. In addition, since the area of 302ch-322ch part of each estimated spectrum is calculated | required at the time of calculation of a scattering rate, it can be diverted by step S4.

  Next, by multiplying the intensity of the estimated spectrum SPE4 by the magnification α, the intensity of the estimated spectrum SPE4 is made uniform (normalized) with the intensity of the estimated spectrum SPE5.

  FIG. 5D shows the estimated spectrum SPE5 and the normalized estimated spectrum SPE4 in an overlapping manner. It can be seen that the shapes of both spectra are in good agreement.

As a statistical index for numerically evaluating the coincidence of spectrum shapes, for example, chi-square χ ² can be used. The intensity of each channel of the estimated spectrum SPE5 x _5, the intensity of each channel of the estimated spectrum SPE4 which is standardized as x _4, chi-square chi ² can be expressed by the following equation (2).

... (2)
Here, the sum is taken over all channels of the PSD. The closer the shape of the estimated spectrum SPE5 is to the normalized shape of the estimated spectrum SPE4, the smaller the intensity difference between the channels, and χ ² becomes a value close to 0. It is determined whether the chi-square χ ² is a reference value, for example, 10 ⁻² or less. The chi-square χ ² for the estimated spectrum SPE5 and the normalized estimated spectrum SPE4 was on the order of 10 ⁻³ , which was smaller than the reference value.

The chi-square χ ² in the case where the spectral shapes were compared between the irradiation amount of 1000 nC and the irradiation amount of 5000 nC was on the order of 10 ⁻¹ , and the difference in the spectral shape was large. The index and reference value for evaluating the difference in spectrum shape are not limited to those in this example.

  If the difference in estimated spectral shape is below the standard, it can be said that the irradiation dose is sufficient from the viewpoint of the spectral shape in addition to the viewpoint of the scattering rate. In this experimental example, it is determined that an irradiation amount of 5000 nC is sufficient from the viewpoint of the spectral shape in addition to the scattering rate. And it progresses to step S5, a measurement result etc. are output to the display etc. of an analyzer as needed, the irradiation of the probe particle to a sample is complete | finished, and a measurement can be complete | finished.

  On the other hand, if the difference in estimated spectrum shape is larger than the reference, the process returns to step S1 again, the irradiation amount is increased, and the measurement is continued.

  As described above, since the light element (oxygen in the experimental example) in the sample has a low scattering intensity, the peak shape is easily affected by noise. The comparison of the scattering rates in step S3 was made with respect to a heavy element (silicon in the experimental example) in the sample. However, if attention is paid to heavy elements, the peak shape of a light element may not be stable at that irradiation amount even if the irradiation amount has a stable scattering rate. In step S4, by evaluating up to the change of the spectrum shape, it is possible to improve the measurement accuracy for light elements and improve the measurement reliability.

  As described above, in addition to the evaluation of the change in the scattering rate, the evaluation up to the change in the spectrum shape can further improve the measurement accuracy and estimate the total irradiation amount of the probe particles.

  In this experimental example, in order to finally confirm a stable measurement spectrum shape, measurement was performed up to an irradiation dose of 45000 nC. The sample used in the experiment was not particularly damaged even when the irradiation dose was increased to 45000 nC, and a stable measured spectrum shape could be observed.

  FIG. 3B shows a measured spectrum SP45 at an irradiation dose of 45000 nC. It is confirmed that a substantially stable spectrum shape is obtained at the stage of the measurement spectrum SP5 with an irradiation amount of 5000 nC shown in FIG. 3A.

  As described above, according to the method of the embodiment, objective measurement end point determination can be performed in ion scattering spectrometry. Measurement with excessive probe particle irradiation can be easily avoided, damage to the sample is suppressed, and measurement reliability is increased. In addition, since the total irradiation amount of the probe particles can be suppressed, the measurement time can be shortened.

  Even if the sample is not easily damaged by probe particle irradiation, the method of the example contributes to shortening of the measurement time and the like.

  In the above-described embodiment, the increase in the irradiation amount of the probe particles is set to a fixed amount. However, if necessary, the increase in the irradiation amount may not be set to a fixed amount.

  Although the present invention has been described with reference to the embodiments, the present invention is not limited thereto. It will be apparent to those skilled in the art that various modifications, improvements, combinations, and the like can be made.

DESCRIPTION OF SYMBOLS 1 Probe particle irradiation apparatus 1a Ion source 1b Acceleration tube 1c Wien filter 2 Chamber S Sample 3 Detector 3a Magnetic field filter 3b PSD
4 Analyzer

Claims (6)

A step of irradiating the sample with probe particles of ion scattering spectroscopy; and
First ion scattering indicating the distribution of the number of detected probe particles for each energy by integrating the number of probe particles scattered by the sample and detected by the detector up to the first irradiation amount of the probe particles. Obtaining a spectral spectrum;
The number of probe particles scattered by the sample and detected by the detector is integrated up to a second irradiation amount larger than the first irradiation amount of the probe particles, and the number of detected probe particles for each energy is calculated. Obtaining a second ion scattering spectrum showing the distribution;
Based on the first ion scattering spectrum, a first scattering rate that is the number of probe particles detected per unit dose of the probe particles for a first energy range corresponding to the first element contained in the sample. Calculating
Calculating a second scattering rate, which is the number of detected probe particles per unit dose of the probe particles, for the first energy range based on the second ion scattering spectrum;
And a step of determining whether or not a difference between the second scattering rate and the first scattering rate is small with respect to a predetermined first reference.   The ion scattering spectroscopic measurement method according to claim 1, wherein the first element is a heaviest element contained in the sample.   The ion scattering spectroscopic measurement method according to claim 2, wherein the first energy range is an energy range corresponding to the heaviest element distributed in the vicinity of the outermost surface of the sample. The step of calculating the first scattering rate includes the step of estimating a first estimated spectrum that is smoother than the first ion scattering spectral spectrum from the first ion scattering spectral spectrum, Calculating the first scattering rate from the estimated spectrum;
The step of calculating the second scattering rate includes the step of estimating a second estimated spectrum that is smoother than the second ion scattering spectral spectrum from the second ion scattering spectral spectrum, The ion scattering spectroscopic measurement method according to claim 1, wherein the second scattering rate is calculated from the estimated spectrum. further,
The number of detection probe particles calculated from the first estimated spectrum, which is the number of detection probe particles in the first energy range, and the first energy range calculated from the second estimated spectrum. Using the ratio of the number of detection probe particles to the number of second detection probe particles to align the intensities of the first estimated spectrum and the second estimated spectrum;
With respect to the first estimated spectrum and the second estimated spectrum having the same intensity, a deviation of the shape of the second estimated spectrum from the shape of the first estimated spectrum is different from a predetermined second reference. The method for determining ion scattering spectroscopy according to claim 4, further comprising a step of determining whether or not the amount is small. A chamber in which the sample is held,
A probe particle irradiation device for irradiating the sample with probe particles;
A detector for detecting the probe particles scattered by the sample;
An analysis device,
The analyzer is
A first ion scattering spectrum showing the distribution of the number of detected probe particles for each energy by integrating the number of probe particles detected by the detector from the start of irradiation of the probe particles to a first dose. And
The number of probe particles detected by the detector is integrated from the start of irradiation of the probe particles to a second irradiation amount greater than the first irradiation amount, and the distribution of the number of detected probe particles for each energy is obtained. Obtain a second ion scattering spectroscopy spectrum shown,
Based on the first ion scattering spectrum, a first scattering rate that is the number of probe particles detected per unit dose of the probe particles for a first energy range corresponding to the first element contained in the sample. To calculate
Based on the second ion scattering spectroscopic spectrum, for the first energy range, a second scattering rate that is the number of detected probe particles per unit dose of the probe particles is calculated,
An ion scattering spectrometer that determines whether or not a difference between the second scattering rate and the first scattering rate is small with respect to a predetermined first reference.