JP7580962B2 - Information processing device, control method for information processing device, and program - Google Patents

Description

The present invention relates to an information processing device, a control method for an information processing device, and a program.

Various additives are added to resins to prepare them with the desired mechanical, physical, and chemical properties. Therefore, knowing the concentration of additives added to resins is important for quality control and material design.

Methods for analyzing additives include extracting and separating the additives added to the resin and analyzing them using analytical equipment such as an ultraviolet detector, an infrared detector, or a mass spectrometer, and directly analyzing the resin itself to which the additives are attached.

Methods for extracting and separating additives added to resins include dissolving and extracting the additives using a solvent and then separating the additives using liquid chromatography, and separating low molecular weight components containing the additives using size exclusion chromatography (SEC). However, these methods are undesirable in terms of environmental impact and worker safety and health, as they use organic solvents such as tetrahydrofuran and chloroform.

Methods for directly analyzing the resin itself to which additives are attached include secondary ion mass spectrometry such as dynamic secondary ion mass spectrometry (Dynamic SIMS) and time-of-flight secondary ion mass spectrometry (TOF-SIMS). Dynamic SIMS is a method for continuously irradiating a solid sample with an ion beam to obtain the distribution of impurities in the depth direction. On the other hand, TOF-SIMS is a method for irradiating a solid sample with a smaller amount of ion beam than the Dynamic SIMS method to obtain information on elements and molecules present on the surface of the solid sample (see Patent Documents 1 and 2). Here, the principle of TOF-SIMS will be explained in detail.

When a solid sample is irradiated with an ion beam (primary ions) in a high vacuum, the components of the solid sample surface are released into the vacuum. Positively or negatively charged ions (secondary ions) generated during this process are converged in one direction by an electric field and detected at a certain distance away. Secondary ions with various masses are generated depending on the composition of the solid sample surface, but in a certain electric field, ions with smaller masses fly faster and ions with larger masses fly slower. Therefore, the mass of the secondary ions generated can be analyzed by measuring the time from when the secondary ions are generated until they reach the detector (time of flight).

The TOF-SIMS method has the advantages of being able to detect components on the order of ppm, i.e., 10 ¹¹ atoms/cm ^2, being applicable to both organic and inorganic substances, and being able to measure components present at a depth of 1 nm or less from the surface of a solid sample. However, in this method, since the solid sample itself is directly analyzed, the peaks of impurities are likely to be mixed into the spectrum obtained. In addition, the ease with which components become secondary ions (ionization efficiency) is easily affected by slight differences in the shape and charge of the solid sample surface, and the secondary ion intensity is likely to vary depending on the measurement point of the solid sample (see Non-Patent Documents 1 and 2). Furthermore, the secondary ion intensity is easily affected by differences in the type and composition of the solid sample. Therefore, the spectrum obtained by the TOF-SIMS method is likely to be complicated, and it takes time to obtain information on the solid sample from only the secondary ion intensity. Therefore, in order to obtain information on the solid sample (resin), a method of analyzing the spectrum obtained by the TOF-SIMS method using machine learning has been considered (see Non-Patent Document 3).

JP 2011-68005 A JP 2007-156093 A

"Recent Trends and Examples of ToF-SIMS Analysis" Surface Technology 59(12) p887-(2008) Materials Analysis by Mass Spectrometry, 2nd ed. ;IM Publications: Chichester, UK, 2013 "TOF-SIMS image data fusion using multivariate analysis and TOF-SIMS spectrum analysis using sparse modeling and machine learning", Journal of Surface Analysis 25(2) 103-114 (2018)

As described above, methods of analyzing spectra obtained by TOF-SIMS using machine learning to obtain information on solid samples (resins) have been investigated (see Non-Patent Document 3). However, no methods have been investigated so far for easily and highly accurately analyzing spectra obtained by TOF-SIMS to obtain quantitative information on the attachments (hereinafter sometimes referred to as "measurement target substances") attached to solid samples (resins).

Therefore, an object of the present invention is to provide an information processing device that can easily and accurately analyze quantitative information of a substance to be measured in a target sample from spectral information of the sample that contains a solid sample and the substance to be measured. Another object of the present invention is to provide a control method and program for the information processing device.

The information processing device of the present invention is an information processing device having an information acquisition means for acquiring quantitative information of the target substance in a target sample, which is estimated by inputting spectral information of a solid sample and spectral information of a sample containing the solid sample and the target substance into a learning model, and is characterized in that the spectral information is acquired by secondary ion mass spectrometry.

The control method for an information processing device of the present invention is a control method for an information processing device having an information acquisition step of acquiring quantitative information of the measured substance in a target sample, the quantitative information being estimated by inputting spectral information of a solid sample and spectral information of a sample containing the solid sample and the measured substance into a learning model, and is characterized in that the spectral information is acquired by secondary ion mass spectrometry.

The present invention provides an information processing device that can easily and accurately analyze quantitative information on a substance to be measured in a target sample from spectral information of the sample that contains a solid sample and the substance to be measured, as well as a control method and program for the information processing device.

1 is a diagram showing an overall configuration of an information processing system including an information processing device according to an embodiment of the present invention. FIG. 13 is a diagram illustrating an example of a flowchart of a processing procedure for generating a learning model according to the present embodiment. FIG. 11 is a diagram illustrating an example of a flowchart of a process for displaying quantitative information according to the embodiment. 1 shows an example of spectrum information of a solid sample and a sample containing the solid sample and a measurement target substance. 1 is a graph showing the change in ion intensity of a specific fragment presumed to be derived from a substance to be measured, relative to the concentration of the substance to be measured in a sample. 13 is a flowchart illustrating an example of a procedure for creating training data to which noise has been added. 4 is a display example of quantitative information of a measurement target substance in an information processing device.

Below, the form (embodiment) for carrying out the present invention will be described with reference to the drawings. However, the scope of the present invention is not limited to each of the embodiments described below.

In the present invention, the following information processing device is used to easily and accurately analyze quantitative information on the target substance in a target sample from the spectral information of a sample containing a solid sample and the target substance. The information processing device has an information acquisition means for acquiring quantitative information on the target substance in a target sample, which is estimated by inputting the spectral information of the solid sample and the spectral information of a sample containing a solid sample and the target substance into a learning model. In addition, this spectral information is acquired by secondary ion mass spectrometry.

This eliminates the need to prepare and input into the learning model spectral information with the same concentration and amount as the target substance in the sample, making it possible to analyze the concentration and amount of the target substance even if the concentration and amount of the target substance in the target sample are unknown.

(sample)
The sample includes a solid sample and a substance to be measured.

[Solid sample]
The solid sample preferably includes at least one selected from the group consisting of a resin, a silicon substrate, and a copper foil, and more preferably includes a resin. The resin preferably has a unit derived from an unsaturated carboxylic acid ester, and more preferably has a unit derived from an alkyl (meth)acrylate. Here, the unit refers to a repeating unit derived from one monomer.

[Measurement target substance]
The measurement target substance is preferably an additive added to a solid sample or an attachment attached to a solid sample. Examples of the measurement target substance include plasticizers, thickeners, viscosity reducers, crystal nucleating agents, crystallization promoters, crystallization retarders, release agents, flame retardant agents, flame retardant assistants, thermal decomposition inhibitors, reinforcing agents, conductive materials, antistatic agents, thermal conductive agents, antioxidants, weathering inhibitors, ultraviolet absorbers, antibacterial agents, colorants, gas barrier agents, antifogging agents, antireflective agents, antifouling agents, and impurities. Among these, the measurement target substance is preferably a plasticizer. Examples of the plasticizer include surfactants.

It is preferable that the group of elements constituting the solid sample and the group of elements constituting the substance to be measured are the same. It is preferable that the group of elements constituting the solid sample and the group of elements constituting the substance to be measured include at least one element selected from the group consisting of carbon atoms, hydrogen atoms, oxygen atoms, and nitrogen atoms.

(Quantitative information)
In this embodiment, the quantitative information includes the amount of the target substance in the sample, the concentration of the target substance in the sample, the presence or absence of the target substance in the sample, etc. In addition, other quantitative information includes the ratio of the amount or concentration of the target substance in the sample to a reference amount of the target substance, the ratio of the amount or concentration of the target substance in the sample, etc.

(Spectral information)
The spectrum information in this embodiment is acquired by secondary ion mass spectrometry. Examples of the secondary ion mass spectrometry include Dynamic SIMS and TOF-SIMS. Of these, it is preferable to use TOF-SIMS as the secondary ion mass spectrometry.

(Information processing system, information processing device)
Next, an information processing system according to this embodiment will be described with reference to Fig. 1. Fig. 1 is a diagram showing the overall configuration of an information processing system including an information processing device according to this embodiment.

The information processing system includes an information processing device 10, a database 22, and an analysis device 23. The information processing device 10 and the database 22 are communicatively connected to each other via a communication means. In this embodiment, the communication means is configured as a LAN (Local Area Network) 21. The information processing device 10 and the analysis device 23 are connected by a communication means of a standard such as USB (Universal Serial Bus). The LAN may be a wired LAN or a wireless LAN, or may be a WAN (Wide Area Network). The USB may also be a LAN.

The database 22 manages the spectral information acquired by the analysis performed by the analysis device 23. The database 22 also manages the learning model (trained model) generated by the learning model generation unit 42 described below. The information processing device 10 acquires the spectral information and learning model managed by the database 22 via the LAN 21.

(Learning model)
The learning model in this embodiment is a regression learning model, and can be generated by machine learning such as deep learning. A learning model is a model constructed by learning a machine learning algorithm using teacher data and making appropriate predictions. There are various types of machine learning algorithms used in the learning model. For example, deep learning using a neural network can be used. A neural network is composed of an input layer, an output layer, and multiple hidden layers, and each layer is connected by a formula called an activation function. When using teacher data with labels (output corresponding to input), the coefficients of the activation function are determined so that the relationship between the input and the output is established. By determining the coefficients using multiple teacher data, a learning model that can predict the output for the input with high accuracy can be generated.

(Analytical Equipment)
The analysis device 23 is a device for analyzing samples, solid samples, etc. The analysis device 23 corresponds to an example of an analysis means. As described above, in this embodiment, the information processing device 10 and the analysis device 23 are connected to be able to communicate with each other. However, the information processing device 10 may be provided with the analysis device 23 inside, or the information processing device 10 may be provided with the analysis device 23 inside. Furthermore, the analysis result (spectral information) may be transferred from the analysis device 23 to the information processing device 10 via a recording medium such as a non-volatile memory.

The analysis device 23 in this embodiment can be any device capable of acquiring spectral information by secondary ion mass spectrometry. Examples of devices capable of acquiring spectral information by secondary ion mass spectrometry include a time-of-flight secondary ion mass spectrometer (TOF-SIMS). TOF-SIMS will be described in detail below.

In TOF-SIMS, when a sample is irradiated with an ion beam (primary ions), elements and molecules are released from the sample surface. Examples of primary ions include Ga ⁺ , In ⁺ , Au ⁺ , and Bi ³⁺ . The primary ions are shaped into pulses with a width of 0.7 nsec or less and a repetition frequency of several kHz to 50 kHz, and then irradiated onto the sample. Secondary ions extracted from the sample by the extraction electric field fly at a constant speed in the field-free space between the grid and the detector, and are then detected by a time-of-flight mass spectrometer. Since the flight speed at that time differs depending on the mass of the secondary ions, mass analysis is performed based on the difference in the arrival time to the time-of-flight mass spectrometer. The reflection type time-of-flight mass spectrometer uses a reflection type flight (drift) tube to perform mass analysis based on the time it takes to fly a distance of 2 m.

The information processing device 10 has, as its functional configuration, a communication IF 31, a ROM 32, a RAM 33, a memory unit 34, an operation unit 35, a display unit 36, and a control unit 37.

The communication IF (Interface) 31 is realized, for example, by a LAN card and a USB interface card. The communication IF 31 manages communication between the information processing device 10 and external devices (for example, the database 22 and the analysis device 23) via the LAN 21 and the USB. The ROM (Read Only Memory) 32 is realized by a non-volatile memory or the like, and stores various programs, etc. The RAM (Random Access Memory) 33 is realized by a volatile memory or the like, and temporarily stores various information. The storage unit 34 is realized, for example, by a HDD (Hard Disk Drive) or the like, and stores various information. The operation unit 35 is realized, for example, by a keyboard or a mouse, and inputs instructions from the user into the device. The display unit 36 is realized, for example, by a display, etc., and displays various information to the user. The operation unit 35 and the display unit 36 provide functions as a GUI (Graphical User Interface) under the control of the control unit 37.

(Control Unit)
The control unit 37 is realized by, for example, at least one CPU (Central Processing Unit) or the like, and controls the processing in the information processing device 10. The control unit 37 includes, as its functional configuration, a spectrum information acquisition unit 41, a learning model generation unit 42, a learning model acquisition unit 43, an estimation unit 44, an information acquisition unit 45, and a display control unit 46.

(Spectral information acquisition unit 41)
The spectral information acquisition unit 41 acquires the analysis results of the target sample including a solid sample and a substance to be measured, specifically, the spectral information of the target sample, from the analysis device 23. The spectral information of the target sample may be acquired from the database 22 in which the analysis results are stored in advance. Similarly, the spectral information of the solid sample and the spectral information of the sample including the solid sample and the substance to be measured are acquired. This spectral information of the solid sample is the spectral information when a single solid substance exists. The spectral information of the sample including the solid sample and the substance to be measured is, for example, the spectral information obtained from each sample prepared by preparing multiple types of samples including different amounts of the substance to be measured, such as 0.2%, 0.4%, etc.

Then, the spectral information acquisition unit 41 outputs the acquired spectral information of the sample to the estimation unit 44. In addition, the spectral information of the acquired sample and the spectral information of the solid sample are output to the learning model generation unit 42.

(Learning model generation unit 42)
The learning model generation unit 42 generates teacher data using the solid sample and the spectral information of the sample acquired by the spectral information acquisition unit 41. Then, the learning model generation unit 42 executes deep learning using the teacher data to generate a learning model. A detailed description of the generation of teacher data and the generation of a learning model will be given later. Then, the learning model generation unit 42 outputs the generated learning model to the learning model acquisition unit 43. Note that the learning model generation unit 42 may output the generated learning model to the database 22.

(Learning model acquisition unit 43)
The learning model acquisition unit 43 acquires the learning model generated by the learning model generation unit 42. When the learning model is stored in the database 22, the learning model acquisition unit 43 acquires the learning model from the database 22. Then, the learning model acquisition unit 43 outputs the acquired learning model to the estimation unit 44.

(Estimation unit 44)
The estimation unit 44 inputs the spectral information of the target sample acquired by the spectral information acquisition unit 41 into the learning model acquired by the learning model acquisition unit 43, thereby causing the learning model to estimate quantitative information of the measurement target substance contained in the target sample. Then, the estimation unit 44 outputs the estimated quantitative information to the information acquisition unit 45. The estimation unit 44 corresponds to an example of an estimation means that estimates quantitative information of the measurement target substance by inputting the spectral information of the target sample into the learning model.

(Information Acquisition Unit 45)
The information acquisition unit 45 acquires quantitative information estimated by the learning model. In other words, the information acquisition unit 45 corresponds to an example of an information acquisition means for acquiring quantitative information of the measurement target substance estimated by inputting the spectral information of the target sample to the learning model. Then, the information acquisition unit 45 outputs the acquired quantitative information to the display control unit 46.

(Display control unit 46)
The display control unit 46 causes the display unit 36 to display the quantitative information acquired by the information acquisition unit 45. The display control unit 46 corresponds to an example of a display control means.

Note that at least some of the components of the control unit 37 may be implemented as independent devices. Also, each component may be implemented as software that realizes its function. In this case, the software that realizes the function may run on a server via a network such as the cloud. In this embodiment, each component is implemented by software in a local environment.

The configuration of the information processing system shown in FIG. 1 is merely an example. For example, the memory unit 34 of the information processing device 10 may have the functionality of the database 22, and the memory unit 34 may store various types of information.

Next, the processing procedure for generating a learning model in this embodiment will be described with reference to FIG. 2. FIG. 2 is a diagram showing an example of a flowchart of the processing procedure for generating a learning model according to this embodiment.

(S201: Analyze a solid sample alone and a sample containing a solid sample and a substance to be measured)
First, the analyzer 23 analyzes a solid sample alone and a sample containing a solid sample and a substance to be measured (step S1). The analysis conditions may be appropriately selected from the viewpoints of sensitivity, analysis time, etc. When analyzing a sample containing a solid sample and a substance to be measured, the concentration of the substance to be measured is changed in several ways. The number of changes required varies depending on the properties of the substance, but it is generally preferable to change the concentration at three or more points. When there are multiple types of substances to be measured, it is preferable to analyze each of them, but if the signals of the substances to be measured can be sufficiently separated, they may be measured simultaneously.

Then, the analysis device 23 outputs the acquired spectral information to the information processing device 10. The information processing device 10 receives the spectral information from the analysis device 23 and stores it in the RAM 33 or the storage unit 34. The spectral information acquisition unit 41 acquires the spectral information thus stored.

As described above, the spectral information resulting from the analysis may be stored in the database 22. In this case, the spectral information acquisition unit 41 acquires the spectral information from the database 22. Furthermore, the timing at which the analysis device 23 analyzes the test substance may be any timing as long as it is performed before the selection of the spectral information in step S3.

(S202: Generate teacher data)
The learning model generation unit 42 generates teacher data using the spectral information of the solid sample alone and the sample containing the solid sample and the measurement target substance obtained in step S201. At that time, the teacher data is generated based on the first spectral information and the second spectral information. The first spectral information and the second spectral information will be described in detail below.

The first spectral information refers to spectral information of the solid sample alone, spectral information of a sample containing a solid sample and a substance to be measured, and spectral information created using a noise spectrum.

The second spectral information is created from the spectral information of samples with different concentrations of the substance to be measured, extracted from the first spectral information group. Specifically, the concentration of the substance to be measured is between the concentrations of the two extracted first spectral information. Hereinafter, the created spectrum will be referred to as an artificial spectrum, and the method of creating this spectrum will be referred to as concentration interpolation.

Next, we will explain how to create training data with noise added. Spectral information consists of the ion intensity measured for each fragment, or the ion intensity normalized by dividing by the sum of the ion intensities of fragments within a specified m/z range, and the numerical value is measured for each m/z obtained by dividing the molecular weight m by the ion charge z.

The ion intensity in the spectral information of the target substance is preferably less than 7×10 ⁵ relative to the sum of the fragment ion intensities with m/z of 10.0 or more in the spectral information of the solid sample and the sample containing the target substance. Even when the ion intensity of the spectral information of the target substance is as small as less than 7×10 ⁵ , it is possible to easily and highly accurately obtain quantitative information of the target substance by using the information processing device according to this embodiment.

Fragments with m/z less than 10.0 are composed of hydrogen atoms and one or two other atoms, so they are present in large quantities and have excessive ionic strength. As a result, changes in fragments with m/z greater than 10.0 are overlooked. For this reason, it is desirable to handle only spectral information with m/z greater than 10.0.

On the other hand, the maximum m/z value of the fragments to be handled is accurate if it is set to the molecular weight of the object to be measured. However, if the maximum m/z value is large, the amount of teacher data to be handled increases, and the learning time is lengthened, so it is preferable that the maximum value can be set by the user according to the expected prediction accuracy. Also, as described later in the embodiment, the larger the maximum m/z value (m/z range), the higher the prediction accuracy of the concentration of the substance to be measured, and the smaller the maximum m/z value (m/z range), the lower the prediction accuracy of the concentration of the substance to be measured. On the other hand, the larger the maximum m/z value (m/z range), the longer the calculation time for prediction, and the smaller the maximum m/z value (m/z range), the shorter the calculation time for prediction. Therefore, the device may have a setting unit (not shown) for the user to set the maximum m/z value (or the m/z range) taking into account the prediction accuracy desired by the user and the calculation time required to obtain the result. Alternatively, the device may be configured to automatically set the maximum m/z value (or the m/z range) based on the prediction accuracy desired by the user and the calculation time required to obtain the result. When setting the m/z range, at least one of the minimum and maximum values of m/z can be set. Furthermore, the device may have a display control unit that displays at least one of the minimum, maximum, and range of m/z on the display unit.

In this embodiment, in the spectral information of the solid sample alone, each ion intensity is multiplied by an arbitrary numerical value between 0.1 and 5.0. The spectrum multiplied by this arbitrary numerical value is called a noise spectrum, and this multiplication process is described as adding noise. Next, a new ion intensity is created by averaging values multiplied by ion intensities having the same m/z value, averaging values multiplied by an arbitrary numerical value, or using the largest value in the spectral information of a sample containing a solid sample and a substance to be measured. This process is performed for all m/z values, and the first spectral information is created.

When generating training data, it is up to you to perform either density interpolation or noise addition first.

(S203: Generate learning model)
Using the generated teacher data, machine learning according to a predetermined algorithm is performed to construct a learning model (step S203). As a specific learning method, for example, a neural network or a support vector machine, which is a general machine learning method, may be used. In addition, as a deep learning method with multiple hidden layers, a deep neural network (DNN) or a convolutional neural network (CNN) may be used. When there are multiple substances to be measured, a learning model is generated for each substance. Then, the learning model generation unit 42 stores the generated learning model in the RAM 33, the storage unit 34, or the database 22.

In this way, a learning model is generated that estimates quantitative information about the target substance contained in the target sample based on the spectral information of the solid sample alone and the sample containing the solid sample and the target substance.

Next, a method for displaying quantitative information will be described. FIG. 3 is a diagram showing an example of a flowchart of a processing procedure for displaying quantitative information according to this embodiment.

(S301: Analyze the target sample)
The analysis device 23 analyzes a target sample containing a solid sample and a substance to be measured, and acquires spectral information of the target sample. The analysis conditions are the same as those in step S201 described above. The analysis device 23 then outputs the acquired spectral information to the information processing device 10. The information processing device 10 receives the spectral information from the analysis device 23, and stores it in the RAM 33 or the storage unit 34. The spectral information acquisition unit 41 acquires the spectral information thus stored. As described above, the spectral information that is the analysis result may be stored in the database 22. In this case, the spectral information acquisition unit 41 acquires the spectral information from the database 22. The timing at which the analysis device 23 analyzes the sample may be any timing as long as it is performed before the estimation of quantitative information in step S302.

(S302: Estimate quantitative information)
The learning model acquisition unit 43 acquires the learning model stored in the RAM 33, the storage unit 34, or the database 22. Then, the estimation unit 44 inputs the spectral information of the target sample acquired in step S301 into the acquired learning model, thereby estimating quantitative information of the measurement target substance contained in the target sample. In addition, as necessary, the estimation unit 44 converts the estimated quantitative information into a format to be displayed on the display unit 36. The format to be displayed on the display unit 36 may be a concentration such as g/L or mol/L, or a ratio to a reference amount (standard amount). If the value estimated by the learning model is in these display formats, conversion is not necessary. Then, the information acquisition unit 45 acquires the estimated quantitative information from the estimation unit 44 and stores it in the RAM 33 or the storage unit 34.

(S303: Display quantitative information)
The display control unit 46 causes the display unit 36 to display quantitative information on the measurement target substances contained in the target sample, which is estimated by the learning model in step 302. At this time, the information may be displayed in a graph or table format. An example of a screen (window) displayed on the display unit 36 is shown in Fig. 7. Furthermore, the level may be displayed according to the numerical value of the estimated concentration, such as "high" or "low".

The present invention can also be realized by supplying a program that realizes one or more of the functions of the above-mentioned embodiments to a system or device via a network or storage medium, and having one or more processors in the computer of the system or device read and execute the program. It can also be realized by a circuit (e.g., an ASIC) that realizes one or more functions.

<Example>
The present invention will be described in more detail below with reference to examples and comparative examples, but the present invention is not limited to the following examples.

Example 1
[Preparation of solid samples]
Methyl methacrylate as a monomer and trimethylolpropane methacrylate as a crosslinking agent (both manufactured by Tokyo Chemical Industry Co., Ltd.) were mixed in amounts of 49.5 g each. Furthermore, 1 g of Omnirad 819 (manufactured by IGM Resins B.V.) as a polymerization initiator was added to the mixed solution. Next, 1 g was taken from the obtained solution and stirred for 5 minutes using a mixer (Thinky Mixer AR-100). Then, 0.3 g was taken from the stirred solution and sandwiched between fluororesin films (Neoflon PFA film, manufactured by Daikin Industries), and mercury lamps (UL750 manufactured by HOYA) were irradiated from a distance of 30 cm from the fluororesin film for 5 minutes. This resulted in a photocured resin, which was used as a solid sample.

[Preparation of a sample containing a solid sample and a substance to be measured]
1 g of the stirred liquid obtained in the preparation of the solid sample was taken, and 0.001 g of a surfactant (Acetylenol E100, manufactured by Kawaken Fine Chemicals) was added as a plasticizer. This produced a sample containing 0.1% of the surfactant, which is the substance to be measured. The method for obtaining a photocurable resin from the solution obtained by adding the surfactant was the same as that for preparing the solid sample.

Furthermore, by changing the amount of surfactant added, samples containing 0.2%, 0.4%, 0.6%, 0.8%, 1.0%, and 10.0% of the surfactant, the substance to be measured, were prepared. Apart from changing the amount of surfactant added, the samples were prepared in the same manner as the sample containing 0.1% surfactant.

[Measurement by TOF-SIMS method]
The prepared solid sample and a sample containing the solid sample and the substance to be measured were each cut into a 1 cm square, and measurement was performed by TOF-SIMS under the following conditions.
Measurement model: TOF-SIMS5 (ION-TOF)
Primary ion species: ^Bi
Primary ion current: 0.4 pA (Pulse current)
Primary ion energy: 25 keV
Measurement mode: Bunching Mode
Detection range: 1 to 800 amu
Electron gun: Electron gun for charge correction used Measured vacuum level: 7×10 ⁻⁸ Pa
Measurement area: 300μm x 300μm
Number of measurement points: 128 x 128 points within the measurement area. Accumulation: 16 times. Measurement mode: Negative ions. Number of measurements: 6 times for each sample.

The above measurements yielded the group of spectra shown in Figure 4. Figure 4 shows an example of spectral information for a solid sample and a sample containing a solid sample and a substance to be measured.

(Comparative Example 1)
The spectrum of the solid sample alone was compared with a spectrum containing 10.0% of the substance to be measured, and the fragments derived from the substance to be measured were estimated. The ion intensities of these fragments and values normalized by the sum of the ion intensities of all fragments with m/z of 10 to 700 are shown in Table 1. Fragments with m/z less than 10 were excluded because the ion intensities of fragments in this range (e.g., H ⁺ ) are too large compared to the ion intensities of other fragments, and the measurement error thereof would have a significant effect on the accuracy of the normalized value.

Next, in Figure 5, the content of the target substance in the solid sample was compared for the fragments, but the ion intensity of each fragment did not show a tendency to increase with increasing concentration, making it difficult to create a calibration curve. The reasons for this are as follows.

The ion intensity of the fragments obtained from the solid sample is less than about 7.0×10 ⁻⁵ . The ion intensity of the fragments obtained from the sample containing the solid sample and the substance to be measured is also less than about 7.0×10 ⁻⁵ . Therefore, the fragment information released from the substance to be measured is buried in the variation in the fragment information released from the solid sample. As a result, it was found that quantitative analysis by human effort is difficult.

Example 2
A method for creating the teacher data will be described with reference to the flowchart of FIG.

[Creation of training data by adding noise in Example 2]
Using the spectral information measured in Example 1, it was prepared in the following manner.

The ion intensities of the spectra of the solid sample alone (six sets in total) were compared for each m/z, and the ion intensities with the highest values were extracted. These highest ion intensities were combined for m/z values from 10 to 700, and then multiplied by any randomly selected value between 0.1 and 5.0 for each m/z value to create a noise spectrum.

Next, the ion intensities of this noise spectrum and the spectrum of the solid sample alone were compared for each m/z, and the ion intensities with the highest values were extracted. An artificial spectrum was created by combining the groups of ion intensities with the highest values from m/z 10 to 700.

The above procedure was repeated 10 times for each of the six sets of spectra for the solid samples, resulting in a total of 60 sets of training data consisting of artificial spectra that do not contain the substance to be measured. Note that the arbitrary values were updated to new values each time (step: S401).

Furthermore, the ion intensity was compared for each m/z for the noise spectrum and the spectrum of the sample containing 0.2% of the substance to be measured, and the ion intensity with the highest value was extracted. An artificial spectrum was created by combining the ion intensity groups with the highest values from m/z 10.0 to 700.0. This process was performed 10 times for each of the six sets of spectra of samples containing 0.2% of the substance to be measured, and a total of 60 sets of training data consisting of artificial spectra of samples containing 0.2% of the substance to be measured were created (step: S402).

Next, for the spectra containing 0.6% and 1.0% of the target substance, a similar process was carried out to create training data consisting of artificial spectra of samples containing 0.2% of the target substance, and 60 sets of training data were created for each.

As a result, a total of 240 sets of training data were created from the first group of spectra (steps: S403 and S404).

[Creating training data by density interpolation]
One spectrum was extracted from each of the teacher data (60 sets in total) consisting of the spectrum of the solid sample alone and the teacher data (60 sets in total) consisting of the spectrum of the solid sample containing 0.2% of the target substance. Furthermore, using formulas (1) to (3) for each m/z, spectra (artificial spectra) of samples containing 0.05%, 0.1%, and 0.15% of the target substance were created.

Here, I ₀ is the ion intensity of the teacher data consisting of the spectrum of the solid sample alone, and I _n is the ion intensity of the spectrum containing n% of the measured substance. The above operation was performed on 60 sets of teacher data consisting of the spectrum of the solid sample alone and 60 sets of teacher data containing 0.2% of the measured substance in a brute force manner, obtaining 60 x 60 x 3 = 10,800 pieces of teacher data (step: S405).

Using a similar method, a total of 10,800 sets of training data were created from a training data group containing 0.2% of the substance to be measured and a training data group containing 0.6% of the substance to be measured, each containing 0.3%, 0.4%, and 0.5% of the substance to be measured.

Furthermore, using a similar method, a total of 10,800 sets of training data containing 0.7%, 0.8%, and 0.9% of the target substance were created from a training data group containing 0.6% of the target substance and a training data group containing 1.0% of the target substance. As a result, a total of 32,400 sets of training data were created from the second spectrum group (steps S406 and S407).

A total of 32,640 pairs were used as training data, and machine learning was performed to generate a learning model. A fully connected neural network was used as the machine learning method, with the relu function and linear function used as the activation function. The mean squared error was used as the loss function, and Adam was used as the optimization algorithm. Repeated calculations of around 1,000 epochs were required to obtain sufficient quantitative accuracy.

Next, six sets of spectra were prepared for samples containing 0.1%, 0.4%, and 0.8% of the target substance, and these were applied to the resulting learning model to predict the concentration of the target substance contained in the sample. As shown in Table 2, the concentration of the target substance was predicted with high accuracy.

Example 3
A learning model was created in the same manner as in Example 2, except that the noise-adding method described in Example 2 was not adopted. Specifically, the method is as follows.

The first spectral information consisted of 24 sets of spectra, including six sets of spectra for the solid sample alone and six sets of spectra for samples containing 0.2%, 0.6%, and 1.0% of the substance to be measured.

[Creating training data by density interpolation]
One spectrum was extracted from each of the six sets of spectra of the solid sample alone and the six sets of spectra of the sample containing 0.2% of the substance to be measured. Furthermore, the above-mentioned formulas (1) to (3) were used for each m/z to create spectra (artificial spectra) of samples containing 0.05%, 0.1%, and 0.15% of the substance to be measured.

This process was performed on six sets of spectra from solid samples alone and six sets of spectra from samples containing 0.2% of the substance to be measured, yielding 6 x 6 x 3 = 108 pieces of training data.

Using a similar method, a total of 108 sets of training data were created from six sets of spectra containing 0.2% of the target substance and six sets of spectra containing 0.6% of the target substance, each containing 0.3%, 0.4%, and 0.5% of the target substance, respectively.

In addition, using a similar method, a total of 108 sets of training data were created from six sets of spectra containing 0.6% of the measured substance and six sets of spectra containing 1.0% of the measured substance, containing 0.7%, 0.8%, and 0.9% of the measured substance, respectively. From the above, a total of 324 sets of second spectral information were created.

As described above, 348 pairs of the first and second spectrometers were used as training data, and machine learning was performed to generate a learning model. A fully connected neural network was used as the machine learning method, with the relu function and linear function used as the activation function. The mean squared error was used as the loss function, and Adam was used as the optimization algorithm. Repeated calculations of about 1,000 epochs were required to obtain sufficient quantitative accuracy.

Next, six sets of spectra were prepared for samples containing 0.1%, 0.4%, and 0.8% of the target substance, and these were applied to the resulting learning model to predict the concentration of the target substance contained in the sample. As shown in Table 3, the concentration of the target substance was predicted with high accuracy.

Example 4
Other than the method of generating training data by adding noise, the method is the same as in Example 2. In this example, the maximum m/z value was set to 300.0 in order to balance calculation accuracy and calculation time.

[Creating training data by adding noise in the fourth embodiment]
The procedure is the same as in Example 2, except that when comparing the ion intensities of the noise spectrum and the spectrum of the solid sample alone for each m/z, an artificial spectrum was created by combining ion intensities from m/z 10.0 to 300.0 when extracting the ion intensity with the highest value.

Next, six sets of spectra were prepared for samples containing 0.1%, 0.4%, and 0.8% of the target substance, and these were applied to the resulting learning model to predict the concentration of the target substance contained in the sample. As shown in Table 4, the concentration of the target substance was predicted with high accuracy.

Example 5
Other than the method of generating training data by adding noise, the method is the same as in Examples 2 and 4. In this example, the maximum m/z value was set to 100.0 in order to reduce the calculation time.

[Creation of training data by adding noise in embodiment 5]
The procedure was the same as in Example 4, except that an artificial spectrum was created by combining ion intensity groups with m/z values ranging from 10.0 to 100.0.

Next, six sets of spectra were prepared for samples containing 0.1%, 0.4%, and 0.8% of the target substance, and these were applied to the resulting learning model to predict the concentration of the target substance contained in the sample. As shown in Table 5, the concentration of the target substance was predicted with high accuracy.

In addition, although Example 5 has a lower prediction accuracy of the concentration of the substance to be measured than Examples 2, 3, and 4, the time required for prediction is short. Therefore, if you want to increase the prediction accuracy, it is preferable to increase the range of z/m (increase the maximum value), and if you want to shorten the calculation time, it is preferable to decrease the range of z/m (decrease the maximum value).

(Comparative Example 2)
A learning model was created in the same manner as in Example 3, except that the teacher data was not created by the density interpolation in Example 3. Specifically, the method is as follows.

The training data consisted of six sets of spectra from a solid sample alone, six sets of spectra from a sample containing 0.2% of the substance to be measured, six sets of spectra from a sample containing 0.6% of the substance to be measured, and six sets of spectra from a sample containing 1.0% of the substance to be measured.

A total of 24 sets of these were used as training data, and machine learning was performed to generate a learning model. A fully connected neural network was used as the machine learning method, with the relu function and linear function used as the activation function. The mean squared error was used as the loss function, and Adam was used as the optimization algorithm. Repeated calculations of around 200 epochs were required to obtain sufficient quantitative accuracy.

Next, six sets of spectra were prepared for samples containing 0.1%, 0.4%, and 0.8% of the target substance, and these were applied to the resulting learning model to predict the concentration of the target substance contained in the sample. As shown in Table 6, it was not possible to predict the concentration of the target substance with high accuracy.

Claims (16)

A learning model acquisition unit for acquiring a learning model;
an information acquisition unit that acquires quantitative information of a target substance to be measured, the quantitative information being output from the learning model by inputting spectral information of a target sample containing the target substance to the learning model,
the learning model is generated using teacher data based on a plurality of pieces of spectral information of a solid sample, a plurality of pieces of spectral information of a solid sample containing the target substance, the plurality of pieces of spectral information each having a different concentration of the target substance in the solid sample, and a plurality of pieces of spectral information obtained by adding noise to the spectral information of the solid sample;
The information processing device, wherein the spectral information is obtained by secondary ion mass spectrometry. The information processing device according to claim 1, wherein the secondary ion mass spectrometry is time-of-flight secondary ion mass spectrometry. 3. The information processing device according to claim 1 or 2, wherein the quantitative information is at least one type selected from the group consisting of the amount of the target substance in the target sample, the concentration of the target substance in the target sample, the presence or absence of the target substance in the target sample, and the ratio of the amount of the target substance in the target sample to a reference amount of the target substance. The information processing device according to any one of claims 1 to 3, wherein the solid sample includes at least one selected from the group consisting of a resin, a silicon substrate, and a copper foil. The information processing device according to claim 4, wherein the solid sample includes a resin. The information processing device according to claim 5, wherein the resin has units derived from unsaturated carboxylic acid esters. The information processing device according to any one of claims 1 to 6, wherein the substance to be measured is at least one selected from the group consisting of plasticizers, thickeners, viscosity reducers, crystal nucleating agents, crystallization accelerators, crystallization retarders, release agents, flame retardant agents, flame retardant assistants, thermal decomposition inhibitors, reinforcing agents, conductive materials, antistatic agents, thermal conductive agents, antioxidants, weathering inhibitors, ultraviolet absorbers, antibacterial agents, colorants, gas barrier agents, antifogging agents, antireflective agents, antifouling agents, and impurities. The information processing device according to claim 7, wherein the substance to be measured is a plasticizer. The information processing device according to claim 8, wherein the plasticizer is a surfactant. The information processing device according to any one of claims 1 to 9, further comprising a display control means for displaying the acquired quantitative information on a display unit. 11. The information processing apparatus according to claim 1, wherein a group of elements constituting the solid sample is the same as a group of elements constituting the substance to be measured. 12. The information processing device according to claim 1, wherein the group of elements constituting the solid sample and the group of elements constituting the substance to be measured include at least one selected from the group consisting of carbon atoms, hydrogen atoms, oxygen atoms, and nitrogen atoms. a learning model generating unit that generates the learning model; and a setting unit that sets at least one of a minimum value and a maximum value of m/z of spectral information of the solid sample and a sample containing the measurement target substance,
The information processing device according to any one of claims 1 to 12, wherein the learning model generation unit generates the learning model using spectral information of the solid sample in the range of m/ z (the value obtained by dividing the molecular weight m by the ionic valence z) set by the setting unit, and spectral information of a sample containing the solid sample and the substance to be measured. The information processing device according to claim 13 , further comprising a display control unit that causes a display unit to display at least one of the minimum value, the maximum value, and the range of the m/z. A learning model acquisition step of acquiring a learning model;
and an information acquiring step of acquiring quantitative information of the measurement target substance output from the learning model by inputting spectral information of a target sample containing the measurement target substance into the learning model,
the learning model is generated using teacher data based on a plurality of pieces of spectral information of a solid sample, a plurality of pieces of spectral information of a solid sample containing the target substance, the plurality of pieces of spectral information each having a different concentration of the target substance in the solid sample, and a plurality of pieces of spectral information obtained by adding noise to the spectral information of the solid sample;
An information processing method, wherein the spectral information is obtained by secondary ion mass spectrometry. 15. A program for causing a computer to function as each of the means of the information processing apparatus according to claim 1.

Images