Mixture component identification method based on terahertz spectrum detection database search

Abstract

The invention discloses a mixture component identification method based on terahertz spectrum detection database search, which combines terahertz spectrum and database search technologies, and can efficiently and accurately identify each component in a complex mixture through matching with known spectrogram information. Comprising the following steps: establishing a high-quality terahertz spectrum database; collecting terahertz spectrum data of a to-be-detected sample, wherein the terahertz spectrum data comprises fingerprint information of the sample; the collected spectral data is preprocessed to improve the data quality; comparing the terahertz spectrum of the sample to be detected with spectral data in a database through a database search and spectrogram comparison step so as to realize qualitative analysis of each component in the mixture; and estimating the relative content of each component by using the intensity information in the terahertz spectrum, thereby completing the quantitative analysis of the mixture. According to the method disclosed by the invention, the terahertz spectrum is combined with database search, and unique substance fingerprint characteristics in the terahertz spectrum and a modern data analysis technology are fully utilized, so that qualitative and quantitative analysis can be systematically performed on a multi-component mixture; and a reliable and efficient technical means is provided for mixture analysis of the terahertz spectrum in practical application. The method can effectively improve the efficiency and precision of component identification, provides a new thought and method for mixture component analysis, and has a wide application prospect.

Description

Technical Field

[0001] The present invention relates to the technical field of spectral analysis, and specifically relates to a method for qualitative and quantitative analysis of the components of a mixture by combining terahertz spectroscopy with a database. Background Art

[0002] The frequency interval of terahertz (THz) spectroscopy ranges from 0.1 to 10 THz. Due to its low photon energy (4 meV@1 THz), THz spectroscopy has characteristics such as non-destructiveness, good accuracy, sensitivity, and rapidity, and has been widely used for identifying substances. With the rapid development of terahertz technology, the unique vibration characteristic signals in THz spectroscopy characterize the "fingerprint" of single-component molecular substances. These unique vibration characteristics of substances in the terahertz band are related to the intermolecular collective vibration modes of the hydrogen bond network and the low-frequency intramolecular vibration modes of molecules. Therefore, these characteristics are highly sensitive to the structure and conformation of molecules and the surrounding environment. Various substances can be directly and rapidly identified according to their unique THz characteristics.

[0003] In chemical analysis, the identification of the components of a mixture has always been an important and challenging problem. Although traditional spectral analysis methods can effectively identify certain chemical substances, there are certain limitations in the identification of the components of complex mixtures. In recent years, terahertz spectroscopy has gradually become an effective analysis means due to its unique substance identification ability. Terahertz spectroscopy can provide unique vibration information between molecules and is suitable for non-destructive analysis of various substances. Using terahertz spectroscopy for qualitative and quantitative analysis of mixtures has become an important application field.

[0004] However, due to the complex components, in addition to having a wide absorption range in the terahertz band, multi-component mixtures also have the phenomena of component spectral overlap and baseline drift. It is difficult to directly obtain the composition and content of a mixture through THz spectroscopy. Therefore, there is an urgent need to develop a new identification method with the ability to identify unknown relationships. Therefore, developing a method for identifying the components of a mixture based on terahertz spectroscopy detection database search, which combines terahertz spectroscopy with database search, by constructing a terahertz spectroscopy database containing a large number of chemical substances and applying modern data analysis techniques, is expected to significantly improve the accuracy and efficiency of mixture component analysis. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for identifying the components of a mixture based on terahertz spectroscopy detection database search, which can efficiently and accurately identify each component in the mixture by matching terahertz spectroscopy data with the known spectral information in the database.

[0006] A method for identifying the components of a mixture based on terahertz spectroscopy detection database search is carried out according to the following steps: Step 1: Establish a high-quality terahertz spectroscopy database, clarify the objectives and usage scenarios of the database. The database includes spectral data of various common chemical substances. Each sample in the database should include detailed spectral data, chemical composition, experimental conditions, and relevant spectral feature information.

[0007] Step 2: Terahertz spectroscopy data acquisition. Collect the terahertz signals of the sample to be measured, and generate terahertz spectra that interact with the molecular substances, which contain the fingerprint information of the sample.

[0008] Step 3: Data preprocessing. Preprocess the original terahertz spectroscopy data of the sample to be measured to improve the data quality and the accuracy of subsequent analysis. Remove the influence of problems such as noise and baseline drift.

[0009] Step 4: Database search and spectral comparison. Compare the terahertz spectrum of the sample to be measured with the data in the database, and use the spectral similarity measurement method and spectral matching algorithm to evaluate the matching degree between the spectrum of the sample to be measured and each spectrum in the database.

[0010] Step 5: Identification and quantitative analysis of mixture components. Adopt the multi-component analysis method to deconstruct the spectral data of the sample, so as to effectively separate the spectral signals of different components. Use the intensity information in the terahertz spectrum to estimate the relative content of each component in the mixture. Finally, summarize the test results and identify the components of the mixture.

[0011] The method for identifying mixture components based on terahertz spectroscopy detection database search provided by the present invention can systematically conduct qualitative and quantitative analysis on multi-component mixtures, and provide reliable and effective analysis and identification techniques for mixture analysis in the practical application of terahertz spectroscopy.

[0012] In order to obtain high-quality terahertz spectroscopy data, in Step 1, to establish accurate reference spectra, collect the terahertz spectra of single chemical substances in a pure state. In addition to single chemical substances, it is also necessary to collect the spectral data of mixtures to reflect the component characteristics in actual samples.

[0013] Preferably, the data in the database in Step 1 includes existing literature data or commercial standard databases and terahertz spectra obtained from experiments. Further, the components of the collected mixtures include standard compounds with known proportions or mixed samples of chemical substances with known various mixing ratios.

[0014] As a further preference, the terahertz spectra of solid chemical substances are collected by the method of pressing tablets. The terahertz spectra of liquid chemical substances are collected through liquid cells, and attention should be paid to controlling the thickness of the liquid. The terahertz spectra of gaseous chemical substances are collected in gas containers.

[0015] Preferably, the frequency acquisition range of the terahertz spectrum is from 0.1 THz to 3 THz, and each test sample should be collected at least three times to ensure the reliability of the results.

[0016] Preferably, after the spectral data is collected, the data is verified to confirm the accuracy and consistency of the spectrogram.

[0017] When establishing the database, the design of the data structure should include information such as sample identification, chemical name and molecular formula, spectral data, and spectral characteristics of experimental conditions. The database should be updated and expanded regularly to increase its coverage and accuracy.

[0018] Preferably, when collecting the terahertz spectral data of the sample to be tested in step 2, the experimental environment should be the same as that when establishing the database in step 1.

[0019] Preferably, the terahertz spectral data of the sample to be tested collected in step 2 is Fourier-transformed into frequency-domain spectral data.

[0020] Preferably, in step 3, the terahertz data of the sample to be tested is preprocessed to improve the data quality and the accuracy of subsequent analysis. The noise components in the signal are removed by the Gaussian smoothing method; polynomial fitting is used for baseline correction to remove the baseline drift caused by the equipment or experimental environment.

[0021] As a further preference, the spectral signal is smoothed to eliminate unnecessary fluctuations, and the spectrum is normalized to ensure data consistency.

[0022] Preferably, in step 4, similarity measurement methods such as Pearson correlation coefficient, Euclidean distance, and cosine similarity are combined to evaluate the matching degree between the spectrum of the sample to be tested and each spectrum in the database.

[0023] Preferably, in step 4, a matching degree threshold is set, and by calculating the correlation between the spectrum to be tested and the database spectrum, the most similar spectrogram is selected as the candidate component.

[0024] Preferably, in step 5, the principal component analysis (PCA) method is used to deconstruct the spectral data of the sample and separate the spectral signals of different components.

[0025] Preferably, in step 5, the standard curve method is used to establish a standard curve between the spectral intensity and the concentration according to the standard sample with a known concentration. The concentration of each component is estimated through the spectral data of the sample to be tested.

[0026] Compared with the existing technology, the method for detecting melamine content based on terahertz spectroscopy of the present invention has the following beneficial effects:

[0027] (1) The method of the present invention creates a dedicated database that can be continuously enriched, improved, and used for a long time. The present invention directly and rapidly analyzes mixtures using terahertz spectroscopy, without the need for sample pretreatment or separation processes, enabling rapid and real-time component identification and greatly improving the analysis efficiency.

[0028] (2) Terahertz spectroscopy has unique spectral characteristics in the characteristic absorption frequency bands of many chemical substances. By comparing the spectra with those of known substances in the database, it is possible to accurately identify and quantitatively analyze multiple components in a mixture.

[0029] (3) At the same time, it can identify and quantitatively analyze multiple components in a mixture in a single analysis, making the analysis of complex samples more comprehensive and in-depth. It can cover various chemical substances and has broad application prospects.

[0030] Generally speaking, the method for identifying the components of a mixture based on terahertz spectroscopy detection database search provided by the present invention has significant advantages in component identification, quantitative analysis, and real-time detection due to its high efficiency, non-destructive, high sensitivity, simple and easy-to-use features. It is particularly suitable for rapid, accurate, and low-cost chemical component analysis and has broad application prospects in many fields such as scientific research, industry, and environmental monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 : Flowchart of the method for identifying the components of a mixture based on terahertz spectroscopy detection database search.

[0032] Figure 2 Absorption coefficient spectra of aspirin, ibuprofen, and a mixed sample.

[0033] Figure 3 Prediction result graph of sample concentration. DETAILED DESCRIPTION OF THE INVENTION

[0034] To make the objectives, technical features, and advantages of the embodiments of the present invention clearer and more definite, the specific embodiments of the present invention will be further described in detail below with reference to the drawings. In the present invention, unless otherwise specified, the devices, reagents, methods, etc. used are all conventional devices, reagents, and methods in the art.

[0035] Figure 1 is the flowchart of the method for identifying the components of a mixture based on terahertz spectroscopy detection database search of the present invention. As Figure 1 shown, in this embodiment, a method for identifying the components of a mixture based on terahertz spectroscopy detection database search includes the following steps: Step 1. To achieve efficient identification of mixture components, a high-quality terahertz spectroscopy database needs to be established. The database target of this experiment is for the analysis of chemicals, especially mixture samples containing multiple common chemical components. The database includes a self-collected terahertz database of chemicals and an online database that can be publicly viewed. This database contains multiple terahertz spectra of common inorganic materials, agrochemicals, biochemistry, pharmaceutical substances (including sugars), and polymers. Each sample in the database includes detailed spectral data, chemical composition, experimental conditions, and relevant spectral feature information.

[0036] In this embodiment, a mixed drug sample (aspirin and ibuprofen) is selected as the data source of the sample to be measured. In the range of 0.1 - 3 THz, the terahertz time-domain spectrometer is used to collect the spectral data of the sample to be measured. To further identify the concentration of each chemical component, in Step 1, different terahertz spectral data of each drug at different concentrations are collected.

[0037] Step 2. In this embodiment, a mixed drug sample is selected, which is a sample prepared by mixing aspirin and ibuprofen in different mass ratios. The sample to be measured is pressed into a circular uniform thin sheet with a diameter of 13.00 mm and a thickness of 3.00 mm, and is measured using a terahertz time-domain spectrometer.

[0038] In this embodiment, the above-mentioned aspirin and ibuprofen are mixed in different ratios to obtain 10 mixed samples to be measured, and qualitative and quantitative determinations are performed on aspirin and ibuprofen. To ensure the reliability of the data, each sample is scanned 10 times at equal intervals using a terahertz time-domain spectrometer, and then the average value is taken. The absorbance of the sample can be calculated using relevant formulas to capture the characteristic fingerprints of the sample. Figure 2 Absorption coefficient spectra for aspirin, ibuprofen, and the mixed sample.

[0039]

[0040] Step 3. The terahertz data collected in Step 2 is preprocessed to improve the accuracy and spectral quality of subsequent analysis. The wavelet denoising algorithm is used to remove noise from the original terahertz spectral data. By setting an appropriate threshold, the effective part of the signal is retained, and the influence of instrument or environmental noise is eliminated. In this embodiment, in Step 3, the Gaussian smoothing method is used to smooth the spectral data, reducing the high-frequency noise in the signal while maintaining the main spectral feature peaks. The polynomial fitting method is used to perform baseline correction on the collected spectral data. The baseline drift caused by instrument or environmental factors is removed to ensure that the spectral data reflects the true characteristics of the sample.

[0041] Step 3: Normalize the spectral data to make the intensity ranges of the data obtained in Step 2 consistent for easy comparison.

[0042] Step 4: Compare the spectrum of the sample to be tested preprocessed in Step 3 with the spectral data of known drug samples in the database. Use a similarity metric method to evaluate the matching degree between the spectrum of the sample to be tested and each spectrum in the database. The similarity metric method is calculated according to the relevant formula of the Hit Quality Index (HQI): where is the library entry being searched, is the unknown spectrum with n data points. This formula gives the best match when the exponent is 0 and the worst match when the exponent is .

[0043] In this embodiment, in Step 4, through a spectral matching algorithm, the spectrum of the sample to be tested is compared with the spectra of multiple standard drugs in the database, and the names of the candidate substances and their corresponding similarity degrees are output. If the matching degree ≥ 55%, it is determined that the sample to be tested may contain this chemical substance.

[0044] Step 5: Mixture component identification and quantitative analysis. Use PCA to deconstruct the spectral data of the sample, so as to effectively extract the characteristic signals of different components from the mixed spectrum. By analyzing the absorption peaks and intensity changes in the spectrum, the characteristic spectra of aspirin and ibuprofen are respectively identified.

[0045] Step 5: Utilize the absorption intensity information of each component in the terahertz spectrum, combined with the concentration data of standard drugs in the database, and adopt a standard curve between spectral intensity and concentration to estimate the relative content of each component in the sample to be tested. Figure 3 is the prediction result graph of the sample concentration.

[0046] In this embodiment, by analyzing the spectral characteristics of the sample to be tested, combined with the results of spectral matching and quantitative analysis, the components and their relative contents of aspirin and ibuprofen are finally identified.

[0047] The above are only the preferred embodiments of the present invention and do not impose any limitations on the present invention. Any modifications, changes, and improvements made within the spirit and principle of the technical solution of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. The present invention relates to a method for identifying the components of a mixture based on terahertz spectroscopy detection database search, characterized in that, Including: Step 1: Establish a high-quality terahertz spectroscopy database; Step 2: Collect terahertz spectroscopy data to generate terahertz spectra containing the fingerprint information of the samples; Step 3: Perform data preprocessing to improve data quality and the accuracy of subsequent analysis; Step 4: Database search and spectrum comparison. Compare the terahertz spectrum of the sample to be tested with the data in the database to evaluate the matching degree between the spectrum of the sample to be tested and each spectrum in the database. Step 5: Identify the components of the mixture and perform quantitative analysis to effectively separate the spectral signals of different components. Use the intensity information in the terahertz spectrum to estimate the relative content of each component in the mixture and identify the components of the mixture. A method for identifying the components of a mixture based on database search of terahertz spectroscopy detection according to claim 1, wherein in Step 1, a high-quality terahertz spectroscopy database is established, including a terahertz database of chemicals collected by oneself and an online database that can be publicly viewed.

2. The mixture component identification method based on terahertz spectroscopy detection database search according to claim 1, wherein, In Step 1, collect the terahertz spectra of single chemical substances in a pure state. Further, the components of the collected mixture include standard compounds with known ratios or mixed samples of chemical substances with known various mixing ratios.

3. The mixture component identification method based on terahertz spectroscopy detection database search according to claim 1, characterized in that, When establishing the database in Step 1, the design of the data structure should include information such as sample identification, chemical name and molecular formula, spectral data, and spectral characteristics of experimental conditions. Regularly update and expand the database to expand the coverage and accuracy of the database.

4. A method for identifying the components of a mixture based on terahertz spectroscopy detection database search according to claim 1, characterized in that, In Step 2, the sample to be tested is scanned 10 times at equal intervals with a terahertz time-domain spectrometer, and then the average value is taken to calculate the absorbance α(w) of the sample to capture the characteristic fingerprint of the sample.

5. The mixture component identification method based on terahertz spectroscopy detection database search according to claim 1, wherein In Step 3, preprocess the terahertz data collected in Step 2 to ensure that the spectral data reflects the true characteristics of the sample. Use the wavelet denoising algorithm to remove noise; by setting an appropriate threshold, retain the effective part of the signal; use the Gaussian smoothing method to smooth the spectral data and maintain the main spectral characteristic peaks; use the polynomial fitting method to perform baseline correction on the collected spectral data.

6. The mixture component identification method based on terahertz spectroscopy detection database search according to claim 1, characterized in that, In Step 3, perform normalization processing on the spectral data for easy comparison.

7. A method for identifying the components of a mixture based on searching a terahertz spectroscopy detection database according to claim 1, wherein, In Step 4, use the similarity measurement method to evaluate the matching degree between the spectrum of the sample to be tested and each spectrum in the database.

8. The mixture component identification method based on terahertz spectrum detection database search according to claim 1, wherein In Step 4, if the matching degree calculated by the spectrum matching algorithm is ≥55%, it is determined that the sample to be tested may contain this chemical substance.

9. The mixture component identification method based on terahertz spectroscopy detection database search according to claim 1, characterized in that, In Step 5, use the principal component analysis method to deconstruct the spectral data of the sample, so as to effectively extract the characteristic signals of different components from the mixed spectrum.

10. A method for identifying the components of a mixture based on searching a terahertz spectroscopy detection database according to claim 1, characterized in that, In Step 5, use the standard curve between spectral intensity and concentration to estimate the relative content of each component in the sample to be tested.

Images