Method for quantitative determination of dimethylacetamide in solid hydrogen-containing sample and application thereof

Abstract

The present application belongs to the technical field of nuclear magnetic resonance testing, and particularly relates to a method for quantitatively detecting dimethylacetamide in a solid-state hydrogen-containing sample and application thereof. The method comprises the following steps: (1) selecting solid-state hydrogen-containing samples containing different amounts of dimethylacetamide as calibration samples; (2) obtaining nuclear magnetic signals of the calibration samples; (3) establishing a nuclear magnetic resonance quantitative analysis model of dimethylacetamide by using a regression analysis method according to the nuclear magnetic signals obtained in step (2); and (4) obtaining the mass content of dimethylacetamide in a to-be-tested sample by using the nuclear magnetic resonance quantitative analysis model in step (3) according to the nuclear magnetic signals of the to-be-tested sample. The method can accurately determine the content of dimethylacetamide in a solid-state hydrogen-containing sample by using nuclear magnetic resonance relaxation analysis technology, and has the advantages of non-destructive, rapid, accurate, non-polluting and strong timeliness, thereby meeting the demand for testing DMAc in actual engineering production lines.

Description

Technical Field

[0001] This invention belongs to the field of nuclear magnetic resonance testing technology, specifically relating to a method and application for the quantitative detection of dimethylacetamide in solid hydrogen-containing samples. Background Technology

[0002] Dimethylacetamide (DMAc) can dissolve a variety of compounds and is completely miscible with water, ethers, ketones, esters, etc. It possesses high thermal stability, is not easily hydrolyzed, and has low corrosiveness, making it widely used. DMAc exhibits excellent solubility for many resins, especially polyurethane and polyimide resins, and is mainly used as a solvent for heat-resistant synthetic fibers, plastic films, coatings, pharmaceuticals, and acrylonitrile spinning. Internationally, DMAc is widely used to produce polyimide films, soluble polyimides, polyimide-perfluoroethylene propylene composite films, polyimide (aluminum) films, and soluble polyimide molding powders; domestically, it is mainly used as an excellent polar solvent for polymer synthetic fiber production and other organic synthesis.

[0003] In the dry spinning of spandex, thermoplastic polyurethane is used as the raw material, and dimethylacetamide is used as the solvent. The polyurethane is dissolved by heating to prepare a polyurethane solution with a mass fraction of 25-35%. Additives are then added, and after filtration, degassing, and degassing, a spinning solution with uniform viscosity is obtained. This spinning solution is then metered and uniformly extruded into a nozzle using a precision gear pump, forming a fine stream and then filaments. After drying, spandex filaments are obtained. However, in actual use, the residual dimethylacetamide content in the finished spandex filaments should not exceed 0.1%. Excessive dimethylacetamide can cause the spandex filaments to stick together during storage and can also cause harm to the human body. Therefore, it is necessary to precisely control the residual dimethylacetamide in the finished spandex filaments.

[0004] Existing methods for detecting DMAc in spandex fibers generally include gas chromatography, near-infrared spectroscopy, and oven drying. Gas chromatography involves extraction with acetone before testing, a time-consuming method requiring chemical reagents. In practical production lines, timeliness is crucial when testing DMAc in spandex fibers, and gas chromatography cannot meet this requirement. Near-infrared spectroscopy, such as the method disclosed in Chinese patent document CN104730028A, uses the toxic solvent acetone. Furthermore, the principle of near-infrared spectroscopy is inversion or scattering, resulting in a spectral penetration depth of less than 10 mm. For compacted spandex samples, only surface signals are detected, neglecting internal signals. This insufficient penetration leads to low test results. Moreover, near-infrared calibration only considers DMAc, failing to account for changes in the physical state of DMAc in solution and after it adheres to the spandex fibers, introducing systematic errors. The oven drying method involves drying at low temperature for a long time to volatilize dimethylacetamide. The residual content of dimethylacetamide is calculated by the difference method based on the mass of the remaining sample. This method is time-consuming, and the dimethylacetamide volatilized during drying can be harmful to the human body. Long-term testing can also cause environmental pollution, and the test repeatability is poor. Summary of the Invention

[0005] Therefore, the technical problem to be solved by the present invention is to overcome the defects of the existing technology in the quantitative analysis of dimethylacetamide in spandex yarn, such as long detection time, high solvent toxicity, and insufficient penetration leading to low test results. Thus, the present invention provides a method and application for the quantitative detection of dimethylacetamide in solid hydrogen-containing samples.

[0006] To this end, the present invention provides the following technical solution.

[0007] This invention provides a method for the quantitative detection of dimethylacetamide in solid hydrogen-containing samples, comprising the following steps:

[0008] (1) Select solid hydrogen-containing samples containing different masses of dimethylacetamide as calibration samples;

[0009] (2) Obtain the NMR signal of the calibration sample;

[0010] (3) Based on the NMR signal obtained in step (2), a quantitative NMR analysis model for dimethylacetamide was established using regression analysis.

[0011] (4) Based on the NMR signal of the sample to be tested, the mass of dimethylacetamide in the sample to be tested is obtained through the NMR quantitative analysis model in step (3).

[0012] In step (2), the T1-MSE sequence is used to determine the calibration sample and obtain the NMR signal of the calibration sample;

[0013] The T1-MSE sequence parameters:

[0014] Sampling toggle time Tau: 0.01ms-100000ms;

[0015] 90° pulse phase and Specifically, Each pulse has a phase of 90°, switching between 90°, 180°, and 270°.

[0016] Sampling time Tac: 0.01us-1000000us;

[0017] Number of sampling repetitions N: 2-1000;

[0018] Sampling bandwidth: 5kHz-5000kHz;

[0019] Sampling cumulative count NS: 2-10000;

[0020] 180° pulse duration 1us-500us;

[0021] 90° pulse duration 1us-500us.

[0022] Further, step (3), the establishment of the nuclear magnetic resonance quantitative analysis model includes the following steps:

[0023] (a) Set the number of calibration samples to n, and the number of sampling data points for each calibration sample to L, to obtain a data matrix n*L; use the normalization processing method to preprocess step (2) to obtain the NMR signal of the calibration sample and obtain the normalization coefficient W;

[0024] (b) The number of principal components Q in the modeling process is determined using principal component analysis.

[0025] (c) Determine that the input to the regression analysis model is Q principal components, and obtain an n*Q principal component matrix; solid hydrogen-containing substances and dimethylacetamide form a two-dimensional matrix, which is statistically represented as an n*2 matrix;

[0026] (d) A quantitative analysis model for nuclear magnetic resonance is obtained by using regression analysis; wherein, the regression analysis method includes basic regression analysis and multi-output regression analysis.

[0027] The correlation coefficient R² for cross-validation prediction of solid hydrogen-containing substances in the nuclear magnetic resonance quantitative analysis model is ≥0.998, and the correlation coefficient R² for cross-validation prediction of dimethylacetamide components is ≥0.998.

[0028] Further, in step (d), the basic regression analysis method is one of principal component regression, partial least squares regression, ridge regression, support vector machine, Gaussian process regression, robust regression, Bayesian regression, limit vector machine, and least squares support vector machine.

[0029] Furthermore, in step (d), the multi-output regression analysis method is a chain regression method;

[0030] Cross-validation is used when constructing the model using chain regression.

[0031] When evaluating the model, 60-90% of the samples are randomly selected as the calibration set, and 10-40% of the samples are selected as the test and validation set.

[0032] Further, in step (4), the NMR signals of T samples to be tested are obtained, and the normalization method and principal component analysis method are used in sequence to analyze them to obtain the principal component matrix of the T*Q sample to be tested. Then, the mass of dimethylacetamide in the sample to be tested is obtained according to the NMR quantitative analysis model in step (3).

[0033] Further, in step (1), the number of calibration samples is greater than 20, and the mass of the calibration samples is 2-100g;

[0034] The mass concentration of dimethylacetamide in the calibration sample does not exceed 10%.

[0035] Furthermore, in step (4), the mass of solid hydrogen-containing substances in the sample to be tested can also be obtained through the nuclear magnetic resonance quantitative analysis model of step (3).

[0036] The solid hydrogen-containing sample mentioned in this invention refers to a solid hydrogen-containing substance containing dimethylacetamide, and the solid hydrogen-containing substance refers to the substance remaining after removing dimethylacetamide from the solid hydrogen-containing sample.

[0037] Specifically, the solid hydrogen-containing material can be spandex containing dimethylacetamide.

[0038] The present invention also provides an application of the above method for the quantitative detection of dimethylacetamide and spandex in spandex samples.

[0039] The spandex sample refers to spandex containing dimethylacetamide.

[0040] The technical solution of this invention has the following advantages:

[0041] 1. The present invention provides a method for quantitative detection of dimethylacetamide in solid hydrogen-containing samples. The method includes (1) selecting solid hydrogen-containing samples containing different masses of dimethylacetamide as calibration samples; (2) acquiring the NMR signal of the calibration samples; (3) establishing a quantitative NMR analysis model for dimethylacetamide based on the NMR signal obtained in step (2) using regression analysis; and (4) obtaining the mass content of dimethylacetamide in the sample to be tested based on the NMR signal of the sample to be tested using the quantitative NMR analysis model in step (3). This method can accurately determine the content of dimethylacetamide in solid hydrogen-containing samples using NMR relaxation analysis technology, and has the advantages of being non-destructive, rapid, accurate, and pollution-free. This method is highly time-sensitive and can meet the requirements of testing samples for DMAc in actual engineering production lines; this method does not have systematic errors and overcomes the problem of low test results due to insufficient penetration in the prior art, and has good repeatability.

[0042] This invention establishes a quantitative analytical relationship between solid hydrogen-containing samples with different residual amounts of dimethylacetamide and their nuclear magnetic resonance transverse relaxation time, thereby obtaining the residual concentration of DMAc in the sample. The calibration and testing process involves no chemical reactions and does not use toxic reagents, improving the environmental friendliness of the testing method and achieving real-time rapid sample analysis. Furthermore, the testing method is simple and has a short overall testing cycle.

[0043] 2. The present invention provides a method for quantitative detection of dimethylacetamide in solid hydrogen-containing samples. This method can simultaneously obtain the mass of solid hydrogen-containing substances and dimethylacetamide in the sample based on the established model.

[0044] By using the T1-MSE sequence provided by this invention and the multi-output regression analysis method, the signals of solid hydrogen-containing substances and dimethylacetamide can be distinguished, an accurate analysis model can be established, and the correlation between the output results can be fully utilized to make the test results more accurate. At the same time, the mass of solid hydrogen-containing substances and dimethylacetamide can be obtained, which is highly efficient.

[0045] 3. The present invention provides an application for the quantitative detection of dimethylacetamide and spandex in spandex samples. Dimethylacetamide and spandex have extremely short relaxation times (e.g., less than 1 ms) and small relaxation differences. Conventional spin echo methods cannot effectively distinguish between spandex and dimethylacetamide. Based on the principle of low-field nuclear magnetic resonance, the present invention uses T1-MSE sequences to acquire nuclear magnetic signals, which can effectively recover and distinguish the signals of spandex and dimethylacetamide, and establish a more accurate analytical model. At the same time, the use of multi-output regression analysis can fully utilize the correlation between output results, making the test results more accurate. This overcomes the problems of overfitting, sensitivity to individual abnormal data, and model failure caused by the failure to consider the correlation of the analyte components in the existing technology of multivariate regression analysis (PLSR), as well as the deviation in component prediction caused by the failure to consider the correlation of the analyte components.

[0046] The T1 portion of the T1-MSE sequence provides spin lattice relaxation information of the sample system, while the MSE portion provides spin relaxation information (T2* and T2). For samples where solid materials like spandex constitute the majority, T1 relaxation is typically slow, making it easy to acquire sufficient NMR signals. However, relaxation attenuates rapidly in the T2 direction, and conventional methods, such as spin echo, tend to lose most of the effective information. The MSE portion provided by this invention can effectively recover the total magnetization vector of the sample system through multiple 90° pulses with different phases and durations, ensuring that the acquired NMR signals include all information about spandex and dimethylacetamide. Combining MSE and T1 for two-dimensional analysis allows for precise quantification and effective differentiation of different components. It eliminates the need to weigh the test sample and allows for direct testing of the residual amount of dimethylacetamide in the sample. Attached Figure Description

[0047] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0048] Figure 1 It is the T1-MSE sequence in Embodiment 1 of the present invention;

[0049] Figure 2 It is the sampling signal of nuclear magnetic resonance in Embodiment 1 of the present invention;

[0050] Figure 3 It is the cumulative contribution rate of the calibration sample in Example 1 of this invention;

[0051] Figure 4 This is a comparison between the actual value and the model prediction value of pure spandex in the calibration sample of Example 1 of this invention;

[0052] Figure 5 This is a comparison between the actual value and the model prediction value of dimethylacetamide in the calibration sample of Example 1 of this invention. Detailed Implementation

[0053] The following embodiments are provided to better understand the present invention and are not limited to the preferred embodiments described. They do not constitute a limitation on the content and scope of protection of the present invention. Any product that is the same as or similar to the present invention, derived by any person under the guidance of the present invention or by combining the features of the present invention with other prior art, falls within the protection scope of the present invention.

[0054] For experiments not specifically described in the examples, the procedures or conditions should be followed according to the conventional experimental procedures described in the literature in this field. Reagents or instruments whose manufacturers are not specified are all commercially available conventional reagent products.

[0055] In a specific embodiment, the method for quantitative detection of dimethylacetamide in a solid hydrogen-containing sample includes the following steps:

[0056] (1) Preparation of calibration samples and test samples: Select solid hydrogen-containing samples with different masses as calibration samples. The mass content of dimethylacetamide in the calibration samples is between 0.001% and 10%, the mass of the calibration samples is between 2 and 100 g, and the number of calibration samples n is greater than 20. The mass of the calibration samples is shown in Table 1.

[0057] Table 1 Quality of Calibration Samples

[0058]

[0059]

[0060] Wherein, ML0 = 0 represents a solid hydrogen-containing substance that does not contain DMAc; MS0 is dimethylacetamide that does not contain a solid hydrogen-containing substance; Mn = MSn + MLn.

[0061] Sample to be tested, for future use.

[0062] (2) Obtaining the NMR signals of the calibration sample and the test sample: Place the calibration sample into a quartz test tube, and then place it into the probe coil of the NMR device. Use the T1-MSE sequence to measure the NMR signals of the calibration sample and the test sample.

[0063] T1-MSE sequence parameters:

[0064] Sampling toggle time Tau: 0.01ms-100000ms Each pulse has a phase of 90°, switching between 90°, 180°, and 270°.

[0065] Sampling time Tac: 0.01us-1000000us, number of repetitions N: 2-1000;

[0066] Sampling bandwidth: 5kHz-5000kHz;

[0067] Sampling cumulative count NS: 2-10000;

[0068] 180° pulse duration 1us-500us: 90° pulse duration 1us-500us.

[0069] (3) Establishment of the quantitative analysis model using nuclear magnetic resonance:

[0070] (a) The number of calibration samples n, the number of sampling data points for each sequence of each calibration sample L, and the resulting data matrix n*L are obtained. The calibration sample NMR signal is preprocessed using a normalization method to obtain the normalization coefficient W.

[0071] (b) The number of principal components Q in the modeling was determined using principal component analysis (PCA), where the cumulative contribution rate of sample information reaching 0.999 was used as the principal component;

[0072] (c) Determine that the input of the regression analysis model is Q principal components, and obtain an n*Q principal component matrix; refer to chemical indicators, and calibrate the solid hydrogen-containing substances and dimethylacetamide in the sample to form a two-dimensional matrix, which is statistically represented as an n*2 matrix;

[0073] (d) A quantitative analysis model for nuclear magnetic resonance is obtained by using regression analysis; the regression analysis method includes basic regression analysis and multi-output regression analysis.

[0074] The basic regression analysis method can be one of principal component regression (PCR), partial least squares regression (PLSR), ridge regression, support vector machine (SVR), Gaussian process regression (GPR), robust regression (Huber), Bayesian regression (Bayesian Ridge), limit vector machine (ELM), and least squares support vector machine (LSSVR); the regression algorithm version comes from the Python machine learning library scikit-Learn 0.24.2.

[0075] The multi-output regression analysis method is chain regression (RegressorChain), and the algorithm version is RegressorChain(base_estimator, order, cv, random_state) from the Python machine learning library scikit-Learn 0.24.2. The base_estimator comes from the optimal regressor in the basic regression analysis method. The order of labels in order is consistent with the order in the n*2 reference chemical index matrix. CV represents the performance evaluation method of the regressor, which is cross-validation here. random_state is a random number that determines the order of the regression chain and is a fixed positive integer. When evaluating the regression model, 60-90% of the samples are randomly selected as the calibration set, and the remainder is used as the test and validation set.

[0076] After model optimization, a nuclear magnetic resonance quantitative analysis model based on chain regression was obtained, which is statistically known as ModelR. The qualified models are as follows: the correlation coefficient R2 for cross-validation prediction of solid hydrogen content is ≥0.998, and the correlation coefficient R2 for cross-validation prediction of dimethylacetamide component is ≥0.998.

[0077] (4) Obtain the NMR signals of T samples containing dimethylacetamide. Normalize the NMR signals of the samples by referring to the normalization coefficient, and then perform principal component analysis to obtain the principal component matrix of the samples T*Q. Obtain the mass of solid hydrogen-containing substances and DMAc in the samples according to the NMR quantitative analysis model in step (3).

[0078] Example 1

[0079] This embodiment provides a method for quantitative detection of dimethylacetamide and spandex in spandex filament samples, including the following steps. In this embodiment, spandex filament refers to spandex containing dimethylacetamide.

[0080] (1) Preparation of calibration samples and test samples: Select spandex yarns of different masses as calibration samples. The mass content of dimethylacetamide in the calibration samples is between 0.001% and 10%, the mass of the calibration samples is between 2 and 100g, and the number of calibration samples n is 25. The mass of the calibration samples is shown in Table 2.

[0081] Table 2 Quality of Calibration Samples

[0082]

[0083] Where ML0 = 0 represents spandex in spandex yarn that does not contain DMAc; MS0 = 0 represents pure dimethylacetamide; Mn = MSn + MLn.

[0084] Five samples to be tested, labeled as Sample 1-5, are prepared for use.

[0085] (2) Obtaining NMR signals of calibration and test samples: Place the calibration sample in a quartz test tube, and then place it in the probe coil of the NMR device. Use the T1-MSE sequence to measure the NMR signals of the calibration and test samples; wherein, the T1-MSE sequence is shown in [link to T1-MSE sequence]. Figure 1 The sampling signal of the calibration sample is shown in [reference needed]. Figure 2 ;

[0086] T1-MSE sequence parameters:

[0087] Sampling flip time Tau: 10ms-3000us Each pulse has a phase of 90°, switching between 90°, 180°, and 270°.

[0088] Sampling time Tac: 500us, repetition count N: 15, corresponding flip times Tau are: 10ms, 15ms, 23ms, 34ms, 51ms, 76ms, 114ms, 171ms, 256ms, 384ms, 577ms, 865ms, 1297ms, 1946ms, 2919ms;

[0089] Each MSE segment has 100 sampled data points.

[0090] Sampling bandwidth: 200kHz;

[0091] Cumulative count NS: 32;

[0092] 180° pulse duration 5µs: 90° pulse duration 2.5µs.

[0093] (3) Establishment of the quantitative analysis model using nuclear magnetic resonance:

[0094] (a) The number of calibration samples n is 25, and the number of sampling data points for each sequence of each calibration sample is L, which is 150, resulting in a data matrix of 25*150. The calibration sample NMR signal is preprocessed using the normalization method to obtain the normalization coefficient W.

[0095] (b) Principal component analysis (PCA) was used to determine the number of principal components for modeling (7). Samples with a cumulative contribution rate of 99.99% were selected as principal components. The cumulative contribution rate of each calibration sample is shown in Table 3. Figure 3 ;

[0096] Table 3 Cumulative Contribution Rate of Principal Components in Calibration Samples

[0097] Serial Number 1 2 3 4 5 6 7 Cumulative contribution rate 97.204 99.903 99.954 99.971 99.981 99.988 99.991 Serial Number 8 9 9 10 11 12 13 Cumulative contribution rate 99.994 99.995 99.995 99.996 99.997 99.998 99.998 Serial Number 14 15 16 17 18 19 20 Cumulative contribution rate 99.999 99.999 99.999 99.999 100 100 100 Serial Number 21 22 23 24 25 / / Cumulative contribution rate 100 100 100 100 100 / /

[0098] (c) The input to the regression analysis model is determined to be 7 principal components, and a 25*7 principal component matrix is ​​obtained; the reference chemical indicators are used to form a two-dimensional matrix of spandex and dimethylacetamide in the calibrated sample, which is statistically represented as a 25*2 matrix.

[0099] (d) A quantitative analysis model for nuclear magnetic resonance was obtained using regression analysis. Regression analysis methods include basic regression analysis and multi-output regression analysis.

[0100] The basic regression analysis method is Bayesian regression, and the algorithm version comes from the Python machine learning library scikit-Learn 0.24.2. The multi-output regression analysis method is chain regression (RegressorChain), and the algorithm version is RegressorChain(base_estimator, order, cv, random_state) from the Python machine learning library scikit-Learn 0.24.2. base_estimator is the optimal regressor from the basic regression analysis method; the label order in order is consistent with the order in the 25*2 reference chemical index matrix; CV represents the performance evaluation method of the regressor, which is cross-validation here; random_state is a fixed positive integer used to determine the order of the regression chain. When evaluating the regression model, 70% of the samples are randomly selected as the calibration set, and the remaining 30% are used as the test and validation set.

[0101] After model optimization, the nuclear magnetic resonance quantitative analysis model was obtained, which is statistically known as ModelR. The qualified models are as follows: the cross-validation prediction correlation coefficient R2 ≥ 0.9992 for the spandex component without dimethylacetamide, and the cross-validation prediction correlation coefficient R2 ≥ 0.9995 for the dimethylacetamide component.

[0102] Comparison between actual results and model predictions regarding spandex and dimethylacetamide in spandex yarn:

[0103] The NMR signals of 25 calibration samples were measured again using the method described above. Substituting these signals into the model, the predicted values ​​of spandex and dimethylacetamide in the 25 calibration samples were obtained. The actual values ​​and model predicted values ​​of spandex in the calibration samples are shown below. Figure 4 The true and model predicted values ​​of dimethylacetamide in the calibration samples are shown in [reference needed]. Figure 5 .

[0104] Figure 4 and Figure 5 This invention demonstrates that the method provided can accurately determine the content of spandex and dimethylacetamide in spandex yarn. The method provides accurate test results and is highly efficient.

[0105] (4) Obtain the NMR signals of 5 spandex yarn samples (containing residual dimethylacetamide). The mass of the samples is shown in Table 4. The NMR signals of the samples were normalized and preprocessed with reference to the normalization coefficient. Then, principal component analysis was performed to obtain the 5*7 principal component matrix of the samples. The mass of spandex and DMAc in the samples was obtained according to the NMR quantitative analysis model in step (3). The predicted values ​​of spandex and dimethylacetamide in the samples are shown in Table 4.

[0106] Table 4 Analysis of Prediction Results

[0107]

[0108] To verify the repeatability of the test method, repeatability tests were performed on sample 1 (5.3268g) and sample 5 (14.2478g) for 11 times. The mass content of dimethylacetamide in the samples is shown in Table 5.

[0109] Table 5 Repeatability Test Results

[0110]

[0111]

[0112] Note: The number in Table 5 represents the mass content of dimethylacetamide in the spandex sample obtained from the Nth test, in wt%.

[0113] The test results above demonstrate that the test method provided by this invention has good repeatability and stable test results.

[0114] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for the quantitative detection of dimethylacetamide in a solid hydrogen-containing sample, characterized in that, Includes the following steps, (1) Solid hydrogen-containing samples containing different masses of dimethylacetamide were selected as calibration samples; (2) Obtain the NMR signal of the calibration sample; in step (2), the T1-MSE sequence is used to determine the calibration sample and obtain the NMR signal of the calibration sample; the T1-MSE sequence parameters are: Sampling flip time Tau: 0.01ms-100000ms; 90° pulse phase φ a φ b φ c and φ -c ; Sampling time Tac: 0.01us-1000000us; Number of sampling repetitions N: 2-1000; Sampling bandwidth: 5KHz-5000Khz; Number of cumulative samplings NS: 2-10000; 180° pulse time 1us-500us; 90° pulse time 1us-500us; (3) Based on the NMR signal obtained in step (2), a quantitative NMR analysis model for dimethylacetamide is established using regression analysis. The establishment of the quantitative NMR analysis model includes the following steps: (a) Set the number of calibration samples to n, and the number of sampling data points for each calibration sample to L, to obtain a data matrix n*L; use the normalization processing method to preprocess step (2) to obtain the NMR signal of the calibration sample and obtain the normalization coefficient W; (b) The number of principal components Q in the modeling is determined using principal component analysis; (c) Determine that the input to the regression analysis model is Q principal components, and obtain the n*Q principal component matrix; Solid hydrogen-containing substances and dimethylacetamide form a two-dimensional matrix, which is statistically represented as an n*2 matrix; (d) A quantitative analysis model for nuclear magnetic resonance is obtained using regression analysis; wherein, the regression analysis includes basic regression analysis and multi-output regression analysis. (4) Based on the NMR signal of the sample to be tested, the mass of dimethylacetamide and solid hydrogen-containing substances in the sample to be tested is obtained through the NMR quantitative analysis model in step (3).

2. The method according to claim 1, characterized in that, The correlation coefficient R² for cross-validation prediction of solid hydrogen-containing substances in the nuclear magnetic resonance quantitative analysis model is ≥0.998, and the correlation coefficient R² for cross-validation prediction of dimethylacetamide components is ≥0.

998.

3. The method according to claim 1, characterized in that, In step (d), the basic regression analysis method is one of principal component regression, partial least squares regression, ridge regression, support vector machine, Gaussian process regression, robust regression, Bayesian regression, limit vector machine and least squares support vector machine.

4. The method according to claim 1 or 3, characterized in that, In step (d), the multi-output regression analysis method is the chain regression method; Cross-validation is used when constructing the model using chain regression. When evaluating the model, 60-90% of the samples are randomly selected as the calibration set, and 10-40% of the samples are selected as the test and validation set.

5. The method according to claim 1, characterized in that, Step (4): Obtain the NMR signals of T samples to be tested, and analyze them sequentially using the normalization method and principal component analysis method to obtain the principal component matrix of the T*Q sample to be tested. Then, obtain the mass of dimethylacetamide in the sample to be tested according to the NMR quantitative analysis model in step (3).

6. The method according to claim 1, characterized in that, Step (1): The number of calibration samples is greater than 20, and the mass of the calibration samples is 2-100g; The mass concentration of dimethylacetamide in the calibration sample does not exceed 10%.

7. The application of the method according to any one of claims 1-6 in the quantitative detection of dimethylacetamide and spandex in spandex samples.

Images