Method for determining trace water content in electrolyte based on near infrared spectroscopy

Abstract

The invention relates to a method for measuring the content of trace water in an electrolyte based on near infrared spectroscopy. Real-time online detection of the content of trace water in the electrolyte production process is achieved through the near infrared spectroscopy.

Description

Technical Field

[0001] The present application relates to the field of detection and analysis technology, and in particular to a method for measuring trace moisture content in an electrolyte based on near-infrared spectroscopy. Background Art

[0002] The electrolyte used in secondary batteries needs to strictly control the moisture content to avoid excessive moisture content affecting the safety of the battery. Usually the moisture content in the electrolyte is less than 0.01% by mass, basically at the ppm level. In the process of producing the electrolyte, it is necessary to monitor the moisture content in time to determine whether the electrolyte has been dehydrated to improve production efficiency. At present, the method for testing the trace moisture content in the electrolyte is mainly the Karl Fischer-Coulomb method. Although this method is accurate in testing, it is impossible to detect the trace moisture content in real time online during the production process of the electrolyte.

[0003] Therefore, it is urgent to develop a method that can detect trace moisture content in real time during the electrolyte production process. Summary of the invention

[0004] In view of the problems existing in the background technology, the present application provides a method for determining the trace moisture content in an electrolyte based on near-infrared spectroscopy, which can detect the moisture content in the electrolyte production process in real time online.

[0005] The first aspect of the present application provides a method for determining trace moisture content in an electrolyte based on near infrared spectroscopy, comprising the following steps:

[0006] A mathematical model is established based on the near-infrared spectral data of the calibration set samples and the measured values ​​of water content;

[0007] Scanning the near infrared spectrum of the electrolyte sample to be tested to obtain spectrum data of the electrolyte sample to be tested;

[0008] The trace water content in the electrolyte sample to be tested is calculated based on the mathematical model and the spectral data of the electrolyte sample to be tested.

[0009] The present application uses near-infrared spectroscopy to achieve real-time online detection of trace moisture content in the production process of electrolyte. Near-infrared detection is the sum frequency and frequency doubling absorption peaks of the characteristic groups of the compound, and the sum frequency and frequency doubling absorption intensity are dozens of times weaker than the fundamental frequency signal, so near-infrared spectroscopy is generally used for the detection of constant substances. For trace substances, due to their weak absorption intensity, coupled with the serious spectral overlap effect of near-infrared, the peak spectrum band is broadened, resulting in some trace substances having no obvious absorption peaks in the near-infrared band. Therefore, it has become a consensus that near-infrared spectroscopy is not suitable for trace substance detection. There is no report on near-infrared spectroscopy in trace moisture detection. However, the present application overcomes this technical bias and selects near-infrared spectroscopy to detect the trace moisture content in the electrolyte. The test results are consistent with the test results of Karl Fischer-Coulomb method, realizing real-time online detection and solving the problems existing in the background technology.

[0010] In addition, the determination method of the present application also has the advantages of high efficiency, rapidity, non-destructiveness, environmental protection, and low cost, and can be used not only for real-time online detection of trace moisture content in the electrolyte production process, but also for offline detection of trace moisture content in the electrolyte. In addition, the determination method of the present application has good accuracy and repeatability.

[0011] In some embodiments, according to the first aspect, a first example of the first aspect is provided, and establishing the mathematical model includes the following steps:

[0012] Sample preparation: Collect multiple electrolyte samples as calibration set samples;

[0013] True moisture content determination: Determine the moisture content of the calibration set samples to obtain the actual moisture content value;

[0014] Spectrum acquisition: Scan the near-infrared spectrum of the calibration set samples to obtain the near-infrared spectrum data of the calibration set samples, and record the corresponding measured values;

[0015] Model establishment: Based on the partial least squares method, the modeling band and spectral preprocessing method are accurately selected to establish a mathematical model.

[0016] This application establishes a mathematical model with good prediction accuracy by accurately selecting the modeling band and spectral preprocessing method, providing an important experimental basis for real-time online detection of trace moisture in the electrolyte production process.

[0017] In some embodiments, according to the first aspect, a second example of the first aspect is proposed, and the modeling band covers 5452cm -1 ~4600cm -1 Band range.

[0018] The selection of modeling bands is very important for the prediction effect of mathematical models. The modeling bands selected in this application cover 5452cm-1 ~4600cm -1 Band range, modeling with such modeling bands is conducive to improving the prediction effect of mathematical models.

[0019] In some embodiments, according to the first aspect, a third example of the first aspect is proposed, and the modeling band is 12492cm -1 ~4524cm -1 Alternatively, the modeling band is 7500cm -1 ~4600cm -1 Alternatively, the modeling band is 9400cm -1 ~7500cm -1 and 5452cm -1 ~4600cm -1 Alternatively, the modeling band is 5452cm -1 ~4600cm -1 .

[0020] By optimizing the modeling band, it is helpful to further improve the prediction effect of the mathematical model. Since the full band contains a lot of information, useless information or baseline noise will interfere with the prediction effect of the model. In order to improve the prediction effect of the mathematical model, this application selects the spectral band with effective information of the water content in the analyte in the near-infrared spectrum, that is, the above-mentioned modeling band for modeling, thereby reducing the intake of useless information and improving the prediction effect of the model.

[0021] In some embodiments, according to the first aspect, a fourth example of the first aspect is proposed, and the spectral preprocessing method includes one or more of the first-order derivative, the second-order derivative, the multivariate scattering correction, and the standard normalization method.

[0022] Since some physical states of the electrolyte sample, such as color and turbidity, can cause interferences such as baseline drift or translation and scattering, making the effective information of the band unclear, this application selects a spectral preprocessing method that can effectively reduce interference to improve the model prediction accuracy.

[0023] In some embodiments, according to the first aspect, a fifth example of the first aspect is proposed, which is to determine the moisture content of the calibration set samples, scan the near-infrared spectrum of the electrolyte sample to be tested, and scan the near-infrared spectrum of the calibration set samples in an environment where the dew point temperature is below -10°C.

[0024] In order to reduce the measurement error caused by the ambient humidity during the test, the actual moisture content of the electrolyte sample and the collection of the near-infrared spectrum can be carried out in an experimental environment with a low dew point temperature. The present application is conducive to reducing the measurement error and improving the prediction effect of the model by optimizing the dew point temperature. In particular, when the electrolyte contains lithium bis(fluorosulfonyl)imide, since lithium bis(fluorosulfonyl)imide has strong water absorption, therefore, by controlling the dew point temperature of the experimental environment, it is more conducive to reducing the measurement error for such electrolytes and improving the prediction effect of the model.

[0025] In some embodiments, according to the first aspect, a sixth example of the first aspect is provided, wherein the dew point temperature is -10°C to -60°C.

[0026] By optimizing the dew point temperature, it is helpful to reduce the measurement error caused by ambient humidity during testing.

[0027] In some embodiments, according to the first aspect, a seventh example of the first aspect is provided, in which the moisture content of the calibration set samples is determined using the Karl Fischer-Coulomb method.

[0028] The accuracy of the measured value of moisture content affects the prediction effect of the model. The present application uses the Karl Fischer-Coulomb method to measure the moisture content of the calibration set samples, and the obtained results can be used as true values ​​for modeling, which can improve the prediction effect of the model.

[0029] In some embodiments, according to the first aspect, an eighth example of the first aspect is provided, and the electrolyte includes an electrolyte for a secondary battery.

[0030] The determination method of the present application is suitable for detecting trace moisture content in the electrolyte of a secondary battery.

[0031] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is an overlay of original spectra of an electrolyte containing lithium bis(fluorosulfonyl)imide according to an embodiment of the present application. DETAILED DESCRIPTION

[0033] The exemplary embodiments of the present disclosure will be described in more detail below. However, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to enable the scope of the present disclosure to be fully communicated to those skilled in the art.

[0034] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application. The terms "including" and "having" and any variations thereof in the specification and claims of this application are intended to cover non-exclusive inclusions.

[0035] For simplicity, only some numerical ranges are explicitly disclosed herein. However, any lower limit can be combined with any upper limit to form an unambiguous range; and any lower limit can be combined with other lower limits to form an unambiguous range, and any upper limit can be combined with any other upper limit to form an unambiguous range. In addition, although not explicitly stated, each point or single value between the range endpoints is included in the range. Thus, each point or single value can be combined with any other point or single value as its own lower limit or upper limit or with other lower limits or upper limits to form an unambiguous range.

[0036] In the description of this article, it should be noted that, unless otherwise specified, "above" and "below" are inclusive of the number itself, and the "multiple" in "one or more" means more than two (including two).

[0037] The above-mentioned summary of the invention of the present application is not intended to describe each disclosed embodiment or each implementation in the present application. The following description more specifically illustrates exemplary embodiments. In many places throughout the application, guidance is provided by a series of examples, which can be used in various combinations. In each example, enumeration is only used as a representative group and should not be interpreted as exhaustive.

[0038] Electrolyte is an important component of secondary batteries. The electrolyte used in secondary batteries needs to strictly control the moisture content to avoid excessive moisture content affecting the safety of the battery. Usually the moisture content in the electrolyte is less than 0.01% by mass, basically at the ppm level. In the process of producing the electrolyte, it is necessary to monitor the moisture content in time to determine whether the electrolyte has been dehydrated to improve production efficiency. At present, the method for testing the trace moisture content in the electrolyte is mainly the Karl Fischer-Coulomb method. Although this method is accurate in testing, it is impossible to detect the moisture content in real time online during the production process of the electrolyte. Therefore, it is urgent to develop a method that can detect the trace moisture content in the electrolyte production process in real time online.

[0039] Near-infrared spectroscopy has the advantages of being non-destructive, rapid, and high-throughput, and has been widely used in real-time online detection in industries such as petrochemicals, food, and medicine. However, since near-infrared detection detects the combined and doubled absorption peaks of characteristic groups of compounds, and the combined and doubled absorption intensities are dozens of times weaker than the fundamental frequency signal, it is generally used for the detection of constant substances. For trace substances, due to their weak absorption intensity, coupled with the serious spectral overlap effect of near-infrared, the peak spectrum band is broadened, resulting in some trace substances having no obvious absorption peaks in the near-infrared band. Therefore, it has become a consensus that near-infrared spectroscopy is not suitable for the detection of trace substances. There are no reports on the use of near-infrared spectroscopy in trace moisture detection.

[0040] The present application overcomes this technical bias and uses near-infrared spectroscopy to detect trace moisture content in the electrolyte. The detection results are consistent with those of the Karl Fischer-Coulomb method, realizing real-time online detection and solving the problem of being unable to detect trace moisture content in the electrolyte production process in real-time online.

[0041] Specifically, the first aspect of the present application provides a method for determining trace moisture content in an electrolyte based on near infrared spectroscopy, comprising the following steps:

[0042] A mathematical model is established based on the near-infrared spectral data of the calibration set samples and the measured values ​​of water content;

[0043] Scanning the near infrared spectrum of the electrolyte sample to be tested to obtain spectrum data of the electrolyte sample to be tested;

[0044] The trace water content in the electrolyte sample to be tested is calculated based on the mathematical model and the spectral data of the electrolyte sample to be tested.

[0045] The present application uses near infrared spectroscopy to achieve real-time online detection of trace moisture content in the production process of electrolyte. In addition, the determination method of the present application also has the advantages of high efficiency, rapidity, non-destructiveness, environmental protection, and low cost. It can be used not only for real-time online detection of trace moisture content in the production process of electrolyte, but also for offline detection of trace moisture content in electrolyte. In addition, the determination method of the present application has good accuracy and repeatability.

[0046] In the present application, "trace amount" means that the water content in the electrolyte is less than 0.01% by mass.

[0047] In some embodiments, establishing the mathematical model comprises the following steps:

[0048] Sample preparation: Collect multiple electrolyte samples as calibration set samples;

[0049] True moisture content determination: Determine the moisture content of the calibration set samples to obtain the actual moisture content value;

[0050] Spectrum acquisition: Scan the near-infrared spectrum of the calibration set samples to obtain the near-infrared spectrum data of the calibration set samples, and record the corresponding measured values;

[0051] Model establishment: Based on the partial least squares method, the modeling band and spectral preprocessing method are accurately selected to establish a mathematical model.

[0052] This application establishes a mathematical model with good prediction accuracy by accurately selecting the modeling band and spectral preprocessing method, providing an important experimental basis for real-time online detection of trace moisture in the electrolyte production process.

[0053] In some specific embodiments, a near-infrared spectrometer can be used to scan the near-infrared spectra of the calibration set samples. Optionally, the near-infrared spectrometer can be a Fourier transform near-infrared spectrometer. Collection method of near-infrared spectra: colorimetric tube transmission liquid sampling. Collection parameters: resolution 2-8cm -1 , scanning times 16-64 times, scanning spectrum range 12500cm -1 ~4000cm -1 , ambient temperature 20-30°C. For example, the resolution can be 2cm -1 , 4cm -1 or 8cm -1 For example, the number of scans can be 16, 32 or 64. The ambient temperature has little effect on the test, and the ambient temperature only needs to meet the use requirements of the near-infrared spectrometer.

[0054] In some specific embodiments, the number of calibration set samples may be greater than 30. For example, the number of calibration set samples may be 30-200, such as 30, 40, 50, 60, 70, 80, 90, 100, 110, 120, 130, 140, 150, 160, 170, 180, 190 or 200. The more the number of calibration set samples, the more conducive to improving the prediction effect.

[0055] In some embodiments, the modeling band covers 5452 cm -1 ~4600cm -1 Band range.

[0056] The selection of modeling bands is very important for the prediction effect of mathematical models. The modeling bands selected in this application cover 5452cm -1 ~4600cm -1 Band range, modeling with such modeling bands is conducive to improving the prediction effect of mathematical models.

[0057] The modeling band covers 5452cm -1 ~4600cm -1 Band range, which means that the modeling band includes at least 5452cm-1 ~4600cm -1 Band range, for example, the modeling band is 7500cm -1 ~4600cm -1 , which covers 5452cm -1 ~4600cm -1 Band range.

[0058] In some embodiments, the modeling band is 12492 cm -1 ~4524cm -1 Alternatively, the modeling band is 7500cm -1 ~4600cm -1 Alternatively, the modeling band is 9400cm -1 ~7500cm -1 and 5452cm -1 ~4600cm -1 Alternatively, the modeling band is 5452cm -1 ~4600cm -1 .

[0059] By optimizing the modeling band, it is helpful to further improve the prediction effect of the mathematical model. Since the full band contains a lot of information, useless information or baseline noise will interfere with the prediction effect of the model. In order to improve the prediction effect of the mathematical model, this application selects the spectral band with effective information of the water content in the analyte in the near-infrared spectrum, that is, the above-mentioned modeling band for modeling, thereby reducing the intake of useless information and improving the prediction effect of the model.

[0060] In some embodiments, the spectral preprocessing method includes one or more of first-order derivative, second-order derivative, multivariate scatter correction, and standard normalization.

[0061] Since some physical states of the electrolyte sample, such as color and turbidity, can cause interferences such as baseline drift or translation and scattering, making the effective information of the band unclear, the present application selects a spectral preprocessing method that can effectively reduce interference to improve the model prediction accuracy. The spectral preprocessing method of the present application can be selected according to the physical state of the electrolyte sample. Of course, the spectral preprocessing method of the present application can also include other spectral preprocessing methods for reducing interference.

[0062] In some embodiments, the moisture content of the calibration set samples is measured in an environment where the dew point temperature is below -10°C, the near-infrared spectra of the electrolyte samples to be tested are scanned, and the near-infrared spectra of the calibration set samples are scanned.

[0063] In order to reduce the measurement error caused by the ambient humidity during the test, the actual moisture content of the electrolyte sample and the collection of the near-infrared spectrum can be carried out in a dry experimental environment with a low dew point temperature. The lower the dew point temperature, the less moisture there is in the air and the drier the air. The present application helps to reduce the measurement error and improve the prediction effect of the model by optimizing the dew point temperature. In particular, when the electrolyte contains lithium bis(fluorosulfonyl)imide, since lithium bis(fluorosulfonyl)imide has strong water absorption, therefore, by controlling the dew point temperature of the experimental environment, it is more conducive to reducing the measurement error for such electrolytes and improving the prediction effect of the model.

[0064] In some embodiments, the dew point temperature may be between -10°C and -60°C.

[0065] In some specific embodiments, the dew point temperature may be -10°C, -20°C, -30°C, -40°C, -50°C, or -60°C.

[0066] By optimizing the dew point temperature, it is helpful to reduce the measurement error caused by ambient humidity during testing.

[0067] In some embodiments, the water content of the calibration set samples is determined using Karl Fischer-Coulometric method.

[0068] The accuracy of the measured value of moisture content affects the prediction effect of the model. The present application uses the Karl Fischer-Coulomb method to accurately measure the moisture content of the calibration set samples, and the obtained results can be used as true values ​​for modeling, which can improve the prediction effect of the model.

[0069] In some embodiments, the electrolyte includes an electrolyte for a secondary battery.

[0070] The determination method of the present application is applicable to detecting the trace moisture content in the electrolyte of a secondary battery. Secondary batteries include lithium ion batteries and sodium ion batteries.

[0071] In some embodiments, after the model is established, the method further includes:

[0072] Model evaluation: The determination coefficient of the mathematical model and the root mean square error of the internal cross-check were used to evaluate the model;

[0073] Model verification: In addition, multiple electrolyte samples were collected as prediction set samples; the moisture content of the prediction set samples was determined by the Karl Fischer-Coulomb method to obtain the measured values; the near-infrared spectra of the prediction set samples were scanned to obtain the spectral data of the prediction set samples; based on the mathematical model and the spectral data of the prediction set samples, the moisture content of the prediction set samples was calculated to obtain the predicted values, and the predicted values ​​were compared with the corresponding measured values.

[0074] Before applying the mathematical model in practice, it is necessary to evaluate and verify the mathematical model. This application uses the coefficient of determination of the mathematical model and the root mean square error of the internal cross-check to evaluate the model. 2 ) is closer to 1, indicating that the model linearity is better. The smaller the root mean square error of the internal cross-validation of the calibration set (RMSECV), the smaller the model prediction deviation. 2 Above 0.99, very close to 1, RMSECV is very small, indicating that the linearity of the model is good and the model prediction bias is small.

[0075] In some specific embodiments, the number of prediction set samples may be 10-50, for example, 10, 15, 20, 25, 30, 35, 40, 45 or 50.

[0076] In some specific embodiments, the moisture content of the prediction set samples is measured and the near infrared spectra of the prediction set samples are scanned in an environment where the dew point temperature is below -10°C. Preferably, the dew point temperature may be -10°C to -60°C.

[0077] In some specific embodiments, a near infrared spectrometer may be used to scan the near infrared spectra of the prediction set samples. Optionally, the near infrared spectrometer may be a Fourier transform near infrared spectrometer. Collection method of the near infrared spectra: colorimetric tube transmission liquid sampling. Collection parameters: resolution 2-8cm -1 , scanning times 16-64 times, scanning spectrum range 12500cm -1 ~4000cm -1 , ambient temperature 20-30°C. For example, the resolution can be 2cm -1 , 4cm -1 or 8cm -1 For example, the number of scans can be 16, 32 or 64. The ambient temperature has little effect on the test, and the ambient temperature only needs to meet the use requirements of the near-infrared spectrometer.

[0078] The moisture content prediction range of the mathematical model of the present application is related to the measured value range of the calibration set samples selected during modeling. In some embodiments, the moisture content prediction range of the mathematical model can be 1 mg / kg-100 mg / kg.

[0079] In some embodiments, the calibration set samples, the prediction set samples and the electrolyte samples to be tested are all samples collected during the production process. The material composition of the electrolyte samples to be tested is the same as the material composition of the calibration set samples, and the water content in the electrolyte samples to be tested is within the water content prediction range of the mathematical model.

[0080] In some embodiments, the modeling software used to establish the mathematical model includes OPUS software from Bruker Corporation.

[0081] The present application is further described below in conjunction with the examples. The following examples describe the disclosure of the present application in more detail, and these examples are merely illustrative, because it is obvious to those skilled in the art that various modifications and variations are made within the scope of the disclosure of the present application. Unless otherwise stated, all reagents used in the examples are commercially available or synthesized according to conventional methods, and the instruments used in the examples are commercially available.

[0082] Example 1

[0083] (1) Collect samples

[0084] 43 batches of electrolyte samples for lithium-ion batteries were collected, and 30 batches of electrolyte samples were randomly selected as calibration set samples, and the remaining 13 batches of electrolyte samples were used as prediction set samples.

[0085] (2) Wet analysis

[0086] The moisture content of 30 batches of calibration set samples and 13 batches of prediction set samples was measured by Karl Fischer-Coulomb moisture meter in a drying room to obtain the actual measured values ​​of moisture content. The dew point temperature of the drying room was -10℃~-60℃, and the ambient temperature was 23℃.

[0087] (3) Near-infrared spectrum acquisition

[0088] In the drying room, the near infrared spectra of 30 batches of calibration set samples and 13 batches of prediction set samples were collected using a Fourier transform near infrared spectrometer to obtain the near infrared spectrum data of the calibration set samples and the prediction set samples, and the corresponding measured values ​​of moisture content were recorded. Spectrum collection method: 8mm colorimetric tube transmission liquid sampling. Collection parameters: resolution 8cm -1 , scanning times 64 times, scanning range 12500cm -1 ~4000cm -1 , temperature 23℃. The near infrared spectra overlap of 30 batches of calibration set samples are shown in Figure 1 shown.

[0089] (4) Modeling

[0090] The OPUS software of Bruker was used to establish a mathematical model based on the partial least squares method, selecting the modeling band, spectral preprocessing method, and the number of principal component factors. The model parameters are as follows: Modeling band range: 9400cm -1 ~7500cm -1 ; 5452cm -1 ~4600cm -1 ; Spectral preprocessing method: first-order derivative; The number of principal component factors is 4.

[0091] (5) Model evaluation

[0092] The coefficient of determination R of the mathematical model 2 The internal cross-validation root mean square error RMSECV is 0.9958, the model has good linearity, and the RMSECV is small, indicating that the model has good predictive ability.

[0093] (6) Model verification

[0094] Based on the mathematical model established in step (4) and the spectral data of the prediction set samples obtained in step (3), the moisture content of the 13 batches of prediction set samples is calculated to obtain the predicted values. The predicted values ​​of the 13 batches of prediction set samples are compared with the corresponding measured values ​​of moisture content obtained in step (2), as shown in Table 1.

[0095] Table 1

[0096]

[0097] Methodological investigation

[0098] 1. Precision test

[0099] A batch of electrolyte samples not involved in modeling were randomly selected, and the mathematical model established in Example 1 was used to repeat the test 10 times, and the model prediction results were recorded. The relative standard deviation RSD value of the 10 results was 0.09%, indicating that the precision of the model was good.

[0100] 2. Repeatability test

[0101] A batch of electrolyte samples that do not participate in the modeling are randomly selected, and the samples are placed in 6 colorimetric tubes. The mathematical model established in Example 1 is used to predict the moisture content of the samples in the 6 colorimetric tubes. The RSD value of the 6 prediction results of the same sample is 0.53%, indicating that the repeatability of the method of the present application is good.

[0102] Embodiment 2-6

[0103] The method described in Example 1 was followed, and the calibration set samples and prediction set samples used were the same as those in Example 1. The only difference was that the selected modeling bands were different from those in Example 1, as shown in Table 2 below.

[0104] The coefficient of determination R of the mathematical model 2 The internal cross-validation root mean square error RMSECV is shown in Table 2 below.

[0105] Table 2

[0106] serial number <![CDATA[Modeling band / cm -1 > principal component <![CDATA[R 2 ]]> RMSECV Example 1 9400～7500；5452～4600 4 0.9958 1.03 Example 2 7500～4600 5 0.9953 1.09 Example 3 5452～4600 4 0.9956 1.06 Example 4 12492～4524 7 0.9943 1.29 Example 5 12500～4000 10 0.2391 13.9 Example 6 9400～7500 1 0.0000 14.7

[0107] Embodiment 7-11

[0108] The method described in Example 1 was followed. The calibration set samples and prediction set samples used were the same as those in Example 1. The only difference was that the spectral preprocessing method was different, as shown in Table 3 below.

[0109] Table 3

[0110] serial number Spectral preprocessing methods principal component <![CDATA[R 2 ]]> RMSECV Example 1 First Derivative 4 0.9958 1.03 Example 7 Second Derivative 7 0.9951 1.12 Example 8 Multivariate Scattering Correction 5 0.9943 1.20 Example 9 Standard normalization method 6 0.9945 1.19 Example 10 First derivative + multivariate scatter correction 5 0.9942 1.21 Embodiment 11 Min-max Normalization 6 0.9910 1.40

[0111] Example 12

[0112] The method described in Example 1 was followed. The calibration set samples and prediction set samples used were the same as those in Example 1. The only difference was that the dew point temperature of the drying room was different, as shown in Table 5 below.

[0113] Table 5

[0114] serial number Dew point temperature / ℃ <![CDATA[R 2 ]]> RMSECV Example 1 -10～-60 0.9958 1.03 Example 12 14～18 0.9696 2.26

[0115] It can be seen from the above embodiments that the present application uses the established mathematical model to predict the prediction set samples, has a good prediction effect, and can provide an important experimental basis for the real-time online detection of trace moisture in the production process of secondary battery electrolyte.

[0116] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for determining trace water content in an electrolyte based on near infrared spectroscopy, characterized in that: The following steps are involved: A mathematical model is established based on the near-infrared spectral data of the calibration set samples and the measured values ​​of water content; Scanning the near infrared spectrum of the electrolyte sample to be tested to obtain spectrum data of the electrolyte sample to be tested; Based on the mathematical model and the spectral data of the electrolyte sample to be tested, the trace moisture content in the electrolyte sample to be tested is calculated.

2. The method according to claim 1, characterized in that Establishing the mathematical model comprises the following steps: Sample preparation: collecting a plurality of electrolyte samples as the calibration set samples; Determination of true moisture content: Determine the moisture content of the samples in the calibration set to obtain the actual measured value of the moisture content; Spectrum acquisition: scanning the near infrared spectrum of the calibration set samples to obtain the near infrared spectrum data of the calibration set samples, and recording the corresponding measured values; Model establishment: Based on the partial least squares method, the modeling band and spectral preprocessing method are accurately selected to establish a mathematical model.

3. The method according to claim 2, characterized in that The modeling band covers 5452cm -1 ~4600cm -1 Band range.

4. The method according to claim 3, characterized in that The modeling band is 12492cm -1 ~4524cm -1 ; Alternatively, the modeling band is 7500cm -1 ~4600cm -1 ; Alternatively, the modeling band is 9400cm -1 ~7500cm -1 and 5452cm -1 ~4600cm -1 ; Alternatively, the modeling band is 5452cm -1 ~4600cm -1 .

5. The method according to any one of claims 2 to 4, characterized in that: The spectrum preprocessing method includes one or more of the first-order derivative, the second-order derivative, the multivariate scattering correction, and the standard normalization method.

6. The method according to any one of claims 2 to 5, characterized in that: The moisture content of the calibration set samples is measured in an environment where the dew point temperature is below -10°C, the near-infrared spectrum of the electrolyte sample to be tested is scanned, and the near-infrared spectrum of the calibration set samples is scanned.

7. The method according to claim 6, characterized in that The dew point temperature is -10°C to -60°C.

8. The method according to any one of claims 2 to 7, characterized in that: The water content of the calibration set samples was determined using the Karl Fischer-Coulomb method.

9. The method according to any one of claims 1 to 8, characterized in that The electrolyte includes an electrolyte for a secondary battery.

Images