Method and system for determining purity of electronic-grade hexafluoroethane

Abstract

The invention discloses a method for determining the purity of electronic-grade hexafluoroethane, and relates to the technical field of the purity of hexafluoroethane, and the method comprises the following steps: vacuumizing a specific sampling container to 10 <-3 > Pa to 10 <-2 > Pa, forming a super-hydrophobic coating on the inner wall, sampling in a temperature-controlled, humidity-controlled and pressure-controlled environment, and removing water and impurities through two-stage pretreatment; enhancing analysis by using a gas chromatograph, a specific carrier gas, a column temperature program, a micro-fluidic chip and an ion mobility spectrometry; then secondary scanning analysis is carried out by using a mass spectrometer double ion source and a calibration system; and finally, learning and matching an algorithm through a data processing system machine. Samples can be accurately collected and pretreated, gas chromatography and mass spectrometry analysis is more accurate and efficient, a data processing system can accurately determine impurities and purity, multiple measurement data can be counted and analyzed, the result reliability is improved, all parts of the system are optimized and matched, the accuracy and stability of the whole process are guaranteed, and the purity is calculated; the method further comprises statistical analysis of multiple measurement data.

Description

Technical Field

[0001] The invention relates to the technical field of hexafluoroethane purification, and in particular to a method and system for determining the purity of electronic-grade hexafluoroethane. Background Art

[0002] In the field of modern electronics industry, electronic grade hexafluoroethane plays a vital role as a key electronic gas. It is widely used in plasma etching, chemical vapor deposition and other processes in semiconductor manufacturing, and has a direct impact on the precision and performance of chip manufacturing. For example, in the plasma etching process, the purity of electronic grade hexafluoroethane is directly related to the accuracy and uniformity of etching. If the purity is insufficient, it may cause defects in chip circuits.

[0003] However, accurately determining the purity of electronic-grade hexafluoroethane faces many challenges. Traditional analytical methods have limitations when faced with electronic-grade purity requirements.

[0004] In terms of sample collection, early sampling containers often did not take into account the special properties of hexafluoroethane. Containers made of ordinary materials may react with hexafluoroethane or fail to effectively prevent the introduction of impurities. For example, some plastic containers may release trace organic impurities, and metal containers may produce metal ion impurities due to corrosion. In addition, the temperature and humidity control of the sampling environment is not precise enough, resulting in the collected samples containing impurities such as water condensed or adsorbed due to temperature and humidity changes, affecting the subsequent analysis results.

[0005] For the pretreatment stage, traditional methods mostly use a single desiccant or a simple filtering method. The choice of desiccant may not completely remove moisture, and may become ineffective after long-term use without being discovered in time. Improper selection of the pore size and material of the filter material cannot effectively intercept tiny particle impurities, which may block the instrument channel or interfere with the analysis results during subsequent analysis. In addition, there is a lack of consideration for factors such as static electricity, which may adsorb impurities and change the original state of the sample.

[0006] In terms of analytical technology, traditional gas chromatography analysis has shortcomings. The types and performance of chromatographic columns are limited, and the separation effect of some complex impurities and hexafluoroethane is not good. The control of the injection port temperature, detector temperature and column temperature is not precise enough, and the program temperature setting is unreasonable, resulting in problems such as peak broadening and reduced separation. Moreover, the control of the injection volume is not accurate enough, which affects the repeatability and accuracy of the analysis. Traditional mass spectrometry also has defects. The ion source type is single and cannot meet the ionization requirements of various types of impurities. The mass scanning range and scanning mode are not flexible enough, which makes it difficult to identify some low-content or special structure impurities. At the same time, due to the lack of a real-time calibration system, the mass accuracy of the mass spectrometer will decrease with time and changes in usage conditions, affecting the reliability of the analysis results.

[0007] In addition, in terms of data processing, traditional methods are mostly simple data recording and manual calculation, which do not make full use of modern data processing technology. For data measured multiple times, it is impossible to effectively remove outliers and conduct accurate statistical analysis, and it is impossible to accurately evaluate the uncertainty of the measurement results, which reduces the credibility of the final purity results and makes it difficult to meet the needs of high-precision purity determination of electronic-grade hexafluoroethane.

[0008] As the electronics industry's requirements for chip manufacturing precision continue to increase, there is an urgent need for a more accurate and reliable method and system for determining the purity of electronic-grade hexafluoroethane to ensure the high-quality development of electronics industry production. Summary of the invention

[0009] 1. Technical issues to be resolved

[0010] In view of the deficiencies in the prior art, the present invention provides a method and system for determining the purity of electronic-grade hexafluoroethane.

[0011] (II) Technical solution

[0012] A method for determining the purity of electronic grade hexafluoroethane comprises the following steps:

[0013] S1. Sample collection and pretreatment

[0014] S1.1. Use a sampling container with a multi-layer composite structure, the inner layer of which is polytetrafluoroethylene, the middle layer is a stainless steel mesh reinforced structure, and the outer layer is a weather-resistant polymer material. Collect electronic grade hexafluoroethane samples in a sampling environment with a temperature controlled at 20°C to 25°C and a humidity controlled at 30% to 40%. The sampling container is equipped with an automatic pressure balancing device, which can keep the difference between the internal pressure and the ambient pressure within the range of ±500Pa during the collection process.

[0015] S1.2. The collected samples are passed through a two-stage pretreatment device, the first stage of which is a drying tube filled with a mixed desiccant of activated alumina and silica gel (mass ratio is 2:1), and the second stage is a polytetrafluoroethylene filter membrane with a pore size of 0.1-0.3 microns, to remove moisture and particulate impurities in the samples; at the same time, a negative ion generator is set in the pretreatment device to generate negative ions with a concentration of 10 4 -10 5 Pieces / cm 3 , used to neutralize possible static electricity and reduce impurity adsorption.

[0016] S2. Gas chromatography analysis

[0017] S2.1, injecting the pretreated sample into a gas chromatograph, the gas chromatograph is equipped with a composite chromatographic column of a PLOT-U type porous layer open tubular column and an Al2O3 / KCl column connected in series, the column lengths are 30 meters and 15 meters respectively, the inner diameter is 0.32 mm, and the stationary phase is a high-purity Al2O3 and KCl coating;

[0018] S2.2, set the injection port temperature of the gas chromatograph to 150°C to 180°C, the detector temperature to 250°C to 300°C, and the column temperature to a programmed temperature increase, with the initial temperature at -50°C, maintained for 5 minutes, then increased to 80°C at a rate of 10°C / min, maintained for 10 minutes, and then increased to 120°C at a rate of 5°C / min, maintained for 8 minutes;

[0019] S2.3. A microfluidic chip is installed in the injection system of the gas chromatograph to precisely control the injection volume within a range of 0.1-1 μL. An ion mobility spectrum enhancement module is installed between the chromatographic column outlet and the detector to enhance the ion migration velocity, improve the separation degree and detection sensitivity by applying an electric field of 500-1000 V / cm. The relative content of each component is determined based on the peak area or peak height of the main hexafluoroethane peak and impurity peak in the gas chromatogram.

[0020] S3. Mass spectrometry analysis

[0021] S3.1, introducing the pretreated sample into a high-resolution mass spectrometer, wherein the ion source of the mass spectrometer is a dual ion source system of an electrospray ion source and a chemical ion source, and the ionization energy is 10-50 eV, which can be switched for use;

[0022] S3.2, the scanning mass range is from 10 to 200 amu, and the data-dependent secondary mass spectrometry scanning mode is adopted during the scanning process. When an ion peak with a specific mass number is detected, the secondary mass spectrometry analysis is automatically triggered to obtain more detailed structural information;

[0023] S3.3. The mass spectrometer is equipped with a real-time calibration system. By introducing an internal standard of known mass (such as perfluorobutane), mass calibration is performed every 10 minutes with a calibration accuracy within ±0.1amu. The impurity components in the sample are identified by analyzing the ion peaks in the mass spectrum, and the type and content of the impurities are further determined in combination with the gas chromatography analysis results.

[0024] S4. Purity calculation

[0025] According to the contents of hexafluoroethane and impurities determined by gas chromatography and mass spectrometry, the weighted calculation method is used: purity = (hexafluoroethane content × weight factor 1) / (hexafluoroethane content × weight factor 1 + ∑ (impurity content × corresponding impurity weight factor)) × 100%, where weight factor 1 is 0.8-0.9, and the impurity weight factor is determined according to the degree of influence of impurities on the performance of electronic-grade products, ranging from 0.1-0.5, and the purity of electronic-grade hexafluoroethane is calculated; at the same time, the uncertainty of the calculation results is evaluated, and the uncertainty range is controlled within ±0.5% considering factors such as measurement repeatability, instrument calibration, and sample uniformity.

[0026] Furthermore, in S1, the sampling container is evacuated before sampling, and the vacuum degree reaches 10-3Pa-10-2Pa, and the inner wall of the sampling container is coated with a super hydrophobic coating, the coating material is a fluorine-containing polymer, and the water contact angle is greater than 150°.

[0027] Furthermore, in S2, the carrier gas of the gas chromatograph is a mixed gas of high-purity nitrogen and helium (volume ratio of 4:1), the purity of which reaches 99.999%, and the carrier gas flow rate is 1-3 ml / min; a gas circulation system is provided in the constant temperature box of the chromatographic column to form a uniform temperature and flow rate distribution of the carrier gas in the column.

[0028] Furthermore, in S3, the mass spectrometer is equipped with a data processing system that can automatically identify and match ion peaks in the mass spectrum with standard mass spectra of known compounds, and the matching threshold is set at 90%-95%; at the same time, the data processing system uses a machine learning algorithm to learn and train a large amount of mass spectrometry data to improve the accuracy and efficiency of impurity identification.

[0029] Further, a system for determining the purity of electronic grade hexafluoroethane comprises:

[0030] Sampling and preprocessing unit

[0031] The sampling container is made of a multi-layer composite structural material, and is equipped with a temperature, humidity and pressure control device and a negative ion generator. It is used to collect samples under a specific environment and is evacuated by a vacuum pumping device before sampling. The vacuum pumping device can evacuate the vacuum degree to 10-3Pa-10-2Pa. The inner wall of the sampling container is coated with a super-hydrophobic coating.

[0032] The pretreatment device is connected to the sampling container and is equipped with two-stage pretreatment components (activated alumina and silica gel mixed desiccant, polytetrafluoroethylene filter membrane) for removing moisture and particulate impurities in the sample.

[0033] Analysis Unit

[0034] Gas chromatograph, equipped with a composite chromatographic column, an injection port, a detector, a temperature control system, a microfluidic chip and an ion mobility spectrometry enhancement module. The carrier gas system provides a mixed gas of high-purity nitrogen and helium (volume ratio of 4:1) for gas chromatography analysis of the pretreated samples;

[0035] A mass spectrometer, having a dual ion source system, a real-time calibration system, a specific mass scanning range and a data-dependent secondary mass spectrometry scanning mode, is connected to a gas chromatograph for mass spectrometric analysis of the samples.

[0036] Data processing and computing unit

[0037] A data acquisition module, used to collect data from a gas chromatograph and a mass spectrometer;

[0038] The data processing module processes the collected data based on machine learning algorithms and specific matching algorithms to determine the impurity composition and content; it also performs statistical analysis on multiple measurement data, removes outliers, and calculates the average and standard deviation;

[0039] The purity calculation module calculates the purity of electronic grade hexafluoroethane using a weighted calculation method based on the processed data and performs uncertainty assessment.

[0040] Furthermore, the sampling container of the sampling and pretreatment unit is connected to the pretreatment device via a heated transmission pipeline, and the pipeline temperature is maintained at 30° C.-50° C. to prevent condensation of the sample due to temperature changes during transmission.

[0041] Furthermore, the interface device between the gas chromatograph and the mass spectrometer of the analysis unit adopts a heating interface with a temperature controlled at 100°C-150°C, and is provided with a pressure sensor and a flow regulating valve to achieve efficient transmission, pressure matching and flow control of the sample, thereby ensuring the accuracy of the analysis; the inner surface of the interface device is coated with an inert coating to prevent sample adsorption and reaction.

[0042] Furthermore, the data processing module of the data processing and computing unit is connected to a remote server, and data can be uploaded to the server for big data analysis and model updating; the server side can optimize the data processing algorithm according to the usage feedback of different users, thereby improving the versatility and accuracy of the system.

[0043] (III) Beneficial technical effects

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] The present invention adopts a special multi-layer composite structure sampling container, equipped with an automatic pressure balance device and a negative ion generator, which can effectively avoid the reaction between the sample and the container, reduce the introduction of impurities and electrostatic adsorption, and accurately control the temperature and humidity of the sampling environment to ensure the quality of the collected samples. In the pretreatment stage, the two-stage pretreatment device can more thoroughly remove moisture and particulate impurities, further improving the purity of the sample.

[0046] In gas chromatography analysis, the use of composite chromatographic columns, optimized temperature programs and injection port and detector temperature settings, combined with microfluidic chips to accurately control the injection volume and ion mobility spectrum enhancement modules, significantly improves the separation and detection sensitivity, and can more accurately determine the relative content of each component. The mass spectrometer design of the dual ion source system, data-dependent secondary mass spectrometry scanning mode, and real-time calibration system has broadened the detection range, enhanced the impurity identification ability and mass accuracy.

[0047] The data processing system based on machine learning and matching algorithms can accurately determine the composition and content of impurities. The machine learning algorithm uses a large amount of data training to process complex feature relationships; the matching algorithm is combined with the standard spectrum library to achieve high-precision matching. At the same time, statistical analysis of multiple measurement data can remove outliers, calculate the average and standard deviation, and perform uncertainty assessment to improve the reliability of the results.

[0048] In addition, the synergistic effect of various parts of the system, such as the efficient cooperation between the sampling and pretreatment unit, the analysis unit and the data processing and calculation unit, as well as the innovative design within each unit (such as the heating transmission pipeline between the sampling container and the pretreatment device, the heating interface between the analysis units and related control devices), further ensures the accuracy and stability of the entire process from sample collection to result output, meets the requirements for high-precision purity determination of electronic-grade hexafluoroethane, and provides strong support for electronic industry production. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 The present invention provides a flow chart of a method for determining the purity of electronic grade hexafluoroethane. DETAILED DESCRIPTION

[0050] Reference Figure 1 The present invention proposes a method for determining the purity of electronic grade hexafluoroethane, comprising the following steps:

[0051] S1. Sample collection and pretreatment

[0052] S1.1. Use a sampling container with a multi-layer composite structure, the inner layer of which is polytetrafluoroethylene, the middle layer is a stainless steel mesh reinforced structure, and the outer layer is a weather-resistant polymer material. Collect electronic grade hexafluoroethane samples in a sampling environment with a temperature controlled at 20°C to 25°C and a humidity controlled at 30% to 40%. The sampling container is equipped with an automatic pressure balancing device, which can keep the difference between the internal pressure and the ambient pressure within the range of ±500Pa during the collection process.

[0053] S1.2. The collected samples are passed through a two-stage pretreatment device, the first stage of which is a drying tube filled with a mixed desiccant of activated alumina and silica gel (mass ratio is 2:1), and the second stage is a polytetrafluoroethylene filter membrane with a pore size of 0.1-0.3 microns, to remove moisture and particulate impurities in the samples; at the same time, a negative ion generator is set in the pretreatment device to generate negative ions with a concentration of 10 4 -10 5 Pieces / cm 3 , used to neutralize possible static electricity and reduce impurity adsorption.

[0054] S2. Gas chromatography analysis

[0055] S2.1, injecting the pretreated sample into a gas chromatograph, the gas chromatograph is equipped with a composite chromatographic column of a PLOT-U type porous layer open tubular column and an Al2O3 / KCl column connected in series, the column lengths are 30 meters and 15 meters respectively, the inner diameter is 0.32 mm, and the stationary phase is a high-purity Al2O3 and KCl coating;

[0056] S2.2, set the injection port temperature of the gas chromatograph to 150°C to 180°C, the detector temperature to 250°C to 300°C, and the column temperature to a programmed temperature increase, with the initial temperature at -50°C, maintained for 5 minutes, then increased to 80°C at a rate of 10°C / min, maintained for 10 minutes, and then increased to 120°C at a rate of 5°C / min, maintained for 8 minutes;

[0057] S2.3. A microfluidic chip is installed in the injection system of the gas chromatograph to precisely control the injection volume within a range of 0.1-1 μL. An ion mobility spectrum enhancement module is installed between the chromatographic column outlet and the detector to enhance the ion migration velocity, improve the separation degree and detection sensitivity by applying an electric field of 500-1000 V / cm. The relative content of each component is determined based on the peak area or peak height of the main hexafluoroethane peak and impurity peak in the gas chromatogram.

[0058] S3. Mass spectrometry analysis

[0059] S3.1, introducing the pretreated sample into a high-resolution mass spectrometer, wherein the ion source of the mass spectrometer is a dual ion source system of an electrospray ion source and a chemical ion source, and the ionization energy is 10-50 eV, which can be switched for use;

[0060] S3.2, the scanning mass range is from 10 to 200 amu, and the data-dependent secondary mass spectrometry scanning mode is adopted during the scanning process. When an ion peak with a specific mass number is detected, the secondary mass spectrometry analysis is automatically triggered to obtain more detailed structural information;

[0061] S3.3. The mass spectrometer is equipped with a real-time calibration system. By introducing an internal standard of known mass (such as perfluorobutane), mass calibration is performed every 10 minutes with a calibration accuracy within ±0.1amu. The impurity components in the sample are identified by analyzing the ion peaks in the mass spectrum, and the type and content of the impurities are further determined in combination with the gas chromatography analysis results.

[0062] S4. Purity calculation

[0063] According to the contents of hexafluoroethane and impurities determined by gas chromatography and mass spectrometry, the weighted calculation method is used: purity = (hexafluoroethane content × weight factor 1) / (hexafluoroethane content × weight factor 1 + ∑ (impurity content × corresponding impurity weight factor)) × 100%, where weight factor 1 is 0.8-0.9, and the impurity weight factor is determined according to the degree of influence of impurities on the performance of electronic-grade products, ranging from 0.1-0.5, and the purity of electronic-grade hexafluoroethane is calculated; at the same time, the uncertainty of the calculation results is evaluated, and the uncertainty range is controlled within ±0.5% considering factors such as measurement repeatability, instrument calibration, and sample uniformity.

[0064] In the present invention, step S1 involves special treatment of the sampling container. Before sampling, the sampling container must be evacuated, and the vacuum degree must strictly reach the range of 10-3Pa-10-2Pa. This high vacuum degree requirement is to eliminate the impurity gas that may exist in the container to the greatest extent, so as to avoid contamination of the electronic grade hexafluoroethane sample collected later.

[0065] At the same time, the inner wall of the sampling container is coated with a super hydrophobic coating, which is made of fluoropolymer. Fluoropolymer has extremely low surface energy, which makes the coating surface super hydrophobic, with a water contact angle greater than 150°. This super hydrophobic coating can effectively prevent trace amounts of water from adhering to the inner wall of the container, ensuring that the sample is not affected by moisture even when the humidity of the sampling environment fluctuates slightly. Moreover, this characteristic also helps to reduce the adsorption of other polar impurities on the inner wall of the container, further ensuring the purity and original state of the sample.

[0066] In step S2, the selection of the carrier gas for the gas chromatograph is carefully considered. The carrier gas is a mixture of high-purity nitrogen and helium, with a volume ratio of 4:1. This mixing ratio is determined through a large number of experiments, and can meet the needs of electronic-grade hexafluoroethane analysis while ensuring that the carrier gas has suitable physical and chemical properties. Among them, the purity of nitrogen and helium both reaches 99.999%. Such high purity can avoid the introduction of impurities into the carrier gas itself and affect the analysis results.

[0067] The carrier gas flow rate is controlled within the range of 1-3 ml / min. This flow rate range is determined based on many factors such as the type, length, inner diameter of the chromatographic column and the properties of the sample. The appropriate carrier gas flow rate is essential to maintain a stable chromatographic separation effect, which can ensure that the retention time and peak shape of the sample in the chromatographic column are optimal.

[0068] In addition, a gas circulation system is set up in the thermostat of the chromatographic column. Through special design and layout, the system can form a uniform temperature and flow rate distribution of the carrier gas in the column. Specifically, the gas circulation system uses a small fan or gas guide device to form a stable airflow circulation of the carrier gas in the thermostat. This circulation can effectively eliminate the temperature gradient and flow rate unevenness that may exist in the thermostat, thereby ensuring that the carrier gas can stably transport the sample through the chromatographic column during the entire chromatographic analysis process, improving the separation efficiency and the accuracy of the analysis results.

[0069] In step S3, the advanced data processing system equipped with the mass spectrometer is one of the key parts of the present invention. The data processing system can automatically identify and match the ion peaks in the mass spectrum with the standard mass spectrum of known compounds, and the matching threshold is set in the range of 90%-95%. This threshold range is determined based on a large amount of experimental data and actual application scenarios, which can ensure accurate identification of impurities that may exist in the sample and avoid misjudgment caused by overly strict or loose matching conditions.

[0070] The data processing system uses a machine learning algorithm to learn and train a large amount of mass spectrometry data. In the machine learning process, a large amount of electronic-grade hexafluoroethane mass spectrometry data of different purities and impurity contents are first collected as training sets. These data cover various possible impurity types and concentration ranges. Then, select a suitable machine learning model, such as a deep neural network (DNN) or a support vector machine (SVM). Through repeated learning of the training set data, the model can automatically extract the characteristic patterns of ion peaks in the mass spectrum, including the intensity, mass number, isotope distribution, and the relationship between them. This machine learning-based method can greatly improve the accuracy and efficiency of impurity identification, especially for some complex structures or low-content impurities, which can be detected and identified more keenly, as follows:

[0071] A. Determine impurity composition and content based on machine learning algorithms

[0072] Data preparation and feature extraction

[0073] Data collection and collation: Obtain raw data from gas chromatographs and mass spectrometers, including peak area and retention time in chromatograms, ion peak mass number and intensity in mass spectra, etc. Standardize these data to unify data formats and dimensions.

[0074] Feature Engineering:

[0075] For gas chromatography data, extract features such as peak shape parameters (including peak width, peak symmetry, etc.), relative retention time differences with known standards, etc.

[0076] For mass spectrometry data, in addition to the basic parameters of the ion peak, the isotope distribution characteristics of the ion peak, the mass difference characteristics of adjacent ion peaks, etc. are also extracted. These features are combined into a multidimensional feature vector for subsequent machine learning model training.

[0077] Model selection and training

[0078] Model selection: Use an integrated learning algorithm, such as a combination of random forest and gradient boosting decision tree. Random forest can handle high-dimensional data and is not prone to overfitting, while gradient boosting decision tree performs well in improving prediction accuracy. In addition, convolutional neural network (CNN) in deep learning can also be combined to process the two-dimensional data structure of mass spectra and extract more complex features.

[0079] Training data preparation: Collect a large amount of electronic-grade hexafluoroethane sample data with known purity and impurity content. These data are calibrated using standard analytical methods and high-precision instruments. Divide the sample data into a training set and a validation set according to a certain ratio (e.g. 70% for training and 30% for validation).

[0080] Model training: Use the training set to train the selected model, and minimize the prediction error of the model on the validation set by adjusting the model parameters (such as the number of trees in the random forest, the depth of the decision tree, the convolution kernel size and number of layers of the CNN, etc.). Cross-validation technology is used during the training process to further improve the generalization ability of the model.

[0081] Impurity composition and content prediction

[0082] For unknown samples, the feature vector after feature extraction is input into the trained model. The model outputs the predicted impurity components and their contents. For each predicted impurity component, the model will give a confidence score, indicating the reliability of the prediction result. According to the set threshold (such as the confidence score is greater than 80%), the final impurity component and content prediction results are determined.

[0083] B. Determine impurity composition and content based on matching algorithm

[0084] Establish impurity standard spectrum library

[0085] Collect various impurity standards that may exist in electronic grade hexafluoroethane, analyze these standards using high-resolution mass spectrometers and gas chromatographs, and obtain their standard spectra (including gas chromatograms and mass spectra) under different conditions (such as different chromatographic columns and mass spectrometry parameters). Store these standard spectra in a database and create an index. The index information includes the impurity name, chemical formula, and main characteristic peak information.

[0086] Matching algorithm design

[0087] Chromatogram matching: An algorithm based on peak matching is used. First, the chromatogram of the unknown sample is peak detected to determine the position (retention time) and intensity of the peaks. Then, these peaks are matched with the chromatograms in the standard spectrum library. The peak retention time deviation (a certain tolerance range is allowed, such as ±0.1 minutes) and the peak intensity ratio are considered in the matching process. The matching score of each standard spectrum and the chromatogram of the unknown sample is calculated, and the higher the score, the greater the possibility of matching.

[0088] Mass spectrum matching: For mass spectra, a matching algorithm based on ion peak mass number and relative intensity is used. The ion peaks in the mass spectrum of the unknown sample are compared with the mass spectra in the standard spectrum library. The mass number deviation of the ion peak (within the mass accuracy range of the instrument, such as ±0.1amu) and the similarity of the relative intensity are considered. At the same time, the characteristic ion peaks in the mass spectrum (such as fragment ion peaks unique to certain impurities) are combined for accurate matching. The degree of matching is evaluated by calculating the matching score.

[0089] Determination of impurity composition and content

[0090] According to the matching results of the chromatogram and the mass spectrum, the matching scores of the two are comprehensively considered. When the matching scores of the two exceed the set threshold (such as the chromatogram matching score is greater than 70%, and the mass spectrum matching score is greater than 80%), the impurity component is determined. For content determination, the impurity content in the unknown sample can be calculated based on the ratio of the impurity peak to the main peak of hexafluoroethane in the matched standard spectrum, combined with the concentration information of the known standard.

[0091] C. Statistical analysis of multiple measurement data

[0092] Data collection and organization

[0093] The same batch of electronic grade hexafluoroethane samples is measured multiple times (e.g., n times, n ≥ 10), and complete gas chromatography and mass spectrometry data are obtained for each measurement. These measurement data are numbered and stored in the order of measurement time, and each data point contains all analysis parameters (such as peak area, retention time, ion peak mass number and intensity, etc.).

[0094] Outlier detection and removal

[0095] Outlier detection based on statistical methods: Calculate the average value of each parameter in each measurement data and standard deviation σ. For a data point, if the value of a parameter exceeds If k is within the range of 3 (usually 3), the data point is considered to be an outlier on this parameter. This test is performed on all parameters, and the data points that are judged to be outliers are marked.

[0096] Outlier detection based on model prediction: Input each measurement data into the trained machine learning model to obtain the impurity composition and content predicted by the model. Calculate the error between the predicted value and the actual measured value. If the error of a data point exceeds the set threshold (such as a relative error greater than 20%), the data point is marked as an outlier.

[0097] Outlier removal: All data points marked as outliers are removed from the data set to obtain a cleaned data set.

[0098] Mean and standard deviation calculation

[0099] Average value calculation: For the cleaned data set, calculate the average values ​​of the hexafluoroethane main peak and impurity peak related parameters (such as peak area, content, etc.). For each parameter, the average value calculation formula is: Where m is the number of cleaned data points, x i is the parameter value of the ith data point.

[0100] Standard Deviation Calculation: Use the following formula to calculate the standard deviation: The standard deviation reflects the degree of dispersion of the measured data and is used to evaluate the stability and reliability of the measurement results. Through these statistical analyses, more accurate results of the purity and impurity content of electronic grade hexafluoroethane can be obtained.

[0101] The sampling container is made of multi-layer composite structural materials. This multi-layer structural design has multiple advantages. The inner layer is made of materials with good chemical stability to hexafluoroethane, such as polytetrafluoroethylene, which can directly contact the sample without reacting. The middle layer is a stainless steel mesh reinforcement structure, which is mainly used to improve the mechanical strength of the container and prevent the container from being damaged by external forces during sampling and transportation. The outer layer is a weather-resistant polymer material, which can protect the container from external environmental factors (such as light, oxidation, etc.).

[0102] The sampling container is equipped with temperature, humidity and pressure control devices and a negative ion generator. The temperature control device can accurately control the temperature of the sampling environment within the set range. The principle is to monitor the temperature in real time through the built-in temperature sensor and adjust it using heating or cooling elements. The humidity control device works together through hygroscopic materials and humidity sensors. When the humidity exceeds the set value, the hygroscopic material automatically absorbs moisture to ensure that the humidity of the sampling environment is appropriate. The pressure control device can maintain the pressure in the container stable by automatically adjusting valves according to the pressure changes during the sampling process. The negative ion generator produces negative ions with a concentration of a specific range (for example, 10 4 -10 5 Pieces / cm 3 ) which can neutralize the static electricity that may exist in the sampling environment, thereby reducing impurities introduced by electrostatic adsorption.

[0103] Before sampling, the sampling container is evacuated by a vacuum pump, which has high-precision vacuum degree measurement and adjustment functions. It uses a vacuum pump to extract the gas in the container and uses a vacuum gauge to monitor the vacuum degree in real time. When it reaches 10 -3 Pa-10 -2 When the target vacuum degree reaches Pa, the pumping operation will be automatically stopped.

[0104] The pretreatment device is connected to the sampling container, and is equipped with two-stage pretreatment components. The first stage is a mixed desiccant of activated alumina and silica gel, which are mixed in a specific mass ratio (for example, 2:1). Activated alumina has a large specific surface area and adsorption performance, and can absorb a large amount of moisture; silica gel has a good adsorption effect on trace moisture. The combination of the two can more fully remove moisture from the sample. The second stage is a polytetrafluoroethylene filter membrane with a pore size within a specific range (for example, 0.1-0.3 microns). This filter membrane can effectively intercept particulate impurities in the sample and prevent them from entering subsequent analytical instruments.

[0105] The gas chromatograph plays an important role in separation and analysis in the entire system. It is equipped with a composite chromatographic column, which is composed of different types of chromatographic columns, such as a PLOT-U porous layer open tubular column and an Al2O3 / KCl column in series. Different types of chromatographic columns have different separation effects on substances of different properties. This combination can improve the separation ability of impurities in electronic grade hexafluoroethane and hexafluoroethane itself.

[0106] The injection port, detector and temperature control system work together to ensure the accuracy of chromatographic analysis. The injection port temperature can be precisely adjusted between 150°C and 180°C. This temperature range is determined based on the properties of the sample such as boiling point and volatility. The appropriate injection port temperature can ensure that the sample vaporizes quickly and enters the chromatographic column evenly. The detector temperature is controlled at 250°C to 300°C. This temperature setting can ensure that the detector has good response performance to the sample components flowing out of the chromatographic column. The column temperature adopts a programmed temperature rise method, with an initial temperature of -50°C and maintained for 5 minutes. This is to allow low-boiling impurities and hexafluoroethane to have enough time to achieve preliminary separation at the front end of the chromatographic column in the initial stage. Then the temperature is raised to 80°C at a rate of 10°C / min, maintained for 10 minutes, and then raised to 120°C at a rate of 5°C / min and maintained for 8 minutes. This complex programmed temperature rise process can better separate impurities and hexafluoroethane in different boiling point ranges, so that they form sharp and well-separated peaks on the chromatographic column.

[0107] In addition, the gas chromatograph is also equipped with a microfluidic chip and an ion mobility spectrometry enhancement module. The microfluidic chip is installed in the injection system. Through structures such as microchannels and microvalves, it can control the injection volume extremely accurately, with an injection volume range of 0.1-1 microliters. The ion mobility spectrometry enhancement module is set between the chromatographic column outlet and the detector. By applying an electric field of a specific intensity (500-1000V / cm), it can enhance the speed difference of ions during migration, thereby improving separation and detection sensitivity, so that some trace impurities that are difficult to separate and detect in traditional chromatographic analysis can be accurately analyzed.

[0108] The carrier gas system provides a high-purity nitrogen and helium mixed gas (volume ratio of 4:1) for the gas chromatograph. The system includes a gas storage tank, a pressure regulator, and a flow controller. The gas storage tank is used to store high-purity carrier gas. The pressure regulator can stabilize the carrier gas pressure within a suitable range. The flow controller accurately controls the flow rate of the carrier gas according to the set carrier gas flow rate (1-3 ml / min).

[0109] Mass spectrometer

[0110] The mass spectrometer is connected to the gas chromatograph for further analysis of samples separated by gas chromatography. It has a dual ion source system, including an electrospray ion source and a chemical ion source. The electrospray ion source is suitable for analyzing compounds with high polarity, and can ionize sample molecules and form gaseous ions; the chemical ion source has a better ionization effect for some non-polar or weakly polar compounds. These two ion sources can be switched according to the properties of the sample and the analysis requirements, thereby expanding the mass spectrometer's ability to detect different types of impurities.

[0111] The mass spectrometer is equipped with a real-time calibration system, which performs mass calibration every 10 minutes by introducing an internal standard of known mass (such as perfluorobutane). During the calibration process, the internal standard enters the mass analyzer after ionization in the ion source, and the generated ion signal is compared with the theoretical mass number. The mass axis of the mass spectrometer is adjusted according to the comparison result to ensure that the calibration accuracy is within ±0.1amu. This real-time calibration mechanism can effectively overcome the mass measurement error caused by factors such as instrument drift during the long-term operation of the mass spectrometer, and ensure the accuracy of the mass spectrometry analysis results.

[0112] The mass spectrometer has a mass scanning range from 10 to 200 amu, which covers the mass number range of common impurities in electronic grade hexafluoroethane. At the same time, it uses a data-dependent secondary mass spectrometry scanning mode. During the scanning process, when an ion peak with a specific mass number is detected, the system automatically triggers secondary mass spectrometry analysis. Secondary mass spectrometry analysis can obtain more detailed ion fragmentation information, which helps to further determine the structure and composition of impurities.

[0113] The data acquisition module is responsible for collecting data from the gas chromatograph and mass spectrometer. It is connected to the gas chromatograph and mass spectrometer through special data acquisition interfaces, which can obtain the signals output by the instruments in real time. For the gas chromatograph, the collected data includes information such as peak area, retention time, peak height and peak width in the chromatogram; for the mass spectrometer, the data such as the mass number, intensity, isotope abundance of the ion peak are collected. The data acquisition module has high sampling frequency and high-precision data acquisition capabilities, which can ensure that the collected data fully and accurately reflects the sample information during the analysis process.

[0114] The data processing module is one of the core parts of the entire system. It processes the collected data based on machine learning algorithms and specific matching algorithms. In terms of machine learning algorithms, as mentioned earlier, models such as deep neural networks (DNN) or support vector machines (SVM) are used to train and learn a large amount of mass spectrometry data and chromatographic data. These models will continuously adjust their parameters during the training process to adapt to the analysis characteristics of different types of impurities and hexafluoroethane.

[0115] In terms of matching algorithms, for gas chromatography data, the chromatogram of the unknown sample is compared with the standard chromatogram pre-stored in the database. During the comparison process, not only the matching degree of parameters such as peak retention time and peak area is considered, but also the peak shape, peak symmetry and other factors are comprehensively evaluated. For mass spectrometry data, the collected ion peak information is matched with the standard mass spectrum of the known compound. The matching process considers the characteristics of multiple dimensions such as the mass number, relative intensity, isotope distribution and secondary mass spectrum fragment information of the ion peak.

[0116] At the same time, the data processing module also performs statistical analysis on the data of multiple measurements. After each measurement is completed, the data processing module will integrate the new data with the previous measurement data. By calculating statistical parameters such as the mean value and standard deviation, outliers are identified and removed. The judgment of outliers is based on statistical principles. For example, if the deviation of a data point from the mean value exceeds a certain multiple of the standard deviation (usually 3 times the standard deviation), it is judged as an outlier and removed. After such processing, the reliability and stability of the data can be improved, providing a more accurate data basis for subsequent purity calculations.

[0117] The purity calculation module calculates the purity of electronic-grade hexafluoroethane using a weighted calculation method based on the data processed by the data processing module. In the weighted calculation process, a specific weight factor (0.8-0.9) is assigned to the hexafluoroethane content, and different weight factors (ranging from 0.1-0.5) are determined for different impurity components according to their impact on the performance of electronic-grade products. For example, for some impurities that have a greater impact on the chip manufacturing process (such as metal impurities, etc.), their weight factors will be relatively high.

[0118] In addition, the purity calculation module also performs uncertainty assessment. During the assessment process, the impact of multiple factors on the purity results, such as measurement repeatability, instrument calibration, and sample uniformity, is considered. By establishing an uncertainty assessment model, the range of uncertainty is calculated and controlled within ±0.5%. This uncertainty assessment can more comprehensively reflect the reliability of the purity results and provide users with more valuable information.

[0119] The sampling container of the sampling and pretreatment unit is connected to the pretreatment device through a heating transmission pipeline, and the pipeline temperature is maintained at 30℃-50℃. This heating transmission pipeline uses materials with good thermal conductivity (such as metal-lined polytetrafluoroethylene), and a heating wire or heating plate is set inside the pipeline. The temperature control system monitors the pipeline temperature in real time through a temperature sensor and adjusts the heating power to stabilize the pipeline temperature within the set range. This can prevent the sample from condensing due to temperature changes during the transmission from the sampling container to the pretreatment device, and avoid the condensate from contaminating the sample or affecting the pretreatment effect.

[0120] The interface device between the gas chromatograph and the mass spectrometer of the analysis unit adopts a heating interface, and the temperature is controlled at 100℃-150℃. The interface device is equipped with a temperature sensor, a heating element and a heat preservation material to ensure the stability and uniformity of the temperature at the interface. At the same time, the interface device is also equipped with a pressure sensor and a flow control valve. The pressure sensor monitors the pressure changes at the interface in real time. When the pressure exceeds the set range, the flow control valve will automatically adjust the opening so that the sample maintains a stable pressure and flow during the transmission from the gas chromatograph to the mass spectrometer, achieving efficient transmission of the sample, pressure matching and flow control, thereby ensuring the accuracy of the analysis. In addition, the inner surface of the interface device is coated with an inert coating, which is usually made of a material with extremely high chemical stability (such as a silanized coating), which can prevent the sample from being adsorbed and chemically reacting at the interface, further ensuring the reliability of the analysis results.

[0121] The data processing module of the data processing and computing unit is connected to the remote server and realizes data transmission through the network communication protocol (such as TCP / IP protocol). The data processing module will upload the collected raw data and processed data to the remote server regularly. The remote server stores a large amount of user data and analysis results, which can be mined and analyzed through big data analysis technology. For example, the server can compare and count the analysis data of different users under different conditions to find potential patterns and problems in the data.

[0122] At the same time, the server can optimize the data processing algorithm based on the feedback from different users. When users find that the data processing results are biased or abnormal during use, they can feedback the relevant information to the server. Based on these feedbacks, the technical staff on the server can adjust the model parameters in the machine learning algorithm, the thresholds in the matching algorithm, etc., so as to improve the versatility and accuracy of the system and enable it to better adapt to the actual needs of different users and complex analysis scenarios.

[0123] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for determining the purity of electronic grade hexafluoroethane, characterized in that: The following steps are involved: S1. Sample collection and pretreatment S1.

1. Use a multi-layer composite sampling container with an inner layer of polytetrafluoroethylene, a middle layer of stainless steel mesh reinforcement structure, and an outer layer of weather-resistant polymer material to collect electronic grade hexafluoroethane samples in a sampling environment with a temperature controlled at 20°C to 25°C and a humidity controlled at 30% to 40%. The sampling container is equipped with an automatic pressure balancing device to maintain the difference between the internal pressure and the ambient pressure within the range of ±500Pa during the collection process. S1.

2. The collected samples are passed through a two-stage pretreatment device, the first stage of which is a drying tube filled with a mixed desiccant of activated alumina and silica gel in a mass ratio of 2:1, and the second stage is a polytetrafluoroethylene filter membrane with a pore size of 0.1-0.3 microns to remove moisture and particulate impurities in the samples; at the same time, a negative ion generator is set in the pretreatment device to generate negative ions with a concentration of 10 4 -10 5 Pieces / cm 3 , used to neutralize possible static electricity and reduce impurity adsorption; S2. Gas chromatography analysis S2.1, injecting the pretreated sample into a gas chromatograph, the gas chromatograph is equipped with a composite chromatographic column of a PLOT-U type porous layer open tubular column and an Al2O3 / KCl column connected in series, the column lengths are 30 meters and 15 meters respectively, the inner diameter is 0.32 mm, and the stationary phase is a high-purity Al2O3 and KCl coating; S2.2, set the injection port temperature of the gas chromatograph to 150°C to 180°C, the detector temperature to 250°C to 300°C, and the column temperature to a programmed temperature increase, with an initial temperature of -50°C, maintained for 5 minutes, then heated to 80°C at a rate of 10°C / min, maintained for 10 minutes, then heated to 120°C at a rate of 5°C / min, maintained for 8 minutes; S2.

3. A microfluidic chip is set in the injection system of the gas chromatograph to precisely control the injection volume, with the injection volume range of 0.1-1 μL. At the same time, an ion mobility spectrum enhancement module is set between the chromatographic column outlet and the detector to enhance the ion migration speed, improve the separation degree and detection sensitivity by applying an electric field of 500-1000 V / cm; the relative content of each component is determined based on the peak area or peak height of the main peak of hexafluoroethane and the impurity peak in the gas chromatogram; S3. Mass spectrometry analysis S3.1, introducing the pretreated sample into a high-resolution mass spectrometer, wherein the ion source of the mass spectrometer is a dual ion source system of an electrospray ion source and a chemical ion source, and the ionization energy is 10-50 eV; S3.2, the scanning mass range is from 10 to 200 amu, and the data-dependent secondary mass spectrometry scanning mode is adopted during the scanning process. When an ion peak with a specific mass number is detected, the secondary mass spectrometry analysis is automatically triggered to obtain more detailed structural information; S3.

3. The mass spectrometer is equipped with a real-time calibration system. By introducing internal standards of known mass, including perfluorobutane, mass calibration is performed every 10 minutes with a calibration accuracy within ±0.1amu. The impurity components in the sample are identified by analyzing the ion peaks in the mass spectrum, and the type and content of the impurities are further determined in combination with the gas chromatography analysis results. S4. Purity calculation According to the contents of hexafluoroethane and impurities determined by gas chromatography and mass spectrometry, the weighted calculation method is used: purity = (hexafluoroethane content × weight factor 1) / (hexafluoroethane content × weight factor 1 + ∑ (impurity content × corresponding impurity weight factor)) × 100%, where weight factor 1 is 0.8-0.9, and the impurity weight factor is determined according to the degree of influence of impurities on the performance of electronic-grade products, ranging from 0.1-0.5, and the purity of electronic-grade hexafluoroethane is calculated; at the same time, the uncertainty of the calculation results is evaluated, and the uncertainty range is controlled within ±0.5% considering measurement repeatability, instrument calibration, and sample uniformity factors.

2. The method for determining the purity of electronic grade hexafluoroethane according to claim 1, characterized in that: In S1, the sampling container is evacuated before sampling, and the vacuum degree reaches 10-3Pa-10-2Pa, and the inner wall of the sampling container is coated with a super hydrophobic coating, the coating material is a fluorine-containing polymer, and the water contact angle is greater than 150°.

3. The method for determining the purity of electronic grade hexafluoroethane according to claim 1, characterized in that: In S2, the carrier gas of the gas chromatograph is a mixed gas of high-purity nitrogen and helium with a volume ratio of 4:1, a purity of 99.999%, and a carrier gas flow rate of 1-3 ml / min; a gas circulation system is provided in the constant temperature box of the chromatographic column, and the carrier gas forms a uniform temperature and flow rate distribution in the column.

4. The method for determining the purity of electronic grade hexafluoroethane according to claim 1, characterized in that: In S3, the mass spectrometer is equipped with a GC-MS data processing system, which automatically identifies and matches the ion peaks in the mass spectrum with the standard mass spectrum of known compounds, and the matching threshold is set at 90%-95%; at the same time, the data processing system uses a machine learning algorithm to learn and train a large amount of mass spectrometry data to improve the accuracy and efficiency of impurity identification.

5. A system for implementing the method for determining the purity of electronic grade hexafluoroethane according to any one of claims 1 to 4, characterized in that: include: Sampling and preprocessing unit The sampling container is made of a multi-layer composite structural material, and is equipped with a temperature, humidity and pressure control device and a negative ion generator. It is used to collect samples under a specific environment and is evacuated by a vacuum pumping device before sampling. The vacuum pumping device can evacuate the vacuum degree to 10-3Pa-10-2Pa. The inner wall of the sampling container is coated with a super-hydrophobic coating. The pretreatment device is connected to the sampling container and is equipped with two-stage pretreatment components, including a mixed desiccant of activated alumina and silica gel and a polytetrafluoroethylene filter membrane, which are used to remove moisture and particulate impurities in the sample; Analysis Unit Gas chromatograph, equipped with composite chromatographic column, injection port, detector, temperature control system, microfluidic chip and ion mobility spectrometry enhancement module. The carrier gas system provides a mixed gas of high-purity nitrogen and helium with a volume ratio of 4:1 for gas chromatography analysis of pretreated samples; A mass spectrometer, having a dual ion source system, a real-time calibration system, a specific mass scanning range and a data-dependent secondary mass spectrometry scanning mode, connected to a gas chromatograph for mass spectrometry analysis of the sample; Data processing and computing unit A data acquisition module, used to collect data from a gas chromatograph and a mass spectrometer; The data processing module processes the collected data based on machine learning algorithms and matching algorithms to determine the impurity composition and content; it also performs statistical analysis on multiple measurement data, removes outliers, and calculates the average and standard deviation; The purity calculation module calculates the purity of electronic grade hexafluoroethane using a weighted calculation method based on the processed data and performs uncertainty assessment.

6. The electronic grade hexafluoroethane purity determination system according to claim 5, characterized in that: The sampling container of the sampling and pretreatment unit is connected to the pretreatment device through a heated transmission pipeline, and the pipeline temperature is maintained at 30° C.-50° C. to prevent condensation of the sample due to temperature changes during transmission.

7. The electronic grade hexafluoroethane purity determination system according to claim 5, characterized in that: The interface device between the gas chromatograph and the mass spectrometer of the analysis unit adopts a heating interface with a temperature controlled at 100°C-150°C, and is provided with a pressure sensor and a flow regulating valve to achieve efficient transmission, pressure matching and flow control of samples to ensure the accuracy of the analysis; the inner surface of the interface device is coated with an inert coating to prevent sample adsorption and reaction.

8. The electronic grade hexafluoroethane purity determination system according to claim 5, characterized in that: The data processing module of the data processing and computing unit is connected to a remote server and can upload data to the server for big data analysis and model updating; the server optimizes the data processing algorithm based on usage feedback from different users.

Images