Abstract: Thermogravimetric analyzer measures the change of the mass of the sample at the program temperature, and is an important instrument for analyzing the physical and chemical properties of materials. The quality of the instrument and the accuracy of the temperature measurement have a significant impact on the equivalence and consistency of the measurement results. According to the verification procedures of JJG thermogravimetric analyzers, two typical thermogravimetric analyzers were verified by using weight, melting point and Curie point reference materials. The verification methods, data processing methods, result judgments and uncertainty evaluations of all technical requirements, such as zero point drift, baseline drift, mass indication error and repeatability, temperature indication error and repeatability, are described through examples. In addition, the data processing methods and uncertainty evaluation of metrology calibration are briefly described. The results show that the metrogravimetric performance of the thermogravimetric analyzer can be scientifically and reasonably evaluated through metrological verification and calibration, and the traceability of the mass and temperature values can be realized, which can ensure the accuracy, reliability and equivalent consistency of the thermogravimetric analysis results. Keywords: thermogravimetric analyzer;Certification;Calibration;Melting Point;Curie Point;Uncertainty CLC Number: o Literature Symbol Code: A Article Number: Thermogr**imetric analyzer (TGA) is an important instrument for analyzing the physical and chemical properties of materials used to measure the change of the mass of samples at the program temperature. The two main parameters of a thermogravimetric analyzer are temperature t and mass mThe thermogravimetric analyzer can determine the reaction temperature [ composition [ water [ ] ash [ reaction kinetics [ ] vapor pressure [ ] of the material. Thermogravimetric analyzers have a wide range of applications in many fields such as energy products, building materials, nanomaterials, polymers, etc. Correspondingly, the International Organization for Standardization (ISO), the American Society for Testing and Materials (ASTM), and the Standardization Administration of China (GB) have introduced a number of standard methods to standardize measurement methods and ensure the equivalence of measurement results [ The quality of thermogravimetric analyzers and the accuracy of temperature measurements are extremely important to ensure the equivalence of thermogravimetric analysis results. The Standard Test Method for Mass Scale Calibration of ASTM E Thermogravimetric Analyzers and the Standard Test Method for Temperature Calibration of ASTM E Thermogravimetric Analyzers specify the mass calibration method and the temperature calibration method for thermogravimetric analyzers, respectively. However, in the absence of a Curie point reference material with accurate and reliable values, the temperature deviation between different instruments after calibration is still large and may reach In addition, due to the lack of metrological verification or calibration specifications, the accuracy of the temperature and quality of thermogravimetric analysis results cannot be quantitatively assessed. Wang et al. developed a variety of Curie point standard substances such as Aleumar alloy, nickel and iron, and used a simultaneous thermal analyzer to determine the standard material in conjunction with many laboratories, and realized the traceability of the temperature value through the melting point standard material calibration synchronous thermal analyzer, which meets the value specification of the reference material and can be used to verify or calibrate the temperature of the thermogravimetric analyzer. JJG Thermogravimetric Analyzer Verification Procedures [Published], which specifies the measurement and verification items, technical indicators and verification methods of the instrument, which can guide the daily verification work and evaluate the accuracy of the instrument. The procedure is concise and lacks verification examples, which is inconvenient for users to understand and operate in depth. In addition, the user can refer to the program to calibrate the instrument, but there is no reliable technical basis for how to evaluate the uncertainty of calibration quality and temperature. To this end, the author carried out metrological verification and calibration on two typical thermogravimetric analyzers based on JJG using weights, melting point and Curie point reference materials, and evaluated the measurement uncertainty of mass and temperature, so as to provide technical reference for the accuracy of the verification and calibration instruments of metrology technical institutions and the evaluation results of general users. 
Test Instruments and Test Materials The two types of thermogravimetric analyzers used in the test are as follows: Q type thermogravimetric analyzer, from the American TA company;Type F thermogravimetric analyzer from NETZSCH, Germany. Test material ( ) Grade F stainless steel weights with a nominal value of mg. The weights have passed the metrological verification, all of which are less than the maximum allowable error (mg), and the correction value is given, and the extended uncertainty u of the correction value is about the maximum allowable error, that is, mg (including factor k) ferromagnetic materials Curie point national certified reference materials, including Aleume alloy (GBW nickel (GBW and iron (GBW Curie point are respectively Extended uncertainty u respectively K) metal melting point and melting enthalpy reference materials (formerly known as thermal analysis reference materials), including indium [GBW(E) tin [GBW(E) Lead [GBW (E) Zinc [GBW (E) melting point is the extended uncertainty u respectively K) There are many technical indicators that affect the accuracy of the measurement results of the thermogravimetric analyzer in the test results and discussions: zero drift, baseline drift, mass repetition and indication error, balance response speed, etcTemperature-related factors include temperature rate indication error, temperature repeatability and indication error, thermometer response speed, etc. JJG listed the main factors affecting the accuracy of the measurement results as metrological verification items, and gave the technical requirements, as detailed in Table [
Mass zero-drift thermogravimetric analyzers often measure changes in sample mass at a constant temperature to calculate the component content of the sample. Therefore, the stability of the mass measurement results at a constant temperature affects the accuracy of the above quantities. JJG stipulates that the use of an empty crucible to keep the temperature in min, and the difference between the maximum and minimum of the mass measurement result is the mass zero drift value [ The zero drift is affected by factors such as the degree of instrument preheating, seismic isolation conditions, and the stability of carrier gas flow. The smaller the zero drift, the better the instrument performance. JJG stipulates that the mass zero drift value should not exceed mg [ The test results show that the mass zero drift values of q and f thermogravimetric analyzers are mg and mg respectively, which meet the technical requirements of the metrological verification regulations. In order to reduce the zero drift value, it is necessary to use the instrument after it is fully warmed up, place the instrument on the seismic isolation table, and avoid touching the instrument or the seismic isolation table during the test to reduce the vibration interference in the laboratory. Mass baseline drift thermogravimetric analyzers often measure changes in sample mass at a certain heating rate to determine the content of certain components. Therefore, the stability of the mass measurement results under elevated conditions affects the accuracy of the relevant values. JJG stipulates that the difference between the maximum value and the minimum value of the mass measurement result in the temperature increase of the air crucible and the rate of min is the mass baseline drift value [ The baseline drift is affected by factors such as buoyancy effect, gas convection, and thermal expansion of the crucible support. Gas convection refers to the upward force of the temperature and the rising air flow on the sample holder, also known as the chimney effect. Theoretically, thermogravimetric analyzers with a down-hanging structure, such as TA's Q thermogravimetric analyzer and Perkin's Elmer's Pyris thermogravimetric analyzer, have crucibles suspended with slings, and the slings are small in size, with relatively small buoyancy effects and gas convection, so the baseline drift is small. Top-dish thermogravimetric analyzers, such as the NETZSCH F thermogravimetric analyzer, have a large sample holder and are characterized by relatively large buoyancy effects and gas convection, resulting in a significant baseline drift. It is usually necessary to obtain a baseline by measuring the sample with an empty crucible at the same temperature program prior to sample measurement, and the thermogravity of the sample is deducted from the base line to eliminate the effect of base line drift. JJG stipulates that the drift of the mass base line shall not exceed MGThe test results show that the mass baseline drift values of q and f thermogravimetric analyzers are mg and mg, respectively, which meet the technical requirements of the metrological verification procedures. The use of baseline exclusion can significantly reduce the impact of baseline drift and help to improve the accuracy of component content measurement results. The accuracy and repeatability of mass measurement results are related to the accuracy of component content determination. Therefore, thermogravimetric analyzers need to be calibrated on a regular basis. Usually the instrument manufacturer provides a standard of known quality, such as a metal ball, which is used to calibrate the quality of the thermogravimetric analyzer. This single-point calibration method usually uses a large mass standard to calibrate the instrument, but when the mass of the measured sample is small, the accuracy of the mass measurement results is not guaranteed. In addition, the manufacturer is not equipped with qualified weights, and there is a lack of accurate and traceable quality values. JJG specifies the use of F-rated stainless steel weights (incl. , mg) to characterize the quality of thermogravimetric analyzers [ This weight covers the common measurement range of thermogravimetric analyzers. The accuracy and repeatability of the mass measurements, the verification and calibration results are shown in the table
As shown in the table, the Q thermogravimetric analyzer measured the mg weights twice, and the results were mg and mg, respectively, with an average value of mgThe difference between the two measurements (mg) is used as the reproducibility value, and the reproducibility limit (mg) specified by the procedure is not exceededThe difference between the average value (mg) of the two measurements and the corrected weight mass (mg) (mg) is used as the indication error, which does not exceed the maximum allowable error (mg) specified in the procedureThe repeatability of the measurement results and the indication error of the weights of the Q and F thermogravimetric analyzers do not exceed the specified limits. According to the JJF measurement uncertainty assessment and representation [ Instruments that have passed the metrological verification can estimate the uncertainty of the measurement results according to their maximum allowable error. For example, when measuring a mg weight, the maximum allowable error of mass is mg, the standard uncertainty u(m) for estimating the mass is the maximum allowable error, i.e., mg, and the extended uncertainty u(m) is mg(k).Similarly, when measuring mg and mg samples, u(m) is mg and mg(k) respectively, and if the user needs to calibrate the thermogravimetric analyzer, the correction can be calculated based on the indication error of the mass. For example, if a weight with a mass of mg (corrected value) is measured with a value error of mg, the corrected value for the mass is mgThe uncertainty of the calibrated mass measurement results includes Class A uncertainty, Ua and Class B uncertainty, UBUA comes from the repeatability of the measurement, taking the sample mass as mg as an example, the range of the sub-measurement results is mg, and the standard deviation is calculated using the range method, and the range coefficient is n) The standard deviation is mg, i.e., UA is equal to mgUB comes from the uncertainty of the weight correction value, and if the extended uncertainty of the weight correction value is mg, then ub is equal to mgAfter variance synthesis, the synthetic uncertainty u(m) is mg, and the extended uncertainty u(m) is mgSimilarly, the extended uncertainty of the calibrated mass measurement results (the average of the measurements) is mg and mg(k) when the sample mass is in mg and mg, and the user can refer to the above results and evaluate the uncertainty of the measurement results according to the reproducibility of the measured results of the sample quality. After calibration, the uncertainty of the thermogravimetric analyzer mass measurement results is reduced. Accuracy of ramp rateThermogravimetric analyzers often measure changes in sample quality at a certain ramp rate, so the accuracy of the ramp rate affects the accuracy and reproducibility of the measurement results. JJG stipulates that the temperature of the sample after min is recorded at the rate of min, from the time of heating to the time when the sample temperature is the starting point, [ The indication error of the heating rate is calculated according to the following formula: δV is the error of the performance of the program heating rate, ;t is the temperature of the sample when min, ;t is the temperature of the sample at the beginning of the metert is the timing interval, min;v is the rate of heating, minAfter testing, the indication error of the heating rate of the Q and F thermogravimetric analyzers is and is not more than the limit specified in the procedure ( The accuracy of temperature measurement and the accuracy of repeatable temperature measurement directly affect the accuracy of the measurement results of the thermogravimetric analyzer, so the scientific and reasonable verification of temperature is the core content of the procedure (JJG) The thermogravimetric analyzer currently on sale can be divided into two types according to the temperature calibration method: one is the instrument represented by the F type of NETZSCH, except for the bottom of the sample crucible In addition to the thermocouples, there is also a thermocouple near the sample crucible, and the differential heat signal (called CDTA by NETZSCH) is calculated based on the measurements of the two thermocouples during the heating processThe other is represented by PerkinElmer's Pyris type and TA's Q type, which have no differential thermal signature. The thermogravimetric analyzer with differential heat signal, when calibrating the temperature, the pure metal is taken as a sample, heated until the metal melts, and the melting point of the metal is calculated according to the differential heat signal, that is, the tangent is made from the inflection point of the rising edge of the temperature difference curve, and the extension line of the temperature difference curve before melting intersects, and the horizontal coordinate of the intersection point is the extrapolated starting melting point, referred to as the melting point tm, in units, as shown in the figure. The average value of the measurement results, minus the standard value of the melting point, is the indication error of the melting pointThe difference between the two results is the repeatability of the melting point. 
The temperature of the F-type thermogravimetric analyzer was verified according to the above method. As shown in the table, the melting point measurements of tin are measured at the melting point and the average value is less than the maximum allowable error specified in the procedure compared to the standard value ( The difference between the two measurements is less than the reproducibility limit specified in the procedure ( Similarly, the melting point error and repeatability of indium are also in accordance with the requirements of the specification. 
For thermogravimetric analyzers without a differential thermal signature, the temperature is usually calibrated using a Curie point reference material of ferromagnetic material. A magnetic field of a certain strength is applied to a sample of ferromagnetic material, and the ferromagnetic material generates magnetism and forms an interaction force with the magnetic field. When the sample is heated to the Curie point at a certain heating rate, the ferromagnetic material becomes paramagnetic, that is, the magnetism disappears, resulting in a significant change in the apparent mass of the sample, as shown in the figure. Extrapolation terminating the Curie point tc,endset. is calculated according to the method shown in Fig
EferenceMaterialMeasuredbyq ThermograrimetricAnalyzer Wang et al. reported that the Curie point is affected by a variety of factors such as the rate of warming and the strength of the magnetic field. Therefore, JJG stipulates that the sample amount of the Curie point reference material is mg, and the position of the magnet is adjusted to change the apparent quality of the sample to about min [This condition is consistent with the Curie point reference material value condition, which is conducive to eliminating the systematic error caused by factors such as temperature lag [As shown in the table, the Curie point measurement results for nickel are respectively and the average value is less than the maximum allowable error specified in the protocol ( The difference between the two measurement results is less than the repetition specified in the procedure Limits ( Similarly, the Curie point indication error and repeatability of iron are in accordance with the requirements of the regulations. Instruments that have passed the metrological verification can estimate the uncertainty of the measurement results according to their maximum allowable error. For example, when the melting point of tin is measured by a thermogravimetric analyzer, the maximum allowable error is uniformly distributed, and the standard uncertainty is the maximum allowable error, i.e., in the vicinity, the expansion uncertainty u(t) of the temperature measurement of the instrument is k) In the same way, when the nickel reference material is measured by the q thermogravimetric analyzer, the maximum allowable error of the Curie point is so in the vicinity, the extended uncertainty u(t) of the temperature measurement of the instrument is k) It can be seen that the uncertainty of the temperature measurement is smaller for the instrument that adopts the pure metal melting point reference material verification. The main reason is that the uncertainty of the melting point identification value of the melting point reference material is small, while the uncertainty of the Curie point identification value of the Curie point reference material is large. To sum up, the two thermogravimetric analyzers F and Q meet all the measurement technical requirements specified in the regulations and are qualified. If the user needs a metrological calibration, a correction can be calculated based on the indication error of the temperature. For example, if the indication error when measuring the melting point of tin is , then the correction value of the temperature is the calibration of the temperature measurement results The uncertainty of the temperature measurement results mainly includes the repeatability of the temperature measurement and the uncertainty of the melting point of the reference material. The melting point uncertainty of indium and tin u(t) is and k, respectively) is estimated by the reproducibility level of the above measurements of the thermogravimetric analyzer, and the temperature is at the same time, and the expansion of the temperature measurement results (the average of the measured results) after calibration is uncertain The temperature u(t) and k) are respectively and k) Similarly, with a thermogravimetric analyzer calibrated with the Curie point, the temperature is at and , u(t) and k) The user can refer to the above results and evaluate the repeatability of the measured results of the sample temperature The uncertainty of the measurement results. As can be seen from the metrological verification and calibration process described above, unlike the user's daily calibration, metrological verification not only examines the accuracy of temperature and mass, but also examines other items that affect the accuracy of the measurement results, such as zero drift, baseline drift and heating rate, and can more comprehensively evaluate the metrological performance of the instrument. Metrological verification and metrological calibration have something in common, both are standards (weights and reference materials) with known measured values with the instrument being tested;The difference is that the metrological verification gives the error of the presentation value, and the measurement calibration gives the correction value;Metrological verification determines whether the instrument is qualified or not, while metrological calibration does not determine whether it is qualified or not. Synchronous thermal analyzers, which have been used since the beginning of the century, function as both differential scanning calorimeters and thermogravimetric analysers, which simultaneously measure the mass and heat flow signal of a sample at the programmed temperature. The differential scanning calorimetry function of this instrument can be carried out for metrological verification and calibration with reference to the JJG differential scanning calorimeter verification procedureThe thermogravimetric analysis function mainly refers to the zero point drift, baseline drift, and the indication error and repeatability of the mass, which can be carried out with reference to JJG for metrological verification and calibration. References: [ General Administration of Quality Supervision, Inspection and Quarantine of the People's Republic of China Determination of ignition temperature of pyrotechnic agents Differential heat Thermogravimetric analysis method: GB t S] Beijing: China Standards Press, National Nanotechnology Standardization Technical Committee Thermal gravity analysis of oxidation temperature and ash of carbon nanotubes: GB t S] Beijing: China Standards Press, plastics—thermogr**imetry (tg)of polymers— part :d etermination of Activation Energy:ISO S] [Standard Test Methods for Loss on Drying by Thermogr**Imetry:ASTM E S] [General Test Methods Branch of the National Technical Committee for Standardization of Rubber and Rubber Products Determination of Composition of Vulcanized Rubber and Unvulcanized Rubber by Thermogravimetric Analysis: GB T S] Beijing: China Standards Press, Zhang Bin, Liu Jianzhong, Zhao Weidong, et al Thermogravimetric experiment and kinetic analysis of lignite spontaneous combustion characteristics[J] thermal power generation, LI Song, Bian Nan Analysis of the failure causes of chip multilayer ceramic capacitors[J] Physical and chemical test (physical fascicle), Wang Ling Application of thermogravimetric analysis technology in the detection of magnesium oxide moisture content [J] Physical and chemical test (physical fascicle), National Chemical Standardization Technical Committee Determination of saturated vapor pressure of chemical products Thermogravimetric method: GB t S] Beijing: China Standards Press, standardtest methodfor massscalecalibrationof thermogr**imetric analyzers:astm e s] [standardtestmethodfortemperaturecalibrationof thermogr**imetricanalyzers:astm e s] [wangt,wang hf,wangf,etal Certification of Reference Materialso***Mel, Nickelandironfor Curie Point[J] Journal of Thermal Analysis and Calorimetry, National Technical Committee for New Materials and Nano Metrology Thermogravimetric Analyzer Verification Regulations: JJG S] Beijing: China Quality Inspection Press, Xu Chaofen, Sun Xuexin, Guo Xin Analysis of the main factors affecting the thermogravimetric curve in thermogravimetric analysis test[J] thermal power generation, Yu Na, Fu Peifang, Zhou Huaichun Test and analysis of TGA basis point drift[J] thermal power engineering, WANG Guan, WU Kang, YUAN Zehong Use and calibration of STA F-type synchronous thermal analyzer[J] Analytical instruments, National Legal Metrology Management Metrology Technical Committee Measurement uncertainty evaluation and representation: JJF S] Beijing: China Standards Publishing House, National Physical and Chemical Metrology Technical Committee Differential scanning calorimeter inspection Procedure: JJG S] Beijing: China Quality Inspection Press, Article**—Materials and Testing Network.