Synchronous thermal analysis (STA, such as TG-DSC/DTA combination) can simultaneously obtain information on the mass change (TG) and thermal effect (DSC/DTA) of the sample, but the data analysis of complex systems (such as multi-component mixtures and reaction intermediates) needs to be combined with thermodynamic principles and experimental design optimization. The following elaborates on the analysis strategy from three aspects: data preprocessing, feature peak recognition, and multi curve correlation analysis.
1、 Data preprocessing: eliminating noise and baseline drift
Smooth filtering
The Savitzky Golay algorithm is used to smooth the TG-DSC curve, and the window width should be selected according to the data point density (usually 5-15 points) to avoid excessive smoothing and distortion of feature peaks. For example, during the thermal decomposition of polymer materials, small mass loss steps may be masked by noise and can be clearly identified after smoothing.
Baseline correction
TG baseline: Subtract the drift of blank crucibles (such as Al ₂ O ∝) through linear or polynomial fitting, with particular attention to baseline tilt caused by instrument thermal expansion in high-temperature regions (>500 ℃).
DSC baseline: Use tangent method or symmetrical baseline method to calibrate the baseline of phase transition peaks such as melting and crystallization, ensuring accurate enthalpy change calculation. For example, the solid-phase transition peak of metal alloys requires precise deduction of background heat flux.
2、 Feature peak recognition: Combining thermodynamic parameters to locate key events
Staged analysis of TG curve
Divide the quality change curve into stages such as dehydration, decomposition, and oxidation, and determine the peak temperature of each stage rate (Tp) through the first derivative (DTG). For example, the thermal decomposition of drug polymorphs may exhibit bimodal DTG curves, corresponding to the transitions of different crystal forms.
The correlation between DSC peak and TG event
Heat absorption peak: often corresponds to melting, sublimation, or decomposition reactions (such as carbonate decomposition heat absorption).
Heat release peak: may be due to oxidation, crystallization, or polymerization reactions (such as exothermic crosslinking of polymer materials).
Synchronous analysis: When the DSC exothermic peak overlaps with the TG mass drop, it indicates an oxidative combustion reaction (such as the combustion of organic matter to produce CO ₂ and H ₂ O).
3、 Multi curve correlation analysis: revealing reaction mechanism and kinetics
Quality heat flow synchronous verification
If TG shows quality loss but DSC has no thermal effect, it may be due to physical adsorption water removal; If both show significant changes, it indicates that a chemical reaction (such as thermal decomposition) has occurred. For example, the positive electrode material of lithium-ion batteries (such as LiCoO ₂) decomposes and releases oxygen at high temperatures, and the mass loss of TG occurs synchronously with the strong exothermic peak of DSC.
Calculation of dynamic parameters
Based on TG data, the activation energy (Ea) was calculated using Friedman or Ozawa methods, and the mechanism function was validated by the relationship between DSC peak temperature and heating rate (such as Avrami equation describing the crystallization process). For example, the calculation of Ea for polymer melt recrystallization can distinguish between first-order phase transition and diffusion controlled processes.
4、 Summary of Skills
Control experiment: Use standard samples (such as indium and zinc) to calibrate temperature and heat flux accuracy, and eliminate instrument errors.
Segmented integration: Calculate the enthalpy change (Δ H) by integrating the DSC peak area, deducting the baseline and correcting for the effect of heat capacity changes.
Software Assistance: Utilizing professional software such as TA Universal Analysis and NETZSCH Proteus for peak fitting and kinetic analysis to improve efficiency and accuracy.
Through system preprocessing, feature peak localization, and multi curve correlation, synchronous thermal analysis data can be deeply analyzed, providing key basis for material thermal stability assessment and reaction mechanism research.