Thermal Analysis UserCom 3

Introduction

Thermogravimetric analysis determines the mass change of a sample subjected to a temperature program and a defined atmosphere. The first derivative of the TG curve, called DTG, is used for the interpretation of the reactions of the sample. In the analysis of the main components of elastomers, the classical extraction processes and also the qualitative and quantitative analysis of the elastomer components with IR spectroscopy or gas chromatography have been virtually completely supplanted by the more elegant thermogravimetric rubber analysis. The main components usually determined are:

1. Volatile components, which are driven off between room temperature and approx. 300 °C. They chiefly comprise added oils and other plasticizers, as well as moisture, solvent residues, monomers and, e.g. stearic acid.

2. Content of elastomers, such as natural rubber and EPDM. Under a nitrogen atmosphere and the usual heating rate of 30 K/min, pyrolysis follows between 300 and 550 °C, depending on the chemical structure of the elastomer molecule.

3. Content of carbon black by burning in air or oxygen (automatic gas switching!).

4. Residue: Inorganic fillers (plus ash). Any CaCO₃ loses CO₂ at approx. 800 ° C. The stoichiometric CaCO₃ content follows from the weight loss.

[…]

Introduction

ADSC allows simultaneous determination of the heat capacity and thermal events in the sample. Here, a periodic signal, the so-called modulation, is superposed on the conventional, generally linear temperature program:

T_p = T₀ + b·t + A·sin(t·2· π/p) where

T_p is the program temperature,
T₀ the start temperature,
b the heating rate,
A the modulation amplitude,
t the time and p the modulation period.

In the evaluation, the differential heat signal is split into a c_p component and a heat of reaction component (thermal event). The question now arises regarding the "correct" choice of the experimental parameters: sample size, mean heating rate, modulation amplitude and modulation period.

As the period of the modulation determines the time resolution and through this together with the mean heating rate of the basic temperature program also the resolution in the temperature range, the aim is to have periods as short as possible. However, the period can not be shortened at will as both the DSC furnace and the sample have a finite thermal conductivity and hence periods which are too short are misrepresented.

Figure 1 outlines the falsification, which is manifested as damping (too small an amplitude) and as a phase shift (time shift compared with the excitation signal). The limited ability of the furnace to handle short cycle times is intuitively easy to understand when the relatively large mass and expansion of the furnace are considered. However, it is surprising that for substances with a thermal conductivity (0.1 to 1 W/(m K)) in the range of typical polymers, the minimal time is determined by the sample and not by the DSC furnace.
In a forthcoming publication [1], the autor describes the simulation of a greatly simplified DSC. In this simulation, the partial differential equation which describes the heat transport in the sample was solved by a finite element analysis. Figure 2 shows a scheme of the simulation. This work has shown that weights of typically not more than 5 mg and cycle times of 2 to 4 minutes are needed for accurate determinations of the heat capacities of polymers.

[…]

References

[1] B. Schenker, F. Stäger, “Influence of the Heat Conductivity on the c_p Determination by Dynamic Methods”, Thermochimica Acta, 1996, submitted for publication