Thermobalance Calibration

Heat effects as such are rather unspecific, especially if the nature of the reactions taking place in the calorimeter is not known. The combination of thermal analysis with chemical analysis (for example with a thermobalance (DTG) or a mass spectrometer which register the rate of advancement of the reaction) enables us to arrive at specific caloric data (i.e., AH/A nt). Absolute values of AH are obtained by calibration only. 

Accurate temperature calibration using the ASTM temperature standards [131, 132] is common practice for DSC and DTA. Calibration of thermobalances is more cumbersome. The key to proper use of TGA is to recognise that the decomposition temperatures measured are procedural and dependent on both sample and instrument related parameters [30]. Considerable experimental control must be exercised at all stages of the technique to ensure adequate reproducibility on a comparative basis. For (intralaboratory) standardisation purposes it is absolutely required to respect and report a number of measurement variables. ICTA recommendations should be followed [133-135] and should accompany the TG record. During the course of experiments the optimum conditions should be standardised and maintained within a given series of samples. Affolter and coworkers [136] have described interlaboratory tests on thermal analysis of polymers. 

Calibration of the TPC runs was performed on a Stanton Redcroft TG750 thermobalance. A 12.5 mg sanple of Co/Ti02 was reduced in flowing H2 at 400°C and then carbided in pure CO at a heating rate of 5°C/min. 

A DSC instrument is very similar to a standard thermobalance with an extra pan for a reference sample (Figure 6.4). For quantitative measurements, the instrument has to be calibrated using standard reference materials. 

Periodic calibration of the thermobalance will prevent errors on the mass axis of the recorder. Many investigators calibrate the instrument before each run by adding a known weight to the sample container. 

The calibration of the temperature of the furnace and/or sample chamber has been discussed by Stewart (12) and Norem et al. (14, 15). Stewart (12) used a conventional thermobalance which contained a thermocouple mounted external to the sample, while Norem et al. (14, 15) calibrated a furnace which used a resistance element for lemperature detection. 

Norem et al. (14, 15) used the third method, as previously discussed, to calibrate the temperature of their type of furnace and/or sample container, A ferromagnetic material was placed in the sample container and suspended within a magnetic field. At the material s Curie point temperature, its equivalent magnetic mass diminishes to zero and the thermobalance indicates an apparent mass-loss. For calibration over the temperature range from ambient temperature to lOOOC, it is obvious that a number of ferromagnetic materials must be used. The criteria which were considered characteristic of an ideal standard were the following (15) ... 

Accurate temperature calibration was necessary for the example above, and this leads on to another advantage of TG-DTA instruments the temperature calibration can be properly carried out as the sample temperature is measured directly, which is not so in conventional thermobalances in which the temperature sensor is not in contact with the sample. A TG-DTA instrument was used to provide accurate transition temperatures for the ICTAC Curie point standards for TG calibration. 

The same MS instrument was also linked to a TA Instruments model 951 thermobalance, which is ideal for EGA work, but is sadly no longer available. This combination was used to investigate the problem of calibration of a TG-MS for quantitative EGA. A schematic diagram of the layout is shown in Figure 10. 

The weight calibration of thermobalances is done using standard weights. The temperature calibration is more difficult. The method using the Curie point temperature, as... 

The fact that volatile substances will vaporise at temperatures far below their atmospheric boiling point, due to the gas flow through the oven of the thermobalance, can be used to perform a simulated distillation [3-1 to 3-4], In contrast to the simulated distillation by gas chromatography GC [3-5, 3-6] no pailition or rectification effect can be assumed during evaporation in a thermobalance. Therefore calibration can only take place using individual substances of known boiling points. It seems reasonable to use the substances which are named for calibration in the standards of simulated distillation by GC [3-5, 3-6]. 

The correlation coefficient for 1 % weight loss is r = 0.986, for 5 % r = 0.994, and for weight losses exceeding 10 % the correlation coefficient increases to / > 0.997. The coefficients of equation 3-3 are different for each weight loss and decrease at increasing weight losses (Fig. 3-18 and 3-19). The ascertained functions are always of the same kind regardless to the type of thermobalance or the series of calibration substances, but the coefficients are, naturally, different. Nevertheless there is a linear relation between the data... 

Nevertheless, a "true value" is obtainable by an independent calibration. This work was undertaken to provide magnetic transition temperatxire values that can be used as calibration points for thermobalances—in particular those in which the sample temperature is not measured directly in the sample or the sample holder. 

Thermogravimetry needs a check of the accuracy of temperature, mass, and time measurements. Practically all thermobalances are capable of producing good data with only infrequent checks of the calibration via a standard mass. Since changes in volume of the sample take place, a buoyancy correction should be done routinely. The mass, m, of the displaced gas can easily be calculated from the ideal gas law (m = pMAV/RT). 

The transition used to calibrate the temperature scale of a thermobalance should have the following properties [1] (i) the width of the transition should be as narrow as possible and have a small energy of transformation (ii) the transition should be reversible so that the same reference sample can be used several times to check and optimize the calibration (iii) the temperature of the transition should be independent of the atmospheric composition and pressure, and unaffected by the presence of other standard materials so that a multi-point calibration can be achieved in a single run and (iv) the transition should be readily observable using standard reference materials in the milligram mass range. Transitions or decompositions which involve the loss of volatile products are usually irreversible and controlled by kinetic factors, and are unsuitable for temperature calibration. Dehydration reactions are also unsuitable because the transition width is strongly influenced by the atmospheric conditions. 

When thermobalances are sold as simultaneous TG-DTA (TG-DSC) apparatuses, temperature calibration is most conveniently carried out using the techniques for DSC calibration described in Chapter 3. Temperature... 

Figure 15.9 Arrangement of thermobalances showing the furnace (shaded), sample position and casing (a) and (c) suspended (b) top-loading (d) horizontal. The magnet position for Curie point calibration is shown in (c).

X-Pert MPD system with Cu Ka radiation. X-ray photoelectron spectroscopy (XPS) measnrements were carried out in an ultra-high vacuum set-up equipped with a Gammadata-Scienta SES 2002 analyzer. A flood gun was used to compensate for the charging effects. The binding energies were calibrated with the C Is peak (284.5 eV). Thermogravimetry was performed with a Calm TG-2131 thermobalance in pure O2 with a heating rate of 2 K/min. 

Calibration of thermobalances by materials showing Curie points at known temperatures is currently being considered by the Standardization Committee of the International Confederation for Thermal Analysis-see McAdie [1972]. 

---