Some Basics En Route to Isoconversional Methodology - Isoconversional Kinetics of Thermally Stimulated Processes - page 21

Fig. 1.15 TGA curves for an
empty pan (blank TG) and
555 μg sample of calcium
oxalate monohydrate placed
in this pan (measured TG).
Corrected TG is obtained
by subtracting blank TG
from measured TG. TGA
thermogravimetric analysis.
(Reproduced from Vyazovkin
et al. [ 42 ] with permission of
Elsevier)
furnace. The temperature lag is accounted for by temperature calibration. Tempera-
ture calibration of DSC is normally done by measuring the melting temperatures
of reference substances, e.g., metals. The same approach is used for temperature
calibration of TGA instruments capable of measuring the differential thermal analy-
sis (DTA) or DSC signal. Alternatively, one can heat ferromagnetic materials (e.g.,
alumel, nickel, cobalt, etc.) in a sample pan while placing a magnet under or above
the furnace [ 43 ]. The temperature of the mass jump that occurs during the ferro-
magnetic to paramagnetic transition is then identified with the Curie temperature of
the material. Another option is to use a dropping weight (platinum coil) connected
to the balance beam by a fusible link, which is a thin wire made of melting point
metal standard [ 44 ].
Temperature calibration is done in a certain temperature range and under certain
heat transfer conditions: heating rate, gas atmosphere, pan material, and sample
substance. These are the range and conditions for which the calibration holds most
accurately. Thus, it should be performed in the temperature range and under the
conditions that are as close as possible to those of the actual kinetic runs. In particu-
lar, this means that when nonisothermal runs are carried out at several heating rates,
the calibration should be conducted in the whole range of the heating rates used.
Using the same pans and gas atmosphere for both calibration and measurement are
easy conditions to satisfy. It is rarely possible to calibrate the instrument by using
substances whose thermal diffusivity is similar to that of the actual sample. Typical
calibration standards are metals whose thermal diffusivity is significantly larger
than that for most of inorganic and, especially, organic and polymeric samples. This
means that the actual sample may still experience some temperature lag unaccount-
ed by calibration.
Another important reason why the sample temperature deviates from the fur-
nace temperature is the sample self-heating/cooling due to the thermal effect of
the process. Due to its limited thermal diffusivity, the sample cannot exchange the
process heat with the surroundings instantaneously. As a result, a temperature gra-
dient arises within the sample. This temperature deviation cannot be accounted for