Page 20 - Handbook of Thermal Analysis of Construction Materials

Section 2.0 - Classical Techniques 3

a larger crucible, surrounded with magnesium oxide and inserted into an
-1
oven. Le Châtelier used a heating rate of 120 K min and recorded the
electromotive force of the thermocouple on a photographic plate at regular
time intervals. As long as no phase change occurred in the clay, the
temperature rose evenly and the lines on the plate were evenly spaced. If,
however, an exothermic transformation took place, then the temperature
rose more rapidly, and, therefore, the lines were unevenly spaced and closer
together. An endothermic transition, on the other hand, caused the mea-
sured temperature to rise more slowly, and the spacing between the lines
was much larger. To ensure that the measured temperatures were correct,
he calibrated his instrument with the aid of boiling points of known
materials such as water, sulfur, and selenium, as well as the melting point
of gold. Since Le Châtelier’s experiment does not fit the ICTA definition
of DTA, his main contribution to the development of DTA was the
automatic recording of the heating curve on a photographic plate. True
differential thermal analysis was actually developed twelve years later (in
1899) by Roberts-Austen. [22]
Roberts-Austen connected two Pt-Pt/10%-Ir thermocouples in
parallel which, in turn, were connected to a galvanometer. One thermo-
couple was inserted into a reference sample consisting of a Cu-Al alloy or
of an aluminum silicate clay (fireclay). The other thermocouple was
embedded into a steel sample of the same shape and dimensions as the
reference. Both the sample and reference were placed in an evacuated
furnace. A second galvanometer monitored the temperature of the refer-
ence. The purpose of the experiments was to construct a phase diagram of
carbon steels and, by extension, railway lines. Since his method was a true
differential technique, it was much more sensitive than Le Châtelier’s. The
DTA design used today is only a slight modification of Roberts-Austen’s,
and the only major improvements are in the electronics of temperature
control and in the data processing, which is now handled by computers (see
Fig. 1).
It took about fifty years for the DTA technique to be considered not
only qualitative, but also as a quantitative means of analyzing and charac-
terizing materials. Moreover, it was only then that the Roberts-Austen setup
was modified by Boersma. [23] The modification was in the placement of the
thermocouples. Rather than placing the thermocouples into either the
sample or the reference, Boersma suggested that they be fused onto cups
and that sample and reference be placed into these cups. This modification
eliminated the necessity of diluting the sample with reference materials and
reduced the importance of sample size. The vast majority of today’s DTA