TA Instruments TGA 2950 Operator's Manual page 204

Appendix B
B–66
The scan in Figure B.17 is a definite improve-
ment over the result in Figure B.16. This is
because the abort limit for the isothermal
segment (0.005%/minute) was chosen to be
equal to the baseline %/minute immediately
preceding the sodium bicarbonate transition
observed in a conventional constant heating rate
°
scan of the mixture at 1 C/minute. The theory
supporting this decision is that if the two transi-
tions are separable then the rate of weight loss
should return to baseline between the transi-
tions. The %/minute limit for the ramp segment
was then chosen to be ten times larger than that
for the isothermal segment.
Although the weight loss result in Figure B.17
seems more reasonable, we are suspicious that
the decomposition of the potassium bicarbonate
(2nd transition) has already started because the
rate of weight loss immediately increases as
heating is resumed at 88 ° C. Another problem is
the torturous 1300 minute time frame of the
experiment.
As can be seen by these results, caution must
always be employed when interpreting results
from stepwise isothermal heating experiments.
It is usually wise to run confirming experiments
by other TGA techniques, particularly if the
sample material is of relatively unknown compo-
sition.
TA I TGA 2950
NSTRUMENTS