Below is a thermogravimetric (TG) scan of commercial grade sucrose. Over the course of the experiment, the compound is quite stable to decomposition to the gas phase under N2 up to the onset temperature of ~226 C. This is indicated by the constant temperature line leading up to the onset. Above 226 C it begins to evolve gas or aerosols.
The furnace temperature ramp is 10 C/minute, the purge gas is nitrogen, and the crucible is platinum. The software that presents the data determines an onset and offset temperature by intersecting lines extending from the slopes of the lines at points chosen by the operator. In this way, the computer “squares” the broad curves and reports a temperature at that point.
Not shown is the % mass loss; from ambient to 500 C the sample lost 78.55 %. The slope of the curve from the onset to offset temperature is 0.69 %/deg C. Multiplying by 10 deg C/min, the cook off rate is 6.9 wt % per minute in the temperature ramp. An isothermal run could be performed to look at behaviour under constant temperature.
The value of information presented by this analysis is somewhat limited to the relevance of conditions in the crucible. TGA can be a big help in determining the onset and extent of dehydration or other decomposition transformations. It can easily detect the ability of a material to sublime as well. And, it can provide information to process chemists and engineers regarding certain aspects of temperature stability and thermal safety.
Lamentations on Chemistry 
So.. if we make the assumption that whats left at the end is very nearly all carbon, it looks like roughly half of the original carbon remains.
Does this percentage change substantially if the furnace ramp speed is changed – instead of 50 minutes, if you set the unit up to proceed across a timespan of 200, 100, or 25 minutes, is the amount of carbon present at the end different?
Also, can radioactive labels determine the likelyhood that any given carbon in the sucrose matrix is going to end up a part of the charcoal goop or a part of the vapors?
Hi Jason,
A TGA experiment is fairly simple minded- it’s just a platinum balance pan in a furnace. The temperature ramp can have an effect on the onset temperature and the slope of the mass loss- that’s why one typically chooses a favorite heating rate for an experimental series. The low sample mass has the effect of offering minimal thermal inertia.
Decomp is likely to occur substantially at the surface at least for some period of time. If decomposition renders the surface impermeable, then conceivably other processes within the bulk material may be favored owing to higher temperature within.
The question about the use of isotopes to monitor how the carbon is chemically bound sounds like a good experiment. It might be more useful to use 13C since it is NMR active and characteristic chemical shifts can be obtained.