In order to be able to perform an error analysis of an experimental procedure, I need to know how accurately the water content of a rock, in particular of salt rocks, can be measured by thermogravimetric means (relative or absolute error). If it is of any importance, the material contains around 0.5 mass-% water.
Information on the accuracy of alternative methods is also appreciated.
I should perhaps add that I mean the water that is present in liquid form, e.g. as a salt solution, not the water bound in the crystalline structure.
As an option to TGA, you can use Karl Fischer coulometer with in line drying oven to accurately measure water content
Thanks, Nazar, and indeed the Karl Fischer method is being considered and in some cases used. But how accurate are these methods? I need some cold hard numbers ....
I am no TGA expert; however as an analyst I would always say that you should try to derive the uncertainty from the actual data. TGA basically means weighing, and weighing is one of the most precise measurement methods we have. But cold hard numbers mean nothing if they don't apply to your system.
Try to split your sample and to reproduce the values you get. Maybe you can even do cycles to test if you obtain a reproducible weight. If you get time-dependent data out of the TGA, fit the constant plateau and calculate with which deviations from constancy your data would be consistent, etc.
Erik, your suggestions make perfect sense for current and future measurements. However, I am stuck dealing with results that have been obtained some time ago, and by others. Hence the need for a statement based on general experience, so that I can put some reasonable upper bounds to the possible scatter of the data.
Jan-Martin, I quickly scanned through some TGA data I found on the web. My impression is that the precision is highly dependent on the system under investigation. 0.5% might be 0.50 +/- 0.05 or even more. If you do not have the original TGA graphs to analyse them I would suggest to find data on materials as similar as possible and to estimate the uncertainty on base of the uncertainties you find in literature. Probably not a very easy work to do, but maybe woth a shot, depending on your goal.
- Badges
- Science method
- Similar topics
- Chemistry
- General Biochemistry
- Biochemical Techniques
- Analytical Chemistry