I am planing to do Kissinger analysis of a certain glass sample to understand the nature of crystallization in the glass by determining it's Avrami parameter. To do so I must know the FWHM in the DTA curve of the glass, but I don't know how exactly to do this? The graph I get from instrument has a tilted baseline (shown green in the attached pic). By doing baseline correction I can get FWHM of this peak, but I want to know is it the right way to do so, because on the y axis we have heat flow which is an arbitrary unit. On the other hand I can find the FWHM without straightening the baseline and considering the red line in the attached picture as a baseline, but this will give an entirely different value of FWHM, so please can anyone tell me what is the correct way to find FWHM without manipulating or disturbing the raw data? One more thing I would ask here is how to put error bars in the data of any characterization curve which I plotted? I mean, in this case if I have 4-5 DTA experiments with different samples and peak temps, how will I know the error in % to put it on the graph (Picture 2).
If you assume that the baseline is a straight line, then you can try the first derivative of the DSC signal versus temperature. In this case the width defined as the distance between the inflection points in the original dependence will be calculated from the distance between the peaks.
You can develop rough estimation following the baseline as your green approximation. Nevertheless, deconvolution of the whole curve should be more precise if you really need to distinguish contributions in the thermogram. Hope this helps.
You are going on in right track with green line and for finding errors you can use ascii format in which all data with different temperature can be accessed in note pad and by applying statistics on it, you can find out errors for all spectra.
I would use the green baseline (actually you could use the resulting curve to calculate actvation energies and mechanisms employing other kinetic methods).
I don´t understand what you mean with "different samples" and I can´t see Picture 2.
Hope to be useful
I would suggest the book "Principles and Applications of Thermal Analysis" Edited by Paul Gabbott (Blackwell Publishing) if you need to fetch more accurate quantitative results. They have shown many... Stick to 'one' procedure for all of your data. Looking at attached figure I would say green line is better choice and for computing you can use ORIGIN.
Dear Anupam, you must subtract your correction curve from your sample's curve. If you want to put errors bars in results you must repeat the measurement for at list 3 different samples of the same experiment.
The simplest way is by extrapolating the left side of the curve until it reaches the horizontal green line while maintaining the normal distribution curve correctly. Than follows the standard procedure to find the FWHM.
I agree with a coment above that u must use a green baseline or else u will loss some details.
You can use Origin 8 to do the fitting as ilustrated:
*sorry distrub your data cause i think it's a hand drawing
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Materials
- Inorganic Materials
- Ceramics
- Glass