In some of papers where carbon concentration EPMA (using standards) analysis is applied, plotted carbon content (line scan) for ferrite is lower than 0.0%. What is your opinion? Does showing lower carbon concentration than 0.0% makes any sense?
As Dr Dusevich said, it is just a shift of the calibration. The conventional way for presentation of such results - just show it "as is" - without a manual correction, but with an explanation of that phenomenon.
Hi
What do you mean when you said lower than 0.0%C. You can extrapolate any experimentally obtained (line scan) Carbon concentration profile to lower than 0.0%C. But that doesn't mean any thing.
Dear Anicet,
for sure you are right about the material. But carbon concentration lower than 0.0% has no physical sense. Why not correct it even manually to 0.0% and put appropriate comment?
Dear Khaled,
please look at figure 7 in Acta Materialia, Volume 61, Issue 8, May 2013, Pages 3120-3129. The carbon concentration is plotted below 0.0%.
I am not asking about materials and methods. Concentration below 0.0 like length below 0.0 has no physical sense. Why showing such values?
OK Grzegorz, after looking at the figure 7 of the paper, I understand that what appears as negative C mass% should be interpreted as loss of C while positive values of the spectra are gains or enrichments (relative to a given standard). Regards, Anicet
In paper your referred to authors made a calibration curve for carbon (a good approach to EPMA measurements of light element). But they used not heights of peaks, but overall intensity (peak+background). So, as a zero concentration for C they accepted not absence of peak above background, but value of a background of specimen with no C. When they performed data acquisition of their specimens, some drift in parameters of their EPMA decreased slightly background intensity (compared to acqusition of standards), and they got "negative" results. Good practice is to check calibaration specimens again after all was done, which they probably skipped. So, their negative values can serve as an systematic error of their measurements. But still their results for higher C concentrations should be rather good for EPMA carbon quantification.
As for "why not correct it even manually to 0.0%" - it is extremely bad idea, one should never do it. Falsification of results...
Dear Vladimir in my opinion it is not falsification when you put appropriate comment that in these points carbon concentration is below detection limit. Do you agree?
Sure I agree. If data manipulation is acknowledged, then it is OK. But - and it is a personal preference - would still prefer to see original result.
As Dr Dusevich said, it is just a shift of the calibration. The conventional way for presentation of such results - just show it "as is" - without a manual correction, but with an explanation of that phenomenon.
- Badges
- Science method