Page 1 of 1

How to calculate S/N on mass spectrum..

Posted: Tue Sep 09, 2008 12:37 pm
by Pegry
Hi everybody,

we have a single quad instrument (Shimadzu LCMS2010) and we need to estimate the Lod in SCAN mode for some compounds. We can measure the signal to noise ratio in 3 ways:

1) Calculate S/N on TIC (total ion current chromatogram)
2) Calculate S/N on EIC (extracted ion current chromatogram)
3) Calculate S/N on mass spectrum.

How would you calculate the signal/noise ratio on mass spectrum? I would acquire the average mass spectra on the peak and measure the intensity of the signal on the molecular ion, but I don't know how to measure the intensity of the noise from the mass spectra.. How would you do that?


Thanks in advance for any answer,

Have a good day!

Davide

Posted: Tue Sep 09, 2008 10:21 pm
by Don_Hilton
Because you are trying to determine the limit of detection - the limit you can trust that you can differentiate between a real signal and zero (random background noise), use the signal you are using to detect the peak. In an MS analysis, this is often the extracted signal.

I would also suggest that you look at methods such as the US EPA methods on determination of LOD's. The signal to noise determination is a starting place and is used to identify the concentration of sample for multiple injections. This sample is injected several times and student's t is used to determine what can really be detected at a particular level of confidence.

Posted: Wed Sep 10, 2008 7:33 am
by Pegry
Thank you Don for your answer.
What we really need is just an estimate for the LoD and not the validation of it. We would estimate the S/N with all the three ways because we are working with drug impurities and we want to get an estimate of the LoD on known impurities to have an idea of the sensitivity also on unknown impurities. The calculation of S/N on EIC is the best if we know the mass of the compound. I would like to calculate the S/N on the mass spectrum to see when a signal can be distinguished from noise also on similar compound but unknown. I found in a recent article that they calculate the S/N for compound A in 3 different ways (but they does'n report how they calculate it):

Comp.A S/N on TIC = 1.3
Comp.A S/N on EIC = 41
Comp.A S/N on mass spectrum = 60

I just want to ask how to calculate the S/N on mass spectrum and why it is higher than that on EIC and TIC.

Sorry if I have not been clear and thanks for the answer..