Chromatography Forum

Hello, I am currently using an Agilent GC-MS system with ChemStation (Enhanced Chemstation, MSD ChemStation E02.01.1177). My supervisor says he remembers that there was a function that could help reduce the noise from non-peaks. However, he said that it was almost ten years ago so he doesn't have much memory from it. From basic googling I assumed it might be the subtract background function (BSB), however I can't tell if it is right and I'm using it wrong or that I'm just completely wrong, as using it causing lower match scores and sometimes wrong library searches from the libraries. Basically I want to know if BSB is the right tool, or if there is anything else that could work? Thanks.

What is your threshold set to in the MS part of the method setup? I use environmental version but should be similar, when you click on the yellow quad icon where you can set scan range and solvent delay and EM volts, there should be a section where you can set scan range, sampling and threshold values. If it is set to zero, you will have a high background noise. Try adjusting it to 150 and see if that helps. Background subtraction that I have used only works to remove background from the spectra of a peak, you select the peak spectra, then a background area spectra and do subtraction. If the section you choose as background has a high abundance you will subtract too much from the peak of interest.

The more I look at the GC-MS data, the more I question if I/he needs to be concerned about the background noise. I checked the methods and they already have threshold values of around 150. Besides that, I don't even think the background levels are high enough to raise issue. The background noise is around 10,000-20,000 abundance compared to the peaks which are 2,000,000+. I've tried suggesting that maybe it isn't something to worry about, but he insists that he remembers doing something. /shrug

EDIT: In regards to subtract background, I click the button and it does subtract background. I'm not sure if that is how yours works, but would it be subtracting the background based on the current window (i.e. if it was zoomed in) or is it based off the whole gc? Because sometimes even though the peak is much larger than any values in my current window, subtract background will almost entirely subtract a peak. Thanks again

Subtract background should take the last spectrum you selected and subtract that from the next to last spectrum you selected. To remove background from the scan of a peak you right click to get the spectrum of the peak, then right click to choose the background spectrum to subtract the click the Subtract menu item. It doesn't subtract the background from the entire chromatogram, at least not in the versions I use.

You may want to consider this paper,

“Optimizing Spectroscopic Signal-to-Noise Ratio in Analysis of Data Collected by a Chromatographic/ Spectroscopic System,” Zheng & McElvain,
Anal. Chem. 1999, 71, 39-45

.....to achieve a better S/N when summing spectra.....only the spectra whose relative abundance is above 38% should be included.

This should work with data that has already been acquired.
For future work, reduce the scan range as low as possible so that background noise will be reduced and your S/N ratio will be naturally enhanced; then apply the 38% criterion.

Regards,
JMB