GC Troubleshooting in Simple Pictures, Part I

October 1, 2016
Dawn Wallace Watson

LCGC North America

Volume 34, Issue 10
Page Number: 818

Five simple pictures that reveal problems with your GC analysis – and how to fix them!

A picture paints a thousand words. The art of gas chromatography (GC) troubleshooting often lies in being able to recognize a problem from the evidence presented in the chromatogram or baseline appearance. In this short series of two articles, we present 10 simple GC pictures that will show you how to recognize the issues, deal with the causes, and prevent them from happening again.

Here, the first five pictures are presented in Figures 1–5.

Baseline Spikes

What type of spike do you have? Zoom into the chromatogram and look at the noise-do the peaks have Gaussian shape and a peak width? If not, then the likely cause is electrical noise. Electrical noise can be caused by a poorly smoothed electrical supply to nearby electrical equipment. This type of noise can be overcome by fitting a power scrubber to the supply, isolating any equipment giving rise to the problem, removing it to a different area of the laboratory, or placing it on a different ring circuit. If the peaks have width and occur in a homologous series, then column bleed is to blame. To fix column bleed, the column can be reconditioned; however, if the bleed profile does not improve you may need to trim the column to remove any contaminated or damaged stationary phase (approximately 5% of the column length) and then recondition. If the peaks have a Gaussian shape, no apparent homologous series, and random peak heights, the column may be mounted too high in the flame ionization detector and may be coming into contact with the flame, which causes the polyimide coating to flake and peel off into the detector resulting in baseline spikes.



Chair-Shaped Peaks

If all peaks are affected, a physical issue rather than a chemical one is to blame. Chair-shaped peaks can be caused by a poor column cut. If the chair-shaped peak is accompanied by a broad solvent peak that exhibits a sudden vertical cutoff, the column is installed too high in the inlet.



Tailing Analyte Peaks

If some, but not all, analyte peaks are tailing, a chemical problem is indicated. The most likely cause of chemical tailing is secondary interactions of polar analytes with active sites, typically silanol groups, in the inlet liner, on glass wool in the liner, or at the edges of a poor column cut. When all analyte peaks tail, a physical problem is indicated (unless all analytes have similar chemistries-that is, they are all basic). For example, a poor column cut can expose surface silanol species that basic analytes can interact with, which can cause tailing, or it can result in turbulent eddies at the column inlet that cause analytes to be held up before they enter the column, leading to tailing. Recutting the column and examining the cut under a microscope to ensure it is smooth and at 90° to the column wall will resolve this issue.




Tailing Solvent Peak

A tailing solvent peak during splitless injection indicates a poorly optimized splitless time. Too long a splitless time causes solvent peak tailing because not enough of the solvent is vented away and instead enters the column. Whereas too short a splitless time results in loss of early eluted analytes. Monitoring the peak area of an early eluted peak and plotting it against splitless time until a consistent peak area is observed will give the optimum splitless time (5–10 s can be added to further optimize this time).



Shifting Baseline Position

A shifting baseline position is different from baseline drift. Here a shift between the start and end of the solvent peak is seen, which suggests that the gas flow is changing during the injection cycle. This flow change could be due to leaks during, before, or after injection caused by a cored or leaking septum. Check your septum regularly for damage.


There may be other causes for these problems, so for further tips visit the CHROMacademy GC Troubleshooter at www.chromacademy.com/gc_troubleshooting.html.