Chromatograms are like fingerprints. If you can “read” chromatograms well, you often can find a plausible cause. In this series, we will show a series of GC-chromatograms that are obtained from users and discuss some potential causes for the phenomena. Then we can move into some solutions for improvement.
Fig.1 Broadened peak shape for volatile compounds because of injection error
Sometimes chromatograms look like Fig. 1. This is analysis of sulfur gases in nitrogen. The concentration is very low and a sample loop is used to inject direct on to the column. The first eluting components show a lot of band broadening. If we would calculate a plate number on these peaks we would become extreme low numbers.
The problem here is, that the large injection volume causes this peak broadening. The impact is especially high on the most volatile components. These components are already “moving” through the column, while the injection is still happening.
If the column flow is 4 mL/min,(30 cm/s) and the injection volume is 400 uL, it will take 6 seconds to transfer all sample into the column. The initial band width for a non retained component is already 6 seconds, which translates in a peak width-in space that is 180 cm wide. If the component has retention, there will be a focusing effect, depending on how much the component is retained.
Ideally the components in the sample must be focused as a narrow band on the inlet of the column. The difference is shown in fig. 2 here a higher split ratio was used and immediate the peaks shape of the H2S and COS improve.
The choice of parameters is therefore important and for optimal focusing the following parameters must be considered;
Fig.2 Using high enough split ratio, the injection band remains small, and efficiency is not compromised
Use a GC column with highest possible retention: the higher the retention, the bigger is the phase-focusing effect and the injection band becomes narrower; That’s why they used 7 microns in Fig. 1 and 2; The same problems also happen in Headspace analysis. Also here the components need to be trapped by the stationary phase. That’s why for headspace, thick-film coated columns are used (like the Rtx-VMS and the Rxi-624Sil MS).
Use a GC column which allows highest injection volume. This will be 0.53mm. The 0.53mm can deal with largest volumes as the length of the injection band in-space, will be the shortest using a 0.53mm.
Start at lowest possible oven temperature. If no cryo is available this will be limited by most GC’s to about 30C.
Inject smallest possible amount that give the required detection limits; This can easiest be done using a split system; 0.53mm columns usually need a split ration of minimal 1:5;
Make sure there are no dead volumes; dead volumes always cause peak tailing which will compromise separation of close eluting components; In case of a direct injection, consider to use the uniliner. If directly coupled to a valve, make sure you use the correct ferrules.
Fig.3 Methane peak is splitted due to inejction error
Fig 3 shows another example where an injection error shows up. Here a gas sample was injected and despite of using small sample volume, and sufficient split, there is a peak-split visible on the first eluting peak, being methane;
In this case there is also an injection problem, which is caused by a double injection.
This can happen for volatile components when injected is done using a gas syringe. To eliminate this one can use fast auto-sampler injection or use a valve setup using a gas-loop.