Chromatography Forum

Hi, i'm looking for a suitable solvent to use in GC-MS (column DB-1 30m x 0.25 um i.d., split injection mode) analysis of a mixture of unknown low weight aliphatic hydrocarbons, basically C5-C10 linear and cyclic alkanes and alkenes coming from ethylene fed glow discharges. i've tried a series of solvents, e.g. acetone, diethylether, dichloromethane, methanol and ethanol. the sampling procedure necessarily needs a certain amount of solvent, preventing me from injecting the mixture alone. But in this way the C5 compounds are completely lost in the solvent peak, regardless of split ratio and flow rate. Any suggestion about a solvent or other method i can use in order to obtain a decent chromatogram? thanks a lot!

May be you can try with a solvent with a high bolling point or even inject the pure sample

Do you think an high boiling point hydrocarbon, undecane for example, may serve the purpose? I will try...

injection of the pure mixture is not possible because of the sampling technique (the analyte is collected in a cold trap to be washed with the appropriate solvent).

My doubt arise from the fact that the solvent is in large excess, so i think that a broad "solvent peak" is unavoidable somewere in the chromatogram.

Thank you davide

I think you can analyse your compound by infrared spectroscopy. If you extract your volatiles organics compounds by a solvent wich contain no C-H bonds you can analyse it by infrared. But in this way you have only the total amount of hydrocarbons. The wave lengh is near 3000 cm-1. If this kind of method interest you i can send you the method.

Davide

You may also want to try a PONA (DHA) column. In my experience they resolve C3-C10 hydrocarbons very effectively, and light nonpolar solvents (such as DCM, CS2) elute early enough that your solvent peak isn't much of an issue.

Your phase is not very thick for the compounds you are trying to determine. If you went to a 1 or 1.5 micron phase thickness I believe you will find the results more satisfactory. In many cases carbon disulfide is used as the solvent for this type of analysis. There are certain safety precautions to take when using CS2, but it will work well.

I think you can analyse your compound by infrared spectroscopy. If you extract your volatiles organics compounds by a solvent wich contain no C-H bonds you can analyse it by infrared. But in this way you have only the total amount of hydrocarbons. The wave lengh is near 3000 cm-1. If this kind of method interest you i can send you the method.

Davide

Thank you very much, davide.
My task is qualitative in principle... i need as much (structural) information as possible, to deduce possible reaction pathways.

However the -high boiling point solvent ( thank you again!)- route seems promising; with undecane as solvent the first region of the chromat. results clean, the solvent peak appears very late, in a "dead land" of no interest.

But:
1- in can't use undecane - too expensive - i should revert to a mixture of high boilig point alkanes or such that.

2- viscosity could be a problem (i have to wash a large volume trap).

3- there's something strange in the peak shape of the analytes, in particular C5 - C6. They eluite as double peaks, a major one and immediately after a small one, about 1/10 of the parent one tall. If i pick their mass spectra they are identical! This problem is dependent on injector temperature. working at 280 °C (instead of 250°C) the two peaks coalesce into a single one but still asymmetric. I'm afraid this artefact is due to a "two stage" vaporization of the sample at the injector level (undecane is so heavy, oh!). Do you have any experience of such phenomenon?



Thank you for any suggestion!

As previously suggested at thicker phase column (1.0u) should enable you to analyse the C5 components using diethyl ether as a solvent. I have analysed pentane (easily) using diethyl ether (use the AR grade solvent with BHT to avoid early eluting oxidation products of the ether) as a solvent with an AT-5ms 30m x 0.25mm, 1.0 u FT. To get the solvent delay time - put a drop (10ul) of each of the solvents into a sealed vial and inject 1ul of the sample head space (you can use an auto sampler - remove the solvent bottles and ensure there is no residual solvent in the syringe and set solvent delay time to 0).

Cheers
Bas

use a GC/FID system direct injection.
if you going to use GC/MS you need a purgeTrap system to work.