Multiple detector GC-MS/TCD

[PPD]

Hello everyone, I am a new guy here.

Recently, my project involves collecting MS and TCD data simultaneously from one capillary column. I checked on-line that the capillary outlet splitter is available but the adjustment of flow rate to different detectors stands on my way since the MS vaccum tends to suck in more air.

I have little experience on that. Could anyone give me some suggestion? Thanks a lot.

By the way, there is a rookie question. Our GC is Varian 3800 came with a TCD. The two pipes I could find on the TCD seemed the outlet ports. Where are the Inlet ports for sample and reference?

[Peter Apps]

Connect the MS to the outlet splitter using narrow bore deactivated silica - say 50 micron internal diameter. You will need to play with thelength of the connection to get the flow into the MS and to the TCD properly split.

Peter

[PPD]

Hi, Peter

Thanks for your suggestion and I will give it a go. Meanwhile, I am so happy to read some ideas you have in the following topic
GC/MS - response factors: http://www.sepsci.com/chromforum/viewto ... =pyrolysis
where you mentioned, if one will be able to identify the compounds with a reasonable degree of certainty, one could then measure their response factors by injecting a known quantity of representative members of each class. This might work better than trying to model the response factors.

In my case, the Pyro-GC/MS is used too. The purpose that the column flow being splitted towards TCD is to quantify the complex pyrolysis products. Response factor is also what we are concerning for those compounds.

With regards to model the response factors or inject the representative standards of each class, which way will provide quantification in TCD case? Or do you have any other ideas of quantitation in my case?

Best wishes, Kai

[Peter Apps]

Hello Kai

With a TCD you have to calibrate with each individual analyte compound - response factors are not uniform within a class of compound. Therefore, you can just as well calibrate the MS response against individual compounds, and then there is no need to split to two detectors (unless you want the TCD data for something alse as well). An FID would work better than a TCD for the approach that you have in mind.

Peter

[PPD]

Hello Peter,

Thanks for your instant reply. The reason we want to use TCD is based on the complexity of the pyrolysis products. And the standards of a large part of the compounds are not available. Therefore, we are thinking of modeling the response factor to get the quantified results.

For my point of view, I also hope can just use MS. Split union to TCD serves as a source of leak problem. But in this case, is that possible to quantify the data just with MS? what's your opinion?

Best wishes,
Kai

[Peter Apps]

Hello Kai

Of all the GC detectors, the one with the most uniform response to different members of a chemical class is the FID - is there a reason why you cannot use an FID for quantitation and the MS for identification ?

The TCD performs no better with complex samples than do any of the other GC detectors.

Peter

[PPD]

Hello Peter,

Thanks for your suggestion of FID. At recent stage, we just have a TCD installed on our GC. That's why we plan to try TCD first to see how things happen. And also we are trying to avoid a lot of standards for calibration.

I read the a book about GC Detectors before. It is said that the response of FID is dependent on the number of carbon atom in each class of compound. Is that right? Sorry, I had little experience of FID.

With regards to use FID for quantification in my case, do I have to model the relative response factors too? or there are any other ways?

Best regards,
Kai

[Peter Apps]

Hi Kai

To a pretty good approximation an FID responds to the mass (moles) of carbon in a compound - so you calibrate in mass and get results in mass. The signal per unit mass is different for different classes of compound, but (reasonably) uniform within a class. For example, in pyrolysis you would need to calibrate with a saturated hydrocarbon, an aromatic, and an unsaturated aliphatic, and if you have chlorines in the samples with some chlorinated compounds. You only need to do one (or preferably two) compounds from each class. Bear in mind that the uniformity of response is poorer for small molecules.

The response factor modelling that I have seen is not particualrly impressive.

Peter

[PPD]

Thanks, Peter

I'll try and see what happens in TCD. If it doesn't work, we will consideration the option of FID.

Best wishes, Kai