Chromatography Forum

Hi everyone, I am a Masters student who is new to this forum and am facing some problems with the Thermal Desorption-Gas Chromatography-Mass Spectrometry (TD-GC-MS).

I am currently using the GERSTEL Thermal Desorption Unit connected to the Cooled Injection System (CIS) where analytes are cryogenically trapped before entering the GC Agilent J&W DB-1 Column. Previously I conducted a recovery study by injecting 1.0uL of a test solution into the upper end of the thermal desorption tube packed with Tenax-TA and compared the peak-area ratio to that obtained by direct liquid injection. The results that I obtained was that as the CIS initial trapping temperature increases, the analyte recovery increases. For instance, at -50 degrees celsius, the recovery was 94% as compared to 55% at -100 degrees celsius. This was a little surprising as I would have expected the recovery at a lower temp to be higher since the analyte trapping would have been better. I tried searching for literature on this but I wasn't able to find any explanation for this.

I would greatly appreciate if anyone could advise me on this. Thank you!

Welcome to the forum.

When you say that you injected the recovery standard on the "upper end" of the Tenax tube, do you mean that it was the end that the gas flowed into while the tube was desorbing ?

If it was, then the higher recoveries at higher trapping temperatures are due to the need for the analytes to migrate all the way through the Tenax before they desorb into the cold trap. Normal practise is to load traps with the gas flow in one direction and desorb with the gas flowing the other way.

Try loading your recovery standard at the bottom of the trap.

Peter

Hello,

Did you remove the tenax liner from the CIS, and spike it?, or did you remove the TDU and put the
insert weldment back on, and inject onto the tenax liner whilst in the CIS whilst being cooled?

What chemical is in your test solution, what is it dissolved in, and is it liable to hydrolysis?
Do you see any hydrolysis products, or unexplained chemicals in one but not the other?

kind regards.
Alex

Hi Peter and Alex,

Thank you for the replies!

Yes, it was the end where gas flowed into as this method was recommended by the GERSTEL Application Chemist. I also took reference from the following literature: P. Werkhoff; W. Bretschneider J. Chromatogr.,1987, 405, 99-106. Could you explain further how does higher trapping temperatures lead to higher recoveries? I am interpreting as the analytes would have to migrate through Tenax all the way at all temperatures, regardless if it is at -50 or -100. I am aware of the normal practice, but I did it the other way because I wanted consistency in each injection by getting the instrument to inject into the Tenax instead of me doing it manually which could have human errors arising.

I was using Tenax-TA in the TDU but the baffled empty glass liner in the CIS. So what happened was that I selected the liquid injection option on the software and the multipurpose sampler gripped the Tenax tube, placed it in the TDU and the solution was injected into the Tenax. The initial temperature of the TDU was held at 25 degrees for 1.5 minutes before I ramped it up to 250 degrees and held it for 10 minutes.

The chemical solution contained a mixture of alcohols, esters, aldehydes, ketones and it was dissolved in Methanol as Tenax has low affinity for methanol. The chromatograms overlaid on each other quite well and there were no unexplained chemicals in any of them.

Interestingly, I found a piece of literature that states

Also the fact that some instrumental variables have been demonstrated as being non-relevant in the whole range under study can have important practical consequences. For instance the instinctive practice of analysts is to cool the injector as low as possible (−120 ◦ C or even −150 ◦ C) in order to guarantee the inter- mediate trapping of target compounds after twister desorption. This study has shown that this parameter is only relevant down to −60 ◦ C and using lower temperatures represents an unnecessary cost in liquid nitrogen.

K. MacNamara; R. Leardi; F. McGuigana Anal. Chim. Acts., 2009, 636, 190-197. This study was however conducted using the Stir Bar Sorptive Extraction method instead of Tenax, and it doesn't explain the reason for this observation.

Thank you once again!

Sincerely,
Yu Ying

I am with Peter. You should make your spikes on the side of the trap closest to the sample during preparation. That will provide you with the most reliable desorption and trapping because it most closely resembles what's actually happening during sampling. How repeatable are your 94% and 55% recovery experiments? Are they just from one shot of each? Perhaps replicate analyses would reveal a precision problem with one or both temperatures?

I would say that your cold trap only has to be as cold as require for the lightest material you are trying to determine. There's no point in going to -120 °C if you're trying to determine toluene in water. I remember "back in the day" my Tekmar P&T sampler was set up for liquid nitrogen. I have since found that liquid CO2 (actually dry ice after it comes out of the T) will trap essentially all of the materials that interest me. You don't need the things you are trapping to be solid on the column, you just need them to condense effectively.

You get higher recoveries at higher Tenax temperatures because at higher temperatures the analytes migrate faster through the Tenax so larger quantities are eluted downstream. A lot of this depends on gas flows and timings.

An additional complication is that the cold trap will be much more effective at any temperature if it has methanol condensed in it to form a temporary stationary phase - the quantity of methanol in the trap will depend on the Tenax temperature.

I am slightly puzzled as to why you would use such low temperatures with Tenax - its main function is its selectivity against water, but at you temperatures you would simply cold trap water along with everything else.

Peter

Hello,

I am trying to understand why you are using a baffled empty glass liner in the CIS.
As the traditional configuration for samples that have been collected on a tenax tube, eg air monitoring, is that you have a tenax liner in the CIS, that you concentrate onto when you desorb the tenax in the TDU.
And once you concentrated your analytes on the CIS, you then ballistic heat the CIS to desorb all the chemical simultaneously so they go onto the front of the column.

and as Peter mentioned, this provides the surface area to concentrate the desorbed analytes.

What is the temperature program on the CIS?

If you are taking it to a much lower temperature you would need to figure in the thermal mass of the baffled liner into the flash desorb temperature timing otherwise it may not be completely reaching the full temperature during the time you desorb from the cold trap.

I am not familiar with the CIS so I am just wondering, does it go into a split mode after heating? If so it could be you lose some analyte to the split flow if there is some lag in the heating of the liner due to the colder temperature.

I think there is some confusion in some answers regarding the CIS. Trapping is done by cold temperatures (liquid nitrogen), so there is (typically) no tenax in the liner. The sample (rather high boiling substances? in solution) is applied as liquid on a tenax tube (solvent removed by gas purging?) which gets desorbed (correct?).

Have you seen this effect, that the recovery is better at higher temperatures repeatedly? Is the magnitude of the effect different, depending on the boilng point of your analytes? The repeatability of thermodesorption is rather poor. Do a calibration with a few levels and use an internal standard. If you still see this effect for different trapping temperatures we can continue discussing the possible root cause(s).

Hi everyone,

Thank you for all the replies!

Yes, trapping is done by cryogenic trapping and I am using an empty baffled liner in CIS. The analytes were applied on the Tenax tube in the Thermal Desorption Unit as a liquid and gets desorbed. Recently I tried increasing the CIS initial temperature to 10 degrees and -30 degrees, and I observed that I lost quite a few of the more volatiles substances (those that elute at 8 mins or before). Therefore, I am sticking with -50 degrees. I used an internal standard and also did a repeatability study on 10 tenax tubes with the same parameters, %RSD that I got for all my analytes ranged from 9-23%.

The temperature program on CIS is -50 degrees to 250 degrees, ramped at 12 degrees per second which is the maximum ramp rate that I can go. It goes into split mode and the split ratio is 10:1 for all the temperatures that I tried.

The reason why we chose to use baffled instead of a Tenax liner in the CIS was because in the past, people in the laboratory observed that using Tenax liner with cooling by a Ultra Peltier Cooler instead of liquid nitrogen caused tailing peaks for Limonene. Following that, they decided to go with liquid nitrogen but could no longer use Tenax as crystallisation forms even more easily in the Tenax liner. Therefore, using a baffled liner with cryogenic trapping was chosen eventually as it solved the problem of the tailing peak.