How to quantify sample from a cold trap

[TheIndian]

Hi,

Firstly, thanks to all, admin and users for the work, this is always a good place for information. Sorry if it is not the appropriate place for posting, I was in doubt between this and sample preparation, and also I look for something similar but find nothing.

There I go. Briefly, I want to quantify some sample previously trapped in a cold trap, basically a u shape tube in liquid nitrogen. I want to raise temperature and quantify what was trapped before, but I have serious doubt about how to do it and do it in a precise and reproducibly way.

Explaining a little more, I want to make some treatment to a sample trough inert flow, it will desorb some gases at ambient temperature, like ammonia, in a small quantity (around 0.5ml STP) wich would be freeze in the nitrogen trap. I have an online chromatograph with a valve injection and a FID detector wich could be connected (or not) to the end of the trap.

I have experience with GC for quantifying from continous reactors and manual injections of liquid samples but when I think in this issue I see obvious troubles. If I raise temperature and push the gases previously trapped they would not be in the loop for the injection (at least not at the same quantity). Taking into account the typically layout of a injection valve(at least for me), while loop is filling with sample flow chrom carrier is going into column, when switch to the other position, carrier pushes sample from the loop into column. So if I raise temperature and inject nothing assures sample goes into column in a reproducible manner. The same thing applies for using a continous flow as standard.

I thought in how to fix it and maybe it could be done in a very unusual way, for me, changing valve ports and making sample inert act as "carrier" of the chromatograph when switch the valve, and inject all the sample into the column but anyone who has change or fix an injection valve knows that is a very annoying stuff, so small and dedicated the tubes, and this is not a operation to do with regularity in order to keep equipment versatility. Chromatograph is an old one, manual gas valves, so I think there wouldn't be electronics/software problems. By the way i would like to ask to you: do you think there would be any column/detector problem by injecting 1STP ml of gas sample into a 1/8" packed column ( for example tipical porapak q, p or N)? Taking into account that there would be not too much resolution problem (only 2 or 3 compounds with diferent polarity)... I think there's no problem, but would be nice knowing your opinion.

On the other hand I thought in calibrate volume of the trap (with pressure system), put a septum and take some known volume sample with a syringe and inject manually but possibly temperature-pressure variations in trap, human errors with volume of sample taken mixed with absolute calibrate makes me mistrust with the precision of quantification.

I have seen something similar for concentrate and quantifying VOCs but only comercial ones and can not find how it works.

Sorry for message length, maybe it is more simple than i am thinking. I try to synthesize to the fullest and make it understable (maybe do not achieve), so omitted some thecnical aspects like flow, equipment, but if you consider it relevant, please do not hesitate in asking and i will try to explain it better. Also sorry about probably technical mistakes

All ideas, advice or reports would be very welcome.
Thanks in advance for your response

[James_Ball]

If you have the ability to cool the GC oven with liquid nitrogen, then you could do your "injection" with the oven cold enough to trap the analyte at the head of the column, then do a ramp of oven temperature as a normal analysis. Also there used to be a device that did the same thing but with a piece of uncoated guard column outside the oven. It was used for doing volatiles from a purge and trap before we were doing them with split injection ports. It would trap and concentrate the flow from the purge and trap into a tight band in the guard column before rapidly warming and passing them onto the analytical column, which sounds like what you would need.

If you feed your cold trap into a normal injection loop that is also in liquid nitrogen you would then trap the analytes in a smaller loop which can then be injected like a normal gas sample loop injection. It might be something you can build yourself without too much trouble.

[TheIndian]

Thanks for your answer James_Ball, really interesting.

I´ve been thinking on it. I knew about cooling oven GC for doing subambient analysis, but seems little strange for me filling all the oven with liquid nitrogen, I do not trust myself in not spread it over the lab.
About the device/guard column it sounds really good. If anyone has some ref or paper where I could take a look, I would be very grateful. Anyway with this disposal the sample inert flow would act as carrier of the chromatograph, isn´t it? or maybe it´s something I´m misunderstanding? After writting the first post I thought in that option and I could do it without touching injection valve, but is really weird for me operating a chormatograph without using carrier inlet. Do you think it is any problem with that? I could use sample flow with an equal flow of carrier and trap is really close of GC.

About cooling injection loop, I am thinking on it but I think that, taking into account flows and quantities, if I could use sample inert flow as carrier it is an easiest way

I don´t know if i´m explaining it well, I could try to do it better if anyone wants.

Thanks a lot

[James_Ball]

For cooling the oven, you don't fill the oven with liquid nitrogen. Most of the GC manufacturers make a valve that will spray liquid nitrogen in a mist into the oven to reach a lower temperature. You only have to go low enough for the stationary phase of the column to trap the analytes you are interested in from the gas flow in the loop, then after flushing from the loop through the column those analytes will be trapped at the beginning of the column, then you do a normal oven temperature ramp to elute them through the detector.

[AICMM]

They used to do this all the time in the really old days. Lots of literature of the old-timers lifting the trap out of the bath and making the injection. If I am not mistaken, the trick is bypass flow while cooling and trap in-line flow while heating. Everything goes on column at that point. I will see if I can find examples and post a pic.

The other option is a cryo inlet like Gerstel or JAS.

Best regards,