Agilent 7890A Cryogenic cooling for sulphur analysis

[mbdxhmc3]

I have an Agilent 7890A fitted with a cryo valve for use with liquid CO2 to cool the column down to -20C. The column used is an Rtx-1 and the detector is an SCD. I am aiming to try and separate and analyse SO2 and COS at ppb level in bulk CO2. As far as I know this can only be achieved using sub-zero temperatures.

I have been able to cool the oven perfectly fine providing the GC is contained within a fume cupboard. However if performed in a lab with "average" ventilation the volum of CO2 exhausted from the GC becomes so great it goes over the 8 hour and 15 min working limits (at about 15C).

Are there any other methods available that could achieve sub-zero temperatures without the need for large volumes of CO2 (or any other asphyxiant gas) to be injected into the GC oven?

Alternatively does anybody know of any methods available that will separate COS and SO2 at ppb level in bulk CO2?

[CE Instruments]

Running CO2 cooling to keep your GC at -20C is not ideal , it takes a lot of CO2. You would be better off using liquid Nitrogen cooling which has a greater efficiency and no toxicity issues.

[mbdxhmc3]

Thanks for your reply.

Would liquid N2 be run at pressure though? Because if it was I would assume the consumption rate would still remain fairly high leading to oxygen depletion issues?

[CE Instruments]

Dry Ice is approx -78C, liquid Nitrogen -187 so you need much less. Air is ~80% Nitrogen, the dilution effect is much lower it should have no effect on the Oxygen. CO2 causes drowsyness at concentrations around 10000ppm compared to the natural level of about 300ppm, I'm not sure of the actual safety limit, 1000ppm ??

[mbdxhmc3]

It's something that we're definitely considering using at the moment but as I say I'm unsure whether we'll need pressurised nitrogen or liquid nitrogen contained in a dewar.

Do you happen to know anything about cryo-traps? I'm wondering whether that may provide a more practical solution.

[lynoguchi]

Did you think about to change your column?
The Agilent has one column for sulfur compounds: http://www.chem.agilent.com/cag/cabu/pdf/gaspro.pdf
This column has a good separation for COS and SO2 even at ambient temperature.

[Steve Reimer]

I run one of my GCs (an Agilent 6890) with liquid nitrogen oven cooling along with a cryo trap on a vacuum distiller.
For your use a dewar nearby is ideal. With a 22 or 50 psi relief valve you should be just fine. The cryo valve for liquid nitrogen is not the same as the one for CO2 so you would have to swap that out.

[mbdxhmc3]

We have tried the gas pro column. Unfortunately we can't seem to get the sensitivity we require (~5ppb).

Thanks for your help. What sort of temperatures can you get to using N2? It looks like we'll need to go to -30C. Using your equipment how much nitrogen would you get through if you were running say 10-15 tests each lasting 10-15 minutes at -30C?

[Steve Reimer]

I use nitrogen to cool my cryo trap to -150 so -30 is easy. You don't even need to have liquid getting to the oven for -30. How fast you go through a 200 liter dewar is dependent on how much loss you have on cooling everything down. The oven doesn't take much to keep cool.
At 3 to 4 hours of a day your dewar might last a week.

[James_Ball]

We have tried the gas pro column. Unfortunately we can't seem to get the sensitivity we require (~5ppb).

Thanks for your help. What sort of temperatures can you get to using N2? It looks like we'll need to go to -30C. Using your equipment how much nitrogen would you get through if you were running say 10-15 tests each lasting 10-15 minutes at -30C?

When I first began in GC/MS 20 years ago we were using an HP5995 with LN2 cooling. We would go through at least a dewar a week for each instrument, with two instruments in a small room. We never had any problems with air quality so you should be ok on that.

The 7890 is much better insulated than that old 5995 an we were using the LN2 to quick cool from 350C down to 50C about every 30 minutes and would run 30-40 samples each day. I would guess that you would use a dewar a week or less for your application. If this test is run isothermal then I would think that once you get all the metal cooled down in the oven it should hold temperature pretty well. If it is only being used for this one test you may even want to add some 1/4 inch styrofoam to the walls of the oven to help reduce the amount of nitrogen needed if it seems excessive.

[AICMM]

mbdxhmc3,

Thought about a Plot-U or Restek micro-packed sulfur column.

Best regards,

AICMM

[ydna1977]

Hi
We have installed a CO2 cooler to HP6890 but only carried out a few test injections and suspect it has used all CO2 gas from CO2 dip tube tank (25KG)
A number of settings on Empower 3 cyro cooling config although not sure which will help to avoid using too much CO2 (ie Cryo timeout enabled , blast cooler ) where temp needs to go to 20C .. any thoughts ?

[ydna1977]

Hi

25kg of CO2 ran out during sample set
Should this happen ?.. or limited to number of injections < 15 ?
Thanks

[Sunny.Singh]

Hello,

We did a similar analysis but on LPG samples for sulphur impurities. The experimental conditions used are detailed below for your reference:

Sample loop 100μL
Channel A 220°C, Split 1:30
Channel B 220°C, Split 1:20
Oven 35°C, 12°C/min to 250°C
(1.25min)
Column A SCION-5, 30m x 0.32mm
x 5μm
Column B BOND Q 25m x 0.32mm
Carrier Helium, 2mL/min
PFPD 200°C
Air 1 & 2 17mL/min, 10mL/min
H2 13mL/min

The instrument we used was a Scion Dual channel GC with PFPD detector.

Please drop me a word at sunny.singh@scioninstruments.com to get full details, results and chromatograms.

[ydna1977]

Hi ,
How many injections were carried out using CO2 (25KG) cylinder ?