Chromatogram on MSA 15%O2,1.5%CH4,60ppmCO,20ppmH2S,N2bal

[davidbrianlove]

I ran some Shimadzu GC8A chromatograms on a cylinder of MSA calibration gas for our Solaris multigas detetors. The gas is supposed to be 15%O2, 1.5%CH4, 60ppm CO, 20ppm H2S, and balance nitrogen. I was using 20ft of HayesepD, helium carrier at 20psi, TCD at 100mA, at attenuation was X1. I had injector and columns at 60C. Other than CO, which was hidden by air and CH4 peaks, I did a pretty accurate job on CH4 and H2S, but got a 2000ppm water peak at 15 minutes. Anyone ever GC one of these MSA calibration cylinders, and have you seen signficant water peaks ?

Dave

[chromatographer1]

I cannot comment on your questions but I do have a question. How many times did you see a water peak? And how long of a supply line from the cylinder to the sampling valve did you have? Do you think the water was definitely from the cylinder or was it from the supply line?

Do you consider a 0.2% error on your calibration significant or are you just curious that you have an impurity where it was not expected. I have had such peaks from commercial mixes before, including a 500 ppm contamination of argon in a 1000 ppm oxygen in nitrogen balance mix.

best wishes,

Rod

[davidbrianlove]

Rod,

Thanks for the feedback. The water could be from a number of sources assosciated with the calibration cylinder - its been laying on a shelf in a hot/humid area with its regulator and a length of tygon tubing for years. I flushed it for a while, but didn't want to fill the area with a slightly rotten egg small (H2S).

The reason for concern is that the project I'm going to test is for sources of water contamination in an argon line. I have the mA and attenuation set for max detectability of water. If I see a peak, I want to believe its real.

I rerun the test a few more times, and see what I get. 20 minutes elution time gets on my nerves. At this sensitive settings of attenuation and mA to the TCD, I can't do temperature ramps easily to speed up water and H2S elution - the columns are manually adjusted for a flat baseline, and the settings for 60C don't produce a flat baseline at 120C. I'm planting seeds for getting a new GC with electronic pressure and flow control.

Thanks for the help.

Dave

[chromatographer1]

Oh, Tygon, so soft, so permeable ! Teflon is no better.

Bake out some metal tubing and make your connections tight.

Water sticks to Steel and copper so you may have to heat the lines to get all the water out. Unfortunately, bare steel reacts with H2S so you may need to get some Fused silica coated SS for your lines.

Water water is everywhere, dang it sometimes. Good luck.

Rod

[AICMM]

DavidBrianLove,

I have to agree with Rod, water is a real bear to get rid of. Very sticky and everywhere. If you want to do high speed water analysis and you don't care about the other contaminants, put a thick film PDMS or phenylPDMS column in there and run isothermal around 50 or 80 C. Water comes off pretty quickly under those conditions with pretty decent peak shape.

By the way, it would not surprise me at all that the water is actually in the standards (again, very hard to get rid of it all....)

Out of curiousity, how did you determine the concentration? Also, have you thought about putting a mol-sieve trap on the end of your argon line and getting on with things?

Best regards.

[davidbrianlove]

Thanks for the added feedback. The Shimadzu GC8A is hooked up to a Chromatopak CR601 integrator - it has a thermal printer and it appears to do an area under the curve count in blocks, and then calculates the percentages based on block count. I don't have any manuals for any of the equipment - it was sitting in a storage trailer at our company.

I ran some more repeats today, and I continue to see a water peak on virtually every test. The level is 2500 to 3000 pm, and the peak is kinda rounded - not sharp at all, maybe this is typical of water peaks at 60C.

As I tried to indicate in the 1st message, my goal is not to evaluate calibration cylinders, but do go throughout our facility and test argon lines from various liquid argon farms, and see how much, if any, of a moisture problem we have. Praxair was kinda sketchy on their tests, so we want to look for ourselves. Moisture may be a bad actor in some of our processes.

There were four 10' hayesep columns in the storage trailer - I have them plumbed as 2 sets of 20'.

My choice of 60C temperature was trying to keep water moving, but not have other permanent gases overlap too much. I would like to keep hydrogen, and air peaks separate from argon peak, and also see of CO2 if there is one. How cold can I go on 20' of HayesepD and still get water peak in reasonable time ?

Thanks again.

Dave

[chromatographer1]

Dave,

Some comments:

if your column is made of stainless steel you will not be able to quantitate water well, as it should measure low. How low is difficult to say, but it will be low.

If you want to measure water the feed lines and column material should be kept quite warm, in the 90-120°C range. Water elutes between ethane and propane on Hayesep D.

Alternately, you could replace the lines and column material with fused silica lined SS tubing. Supelco could supply it. (pardon the commercial)

I would do water analysis via TCD with a Carbowax 20M packed column or 0.53mm ID capillary column.

You wrote:

"My choice of 60C temperature was trying to keep water moving, but not have other permanent gases overlap too much. I would like to keep hydrogen, and air peaks separate from argon peak, and also see of CO2 if there is one. How cold can I go on 20' of HayesepD and still get water peak in reasonable time ? "

H2 will be separated well, but air (nitrogen and oxygen) will not separate from argon on this column. Ar and O2 coelute. Methane and CO2 will separate but the R polymer will do a better job of separating the CO2 and give a better peak shape for the water.

The rounded shape of the water indicates that absorption is taking place and you are not measuring water accurately. I would measure trace amounts of water with Drager tubes from Mine Safety, or with portable Karl Fisher titrator in going from site to site.

Otherwise, use one column to measure water, and another column (MS5A) to measure the gases subambiently would be my recommendation.

best wishes,

Rod

[JI2002]

I use same column, but different GC with different oven temperature program. I have a tailing water peak at the end of run in every run, even in instrument blanks. The water peak gets bigger with longer intervals between instrument blanks. My guess was that the water comes from carrier gas (we don't have water moisture trap for carrier gas).

[chromatographer1]

By all means, get an impurity indicating trap. At least put an old mole sieve column in line to your GC. It will extend the life of your packed and capillary column and save you money in the long run.

But having a trap before the GC many not be enough. Many new GCs have an EPC module to control carrier flow and pressure.

This module may have a diaphragm reference to atmospheric pressure.

This diaphragm may be permeable to water in the atmosphere and contaminate the carrier gas stream at low but detectable levels.

I see it on my TCD when I install a freshly conditioned mole sieve column. The signal drops measurably as the water is trapped on column.

best wishes,

Rod

[JI2002]

Are you sure the peak at 15 min is water? I did some tests on my system, water peak doesn't elute at 60C after 30 min. run.

[davidbrianlove]

Actually, no. Its more based on prior experience, chromatagram databases at Grace Discovery Sciences. I don't have a mass spec., and I don't have a reference with a known quantity of water. I have "blown" in the sampling loop, and "blown" into syringes, but haven't done it with the current column setup (20 feet on each leg) (perhaps I should). If you hold your breath long enough, you get N2, O2, CO2, and H2O peaks. I'll have to try this stuff again to get reference peak times. I got worried that I was damaging the columns by putting so much moisture in. I do have some syringes, but the are a little dried out from years in storage trailer, so I've hesitated from doing a lot of injections.

Is there a column that at temperatures above boiling that would have good separation on permanent gas components (H2, N2, O2, Argon, CO2), and still have a good water peak in a decent amount of time ? Please remember that I have a 20-some year old Shimadzu GC-8A-TCD, no manuals, fully manual pressure setting, no flow control, and a thermal printer Shimadzu CR601 integrator.

Thanks!

Dave

[chromatographer1]

Dave,

I sent you an email with an attached 20 page document detailing fixed gas analysis and the various columns which are capable of performing the separations you require.

If you have any questions you are welcome to call or email me.

best wishes,

Rodney George
Senior Research and Development Scientist
Gas Separations Research
Supelco
595 North Harrison Road
Bellefonte, PA 16823

814-359-5737 voice
814-359-5459 fax
rodney.george@sial.com