Missing compounds using air bags and sample loops

[yuttingzheng]

Hi, I just got a really difficult problem about injecting gas samples by using air bags and a gas sampling loop system. Looking for any suggestions to figure out this problem. Thanks!!

I squeezed an air bag to inject a gas STD into a sample loop system. The gas STD including benzene, toluene, ethylbenzene and p-, m-, o-xylenes (only 10 ppm for each compound), balanced with Nitrogen. There is only benzene peak showed up. Toluene sometimes showed up with weak intensity. All others never showed up. If I manually inject the STD from injection port by using a 50 micro-L syringe, I can see all compounds. My question is, where are those missing compounds when using sample loop system?

The loop system is a little complicated:
1. When filling the loop by squeezing an air bag, three loops are supposed to be filled: Loop 1 (1 mL) -> Loop 2 (0.5 mL) -> Loop 3 (0.5 mL) -> vent
2. When start the GC program, a 10-way and 3 6-way valves are turned and gases in the loop system are carried into the GC columns
Loop 1 -> Hayesap series columns -> TCD
Loop 2 -> Rt-Alumina column -> FID 1 (split ratio = 50)
Loop 3 -> Rtx-1 column -> FID 2 (split ratio = 39)
Aromatics are expected to be shown in FID 2

The loop system is in a heating box with 160 (also tried 90) degree C. So, it is impossible that these aromatics are condensed. Oven temperature = 90 degree C and is programmed to increase to 200 degree C.

Benzene peak looks good and reproducible, but not for the others.

GC: Shimadzu 2014
Carrier gas: Helium
Injection port T = 150 degree C
Mode: constant pressure (due to high pressure, I only can use constant pressure... also tried constant flow but didn't help)

Thank you very much!!

[Johnny Rod]

These components do not have high vapour pressures and are prone to absorbing into rubbers/plastics. How are you connecting the gas bag to the GC? If you are not concerned about a bit of air in your sample then I would prefer to withdraw a sample from the bag with a large gas-tight syringe (20-50ml) and fill the loops using a luer adapter. You can get these from Sigma etc.
http://www.sigmaaldrich.com/catalog/pro ... ch/z261637
then you need a NPT adapter to fit your GC tubing (e.g. 1/8" female NPT to 1/16" Swagelok). Much easier than squeezing bags.

[yuttingzheng]

I'm using a Tygon flexible tubing to connect air bags and the loop inlet (a 1/8" SS tubing). I also used the Air-Tite syringe (30 mL) but still the same thing happened. These aromatics are only 10 ppm in nitrogen and the heating box of the sample loop system is 160 degree C. Can this possible that these aromatics are absorbed?

Anyway, thanks for your suggestions!!

[AICMM]

yuttingzheng,

Have you tried making them in a metal can? I have made these components in the gas phase by putting a small amount of liquid standard into an evacuated can and then back filling with compressed gas. Works very well for a poor man's standard. I used metal tubing but it was all at ambient temperature.

Best regards,

AICMM

[yuttingzheng]

Not exactly the same. What you're talking about is sample preparation. In addition to use purchased standard gases, I also prepared standard gas by myself. I injeced 1 micro-Liter volatile comopund... e.g. benzene into air bags with 500 mL inner gas inside.

I think the sample itself is good. Manual injection gave good results. However, Inject using sampling loops gave no comopunds after benzene.

[Peter Apps]

There is a good chance that the missing compounds are boing absorbed by the Tygon tubing - you need to find another way to connect the bag to the valve, and use a syringe as Johnny Rod suggested.

Peter

[chromatographer1]

Fused silica coated SS or (more expensive and fragile) glass lined SS tubing will help your problems.

Bare metal tubing is NOT recommended.

best wishes,

Rod

[Johnny Rod]

How so?

[yuttingzheng]

We're still trying to figure out this problem. We changed to use SS tubing but it did not help. The Tygon tubing is very short (<3 cm) and is thought not the reason.

What we're thinking now is that the Liner caused this problem. The liner we're using is a regular liner. We will change to use a liner with really small volume (small I.D, for example, SPME type). The sample loop is using 1/16" SS tubing. Once the gas goes to the liner with much bigger I.D (regular liner) than the 1/16" SS tubing, the gas may go to split valve instead of column.

We're waiting for the liner... Could this help?

Thanks for all the suggestions!!

[chromatographer1]

How in the world are you putting sample into a sample loop valve through an injection liner?

You should have the sample source line directly plumbed into the valve to fill the loop.

Have you checked the proper connections for the valve and the proper rotation of the rotor within the valve?

Do you have the rotor installed 180 degrees (backwards) incorrectly?

GOOD LUCK,

Rod

[yuttingzheng]

Oh... seems I made you confused. Gas samples are injected into the valve system with 4 sample loops. The gas will flow through those loops and then to the ventilation. Once the valves are turned, those gases in the sample loops are carried into an FID injection port where the gas sees the liner, and to the column and to the FID.

I think the configuration is correct. But we cannot see compounds after Toluene. If we do manual injection we can see all compounds.

Thanks for your reminding~~

[Johnny Rod]

What are the gas flows and what are the valve timings? I don't see how the liner matters if the gas stream from the valve has to pass through the inlet port anyway - same as when you inject manually.

[chromatographer1]

The problem is missing peaks, and they are not getting onto the column (this is proved).

The most likely explanation is they are absorbed in a cooler spot somewhere along the sample path. Discrimination of injection seems unlikely. An error in plumbing the valve and valve rotor is possible, but only one on site can make that determination.

Figure out where in your sample path it is cooler than it should be, since you are using a separated heated box for the valves and loops and THEN you are passing the sample to the inlet ports SOMEHOW. There is a surface that absorbs the higher homologs you are sampling either due to temperature (physical chemistry) or chemical composition (chemistry).

I wonder, why you are doing a split injection using sample loops? when you could do a simple injection directly is beyond my understanding. I never did this in my work for what I hope are obvious reasons. But if you insist on using 0.25mm ID columns then I can see your reason for using a split injection. But that choice is yours and many would not follow you in that choice.

Without seeing your configuration not much else can be added that would help you. I hope you find your problem and can learn from the error.

best wishes,

Rod

[JI2002]

I recommend making the standard in a canister or cylinder. Squeezing a bag is not a good technique.

[chromatographer1]

I concur.

I wonder what temperature the bag is kept? And the chemical nature of the bag. (hint hint)

Evacuate a cylinder. Insert a calculated amount of aromatics by syringe into a line leading from a pressurized gas source. Connect the line to the cylinder. Fill the cylinder with a calculated amount of gas to achieve the desired concentration as vapor. Warm the cylinder adequately to avoid dewing the aromatics. Use cylinder as you would the mixing bag.

Good luck,

Rod