CO2 peak area changing with balance gas

[mdGC2347]

Hello,

I am having a strange problem and I can use any help I can get... My company uses calibration standards for our various GC's and I realized that the gas standards that we are receiving have inconsistencies in the CO2 peak.

For example:
When I receive a cylinder with 10% CO2 in balance N2, I can calibrate my GC to 10%
If I then use a cylinder with 10% CO2 in balance H2 and CH4, the peak area is larger, and my concentration comes out to be closer to 13%!

At first, we believed it was an issue with the gas cylinders, but I have since tried multiple gas suppliers and find this same trend.

I cannot find any information online that says that H2, CH4, or N2 can impact the CO2 reading, but has anyone else ever experienced this or have any ideas?

For reference, we are seeing this on 3 different GC's all of different types of Agilent Micro GC with both He and Ar carrier gases.

One of the GC's in use is a 490 Micro GC using a PPU column with a TCD detector and He carrier gas. There is excellent separation between peaks.

The plot then thickens when our gas providers show us that they use a GC with TCD, but only using a packed column instead and they don't see the inconsistency.

Thank you for any help!! This is driving me crazy!

-Matt

[chromatographer1]

Your MicroGC injection system is quite different from a fixed loop packed column no split injection system GC.

You are getting more sample injected with a Micro GC injector with a H2 C1 balance than with a N2 balance.

How do I know? Your CO2 peak is larger when you inject a tested standard.
Detectors don't lie. Now, why is the question. It is not a mechanical defect, it an effect of physical chemistry.

Look at the physical conditions of the injector and figure out why the balance gas makes a difference. The packed column GC uses a fixed volume loop where the sample is equilibrated to atmospheric pressure.

Does your MicroGC do that? What gas conditions exit when you inject using the MicroGC? Another hint: Why are the gas equations called Ideal Gas Laws? What gases are known to be Ideal?

best wishes,

Rod

[mdGC2347]

Rod,

Thank you for your response. You brought up a great point...I had not considered the non-ideality of hydrogen interactions with CO2 could allow for more CO2 molecules to be injected to the detector.

I will spend some time today looking into the differences between the MicroGC injection system and the fixed loop packed column no split injection system, since I am not very family with the inner workings of either. If you could point me in the right direction of a good website (or had some time to give a brief explanation), I would appreciate it. My injection settings are simply temperature (I use 110°C) and Injection time (I use 40ms).

I have primarily worked with MicroGC's and I had never been aware of this issue. Additionally, when I asked Agilent about this, they said they had never come across this problem and that it must be the gases. Is this a known problem with MicroGC's in the industry? Have you encountered this yourself?

Thanks again,

-Matt

[mdGC2347]

I have been doing some research on the injector types and the gases and I am getting stuck on the overall premise of the boosted flow rates through the timed injector into the detector due to the presence of these gases. Hydrogen is considered just as ideal of a gas as N2 at STP. Not only that, the compressibility of H2 is very comparable to N2 at conditions even more extreme than ours. If the issue was the presence of methane instead, methane its still quite ideal at our pressure and temperature, but if anything is less compressible, and therefore would result in the opposite effect and I would be getting less CO2 in that mixture.

At 130°C (my injector is set at 110°C) and 5 bar (my injector is at slightly higher than 2 bar) the compressibility factor (where 1 is ideal) of the gases are:
H2 = 1.003
N2 = 1.001
O2 = 1.000
CH4 = 0.998
CO2 = 0.991
Ar = 1.005

If this was caused by intermolecular forces between either methane or hydrogen and carbon dioxide, giving the mixture a higher molar density, then this would be seen in the fixed volume injector as well, which it is not.

Can you help me see what I am missing? Any help is appreciated!

-Matt

[plantrob]

This issue was recently encountered in my lab as well. The key is in the notion of an "injection time" - the micro-GC injects not the full contents of an injection loop, but allows sample to flow for a fixed amount of time to define the sample volume. The amount that flows depends on physical properties of the gas being sampled, most notably viscosity. Hydrogen has a different viscosity than nitrogen.