Can He or H2 be used as carrier gas for ECD?

[Anthony_Ng]

Hi all,

We are using N2 as carrier gas for ECD.

I know that He and H2 can be used in GC-MS for their lower "viscosity" and hence have a better resolution. What about ECD?

And what is the possible choice of carrier gas for other detectors? (e.g. NPD, FID......)

Thanks in advance.

[fsistere]

Hi
I use He as carrier gas with ECD since... ... and there is no problem. The make up is N2.
I use too He with FID and NPD.

[AICMM]

Anthony_ng,

I have never used H2 with ECD so I will be interested to hear if you try this. I have very often used He with ECD and you will probably find overall better performance with this as carrier over N2. The problem nowadays seems to be the price of helium.

Regarding FID and NPD (and FPD for that matter), H2 would be a very good choice of carrier for these detectors. Keep in mind that you would decrease the detector reactor gas accordingly with the addition of hydrogen as carrier.

Best regards.

[Ron]

The use of helium carrier depends on if you are using a packed column or a capillary column. If you are using a capillary column, there will be no problem with helium as carrier gas as long as you use nitrogen as the makeup gas. If you use a packed column using helium will probably result in a significant reduction in sensitivity, and you will have to use nitrogen as a makeup gas.

[GasMan]

I have found that if helium is used without nitrogen being present, as when using packed columns, we got a full scale signal for baseline, which suggests to me that the helium was ionising. So I would say NO for packed columns, but as mentioned in other posts, you can use with capillary providing that nitrogen is used as makeup and purge gas.

Gasman

[lmb]

I know that He and H2 can be used in GC-MS for their lower "viscosity" and hence have a better resolution

Dear Anthony_Ng,

The following info might be useful.

1. Viscosity of He is not lower, but slightly higher than viscosity of N2. In relation to viscosities of He, viscosities of (H2, N2) are (0.45, 0.9). In other words, of the three gases, (He, H2 N2), He has the highest viscosity. Viscosity of H2 is 45% of viscosity of He, and viscosity of N2 is 90% of viscosity of He.

2. Carrier gas has no effect on resolution. Take any capillary column analysis with N2 as a carrier gas. Replace N2 with He or H2 and translate the method parameters using Agilent’s GC Method Translation Software (available free of charge at http://www.chem.agilent.com/cag/main.html#mxlator). Resolution of ANY peak pair will be almost identical in all three cases.

3. Analysis time is a different story. H2 is the fastest carrier gas and N2 is the slowest. The difference in the analysis times depends on the degree of the gas decompression along the column.

3a. In GC/MS (vacuum at the column outlet) and in all other cases where inlet pressure is much larger that outlet pressure (long, small-bore columns with outlet pressure at or below 1 atmosphere) analysis times of (H2, He, N2) relate approximately as (1, 1.7, 2.9). Thus (while providing exactly the same resolution in all three cases), the times of analyses with He and N2 would be 70% and 190% longer than the analysis time with H2.

3b. When gas decompression along a column is small (short, wide-bore columns with outlet pressure at or above 1 atmosphere), the analysis times of (H2, He, N2) relate approximately as (1, 1.25, 4.2), i.e., (while providing exactly the same resolution in all three cases), the times of analyses with He and N2 would be 25% and 320% longer than the analysis time with H2.

Hope that you find this helpful.
lmb

[Anthony_Ng]

Thanks for all your info and suggestions

Yes, I am using a capillary column and ECD as the only detector at this moment. N2 is used as carrier, make-up and detector gas.

Recently I saw some articles saying that if I use He or H2, according to the Van Deemter equation/curve, I can get a lower plate height (H) than N2, even with faster flow rate. Does it imply a better separation and faster run time?

Another issue is that: if I use a capillary column with smaller id while keeping phase ratio unchanged, can I get a better separation?

Thanks again for correcting me the viscosity of He>N2>H2

[Don_Hilton]

Yes, if you use a smaller ID with the same phase ratio you can get a better separation, but the loading for the column is less. So, if you have a lot of matrix in the injection, you may have to cut back on injection size.

[GC_User]

Yes you can use H2 as the carrier gas for an ECD for a capillary column. It's cheaper than He and you can get faster analysis times. H2 also works as the carrier gas with FIDs no problem. If you set up multiple instruments to work with H2 as a carrier gas you might want to look into purchasing a H2 generator if you don't have one already. It's cheaper in the long run than constantly buying cylinders.

[lmb]

Anthony_Ng: "He or H2, according to the Van Deemter equation/curve … get a lower plate height (H) than N2, even with faster flow rate."

More relevant to the capillary columns that we discuss is Golay formula/curve based on Golay’s solid theoretical work. (Van Deemter formula/curve results from semi-empirical study of packed columns.) According to Golay formula, carrier gas has NO EFFECT on the lowest plate height (Hmin) that can be obtained from a column. As a result, carrier gas has NO EFFECT on the best obtainable resolution either.

Speed of analysis is another matter, as I mentioned in my previous posting. Carrier gas flow rate at which the plate height is the lowest (i.e. H=Hmin) is known as optimal flow rate (Fopt). Fopt of H2 is 1.25 times higher than Fopt of He and 4.2 times higher than Fopt of N2. Higher Fopt together with its lowest viscosity makes H2 the fastest carrier gas.

I again forgot to mention another advantage of H2. Peter App brought it up in a previous set of postings. Peaks out of a column with H2 are the sharpest and the tallest. Therefore, H2 offers the lowest (the best) detection limits.

Anthony_Ng: "if I use a capillary column with smaller id while keeping phase ratio unchanged, can I get a better separation?"
Don_Hilton: “Yes, if you use a smaller ID with the same phase ratio you can get a better separation.â€

[Peter Apps]

Hi Anthony

If the detector uses hydrogen as a fuel (e.g. FID, NPD, FPD), and the flow through the column is a significant fraction of the fuel gas flow, as it will be with 530 um columns, you can get serious baseline drift because the volume flow rate through the column decreases during a temperature programmed run if the inlet pressure is constant. Most (all ?) modern GCs allow you to set a constant volume flow (ml/min), which solves this problem.

Peter

[Anthony_Ng]

[/quote]

More relevant to the capillary columns that we discuss is Golay formula/curve based on Golay’s solid theoretical work. (Van Deemter formula/curve results from semi-empirical study of packed columns.) According to Golay formula, carrier gas has NO EFFECT on the lowest plate height (Hmin) that can be obtained from a column. As a result, carrier gas has NO EFFECT on the best obtainable resolution either.

Oh! Probably I misused the concept of fast HPLC, simply plug the concept into fast GC. (i.e. Using a smaller particle size, smaller id and shorter column, it will have similar resolution but faster run time)

Back to GC, if I switch to a smaller id, shorter and same phase ratio column already, can I go a step further to change carrier gas from N2 to He or H2, making it a higher linear velocity, so that I can speed up the analysis? (But keeping the separation/resolution comparable to the larger bore column with N2 carrier gas)

By the way, is there any suggestion on books/articles/web sites that I can re-learn the topic something like "speed up from standard GC to fast GC"? Thanks again for all of your advice!!!

[Peter Apps]

Hi Anthony

The simple answer to your question is "yes" - make the changes that you list and you will get the same analysis done in (much) less time. Beware though (as has been already mentioned I think) that you cannot put as much sample onto a narrow bore column as one to a wide bore one without running into problems with concentration and/or volume overloading, and the proper operation of the injector and inlet become more critical as column diametrer decreases.

Good luck with trying it.

Peter

[lmb]

Anthony_Ng:

is there any suggestion on books/articles/web sites that I can re-learn the topic something like "speed up from standard GC to fast GC"?

Try this: M. Klee and L. Blumberg, "Theoretical and Practical Aspects of Fast Gas Chromatography and Method Translation", Journal of Chromatographic Science 40 (2002), 234-47.

[Victor]

Lionel,

In liquid chromatography the viscosity of the mobile phase has a profound effect on the diffusivity of sample molecules, which is higher in liquids of low viscosity. Clearly from your arguments, this is not the same in gases.

What is the fundamental physical property of gases which affects the diffusivity of the sample molecules? I presume this is the density of the gases. Can you offer us any more information on this subject?