Chromatography Forum

Hi folks,

I have a method for a FAME (fatty acid methyl ester) mix that is well established in our lab using helium carrier. Recently, we bought a couple new Agilent 7890 GC's and decided to set one up with hydrogen carrier for now, with the idea in mind that we will eventually be moving in that direction. The FAME method is one we run a lot and that I would like to get transferred over to hydrogen first. It's ideal because we still have to perform the validation, so we can demonstrate equivalency now, at a convenient time. We have no experience in working with hydrogen carrier, however.

I had some samples prepared and divided between the two 7890's, including a previously-analyzed lot, so that I could compare exactly the same solutions with helium and hydrogen carrier on the two new, otherwise-identical systems.

On the 7890 running helium carrier, the previously-analyzed sample gives exactly the same results as on our older 6890 systems.

For the system running hydrogen carrier, I tinkered a lot with it, and found that I could almost exactly reproduce the selectivity and resolution, but with half the run time. What seemed to work was just scaling the oven program (doubling the ramp rate and halving the hold time), and then playing with the initial linear velocity until a key reference peak eluted at the "expected" analog of its place in the helium chromatogram, at the "top" of the oven ramp.

The only problem is that in the hydrogran chromatogram, there is a lot of small "junk" close to the top of the oven ramp over a space of several minutes that is not present in the same sample run on the 7890 with helium carrier, or the original release chromatogram. The blank diluent injections do not contain this "junk," and the column compensation runs are extremely clean, but all of the FAME samples, regardless of origin, contain the same junk.

This is a deal killer for me, because my method requires me to identify a large number of very tiny peaks, and quantify by area percent. I'm thinking of changing the carrier to helium on this system and giving up for now because I'm very pressed for time, but I was hoping that if I put my problem out there, someone with more experience in working with hydrogen carrier might be able to tell me something to look for.

Thanks for any advice you can offer,

Stephen

Hi Stephen, I know that with mass specs that run H2 carrier gas you will see additional noise for several days as the combination of hydrogen, temperature, and metal surfaces acts like a self cleaning oven and cooks off trace materials. I've seen the baseline at install, and then after 3 or 4 days, and its a dramatic difference.

Because the junk is only present in the FAME samples it sounds as if you're getting some level of reaction between the H2 and the sample. As the system is used more the active sites where these reactions are occuring may get used up and the reaction may stop. I am not familiar enough with FAME/H2 carrier to say for certain.

I would give up for now and then when you have time to come back and try again. Its certainly doable, but you're right that there are method/application specific issues that will most likely need to be dealt with.

At least the run finishes in half the time, a nice reward if you manage to fix the problem .

Thanks for your advice. You're right, it is quite a nice reward, if (when!) I finally get it to work! Your assessment sounds right to me. I wonder if I might benefit from performing the Agilent-recommended inlet cleaning procedure with the gun brush, but on the other hand, maybe that would be counter-productive and what I really need to do is just blast it with FAME mix for a while to try and use up the active sites. As much as I am dying to play with it, I am so very, very pressed for time and I am probably going to have to do as you suggest and bail, going back to helium.

Hi Stephen, I know that with mass specs that run H2 carrier gas you will see additional noise for several days as the combination of hydrogen, temperature, and metal surfaces acts like a self cleaning oven and cooks off trace materials. I've seen the baseline at install, and then after 3 or 4 days, and its a dramatic difference.

Because the junk is only present in the FAME samples it sounds as if you're getting some level of reaction between the H2 and the sample. As the system is used more the active sites where these reactions are occuring may get used up and the reaction may stop. I am not familiar enough with FAME/H2 carrier to say for certain.

I would give up for now and then when you have time to come back and try again. Its certainly doable, but you're right that there are method/application specific issues that will most likely need to be dealt with.

At least the run finishes in half the time, a nice reward if you manage to fix the problem .

At a recent Agilent seminar they said that it takes about 2 weeks of frequent injections to clear out the additional background that results from a switch to hydrogen, which acts as a very efficient cleaning agent to scour all the little areas the helium did not penetrate to.

You can certainly decrease run time with hydrogen (30% reduction in time is typical they said) as the optimum in hydrogen's Van Deempter equation occurs at a higher gas flow, with higher effective plates, so the sample moves through the column faster.

Agilent has a method conversion app that is designed to help you convert from helium to hydrogen. This can be downloaded from their website. Your local Agilent rep can likely send it to you.

You can also decrease run time with Agilent's smaller bore columns (we are about to go from 0.25mm to 0.18mm ourselves). For added speed (at some cost) look into their backflush capability (part of their Capillary Flow Technology). We will be adding backflush to our Agilent GCMS to reverse the flow as soon as our last FAME peak leaves the column so we do not need to wait another 20 minutes for the cholesterol peak to go through.

Performing the Agilent recommended inlet cleaning procedure turned to to give dramatic improvement, and time and injections took care of the rest. Works great now - run time cut in half for this method!