Shouldered Ethanol Peak?

[JaceAssured]

My partner and I are new to the gas chromatography world and thus, we are having some trouble getting one of our GC methods to run properly.

We are currently trying to make a calibration curve and our standards consist of ethanol, water, and n-butanol (internal standard). We are operating a 7890B Agilent GC with an ALS. It contains a HP-5 general use capillary column [length=30m , ID= 0.320mm], and has an split/splitless injection.

Our method parameters are:
Injection Volume = 0.5 uL
Inlet Temp = 220 C
Inlet Pressure = 7.5492 psi
Split Ratio = 100:1
Oven Temp = 50-100 C with a 7 C/minute ramp
FID Temp = 250 C
Air Flow = 400 mL/min ; H2 Flow = 30 mL/min ; Make-up Flow = 25 mL/min ; Carrier Gas Flow (He) = 1.5 mL/min

When running our standards we observe a split/shouldered peak for ethanol but the butanol peak is sharp and repeatable. We are unsure of what is causing our ethanol peak to split/shoulder while our butanol peak remains a single sharp peak. When testing 99.5% Ethanol, the shouldering does not occur. Can anyone please help?

[JaceAssured]

Here is a photo of one of our chromatograms.

[Peter Apps]

With a 100:1 split solvent effects are unlikely, and so this is probably due to flow disruptions from the pressure pulse in the inlet as the water flash vaporizes.

For GC in general is a poor choice of solvent, avoid it if you can.

If you are stuck with water as the sample matrix then you need to play with injection volumes (you can risk going to small volumes because the internal standard will correct for volume fluctuations), inlet temperatures, and carrier pressure (but try to keep close to pressures that give optimum flows). Even a change of inlet liner design might help.

Peter

[Jesse Robinson]

As someone said, water is a poor solvent for GC. If possible you should try a wax column such as the DB-Wax. It can better handle the water without damaging the phase. One more thing you could try would be to slow your injection speed down. I had the same problem with Methanol. Slowing the injection speed down has fixed this issue for me.

[amazhad]

Using one of the Agilent solvent vapor vent calculators with your method parameters, a typical split liner (Agilent P/N 5183-4647) is filled to 85% capacity during the vaporization process. Water has a very large expansion co-efficient compared to most other solvents and the problem might be related to overloading the liner.

It is recommended to keep the expansion to under 75% of the volume of the liner you are using. You can either adjust the inlet temp down, decrease the injection volume, get a larger volume liner type or switching to a pulsed split method with an elevated pressure.

I noticed both peaks are tailing which can indicate a poor column cut or installation. This can also cause split peaks due to gas swirling at the bottom of the inlet. I would recommend removing the column from the inlet, re-cutting it and re-installing.

[Bigbear]

If you are injecting water the inlet @ 220 is too high. Try it @ 125.

[Consumer Products Guy]

Ethanol does not have a great peak shape on HP-5 type columns, those are not very polar. But you likely can make it work, we used HP-5 for ethanol in a validated method for many years (our non-analytical scientist supervisor was concerned that we could find any IPA contamination, because once before start up the facility had used IPA to clean a line).

[Klaus I.]

our standards consist of ethanol, water, and n-butanol (internal standard). ... It contains a HP-5 general use capillary column

What exactly are you trying to do?

[JaceAssured]

We are trying to measure ethanol concentrations in aqueous samples. So far we have tried adjusting inlet temp, carrier gas pressure, injection volume, and basically all parameters and nothing seems to help.

We now believe that the issue is not method related and believe that water expansion is not an issue. We ran several 100% EtOH samples from different stock solutions and the double peak still appears. This may be due to poor column installation, as someone had mentioned earlier. We are going to re-cut the column and re-insert it into the inlet to see if that helps.

[Consumer Products Guy]

[quote="JaceAssured"]We are trying to measure ethanol concentrations in aqueous samples. quote]


Use a PEG column !!! Make your life easier. We do this assay all the time. Use a PEG column, keep your injection volume low. We've used n-propyl alcohol and ACN as the internal standards, most common ones for this in industry (unless you're measuring for alcohol in dead bodies or drunk drivers or something).

Make your boss get the cobwebs out of his wallet !!!

[Klaus I.]

We are trying to measure ethanol concentrations in aqueous samples. So far we have tried adjusting inlet temp, carrier gas pressure, injection volume, and basically all parameters and nothing seems to help.

We now believe that the issue is not method related and believe that water expansion is not an issue. We ran several 100% EtOH samples from different stock solutions and the double peak still appears. This may be due to poor column installation, as someone had mentioned earlier. We are going to re-cut the column and re-insert it into the inlet to see if that helps.

Maybe the column-type is the problem? Try a 1301 or 1701 type designed for organic solvents.

[Bigbear]

When re cutting the column change the gold seal and liner as well. Colummn end 5mm up from the tip of the ferrule.

[Klaus I.]

When re cutting the column change the gold seal and liner as well. Colummn end 5mm up from the tip of the ferrule.

The butanol peak seems to be ok. Nevertheless retention time is absurdly low.

If you are injecting water the inlet @ 220 is too high. Try it @ 125.

Most properly I'm meanwhile to far away from gc/ms for some years, but can you please describe in detail this recommendation?

[Consumer Products Guy]

What are you using as your ethanol standard? Even HPLC grade can have 5% methanol and 5% IPA in it, you need to use 99.9% ethanol as the standard.

[osp001]

What are you using as your ethanol standard? Even HPLC grade can have 5% methanol and 5% IPA in it, you need to use 99.9% ethanol as the standard.

I was going to recommend running a sample spiked with methanol, see if the peak gets larger. Just a hunch.