Chromatography Forum

Hello everyone, I am a graduate student with little background in GC. I am trying to separate IPA and acetone (100 mM and 1 mM in MilliQ water correspondingly) using our lab's Trace 1300 GC with a HP-INNOWAX (20 meters, 0.180 mm diameter and 0.18 film thickness, the only polar column I could find in campus from another group since funding situation is not good). Two days ago I conditioned the column as I noticed some background ghost peaks since other student had to use it for other samples (not IPA or acetone). I turned the purge flow to 1.4 ml/min with the oven at 35 C for 20 minutes, and later I increase the temperature to 240C (temperature limit of column is 260C) for 2 hours. After that conditioning I have been constantly seeing my peaks of acetone and IPA not separating anymore. Before conditioning it had a nice 1.1 minute clearly defined separation (Acetone 2 min, IPA 3 minutes) but now both appear at about 1.5 minutes where the acetone peak is at the shoulder appearing at earlier time than IPA, but is not that well separated anymore. My question now is whether the column can be fixed, as my only conclusion is that the conditioning in fact made worse the polarity of the column. Also is there a quick way to test it is indeed because of the column being the problem? I read somewhere we can do some polarity test mix but my PI is on vacation and he is the only one able to make chemical orders. What is your suggestion? Unfortunately we are the only group who works the most on GC, I tried asking labmates, friends, other groups and technicians but no one knows very well or they mention they know less about GC than our group does.

Method used: 220 C SSL inlet with split ratio 11, and constant flow of 1.800 ml/min, Oven starts at 35 C for 2 min and ramp to 125 C at 20C/min while holding for 2 minutes. FID temperature 250

The fact that you can't separate your analytes is enough of a test mix to say that your column has been compromised. You used to be able to separate them and now you can't. I'm not optimistic that it can be repaired/regenerated.

Make sure that you don't have oxygen in your carrier gas. It is public enemy #1. Constant exposure of your stationary phase to oxygen - especially at elevated temperatures - will kill it.

I would recommend a leak detector before conditioning a column. They might have swapped yours out and did a poor job reinstalling it. That column is cooked. If the random peaks were outside your target RT windows I'd just ignore them in the future.

I suspect that you have damaged the column, but you may be able to salvage something.

I presume that you are injecting 1ul of the water samples.

That gives you about 6ng of isopropanol on the column, which is plenty for an FID to detect.

First step is to cut 2m off the inlet end of the column and re-install it. Any damage from the conditioning, (and from injecting water) will be concentrated at that end.

Then reduce your carrier gas flow rate to give a linear velocity of 30cm/s if you are using helium, and 50cm/s if you are using hydrogen (15 cm/s in the unlikely ecen you are using nitrogen). There are flow and pressure calculators online - the column and GC suppliers websites have them.

Increase your split ratio to 20:1; that will halve the amount of water your are inflicting on the column, but you will still have an easily detectable 3 ng of isopropanol.

Peter

I presume that you are injecting 1ul of the water samples.

Water expands A LOT in the inlet. We never injected more than 0.5 microliters when water was our solvent, like with hand sanitizer samples, and we used a wide bore capillary and higher carrier flow.

Increase your split ratio

And/or do this.