GC 6850 peak tailing

[rb6banjo]

Is that methanol at 0.8 minutes in your chromatogram? Since the b.p. of EO is only about 11 °C, I think the peak identifications are correct.

[SmitaKamal]

I beleive it is DB WAX ...PEG.

Approximately at 3.4 is Ethylenne chlorohydrin that comes out. yes, at 0.8min it is methanol as EO standard mix with Methanal straight from vendor.

BP for EO is 58C
ECH is 130C.


I just changed the initial oven temp to 45C and looks likt eh USP tailing is pretty good for all concentrations 100ppm-1ppm.

[rb6banjo]

I think you're misinformed about the EO b.p. It would be liquid at RT if the b.p. was 58 °C. It is not. My old Lange's Handbook says 10.7 °C, my CRC Handbook of Chemistry and Physics says it's 13.5 °C (at 746 torr, 59th Edition), and Sigma-Aldrich says it's 10.7 °C. It's quite a bit lower than methanol.

[SmitaKamal]

Yes you are right. I meant BP for EO is 50F not C.

My bad. You seem to be an expert. Where are you based?

[rb6banjo]

In the US, Missouri. I used to work for Union Carbide in West Virginia. We made a lot of EO based products "back in the day".

[SmitaKamal]

Thanks for your help. Iam based in Texas. Union Carbide and i have long connection since 1984 when one of its plant exploded and took away many inoccent lives

[Peter Apps]

There is another complication - you mention both water and methanol as solvents. Neither of these are particularly method friendly, but water is by far worse than methanol.

Under the conditions that you give both of them can condense on the column and produce peak distortion due to solvent effects. Water can also damage the stationary phase. The solvent effects are likely to be worse with the very high gas flows that you use - seeing shoulders on the peaks is a classic sign.

I cannot see any benefit to gas flows that high (or none that cannot be more easily achieved by other means) so unless there is some definite reason (and not "that was how it was set up when I got it") for flows that high I think that your first step needs to be to reduce the gas flow to about 10 ml/min. You then need to give the FID 30 ml/min of fuel hydrogen.

What kind of liner is in the inlet ?, and when did you last do an inlet clean ?

Problems with solvent effects are most easily solved by split injections. What is the signal: nose for your peaks of interest under current conditions, and what is the lowest S:N that you can work with ?

Peter

[SmitaKamal]

I reduced the gas flow to 30ml/min and gettin shoulder peaks which I ma not getting when Flow was 179ml/min and same standard vial.

I have tried pulse method as well. Tried split and nothing will give me a better peak then at 179ml/min. I am in a method development phase therefor trying lot of combinations but nothing seems to be working except for already set method.

[rb6banjo]

EO should be well behaved chromatographically and somewhat immune to any problems with activity in your inlet. It's an ether. Is it possible that you have some acetaldehyde in the standard? I can't imagine why you would have shoulders unless you have coeluters.

[SmitaKamal]

Vendor claimed standard says EO in methanol and purity is 99.9%. Concentration is 50000ug/ml. Other standard I use is 2-chlorohydin in methanl and is 99.9% pure and conc is 2000ug/ml. Both come in 2 separate 1ml vial.I dilute both of them in HPLC grade water to get 100ppm in 10ml. So that in 1 shot both componds can come out.

I am not sure if there is any reaction taking place inside the vial.

[rb6banjo]

Doesn't seem likely. Maybe it's as Peter said, the water is causing trouble for the analytes to lay down nicely at the head of the column. Injection of water is usually something most of us try to avoid.

What if you dilute the standard in something heavier than the 2-chlorohydrin? Something like n-butanol?

[Peter Apps]

I reduced the gas flow to 30ml/min and gettin shoulder peaks which I ma not getting when Flow was 179ml/min and same standard vial.

I have tried pulse method as well. Tried split and nothing will give me a better peak then at 179ml/min. I am in a method development phase therefor trying lot of combinations but nothing seems to be working except for already set method.

But the set methods doesn't work does it ? - it gives you an unacceptably tailed EO peak, otherwise you wouldn't be posting.

The reason that your excessively high flow rate looks as if it is working is that the shoulders on the peaks are squashed into the peak because the peaks are so narrow in time. The "tailing" on the EO peak is actually a shoulder caused by solvent effects. You have painted yourself into a corner by using excessive flow to make it look as if a fundamental problem with injection conditions has been solved.

30 ml/min is still too fast. You need to use 10 ml/min so that you can address the solvent effect problems and get to a coherent set of conditions.

At "100 ppm" a 1 ul injection gives you 100 ng on column. Can you confirm this ? If you are mixing the two vials and topping up to 10 ml with water then you do not have "100 ppm".

Even with the poor FID response to EO you should be able to set a 10:1 split, and still have a very easily detected 10 ng on column. The split will reduce the quantity of water and methanol that you put onto the column, which should stop them condensing as droplets that disrupt the starting bands of the analytes.

So: use a 10 ml/min flow rate. Set a 10:1 split. Inject 1 ul. Do not change anything else.

Peter