Page 1 of 1
High RSD in headspace GC
Posted: Fri Jul 12, 2013 1:59 pm
by MestizoJoe
I'm testing ETO which is prepared in water near 20 ppm. I've been getting high RSDs since I developed the method (~3%). Now I'm getting 4-7% RSD.
I thought that I was not crimping tight enough on the headspace vial caps, but crimping tighter didn't solve the RSD problem.
Some notes on the method:
HS oven temp 90C loop 105C transfer line 110 inlet 200C
What can be done to improve thits? Could it be as simple as replacing a gold seal and liner?
Re: High RSD in headspace GC
Posted: Fri Jul 12, 2013 4:13 pm
by MestizoJoe
I need to mention also that The RSD is fine for the less volatile sample which is ethylchlorhydrin (ECH). The RSDs for ECH have been consistent and always below 5%.
I thought the caps were not crimped tight enough since only the more volatile ETO has high RSD.
Re: High RSD in headspace GC
Posted: Sat Jul 13, 2013 3:07 am
by chromatographer1
EtO is quite reactive. Are your numbers high or low?
I can see you need to have temperatures somewhat elevated for the ECH, but why the high inlet temperature? Would 120C be adequate? But then I am used to not using an inlet and I never had this issue. I never used a splitter and my results were always good when I left that variable out of the equation.
But I would replace the inlet and any other sites you might consider to be reactive.
best wishes,
Rod
Re: High RSD in headspace GC
Posted: Sat Jul 13, 2013 4:42 am
by MestizoJoe
Thanks for the reply Rod.
The numbers are correct if I take the average. I observe that after using my check standard. But they vary. I don't believe it decomposes but I may be wrong.
I'll change the liner and gold seal.
A newly realized variable that might be causing problems is the lab temp during sample prep. Been a hot few days here in the G and our HVAC system isn't too notch. I'm going to do sample prep in a better controlled environment from now on.
As for the temperatures I use in the method... When the method was developed my skills were novice. The conditions are not optimal. However during method development I saw good reproducibility.
Have a great weekend. And keep it classy Kansas City.