Maybe this is a dumb question but,the headspace vial is pressurized w/ He prior to injection so, obviously some is injected with the components in the headspace is it possible that sometimes less He and therefore more of the componets are injected. i am getting area counts on the occasional standard that are 1.5 time the rest of the sample set (8 repilcate injections of a standard) there is no carryover seen in my blanks and in the most recent case the jacked up number was injected immediately after a blank. any comment will help Thanks
By Rodney on Wednesday, February 19, 2003 - 06:15 am:
Reproducible headspace analysis requires that sealing and heating be very consistent.
Any leaks or preparation errors will be visible in the analysis of standards.
Since you are seeing these errors you need to find the source of these errors.
I have experience with both fixed loop and timed injection HS analyzers. Feel free to discuss your situation with me by phone or email.
Senior Research and Development Chemist
Gas Separations Research
595 North Harrison Road
Bellefonte, PA 16823
By Anonymous on Thursday, February 20, 2003 - 01:34 am:
Headspace troubleshooting usually begins by taking a critical look on your preparation steps. The most important (but not all) causes for poor reprodicibility are:
- Sealing the vial tightly (but not too tight, so that the seal becomes crumbled)
- Heating procedure (condensation !!)
- Hardware problems such as varying gas flows
- GC injector (Liner, split line)
- Heating of the transfer line
and many more..
hope that helps
By Anonymous on Thursday, February 20, 2003 - 09:13 am:
I have always preferred internal standard methods for headspace analysis to compensate for small variations in response due to vial sealing. If there are major variations, you need to address the basics of the method as noted above.
By Anonymous on Wednesday, February 26, 2003 - 07:59 am:
Thank you to everyone who posted i have figured the problem out. the transfer line was bent at a severe angle resulting in restircted flow into the inlet. This restricted flow was causeing the sample to be introduced in much too wide of a plug. Ths was sending more sample out the split vent and giving me very broad peaks. My areas are now consistently what they should be in the range of what i thought were the abnormal counts but, were in fact the area i should have been seeing and the peaks are much sharper.
By Rodney on Wednesday, February 26, 2003 - 09:09 am:
Thank you Anonymous for giving us the end of the story. Many times posters forget to do that. It is very appreciated. thanks again.