carryover of organic acids by GC

[djletin]

I have been analyzing a series of monocarboxylic acids and naphthenic acids (C11-C20) by GC-FID using a Perkin Elmer Autosystem with Programmable Splitless (large volume 20-30uL) injections.
The solvent is 1% formic acid in methanol. Direct analysis with no dervitization using an Agilent PEG column.
concentration range is ~ 1 ppm - 40 ppm.

I am having signifcant problem with injection-to-injection carryover. Blank GC runs (no solvent injection) are clean.
Checked cleanliness of solvents but they too are clean. Have switched out syringe and inlet liner and that also is not issue.

Also observing "priming" phenomena where I need to do a number of injections of same sample or standard to achive reasonably constant response.

[Peter Apps]

What about the septum - did you change that ?

Was there carryover into the very first injection with a new inlet liner ?, and with the very fist injection with a new syringe ?

What kind of inlet liner are you using ?, and how is it deactivated ? 20 - 30 ul is a very large volume, and without fancy operating conditions and special liners there is a good chance that sample is going in all directions and subsequently bleeding back. What inlet programme in terms of temperature and flows are you using ?

I presume that 1 ppm is 1 ng/ul, so why are you using such large volumes ? You should be able to see 1 ng as a good clean peak. Doing direct injections with a siltek deactivated Unilier and a Restek PEG column is get peaks for 0.2 and 0.1 ng of free acids at the chain lengths you are working with.

Peter

[chromatographer1]

Peter is correct in noting that the vapor cloud of the injected sample can go into places where it is not intended to go, especially if the cloud produces a volume beyond the capacity of the inlet design. This is called backflash. A thorough cleaning or replacement of the inlet pneumatics is a solution. Slow speed insertion of the liquid sample into the inlet can minimize this problem.

The other issue is that you are using a PEG column, not a FFAP column. PEG has a 35% basic character if I remember my course work from thirty years ago correctly. So losses (followed by ghosting) of strong organic acids at trace levels is not surprising.

This of course, does not begin to describe the possibility of basic deposits on the column from previous injections.

Peter other comments are well said and I need not repeat them here.

best wishes,

Rod