Page 1 of 1
Poor injection reproducibility
Posted: Wed Jun 28, 2006 11:56 am
by mk65
We are running a terpenoids GC method on an Agilent 5890. For years the method has been great but recently we are having probelms with poor reprducibility (>10%rsd cf <2%rsd). Any suggestions as to where to start on trying to find the route cause of the problem?
mk
Posted: Wed Jun 28, 2006 12:40 pm
by leadazide
Maybe this is to basic but have you tried changing the injection needle and/or septum? I have had trouble with injection RSD% myself because of these factors.
It could also be the linier in the injection port that is either (very)dirty or broken.. I have earlier broken liniers in the injection port by tightening the septum knot too tight... When the manual says finger tight... they mean ONLY finger tight
Hope this helps.
Posted: Wed Jun 28, 2006 2:16 pm
by DR
Syringes wear out too.
Posted: Wed Jun 28, 2006 3:19 pm
by Peter Apps
It may not even be poor injections - there might be a leak at either end of the column, or a problem with the detector such as dirty electrodes of erratic gas supplies.
Could also be that the splitter line is clogging up with sample condensate.
Are you measuring peak area, relative peak area, or is the variability in the final result ? If in the final result, how are you calibrating ?.
Is anything else changing - baseline drift, retentions, peak sizes ?
Peter
Posted: Wed Jun 28, 2006 3:25 pm
by Consumer Products Guy
I tend to agree with Peter, and agree: don't overlook the split line. We've had to clean these, or replace with 1/8 inch tubing.
Split line
Posted: Wed Jun 28, 2006 8:58 pm
by ingochrist
We just had a similar problem and with cleaning the split line with MeOH helped a lot.
Posted: Mon Jul 10, 2006 11:16 am
by mk65
Thank you everyone for your help! We have been through most of your suggesting and one final one we came up with was a solvent issue. Has anyone else every had a problem with DCM affecting the response on FID? We have used the same supplier of DCM for the passed few years but it seems the likely cause at present. If we prepare our samples/standards in chloroform instead the %rsd observed is acceptable (<2%), but in this one batch of DCM it's >7% rsd.
Regards
mk
Posted: Mon Jul 10, 2006 3:52 pm
by Matt Savage
Read the contamination section of the link below....
http://lcgceurope.adv100.com/lcgceurope ... &&pageID=6
To be honest though we've been using CHCl3 with no problems of soot in the FID.
Matt