Chromatography Forum

My dear colleagues ;

I have a linearity problem with a autosampler GC method and its story is stated below ;

I have been asked to make partial gc method validation of a drug master file method which contains solvents as petrol ether(70-90°C), acetone, methanol,dichloromethane,acetonitrile. Methold conditions are listed below ;

GC HP 7890
Column : hp-wax 30mx0.53mm 1µl ( I used varian cp wax 52 cb and restek mxt-wax with same dimensions)
diluent : NNdimethylacetamide
GC autosampler
inlet temp : 140°C
1 µl injection volume
3 psi constant flow
splitless
oven program : 40°C 20 min 6°C/min 165°C 20 min
sample con. : 20mg/ml
solvents limits : petrol ether(70-90°C)50ppm, acetone 250ppm, methanol 300ppm,dichloromethane 300ppm,acetonitrile 200ppm

Due the bad peak shapes I made little modifications like
4:1 splıt , 45°C starting temp. and 3,6 psi constant flow

When I had started method validation a unexpected problem occured. Petrol ether and acetone were linear but rest solvents are nonlinear. Their linearity curve is more like Nike sports wear logo a little wave.

No matter how many times I trıed (different splıt ratios even injection volumes) again and again results did not change.
But when I increase the sample concentration like 2 fold and respectively solvent concentration and if I make analysis splıtless with awful peak tailing ,suprisingly they were all linear,
I have tried 2 different liner types, 3 different GCs, 2 different GC columns which were brand new but resuts are same.

Any comments will be highly appreciated.

Best Regards.

The first thing that stands out is the low inlet temperature - is there a specific reason to use 140C ?. If not increase it to 220C or so. Also the inlet pressure is low, which makes the flow through the inlet more vulnreable to disruptions during injection - increasing the pressure (if you are setting the carrier flow than increase that) might help.

Peter

As I stated above ; it is a drug master file method of an API so inlet temperature comes from there. According to their validation data they have no problem with linearity.
And again 2 of solvents (petrol ether and acetone) are linear in this circumstances. MEthanol, dichloromethane and acetonitrile are not linear.

Thanks and regards.

If you cannot change method parameters then I am puzzled as to how you expect to fix the problem.

Nonetheless - please post an example of the poor peak shapes and a non-linear calibration curve - instructions are in a sticky at the top of the GC page.

Peter

USP <621> allows for modifications to conditions. Don't repeat poor chemistry just because it's written somewhere.