injector capabilities

[pstaunton]

currently running a method that requires a 0.5ul injection volume with a 5ul syringe.

the standard I inject contains heptane and toluene in dichloromethane at concentrations of around 0.6ppm.

the problem is my system suitability fails for %rsd on the three standard injections. two areas will stay consistent and then one will decide to jump up or down usually giving me an rsd of above 10%.

this is a validated method so please don't ask me to increase injection volume or anything like that as I would not be allowed.

running on an agilent 6890N in splitless mode,
split time 1.5min, 25ml/min
7ml/min He constant flow,
injector temp is 245
detector temp is 300
program: 35 for 8min, ramp 15C/min to 290C, hold 25min
column is DB-5 30M x 0.53umx1um film thickness

just to point out the heptane and toluene peak areas will both change to the same degree.

any ideas why this happens or could be improved would be appreciated.

[tkubowicz]

Hello

I see 2 possible problems:

1. Leak in GC:
-check inlet for leaks. Check split vent trap connection/septum.
-check syringe for leaks. Replace it if it is not tight enough

2.GC method. I'd change purge time to 0.5-0.7 min and increase split flow to 40-60ml/min

Regards

Tomasz Kubowicz

[Peter Apps]

You might be getting solvent condensing in the column - 35C is just below dichlorormethane's boiling point (at atmospheric pressure).

But since you can't change the method, how does it help to know the cause of the problem ?

Peter

[pstaunton]

also to say i have tried this method on three different agilent systems in the lab and they all give the same %rsd failures.

sometimes this method can be run for weeks at a time without these issues and then all of a sudden it is gaurenteed to fail.

[tkubowicz]

Hello

I'd check method setting for injector:
-number of "pumps"
-filling volume
-drawing/injection speed
-washing solvents
Perhaps there is air bubble in syringe and that is why injection volume is slightly different every time.
You have this problem for 3 GCs so it is very uncommon that there is leak or liner problems for all GCs at the same time.

And also I'd review inlet setting and oven temp for your method. Even if it is validated method it is worth to try.

Regards

Tomasz Kubowicz

[James_Ball]

Are you using teflon tipped gas tight syringe or normal stainless steel plunger?

[pstaunton]

using an agilent stainless steel plunger syringe.

on the flipside my run decided to give nice consistent results in the meantime.

luck should be the last system parameter given in the procedure.

[Peter Apps]

You are putting 0.3 ng on the column. This is enough for an FID to see, but you are in the region where baseline noise, either short or long term will make a big difference to integrated area.

Check where the integrator is putting peak starts and ends, and is it consistent ?
Peter

[pstaunton]

hi peter,

we integrate manually for our GC assays and i have ensured that all integrations are consistent.

when i overlay my working standards its clear that one of the standard injections has either gained or lost alot of area somehow.

just to add i run at a sampling rate of 5.0, is that enough data points for this method?

[Peter Apps]

How fast you need to sample depends on how wide the peaks are - you need at least ten points across the peak, and the more the better. There is no upper limit to sampling rate with an FID.

Have you checked for leaks etc as suggested by other answers ?

Peter