Reproducibility problem

[tabra]

Hello Chromatographers!

I need your help! I have some experience in LC, but total beginner in GC.
Working on Thermo GC Ultra, TriPlusAutosampler, GC/FID

Column - TR-5, 7mxi.d.0.32, 0.25microm
Start Pressure - 19.2 kPa (temp - 150 degree C)
Inj. volume: 1 microL (phthalate esters, methilen chloride solutions)
Splitless - first 0.8 min, then split 1:60
Carrier gas: nitrogen - flow - 1mL/min
Do not know if I forgot something...

Same vial, 10 times for standard solution - very bad reproducibility! Do you have idea where am I going wrong?

Thanks[/i]

[tabra]

It is peak area reproducibility. RSD% is about 10% using autosampler. Very bad, is not it?

[Schmitty]

One suggustion: Don't use the same vial with DCM as the injection solvent, as it is prone to rapid evaporation.

Are the peak areas increasing as time goes on?

[tabra]

There is no rule, sometimes peak areas increase, sometimes decrease... Obviously, DCM rapid evaporation is not the only problem.

[chromatographer1]

The most common problem is removing the sample from the vial consistently when the vial is very tightly sealed. It doesn't work the DCM sample may pressurize the vial if the lab is warm.

suggestions:

You may be overloading the injection liner volume. Reduce the size of your injection. If you have backflashed sample into your pneumatics it may take several injections of blank solvent to rinse them clean.

Then start injecting samples again at a smaller volume.

Loosen the vial cap (slightly) so the vial can be at a constant ambient pressure (without loss of vapor through evaporation, of course).

The other issue could be column installation position in the injection port.

Good hunting and best wishes,

Rod

[JI2002]

A few suggestions:
1) 7 meter column seems short, especially if the i.d. is 0.32 mm.
2) Why not use H2 or helium as carrier gas?
3) starting temp seems high, try initial temp at 50c.

[tabra]

Thank you very much for useful advice. I'll try to solve the problem - changing parameters one by one....

Er... I had problem with reproducibility evan when I was worked with 60m column. Analysis was time-consuming so I decided to try with test column – 7m to figure out what is going on.

[Schmitty]

Maybe you can't tell us, but if you want to let us know what your compound is, that may help identify a reason why it is so variable.

One further suggestion: Did you replace the inlet liner? What type are you currently using? Is it perhaps activated or has become activated based on a property of your compound of interest or matrix present in your injection solution?

[tabra]

Of course, I can. We are working on development of in house method for phthalate esters determination in toys. Matrices are very complicated and glass liners (high deactivated borsilicate glass - Trace&Focus 5 mm ID SSL - 105 mm x 8 mm OD) become dirty again and again, so I change it very often.

[chromatographer1]

I would change solvents and see if that makes a difference.

10 percent variation is high but since your samples are dirty and you are doing splitless it may not be unreasonable, especially if you are looking at very low levels which apparently you are. You may have some water in your solutions? If you do then you may not have a stable sample. Your results may be normal.

Perhaps using a smaller injection when operating in a splitless mode will be helpful.

Hope you find a solution.

best wishes,

Rod

[AICMM]

tabra,

I would go more along the tack of JI2002. I would lower my start temp to 50C (improve re-focus) and I would cut my split flow down to around 30:1 or so and I would consider delaying my split time. What is your calculated or measured flow rate? Under these conditions, I would not be surprised that you have not swept your liner clean in the time alloted.

Best regards.