tailing peak

[mmegliol]

Hello everybody,

Hope somebody can help me because the GC is driving me craezy!!!

I'm try to set up a method to quantify methanol in wine sample (after distillation).
I'm using a GC shimadzu2010 with installed a column from Supelco Supelcowax10 60m x 0.32 id x 0.25; I collect the data trough the anolog signal and I use PE Totalchrome to integrate the peak.
I'm injecting 1 microliter manul.
inj 220 det 280

I start trying to keep the set up as much standard as possible, the flow in the column is 1.8 ml/min, split at 30 and I'm working with Velocity selected for what I understood the gc keep the same velocity and change the press on the inlet that at begginig is around 15 psi.

The ramp for the temperature I'm using 5 minute at 45 and 5C/min until 110 after that I ramp at 25C/min till 240 to clean all the other alchols that I don't need.

The problem is the shape of my peak they are all tailing, and I can't have a good reproducibility on my area.
I try to change ramp, flow, inlet press, split but nothing will effect the shape of my peak.

Tonight as a last test I tried to remove 50 cm from the column and clean the glass insert changing the wool, I left the gc on with temp at 230 overnight to clean well the column.

Anybody have an idea on what could be the reason of the tailing ( supposing that there are no leaking in the system!).

thanks,

Matteo Meglioli.

[Peter Apps]

Hi Matteo

Most likely a dirty or otherwise active inlet liner is the cause of the tailing, and you may have found an improvement after you changed the liner.

Poor repeatability of areas is probably related to injection techique and the large volume of vapour that you get when you inject ethanol into a hot inlet. It is well nigh impossible to train someone how to do manual injections over the internet - try pushing the needle through the septum, counting to three, smoothly depressing the plunger, counting to three and withdrawing the needle. Do not try to copy the fast cold needle injections that some autosamplers do.

Peter

[matthieu]

Hi,

Sometimes a peak tailing can occur when the column is not properly fit into the injector...(I dont like this splitted nut system from schimadzu!!)
Did you use column guide to determine how much of the column you have to insert into the injector?

Regards,

Matthieu

[mmegliol]

The injection tecnique I used is insert the needle push the plug down, click start on my system, counting until 3 and withdrawing the needle. i'll try using the tecnique suggested.

I don't use the splitted nuts, I used regolar with the tube guide to have the right length inside the injection port.

Wolud be better clean the FID, I think from last time i cleaned I did 60-70 injection.

[Peter Apps]

If you are injecting wine distillate there is nothing that can have contaminated the FID - except when you inject chlorinated solvents or silicon containing derivatives the FID will stay clean for thousands of injections.

Can you post a chromatogram (instructions at the top of the LC page) there are different forms of tailing tha have different causes and cures.

Did cleaning / changing the inlet liner make any difference ?

Peter

[mmegliol]

Hi Tom,

Cleaning the glass insert did not change anything, the shape of the peak is exactly like before and how you predict even cleaning the FID did not get me anywhere.

When I cleaned the inlet i cut off part of the column, just to be sure, but no results!

this is my chromatogram

http://i29.tinypic.com/m7s1i0.jpg

yesterday night I was looking at the chromatogram, is it possible that I'm looking at the chromat to much magnify?

i try to run a sample for all the fuses oils and on a smaller time scale looks much better!


thanks for your help.

[mmegliol]

Hi Tom,

Cleaning the glass insert did not change anything, the shape of the peak is exactly like before and how you predict even cleaning the FID did not get me anywhere.

When I cleaned the inlet i cut off part of the column, just to be sure, but no results!

this is my chromatogram

http://i29.tinypic.com/m7s1i0.jpg

yesterday night I was looking at the chromatogram, is it possible that I'm looking at the chromat to much magnify?

i try to run a sample for all the fuses oils and on a smaller time scale looks much better!


thanks for your help.

[Peter Apps]

If this is a chromatogram from a sample then I do not think that you have a problem that it is worth your spending a lot of time in solving - all your peaks are baseline separated and the tailing is certainly tolerable.

With a separation like this, any problems with area repeatability are almost certainly due to injection technique. If you can use an internal standard it would help to reduce this source of variation.

Paradoxically you might get a better separation on a much shorter column becuase the tailing will probably be less. Do you have a 20 or 30 m column with the same stationary phase that you could try ?

Peter

[SemiAuto]

I'm not an expert of experts - but I would pay most attention to injection technique to reduce your problem. If you must hand inject I would try and mimic an auto injector for quick injections. The faster the sample is injected into the inlet - and the faster the syringe is withdrawn the best chance of all of the components not to tail. The time that you set with the syringe needle slowly volatilizing out the remainder of what is in the syringe needle as it heats up while the bulk of the sample is already on its way down the column would seem to be a bad idea for your type of compounds. Remember that there is sample in the steel of your needle at all times and that's OK... you just want to inject a total 1-uL and get it on its way.


* All of this is assuming 1) that you actually don't have a leak in the septa nut being too loose very common or inlet cap not seated correctly and/or tight enough OR 2)assuming you have a column designed for alcohols which tend to tail on a generic DB-5 column.