GCMS peak broadening with spikes at apex

[Edde]

I ran a usual method for ethylene glycol quantitation with HP-5MSi column after changing a brand new column. The peak shape of ethylene glycol was unusual with peak broadening and some spikes in its apex. Ethylene glycol cames out at 4.8 min. Other glycols such as 1,2-propanediol came out later at 5.2 min with sharp peak. The ethylene glycol peaks did not increase proportionally at the 2nd highest and highest calibration points. In fact, the peak areas were almost the same (area 1.5e7, which should not saturate the detector).

I remember I just ran carrier gas through the brand new column but had not conditioned the column after installation. After that, I used the column to ran a few batches of other gradient analysis with max temp up to 320 deg C (with short hold time). Plus another batch of ethylene glycol quantitation with max temp up to 300 deg C (hold time 5 min).

Is the poor peak shape due to failure to condition the column after installation?

I changed new liner, septum, cut 12 cm column head and cleaned ion source to make sure other variables are optimal. Then condition the column again at 280 deg C 3 hours, then 300 deg C 1.5 hr. Injection between each conditioning showed tiny improvement.

Can I improve the column condition now by conditioning for longer period of time?

Please kindly give your expert comments.
Thank you.

[Mr.Brown]

I do not have any experiance with ethylene glycol quantitation.

However conditioning of the column is always a good idea
I usually start at 60°C and heat at 3°C/min up to the max allowed temperatur
and keep it there over night.

Bad shaped peaks is in my experiance mostly a problem of the injection and the transfere of the sample from the injector onto the column.
You should check once more if the column is installed correctly and is straight cut.
You sould also check for Leaks in the injector

[Edde]

Thanks Mr. Brown.
On repeated baking at near max. temp. (300 deg C) for 4 hours, the peak seems improved but still spiked at apex. I inspected the inlet column and found the column nut over-tightened causing the column neck to be squeezed.

I am not sure if the column without initial conditioning step is already irreversibly damaged and cannot get optimal after baking later?

Anyway, I replaced another brand new column and programmed an injection of 1uL ethyl acetate increasing the temp from 60 deg to 300 deg C (instead of max column temp 325 deg C), hold for 4 hours.

I later doubt if injecting ethyl acetate before completion of conditioning is ok.

Is 4 hours sufficient?

Thanks.

[Mr.Brown]

Restek has some nice guidelines on his homepage
http://www.restek.com/Technical-Resourc ... AN1533-UNV
http://www.restek.com/Restek-Capillary- ... ation#cond

You should also talk to your supplier what they recommend.
Conditioning is done to get ride of some loose compounds from the column film.
I am using MS as detector so my supplier recommended to condition over night.

Conditioning of a column after several injections should not damage the column.
If I have some nasty stuff on the column and I have to cut of a peace I usually do again a conditioning.

Usually for conditioning one does not inject anything.
The column should not be connected to the detector
and only the temperature program is started


There are also some troubleshooting guides out there for example
http://www.mn-net.com/tabid/10583/default.aspx
http://www.msp.ch/cs/Reference_Guide.pdf
Some of them sometimes help

[Edde]

Mr. Brown, your comments are indeed very useful.

Some progress of my problem:
A sample injection still showed a flat and broad peak.
Then, I tried injecting the calibration curve and pending results tomorrow.

I will keep you posted of any update.

[Edde]

Although the peaks are much sharper this time, the calibration curve was not satisfactory. One of the calibration points had exceptionally high IS response. Estimated from the diluted sample, the neat injection should overload the detector but I saw no chopped peak apex. I did not lock the RT. The flow rate is 1 mL/min instead of the 1.1 or 1.2mL/min in past successful runs.

I use splitless mode, open split valve at 0.67 min. My first peak comes out at 5.1 min. I am not sure if flow rate is the culprit for broad peak and disproportional response. But I increased the flow rate to 1.2 mL/min and see if there is any improvement.

Any comments are welcome. Thanks.

[Mr.Brown]

In splitless mode the flow rate is for sure an issue.
During splitless only the flow rate is the main factor how fast your sample gets on to the column
in principle the faster the sample gets on to the column to more focussed are the peaks.

In splitless mode i additionally use the high pressure mode
this increases the pressure in your injectore and therfore the flow rate during the splitless time
which gives much more focussed peaks.

[Edde]

Should constant flow rate be used all along regardless of the column length? I mean when I cut the column head, the retention time should be shorter and may get to the solvent front before data acquisition.

Am I correct that high pressure mode or equivalent is not present in Agilent older GCMS model?

Thank you very much for your kind sharing.

[Mr.Brown]

I usually use a contant flow rate.
Of course the more you shorten the column the shorter the retention time will be.
However you have to cut of a lot of the column so see a huge effect.
This usually is only a problem if your analyts peak is already very close to the solvent peak.
If this is an issue I would suggest to adjust the temperatur programm as well as the flow rate.

high pressure mode is probably only available at newer machines
however I do not have much experience with Agilent systems.

[Edde]

Thank you very much indeed for your enlightenment.

Please forgive me for bothering you with so many questions.
May I ask how to adjust temperature programme together with flow rate?

Thanks.