Tailing problem with PAH analysis

[M-KW]

Hi,

I was using a Zebron 35HT (30x0,25<0,25) column for about 3 months before it started tailing on my GCMS System. I was using a deactivated guard column (2,5m x 0,25) which I had to shorten / exchange periodically to get acceptable tailing. I changed the liner about once a week due to increased (tolerable) tailing. however, after 3 months tailing wouldnt get better with that 35HT column even though I changed septum, liner, guard column and syringe. Even injecting directly on the 35HT didnt resolve the tailing problem. (Splitless directly on the analytical column doesnt work accordingly does it?)

So I tried using a 30x0,25x0,25 HP5ms which is now showing excessive tailing after only about 30 injections. (using a guard column connected by agilents ultra inert glass press fittings)

I am using a Shimadzu QP2010 Ultra with a glass insert liner with silanized glass wool. Thermogreen septa (Supelco). The splitless injection is done at 320°C liner temperature for 1.5 minutes at a pressure pulse of 300kPa at 70°C Column temperature. Solvent is n-hexane. I am using a constant flow method of 50 cm/sec. Temperature ramps are optimized for benzo-<->benzo[k]fluoranthene- separation, with a maximum temperature of 340 °C. (quite fast if anyone is interested ) The MS is a quadrupole with Turbopump, the EI Source needs a bit of a cleaning I guess...

Anyone an idea why my system is unstable? I would expect a higher life time of my columns and less changing of the liner/Guard column. Samples arent highly contaminated, with low matrix contamination.

Is it a problem that I am simultanously checking for PCB compounds?

I used auto-tune to check for an air leak, m/z 28 was in appropriate range...
I also fasten the nuts inside the GC periodically (using graphite/vespel ferrules).

Related to this: how do you clean your glass insert liners and which kind of glass wool do you use? (and how much (pictures appreciated))

[Rndirk]

My experience with PAH on GC-MS is that the signals always show tailing, to some extend. It's subjective but I know what is an acceptable chromatogram for PAHs and what is not for my system. Maybe you can include a chromatogram?

Any point of activity in your system, starting from the inlet to the MS, contributes to tailing. If changing your liner/septum & trimming the guard column doesn't improve the signals then we have to look elsewhere.

- Do you clean your inlet?
- If you use a guard column, the connection between that column and the analytical one is a point I would suspect to cause tailing. I have experienced that overly tightening the ferrules in this connection can lead to excessive tailing. How is this connection made in your case?
- What is the MS source temperature? Increasing it (if lower) to 300°C can certainly help for PAHs.


By the way, you don't have to auto-tune to check for leaks. Your system should have a check tune or air/water check, which is shorter and doesn't change the MS parameters every time.

Having a leak detector to manually check the GC connections is also very useful.

[M-KW]

I clean my liners; at the moment on the other hand I use a fresh new liner from Restek with glasswool

As I wrote the connection is a universal press fitting (glass) from agilent. MS Source temperature is 280 °C.

Maybe the injection temperature is too high? I am using 320°C and changing the liner (from a washed one to a new one) did decrease the tailing now.

I will reply in a second with two chromatogramms so you can see my problem.

[M-KW]

5ppb Standard: SIM-Chromatogramm for Benzofluoranthene and Benzo[k]fluoranthene

<-- at start
<-- 24 injections later

[markf]

Doesn't look like just a tailing problem, sensitivity loss maybe? but why...

Or the areas are okay and the peaks just "flatten out" with time?

[M-KW]

areas are lower by about 40% which I somehow linked to the tailing

[Rndirk]

5ppb Standard: SIM-Chromatogramm for Benzofluoranthene and Benzo[k]fluoranthene

<-- at start
<-- 24 injections later

If this is the same sample/standard then tailing is indeed not your problem, loss of sensitivity and/or loss of analyte is.

My guess is that after a certain amount of thermal cycles something starts to leak or loosen in your system, causing analytes to escape or condense somewhere. Also if this happens, it is probable that oxygen enters the system and analyte peaks decrease even more .

Take a close look at the inlet-column connection, column-column connection and column-transfer line connection. Use a leak detector, increase the inlet pressure if needed, and carefully sniff those. This doesn't work for the column-transfer line connection due to the vacuum, here you can use some kind of spray and check the MS.

I can't give tips for glass press fittings, never used these.

EDIT: the fact that changing your liner resets the problem doesn't fit in my story....

[Peter Apps]

This looks like symmetrical peak broadening rather than tailing - the peaks are getting twice as wide and half as high.

Dirt anywhere will cause both broadening and tailing, liner swaps and guard column trimming need to be regular and frequent. If the guard column really is catching crud that escapes the inlet, then as it gets shorter from trimming it becomes less effective, so you may have got some crud into the separating column itself. You could try trimming the first 50 cm or so off the separating column.

What deactivation do you have in the guard column ? - your oven temp is just right to get hexane solvent recondensing, which can cause all kinds of peak distortions if the liquid does not wet the walls. This may also be an issue with the guard being trimmed too short - with liquid getting onto the column and disrupting the phase.

Symmetrical broadening can be caused by non-optimal gas flows, which will be generated by the pressure pulse during injection. Do you really need this ?

Peter

[James_Ball]

5ppb Standard: SIM-Chromatogramm for Benzofluoranthene and Benzo[k]fluoranthene

<-- at start
<-- 24 injections later

Im betting the areas are about the same, but the heights are lower due to the broader peaks. This is a common problem with benzo b and k.

You said you trimmed and replaced the guard column but did you trim any of the analytical column?

Once a contaminate gets past the guard column it will stick to the head of the analytical column and sometimes you have to remove the guard and take a loop out of the analytical column then install a new guard to fix things.

It is usually high boiling hydrocarbons that build up on the head of the analytical column that will cause this problem. If it happens by simply injecting clean standards then you are probably getting air in the system and degrading the analytical column over time.

[M-KW]

I use high pressure injection since it helped get less tailing on the naphthalene peak. I will try without...

However I changed the insert liner again and it worked again for some injections. Then the tailing increased again. however the first peaks are good still the last ones are now broad and even retention time has increased.

I hardly injected real samples, mostly standard mixtures right now...

I use a non-polar deactivated guard column 2,5m x 0,25.

EDIT:
I changed the liner (to a brand new one) and found out that performance is better, but after a short while I saw tailing again.
50 ppb samples: Benzo[a]anthracene and Chrysene:
<-- Directly after Liner change
<-- 17 injections later

EDIT 2:
Should I clean the EI source? could the broadening/tailing be in the MS?
Or could the septum be the problem? I found parts of the septum in the liner a few times already.

[Rndirk]

I use high pressure injection since it helped get less tailing on the naphthalene peak. I will try without...

However I changed the insert liner again and it worked again for some injections. Then the tailing increased again. however the first peaks are good still the last ones are now broad and even retention time has increased.

I hardly injected real samples, mostly standard mixtures right now...

I use a non-polar deactivated guard column 2,5m x 0,25.

EDIT:
I changed the liner (to a brand new one) and found out that performance is better, but after a short while I saw tailing again.
50 ppb samples: Benzo[a]anthracene and Chrysene:
<-- Directly after Liner change
<-- 17 injections later

EDIT 2:
Should I clean the EI source? could the broadening/tailing be in the MS?
Or could the septum be the problem? I found parts of the septum in the liner a few times already.

The septum can be a problem yes, if it is not inserted correctly (screwed too tightly for example). Still i was under the impression that (see my earlier post) your peak areas are decreasing a lot.

[Peter Apps]

If changing the liner fixes the problem then whatever is happening to the liner is the cause of the problem, so you do not need to clean the ion source.

Septum particles in the inlet are a disaster for these high boilers. what kind of syringe needle do you have ? - if it has the default bevelled tip change it for one with a conical tip or a side-hole.

Increasing the inlet temperature will help.

Why the liners are deteriorating so rapidly is a puzzle - do you have an air and moisture scrubber on the carrier gas line ?

Peter

[M-KW]

Increase inlet temperature? Its 320°C already (608 Fahrenheit) further increasing led to improperly high septum bleeding :/

I am using an air and moisture scrabber which I now controlled and its still only a little bit worn.

The MS filament is pretty old right now so that might be a different problem soon.

I will now once again change insert liner and cut the guard column... Maybe the guard column is too short? its only 2.5 m at the moment with now decreasing length.

[Rndirk]

Increase inlet temperature? Its 320°C already (608 Fahrenheit) further increasing led to improperly high septum bleeding :/

I am using an air and moisture scrabber which I now controlled and its still only a little bit worn.

The MS filament is pretty old right now so that might be a different problem soon.

I will now once again change insert liner and cut the guard column... Maybe the guard column is too short? its only 2.5 m at the moment with now decreasing length.

Inlet temperature is OK in my opinion.

Guard column length => if you aim for recondensation of the solvent on the guard column (start oven temperature < boiling point solvent) then the rule of thumb is 1µl (liquid) / 1m. So if you inject 1µl this should be OK too.

[Peter Apps]

There is something very odd about your peak sizes - the peak height at 5 pp is about 2, at 50 ppb it is only 4.

What volume are you injecting ?

Peter