I’m assuming this is a standard split-splitless inlet. Is that correct? It’s possible that you have active sites or something else happening in the inlet body (the inlet weldment). If you’re willing to change the inlet weldment next as part of your troubleshooting, the last attempt before you do that might be to clean it first with a gun brush. Please note that this is not normally recommended at all any longer due to the fact you’d be introducing micro-abrasions on the inlet, creating a higher surface area, and wearing away the inert coating – thereby further compounding the problem in the future; however, if you would like to see whether this improves the peak shape in the short term, it would confirm what I suspected and you can change the weldment the next time you do maintenance. Again, for anyone reading this – cleaning the inlet with a gun brush is NOT recommended as part of regular maintenance.
We have moved away from that practice for that reason of the micro abrasions wreaking havoc with results.
We have no free weldment assmebly avaiable. See the comment I reply to below for more info.
Is the GC a 6890 , 7890, or other? I assume this problem is new and hasn't existed throughout method development? You may need to consider cleaning the inlet by a different method. I have attached one as a reference. If that doesnt work, I have seen cases where the inlet needs to be replaced. If the model is a 7890, you then have the option of moving to an inert inlet. If its a 6890, inlet replacement is an option, or you can try the inert gold seal 5190-6144 .
Cleaning Inlet.pdf 65.2 KB
Thank you James.
it is a 5975T LTM system with a split-splitless injector (shoudl of mentioned in first comment). We have 3 of them showing the exact same pattern from different laboratories all noticed recently.
We want to avoid the gun brush as the abrasions have caused issues with us in the past.
We have also checked for any cold spots on the LTM column-column cage connection and it's OK. my colleague has sent pictures and it's fine.
We cannot fathom why it's only in this one matrix (which we need to analyse), why it's on 3 instruments at the same time, and after changing every part and cleaning the inlet as Herbs video described on youtube, that it is still causing issues.
Could it be possibly the EPC on all 3 instruments? As they all show the same fronting, and splitting pattern. this is 3/7 instruments in the past few weeks with different samples of the same matrix.
Bare in mind the GCMSs are several years old.
This is a tough one. I wouldn't suspect that you would have three EPC issues on the different instruments , possible, but not probable . Any change in lot numbers of the solvent that you use that could be tracked back to all three?
Another thought since you mention pulsed splitless at 320 C, can you share the solvent, pressure at injection , liner and column size? I would like to work out the vapor volume, unless you have already done that.
You have a puzzle on your hands.
I can't tell you the solvent but the liner is a splitless liner (with glass wool at the bottom focussed on the column, can't find part No. right now for it).
The column size hasn't changed either it's still 0.25 um 10 M column and the pressure at injection is 18 PSI with an increase to 50 PSI for several minutes.
i can't give you the exact numbers due to inteligence property issues, but this is only a recent thing and we have been running this method for years.
The same standards are being used on all 7 instruments but only 3 are exhibiting this behaviour, which is random from injection to injection. It's not even trending, some look Ok and tolerable while others look almost 50:50 split peaks.
We are currently trying to soak inlet top where the septum lies in methanol and then in toluene in an attempt to remove any debris (DCM wasn't avaiable in the lab).
I agree with James that it’s improbable the EPC modules are all causing this, especially if you’re not also seeing retention time issues. If all 3 have been running the same method for a long time, I would really suggest either buying a new inlet weldment to test or cannibalize one of the other 7 instruments for a temporary swap to see if the problem moves along with the inlet.
Is the liner an Agilent UI liner? That maybe worth a try. Without getting into details, liner performance can vary by lot , manufacturer etc. as well.
I don't know why 3 of your instruments do this and 4 are OK unless it is something like the column position inside the injector?
Looking at your IS peak, it seems that the system is overloaded and you have signs of ether band broadening in "space" or "time". These problems tend to be worse when the polarity of your solvent, analyte and stationary phase are different (Like, say, propanol/hexadecane; or hexane/pentachlorophenol on different stationary phases like DB1 for hexadecane, PEG for pentachlorophenol, or DB17 for both - The different polarity solubilities with DB17 will be better, but not optimal). These systems are potentially very unstable, and a small change can produce large effects.
I appreciate that you cannot give more info, so the questions to ask yourself are:-
Does reducing the amount injected or analyte concentration help?
Is the problem worse/better for materials with different boiling points and (in particular) polarity.
With a splitless injection: If the underlying problem is caused by, say, a polar solvent and relatively apolar stationary phase and/or analyte you will probably need a "retention gap" - For a different polarity solvent/analyte/column combination, a medium polarity deactivation is usually OK (Agilent Part No CP8017). A retention gap is not a "guard column" or "pre-column" as it contains no stationary phase. With a normal coated column, a different polarity solvent will condense in droplets along the start of the column without being adequately sorbed by the stationary phase - When the oven is ramped the solvent evaporates leaving the analyte spread unevenly along the column, potentially showing the sort of effects that you see. This is usually worse as you increase the amount of analyte. With a retention gap your solvent and analyte condense together on the "empty" surface - When the solvent evaporates the analyte is still spread (randomly?) along the empty column but is not retained (no stationary phase) until it travels to the "coated" column where it is focussed in a narrow band. This works best if the initial oven temperature is >20C below the boiling point of the solvent (Don't forget the nominal BP of your solvent will be higher in the pressurised column than normal - i.e hexane: 69C/~85C/~96C or DMSO: 189C/~220C/~240C at 1/2/3 atmospheres respectively).
I am a first time contributor so I hope this meets the necessary guidelines.
I agree with your decision to not gun brush the inlet. You will end up replacing the inlet anyway as you have seen. It is possible that deposits on the inlet shell weldment are causing your problem. Injections that allow expansion outside the liner will leave deposits on the walls of the inlet. Over time very small deposits will build up. When injections of heavier matrix or higher boiling solvent, with analyte, contact these deposits they can stick and then release. Symptoms include non-reproducible area counts, peak tailing, peak fronting and peak splitting. So a dirty inlet may be your problem. You can clean it with a swab and solvent, and then inject less to minimize vapor volume expansion outside the liner. If expansion is very large compared to the liner volume, vapor can travel out the septum purge or backwards in the inlet carrier line. Deposits will form very rapidly inside the carrier line as it is unheated. Subsequent injections can exhibit split peaks.
If you are fairly confident that your injections do not exceed the liner volume, then I would look at injection parameters as you have already tried consumables. Timstrutt covered polarity mis-matching very well . This makes sense as the problem matrix is different than (all?) others. The simplest thing to try is his suggestion of reducing injection volume, say 0.5 or 0.2 uL if the current volume is 1 uL. An alternative is to eliminate the injection pulse or limit its duration to a very short time, say 0.1 min. Many split/mishapen peak problems have been traced to pulsed splitless injections. The factors are pulse time compared to purge valve on time, pulse pressure, solvent polarity, injection volume, etc.
Any other information you can provide would be helpful in solving the problem long term along with the current symptom. Can you tell us:
Is the injected volume primarily this problem matrix or is it a solvent ?
If solvent, is it polar ?
The pulse time
The pulse pressure
The purge on time
Can you provide a screen shot of the IS ion trace and one of the analyte ion trace, both good and bad in order to see the severity of the problem ?
We tired a methanol wash (no DCM in the labs), around 25 injections, as well as soaking the inlet and top of the inlet weldment assembly in methanol over night to try and clear any heavy compounds/matrix away.
We have managed to find a temporary measure in lowering the injection volume by half from 1 ul sandwich injection (ISTD and Analyte matrix, 1 ul each) to 0.5 ul each.
We have also purchased an inlet weldment assembly and top of the weldment assembly (that connects to the EPC) to see if our issue can be resolved.
We think that a mixture of heavy compounds may have flashed back into the top of the inlet at the septum purge piping and causing the problems.
I marked this as assumed answered. Given the limited information that can be made available in this case and the quality answers below, this is about as far as the thread can go. If more discussion needed, certainly feel free to post again.
1 person found this helpful
Yeh we have a temporary solution by halfing the injection volume it resovled teh issue. We don't know how it started but that has fixed it. We have ordered parts to replace teh inlet weldment and the septum weldment assembies. So hopefully replaceing one or both of these parts will resolve the issue and we can go back to normal injection volume.