Ghost peaks, where are your origins?

[lcrandall]

Hi all,
I am having some major frustrations with a group of ghost peaks and would love any suggestions.
Method is gradient elution at 220nm
MPA H2O with 0.05% TFA
MPB ACN with 0.05% TFA
85:15 to 5:95

So, I start seeing peaks right in the middle of the run (this is a stability indicating assay) and and having a hard time tracing their origin.
I've tried:
different brand H2O
no TFA
blank, no injection runs
two instruments
three columns (two of which have never seen the compound)
mucho washing
the area of the peaks doesn't seem to scale up with initial equilibration time
different glassware not washed with Extran (for example, the bottle is new, rinsed with H2O, ACN, H2O, H2O, then filled with H2O)

I am waiting for the delivery person with a different brand ACN

Does anyone have any other suggestions?
Thanks,
Lisa

[Bruce Hamilton]

My guess would be something associated with mobile phase or it's preparation, given the problem is appearing on two instruments.
If you have a DAD, check the spectra of the ghost peaks, as that can help identify possible sources, and also ensure that your eluting solvent isn't absorbing strongly at 220nm.

As usual, the best solution is to have alternative brands of each of the mobile phase and sample solvent components available to try, along with another instrument. Then you just have to continue to be persistent and systematic as you have been.

My current ghostbusting routine is to perform a couple of maximum volume injections of pure acetonitrile and 1:1 H2O:CH3CN ( 100 ul on my Agilent 1100 ) using the final mobile phase composition and short runs ( < 10 mins ) onto a column temperature about 10C above the method. That cleans the injector and gives any column junk a fright.

If I see a lot of peaks, I check their spectra to see if I can link them to recent samples ( I usually can't ), and I flush both mobile phase channels with 100% water and 100% acetonitrile - without TFA, before using the mobile phase with TFA. I have a suspicion that membrane degassers make good residences for homeless ghosts.

There are many other possibilities, and other ghost-afflicted people may have better suggestions.

Please keep having fun,

Bruce Hamilton

[Kostas Petritis]

You have done really well in your efforts to try to pinpoint the source of these peaks. If the area of the peaks doesn't seem to scale up with initial equilibration time that pretty much eliminates mobile phase impurities. Propably the source is from your injector, although you indicate that no injection runs exhibited the same behaviour.

I would guess that without a chromatographic column you do not get any of these peaks? What brand is your injector?

[lcrandall]

Hi
Thanks for the suggestions.
Yesterday, we pulled the needle out of the sheath and washed it and the needle seat. (It's an Ag 1329A autosampler) I ran a bunch of blank and ACN injections. It looks better, but still not where I need it to be.
Goodness, this compound! Instrument #2 probably saw less than 5 injections of this stuff. If it's contaminating the needle that quickly, even with a needle wash..GEEZ.
I will try the suggested injections (at 100uL) below and wait impatiently for the Fed Ex truck.

Does anyone know how mobile phase runs through these autosamplers? Is it always running through the loop, or is the loop only used during injections?
Thanks,
Lisa

[Kostas Petritis]

You might want to make some blanks by by-passing completely the injector. If these are clean, new ACN won't help you. What might help you is excessive needle and loop cleaning with different polarity solvents and maybe one high salt one -1M- (use water for flushing before and after the high salt). The best combination will come up after some trial and error...

[HW Mueller]

In other words, are these ghost peaks or carryover?

[lcrandall]

They are true ghost peaks, showing up in blank (no vial) injections. I tried the 100uL injections with heat and it looks okay for the most part. But, there is a new little friend eluting early, right where I have two impurity peaks expected.
So, we just realized...there was painting going on near our lab recently. It still smells like paint in here. We leave the lab doors open during the day. I close them in the evening, but the janitor always leaves them open. It's not unheard of, that it could be paint volatiles. My problems started about the same time the drywall and paint were being worked on.
If I can just get rid of the new 7min hump, I would be good to go.
Lisa

[HW Mueller]

If I understand what you did it could be carryover. If it is it should diminish by doing consecutive blank injections, a ghost would persist until you eliminated its source, for instance, by eliminating contact with paint volatiles.

[lcrandall]

Hi,
The peaks do not diminish with subsequent blank injections.
I will prepare everything now in the hood with fresh bottles and keep parafilm around bottle caps. Hopefully, if it is the paint volatiles, this will keep it out of my solutions.

Lisa

[Rob Burgess]

Hi Lisa,

I'm just wondering if you ever got to the source of your problems on this one? I'm experiencing pretty similar problems, and I must agree wuth you, it is infuriating!

Rob

[Tomasz]

If you have Agilent autosampler (Ag 1329A=Agilent 1329A?) the mobile phase runs always through the loop except the time needed for the preparation for injection. I don't know other systems but they probably work in the same manner.

[Rob Burgess]

I've solved some of my problems - it appears a batch of sample diluent that was made up had an extra peak in it - source unknown (plasticizer?).

However I'm still experiencing extensive carryover problems with my Agilent 1100 autosampler. I've cleaned the needle (no joy) washed out both lines/sample loop with IPA wash (no joy - although it has improved the baseline somwhat).

I'm injecting a sample of diluent (blank) and carryover from the main sample peak persists even after 20 overnight injections. This carryover peak does not diminish in size. I just can't seem to get rid of this carryover peak - this has happened to me before. Could it be a rotor seal problem? Worth changing I suppose as an etched damage one would inject a small amount of highly absorbing "trapped" compound each time the rotor was turned wouldn't it?

[HW Mueller]

Rob, did you estimate how much stuff would have to be "trapped" to give your results that "see" no change in carryover after 20 blank injections?