Ghost Peaks?

[Apple_insci]

We had a bunch of junk peaks eluting in the late of the gradient with blank injection.
Mobile phase A is Phosphate buffer at pH 3.0; B is pure acetonitrile. The blank is ACN/H2O 75/25.

Are these junk peaks ghost peaks? What might be most suspicious cause of this?

Thanks

Apple

[danko]

Hi Apple,

It’s not at all unusual to observe junk peaks eluting under gradient elution. The cause is mostly impure chemicals used for preparation of the mobile phases. Or it just might be the mobile phase reservoirs, that weren’t washed adequately. These impurities are retained in the beginning of the run, while the mobile phase is mostly aqueous and eluted as a consequence of more and more ACN is mixed in the flow.
First you need to check whether you’re using HPLC gradient grade chemicals and if not, you should consider this possibility in the future. Secondly, you need to check whether or not you’re introducing these impurities by injecting “somethingâ€

[Apple_insci]

Thanks Danko. I tried to inject nothing with the mobile phase, but the junk peaks were still there. So I used pure water (tried both MilliQ and HPLC water) as M.P.A and injected nothing, still found the junk peaks. Can I conclude that the junk peaks were caused by water? Changed ACN brand too, but didn't change anything.

The MilliQ we have is Biocel. Is it good for analytical analysis?

What else I should try to get rid of the junk peaks?

Thanks!
Apple

[danko]

OK, it requires more comprehensive troubleshooting than I expected. First you need to confirm that the junk peaks have the same pattern regardless the ACN brand. If the answer is yes, then the ACN is “clearedâ€

[DR]

The way to see if the junk is from the water is to let the system equilibrate for a long time at your initial conditions. Then run back-to-back blank injections. If there is a lot more junk in the first blank, it's coming from the water (usually).

To get rid of the junk, pretreat your A-phase destined water with Empore C18 Extraction disks. If you search the forum, you will find that I (and others) have previously covered this in greater detail (maybe Tom could assemble a "best of late gradient system peaks" thread and make it sticky?).

[danko]

Excellent suggestion DR

Best Regards

[Apple_insci]

is there any possibility the column might be the cause?

Apple

[danko]

Hi Apple

I doubt it very much. Of course there can be some column bleeding, but you won’t see it as peeks.
Really, you should try to conduct the test DR suggested: Equilibrate the column with the start conditions for an hour or two and then inject 0 μL twice. If it’s the mobile phase that causes the trouble, you’ll see, as DR pointed out, much more junk in the firs chromatogram than in the second.
If it is the column, the junk would be identical in both runs.

Best Regards

[tom jupille]

(maybe Tom could assemble a "best of late gradient system peaks" thread and make it sticky?)

Good suggestion; I'll put something together and add it to the FAQ.

[Apple_insci]

I tried the way DR proposed. After equilibrated for 1.5hr at initial condition, the first injection (nothing) did have worse junk peaks than the second inj.! Yeah! Thank you very much DR and Danko!

I know nothing about Empore C18 Extraction disks. But will search DR's old posts. Any other quciker option I can try since ordering the disks may take longer time?

Thanks again all!

Apple

[Apple_insci]

I found DR's old post, it says "feed the disk 1L of water, then flush it with a little MeOH, rinse, repeat". Does MeoH cause contamination? Can I repeat with brand new disk

Thanks!
Apple

[Bruce Hamilton]

I know nothing about Empore C18 Extraction disks. But will search DR's old posts. Any other quciker option I can try since ordering the disks may take longer time?

I've never tried it, but I don't see why using your C18 HPLC column won't perform the same task if you only want to perform a few runs..

I would just use an initial mobile phase that was compatible with your column ( assuming it doesn't like 100% water ), and pump isocratically through the column to obtain sufficient volume, and then use your gradient method to clear the all the ghosts from the column.

If you have a C18 column that likes 100% water, you could pump you buffer through that first, using a gradient to clean it. If this is a really silly idea, I'm sure others will quickly tell you

I'd actually be wanting to permanently fix the water source, rather than apply another filtration step. Ghost peaks often appear after water purification systems have had cartridge exchanged, maintenance, etc. The recommended flush volumes are often too small.

Please keep having fun,

Bruce Hamilton

[tom jupille]

Another trick (that only works if you have a high-pressure mixing system) is to put a C18 guard cartridge in line between the "A" pump and the mixer.

That said, I agree with Bruce that the best approach is to find and eliminate the source of contamination.