Help - my separartion has just gone terrible

[Prim]

Last week I had a nice peak at 1.88 minutes (40 seconds across at base) and a clear chromatogram upto 6 minues) Today I have a lump of mess. An extrememly spiky peak ranging across 2 minutes with an additional peak turining up at the end (again wide)

This is a new BEH amide column
pH 9 (using ammonium hydroxium in ACN 5% MeoH) and (100% MeOH 0.1% ammoium hydroxide) (20 % max MeOH in gradient)

I am not sure what has gone so terribly wrong. I have made a new sample and the problem remains..... I fail to see how acn and MeoH go off.


Any help much appreciated

[mattmullaney]

Hi Prim,

A bit more detail would help. What are the column dimensions and eluent flow rate, and maybe the temperature at which the separation is carried out. Or, the retention factor of the analyte...that's what I'm after anyway. Offhand, the eluent is kind of basic, and the peak you were obtaining even last week seems really kind of broad in a six-minute gradient.

By chance, can you install a sacrificial silica column in between the pump and the injector in your LC system? This can help a lot in the way of preserving the stationary phase in your analytical column--and it won't interfere with your separation at all.

[Gerhard Kratz]

What happened with the column between last week and this week? Could a colleague have used it? What was the storage conditions?
What Mattmullaney pointed out is called a saturator column, and you can use every available silica material, but make sure that you install the column between pump and injector. Good luck

[Prim]

It's injected in neat urine so the matrix is super dirty, I suppose that could account for a wide peak. Maybe.

Nothing has happened to the column, as in, no one has used it. Though it was left in 99.9% ACN at a pH of 9 for a week. Not my normal practice but due to illness.

I will try to equilibrate the column according to Waters conditions and hope it helps.

I do use a guard column. Any ideas as to why my peak is so broad? This compound is difficult to retain, less than 1 minute on a C18 BEH and T3. Also I require negative ionisation in the mass spec, with optimum ionisation in high organic.

I think this assay is the most frustrating thing I've ever worked on.....SPE appears a total no go in regards to sample clean up as retention of the analyte is near impossible.

Thanks for the replies

[mattmullaney]

I still need to know a bit more about column dimensions/eluent flow rate/temperature/injection volume to speak to the breadth of the analyte peak. Another question comes to mind...if the analyte is so polar and doesn't "want" to be retained on a C18-type phase (Atlantis T3) or a hybrid amide phase, why not try HILIC? You'll certainly have high organic for the MS conditions you require.

A guard column, as opposed to a saturator column, has a bit of a different purpose. The saturator column "takes the hit" from the basic eluent rather than the analytical column, which is why it is placed into the flow path between the pump and the injector rather than immediately prior to the analytical column (where a guard column would normally be placed). The stationary phase of the saturator column is sacrificed--basic eluent dissolves the silica in the saturator column so as the silica in the analytical column does not suffer as much of the same fate, and the chromatography of the analyte(s) is unaffected.

Also keep in mind--the effective pH of the eluent, after MeOH enters into the elution program, will be more basic than what it is in the aqueous buffer.

[HPLCaddict]

A BEH column normally should withstand pH 9, but one never knows . And pH 9 in water is different from pH 9 in Acetonitrile...
What's the retention factor of your analyte peak? What are the column dimensions? If this is some sort of UHPLC-like hardware, could it be that your peak is widened because of extra-column band broadening?
Concerning the ugly peak you're getting now, I'd risk a injection without guard column...maybe it's time to change it.

@Matt: This actually IS HILIC

[mattmullaney]

Good thing there is a HPLCAddict around to clean up my errors, so my thanks for that...apologies to all for failing to recognize a gradient program that clearly is designed for a HILIC column...and failing to see that BEH Amide IS a HILIC column.

Ugh...!

Otherwise...retention factor, column dimensions, etc., no doubt they can help for to solve the peak breadth issue. The nature of the extra-column volume, too...another concept I missed before.

[HPLCaddict]

Another thing just came to my mind . You're injecting neat urine, right? That's of course a purely aqueous sample...and the worst sample "solvent" for HILIC. Maybe that's the reason for the peak being so broad?

[mattmullaney]

Another excellent catch by HPLCAddict...could very well be that the matrix (urine) is just too polar here. Easy checks are to cut the injection volume down and/or to dilute the sample with the initial eluent, say one part sample and one part eluent and inject twice the volume.

Though it still would be nice to know what the injection volume is at the moment...