LC-MS/MS Column Carryover

[MJS2008]

I'm working on a reverse phase, negative mode LC-MS/MS method for a proprietary phosphonate and it's phosphonic acid metabolite. Sample prep is a simple SPE using Waters uElution plates, however I'm seeing huge amounts of carryover or interference noise at the retention times of the compounds. I've tested the method without the column in place, and the carryover following a high-concentration injection is negligible, so the carryover is not coming from the autosampler. Anyone have experience either with analysis of phosphonates or with columns or mobile phases which can reduce column carryover?

[Uwe Neue]

If the carryover is not coming from the autosampler, it must be related to the column. What column are you using? What is your mobile phase?

[MJS2008]

I'm using a Precision C18 column manufactured by Mac-Mod, 2.1x50 mm dimensions, 5 um particle size. The mobile phases are (A) water with 5 mM ammonium formate and (B) methanol with 5 mM ammonium formate and 0.1% ammonium hydroxide, running on a gradient from 25% to 100% B over about 2 minutes. Both compounds elute shortly after the gradient has reached 100% B. This isn't the best method I've ever seen, but it's what has been used for these compounds in the past so I'm treating it as a good starting point.

I've also tried a variety of other C18 and C8 columns (Waters Atlantis, Agilent Zorbax, Phenomenex Luna). All show similar carryover. Even short C18 guard columns exhibit carryover of these compounds.

[Kostas Petritis]

Propably your compounds are quite sticky and accumulate in your frits or other parts of your column. I would try to passivate the column by using 40 mM EDTA.2Na, overnight at low flow rates. Be careful to use the .2Na version of EDTA and not any other versions of it...

[HW Mueller]

We have been through this before, there is a problem with having part of an analyte stick and then suddenly unstick (in normal peak shape) the next time one injects something (a blank) unless that blank´s solvent does something, or the mechanical aspect of a gradient does something......

[Keaka]

With column carryover you only have a limited number of choices as far as how you want to tackle it:

1) Modify your gradient
2) Change your mobile phase
3) Try a different column

For 1) Sometimes there is a mechanistic effect with carryover such that extending the 100%B portion of your gradient to flush the column isn't enough, you actually have to cycle back to initial conditions..and go back to 100%B again in order to shake the carryover out. But since your carryover is huge I don't think any adjustments to your gradient will offer a lot of improvement.

For 2) Try hitting it with 50mM salt and see what it does to your carryover. I'd recommend tighter pH control of your mobile phases.

If all else fails, you'll have to try other columns. Since you've tried a number of silica columns already, you might want to look into something like a Onyx Monolithic C18.

[sassman]

I think the Onyx columns are also silica based. Maybe you can try one of the Supelcogel columns.

[Kostas Petritis]

Sometimes if you have some void volume in your connection on the head of the column (especially true in capillary column applications) the carry over goes away only after dilution with your mobile phase. In your case, due to the very fast gradient you might have residual carry over that is not able to be flashed away and you get it in the next injection. Try either a lower gradient slope (longer gradient) or leaving the strong solvent a little bit longer at the end of the gradient to see if this fixes the problem.