LC triple quad versus LC Trap

[mckrause]

We are trying to move to MS/MS for LC work. We have been working with a single quad Agilent system for several years but find ourselves needing the second ion confirmation. I know I can get both ions from either a QQQ system like an AB Sciex 3200 or from an ion trap like the Agilent system. From what I can tell most people use a QQQ, euphemistially called a "linear ion trap". Is there a reason that we would NOT want a true ion trap?

We run mostly single compound quantification, low level (low ppb). We always use ESI, almost always in positive mode. We do some screening work but it is all done in SIM mode and quantified against known standards. I am not doing, and don't have plans to do, full scan MS/MS with library searches. Our quantification is for regulatory purposes, so you're talking about multilevel calibration curves (7-10 levels) with curve fits of >0.995.

I like the Agilent software, probably because I just don't know any better. I am not married to Agilent as a system, just used to it and its quirks.

Any feedback that y'all can provide would be most welcome.

[lmh]

Traps and triple quads are different. The term linear trap is used by companies such as Thermo to mean a trap designed with a linear quadrupole-like element blocked electrostatically at the ends to stop ions from escaping, as opposed to a 3d trap, which has two dish-shaped electrodes and a ring electrode (typical geometry in Bruker, Shimadzu and older Thermo instruments). The idea of a linear trap is that it should be able to contain more ions, which should boost sensitivity. Bruker have equally good arguments why a 3d trap is a good thing, but I can't remember them!

In general, the sensitivity of a trap is limited by how many ions can be held in the trap. This means that changing to SIM or MRM mode may not make much difference, depending on how clever the manufacturer has been about filling the trap. Traps are excellent in scan-mode, and don't gain much by looking at single ions. It is, however, possible to combine a trap-like behaviour with a quadrupole, so that ions are first accumulated in a trap-like thing, and then scanned out of the trap into a quadrupole approximately coordinated so that the ions leaving the trap are about the right mass for the quadrupole scan. This should mean that instead of the quadrupole getting the ions it would see in a usec of scan-time as it passes through the right mass, when scanning, it actually sees a vastly enriched ion-beam and gets much better scan data.

This is what Waters are doing in their top-end triples.

On the whole, if you want to do MRMs with multiple transitions to confirm identities of target compounds, traditional trap fragmentation is not what you want. True traps usually fragment by resonance, which means they apply a frequency to the trap that causes ions of the precursor mass to whizz round the trap fast enough to collide with collision gas and fragment. The fragments are smaller and have a different frequency, so they are no longer excited. The consequence is that if you are unlucky, water drops off, the ion loses 18, and that is the only fragment you get - then you need MS3 (or some sort of wide-band activation or second activation) to see any interesting ions at all. In a true triple-like instrument, fragmentation is by general acceleration through a collision cell, the fragments being accelerated just like the precursors, so it gives a richer range of fragmentations.

If you're buying an instrument, get all the manufacturers to tell you just why theirs is the best approach, and then don't trust them totally: make them show you with some of your samples and target compounds.