The LTQ ion trap on our Orbitrap was replaced last week. Just this inner component, the essence of the linear ion trap and it has done us well since 2003 when we at TSL, first purchased the LTQ. A few years later, in 2007 we upgraded it to an orbitrap in collaboration with JIC and IFR. The LTQ part was taken away and fully overhauled but I think this little ion trap was exactly the same one.
So this in what those inner quadrupoles look like and the gloved hands give a nice sense of scale. I would have liked to keep this for teaching purposes, but sadly the old component had to go back and instead we have a little set of photographs (thanks to Jan). I think the LTQ is a great robust design, it has simply worked (well, nearly all of the time, as much as one can reasonably expect from a complex instrument in 24/7 operation).
I would have liked to link to a nice little teaching video but can only offer you the obvious wiki page (http://en.wikipedia.org/wiki/Mass_spectrometry) or a highly detailed paper (http://www.chem.mun.ca/courseinfo/c4151/references/Linear%20ion%20trap.pdf). This wiki gap should be plugged really - ion traps are so well established and such a fundamental part of research and analytical sciences.
So instead I will enthuse briefly and (I hope) at a reasonable level of detail:
Looking at the trap reminds me both of the achievements of mass spec; that signals as low as in the order of hundreds of thousands of ions (charged molecules) may be detected. That is pretty amazing, 500,000 is a number that I can quite clearly imagine - and much smaller than current economic crisis figures for example (http://xkcd.com/980/) - and only a tiny fraction of a mole (chemical unit 6x10^23). Yet with all the current excitement about single molecule detection in microscopy and DNA sequencing I enviously think of what might be a hard limit for mass spec. You see, a mass spec must actually move ions physically in order to detect them, over really quite considerable distances, and in doing so loses some en route. Furthermore, when looking at the fragmentation of a molecule you need to have a population of ions: if it breaks into two pieces each piece must retain at least one charge to be detectable – uncharged molecules are invisible to a mass spec. So to fully sequence a peptide of eight amino acids one needs at least seven specific fragments (as one knows the total mass), preferably more – observing matching parts of a fragmented molecule (b and y ions to use some jargon) is much more convincing than observing just a few pieces. A spectra is rarely completely free of noise, so many fragments need to be recorded to give some reasonable intensity. So I find it hard to imagine, even with perfect collection and trapping of ions, that this limit of needing a population of ions to fragment can even be over come by mass spec.