Structure elucidation from GC/MS with Electron Ionization

[adam]

Hello

I used to do a lot of GC/MS but it's been quite a long time. So I am hoping someone can help with the following - fairly basic -question.

My understanding is that, when you use electron ionization, you typically get pretty complex spectra which can be interpreted by matching against library spectra. Conversely, my understanding is that it is usually pretty difficult to interpret electron ionization spectra by using the "detective approach" i.e. where you look at the different masses and try to reconstruct what the molecule was - like the procedure commonly used with LC/MS/MS.

One of my colleagues used the following analogy: He said Fragmenting a molecule by LC/MS/MS is like breaking a bottle with a hammer into five pieces; you have a reasonable chance of putting the pieces together and figuring out what the bottle looked like. But with electron ionization it is like smashing a bottle with three sledge hammers. It becomes far too complex to "put the pieces together" - but is more effective from the standpoint of the library searching approach.

Are we on the right track. I would be very interested in hearing any feedback.

Thank You
Adam

[lmh]

This probably doesn't help, but I'd like to enlarge on the LC-MS side of your analogy. Firstly there are different flavours of fragmentation available for LC-MS; even straightforward collision induced dissociation is different in some ion traps to how it works in a QqQ or Q-ToF. In many ion traps, only the parent ion is accelerated (by resonance), so after a single fragmentation, the fragmentation process stops (if you're unlucky you've discovered that an ammonium adduct contains ammonia). In a collision cell of a QqQ, fragments continue to be accelerated through the collision cell and fragment further.

More importantly, ions in LC-MS CID often undergo rearrangements, so the fragments you get out can be made up of subunits that were not actually neighbouring, attached parts of the original parent molecule. There are several ways to imagine why. For example, at low energy, it's unlikely that enough energy will randomly be found in any one bond to break it (so the QET mechanism doesn't happen), but given time, it's possible that the molecule will twist round into a conformation where a lower-energy intramolecular reaction becomes possible, a reaction that results in fragmentation. Some of these reactions can be easy to guess, but some, at the moment, are quite hard. Of course not all molecules indulge in non-intuitive fragmentations.

The upshot of this is that your colleague's analogy needs a bit of modification: in LC-MS, you might (or might not) find the bottle's neck attached to its bottom, but that doesn't mean that they were attached before you hit it with the hammer. I'm, personally, rather sceptical of mass spectroscopists who claim to be able to solve structures by fragmentation alone (or even with accurate mass). They usually turn out to work on a small subset of bottles (a particular family of closely-related molecules), and even then, I feel that full structural identification (in the absence of a library or standard for comparison) is usually going to be a multi-disciplinary technique, usually needing NMR.

[AICMM]

adam,

If I am not mistaken, McLafferty's book was done with mostly EI instruments. It has been a long time since I have looked at the book (lent it out and never got it back) but I am still pretty sure that is the case. There is a good deal on structural elucidation in that book.

Best regards,
AICMM

[Don_Hilton]

When you try to read McLafferty's book that therear principles by which molecules fragment. And there are other diagnostic bits of information you can pick out of a spectrum. But, much of this helps you to figure out what could not have happened to get to the spectrum before you. This is like working a puzzle. Using the bottle analogy, you nave a number of fragments that may have been a bottle. You are able only to see them poorly,so you have to make assumptions as to how to fit together. And you may not besure the fragments are all from the same bottle - or even that it was a bottle that was broken. External clues will help.

There is a story that I may have posted in the past. It is one that I heard as an undergradute - so it is a number of years old. And it may have some truth to it. The story goes like this:

McLafferty is said to have caimed that a good mass spectroscopist could identify any compound by careful examination of mass spectra. People would give him compunds for identification and he would give them to his students. His students were known to haunt the halls of Cornell at strange hours of the night, borrowing instrument time, obtaining IR and NMR spectra. The students, having worked out the structure, would rationalize the mass spectrum and show how it unambiguously revealed a single strucure. And McLafferty had another example that proved the claim.

The story may be a bit far fetched, actuallly, because one quickly learns that the neutral loss can be fatal to that unambguous solution.

For some of us, the many small fragments are much more promising becaues the question is how to get detail about the structure. The more places the molecule breaks, the more information you have about how parts are connected. The solution of a structure is all about gettign sufficient information - and being able to use it.

[lmh]

Anyone care to distinguish L-glucose from D-glucose by MS alone, then?