Interpretation of Mass Spectra, online course 17oct2011

[scitrowske]

Hi, my name is Scott Citrowske I am an analytical chemist at Boston Scientific in Maple Grove MN.

What I would like to get out of the MS interpretation class; is a solid technique to determine unknowns using accurate mass measurements along with MS/MS fragmentation.

I have about 18 years of laboratory experience and about 7 of them doing some type of MS work. Currently I am working on LCMS primarily. In our lab we have 4 UPLC MS systems. 3 are Waters Triple Quads and 1 Agilent 6510A QTOF. I mostly do method development work for GLP type anaylsis but also perform non-routine analysis for identification of unknown compounds, drug degredation ID, and leachable extractable analysis.

[tom jupille]

Here's a link to the one taught by David Sparkman:

http://www.mass-spec-training.com/msi.html

[Alexandre]

Dear Scott,
With your profile you really don’t need courses. There are labs in research institutes and transnational corporations that try to do elucidation of unknowns. It goes slow… So most people are at the same stage as you.
There is good progress by many MS manufacturers in terms of hardware and software.
Basically, at this stage of complexity of many modern problems the computer assistance is essential. Of course, you have to make the final judgment and… verification. You will be lucky if not by de-novo synthesis.
From my experience, try to research your area as much as you can. Purchase many homologous standards, investigate them at all possible CV CE ionization modes and types, try no adducts and with.
I had luxury to play with identification of sugars and had even Orbitrap. Surprisingly not many papers on bloody simple sugars. I assumed that on something as simple as sucrose I should find hundreds of MS papers, no way! On Orbitraps – forget it.

------------------------
"If your experiment needs statistics, you ought to have done a better experiment." Rutherford

[Jetjamnong]

I agree with Alexandre.

Jetjamnong

[lmh]

Sugars are a prime example of something that is phenomenally difficult to sort out using LC-MS. All hexose disaccharides have exactly the same mass so it makes no difference whether you've got an orbitrap or not (except it's nice to have confirmation that it really is C12H22O11+adduct). Since a typical hexose such as glucose has 5 possible hydroxyl groups by which it could be condensed to another sugar, the number of hexose disaccharides that theoretically could exist is large. Adding further sugars makes the situation worse and worse. Many of them will fragment very very similarly because the bonds that usually break are the glycosidic linkages, so all you find is that your disaccharide contains... two sugars! wow! There is absolutely no shame in finding it hard to identify unknowns from chemical deduction using LC-MS data.

[Mattias]

Agree,

I have mostly worked with MS/MS of peptides, which is very easy since you know exactly where the molecule will break. Lately I have had some tasks to identify small organic molecules, which is much harder.

- Exact/Accurate mass can help you to find possible molecular formulas.
- Databases like http://www.massbank.jp can also give assistance.
- If the UV-spectrum of the peak is aromatic you can add a "double bond equivalents" filter (an aromatic molecule must have at least 3 as DBE)
- Look for isotopic pattern and loss of water or ammonia.
- Does the molecule have response in both positive and negative mode?

After that you will need some kind of information of what substance it could be (e.g. is it a leachable or a degradation product).

Using the fragmentation pattern to directly see the structure is (I would say) almost impossible for small organic compounds. It is the luxury you can afford when you have found the compound (but I usually don't give much for explainations that uses more or less exotic rearrangments, it is just a way finish the puzzle when you know the answer).

[scitrowske]

Thanks for all of your input! I am currently enrolled in this class to try and see if it worth it for some of my other Jr. Colleague’s. I have done some unknowns analysis using the QTOF with some success. I have in my arsenal a good NIST data base and Mass Frontier fragmentation generator software. I have found that the Mass Frontier is good at helping choose which ref materials to buy once you think you have the unknown identified.

I have also used deuterated MeOH in my mobile phase to help confirm some isomeric degradation products of our pharmaceutical ingredients. if you have a compound with OH groups on it the Deuterium will replace the H on the OH and you can see the differences in the fragments that are generated in MS/MS. Pretty cool...

Thanks again, and if anyone has any other good tricks for determining unknowns it will be appreciated.

As always if you can use other techniques to help ID like GC/MS and FTIR that is a plus. But some times you matrix or compound may not allow the use of them.

Scott

[sdguy]

Thanks again, and if anyone has any other good tricks for determining unknowns it will be appreciated.

Form a friendship with your local NMR guy. The combination of NMR and MS is the most powerful structure determination tool, with the possible exception of single-crystal X-ray.

[Gaetan Glauser]

Just keep in mind that while NMR is very powerful for structural characterization, it is much less sensitive than MS, even with the use of micro-probes. One will generally need high amounts of purified compound, which in many cases will be impossible or incredibly time-consuming.