small molecule identification using LC/MS

[meillid]

Hi, I try to use LC/MS to identify some unknowns degradents with the parents compound have masses of 291 and 311. Hopefully the degradent will have masses below these numbers. However, even when I run only mobile phase (0.1% FA in 20:75:5/H2O: MEON: ACN), there are so many very strong peaks (around e5) within this mass range, such as 163, 191, 193, 225, 257, 275 etc, Is this normal for your mass spectrometer? If it is true, how do you make use LC/MS for the identification? Thanks in advance.

[Ary]

Hi Emily

These could be solvent clusters. Try pushing your source temps up in 10C steps and see if they go away. Also increasing your gas flows will help. Depends on the instrument you are using exactly what helps most if it is clustering. If its not clustering then its probably impurities in your solvents and the most likely culprits will be the FA and your water. Try looking at these things first as the clustering tests are pretty quick to do, and changing to different solvent sources should fairly easy too. If you are still having problems after that then let us know. Good luck

Steve

[Noser222]

Check your fragmentor setting if you think it is solvent clusters. A background subtraction would help if you are not doing that already.

[yangz00g]

Unkown metabolite analysis is a very difficult task for even highly experienced analysts.

You need get some first-hand knowledge of the possible metabolizing and degrading pathway of the parent compounds in a certain media (please be noted that in a compex matrix the metabolites can be an adduct). If the compound has a special stable function group or a special element, you may want to pay attention to this kind of " tag".

Put simply, do a full-scan, and then a data-base match to single out the known compounds, and then determine the unmatched peak to see if they are real unknowns, and then ......Remeber that known compounds could be unkown metabolites.

Sometimes LC-MS or LC-MS/MS alone cannot handle unknown analysis typically in nature product analysis and drug metabolite profiling because of complex matrix, ion suppression, co-eluting peaks, etc, you may have to need additional detection method such as NMR to help the identifciation.

Hope this help

[james little]

We identify a lot of compounds by LC-MS with a TOF-MS and LC-MS/MS with a triple quadrupole.

We run 5 experiments with one injection. One scan is set up at low skimmer voltage in postive for MW info. One scan at high skimmer voltage in postive mode for substructural information. The lower skimmer voltage is setup to sample m/z 125 and above and the higher voltage one to scan to the lowest mass detectable with our instrument (~m/z 59). Can scan lower at higher voltage since chemical noise from solvents decreased significantly.

Two experiments are also setup in negative ion mode.

The fifth experiment is the diode array from 190-900 nm's.

Most of our compounds have UV chromophores, thus we use the UV trace to locate the peaks of interest and guide the process of background subtraction.

With just a little knowldege of the chemistry, you can identify unknowns readily. We add the spectra (plus associated structures) of both high skimmer voltage negative and postive to a library. The library is used to find model compound fragmentation by searching the database by structure and for identifying unknowns by searching the spectra.

The library software and appropriate libraries are used for EI GC-MS and LC-MS/MS and in-source LC-MS.

The library software and approach is described at:

http://users.chartertn.net/slittle/hplib.html