[Urbanium]

Hello, I never had a lot of experience about MS, when I was undergrad student one postdoc in my previous group acquired all the MS spectra for us, and now I'd like to know if there is some rule of thumb or a brief set of rules or a flowchart that can be used when deciding about the ionization method for MS (of course based on the structure of the investigated molecule)?

We have available EI, ESI and CI on our instrument. Which of these is suitable for what kind of molecules?

Also, why is it said that solvents like DMF and DMSO are not suitable for dissolving MS samples because they lead to poor ionization? How do they affect the ionization if they should be vaporised at high temperatures?

[MOTOBALL]

Very generally, covalent organic compounds can be categorized on a volatility scale.

1) Very volatile----chloroform; nitrobenzene; naphthalene

2) Some volatility----dioxin; PFK

3) Non-volatile-----glucose

EI and CI require that the compound be introduced into the ion source in the vapor phase for ionization to occur (introduced via GC or solids probe).
If you want a MW determination, use CI.
If you want a "fingerprint" mass spectrum of fragments to compare to a (computerized) library of spectra for identification of an unknown, use EI.

ESI is used typically for samples that are water-soluble, such as biomolecules separated by HPLC.
A disadvantage of ESI is that the lower m/z region is usually dominated by very intense background cluster ions derived from constituents of the mobile phase; this can make detection of low-level analytes below about m/z 150 problematic.
Another, major disadvantage is that ESI usually provides very little structural information---just MW, and signals at m/z that correspond to losses of small neutrals (H2O, NH3 etc) from [M+H]+
You will generally need ESI-MS/MS to generate structurally significant signals.

[MOTOBALL]

DMF & DMSO--relatively high b.pt. liquids & viscosity.

These are both OK for GC-MS, because the solvent will be flash vaporized in the injection inlet and will precede the analyte through the column---no interference with ionization.

One model of ESI depends upon the successive formation of ever smaller charged droplets from larger ones---since this rate is affected by viscosity of the liquid, the presence of high viscosity liquids (DMF, DMSO) in the liquid will reduce the ease of formation and therefore reduce sensitivity.  This is why greater sensitivity is observed at the end of a gradient HPLC run (low viscosity MeCN at high concn. ~90%) compared to at the beginning (high viscosity H2O @ high concn ~90%).

Also, DMSO can form very intense cluster ions (I have not used DMF-based solutions).