Are these adducts between the precursor and fragment ions? Adducts with the solvent? I am doing direct infusion with an APCI source (negative mode) and I am using pure acetonitrile.
My first guess would be that you have double charged (+2) species in the ion that you are isolating. Did you look at the isotopic distribution of the ion? There is no way that the fragments can have adducts unless the parent ion also has those adducts. There has to be some contamination in your parent ion.
Is this on an ion trap instrument? If so, then you could see ions at higher m/z values due to ion-molecule reactions during isolation and CID. This commonly happens in an ion trap. You can test this by turning the energy for CID to 0% and then increasing the isolation time. As you increase the isolation time, the higher m/z value would start to increase in intensity due to increased time for the ion-molecule adduct to form. For APCI, you typically don't get multiply-charged species, but if you are getting a multiply-charged ion, then you would get some singly-charged products.
How much higher in mass are the fragment ions observed? What instrument are you using and what collision gas is used? It could be adducts of your compound with solvent or collision gas molecules. If your collison gas is room air and the relative humidity is high, you could possible see water adduct ions at 18 Da higher.
I have also seen this when acetonitrile is used as a diluent in the past. Also, are you certain that you have a singly charged precursor ion ( I assume it is the deprotonated form of your analyte and that the isotopes are 1 Da apart in full scan mode)
I agree with Kerry Hassell, you should look at your isotopic distribution. In the case you have less than 0.5 m/z difference between your parent ions isotopic peaks it must be that you have contamination with multiply charged compounds. You may get rid of it by using higher collision energy (the multiply charged compounds should fragment at lower collision energies). We have developed a simple method that enables structural study in the case of isobaric contamination:
We have also encountered this type of phenomena in the past and we do believe it was fragments forming adducts with solvent, in our case water, in the collision cell. As we were working on a QTOF at the time, we were also able to mass measure the ions formed to give us greater confidence in the conclusion. It turned out to be [M+H-MeOH+H2O]+. This effect was also discussed in a paper published in JASMS a few years ago - Tuytten, Robin et. al. [J. Am. Soc. Mass Spectrom 2005, 16, 1291-1304]
What kind of compounds are you looking at? And which m/z range? Some larger (i.e. >700 m/z) molecules might indeed bear two charges. I saw it for tetraglycsosilated phenolics in negative mode ESI Ion Trap-Orbitrap measurements.
First of all, let me give you some additional info: I am using an ion trap instrument (Thermo LCQ) and my analyte is a triterpenoid acid with MW ~ 457 Da and the collision gas is He. The fragment I observe go up to m/z ~500. I am quite sure that the isolated ion is not multiply charged as I observe it at [M-H] and yes, the isotopes are 1 Da apart. In the case of multiple charged ion I should also observe an ion at m/z ~ 913, though probbably of low intensity, right? Either way, I assume that the molecule is too small and it lacks functional groups for multiple charging.
@ Timothy Garrett: interesting... I will try doing that for sure. Have you had any similar experiences in your studies?
The contaminants from the collision gas is also a possibility, but we have not observed this fenomenon in the past.
Graham Lawson De cMontfort Uniuversity. I would look for adduct ions based on the components of the LC mobile phase. We have seen multiple water clusters, combinations of water and methanol clusters and methanol clusters alone. These have been observed in both ion trap systems and old triple quad systems and were produced deliberately by increasing the source pressure ie increasing the number of ion molecule collisions.
A presentation about the APCI, how it forms ions and a list of adducts that can be found in both positive and negative mode: http://www.chemistry.uoguelph.ca/educmat/chm343/martos%20old%20pages/LC%20MS.PDF
@ Alen, yes, I have seen ion-molecule reactions in my studies. Some ions will pick up water, CO2, or even part of the analyte can form an adduct with itself. Ion-ion interactions are typically more exciting as is the case with ETD (electron transfer dissociation). Hope you get your answer.
@ Timothy, looks like you were right... I did the experiment with the isolation time and at really long isolation times some fragments indeed appear at higher m/z values compared to the precursor ion (and @ CID = 0%).However, these ions do not all appear in the spectra of my CID experiments in which the energy is set to 40 %. Therefore, I assume that these ions might actually be some precursor-fragment adducts. Oh, and I do not obtain any ions/adducts with acetonitrile.
Again, thank you all for your help and suggestions.
Your have 43 Da of adduct formed in APCI mode.It seems either in triterpenoid acid there is some impurity detected while the sample is ionize is mass spectrometer. Or some adduct of 43 Da is formed. It depends on buffer and solvent use for solubility of
triterpenoid acid. These are possibility that we have to studied carefully.