The signals shown in the attached file, lower panel, appeared as more or less constant background signals in LCMS analyses (ESI, negative mode, qTOF). Conditions and materials have been routinely used with that instrument without problems: binary gradient of water and acetonitrile, each with 0.1% formic acid, at 0.4 mL/min over a Dionex Acclaim Polar Advantage II 2.1x 100 mm column. Steel tubing in front of and PEEK tubing behind the column. The contaminant does not appear in positive mode. As we measure in positive mode most of the time it is not possible to precisely find out when the contamination took place. The first analyses that have shown the signals were acid hydrolysates of plant polyphenols (proanthocyanidins) which contained 4 M HCl. 5 µL were injected of six samples (first in positive, then in negative mode).
I have tried to learn more about that contaminant by trying some simulations of the isotope pattern (upper panel). The pattern at lower m/z values (m/z 160.8) matches 3 chlorine atoms, the other one 4 chlorine atoms, and the difference between their supposed monoisotopic signals at m/z 160.8 and 195.8 is 35.9698 u (HCl). The simulations are for Si2Cl3 (m/z 160.8) and Si2Cl4 (m/z 195.8), which was the best match so far but the m/z deviation of about 0.03 is far to big and the isotope pattern does neither fit well.
Maybe someone is familiar with this contamination or has some idea about it.
Greetings and sincere thanks in advance,
Jandirk Sendker
Guess what, Jandirk, your spectrum in negative mode matches Fe(II)Cl3 and Fe(III)Cl4 near perfectly. I am willing to bet 2 cents on that.
Probably one of the most common LC-MS contaminates, dimethyl phthalate (I know the mass does not align, but it is one we have seen between 194-198 at various times).
Have you performed maintenance on your HPLC ? Esp the A/I rotary valve seal? When worn, they can catch and release all kinds of garbage back into the source. Dirty inline pump filters, solvent pickup filters and bottles do the same. Replace solvent pickup filters with new ones, replace bottles (or wash in 30% nitric and dry with pure Nitrogen gas) as a start.
Guess what, Jandirk, your spectrum in negative mode matches Fe(II)Cl3 and Fe(III)Cl4 near perfectly. I am willing to bet 2 cents on that.
Thank you all for your contribution, I am very sure that Hanghui is right. Iron would never had occurred to me but the simulation of FeCl3/FeCl4 also perfectly explains the tiny isotope peaks at m/z 193, 196, 198 etc. Thanks for that hint! The low relative intensity of m/z 196 should also rule out phthalides (For this time! I am sure William's hints will be useful at another contaminant search.)
It appears that the small HCl amounts from the samples suffices to corrode steel parts? Please share if somebody has general information or experience about this at hand.
Greetings & Thanks,
Jandirk
*I could not see the fragmentation patterns on my phone and just went on the mass as a best guess.
I agree, looks like iron and that is another common contaminate, esp when acids (HCl) are used. The stainless steel used in HPLC systems will still show iron contaminates in various forms (FeCl variations are certainly one of them) and now that I can see the fragment pattern would confirm it. Good job.
Bill is certainly right on ion contaminants in various forms. A contaminant ion at m/z 116.928 in negative mode was regularly observed on my LC-MS system (H2O/ACN w/0.1% formic acid as the mobile phase), and the accurate mass is consistent with the formula FeO(HCOO)-
I would agree with Hanhui. My first thought when reading your description was that you are introducing fairly concentrated hydrochloric acid into a stainless steel system. Admittedly it is a small slug of material but it could be enough to cause issues.
- Badges
- Science method