How do we get rid of/identify contamination with m/z 248.9737 (negative mode)?

The problem: We see a persistent background mass signal at about 249 (exact mass 248.9737) operating in negative mode. It shows up most clearly when we are running with acetonitrile, and does not show up in water.

Our system: Agilent BioInert LC connected to Thermo Fisher single quadrupole mass spec.

Locating the problem:

  • Connected the mass spec to a different LC and to a syringe pump – no contaminant seen in either case. The mass spec is not contaminated/the source of contamination.
  • Connected pump directly to the mass spec – still see it. Contaminant is not from samples run on the LC.
  • Took solvent samples from various points along the flowpath and injected them directly into the mass spec. Solvent directly from the solvent bottle is clean. Solvent drained from the capillary coming out of the gradient valve is also clean. Solvent taken from the purge valve exit is contaminated.
  • Tested solvent from the purge valve exit on a different mass spec (HRMS) to get exact masses. The contaminant’s main peak is at 248.9737, which breaks down into 183.0136 and 119.0507 at higher voltages. This is all in negative mode – there isn’t a corresponding signal in positive mode.
  • The mass breakdown does not match the profile for the most common contaminant that appears around 249, trifluoroacetic acid.

Things we’ve tried:

  • Replaced all solvent bottles and solvent filters, replaced one test line with Agilent ultra-clean bottle/filter/tubing kit.
  • Flushed pump overnight/over weekends with various solvents (none has had any effect on the magnitude of the contaminant signal).
    • 50/50 isopropanol/water
    • 100% isopropanol
    • 100% acetonitrile
    • 100% water
    • 20mM ammonium acetate in 50/50 water/acetonitrile (pH ~7-8)
    • 10% acetic acid (pH ~2-3)
    • 75/25 methanol/water
    • Agilent flushing solvent (50% IPA, 25% ACN, 15% cyclohexane, 10% DCM)
  • Replaced the pump head, pump capillaries, check valves, damper, purge valve (all parts in the pump after the MCGV).
  • Used a different manufacturer’s acetonitrile (we’d still like to try an ultra-pure ACN, ordering now).

It is bizarre to us that the contaminant signal is so consistent throughout all the flushes and part replacements (our noise right now is around magnitude E02, and the contaminant signal is consistently around magnitude E05). The most recent suggestion of our Agilent representative was a resource they found noting an exact mass of 248.9746, matched to a gold complex [Au(CN)2]- that can stem from impurities in the ACN reacting with gold parts in the BioInert LC like the gold seal(s). In that case they said that other brands and lots of ACN reduced the signal but didn’t remove it. Any help would be greatly appreciated!

Was this helpful?