Chromatography Forum

I'm using an HP 5971 with EI, DFTPP tune and I've been having a problem matching alkanes during my RI calibration runs.

Although the match probability is high with decane (97%), it decreases as the alkanes get longer, so that when I reach eicosane, the match probability is down to around 83%. The problem seems to be pretty consistent, that the abundance of ions 41 & 43 (C3H5-6) is too low by 20-30%, while the hydrocarbon fragments above this are too high, although not by much (<10%). Since the molecular ion is also too abundant, my guess is that for some reason my EI source is not fragmenting aggressively enough. My vacuum level is good, around 9x10-06, the background noise is good. Can anyone think of a reason for this?

I think either the tuning could be a little bit off, or the tuning conditions were not the same as the conditions in your calibration runs (temperature, electron energy, vacuum level, presence of air/water ..).

One thing you can try is tune and measure on the other filament (if there is one). Another simple task is to bake-out the MS parts: put it a couple of hours on the maximum temperature (source and quad), cool down, tune and measure.

I think either the tuning could be a little bit off, or the tuning conditions were not the same as the conditions in your calibration runs (temperature, electron energy, vacuum level, presence of air/water ..).

One thing you can try is tune and measure on the other filament (if there is one). Another simple task is to bake-out the MS parts: put it a couple of hours on the maximum temperature (source and quad), cool down, tune and measure.

The problem with the 5971 is the source is heated only by the transfer line. If you have the transfer line set to 280c the source will only ever read about 170C.

The other thing which would help solve the problem would require at least a 5972, and that would be adjustable emission current. With that you can bump up the uA current to the filament and gain more fragmentation to shift the spectra to the lower masses, but unfortunately the 5971 lacks that.

The best thing is to inject known standards and build a reference spectra for each that matches what the analyzer produces, this will compensate for the tune and lack of fragmentation. As long as you can consistently reproduce the spectra from a reference standard with those of analytical samples you should be ok.