Reverse Phase Ghost Peaks

I am experiencing ghost peaks in reverse phase lc applications which I believe are due to potential solvent mixing/outgassing issues during an aggressive fast gradient.


The current gradient I am running is a 100% aqueous buffer solution for 0.5 minutes post injection following by an aggressive ramp to 44.5 % acetonitrile-based buffer at 0.51 min for a 3.5 min hold followed by a ramp to 100% acetonitrile-based buffer over 1 min with a 5 min hold for column cleaning. UV detection is done at 260 nm w/ a bandwidth of 4 nm with a reference wavelength of 360 w/ a bandwidth of 80 nm


Following a 30 min purge at 5 mL/min of all buffer lines (to ensure purging of vacuum degasser as per instrument manual) I have completed multiple system blanks with the following results:

- A run of the normal method resulting in 2 peaks with retention times 3.232 and 3.979, areas 69.4 and 39.4, and heights 5.2 and 1.7 respectively

- A run with the initial buffer b ramp increased to 60 % resulting in 2 peaks with retention times 3.073 and 3.635, areas 56 and 50.4, and heights 5.2 and 2.9 respectively

- A run with the initial buffer b ramp increased to 75 % resulting in 2 peaks with retention times 2.977 and 3.372, areas 38.5 and 38.5, and heights 4.4 and 3.2 respectively

- A run with the column bypassed via capillary tubing with only one NEGATIVE peak at time 1.361, area 26.3, and height 2.2

In summary, runs with the column resulted in a larger second peak as the percentage of the initial increase in acetonitrile-based buffer also increased, leading me to believe that the peaks are due to the aggressive mixing during the protocol.


Peak shapes when the column is attached are consistent with the literature regarding outgassing issues with 2 "shark-fin peak shapes" where the leading edge of the peak is quite sharp with a linear decline back to baseline (although these peaks never return to baseline).

Unfortunately these peaks are not consistent whatsoever across runs so a simple blank subtraction will not suffice for this particular method

I was hoping to get some suggestions as to ways to confirm the proper function of the on-line degasser, add back pressure post column to prevent outgassing if that is in fact the issue, or any other methods that have worked for eliminating ghost peaks in the past.

  • Hello,

    According to your blank results, it seems that you may have 2 peaks that are stuck in the system somewhere. I am not sure if those 2 correspond to the peaks of your samples or not, but that might be what you are seeing. 

    This document may be a good source of information in your case  Ghost Peaks using ACN and Water 

    Also, we have a phone app that is called HPLC Advisor that can help you with troubleshooting various things and ghost peaks is one of them. I suggest you look at that

    Now as far as the issue goes, if you are going high organic, and you flush the system for some time like that (can test for 15-20 mins to start), do those peak reduce in height? You could also look at the rotor seal of your sampler, to make sure there is no scratches on it, that is a source of ghost peaks since the sample stays in the scratches. Another thing to try would be to prepare fresh new mobile phase and see if without the column you still see peaks (whether they are the 2 ghost peaks or that negative peak)

    You can also look into backflushing your column, but that is not possible with all LC columns, so you would have to make sure you can do it with yours without damaging it.

    Kind regards,


  • Thanks for the reply. I have backflushed the column directly before past runs with the peaks still persisting afterwards. I have also completed a 30 minute flush of organic buffer without the column attached recently before running a blank sample and the peaks persisted as well and did not change in size. 

  • do a flush of the system and in the software do some injection valve switch (right click on sampler and switch valve to bypass / mainpass) a bunch of times.

    If it is still having ghost peaks after that, then I would look at my capillary connections throughout the system. 

    If you do not see ghost peaks or they dropped significantly, look at the rotor seal to make sure it is not damaged and replace it if needed

Was this helpful?