I worked today with a CH3:H2O (1:1) mobile phase. 8 minutes after I injected 200 ul of distilled water, a ghost peak appeared. Then, I injected other 200 ul with a NEW mobile phase and the peak didnīt show up. If there is an impurity in the mobile phase, it should go through the column constantly, disturbing always the baseline, but it didnīt.
I would appreciate if you tell me why this happened.

Regards

Mariel

It might have been a dirty injector/needle. The question is whether the ghost peak is a reproducible phenomenon--John Dolan's Rule of Two--make sure the problem recurs before taking steps to remedy it.

In fact, the water was the last thing I've injected; previously I've been injecting dissolution samples (each one with a ghost peak at TR = 8min, interfering the main peak integration)and after that, 200 ul of dissolution medium ,finaly the 200 ul of water, always with the same extra peak. Only changing the mobile phase I resolved the problem. I don't understand how an organic impurity in the mobile phase (the only thing I've changed) can produce that periodical peak. It doesnīt seem to be dirty injetor/needle, does it?

I appreciate any of your help. Thanks.

Try injecting a sample of the mobile phase and see if the peak still appears.

If there is a UV absorbing contaminant in mobile phase, an equilibrium will be finally formed on the column and a stable baseline will appear. However, by injecting a sample dissolved in solvent other then mobile phase this equilibrium will be disturbed and a ghost peak will appear at the time corresponding to the RT of the contaminant. Depending on the conditions, the peak may be both negative or positive.

Here are some question to the anonymous message from August 15. Have you experience with those ghost peaks? In witch time range do they appear and how different have to be the solvent used to produce ghost peaks on this way?
I'm very interest because I have a topical problem with a ghost peak at Rt = 4 min.

what HPLC system are you using?

Here is another question for anonymous from August 15, 2000. If there is a UV absorbing contaminatnt in mobile phase, will consistent injections of the solvent other than the mobile phase clean out the column? Or will it only reduce the contamination to a lower level and never fully clean it out?
I have a ghost peak that shows up only in the mobile phase and diluent sample solution. The mobile phase injection exhibits the ghost peak at a consistent RT and increases as I run my spiked samples. Does this increase in area of the ghost peak disprove the possibility of contamination in the column by the mobile phase? And does it help to promote the possibility of a dirty sample loop or injection system?