I am running a gradient method for the determination of related substances of a pharmaceutical compound. The method was supposedly validated previously. The method is 40 minutes long with only water and organic-- no buffer or additives. However, late in the blank gradient, I am seeing a peak which interferes with a known related substance. This peak is small, but bothersome, nonetheless. It is also present independent of changes to organic supplier and batches of columns. I do not have the freedom to change the mobile or stationary phase chemistry, and small changes in gradient strength around the elution time to not enhance the separation of the interfering peak and the RS. a 5 degree change in temperature also has no effect on resolution of these peaks. I am looking for some suggestions to resolve this related substance. Am I missing something obvious?

I assume that the peak had not been observed previously. I would check if the source is the water. Change the equilibration time with the "A" mobile phase. If it increases with longer equilibration times, the source is related to the water supply (this includes tubing etc.). If it doesn't we need to scratch our heads again.

Does your organic have a stabilizer/antioxidant? We have run into problems with amylene in chloroform and BHT in THF. They both produced a peak in our normal phase and reverse phase analyses.

The gradient components are most likely the source of the peak. The water could have an organic impurity that will deposit on the column in high aqueous low organic MP. At some point the concentration of the organic washes it off. Likewise the organic may have a dissolved organic impurity that is insoluble under high aqueous, low organic MP conditions. Again it deposits on the column and is washed off when the organic strength of the MP makes it soluble. Try a 30 minute 100% organic flush. Follow it with a blank after allowing exactly 10 minutes equilibration at starting conditions. Allow the system to maintain the starting conditions for 1 hour. Run a blank again. After allowing the system to equilibrate at starting conditions for exactly 10 minutes after the second blank, run a third blank. If in the first and third blanks the interfering peak is small compared to the second blank, then the peak is depositing on the column from the starting MP. Try vacuum filtering your water and organic through a 0.2um filter. I would be willing to bet the filters would not be as white after filtration has been completed. I have seen HPLC, and better grade, acetonitrile discolour filters anywhere from light yellow to dark brown. Some of them have plugged 0.2um filters and I have had to use larger pore filters.

I'd be looking at the water. We are consistently bothered with an impurity peak eluting at ca 90% MeOH, no matter what the source of the water (purchased HPLC grade or >18 MegOhm RO/deionised water). The only way we can get rid of this peak is to run the water through the RP18 column first! The peak is small (typically < 1 mAU), but a pain in the neck when your analyte peak is of the same size.

Over time, these small nasty little peaks have a habit of growing in peak height and area counts. If their growth does not cause the chromatographer problems these insidious little peaks tend to move their retention times to those of your peaks of interest. They have so many different sources that I am beginning to believe that they are an alien life form invading earth from other planets.