Water related peaks in blank injection

[Vivian Huang]

We have one RC method used for two sites. Blank injection from one site is always clean baseline while baseline from another site always has three unknown peaks. One of the peak eluted at the same retention time as active. We tried different HPLC system and different columns, the same senario happened. Once we moved the clean baseline mobile phase to another site and run on their system and column, the baseline was clean agian. All the mobile phase was prepared from pre-treated water and filted. ALl the reagents were kept as the same bottles. We are suspecting water is the cause of this problem. How can we prove it and what can we do to resolve the problem? Thank you!

Mobile phase A: 98% phosphate buffer and 2% of ACN, B: 60% buffer and 40% ACN, Gradient from 100% A to 100%B. Wavelength: 210nm

[Peter Apps]

I assume that you mobile phase is water plus buffer plus organic.

At one site all these materials are clean, at the other at least one of them is contaminated.

Take a batch of clean components (i.e. not made up into mobile phase) from the clean site to the dirty one. Make up sets of mobile phase using all the combinations of clean and dirty until you find the culprit;

clean water plus dirty site buffer + dirty site organic
Clean water + clean buffer + dirty site organic
dirty site water + clean buffer + clean organic

etc etc

Good luck

Peter

[Uwe Neue]

It's a gradient... Therefore the quickest way to prove that the stuff comes from mobile phase A is run a bunch of blank gradients while varying the equilibration time with A from run to run (0 minute, 5 minutes, 10 minutes). If the peaks increase with the pre-gradient equilibration time, they are somehow associated with the A mobile phase... the water, the lines, the pH meter etc...

[AA]

Uwe is exaclty right. I will say that when I have to prove tis to customers, I usually do a 5 minute at 100% water (or 99%, some columns dont like 100% water) and then leave one running at 99 or 100% over lunch over lunch and then run the gradient, if its the water (it almost always is) you see it has a HUGE difference in the mystery peaks.