I am not only puzzled, but also confused. Did the retention time always change up and down, or sometimes only down? Did the rt variation (up or down or both) occur with all the columns and mobile phases tried? The corticoids were always present and never showed this rt variance?
Also, what is this 50mmol phosphate buffer + sodium phosphate?
Your analyte is a mixture? Did it give the same pattern of peaks under all conditions? Did the pattern change with rt changes under the same conditions?
Are you shure that the corticoids didn´t have a short enough rt to obliterate rt changes for them?
I don´t know anymore whether these were isocratic runs, if they were, why dont you premix the mobile phase and see what happens?
Another side remark: It is very puzzling to see many times, as here also, that people ask for help, but dont bother to answer all the question of people who are willing to assist.
One more question:
Do you have a possibly strong incompatibility between the pH/ion strength of sample solvent and mobile phase?
Lot of questions, I will answer them one by one:
In the past, with the original method, the retention time changed sometimes down, and then increased. Most of the time we only see a decrease. In my test runs, I only test 4 times 40 minutes, which always gave a decrease for my Polymyxine peaks (4 peaks), but not for the corticoid (which is inbetween the 4 peaks).
The rt variation occured with all the columns tested, but as said before with the mobile phase without the sodium
sulphate I didn't saw the decrease, but maybe the test time was too short. The runs were put on the HPLC, I will post the results on monday.
50mmol phosphate buffer + sodium
sulphate:
6.8g KH2PO4 + 900ml H2O, brought to pH 2.3, added 4.6g sodium sulphate and then added H2O till 1000ml.
(I really added sodium sulphate instead of the sodium phosphate I mentioned in 1 post, sorry for the confusion).
My analyte is a mixture of 4 polymyxines and a corticoid. I'm not sure what you mean with the pattern of peaks, but the peaks were more or less consitent within all the experiments, no dramatic shift from one peak to another.
As the corticiod is in the middle of the 4 polymyxine peaks, I had noticed the difference for the rt if it was shifting, but all the peaks around the corticoid were shifting, but not the corticoid.
It are isocratic runs, but in the future this method will be used on a more complex mixture which needs a gradient after the isocratic part, that's why we are mixing with the pump. Also, if it was a system failure, the corticoid should have been shifting.
About your side remark: I was pretty sure that I answered the whole question you are talking about, but after reading back, I see that there is a part of my post missing. I do apologize for this.
The sample solvent is 100% water, so the possibility of a ion strength incompatibility is available. I will check this on monday.
Thanks for your time
Ace
