Chromatography Forum

Hi.

Sorry in advance for the long winded explanation...

I'm trying to separate about 12 known impurities from a two component drug product. After some method development I decided on developing a gradient method using ion pairing agents. Below is the full chromatographic condiitions.
Everything started going well during validation, multiple columns and multiple systems were reproducible. Then all of a sudden we started injecting LOQand resolution and we were getting a TON of noise making integration difficult.

Now I have a very concentrated solution of one active and reasonable concentration of the other (65,000ug/ml compound a, 500ug/ml compound b). Consequently, the placebo load as you could imagine is quite high.

I've heard that making gradients with odsa Is not recommended however I didn't have much of a choice. I needed enhanced selectivity and the main component and impurities are ketones and according to what I've found online exhibit extreme fronting due to gem diol formation and hemiketal formation.

I also noticed that when sample was injected prior to system suitability that the noise seemed to reduce. I tried making 3 quick injections of selectivity soliton followed by the normal gradient and then followed by 1 more injection of selectivity and then start into my system suit. When I did this however I am only seeing solvent front and component a, then no other peaks. Noise seems to have been reduced but again I don't see anything except solvent front and gradient baseline shift

1. Out of general curiosity, was my approach towards making a gradient method good? The idea was to as accurately as I could get the SAME concentration of ODSA in mpa and mpb? I was noticing when I first was playing around with this that I was getting some baseline noise but it was tolerable but then recently something happened and now I peaks are masked by an extremely noisy baseline.

2. In my attempt to "condition" the column, could I have gone overboard and that's why I'm seeing no peaks (other than the 65,000ug/ml component)?

3. As a side note, during the development we had also observed that different manufacturers of ODSA gave slightly different selectivity. This was confirmed by a colleague of mine and I was wondering if anyone had any ideas about why this is? We ended up during validation to stick with the same manufacturer of ODSA, and had opened a few different bottles from that same manifacturer but then we opened up smoke new one and we started observing the baseline noises.

Stock buffer solution: 10grams octanesulfonic acid, 2.5 grams potassium phosphate mono basic , 4ml tea per liter ph 3.0

Mobile phase a: 460ml buffer: 500ml water: 40ml acn
Mobile phase b: 460ml buffer:40ml water:500ml acn

Chromatographic conditions:
Injection volume 10ul
Flow:1.5ml/min
Temperature:50c
Column: Luna c18(2) 100x4.6, 5um
Wavelength: 235nm

Gradient:
0->15 minutes 100%a
15->16minutes 93:7
16->25minutes hold 93:7
25->45 50:50
Return to initial

I don't see much that needs correction. The only thing that could be a problem is that the formulation could build up on the column under these chromatographic conditions and then dumped into the injection. Try reversing and washing the column with 50% IPA.

Then add a washing procedure with 50% IPA after every injection. Then re-equilibrate the column with Mobile A before the next injection. Always use HPLC solvents.

hi Geof235:
can you attach some chromatograms for illustration , i am sorry , but i think chromatograms will explain the problem more easily and so we can think with you

In my opinion, the best approach for gradient with IP reagent is not to use ion-pairing reagent and explore mixed-mode chromatography where IP is attached to the surface of silica gel

No need to dedicate column to be used with IP reagent, much shorter equlibration and full compatibility with LC/MS and prep chromatography.

Please contact me if you want my help (http://www.helixchrom.com)