shorter column and smaller pore size?

Dear all

I have a very crude method with a Kinetex C18 100 A (100 x 4.6 mm, 2.6 um) running at 2 ml/min at 40'C with pressures ~ 320 bar. We have lots of samples running a day though this column with a 'quick' 10 min. scouting method and I wanted to reduce the amount of solvent used but keep the time as short as possible.

My question is what should I really change on the column - the length or the particle size? I assume changing the length would reduce the time and solvent but not improve resolution and if I reduce the pore size of the column, and length I could reduce flowrate but I am concerned about excess pressures - the instrument is not UHPLC.

Would anyone please be kind enough to suggest a smaller column (either length ID or both) that would be worth trying?

Would a Poroshell (699675-902 (1.9µm)) be suitable for non-UHPLC?

thank you all

  • Hi Mark,

    Welcome to the Agilent Community! I moved your question to the Consumables User Forum for better visibility.



  • Hi Mark,

    shorter columns mean less plates, in your case 100 mm to 50 mm would half the number of plates. In the resolution equation you have the factor √ Nth, that means half the plates = 30 % less resolution. 

    Looking at all dependencies of the resolution equation shows this:

    And your selectivity is fixed for your choice of phase.

    So, what now? Back to the Van Deemter equation (or Knox equation). You need to lower your diffusion factors to get less band broadening and therefore smaller widths peaks, that will increase resolution. Shorter columns is one way, so you might not loose 30% plates after all; another way is using superficially porous materials (and you already do this), smaller particles works also and then there is average linear velocity (flow). For superficially porous materials (and small particles) the results of above mentioned equations show that you can work outside the theoretical optimum flow and go up. Yes, you would loose some plates, but the positive diffusion effects might be a lot bigger. Even with a standard LC system you might want to test higher flows up to the max. pressure of the system.

    Also keep in mind all those changes work best if your peaks of interest elute somewhere between retention factors 2 to 5. Maybe adapt your gradient accordingly.

    You also see there is no easy answer as everything is connected and the best chance would be a better selectivity (phase).

    I still hope this helps a little.



  • Dear Norbert

    thank you sincerely for the excellent reply and logic/maths behind such variables; this will prove invaluable in furthering my limited knowledge on the subject.

    The key, basic thing I would be grateful for reassurance is if I went for a smaller pore size (such as the Poroshell 1.9 um instead of the current 2.6 um) would this be suitable for standard HPLC pressures? I'm very nervous to purchase this only for it not to be suitable due to having to reduce the flowrate to compensate for increased pressures and thus longer runtimes

    thank you

  • Theoretically, if you go from 100 mm x 2.1 mm x 2.6 µm to 50 mm x 2.1 mm x 1.9 µm, the pressure stays the same at the same flow rate and the same phase. Kinetex and Poroshell have a slightly different rho-value, but that would only do a +/-10% change in pressure.

Was this helpful?