Step gradient elution Vs Theoritical plates

[Ramachandra]

Hi all,
I have developed a step gradient method, separating 8 solutes in 10 mins on 25mm column containing mobile phase dilute acetic acid (ph:4.0) and methanol. I am getting less theoritical plates ~200 at initial (~ 20% methanol), ~40000 plates at middle ( ~80% methanol ) and ~15000 plates endding stage(~25% methanol. What is the reason for large variations in theoritical plates? whether these type of variations due to viscosity change of mobil phase during the run? whether the variation in plate count is common for all step gradents or only for my method?

[Hollow]

be aware that the theoretical plates aren't defined for gradient separations, and therefore not applicable!

in gradient elution the peak width stays +/- the same for every peak, and thats why the result of the plate count calculation will increase with retention time and just gives meaningless numbers

[Ramachandra]

Thank you Hollow,
For the same method, i have added different additives from 2mM to 10mM concentration. i have observed that the change in the retention factors of each analyte (2-10mM) is <0.5 only, not like isocratic method. Why i have not got large retention factors in gradient method? whether this phenomenon is common for all gradients or due to less additive concentration?

[tom jupille]

Gradients are different from isocratic. Just as the isocratic plate number formula can not be used with gradient data, likewise the isocratic retention factor cannot be used with gradient data. Look at it this way: your gradient is making a *huge* change in solvent strength which dwarfs the effect of your additive.

[Uwe Neue]

Also, I find your step gradients a bit strange: 20% to 80% to 25% ??????

[Ramachandra]

Thank you, tom jupille and Uwe neue for your constructive suggestions
To separate polar (logp<1.0) and non polar(logp>3.0) analytes on 25 mm column, i employed huge change in orgnic concentration.
could you please, suggest me right formulae for retention factor, asymmetry factor and resolution factor for gradient method. is there any formulae to check, the developed method is right one in gradent? (even though we got all chromatographic parametrs in acceptable range).

[Uwe Neue]

Peak asymmetry factor is determined from a single peak and can be used in isocratic separations, gradient separations and step gradients.

Resolution is good for two (commonly adjacent) peaks, and can be used in isocratic mode, gradient mode and step gradients. I would be leary about calculating resolution for peaks that elute in two separate gradient steps, but also do not see a reason why one can't do that.

Retention factors can be calculated for isocratic methods, gradient methods, and even for step gradients.

[Ramachandra]

Hi Uwe neue,
i think, you are differing the statement given by tom jupille " the isocratic retention factor cannot be used with gradient data". Even, I am not found separate option for gradient method to calculate chromatographic parameters in HPLC softwares like Class VP, LC solutions. Still, i have confusion on 'the change in the retention factors of each analyte (2-10mM) is <0.5 only'. Is it reliable change? if yes, is there any specific reason (like viscosity/equilibrium disturbance....) for less variations in retention factors in gradient elution with increase of additive concentration.

[Uwe Neue]

I am sure that Tom wanted to say that there is a difference between gradient retention factor k(g) and the classical isocratic retention factor. However, you CAN calculate retention factors in gradients.

The judgement about permissible changes in retention factor depends on the method, and the needs of the method. I would not put generic values on this, especially if you are comparing gradients and isocratic separations. For example, for isocratic separations, I would set limits on the %change, while for gradients I would specify absolute changes that are allowed.

May I suggest that you get yourself some good textbooks, like Lloyd Snyders various book (which includes one that is dedicated the gradient analysis)?

[Ramachandra]

Thanks to every body, who involved in this discussion.