Octanol in Mobile Phase

[DJ]

DJ, you are saying that the retention time depends only on the dissociation constant?
(Even less sure about some of your other statements, but too lazy to check into it).

No. According to popular LSS theory, one could conceivably predict retention time under any gradient or isocratic program if constants: logKw and S are known. Even if one had knowledge of pKa, I'm not sure how useful it would be for predicting retention time.

[HW Mueller]

OK, I didn´t see that your k is a retention factor, thought you are talking about dissociation constants. But now I understand this even less: If you know a retention factor at a certain conditions you can predict the rt at slightly different conditions in the same system? Everybody in HPLC has always done this.

[DJ]

Being able to predict retention times would have value for developing a method for a multi-component sample. Computer modeling may not be perfect, but I think the technology is sufficient to land one in the ball park.. as opposed to the trial and error approach ("randomized empiricism", as Uwe Neue would say) to method development.

Since regression coefficient "S" varies for different molecules, the "just flatten the gradient" approach doesn't guarantee improved resolution (or that order of elution will be preserved for that matter).

Drylab is a popular HPLC simulation program. I believe it based on LSS model, though I have not used it. Last time I checked, the license, registration, and decoder ring needed to run this software was a little pricey.

[HW Mueller]

It seems that the utility of predicting rt would depend on what you use as basis to this prediction. Is that all that Dry Lab does? Predict rt or k´ after one determined these empirically with one condition of a system?

[DJ]

I'm not familiar with Dry Lab software. I know it predicts retention, resolution, peak widths: all straight forward calculations.. However, it also spits out a hypothetical chromatogram, and this is something my calculator cannot compete with.