Application of CADET into optimization of normal phase HPLC gradients

[pklochowicz]

Dear CADET Community,

I am currently working on a project that involves performing and analyzing organocatalytic reactions in an autonomous manner. However, I am struggling with optimizing the HPLC analysis, which I would like to be completely independent of human intervention. The HPLC module I am using is rather classical, with a column size of 250 mm x 4.6 mm, and the eluents I am using are a mixture of hexane and isopropanol. Unfortunately, I cannot change the eluents since the other column with chiral filling is integrated into the module. The analytes are small (100-400 u) and, in most cases, non-polar (with an average amount of isopropanol used in isocratic experiments being 2%). My goal is to find the gradient that will allow for the separation of 2-4 analytes. The separation does not have to be very precise; rather, I care about its reliability.

I found CADET to be a promising tool that can help me, but as I am not a specialist in analytical and physical chemistry (I mainly work in the fields of organic chemistry and computer science), I have some questions that I hope you can answer:

1. Is it possible to build a model that can predict the retention time of a single analyte under a given gradient?
2. If yes, could you please advise me on how many preliminary experiments I would need to conduct to find the parameters for the absorption model? Also, should the data be from isocratic or gradient experiments?
3. What absorption model is suitable for such experiments? I have read that the Langmuir isotherm, for example, cannot be used in gradient conditions.

If these questions seem trivial or pointless, I would appreciate it if you could recommend some literature that could be helpful in this task, especially for experiments with classical normal phase HPLC. Most of the literature I have come across has been about ion-exchange chromatography or RP-LC.

Best regards,
Patryk

[j.schmoelder]

Hi Patryk and welcome to the forum!

@pklochowicz

Is it possible to build a model that can predict the retention time of a single analyte under a given gradient?

Unfortunately, it’s hard to give a general answer. In principle, such things are possible but there are some requirements:

• Can you accurately measure the concentration profiles of individual components at the column outlet?
• Do you understand the retention mechanism, including the influence of the buffer on the adsorption?
• Have you characterized your system (i.e. column parameters, periphery of the column)?
• and some more…

@pklochowicz

If yes, could you please advise me on how many preliminary experiments I would need to conduct to find the parameters for the absorption model? Also, should the data be from isocratic or gradient experiments?

First thing would be to measure porosities (bed and particle).
Then, you need to find out if dispersion and diffusivity play a role. Usually, on HPLC that’s not the case…
But if so, doing single component experiments can be useful.
Then, you need to select a suitable adsorption isotherm and find out how the buffer composition influences the isotherm.
Then, you will need to run different experiments to capture these effects and fit the isotherm parameters to those experiments.

@pklochowicz

What absorption model is suitable for such experiments? I have read that the Langmuir isotherm, for example, cannot be used in gradient conditions.

There are different models that can do this and it really depends on the system. For example, for ion exchange, the SMA isotherm is a good starting point. In your case, a Mobile Phase Modulator could also be suitable. Both are just modified Langmuir isotherms that one way or another allow the buffer to influence the adsorption process.

Hope that gives you a good primer and let us know if you need additional infos.

[alters]

Seconded on using the Langmuir model with dependence on a mobile phase modulator, assuming that the adsorption behavior is Langmuirian in nature.