Hello there, I will shortly be developing a new experiment for teaching HPLC in a second year undergraduate analytical laboratory. The new system that we will be running it on is an Agilent 1100 with 5-channel UV/Vis detector. We'll be customising it a bit by adding a valve to choose between one of two columns (one is a C-18 and I think that the other one has a nitrile stationary phase). We are trying to come up with some model mixtures that can be used to demonstrate what happens to a separation when you change your eluent, or the column.
What would be some good analytes to use (preferrably inexpensive to purchase) that would coelute on one of the columns, but be well resolved with the other column using the same mobile phase? Also, a pair of analytes that are very poorly resolved on the column, but have sufficiently different absorbtion wavelengths so that they can be resolved in the spectral domain would be useful.
I'm really a GC person which is why I can't pull model mixtures off the top of my head.
Thanks all!

Your best bet would be to contact the column manufacturer and run this scenario by them...

Just a thought off the top of my head...

Sunscreens or vitamins (oil soluble) could be fun as some of them can be tricky to resolve. Selectivity of a column with regard to these compounds really can be altered by using different solvents. I have seen elution order of some of the compounds listed below change radically on a single reversed phase column under different solvent systems. They could also possibly help with your second problem of quantifying poorly resolved species using spectral differences. Compounds such as eithyhexylcinnimate, octyl salicylate, benzophenone-3, butylmethoxydibenzoyl methane, maybe even old type sunscreens such as PABAs (p-amino benzoic acids) could be fun. They're cheap and if you're sneaky, you can probably call bulk manufacturers of these components and get them to send you samples (ca 100 gm) for free. But you'll probably have to be sneaky.

This just occurred to me...Watch out for butylmethoxydibenzoyl methane. It is a diketo molecule that undergoes a keto-enol tautomerization rather slowly. Both species can actually be seen and quantified (the diketo has a lambda max at 266 nm and the enol at 360 nm), though I've never been able to get them completely resolved.

Another thought...nitrile columns are not the most rugged on the planet. If you absolutely need to go with a NP column, maybe an amino column would be of greater utility, though I have no idea how one would act with regard to the analytes above. If you would like to compare the same mobile phase between two columns, you might want to choose two columns that are both reversed phase, but are different supports. Columns such as an Inertsil C18 and Supelco RP Amide C16 or even an old style kromasil column will likely give quite different selectivity when used with similar mobile phases.
Anyhow, feel free to contact me at christopher.judd@collabo.com or bmw327@optonline.net or by phone at 631-689-0200 ext 2807 and I'll be happy to share some info.

Best of luck!

Chris

James,
Perhaps approaching the problem from the other side might work. Take a guess at suitable compounds, that is two that might elude at different RT on the two columns, then force the mobile phase to coelude them. The other column will probably separate them.
Jim

I am a laboratory trainer for a pharmaceutical company. I use Asprin to train my chemists. I use a C18 column and follow the United States Pharmacopeia method for Aspirin Tablets. The only draw back is that it is Isocratic.
Glenn