Two different columns with different functional groups

[HippyLabRat]

I'm currently attempting to develop a method where I can simultaneously detect two fat soluble vitamins (Retinyl Acetate, Tocopheryl Acetate) and a highly water soluble vitamin (Ascorbic Acid) in a single injection. I found a research paper that claims to have done this very thing isocratically on a C18 reverse phase column, clear back in 1999: [Journal of Pharmaceutical and Biomedical Analysis 21, 399-406; Paulo, Marques, Morais, Almeida; University of Lisbon, Portugal]

I can't replicate their results on my Agilent 1290 with my Agilent Zorbax RRHD 2.1x100, 1.8 micron particle column. All of my ascorbic acid gets dumped with the void. They recommend using 100% Methanol as their mobile phase. I've also tried combinations of more non-polar solvents (ACN, IPA, MeOH) with no success at slowing the Ascorbic Acid. And can't get results that appear anything like what they have from the paper.

My questions:
1) Can I use a short column (50mm) with a polar stationary phase (Amide, Nitrile, or some type of HILIC column) inline before my C18 column? 2) Could the 1290 quat pump handle pushing though two columns? (Is this any different than running a single 100 or 150mm column?)
3) How different can the two columns be with regard to particle size, inner diameter, etc.?
4) I can't sem to find UHPLC steel column-to-column unions to do such a thing. Do they exist?

Thanks for your time.

[tom jupille]

Ok, a couple of preliminary comments:

First of all, there have been well over 500 different C18 columns marketed over the years. The differ in things like pore size distribution, suface area, surface coverage, silanol activity, and bonding chemistry and are *not* generally interchangeable. If you are trying to reproduce a published method, you have to start with the *exact* column used. And even then, many published separations are run on variously aged / deteriorated columns and are thererefore difficult to reproduce.

Second, if your problem is non-retention, you need to add more water to your mobile phase. Water is the weak solvent in reversed-phase, so by going more non-polar you are moving in the wrong direction.

Now to answer your questions:
1. Yes, you can couple two different columns together.
2. Yes, a 100 mm + a 50 mm column is equivalent to a single 150 mm column as far as pressure and flow are concerned (assuming same diameters, particle sizes, etc.).
3. As far as particle size goes, it really doen't matter except to the extent that you will be throwing away any efficiency benefits of the smaller particle size column while still paying the price in pressure. As far as diameter goes, you really should match the in order to keep a consistent linear velocity.
4. A column to column union is simply a short stub of tubing with nut/ferrule assemblies on each end. There is nothing special to make, just keep the tube narrow and short.

All of that said, I suspect that what you're proposing will be a black hole that sucks in your time & effort. If this were my problem I would look at a gradient, a mixed-mode column, or ion-pair before trying coupled columns.

[Gerhard Kratz]

I agree 100% with Tom. My recommendation is to check out application books from different manufacturers who published methods to run water and fat soluble vitamins in one run, on one column. But it is challenging. Good luck.

[HippyLabRat]

Thanks, you guys. You were HUGELY helpful.

[HippyLabRat]

One more question:
Does anyone know what the "proprietary functionality" is in an Agilent SB-Aq column?

Scoured the web, and can't find any information other than Agilent's which doesn't give any specifics.

[tom jupille]

Most "AQ" type columns use an embedded amide or carbamate a few carbons up the chain from the link to the support. I don't know about that specific one.

[HippyLabRat]

One more statement:

I've tried grandients ranging from 1-10% water, and included Methanol (20-50%), Acetonitrile (20-40%) and Isopropanol (0-30%).
The Retinyl Acetate and Tocopheryl Acetate elute reasonably, but the Ascorbic Acid still just dumps in the void no matter what combination of solvents I use. I've used flow rates from 0.75 to 1.0mL/min. (I think I ran a 0.5 mL/min run Monday, but have yet to see the results.)

[HPLCaddict]

Don't use plain water, you have to be acidic to protonate the ascorbic acid. Only in the protonated, neutral state you have a chance to retain ascorbic acid by reversed phase. Even then, you have to stay low with the aqueous amount of the eluent.

[Creal]

Additionality, ascrobic acid is very polar compound, so anyway the retention on C18 column will be low. Especially on Zorbax RRHD which I'm assuming have very good endcapping. I don't have access to read the originally paper but I suspect that they used old type column with worse that Zorbax RRHD endaping so there was a possibility to retain ascorbic acid with silanols activity. Also I'm agreeing with statment above, ascorbic acid has to be protonated in order to retain on column C18 type.

[Vlad Orlovsky]

We have developed a column which separates fat- and water-soluble vitamins in one run. The column should be available for purchase in January 2015. There are 4 mechanisms of interaction you can explore on this column at the same time - reversed-phase, HILIC, cation- and anion-exchange, which basically covers all vitamins. I am working on a newsletter, which will most likely be published in December. Contact me if you want to see preliminary data.

Also your particular set can most likely be done on Primesep D or Primesep SB column, which is RP/anion-exchange column. You fat soluble vitamins will retain by RP and ascorbic acid will retain by anion-exchange mechanism. Both columns are commercially available since 2004.

[Vlad Orlovsky]

Just managed to develop a method in less than 2 h for Vitamin C, Vitamin A and Vitamin E . Will post it here soon.