Suggestions on 'scout gradient' and solvents for organic lab

[seravitae]

Hi guys,

I have inherited a Varian ProStar HPLC setup (with PDA) for my organic chemistry laboratory. I will admit that while I have decent experience with flash and thin layer chromotography, HPLC is a new frontier for me.

As far as I am aware, all of our HPLCs around campus run acetonitrile/water or methanol/water mixes with C18.
However, our laboratory does not deal with peptides, and most of our products are isolated via hexane/ethyl acetate to methanol/chloroform mixes on normal phase silica.

We do have normal phase and C18 bonded columns for the ProStar.

I want to set the HPLC up as a 'scout', ie, a fairly generic profile as a 'next step up' from TLC, hopefully covering most or all of the range of solvents we use for flash isolation.

I was wondering if anyone had advice in this situation as to what phases and solvents I should use in this case. Should I stick with the generic C18 acetonitrile/water mix? Solubility might be seriously against us in these very polar conditions, and also makes it difficult if the chemist must purify on NP silica but perform HPLC scouts on C18. That said I have also heard that it can be dangerous to run HPLC's with fully organic mixes due to static, would I need to ground my columns for hexane/ethyl acetate/normal phase? [apparently this is not an issue in RP due to the high water content making it very difficult to ignite an acetonitrile/water mix.

Any advice as to what direction I should go would be most appreciated.
Kind regards,
Sebastian

[Consumer Products Guy]

I'd guess 90% or more of HPLC is done on reverse phase these days, I start there.

[tom jupille]

If you want to use the HPLC as a pilot for flash, you need to use a columns whose chemistry is as similar as possible to that of your flash columns. If silica gel, try to match the pore size distribution and specific surface area. The most general screening approach would be to run gradients starting with 0.1% isopropanol/hexane and running up to 100% isopropanol (IPA is about as polar a solvent as you can get and still maintain miscibility with hexane). If everything elutes at the beginning of the gradient, repeat with less polar solvents (e.g. THF, MTBE, EtOAC, MeCl2, etc.). The catch to normal-phase gradients is that (unlike reversed-phase), eluant strength is a non-linear function of composition; you get (very roughly!) about the same strength change going from 0-1% as from 1-10% as from 10-100%.

Another (and to my mind potentially more useful) use of the HPLC is as an orthogonal analytical technique using reversed-phase conditions. Because RP and NP have different retention mechanisms, it is very likely that coeluting peaks in one system will separate in the other. Thus, you get a better picture of the purity of your NP fractions if you analyze them by RP. In that case, the approach would be a "conventional" gradient from 5 - 100% acetonitrile in water. For analytical purposes, solubility is usually not that critical a problem. You would typically be injecting a few micrograms of sample in 10 microlters of solution. The biggest catch is miscibility; you would probably have to dry your fractions and redissolve in something that is water-miscible for RP chromatography.

[unmgvar]

flash with Si is still the cheapest way to do fast and good cleaning of your chemistry trials
HPLC is a very good way to get the best answer for purity after flash but i see many organic chemist with too little knowledge to correctly do this
one first mistake is see is that they overload the sample signal and think the %area result they get is correct

also with most flash glass columns and also many flash instrument 60-40u particle size is used and this cannot compare to 5u of the HPLC column. so i would stick to TLC for development of method for the flash system

remember the most important factor, as the chemist you need to know your compound
those you started out from and those you expect to get to see
your good level of chemistry will let you have little impurities and so chromatography will be also easy to solve

[Hollow]

If you want to use the HPLC as a pilot for flash, you need to use a columns whose chemistry is as similar as possible to that of your flash columns. If silica gel, try to match the pore size distribution and specific surface area.

or at best, have an analytical column packed with the same material as you use for your flash columns. Either pack by yourself or ask some column-supplier for assistance.
Doing so, you should be able to develop your separation on small scale, with less solvent consumption and then scale and transfer to the bigger scale.

[seravitae]

Thanks for the comments guys, I really appreciate it.

I guess I should clarify, that I wasn't planning to use HPLC as a replacement for TLC, just to confirm purity on a more sensetive scale. Usually analysis by TLC (with MS/NMR as well) is enough for intermediate molecules. Our products end up in biological systems, so there is the potential desire to do semi-prep or prep runs (we have columns and kit for these) on final products for bioassay. In that respect, I think i might set it up in normal phase with hexane/IPA by default, as in RP solubility will definately cause issues on semi-prep/prep scale.

Sorry I forgot to mention the desire of semi-prep/prep, which is obviously a pretty big factor in choice.

I should also point out regarding matching silica dimensions/porosity - that we would not use the HPLC as a strict guide for flash seperations, that is, we would not grab a HPLC trace first and use that to decide flash conditions, we would always run a TLC first, which has similar/identical silica grades.

I can always switch to RP if we need to do very polars or peptides, but we dont do this often enough to keep it in RP mode.

I'll keep in mind the issues about nonlinearity of organic mixes.

Edit: removed comment about using Hexane/Ethanol, turns out my solvent miscibility charts rule miscible solvents even if they are only partially miscible.