Suitable solvents for column chromatography

Why are 100% hexane or 100% ethyl acetate never used as eluents in column chromatography?

• You can. Why do you assume that you can't? – jerepierre Oct 12 '15 at 21:06
• I’ve done it. Many times. – Jan Oct 13 '15 at 7:58

Why, of course you can, but most of the time one of them would be "too strong" and the other "too weak", so neither would work well for you. That's why you have to mix them to the right proportion (which depends on your system, and is found by trial and error).

• Not "trial and error", please. Organic chemistry is still science! – Karl Oct 12 '15 at 21:11
• @Karl Trial and error is a scientific method. It’s how we arrived at today’s chemistry. There would be no Wittig reaction (just to name one) without trial and error. – Jan Oct 13 '15 at 7:59
• But you can do a lot better that trial and error for a column chromatography. Do TLC, check polarity of expected products. – Karl Oct 13 '15 at 21:10
• @Karl That is exactly what I understood by trial and error, and precisely what I assume Ivan meant by trial and error. – Jan Oct 14 '15 at 7:37
• @Jan Well, I think you're misusing the expression. You do one educated guess for a TLC solvent mixture, measure the Rf value, and adjust the mixture accordingly. Probably repeat the last two steps, then run your column. I see no possible "error" here. – Karl Oct 27 '15 at 0:55

Your comment is not entirely true. The solvent system used depends upon the behaviour of your (crude) product on silica gel.

Neat hexane (or a substitute such as petroleum ether or cyclohexane) is often used to wash 'grease' (non polar compounds) off the column, whilst neat ethyl acetate (or ether) is often used to elute highly polar compounds.

Ideally, you choose a solvent system in which the compound you want to isolate has an Rf of around 0.3 (please read the original flash column chromatography paper for more details), however, in practice, many chemists run 'gradient' columns. This entails starting with neat hexanes, and then slowly increasing the polarity of the solvent until the compound of interest elutes. This is also the way HPLC and automated flash (eg. CombiFlash/Flashmaster) works, using gradient elution.

• ‘many chemists run gradient columns’? I haven’t seen too many ‘many chemists’ then … – Jan Oct 13 '15 at 8:01
• Anyone involved in serious organic synthesis intuitively runs some form of gradient column if they're doing anything more complicated than a simple reduction or protecting group manipulation. Starting with a low polarity solvent to flush off any grease then slowly working up to something strong enough to elute the product. In total synthesis, this is very commonplace, though potentially not so much if you were doing methodology. – NotEvans. Oct 13 '15 at 17:33
• I can’t agree. I go looking for the solvent mixture that gives my product an $R_\mathrm{f}$ value of $0.3$ (approx.) and just go with that. Then I find pure product at around the seventh-ish fraction. And yes, I do natural products chemistry and yes, most of my colleagues here do the same thing. The only exception is if I really don’t know what happened and I get both unpolar and polar components. – Jan Oct 14 '15 at 7:36
• Running a gradient column provides additional understanding. To only focus on the compound of interest is to miss a trick. This would not be normal practice in an industrial setting whether the chromatography is normal phase, as in flash or biotage , or reverse phase preparative or analytical. Collecting and identifying the other components helps build up a mass balance or mechanistic understanding as required. – Beerhunter Oct 16 '15 at 23:56

Simply speaking you would often only use pure hexane (or better cyclohexane, as the former is somewhat toxic, or (cheaper) isohexane) to wash something completely unpolar off the column before further elution steps.

Pure ee you would use to just wash out everything that is still on the column after your chromatography.

Of course there can be the (i think rather rare) case where the polarity of either one is just what you want.