Troubleshooting Flash Chromatography

• Problem: My compound is not stable to silica gel (if you want to find out if your compound is stable to silica, see TLC tip #4. How can I purify it by chromatography if it decomposes on silica?
• Solution: If the separation is an easy one, you could try purifying on florisil (200 mesh) or alumina. For more challenging separations, it is possible to deactivate silica gel (reduce its acidity) so that it is less damaging to your compound.


• Problem: Your compound should be off the column by now, but you're still collecting fractions with no sign of it.
• Solution: A number of things could have happened here.
  • The compound decomposed on the column and will never come off. Test your compound for silica stability.
  • You are not using the solvent system you thought you were using. Double-check the bottles you used to prepare the solvent system and consider whether you could have reversed the polar and the nonpolar element during preparation.
  • Your compound came off in the solvent front. Check the first fraction.
  • Your compound did come off, but the fractions are so dilute you haven't detected it. Try concentrating fractions in the range you expected see your compound, then you will be able to detect it.


• Problem: I cannot separate the components of my reaction mixture using column chromatography, even though there is a large Rf difference. All the fractions are mixed. I don't understand why this is happening.
• Explanations and Solutions:
  One reason for this behavior is that you are being misled by thin layer chromatography. You see two compounds, but one may be the degradation product of the other, and the degradation occurs on silica gel. During a column, this process goes on during elution, and all fractions have both product and degradation product. Check that your compound is stable to silica (2D TLC).
  Sometimes the wrong choice of solvent will also encourage this behavior- if the compound with low Rf dissolved well in the eluant, and the compound with high Rf does not, you may observe this behavior. Try to find a solvent system that dissolves both compounds well.


• Problem: My compound is very nonpolar and does not have an Rf as low as 0.3-0.4 in any solvent system. What do I do?
• Solution: If it's an easy separation, a high Rf might not be a problem. If it is, you will want to consider other purification methods. If the compound is a solid, you may be able to purify by crystallization, or you are working on large scale, you could use distillation. If none of these methods work, but your compound is mostly pure, think about the future of the compound. Are you about to convert it to a more polar species that will be easier to purify? Depending on the chemistry, sometimes it's possible to go on to the next reaction with your mostly-pure compound and purify successfully at a later stage.


• Problem: My compound does not have any major impurities in it and I hate chromatography. Do I have to run a column on it?
• Solution: It depends on the next step of the reaction sequence. The best answer is to try the next step without purification, on a small scale, to see how it goes. Another option is to run the compound through a short plug of silica, which will remove baseline impurities and call allow you to skip the fraction-collecting stage of chromatography.


• Problem: My compound is very polar: it does not run up the plate even eluting with 100% EtOAc. How do I purify it by chromatography?
• Solution: You could use reverse-phase silica for your column, or try a more aggressive solvent systems to move it off the baseline. Solvent systems containing ammonia can be useful in this context: prepare a stock solution of 10% ammonium hydroxide in methanol, and try using 1-10% of this in dichloromethane for very polar compounds.


• Problem: I have less than 50 mg of compound, how do I purify it?
• Solution: You can run a miniature flash column by packing a short (6'') pipette with cotton, sand and silica, just like you'd run a normal column. Choose a solvent system that places your compound at an Rf of ~0.2. You will have to add solvent to the column every 1-2 fractions, but it's simple to apply pressure with a pipette bulb, and you'll be finished in no time. This method works better than you might expect.


• Problem: Your compound takes FOREVER to come off the column, it starts coming off at a reasonable point in the elution, but it doesn't stop coming off until many many fractions later.
• Solution: When the compound starts coming off the column, increase the polarity of your eluting solvent. If there are no impurities at lower Rf, you can increase it significantly and avoid the annoying trailing effect. It is important to keep the same two solvents, and just increase the percentage of the polar component.


• Problem: I have two compounds very close in Rf, how do I separate them?
• Solution: Although it's not very much fun, the best way is to use a solvent gradient during elution. You begin with a solvent system in which your compound runs at an Rf of about 0.2 (or less) and increase the percentage of the polar component of the solvent system slightly each time you replenish the solvent in the column. You will need to experiment with degrees of gradient for each difficult case before you find conditions that give optimal separation. Some people find it's easier to run two columns instead of one. For difficult separations, you will often have some mixed fractions- rather than throw them out or contaminate pure fractions, keeping mixed fractions "for a rainy day" is recommended. You can always purify your mixed sample later, if you need it.


• Problem: Your crude reaction mixture is not soluble in the solvent system you plan to use to elute your column. You don't know how to load it on the column gracefully. (note: this happens most often with the ethyl acetate/ hexane solvent system, and usually is only a problem on large reaction scale.)
• Solutions:
  • You can switch another solvent system like dichloromethane/ hexane or acetone/ hexane.
  • A (rather risky) method is to dissolve your sample in very small volume of another solvent (like dichloromethane). There is a fine line here between success and having to repeat the column due to poor elution behavior.
  • If you suspect that a reaction byproduct is the cause of the solubility problems, you can try a silica filtration to remove the offending solid. Use a short plug and a solvent that efficiently dissolves your product but not the impurity. You may want to collect big fractions to achieve the best rough separation possible, before running the real column.
  • You can infuse your crude reaction mixture into a small quantity of silica, and load your sample as a solid. Pack your column as usual with solvent-infused silica, but do not top with a layer of sand. Using a large round-bottomed flask, dissolve your compound in an appropriate solvent and add enough silica gel to form a thin (not thick) layer on the top of your column. Remove the solvent by rotoevaporation. You should have a flask of free-flowing silica gel that you can pour on the top of your column. Add a layer of sand and elute as usual.


• Problem: Your compound has clogged the silica somehow- elution is very slow
• Explanations and "Solutions": This is not very common, but when it happens, it's a major headache. The problem arises when either the compound or an impurity crystallizes in the column, forming a solid barrier that blocks solvent flow. The prognosis is not good- console yourself by noting the problem so you can avoid loading the loathsome mixture onto a normal column in the future. You will either need to employ some pre-chromatography purification technique or use a very wide column and a lot of silica.
  • Take a thin wire and push up on the cotton plug from the spout of the column. If you're lucky, the blockage is at the narrow point where the cotton is and can be cleared.
  • Take a long glass pipette, rod or similar implement and try "stirring" the silica slurry in the column, just to get the solvent moving again. You'll have to start over to purify it, obviously.
  • If these methods fail, you'll have to pour the whole mess out through the top of the column and perform a crude filtration to extract your sample from the silica. It is possible this step will remove the insoluble material, and you can carry on with chromatography.