I would suggest you to soak the silica with your solvent (used for making slurry) for at least 2 hours. This will remove all the air spaces and fill all the voids which are arising during instant use. When you add the slurry into the column, add it in portions such that silica and solvent both go simultaneously inside. Let the silica settle and open the knob to remove the solvent from below till it reaches the level of silica, then again add remaining silica in the same way. I hope you understood. Just keep the silica submerged in the solvent while in the column. Use a long glass rod and use it for vortexing manually from above. I mean just stir the solvent above the silica and you will see the silica is vortexing and settling as a pile at the uppermost end. Now add your sample.

Sample is also to be prepared in silica if possible. I mean take the solution of the sample in a china dish and add some silica in it. Allow to stand for about 30 min for silica to adsorb the sample. Then allow to dry to get a final consistency of free flowing silica. You can now add the sample. When done with this, again follow the vortexing technique. Finally keep cotton over the sample bed to absorb excess solvent. If there is still excess solvent, use a pippette to remove it. In this case, tapping is not required in the whole process because the column is packed itself by gravity and no air spaces are present for the column to break.

Run the column as usual. If bands are separated it is well and good, otherwise decide a volume, say 50 ml, and collect fractions of this decided volume. Test each fraction on TLC for its profile and pool the similar fractions. Hope you will get single spots in some fractions. Collect and pool such fractions together and concentrate the fraction and crystallize it by keeping in a refrigerator overnight.

Hope this helps.

Guc Luck and regards

if the bands are getting slanted over period of time or you are changing the mobile phase which some how is changing the stationary phase as well. I would check the level of column using the bubble leveler. probably somewhere during the packing the column got packed incorrectly with tight or loose. repack the column. what else?

ProperPacking of the column is prerequisite for running the column. While loading the silica gel column we used to keep the cotton at the bottom part of the column and mix the silica gel with the hexane and add slowly through funnel at the top while opening the stopper at the bottom.

As mentioned above, check that the column is perpendicular to the ground when packing, add the silica as a slurry, you can tap on the column as the silica is settling to help it settle a little better. I use a cork ring for a round bottom flask for tapping because it isn't hard and I won't damage the glass.

To some extent, this is an issue with flash silica due to the relatively large, irregular particles which tend to create small channels where the solvent and sample in solution might more easily run. It takes time and patience to pack a column that minimizes band-tilting.

I would suggest you to soak the silica with your solvent (used for making slurry) for at least 2 hours. This will remove all the air spaces and fill all the voids which are arising during instant use. When you add the slurry into the column, add it in portions such that silica and solvent both go simultaneously inside. Let the silica settle and open the knob to remove the solvent from below till it reaches the level of silica, then again add remaining silica in the same way. I hope you understood. Just keep the silica submerged in the solvent while in the column. Use a long glass rod and use it for vortexing manually from above. I mean just stir the solvent above the silica and you will see the silica is vortexing and settling as a pile at the uppermost end. Now add your sample.

Sample is also to be prepared in silica if possible. I mean take the solution of the sample in a china dish and add some silica in it. Allow to stand for about 30 min for silica to adsorb the sample. Then allow to dry to get a final consistency of free flowing silica. You can now add the sample. When done with this, again follow the vortexing technique. Finally keep cotton over the sample bed to absorb excess solvent. If there is still excess solvent, use a pippette to remove it. In this case, tapping is not required in the whole process because the column is packed itself by gravity and no air spaces are present for the column to break.

Run the column as usual. If bands are separated it is well and good, otherwise decide a volume, say 50 ml, and collect fractions of this decided volume. Test each fraction on TLC for its profile and pool the similar fractions. Hope you will get single spots in some fractions. Collect and pool such fractions together and concentrate the fraction and crystallize it by keeping in a refrigerator overnight.

Hope this helps.

Guc Luck and regards

Thank you all so much for wonderfull guidance!!

Dear Mr. Pandey, As i understood, you mean to use hexane to make silica slurry with no matter my solvent sytem is??

I'm not Dr. Pandey, but I'd pack the column in whatever solvent I'm starting with.

Also note that "dewetting" happens with silica columns as well as with TLC plates. If you pack the column with hexane and you run a step gradient with ethyl acetate, you may see further air with the ethyl acetate gradient as the ethyl acetate gets adsorbed on the column. I'd consider packing the column with 1 or 2 percent of whatever "B" solvent I'm running to avoid this, unless there are compounds that elute in hexane (or whatever you will use for "A" solvent). I've seen this running hexane/ethyl acetate and dichloromethane/methanol.

Jack her response was directed toward one of the posters, it is a good suggestion to use 1-2% of solvent B to pack the column

Dear Sharad, But, my solvent system is isocratic.

Keep in mind that as silica gel dissolves in polar solvent it heats up, which can cause cracks and bubbles to form. When I was preparing my sample to load on the column I would put a very small amount of silica in with the sample in a round bottom, dissolve it and then rotovap so that it would homogenize.

@ Alex- you are describing the heat of adsorption for the solvent and is related to the "dewetting" I described earlier. That's why, if running a gradient, I'll often pack the silica in a small amount of the "B" solvent. If running isocratic, I'll pack the column in the solvent mixture. I'll add the solvent slowly when making the suspension- I learned this the hard way when I started in this business. I dumped dichloromethane in about a kilo of silica and the DCM got so hot it boiled and shot the silica out of the beaker.

I also found that solid loading, as you described, gives the best resolution. If you can get a plastic frit to fit your rotovap trap, it does a good job of keeping the silica in the flask while allowing a good vacuum (see attached picture).

Good tips :). Yeah, I did about a billion of these gravity columns back in college and never want to do it again. Not an easy technique!

I got a column with a ground glass fitting and solvent bulb on top, and had an older professor do some glass blowing to create a little manually adjustable flow controller that we would hook up to the hood's nitrogen supply in order to achieve higher flow rates. Not something you can keep consistent from run to run or even over the course of the separation, but if you were careful it definitely helped!

All the tips given are all good, but as Alex pointed out above, it happens rightly that silica gel dissolves in polar solvent and get the column heated up, this can results in cracks forming and bubbles. This heat and cracks can be stopped by wrapping a strip of tissue paper on the affected area where the cracks occur and constantly wet the tissue with acetone and the heating and cracks will automatically stopped.

Just one little "nit"- the silica isn't dissolving, at least in most solvents that we would run. It is adsorbing solvent and generating heat as a result, but isn't dissolving. If running the column under basic condition with high polarity (packing a silica column in 7N ammoniated methanol, for example) would cause it to dissolve, but we probably aren't discussing such solvent systems.

Pardon my ignorance, why would silica dissolve? It is a solid material.

I probably said it wrong, I've just noticed silica make it through the sand or cotton at the bottom of the column and figured with the heating/disruption of the structure that some of the silica is being surrounded by polar molecules and in effect being solvated.

I'm guessing small particles make it through, at least partially, through the cotton and sand. A very little amount can be solvated in methanol and water below pH 8 (above this pH it does dissolve), but this isn't enough to affect our purifications. I've kept high purity silica in LC-MS grade ethanol and 17.8 megohm water for a week, and still only found PPM levels of silica in the liquid afterwards. They do size the silica before packing, but small particles result from from mechanical agitation from shipping and even packing the column. Please note that bare silica is used in reusable HPLC HILIC columns.

Please note I'm limiting my remarks to pure silica, not Silica G which contains around 10% calcium sulfate.

Honestly my technique was probably just bad :P. I (thankfully) haven't had to pack any columns for a few years.

I wish I'd been able to see this comment thread back in 2006-2007.

the silica G is for TLC, cannot be used for column chromatography and should not be the topic of discussion here

Without saying anyone is correct for using it, or not using it, it seems silica G does get used; this purification used silica G columns: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0034945

This company seems to sell "column chromatography grade silica G": http://www.ashirwadchemicals.net/chromatography-grade-silica-gel.htm Note the name at the top of the specifications table.

Since people do seem to use silica G in columns, I merely state that my comments do not apply to that material.

As Alok Nahata told you, follow him. But along with it one more thing is YOUR SOLVENT or SOLVENT SYSTEM. Your solvents must be very pure and dried. There should be no moisture in column. For removing the moisture use Sodium Sulphate or any other dehydrating agent, which works best.

Pack the column using very low polar solvent Hexane and apply your metabolite in dried form mixed with silica gel. Try to use mid polar solvents to elute your metabolites. Then slowly increase the polarity. It may helpful for you to elute the metabolites.

Dear Nithya, before running the column we should dry the column and Silica (200-400 mesh size) at 70-80 degree temperature in hot air oven. Take dried silica in 500 ml glass beaker and make slurry with hexane. Before putting slurry in column, saturate the column with running hexane 2-3 times. Keep the column 85% fill with hexane and pour the slurry in fractions and tap slowly, silica will settle down after some time. After settling silica run it with hexane for at least 1 hour. Then After Rota evaporate your sample with silica (60-120 mesh size) dry it and make slurry in hexane and add slowly and leave it for settling down. Keep 10 % hexane in column and add your mobile phase. After following this procedure you will never face any problem regarding Column break and slanted condition during column chromatography.

All the best !