Nov 30, 2009

RBF Teaches: Column chromatography

All right! Listen up kiddies,

Column chromatography is much the same as TLC. Except this time the solvent front is forced down instead of being drawn up. All of the principles remain the same. So let's get straight into how to set up a nice column that should provide you with great separation of your compounds and be a great aid to you in your (awesome) organic life. The guide here is for a simple Flash column (forced with compressed air) but is translatable to a gravity forced column (which isn't really used much anymore by the stressed PhD student with a slave-driving supervisor). Aso guys, bear in mind that as you do more and more columns you get a feel for them and they do become like second nature.

IMPORTANT NOTE: Silica dust is toxic to inhalation (similar to asbestos) and all measuring of silica should be done in a fume hood and FOR GOD'S SAKES DON'T BREATHE IT IN!

Step 1: Choosing the column

There is a rule-o'-thumb in the lab that goes something along the lines of if you have a good separation (on TLC) of your desired spot from the undesired components you can use a ratio of ~20:1 (silica:mixture) or if its much more close running then you increase the ratio to ~100:1 (this should separate even the closest spots). The easiest way to measure out your silica is calculate what you need first then weigh out approximately what you need (in a clean conical flask/beaker), weigh it and so forth. If you have a bit too much or too little you can probably get away with it in the end anyway. But bear in mind the supervisors don't like forking out money and silica is exy $$$ (unless you buy cheap Chinese stuff like us :P). [MSG edit: With practice and experience, choosing a column becomes second nature.]

This leads us to the important bit: the actual column. So APOC - a book I regard as a mini lab bible for undergrads as well as honours students (Read Chapter 11.6.2)- recommends choosing a column that will fit the silica you want to use up to a length of 18cm. This is a good guide but for me I basically like to do the following:
  1. Take a long pasteur pipette and attach a rubber dropper to it.
  2. Insert it into a column (that looks about the right length and width to fit the silica) and put it in until your fingers (holding the rubber dropper) are at the lip of the column. Mark the tube with a marker pen.
  3. This is the minimum height that your silica must reach. The maximum height is about 3cm from the bulb (solvent reservoir) of the column (if it has one) or the same distance as in step 2 if it doesn't.
  4. Pour in the silica dry and tap the column gently to settle the silica. Check the height of the silica if it is above the line in step 2 you are good to go if not just pour in a bit more silica until you reach it. If it is way off grab a different column. (Sometimes you have to make do with what the lab has to offer). I suppose I should mention here that if you have a column that has a glass frit then YAY and if not then you have to stuff some cotton wool in from the top of the column into the bottleneck of the column just above the tap.
[MSG edit: everyone has their own way of packing the column. I don't do any weighing or anything but again, with experience, you will get a feel for the amount of silica required to achieve a given separation using a given solvent system.]

Step 2: Preparing the column

This is a bit tricky but I am sure you will be able to do it if you take a bit of time to read through the instructions and don't rush into anything, BUT, don't be afraid to try it. It isn't terrifyingly difficult.
  1. Once you have the right amount of silica to fit the column to the minimum height pour the dry silica into your conical flask.
  2. Using the solvent system you figured out from your TLC pour enough solvent onto the dry silica and swirl the flask/stir with a clean glass rod until you have a free moving slurry. By this I mean you want the silica to move completely without lumps. In this case more solvent is better than too little as you are going to pour it so think of it like pouring soup (you would find it hard to pour a thick soup right? and yes that is rhetorical)
  3. MAKE SURE YOUR COLUMN IS (vertically) STRAIGHT (in both dimensions) and the tap is closed. Pour a small amount of the solvent into the column (3-5cm is plenty) and then add a small amount of sand (about 1cm thick) and tap the column to get it to settle nice and flat. Wash any sand that sticks down with a bit more solvent (use a pipette). [MSG edit: Alternatively you could put the sand in first then add solvent so that you don't have to wash down the grains of sand. Give the sand layer a few taps with your hand afterwards to liberate any gas bubbles.]
  4. ONCE AGAIN, MAKE SURE YOUR COLUMN IS STRAIGHT then pour the slurry (for heaven's sake use a funnel) into the column. [MSG edit: make sure your solvent level in the column is high enough so that the sand layer isn't disturbed when you pour in your silica slurry.]
  5. Open the tap and pour all your slurry in but don't overfill the column. REMEMBER the tap is open so you will have some solvent dripping through. In this case I usually have my solvent mixture bottle underneath the column but if you haven't poured all the silica into the column you can put this flask under the column collect a bit more solvent then pour it into the top of the column again.
  6. Once it is all in, gently start tapping the column. I use a piece of metal rod wrapped in rubber hose but you can just use your hand or some rubber tubing. Whatever makes you feel good.
  7. As I said earlier this is Flash chromatography so you can whip out your gooseneck and attach it to a compressed air line and gently OH SO GENTLY push compressed air into your column forcing the solvent through the silica much faster. I personally keke-clip my gooseneck to the top of the column have a long compressed air line with a three way tubing adapter halfway along that allows me to regulate the flow of air with a thumb gently covering the open port of the three way but MSG likes to regulate the air flow with his hand holding the gooseneck. I like my way for sitting down and I haven't had a problem when I use this correctly but neither has MSG. Too much compressed air into the column is not a good thing! [MSG edit: Running flash columns are one of the safer operations done in an organic lab, but you still have to be careful. One of the obvious dangers is exploding columns (because after all, you are forcing air into a semi-sealed glass tube). So whatever method you choose, make sure you're comfortable with it. Safety first! Personally I regulate airflow with my hand. Apart from the fact that you probably get more tired using this method, it is superior in almost every other way. So even if you find this method more difficult, I would recommend running the column in this way. As a student you're here to learn, and so don't worry about making mistakes or stuffing things up: it's all part of the process!]
  8. As the solvent is pushed through the silica is also packing into a highly ordered form and if you have done the previous steps correctly the silica should form a flat surface at approximately the mark you made earlier. HOWEVER, you DO NOT want to push all the solvent through, you want to have a good amount of solvent covering the silica at this point. If you see the solvent is running out then use a pipette to add some more, it is ok to shut the tap of the column as well. You may also have some silica stuck to the sides of the column and while it doesn't really matter you can wash this down with a bit of solvent (using a pipette).
  9. Finally, with about 3cm of solvent remaining above the silica level you can add about 0.5cm of sand to the top of the column (THIS IS OPTIONAL BUT ADVISED). This just prevents the disturbance of the nice flat silica (you want your compounds to all start at the same point as in TLC) when you pour in more solvent. [MSG edit: I usually do this step AFTER loading the column (see below)]
Step 3: Loading the column

This is quite easy to do but don't go to sleep yet.
  1. After adding the MINIMUM amount of solvent necessary to dissolve your sample (generally you use the same solvent as the column but sometimes DCM is necessary) take a tiny amount for TLC comparison. [MSG edit: if you do use DCM, don't be surprised if you don't get as good a separation as you intended. When dissolving your sample, use at least the same amount of solvent as your sample. That is, if you have 1g of sample, use at least 1mL. This is more relevant when you have a liquid as your product. On a volume ratio, if the amount of solvent is less than your sample then technically your solvent is dissolved in your sample rather than the other way round]
  2. Allow the solvent level in the column to reach the same level as the top of the silica.
  3. Using a long pasteur pipette very carefully transfer the compound across. (Now do you see why we made the minimum height mark?) It is wise to load the sample on by running it down the side of the column but do this as close to the top of the silica (or the 0.5cm layer of sand if you have it).
  4. Once you have added all of your sample run the column down to level with the top of the silica once more and rinse your product flask with the solvent.
  5. Add this to the column again and repeat step 4.
  6. Very carefully add ~1cm solvent to the top of the column using a clean pipette and repeat step 4 once more.
  7. Carefully add solvent to the top of your column with a pipette to ~5-10cm then you can try gently running the solvent down the edge of the pipette to prevent the disturbance of the silica.
  8. Once you have ~25cm of solvent in there you can carefully pour the solvent mixture into the column using a funnel until the column is full.
[MSG edit: After ensuring the silica level is completely even, I load my sample solution straight onto the top of the silica, without the sand layer. The most important part of getting a good separation is a loading of a thin even layer of sample. I add my sand layer after I am finished with adding solvent to my column. This saves having sand stuck to the side of the column, but actually it doesn't really matter anyway.
Step 4: Running the column

Now for the easy part. While it is probably unecessary at this point it is still prudent to place the first collection vial underneath your column. Sometimes compounds can just fly off the column and it is easy to lose them in your waste solvent beaker if you are not careful.
  1. Open the tap
  2. Using compressed air you can force the solvent through or simply wait for gravity to do its job. It is important in both these cases that the flow rate remains roughly regular. If you are using compressed air for Flash chromatography a fast flow rate is important for good separation. You want the solvent to be pouring in a continuous flow from the tip of the column not dripping.
  3. Change your collection vials at about 2/3 full (a rule-o'-thumb is each fraction is half the weight of the silica e.g. 20g silica column = 10mL fractions) and rinse the tip of the column with a small amount of the solvent mixture before exchange (or you can dip the tip of the column in the solvent in the collection vial if it reaches).
  4. I personally relieve the flow rate when I am exchanging collection vials but like I said the quicker the better.
  5. Run TLC's on the fractions (for larger columns TLC every 2nd or 3rd fraction to save a bit of time). I generally take 10 fractions then TLC and while the TLC is running I collect another 10. I am spotting the TLC plates while forcing the solvent through the column. (This comes with experience; some people are even more insanely awesome LabChimp)
  6. Analyse your fractions fully (when you find a spot in your "every 2nd fraction TLC" it is often necessary to test some in between fractions (e.g. spot appears in 8 ends in 12 but need to check 9 and 13) against the reference sample you took) and group similar fractions together.
Step 5: Workup

I know, I know the whole process of columning is the workup but I couldn't call it "rotavapping".

After you have fully analysed your fractions you can now put them in a PRE-WEIGHED RBF. Chuck that bad-boy on the rotavap and remove the solvent as best you can - sometimes the volatility of the product can cause problems here. Dry your compound, either under high vac. or in the case of the undergrads here under a stream of N2. Then weigh your product, run NMR do the science thing and go home satisfied you did a column yay!

Once again I am making a video of this WOOPEE! and some nice pics might happen to add in here one day if I get a chance. Hopefully, MSG and I can comment on this one too once it is done.


  1. Don't listen to MSG when he says it is superior in almost every way. He forgot the part where I can sit to run my column and he has to stretch his arm all the way to the top of the column. But for beginners MSG's method is best. When you are more confident and experienced then you can take a load off. Better to be safe then covered in shards of glass.

  2. it's well written and good information. Only thing you have missed is about dry loading. Hope to see pic or videos soon though its already a month, Hehe

  3. Specific application of column chromatography in pharmacy. how column chromatography useful in separation and identification of natural products.can u explain with examples?