Friday, April 04, 2008

Dry Column Vacuum Chromatography (DCVC) - revisited

I get a lot of questions about Dry Column Vacuum Chromatography (DCVC) so I believe it is time for another post on the topic. DCVC is really taking off and the paper now has >70 journal citations.
Firstly, let me reiterate that like most things in life DCVC is best learnt by doing. Some of the questions I get are very detailed and specific and I can't provide clear cut answers. With experience you'll have an idea what to do and you'll improvise along the way and get it right.
Q&A session with the true believers:
(1) ...did you know that your paper had the"honor" of being "html-ed" by Rhodium....
- I didn't know that. I guess that everyone is purifying their illicit drugs in the garden shed using DCVC these days.
(2) ...searching for a good source for the silica gel, right now it is about 4-5 times more expensive than our flash gel source...
- That is a huge difference. You have a very sweet deal on flash silica. In my case it costs about 40 % more but considering how little you use compared to flash it works out as a big saving. If you find a cheap source of DCVC silica please let us know.
3) ...In the comments, you mentioned that you have done DCVC with 50-100 mg, and I was wondering what size of fritted funnel (frit diameter) you would use for that. Just calculating the size using the approximate amount of silica gel gives me something ridiculously small...
- The smallest sinter I use has a diameter of 1 cm. I have run 20 mg columns with 5 ml fractions on this column with no problems. Smaller than that would be impractical.
(4) ...Also, just to help me chose the size [Column] if I'd get one or two funnels made, what diameter(s) would you use for say 500 mg to 5 g?...
- I currently have the following four columns: 1, 4, 6 and 8 cm. My favourite (that I use 80% of the time) is 4 cm. It's a good size to work with and it's good for 20 ml fractions. I'll do anything from 100 mg and up to 5 grams or more (depending on the separation) on that column. For your requirements I'd say get a 4 and 6 cm column.
(5) ...frit porosities - the notation 1 - 4 is what is used in Europe, here in the US they have C, M and F, which actually don't directly correspond to the 1-4 sizes. "3" is a size where there is no direct equivalent, which is too bad because it seems ideal also to have in a flash column. The actual numbers are below, and I was a bit surprised because it seems that the P3 should clog up with time when using the Merck 15111 silica gel with its size range of 15-40 micron.
American standards - (Kimble & Corning, ASTM) nominal pore size, in microns, Medium 10-15 µm, Fine 4-5.5 µm
European standards - (Robu & Schott, ISO 4793) nominal pore size, in microns, P3 (P40) 16-40 µm, P4 (P16) 10-16 µm
- I had no idea about all this. Why on earth can't we just standardise these things. Anyway, thank you very much. I have always wondered exactly what the pore sizes represented.
(6) ...What is a least polar couple of spots with delta Rf 0.05 (often called "eight", "8") which you would separate via DCVC?...
- I only do DCVC. If it fails the next stop is prep. HPLC. Fortunately it still hasn't failed. As with all normal phase chromatography super non-polar compounds separation sucks. That said I have often columned stuff (not super non-polar) that appears to be one compound by TLC and managed to get two compounds of the column. Behold the power of slow step gradient elution. I have achieved truly mind boggling separations over only 20 fractions. The worst column ever occurred 1.5 years ago. I kept getting a fair bit of each diastereoisomer clean but it took 5 columns to get it fully separated. Still it was easily done in one day by using the same column 5 times and only collecting 30 fractions per column.
(7) Do you always start from heptane even for very polar mixtures?
- Yes! It gets the stuff off the Celite on to the silica and wets the column so that it runs well. Generally I do 4 x hexane, heptane or pet. ether first.
(8) ...can I use toluene/ethyl acetate, chloroform/ethyl acetate or chloroform/methanol mixtures successfully?...
- Yes! You can even use really low boiling solvents such as acetone, dichloromethane, ether etc. but due to evaporation it is easier to work with higher boiling solvents. Where I work now hexane and heptane has been replaced with 40-60 petroleum spirit which I use for my columns without too much difficulty. Remember to have the pump exhaust in the fume hood, especially with the low boiling solvents.
(9) ...Does it work substantially better [with Celite] than preadsorbtion on silica gel?...
-Yes! Celite is easier to handle and it doesn't affect resolution.
(10) ...Does this trick with Celite work even if a sample is only sparingly soluble in an eluent?...
- It doesn't matter what your eluent is. Dissolve your compound in something polar that is easy to get rid of, for example ethyl acetate or methanol. Add Celite, concentrate in vacuo and load it on the column. Ensure that you have removed all solvent prior to loading. A lot of say methanol in the Celite will compromise resolution.
I hope that helps. If you are new to this area please read the previous post and check out the following paper (a copy can be supplied upon request):
Dry Column Vacuum Chromatography, D.S. Pedersen and C. Rosenbohm, Synthesis, 2001, pp. 2431-2434.
Let me know if you have more comment, suggestions or questions. D!


Ψ*Ψ said...

Sooo much faster than running a column! We use something similar, but with sidearm Erlenmeyers collecting the fractions. Is it that much faster with the adapter?

Daniel Sejer said...

Are you kidding? I'd say that it is at least twice as fast with the alternative set up. We used to use Buchner flasks for this but it was driving us nuts having to pull the thing apart and reconnect things again for each fraction. D!

Seansheep said...

The DCVC revolution is slowly arriving!

Joanna said...

You convinced me. I'm going to try it out!

Anonymous said...

I'm a big fan of this technique and have been using it upwards of 10 years (back then I was following a reference from the Journal of Chemical Education!). I've never gone much above 12g per run but it is good to hear the sky is the limit though I'd be interested to hear what size of column (sinter) would be required for a 100+g run. I'm currently making lots of primary and secondary amines and I am currently having difficulties with DCVC for this type of system since I have to use a phase modifier such as TEA (0.1-1%) to try and keep the bands from tailing to infinity. However the TEA curiously seems to have the effect of "bunching" everything together so everything comes off in one or two fractions even if this is 15+ fractions into the column where you might expect things to have "spread out" by then. Anyone got any ideas as to why this occurs and what could be done to improve matters (removing the phase modifier isn't an option since you just get one long tail... amines really are a pain)? The elution mixture I typically have tried is 1% TEA / 0-x% MeOH / 99-x% EtOAc (sometimes I use less TEA but doesn't make much difference).

Snarfenhaggar said...

regarding phase modifiers and a lot of chromatography on amines, have you considered using amino-silica? (aminopropyl-modified silica gel, typically 1-10% modification)

it circumvents the requirement for TEA, but costs more than silicagel. something to consider if you're already paying more for your chromatography media.

If you can reuse the media by dryloading to form a guard, or flog your crude material over a quick silica plug to catch any coloured nastiness beforehand, the one column worth of media should last a long time.

Milo said...

When I was a postdoc, there was a guy who worked in the lab on Fridays. He was a retired chemist from Shell and needed his chem fix. He showed me this technique and said he had been shown this when he was a grad student.

Interesting how some lost ideas tend to resurface periodically.

Daniel C said...

Excuse me for my ignorance, but could somebody explain precisely the difference between DCVC and normal flash chromatography? Is the only difference that you adsorb the compound onto silica before eluting in DCVC?

Daniel Sejer said...

There are two posts on DCVC on the blog. Read them both and you will understand. D!

ejlmp said...

You can get better results with VDCC (your name dcvc) by reading my original paper on VDCC published in
Vacuum Dry Column Chromatography. J. Org. Chem., 1982, 47, 4592-4594
There is a typo in the first sentence but, other than that the techniques detailed there in blow the current method of using a Buchner funnel out of the water. You get better results with a Nylon column. If the separation is for closely running compounds, pick a solvent that has a low Rf value for the mixture and do an extended eluation by collecting several column volumes of eluant until separation is obtained.

ejlmp said...

There are two ways of running VDCC with nylon tubing removes the need for a fraction collector or the need collect fractions. With a nylon tubing filled with absorbent large scale limits of using a filter funnel fora column are absent. In VDCC a nylon tube after the chromatography run is used to puncture the column in a horizontal position with a syringe of 10 microL and spot on a TLC to find your sample. The compounds run on the column exactly as on a TLC plate so you will know where to sample. So you can forget about the need for test tube collection. All you need is a larger TLC plate for convenience.

stanner2 said...

I'd like to request a copy of the Synthesis article please.
Thanks! -SGT

Daniel Sejer said...

SGT, please email your request to and I'll send you the paper. D!

kah hoe said...

Hi Daniel,

In the Synthesis reference, the authors mention about doing gradient elutions and in your blog, you mentioned that you usually use 4 cm columns for your DCVC runs.

I am rather interested to try out DCVC myself but I am not sure how to go about preparing the elution solvents.You mentioned in your blog that you do step gradients. Take an example of 2 g crude reaction mixture with my product giving a Rf of 0.3 using 1:2 (EtOAc:Hexane). Using a 4 cm column, how do you gauge the volume of 100% heptane or hexane do you typically use before changing to 95% hexane to 5% EtOAc?


Daniel Sejer said...

Kah Hoe,
The best advice I can give about column chromatography is to try it and see what happens. What you describe sounds like a 0-100% EtOAc in heptane job. Try 5% steps and 20 ml fractions on a 4-6 cm wide column. D!

Juda Weisblum said...

Hello, I wanted to ask, does the type of celite matter? Sigma for example lists all kinds of celite, calcined, treated with calcium carbonate, acid washed etc... Does it matter?

Daniel Sejer said...

@Juda, Yes it makes a difference what Celite you get. You want to make sure that the stuff you have won't dissolve in your solvents and contaminate your product. Currently, I'm using Fluka Celite 545 coarse (Cat. # 22140). It works pretty well but avoid acids. A lot of it dissolves if it gets exposed to for example acetic acid. D!

Anonymous said...

Very neat technique.

I have been trying it on 3 g of a crude reaction mixture.
My compound has a Rf of 0.3 in 100% EtOAc), and when I run the DCVC on a 4 cm diameter column (5 cm of silica) with the crude pre-adsorbed on Celite 545 and with gradient elution (5% inrease in EtOAc per fraction), all the components of the mixture elute together when I reach 55% EtOAc 45% Hexane.
Am I doing something wrong here?
I pack the column very tight with a Buchi diaphragm vacuum pump while applying pressure with a glass stopper. The silica is Merck 60 (15-40 um) type 15111.

The mixture is composed of pyruvate esters and polyethylene glycol ethers.

Thanks in advance!

mamid said...

I've been doing flash only so far, but I'd really like to try this. All I need is the appropriate silica gel. What do you think about 25-40 micron silica: is it worth buying if cheaper than 15-40? I think it is, but I want a more competent opinion.

Daniel Sejer said...

@mamid: You can use the smaller particle size if you want to but it is not as good. Also the distribution of particle size is important. If 90% of the particles are 40 micron then its no good. I always buy the Merck silica because it works really well but I'm sure there are other equally good brands out there. D!

Aleksandr said...

Very impressive technique. I am going to try it with separation of lipids and lipid-PEG conjugates. My amounts are about 250-500mg, max 1g, so I need ~10-15mm sintered funnel. Could you tell me, please, where do you get sintered funnels with above mentioned diameter and 6-7cm of length (from the sinter to the lip). I dug over Sigma-Aldrich and Termo-Fisher catalogs, but found only funnels with 2-3cm length at the 10-15mm diameter, which is less than I need to apply even 5ml fractions. Do you know some companies that provide required size?

Daniel Sejer said...

@Aleksandr. We normally have the sintered glass funnels made by a glass blower to our specifications. However, I seem to recall that we used to buy them directly from Pyrex and Duran when I was working in the UK. D!

Eli Rosen said...

DCVC is a great method! Ive had great experience with normal phase silica. Has anyone ever tried running a column with RP silica?

mamid said...

@Aleksandr. Usable funnels are available in at least one size (diameter 20 mm, height above disc 100 mm). Check vendors' pages for "filter tube" (Allihn). See for example Witeg or Carl Roth.

Steffen said...

I have been using the DCVC (following your article + later improvements) for a while now - I must confess that I am still new to column chromatography - and I can't really replicate what I think the packed column should look like.

Any pro tips?

I have much trouble with the top few millimetres of the silica; everytime I try to compress it, it just flies up on the other side of the column. I have been trying to use a trick from another similar article by closing the top of the column, let the pump draw some good vaccum, and then open it again very suddenly in the top - the 'shock' should compress it well and it works sorta.
Also, when testing the column the eluent front is actually nice and horizontal, but the silica does have a very heterogeneous look.

Maybe some pictures of a running column would help :)
Thx in advance.

Daniel Sejer said...

@Steffen: It sounds to me like you are doing everything right. It sounds to me like you are putting too much effort into packing the perfect column. The top few mmm are unlikely to compromise your resolution. Just give it a go and see what happens. On average we spend 10-15 min packing a column (for the entire process). D!

Mikkel M said...

I am trying to get Dry Column Vacuum Chromatography to work and have really had a lot of trouble with everything just coming out together way too early. Today i discovered that the tapping on the top silica during packing was extremely important. I even got a small amount of separation, which has never occured before.
I ran a 5 cm (inner diameter) column using 20 ml fractions. The gradient was from DCM to 30 % EtOAc in DCM (where Rf = 0.4-0.5) with 0.5 % steps (yes, a lot of steps). Crude was 6 grams. Compared to Flash i did not get an impressive result. My compound started to appear at 3 % EtOAc! on TLC it practically doesn't move in that eluent.
I really want to make this work (if the rest of the world can then so should I be able to, right?) so now i wonder, is smaller fractions or lower gradient the best way of getting separation? Or is 20 ml too small fractions?

I hope you can help me out, im a bit desperate to find out what i do wrong.

Daniel Sejer said...

@Mikkel: Right, this is a complicated question to deal in writing. Your Rf sounds sensible so if you are using the right silica and packing the column as described around here (and in the paper) that isn't the problem. You may be overloading your column so try doing 1 gr on the same column and see what happens. Often when we have this problem it is due to residual polar solvent in the crude compound. DMF, DMSO and pyridine are the worst. Even quite small amounts can mess things up. Residual MeOH and other polar solvents are also a problem but these are easier to remove under vacuum. If you have something like DMF in there stick your stuff on a high vacuum system/freeze drier overnight. Those are my best suggestions based on what you write. I hope it helps. For your situation 20 ml fraction or even 50 ml should be OK. It depends on how difficult the separation is if you can go higher than 20. Maybe you need to use a larger diameter column. Try 1 gr first on the 5 cm column and see if this helps. Generally, we do something like 20-30 fractions in total so I wouldn't recommend the very slow gradient you use as it simply takes too long and probably doesn't improve things very much. D!

Alex said...

Dear Daniel,

I am trying to separate the mixture of polymer and its derivative (the product of conjugation of polymer with small molecule). The mixture is about 50mg and I have 3.3 cm diameter column. Do you thing it is possible to run such a small amounts on this type of column? And have you ever heard someone separate polymer mixtures using DCVC?
Thank you.

Daniel Sejer said...

@Alex: That small amount of compound on that size column means your stuff will be very dilute in the fractions and could be difficult to detect unless it is very UV active or sensitive to TLC stains. Other than that I don't see any problem except that your compound may be difficult to purify. Give it a go. I would do relatively small fractions (10-20 ml). D!

ejlmp said...

Are you using Nylon tubing? If so try to use a narrow column for small samples. Nylon tubing was available when we developed dry column chromatography enhanced with vacuum. see The Journal of Organic Chemistry, 1982, 47, 4592. I take a developed Nylon column, lay on side and puncture the nylon column with a small syringe to withdraw samples to locate material for analysis with tlc or VPC, NMR, etc. I puncture the Nylon tubing at areas that approximate Rf values on a TLC plate using the same eluting solvent as used in the Nylon column. When the area with desired material is found, cut the nylon tubing and extract with solvent. Questions? Eric

ejlmp said...

If you use UV active silica in a dry column (vacuum or gravity), the substrate will blank out the uv silica and aid in the location.