[nox]

Maybe it's just bad luck, or the fact I'm picking up synthesis again after a 4 month hiatus, but I've managed a week long streak of running crappy columns where stuff just co-elute all over the damn place.

And we're not talking about tight spots either, we're looking at things that have an R_f difference of at least 0.2. The separation looks so tantalizingly perfect on TLC, yet when I actually run the column, I get partial separation at best.

I slurry pack my column in whatever eluent I plan on using, and wet load my crude using toluene -- used to load with DCM until I had several bad experiences where it just flushed everything out within 5 fractions. After speaking to various people I've been told the following: pack column using hexanes only, do a gradient rather than an isocratic run, pack a longer column (I usually use about 6 inches of silica), and use dry loading rather than wet.

I usually get complete separation on running the 2nd column, only difference being I pack more silica and use a less polar system. I actually suspect it's what I use to load my column that's giving me these headaches. Right now my project involves making huge, greasy, polycyclic heterocycles, and I'm starting to think that these run merrily along with toluene. They dissolve in DCM just fine, but I get that flushing problem mentioned above.

For the record my columns usually use 5-15% EtOAc in Hexanes, and the R_f of the compound always runs around 0.3 on TLC in whatever system I'm running my column in.

Any insight is appreciated.

[Dan]

I've been told the following: pack column using hexanes only, do a gradient rather than an isocratic run, pack a longer column (I usually use about 6 inches of silica), and use dry loading rather than wet.

I was going to suggest all of these things as well. Also, after loading I always run down one column volume of the weak solvent (usually hexane) to flush out the load solvent with minimal streaking.

Right now my project involves making huge, greasy, polycyclic heterocycles, and I'm starting to think that these run merrily along with toluene.

Yeah, toluene may be moving your products - have you tried a TLC with toluene? I've columned really greasy diamine in 50-100% toluene/hexane gradient before.

For the record my columns usually use 5-15% EtOAc in Hexanes, and the R_f of the compound always runs around 0.3 on TLC in whatever system I'm running my column in.

Maybe try a system that puts your product at 0.2 (or less). The silica on the TLC plate is not exactly the same as the silica in your column, so you may have to adjust a bit - especially if there was a supplier switch in the 4 week hiatus you mentioned, then you'll need to run a few columns to get a feel for the new batch. That said, I would strongly advise gradient over isocratic elution, it has always given me much better results.

[azmanam]

If you haven't already, read this paper:

http://dx.doi.org/10.1021/jo00408a041

[Doc Oc]

Maybe try a system that puts your product at 0.2 (or less). The silica on the TLC plate is not exactly the same as the silica in your column, so you may have to adjust a bit

This.  All of the other stuff you mentioned might work, but I never had any problem with those.  I did, however, run into issues where I'd run a TLC in a certain solvent system and then have worse separation on the actual column.  One of my colleagues suggested I always run my column in a lower polarity than my TLC (ie; TLC 9:1 hexane/EtOAc, column 95:5 hexane:EtOAc).  That always seemed to work for me.

Rather than loading your compound with toluene or DCM, you could try dry loading it, that worked out amazingly well for me with a lot of my compounds.

[Honclbrif]

Always load in the solvent you are going to run with. If it won't dissolve in that, dry load it. If you want to do the toluene tracer trick (T³?) push the toluene band a few cm from the top before you load. So you pick up a few extra dead fractions? I'd rather have them than bad/no separation.

[the.khemist.ds]

I agree with Honclbrif, always load in the same eluent or dry load. (I've also started by packing and loading the column using a much more non-polar eluent system and then moving up to the eluent system I've chosen - It sometimes works)

Is the impurity the faster eluting compound? If so I'd suggest picking an eluent system that brings this down to Rf 0.3 and then move to an eluent system that brings your  product out.

Alternatively, try a different eluent system - perhaps swap the ethyl acetate out and use a DCM:hexanes mix.

[nox]

One thing about dry loading I've heard is that the separation tends to be not so great compared to wet loading, is there any truth to that? Because frankly,  that's pretty much the only thing holding me back from dry loading all my columns.

[Dan]

One thing about dry loading I've heard is that the separation tends to be not so great compared to wet loading, is there any truth to that?

I actually find the reverse is true - that the bands are thinner from a good dry load compared to a wet load because dry loads don't streak, and it's normally easier to keep the load volume down.

[Honclbrif]

Dry loading is a bit of a balancing act. You want to keep the loading band as narrow as possible, but you want the material adsorbed to the chromatographic medium and not just caked on. All I can recommend is start with a little less medium than you think you'll need, vap off the solvent, and if it looks all gross and cake-y, add a little bit more until it looks like its dispersed when dry. If you've added too much, either bite the bullet, run the column, and get ready to try again, or strip the material off with a good solvent and use less medium next time. I'd give better advice on how to hit this happy point if I could but experience is really the best teacher.

[nox]

I feel so much better after today.

Was repeating a reaction from one of the postdocs (in order to assess reproducibility/set my own baseline). Took a TLC and lo and behold -- 8 spots. 3 above, 4 below my product, all chasing each other (but thank god they were at least resolved) Curse at TLC and crude for 5 minutes, go for lunch break/wallow in misery.

After lunch, decided to man up and run the damn column. Incorporated every piece of advice -- packed 10" column in hexanes, added toluene tracer and pushed in a good 5 cm, loaded in absolute minimum amount of DCM (~1 mL), flushed with 1 column volume of hexanes, pushed through another volume of 5% EtOAc/Hex, then finally eluted with 10% EtOAc/Hex (product runs ~0.15 in 10% EtOAc).

Product came out over 8 fractions, 1st fraction had some co-elution of upper spot. last fraction had some co-elution of lower spot, middle 6 clean. Pooled all 8 fractions, NMR sees barely any impurities (apart from residual solvent and a bit of grease).

So even with such a nightmarish TLC, I was still able to get relatively decent separation. Clearly I've been running all my columns wrong the past week.

[OC pro]

Yeah, separating pure product out of a wild mixture gives a good feeling 
Learning only by doing...

[chrisr789]

Always load in the solvent you are going to run with.

As in the less polar of the two? So in a pet/EtOAc system you'd try to load in pet?

[Dan]

Always load in the solvent you are going to run with.

As in the less polar of the two? So in a pet/EtOAc system you'd try to load in pet?

No. Say you are going to run your column in 1:1 pet/EtOAc (isocratic elution), then you load you compound in 1:1 pet/EtOAc. Lots of people do columns this way.

Personally, I wet load in either - in order of preference, depending on solubility - toluene, DCM or chloroform, followed by a column volume of the weak column solvent (normally hexane), and then begin gradient elution. If the separation on TLC is <0.1 I dry load.

[chrisr789]

^Yeah I usually opt for DCM loading in a Pet/EtOAc column. I must have had my head in the sand as I've never heard of people loading in their isocractic eluent before. Scrap that, I've just asked about in my office and a couple of people do. I'll give it go later and see how it compares.

[da692]

for you dry-loading fans out there: celite works as a good replacement for silica for dry loading medium. it has helped me apply particularly polar mixtures to the column as tight bands. be sure to use acid washed celite. when you wash your celite with a little aq. HCl (and subsequently H2O until neutral, then solvent and dry under vac.) you will see why you should wash your celite.

Although i've used dry-loading and it has saved the day for me in the past; however, I still prefer to wet load mixtures onto columns.