[TheUnassuming]

Does anyone know of a good paper or resource for silica gel solubility in various solvents/solvent systems?
I'm currently making a number of very polar compounds and that information would be amazingly handy for purification.  I've searched for a while but haven't come up with anything relevant.  I can't imagine no one has gone through and figured it out, but perhaps the paper is so old it is effectively buried.
Thanks for any *delete me*

[stewie griffin]

I've never come across a paper on that topic either. All I've ever heard was don't use more than ~15% MeOH or you start to dissolve the silica. I've pushed the limit with the MeOH before in desperation - just filter my isolated material post column.
If your compounds are so polar, perhaps it would be better to explore using reverse phase chromatography...

[TheUnassuming]

I've pushed up to ~20% MeOH/EtoAc (~15% with DCM) with barely any signal in NMR analysis of compounds, but thats about all I know.   
For the final compounds that will be sent out for biological testing, reverse phase HPLC is/will be done if for no other reason than the biologists always like to see a clean LC spectra before they dose their critters/cells.  So that is always an option, I just want to avoid using it when I have several steps left to go. 
I've never done alumina columns before, but that is worth ordering up to try.

[stewie griffin]

That's true you could try alumina. I've only used it once when I had to, and it was a lit prep. There are different  different levels of deactivated alumina (Brockman levels, set ratios of water to alumina). I haven't seen the recipe for Brockman levels in the lit (not that I ever looked) but we had the ratios floating around labs.
I guess you could also deactivate your silica with some eluent containing triethylamine.

[Babcock_Hall]

I feel your pain.  I work with polar compounds a good deal, and I use various purification methods.  Sometimes I think that reverse-phase sep paks might be one way to go, but I have not tried them yet. 

[Dan]

I suggest trying eluents based on EtOAc/IPA/water, 45:5:1 up to around 7:2:1 does not seem cause major silica dissolution.

I have columned free sugars on silica using these eluents.

If your compounds have acidic/basic groups, you may want to look into doping the silica.

[Babcock_Hall]

Dan,  Do you know of a good reference that deals with doping silica?  Thanks in advance.

[Dan]

No, but I can tell you from personal experience.

Triethylamine (TEA) doping for purification of amines:

1. Slurry pack column in the weak solvent of your eluent (I will call it hexane hereafter).
2. Run through approx. 10% TEA/hexane. You will notice a TEA front running through the column as you do this - the TEA-doped silica has a whiter, more opaque appearance than the untreated silica.
3. Repeat step 2 until the entire column is visibly doped.
4. Run through 1 column volume of 1% TEA/hexane.
5. Continue the column as normal, but add 1% TEA to the eluent.

For TLC, soak the TLC plate in 10-20% TEA/hexane for approx. 20 seconds and dry with gentle heating (or air dry). Run TLC as normal, but with 1% TEA in the eluent.

AcOH doping for purification of carboxylic acids:

I have not done it personally, but I suggested to a student working with carboxylic acids to follow the TEA procedure with AcOH in place of TEA - the results were very good, complete elimination of streaking on silica.

[BobfromNC]

I recently used HOAc as the modifier in some columns for acidic compounds.    Using about 1% or less HoAc in EtOAc for 0-100% gradients of hexane/EtOAc worked pretty well, but I got better luck with a 0-10% MeOH in DCM gradient, where I added about 3-5% HOAc to the MeOH such the the mix where the peaks were coming off was about 90:10:0.5 of DCM:MeOH:HOAc, and that worked pretty well for a few of the acids I was trying to make.   

But I have found that keeping the acid protected as an ester, then purifying carefully at the ester stage and then deprotecting them gently at the last step is usually the best solution, as then often the final acid can be extracted into water, the aqueous acidified and the acid either crashes out, or can be extracted into EtOAc or DCM, dried with Na2SO4 and then concentrated to give very pure acid, which does not need further purification in many cases.   

[quimorg]

I been using this tetramixture to move very polar compounds and it has work nice, the only difficulty part is making a gradient.  The mixture is 37.5% Hexane, 12.5% Et₂O, 25% CHCl₃, 25% MeOH

[IsotopeBill]

Here's a good site that shows the author's experimental results:
http://chemtips.wordpress.com/2013/02/01/is-silica-soluble-in-methanol/

[Babcock_Hall]

Our desired product is phosphonate diethylester that also has a secondary amine group present, as well as a carboxylic ester.  Based on one paper, it is said to be prone to transesterification on silica in a methanol-based solvent (0-3% MeOH in CHCl₃).  We are considering switching from methanol to ethanol to avoid this transesterification.  There are three major products by P-31 NMR, and the presumed product is about 60% of the total by integration of the P-31 spectrum.  One other problem with this synthesis is that it has been difficult to identify the product from a crude reaction mixture; therefore, we are not entirely sure how polar our system will have to be in order to move the desired product.  There are a total of 2.5 equivalents of TEA present, much of which is as the conjugate acid.  One equivalent to convert the hydrochloride of ethylglycine to a neutral form and the other to pick up a proton to facilitate the nucleophilic attack on diethyl bromoethyl phosphonate.

(1) At what percentage of ethanol does silica start to dissolve?

(2) is there a rule of thumb for what percentage of ethanol is equivalent to a given percentage of methanol in terms of polarity?

(3) Should we consider a TEA-based chromatographic system, such as the one above?

[discodermolide]

So you are trying to alkylate the N of glycine ethyl ester with bromoethyl diethylphosphonate?
This may be a messy reaction, it could give you a dialkylated product and the mono-alkylated product may well react with the ester.
A reductive amination may be the way to go.
As to the polarity. I have routinely used DCM /MeOH mixtures, up to 10% methanol with only traces of silica gel. So I guess that with 3% you should be ok and up to 5% should be ok. If you want to avoid transesterification of the carboxylic ester then use isopropanol as a solvent.

[Babcock_Hall]

My idea about transesterification was to make it an identity reaction--that is why I thought ethanol was a good idea.  But I like your idea of isopropanol, as well.  The original synthesis is from Bigge et al, J. Med Chem., 1992,35, 1371-1384.

[Dan]

On topic but not directly relevant to your specific question, here is an interesting article about silica solubility in methanol:

https://chemtips.wordpress.com/2013/02/01/is-silica-soluble-in-methanol/