Anhydrous Solvents Part 3: Acetone and Molecular Sieves - Bad idea!

I have discussed anhydrous solvents a couple of times and have been advertising the use of molecular sieves (MS) quite strongly. During my MS crusade I have pointed out that MS are no good for drying THF but that pretty much all other standard solvents work well with sieves. As it turns out this is incorrect and I have received a terrible punishment from the MS God. It's all rather embarrassing as a PhD student in the lab was fully aware of the particular problem I'm getting to shortly. The deal with MS is that they are weakly basic. I take advantage of this by always adding some MS to my CDCl3 which keeps it dry and mops up any HCl formed by the slow decomposition of CDCl3. The other day I was running some of 1H NMR and to my great pleasure I had finally (after months of struggling) made a very important target molecule. I had split the fractions from a column up in three batches to be on the safe side. All three 1H NMR spectra were great so I decided to combine them in one flask. I was running NMR in acetone-d6 and decided to use some acetone for the transfer. I couldn't find the HPLC acetone we normally have standing around and was getting a bit frustrated when I remembered that about a year ago I had made a bottle of acetone over MS (This is were all the alarm bells go of with the experienced chemist). I managed to find the bottle and proceeded to transfer all my stuff into a new flask. However to my utter surprise I was unable to remove the solvent on the rotary evaporator. On the high vacuum pump with a fair bit of heating most of it came off but by TLC there was a new UV acitve (and quite polar) compound. I had to re-column my product but it still wasn't pure! Currently I am attempting to crystallise it from the impurities. A fair bit of yelling and acussing people of sabotage took place. Fortunately it was late and there was only one other person in the lab.

Before I proceeded to clean up my compound I decided to figure out what the source of the problem was and I quickly discovered that the acetone smelt funny. Initially, I thought it was contaminated with benzaldehyde but when more dilute it had a floral/perfume scent that reminded me of ketones/esters. At some point the PhD student in the lab realised that I had been adding MS to acetone and mentioned that as far as he knew that was a no go because it goes Aldol in the presence of the weakly basic MS. I cannot believe that this hadn't occurred to me. As it turns out it is well known that ketones go Aldol when exposed to MS and many different compounds are formed. A selection of possible products from acetone are shown in the Scheme above. The polar compound I removed by chromatography is the acetone trimer with one hydroxy group.
In my defence I'll say that acetone can be dried with MS provided you use the acetone within a few days. I successfully used the anhydrous acetone during a week of experiments back in July 2009. The take home message is to bin it after a week and not use it a year later as I did. In fact just don't add MS to acetone but instead dry it with MgSO4 as described here. D!