aka “Clowns to the left of me, jokers to the right”
This post isn’t about the “best” way to run an extraction – that will be a whole series of posts, sometime in the distant future – it’s about a specific method. A couple of years back I needed large quantities of a certain reagent (>10g), which wasn’t really commercially available. It could be synthesized in a single step (below), but unfortunately the purification was a bit of a pain. The literature called for flash chromatography in a chloroform/methanol/ammonia mix; a nasty but serviceable. But the compound bled on the column, and after several hours of work and several litres of carcinogenic solvents all we had was a small amount of the needed material. To add insult to injury, the product is rather efficient at removing CO2 from the atmosphere. Let it stand for a little while and you end up with a less-than-useful white sludge. 
So, some modifications to the procedure were needed. Pre-column the crude material was composed of only three compounds: unreacted 1,4-diaminobutane (putrescine, named for its distinct odor), tert-butyl 4-aminobutylcarbamate, and di-tert-butyl butane-1,4-diyldicarbamate (according to chemdraw). The Boc2O was consumed in the reaction, and both the t-butanol byproduct and triethylamine were readily removed under vacuum, though we normally removed the butanol in an extraction pre-column. Stretching back to second year organic chemistry we supercharged the extraction step, and chem up with this gem of a flow chart.
It’s not something you think about every day, but each of the three common extracting solvents has a different extraction efficiency. Ethyl acetate is the strongest, able to pull highly polar compounds out of simple water, and the rest flow down in polarity, like so:
EtOAC > DCM > Et2O >> Hexanes
It took a little trial and error to find the appropriate choice for each step (ethylacetate, for example, will draw out putrescine as well as the desired product in the last step), but in the end that was time well spent. In the updated procedure there’s no column, the whole process takes about thirty minutes, and we obtain at least 90% of the mono-Boc product at the end, pure by NMR. There’s also the option of getting the two impurities as pure compounds along the way, should we want them.
The double extraction requires at the least an acid or base, so it’s not generally applicable. It does make a good ace in the hole though, when little else works.
Chemtips 
Some day I statistically propargylated a certain hydrophilic diol. The resulting mixture was dissolved in brine, the dipropargylated product was extracted with Et2O, and the desired product – with CH2Cl2. The substrate remained in aq. medium. Sometimes those methods work, so statistics is not the reason not to try.
I work with polyamines, so require thus precursor a lot, and have regularly made it on laege scales.
The mono-Boc product is soluble in water, the bis-Boc is not, so I remove all volatiles, then filter.
Extraction with DCM then gives you just the desired product, as the free diamine prefers the aqueous.
That’s what I like about organic chemistry. Always more than one path to success 🙂
Thanks Maksim, I’ll have to keep that in mind the next time I work with diols. I’m used to thinking in terms of protonation/deprotonation, but a big enough difference in polarity should allow a double-extraction as well.
I want to edit my comment,reason – grammatical mistakes.
Can author of the blog answer my stupid question please?
why does only one NH2 group react? what is the factor which decides this selectivity?
Statistics, mostly. There’s little difference in reactivity between putrescine and the mono-Boc protected variant, so you get a roughly 25/50/25 style distribution of compounds.
Yields are usually around 60%. Not exactly a great, but the starting materials are relatively inexpensive.
Any synthesis that requires altering one of two similar sites is a bit of a gamble. In the best case the product is far less reactive than the starting material, and you approach 0/100/0 distributions. I’ve had it go the other way as well, recovering essentially 50/0/50 (one equivalent of the protecting group).
Brandon, I found this workup to be a really eligant alternative to the more laborious column chromatography purification. Unfortunately in my hands the reaction did not get anywhere near the 90% you state above… I was wondering if you’d be able to send me the exact reaction conditions that you employed? Any advice would be greatly appreciated!
Sorry, my language was a little inexact. The yield isn’t 90%. I’m say that of the monoBoc produced, 90% will be recovered during a double extraction.
Depending on stirring and addition rates your true yield will vary, but it is usually between 25 and 45%.
Because the monoBoc is about as reactive as putrescine, the highest yield you can theoretically get is just north of 60% (it would be exactly 50%, but I use 1.5 molar equivalents of putrescine per equivalent of Boc anhydride). This is never going to be a high yielding reaction.
Pingback: While I Was Out… | Chemtips
Pingback: Open Letter to a New Student | Chemtips
Informative blog post , I learned a lot from the specifics – Does anyone know if I might be able to find a blank a form version to use ?
I built that in powerpoint, using the various shape functions.