Total Synthesis Strychnine - Step 6

We are almost to the good stuff in this total synthesis.  And by good stuff, I am talking about the very important stereocenter forming step.  But before we get there, you will notice we are still a couple carbons short of making one of our rings of strychnine.  As we stand in our synthesis so far, we have been planning along the way, that the beta carbon of indole ring would act as a nucleophile, reacting with an electrophile to form a new carbon-carbon bond with the proper stereochemistry.  It seems to me that these types of carbon bonds for chiral centers, usually take place with some sort of carbonyl derivative.  With this in mind, by reacting our primary amine with ethyl glyoxalate, we can set up our reaction for a nucleophilic attack on an electrophilic imine.  Below is the scheme with the nucleophile in red, and the electrophile in blue, which will be reacted in the next step.

Overview:

What we are doing is this step is a condensation between the primary amine and the aldehyde.  From Organic II, you likely learned about the reactivity of aldehydes vs. ketones and other carbonyl derivatives.  The aldehydes are significantly more reactive than esters for two main reasons.  As is always, these two reasons come from electronics, and sterics.  The steric argument comes from the hydrogen rather than another alkyl or heteroatom.  The hydrogen is much smaller and so allows approach of the nucleophile with much ease.  Compared to the ester, which has a resonance structure with the oxygen, it is also more bulky.  For these reasons, actually the ester is much less reactive than the aldehyde.  so much so that we see attack of the aldehyde moiety in place of the ester, even at reflux temperatures.  One must really understand the depth that these seemingly small differences have on the reactivity of the compound.  Even a difference in reactivity of only 2 orders of magnitude can really allow selectivity of one group over another.  Again, it's not that it's impossible for an amide to form, but it's less likely.  Another reason why, the leaving group would be an ethoxy anion, which, in benzene which is a non-polar solvent, may not solvate the anion well, and it's actually a poor leaving group in the first place.  The mechanism for this reaction is shown below.

While it may be possible that the leaving of the hydroxide group may occur concomitantly with the removal of the last proton, I put them as two separate steps for clarity.  In fact, because benzene is a non-polar solvent, it is more likely that water is the leaving group.

Workup:

One important consideration in these types of condensation reactions is the production of water.  It can be difficult to push the equilibrium toward the products if the water is not removed.  To remove this water and aid the equilibrium, there are a couple of tricks a chemist can use.  One, is to place molecular sieves in the reaction flask.  Molecular sieves are one such option.  These are simply some microporous material which has small pores to let small molecules fit in, but not larger molecules.  In this way, water will work its way into the sieves and essentially be removed from the reaction.  Woodward, however chose to use another, more popular technique.  This involves using a Dean-Stark trap.  This trap collects vapors during the reflux reaction, which contain benzene and water, and condenses them in the typical reflux condensor, however, the mixture is not dripped back directly into the reaction flask, but rather, into a "trap" where the water will sink to the bottom, and displace benzene which will fall back into the reaction.  It's a little hard to explain, but here is an image which should help.  

Lastly, the workup.  Refluxing for 5 hours is necessary to collect the water, and to help drive the reaction forward.  Refluxing is very common in condensation reactions.  To work up the material it is actually quite easy.  The material precipitates out of solution in the refluxing benzene as yellow crystals.  However, this often doesn't mean all of the material has been precipitated.  To ensure this, Woodward had cooled the flask overnight to aid in precipitating all of the material.  Even so, Woodward was clever enough to know that there was likely more material in the flask, even after filtering the solid away.  Nowadays we would use TLC to see if more material remained in the organic layer.  To get the rest of the material out of the benzene, he added lots of diethyl ether (which our desired compound should not be soluble in, as diethyl ether is pretty non-polar).  This helped to precipitate actually the largest amount of material out of the flask.  Concentrating the mixture either by boiling or by rotary evaporation, and then re-diluting with ether, again, gave more product (because some of the benzene was removed) and then, adding petroleum ether (really just a mixture of alkanes) to the mixture, precipitated another smaller batch of material from the solution.  A final recrystallization from benzene yielded the yellow crystalline powder in good yield (92 %).

Overall, a very simple and common reaction, but as I always say, there's a lot that goes on behind the scenes.  See you guys next time.