The direct Akabori with alanine and benzaldehyde requires a few things in order for it to work well.  Distill your benzaldehyde immediately before, for best results.  Also, grind up the alanine to the finest consistency possible, to aid in reaction rate.  I have attempted the akabori once before, but my benzaldehyde was slightly impure at the time, as well as the alanine having crystals about the size of table salt.  Needless to say, I achieved almost no product, which I could predict before even performing workup, due to large amounts of unreacted alanine remaining at the bottom of the flask.  I've heard of people using a dean stark apparatus, with xylene as the solvent, obtaining better yields (but still, not fantastic). 

Can I just say...

There seems to be two different theories and procedures for what is happening in this reaction(the old and the new). And in all fairness there is two ways to approach this the 1942 method of refluxing the aldehyde and amino acid with pyridine or solvent until CO2 evolution ceases with a 16% yield. Nicodem posted a mechanism for this method that is very close to what was proposed to happen in that case,


Now, the Japanese chemists who wrote the paper below use a different method because after analysing the products via NMR spectroscopy they discovered the formation of an unstable oxazolidine ring structure which could be opened via hydrolysis to form the desired product. They heated the aldehyde and the amino acid to 130*C for 1 hour, which formed an unstable oxazolidine intermediate as well as the 16% yield via the nicodem style mechanism, the ring can then be opened via a 3 hour reflux with 5% aqueous Acetic acid boosts the yield from 16% up to 48%.


A quote from the Japanese;
"The N-methyl alanine (1) and benzaldehyde (4, R=H) with the non solvent 130? and 1hr when it heats,
 the ephedrine (6, R=H) was obtained same as Akabori and the like report at only low yield, but the reaction blend was scrutinized, and it became clear from the measurement of NMR the [ro], the Oxazolidine body (5) to have formed densely.This chemical compound (5), being unstable, from the fact that isolation is difficult, when adding water it disassembled the reaction liquid directly with the AcOH aqueous solution, improved."

Old style that is consistant with nicodems proposed mechanism;


New style, now that oxazolidine ring formation was discovered;


The paper in question;
http://www.sciencemadness.org/talk/files.php?pid=153366&aid=7798

thanks for no1uno for uncovering that gem!

no1uno your link didn't work for me so here it is if someone else had that problem.
http://pubs.acs.org/doi/abs/10.1021/jf60186a007
Not to know what I'm talking about but I think it would be easier to make N-methylalanine first. If pyruvic acid and methylamine works that would be sweet because I believe pyruvic acid can be made via oxidation of propylene glycol (antifreeze) by sodium hypochlorite (bleach).
C4H8O2(l)+NaOCl(aq)--->C3H4O3(l)+NaCl(aq)+2H2(g)
Can we please have more information on the N-methylalanine via pyruvic acid and methylamine reaction? Nice find Naf1 & no1uno.
Also can pyridine be replaced? Not very OTC.

"Because if you use plain alanine then you are likely to get the N-benzylidene-a-methyl-b-hydroxyalanine"

But Akabori reactions with Alanine and benzaldehyde and other aldehydes are well known, ie the highly linked Cycloknight post at SM where he produced PPA from benzaldehyde and alanine;

http://www.sciencemadness.org/talk/viewthread.php?tid=5979

The mechanism Nicodem posted, and I referenced above uses Alanine and benzaldehyde, have a look!! Are you referring to heated solutions of Glycine-copper or heated alkaline catalyzed reactions, I have no doubt you could promote formation of the afore mentioned compound but just heating alanine and benzaldehyde will proceed via akabori!  Furthermore Alanine is quite able to form the corresponding oxazolidine ring with benzaldehyde and there are many examples in the literature. The new enhanced Akabori will work on Alanine it is just whether that is more feasible than acquiring methylamine and pyruvic acid. A testament to its oxazolidine forming abilities should be seen in the yields in Cycloknights post!

The Erlenmeyer-Plöchl azlactone and amino acid synthesis, does not concern us ephedrine or phenylpropanolamine are not an azlactones or amino acids!

This is from the textbook definition of an Akabori;

3.Form alkamines by heating aromatic aldehydes with amino acids. No reaction observed with teriary amino groups.

http://www.chempensoftware.com/reactions/RXN006.htm

The pdf you referenced, uses a phosphate buffer, dimethylformamide and L-TA enzyme. I said before it is possible to create the conditions necessary for the above reaction. But just refluxing Alanine and benzaldehyde will proceed via Akabori.

The paper you referenced;
http://studentsupport.syr.edu/upload/Tracey-Ann%20Samuels%20Full%20Text.pdf

And as azlactones nor amino acids are produced the Erlenmeyer-Plochl reaction is not taking place. Read the sciencemadness thread i referenced! If you want confirmation from Nicodem read this!!!!!!!!! He is involved in this discussion!!!!

http://www.sciencemadness.org/talk/viewthread.php?tid=5979

quote from NICODEM in the above thread entitled Amino alcohols via Akabori trial run, where cyclokinght documents he akabori attempts by refluxing alanine and benzaldehyde, heres a quote from Nicodem;

"N-methylalanine with benzaldehyde gives a mixture of the ephedrines, alanine with benzaldehyde gives a mixture of the norephedrines, as simple as that. The chirality of the alanines is irrelevant. Check the original papers for other details. The reaction with alanine works just like CycloKnight described. The yields however vary a lot and you might not be able to reproduce his 21% yield in the first attempt. Stirring is important since the reaction mixture gets very viscous and will not mix well by itself. Also, lots of CO2 evolves which results in foaming, which I imagine, would carry the mixture on the top of the condenser if not tempered with efficient stirring."

edit; The Japanese paper referenced must not have been in circulation when that SM thread was being written as the topic of oxazolidines is not brought up. And the proposed mechanism written by Nicodem indicates the old style method before the oxazolidines had been discovered by NMR spectroscopy by those Japanese in that paper.

"I know about the Japanese paper in question (to the best of my knowledge I'm the one who found it)..."

Yes I gave you credit in my post above, something like' thanks to no1uno for uncovering that gem.

"I was also involved in the said discussion with Nicodem"

I also know this, the link that I used for the paper above came from your original post at SM, this also confused me more as to why you were saying alanine and benzaldehyde wont participate in an Akabori. After being in the Akabori thread about alanine and benzaldehyde reflux to produce PPA??

"Quite frankly, we have 3 procedures, using essentially the same reagents (variations relate to the existence or non-existence of the a/N-methyl or both)."

There is not three procedures, there is the older accepted Akabori with 16% yields, then there is the newer Akabori including hydrolysis of the formed oxazolidine to double the yield to 48%. Thats really one procedure with a modification to increase yield. Now Nicodem tried to tell you and here is excerpt from one his responses to you!

"There is however an important, even though mostly formalistic, consequence (unless the oxazolidine forms only when using N-methylalanine but not with alanine, which is however unlikely). Namely, the Akabori reaction gives a much better yield when the correct stoichiometry is accounted for. Much of the fame of its low yields is thus due to using the wrong stoichiometry in calculating the yield and/or the belief of having used alanine as the limiting reagent. For example, CycloKnight always used substoichiometric amounts of benzaldehyde in his experiments, even in those cases where he believed of having used an excess of it (unless I missed some example while rapidly skimming trough the start of the thread). So it is yet to see what the yields are when using alanine as the limiting reagent (maybe this is explained in that paper, but since I don't understand Japonese…)."

That was a direct cut and paste from a Nicodem post from here;

http://www.sciencemadness.org/talk/viewthread.php?tid=5979&page=6

"I sincerely doubt Nicodem would be quite so certain of the fact that the desmethyl precursor's would proceed by the exact same reaction as the N-methyl variant."

About half way the page, where you said the same thing. And he tried to tell you then.
And one more Nicodem quote for you! This was written to YOU by Nicodem, regarding this whole conversation.

"Seems like I have to say everything three times before you actually get to read it. I must say that your old habits die hard. So I'll try one more time, more concisely, before I give up.
You say "if alanine reacts with one mol of benzaldehyde to give an azlactone" and I say: benzaldehyde and alanine do not react to give an azlactone!
Please read the text concerning the scheme where that compound 193 is mentioned and more so the chapter on the synthesis of azlactones from N-acyl-amino acids. I already told you that once you read it you should clear up your confusion.
Besides, why don't you just try to draw the condenstation between alanine and benzaldehyde so that you can see it can not give any azlactones unless the reaction involves an oxidation (always check the oxidation states of the left and right side of the equation!).

Benzaldehyde has nothing to do with the Erlenmeyer or related reactions, utmost it can be used as an aldehyde in a condensation reaction to prepare the 5-benzylidene derivatives, but not for the synthesis of the azlactone itself.

First of all the conditions are anything but similar and the products are not even closely related. Not to even mention that in the Erlenmeyer reaction benzaldehyde has no role while in the Akabori reaction one equivalent is necessary in order to form the amine/imine for the alpha-CH group activation so that this can participate in the condensation with another equivalent of benzaldehyde as well as to allow the decarboxylation of the -COOH group.
In short: for the cyclization of N-benzoyl-alanine you need zero equivalents of benzaldehyde; while for the formation of the oxazolidine end product of the Akabori reaction you need two equivalents of benzaldehyde. (your homework: draw the reactions and balance the equations)
I really can not explain better. It is your turn to read all the posts all over again, do some reaction drawing and some reading, because I really can not explain in any simpler words. "

OK, I finally see where you are going wrong!

Akabori reaction has three different varients;

The first variant is able to;
1. Oxidize an alpha amino acid by heating with an oxidizing sugar.

The second variant is able to;
2. Reduce alpha amino acids and esters by sodium amalgam and ethanolic HCl to alpha amino aldehydes.

And the third this one we are interested in, read carefully it is very literal;
3. Form alkamines by heating aromatic aldehydes with amino acids. No reaction observed with teriary amino groups.

That is the textbook definition of the Akabori and variations of. Note N,N-dimethyl amino acids cant react in the third example, but pretty much all others INCLUDING GLYCINE CAN!!! Yes look it up Glycine reacts with benzaldehyde via the Akabori mechanism, to form phenylserine.

http://www.jbc.org/cgi/reprint/237/10/3229.pdf

quote from the above paper;
"DL-Phenylserine-2-C was prepared by the condensation of
benzaldehyde and glycine-2-C as described by Akabori et al.
, and the products of the reaction were separated by chromatography
in isopropanol-acetic acid-Hz0 (70:5:25) after removal
of the salts on a Dowex 50-H+ resin and elution with ammoniacal
ethanol . This synthesis produces the erythro
and threo forms of dl-phenylserine. The composite of all four
forms was used in the experiment described. Synthesis of p-hydroxyphenylserine
was attempted by this procedure but without
success."

refs;

Arch. Biochem. Biophys, 83, 1 (1959)


Now where you are going wrong and Nicodem did point this out, that for an Erlenmeyer type of reaction to proceed oxidation has to take place. Oxidation can take place in the presence of excess metal hydroxide.Like here;

On the Erlenmeyer reaction.I.Mechanism of Threo-B-phenylserine Formation

http://www.journalarchive.jst.go.jp/jnlpdf.php?cdjournal=bcsj1926&cdvol=34&noissue=9&startpage=1314&lang=en&from=jnlabstract

The Preparation of Hydroxyphenylserines from Benzyloxybenzaldehydes and Glycine

http://designer-drugs.com/pte/12.162.180.114/dcd/pdf/akabori.phcho.glycine.pdf

Just to clarify amino acid and benzaldehyde heated together proceeds via Akabori (except amino acids with tertiary amino groups).

Amino acid and benzaldehyde in the presence of excess metal hydroxide and usually alcohol will proceed via Erlenmeyer.

Two completely different reaction conditions (neutral vs very basic) to completely different reaction mechanisms two completely different products! Two completely different reactions!

I must say no1uno, you bring up some excellent points. I was actually a bit confused before and did not understand exactly what you meant. Now I do, they are very reasonable arguments.

"Nicodem certainly has not, to the best of my knowledge, come up with a clear reaction mechanism yet including the latest version of the "Akabori" paper (which quite frankly, despite the products and reagents, does not necessarily fit the jacket of what is known as an "Akabori" reaction)."

But to be fair,he did the mechanism of the original Akabori and that still takes places (with only 16% yields) so that is still feasible. But there is also an appreciable amount of oxazolidine that forms and can be hydrolyzed to yield an Akabori 'type' product. The only thing I will fault with Nicodems comments on the Akabori, he says that the new Japanese paper did not survive Occams Razor. But when your NMR analysis tells you an oxazolidine is in solution so you hydrolyze it and indeed get an increased yield Occams razor goes out the window.

This is a selected quote from wiki describing Occams razor in science;

"not an irrefutable principle of logic, and certainly not a scientific result.[17][18][19][20] As a logical principle, Occam's razor would demand that scientists accept the simplest possible theoretical explanation for existing data. However, science has shown repeatedly that future data often supports more complex theories than existing data. Science tends to prefer the simplest explanation that is consistent with the data available at a given time, but history shows that these simplest explanations often yield to complexities as new data become available.[8][18]"

So the mechanism Nicodem posted is still valid for 16% of the product. The other 32% is formed via oxazolidine, if you google you will see that (probably since that paper) that oxazolidines and Akabori have become somewhat synonymous. NOW the Akabori mechanism can be determined by Nicodems original drawing, and by oxazolidine ring formation. Below is the constitution of an oxazolidine, adding any double bonds, oxygen, nitrogen etc... And it is no longer an oxazolidine, it has to be of the constitution below.


Akabori style oxazolidine formation, starts with nucleophilic addition of benzaldehyde and alanine to form a hemiaminal, followed by dehydration to the imine. Once the somewhat unstable imine forms all that is needed is HEAT to push the decarboxylation of the imine forward, a zwitterion is left after decarboxylation. The zwitterion can participate with another benzaldehyde to form an Aldol type addition. And as you can see below that the oxazolidine ring formation would happen very quickly after the aldol addition as the electron movement stabilizes the compound forming an oxazolidine.



"the original "Akabori" itself, was conducted in neat pyridine was it not (see scheme 1 of the Japanese paper, the 1942 procedure)? So exactly how was it any less basic than the one done in a solution of hydroxide?"

Yes it was, but although a base pyridine it is not quite the same as a metal hydroxide is it? For example it can catalyze the Knoevenagel condensation which is remarkably similiar to what you want to happen in an Akabori type situation. The Knoevenagel condensation is catalyzed by a weakly basic amine (pyridine and many others).

"We have numerous reactions which proceed from the formation of the N-benzylideneamino acid"

No as stated above the Akabori type rings are oxazolidines. The mechanism that Nicodem wrote and still accounts for 16% of the yield does not even form a intermediate ring structure.

But condensing benzaldehyde and glycine in the presence of strong base affords Benzal glycine. As per the starting material in this machanism;


Also the mechanism for benzaldehyde and glycine via the Erlenmeyer reaction is documented here (again);
On the Erlenmeyer reaction.I.Mechanism of Threo-B-phenylserine Formation
http://www.journalarchive.jst.go.jp/jnlpdf.php?cdjournal=bcsj1926&cdvol=34&noissue=9&startpage=1314&lang=en&from=jnlabstract

As you can see the corresponding ring structure in that example of the Erlenmeyer reaction is not an oxazolidine! It is an azlactone, and further more the Erlenmeyer reaction between glycine in alcoholic hydroxide with benzaldehyde, no intermediate ring structure is formed at all!!!! This mechanism would be much more appropriate unless Acetic anhydride was used;

That is true - but the a-alkylation of alanine by either the Akabori (with or without pyridine) or the route described by Belokon et al that I cited above, not to mention the one resulting in the phenylserines, all start from the benzylidene? I know there is proof that the benzaldehyde can be recovered from that involving glycine, which does NOT form a 5-membered ring (for clarity - we have azlactones, oxazolidine & oxazolidonones here so far)...

There is to date, not a single bit of proof (of which I am aware) that the excess benzaldehyde in the "Akabori" etc. is recovered unchanged, or even whether the initial imine is not somehow reduced to an amine - allowing the carbon bearing it (from the 2nd benzaldehyde) to bond with the OH from the alanine (which given the absence of the double-bond in the intermediate postulated by the Japanese authors), not probable perhaps, but the only thing I can think of on the fly.

Occam's razor is a tool not a rule, it is used to determine the probability of various results, it does not rule them out completely, especially when the seemingly anomalous is what we are trying to explain.

PS the pyridine in the original "Akabori" paper is still a base, quite a strong one and quite capable of forming a salt with the carboxylic acid functionality on the alanine - as is the alkali metal hydroxide in the glycine variant.

Well, technically the first step of an Erlenmeyer type reaction  condenses benzaldehyde and glycine in alkali(sodium hydroxide), to produce the alkali salt of benzalglycine a 'benzylidene'. Sodium hydroxide forming the alkali salt and removing alot of positive character from the reaction thus, it favours Erlenmeyer conditions of benzylidene formation over the positively charged schiffs base of an Akabori. If the erlenmeyer is not conducted in excess alkali then yields are severely reduced by the competing Schiffs base formation(that is Akabori). Another benzaldehyde is then able to react with benzalglycine in base, to form the to two stereoisomers of N-benzal-phenylserine. A standard deprotection using H+ hydrolysis to remove the benzylidene leaves phenylserine.   

And the first step of the Akabori condenses benzaldehyde and glycine in a nucleophilic addition, to produce a Schiffs base(with a positive charge on the Nitrogen). When heat applied to this relatively unstable imine, decarboxylation is promoted which leaves a zwitterionic compound. The zwitterion can react with another benzaldehyde to form an aldol adduct. The oxazolidine ring forming abilities of glycine will determine whether an enhanced yield via oxazolidine will be feasible.

You are correct Pyridine is quite basic, and will form a salt with glycine also. But pyridinecarboxylic acids are known to decarboxylate readily. Maybe that was Akabori's original thoughts that creating the pyridinecarboxylic acid in situ would aid decarboxylation? But instead was probably the cause of his low yields by benzylidene condensation among others.

"That if one refluxed benzaldehyde and l-alanine for 1hr and then added AcOH to make the solution 5% AcOH, then the yeild of PPA *COULD* be doubled?"

If you are planning it, come back and tell us how it goes. The Japanese paper hints that alanine wont work as well as N-methylalanine. But cylcoknight at SM got decent yields using just reflux so there is hope the an increase will occur! But cannot say for certain.