Amino alcohol via Akabori, trial ru

Amino alcohol via Akabori, trial run


Recently I’ve been experimenting with the Akabori reaction, just curious to see if the process for making amino alcohols can be improved upon. I thought it would be quite an interesting amateur science experiment.
So far its been an interesting challenge !

The Akabori refers to the reaction between an aldehyde & amino acid to yield the corresponding amino alcohol.


For this run, the following chemicals were used:
83 g l-alanine (considerable excess)
40g benzaldehyde
250ml xylene as solvent (convenient & OTC)

The conventional low yielding approach is to simply heat benzaldehyde & l-alanine at 150-160 C until CO2 evolution ceases.
The objective with this variation is two fold: to keep the l-alanine in excess at all times by slowly dripping the aldehyde into the reaction mix, the idea being that this will increase yields wrt to consumed benzaldhyde, & to keep water out of the reaction.
When water gets into the reaction, by-products are immediately formed and the reaction mix will turn red (tar) quite quickly. The boiling temp of xylene is 139 C, therefore if water drips back into the boiling solution, the results can send foam as far as the condenser & beyond.
Foaming is another issue with this particular trial run, higher dilution is required, higher than what was used in this trial run. By the end of the reaction, due to tar formation, foaming was getting a bit out of control.

The reaction was maintained at high reflux to help flush out generated moisture, this was maintained for 1 hour before the aldehyde addition. To assist with this task, a DIY Dean Stark trap was hastily thrown together, to collect generated water and help keep it out of the mix.

First, the l-alanine was finely powdered as much as possible.


Vegetable oil was used for the oil bath. Stir bar spinning.


General setup.






Moisture starting to collect in the trap.




For the first aldehyde addition, 26 ml of aldehyde was slowly dripped into the solution. Later (see below), an additional 14 ml was added; 40 ml benzaldehyde added in total.







Aldehyde addition is begun, 1 drop every 2 seconds.


Byproducts collecting, the red drops are water.






1 hour after the addition. The addition of the 26 ml aldehyde took 1 hour 15 min to complete.


The reaction was run until water stopped condensing, then the addition of the rest (14 ml) of the aldehyde was begun.




Benzoic acid forming in the aldehyde dropping funnel…


2 hours after the aldehyde addition was started.








Total collected condensate at the end of the reaction


Water accumulation ceased, so the reaction was stopped.
Reaction time = 4 hours.




The reaction mix was put in a freezer to cool, and precipitate impurities.
Once no more impurities (l-alanine mostly) were seen to fall out of solution, the mixture was carefully removed, 3* 50 ml cold toluene was used to rinse the unreacted l-alanine, the pooled toluene extracts were then shaken with 15% HCl until the acid remain acidic upon settling. Temperature before acid addition was 10 C, after addition the temperature had increased to 26 C. A little extra dH2O was added to help with the interphase mixing.
Tar has settled to the bottom, aqueous layer is in the middle, with toluene/xylene on top.


Tar dripped out.


Aqueous fraction extracted 5 times with 50 ml toluene. 10 minutes settling time between each extraction


Followed by 5 extractions with 50 ml dichloromethane


Followed by 5 extractions with 50 ml ethyl acetate.


Aqueous solution was then placed in a pyrex dish and placed on a small pot boiling water to maintain 100 C until the level of the solution had boiled down into a paste. The paste was kneaded with a spatula until the HCl odour had mostly disappeared.




Ethyl acetate has been found to work well as solvent for crystallising the final product (Eureka, finally a means of cleaning up the goop.).

Ethyl acetate washing up. Don’t do as shown – its unnecessary and more difficult, just place the paste in a suitable container and then mix in the ethyl acetate (approx 80 - 100 ml).


The ethyl acetate soaked paste is poured into a 250 ml RBF and heated until all dissolved, then set aside to slowly cool.


After 8 hours, no more crystals appear to be forming. The flask was then put in the freezer for a few more hours.
The crystal mass fills half the flask, not only lining the glass edge as shown in the picture.






Then as fate would have it, luck ran out. Although the acetone used for the washup had been dried with a couple hundred grams of anhydrous magnesium sulphate, the crystals still dissolved somewhat. This was only realised after the acetone extracts were placed in the same container as the previous run’s acetone extracts, so I can't work out the exact yield (Doh!).
It would appear that either: the PPA.HCl salt (phenylpropanolamine.HCl aka norephredine.HCl) is slightly soluble in acetone, or the acetone was wet. SWIM has no idea, yet.
Either way, the dried yield was found to be 10 grams (not including the amount still in the acetone).
Judging by previous runs, the actual yield is probably at least several grams higher.

In the workup it was obvious that some benzaldehyde has remained unreacted, indicating that a higher dilution should be used in future runs.
Any suggestions on how to improve this process further?

Additional note: the ethyl acetate used was distilled from "acetone free" nail polish remover, and was later found to still contain a little isopropyl alcohol.


[Edited on 10-6-2006 by CycloKnight]
View user's profile View All Posts By User
Vitus_Verdegast
National Hazard
****



Posts: 297
Registered: 5-12-2004
Location: Ottoman Empire
Member Is Offline

Mood: tea time

[*] posted on 10-6-2006 at 08:39


By Jove!! Is that the Daily Mail you're reading ?! :o:o:o




Just kidding.... Good work!
I'm afraid I haven't read up enough on this reaction to give any suggestions, only the reports of failure that were posted a couple of years ago by somebody named 'DRIVEN'.

I think Beilstein should be able to give you PPA solubility info or at least where to find it.

[Edited on 10-6-2006 by Vitus_Verdegast]



Sic transit gloria mundi
View user's profile View All Posts By User
CycloKnight
Hazard to Self
**



Posts: 68
Registered: 4-8-2003
Location: United States of Britannia
Member Is Offline

Mood: Still waiting for the emulsion to settle.

[*] posted on 10-6-2006 at 13:56
Akabori runs


........only the reports of failure that were posted a couple of years ago by somebody named 'DRIVEN'.

Its easy to carry out a sucessful reaction, the workup however is quite tricky.

The general procedure for amine extraction cannot be used here. The general procedure being:
-extraction of final reaction mix using dilute HCl
-basification with 25% NaOH solution
-extraction with organic solvent
-washing organic extacts to remove dissolved NaOH
-drying & gasification with HCl to give the amine salt
-washing the salt in suitable solvent

The reason is mainly because this freebase dissolves quite well in water, prohibiting the organic freebase extracts from being washed prior to gasification, resulting in lots of NaCl formation (and other garbage) in the product. The dissolved NaOH also prohibits the use of vacuum distillation for distillation of the freebase directly, since the NaOH will destroy it at the temperatures required for the vac distillation.
If the freebase/organic solvent mixture is washed to remove the NaOH, the product ends up in the aqueous washes and is very difficult to remove.

Even if the aqueous washing step is skipped in the general procedure, the final crystallised product (and salt!) forms a suspension in the gassed toluene solution, and forms an orange soup-like mix that still requires further purification.
The coloured impurities remaining from that step do not dissolve well in toluene, but do in acetone (with some loss of product).
If one chooses to use the photo procedure posted earlier and skip the HCl gassing/freebase step, then the final brown paste cannot be purified using acetone alone. Acetone was used only to purify the ethyl acetate soaked crystals (BTW, this was because SWIM had run out of dried ethyl acetate)

The use of ethyl acetate for final crystalisation has given the best results seen so far, definitely a step in the right direction for OTC workup.

The more conventional procedure is listed in Chemical Abstracts, Volume 47, column 3347 uses N-methyl-d,l-alanine instead for the synthesis of d,l-ephedrine. In that procedure they use 50 grams of benzaldehyde & 20 grams of the amino acid, mixed together and heated (without solvent). 12 g product are claimed.
When that procedure is used with alanine there is little or no benzaldehyde recoverable from the mix, it is simply mostly converted to tar, PPA, water & CO2 during the reaction.

With the xylene method, about half of the benzaldedhye is recoverable from the solvent !!
Therefore, taking only the REACTED benzaldehyde into account, the reaction efficiency is increased somewhat.

On the other hand, recovering the aldehyde from the xylene mixture by vacuum distillation isn't exactly ideal so what SWIM is aiming to find is the optimal dilution & reaction time that ensures that all the aldehyde is reacted thus giving (in theory at least) the highest reaction efficiency, without the hassle of recycling unreacted aldehyde.
During the xylene run posted ealier, solution colour began to change from light brown to orange after ~20 ml aldehyde had been added. It also appeared that the last 10 ml (30-40 ml addition) aldehyde was simply getting converted to tar.
This could indicate a target dilution, as well as an aldehyde to amino acid ratio, to aim for in order to decrease aldehyde consumption wrt yields.


As for product purity, this is what is known so far:
1) The light cream coloured product does react in warm NaOH solution to yield a light oily layer floatin