OTC pyridine production

[megalomania]

In the formaldehyde thread:
http://roguesci.org/theforum/showthread.php?s=&threadid=2089&highlight=pyridine
We touched upon some possibilities of obtaining pyridine from OTC sources such as extract of black pepper. Unfortunately that sources seems rather tedious and possibly expensive. Some industrial type catalytic process may give results, but those are always difficult to reproduce. I cited a few patents in that thread.

I have found a chapter in a book that prepares pyridine derivatives through various condensation reactions and I would like some users input on the applicability of these reactions. The specific information is a series of reactions called “Hantzsch’s Collidine Synthesis” originially published in 1882. Collidine for those who don’t want to look it up is trimethylpyridine. The reaction proceeds thusly:

Acetaldehyde and ammonia combine with 2 moles of ethyl acetoacetate (prep given on orgsyn) to form an alkylidene bis-acetoacetic ester which undergoes ring closure during the reaction to form the cyclic pyridine derivative ethyl dihydrocollidinecarboxylate (rather like piperidine but with 2 double bonds). Subsequently the ring is made aromatic by reacting with nitrous acid. The carboxylate portions are removed by neutralizing the compound to its potassium salt with KOH, and then heated with lime in the same way benzene is made from benzoic acid.

Now then, collidine isn’t exactly pyridine, but it is quite similar to trimethylbenzene (mesityline), of which I have been studying of late to convert to benzene. Theoretically the mesitylene (or in this case collidine) can be oxidized to convert all those methyl’s into carboxylic acids, and they can be removed just like I said above, by heating with lime. Of course there seems to be some disagreement in the literature I have read as to whether or not one can actually oxidize all 3 methyl groups at the same time. It may take an additional step to oxidize any dimethyl or methylpyridine that remains. Also some stronger oxidizing agents other than potassium permanganate may be needed.

Of course another alternative is to conduct this reaction using different starting materials. Ethyl acetoacetate could be replaced by something without the extra methyl group. I don’t know what this is called, it’s not in Merck, but it looks like this:
O=CH-CH2-COOR where R is the “ethyl” of ethyl acetoacetate… hmm, why don’t I just draw a diagram… see attached image.

With all those double bonds shifting about it is quite possible those extra methyl groups are necessary for resonance :( so we may be unable to proceed directly to pyridine using formaldehyde and ethyl-acetoacetate-less-a-methyl.

If anybody thinks I’m on the right track here I will post the details of the collidine synth.

[Marvin]

I am interested in the synth, Id like you to post it though I dont consider it useful as an OTC method. Ive been after pyridine for some time, and done a fair bit of research.

Using formaldehyde instead of acetalydehyde will certainly work, and Vogels practical organic chemistry has this synthesis. Unfortunatly its in the 5th Ed, 1989, so I dont have it. In removing the methyl group from ethylacetoacetate though you are turning a ketone (more properly a beta keto ester) into an aldehyde and I would expect this to mess up the initial crossed aldol becuase self condensation is now more favoured.

Ethyl Acetoacetate can be made from Ethyl acetate by an ethoxide 'catalysed' reaction, but all the syntheses Ive seen end up using up large amounts of sodium metal. Allthough 'in theory' you can make this by electrolysing molten sodium hydroxide its at this point the whole synthsis looks unfeasable to me for my pyridine requirements.

If you can find somewhere that will supply bulk B vitamins, you can always buy niacin and decarboxylate that to pyridine. Not the cheepest way, but certainly OTC, and about as simple as its going to get. Note that the carboxyl group is on the beta carbon, and that these are more difficult to decarboxylate than alpha or gamma which go pretty easily.

What software are you using to do the diagrams btw?

[megalomania]

I'll get the reference from Vogel's 5th edition Monday, it's at my library.

I have the distinct impression there is a way to make sodium ethoxide without using sodium metal, if that is the usage of sodium metal you were referring too. I can't remember where I saw that info, i"ll have to check my stack of notes...

That niacin route sounds interesting, but probably not economical. I buy B vitamins for myself and they can get quite pricy. Of course the reaction does seem to be an expediant OTC route if one does not require large quantities of pyridine. The decarboxylation of niacin does look pretty easy, and despite the function group of interest being on the beta carbon, I would say it will snap off easy enough with a little extra reaction time. Have you ever tried the reaction?

Lets see, one could get 40 g of niacin for around $14.00 (400 100mg pills). That's only 0.325 moles, and even assuming 100% extraction from the pills and conversion to pyridine that would only give us about 26 mL. The cheapest grade of pyridine in the Aldritch catalog is $31.80 for 500 mL.

I made the graphic in Chemwindow 6.

I am seriously considering setting all my other projects aside and concentrating on alkali metal production. It seems all roads lead to such metals when you try to synthesize things. I tried to electrolyze some molten salts a few days ago and my metals just ignited as soon as they were formed, probably because of the magnesium in the mix. Industry makes it look so easy... bastards.

[Marvin]

You can make sodium ethoxide, and indeed sodium itself fairly easily if you can get magnesium metal. But its just another chemical that cant be recycled at home. My intention is to try a few interesting reactions, make the odd interesting or useful chemical as and when I feel like it, rather than to try recreating the last 200 years of industrial chemistry in my back garden.

Sodium can be made in quantity, but its a matter of buying the right bits and setting them up as if you were doing it industrially. Most people dont want to spend 6 months of their life setting up the equipment and at the end just being able to produce 1 chemical. Unfortunatly this is the kind of dedication kg quantities of sodium require, though it would be very cheep compaired to lab supply places. I would think a stainless steel pressure cooker would do quite well with nickel gauze diaphram and welding grade nitrogen gas flushing. 25kg sacks of NaOH can be bought even here quite cheeply and the power amounts to less that this. Somewhere I have the estimated cost of sodium that I worked out a while ago. The problem of a rigid insulator that will withstand attack by molten caustic soda rears its ugly head though and I wasnt able to solve that satisfactorily. Ive made tiny amounts in a very simple setup, but nothing synthetically useful so far.

The collidine synthesis decarboxylates at the beta position, so Id expect conditions to be hopefully similar. I havnt tried the niacin route and I dont have specific conditions which is one of the big reasons I'm interested in the synthesis you have and the similar synthesis in Vogels which may or may not include a decarboxylation in a later step. I am still hoping to find someone that will just sell me pyridine, but in the event I cant niacin will probably be how I'll go. I want to do a coupling reaction that has to be done in molten pyridine hydrochloride, but nothing most people on this forum would be interested in.

After doing a quick search on US sources for niacin a reasonable estimate would seem to be $35 for 500g of niacin. Tablets can be found that are 500mg up to 2g typically and are only slightly above this price for 500g and 'bulk' powder prices similar for this amount. Its probably possible to buy it in multikilogram quantities and have the price fall much furthur.

[megalomania]

After reading about the Hantzsch pyridine synthesis in Vogel’s I have come to believe that this reaction can be done with an aldehyde rather than a ketone to get an unmethylated version, aka pyridine itself. I believe the reaction proceeds via some sort of imine-enamie pathway that works with either aldehydes or ketones as it is the carbonyl that is important, not that which may be attached to the carbonyl. The ester is still quite important to the reaction.

Instead of using ethyl acetoacetate as the starting material (giving us dimethyl pyridine) we can start with ethyl formylaldehyde. Ethyl formylaldehyde is just like ethyl acetoacetate except it’s an aldehyde, not a ketone. Ethyl formylaldehyde should give us pyridine when reacted with formaldehyde.

I found 9 references for the synthesis of ethyl formylaldehyde, but none seemed all that good. The one patent there was is in Japanese. All I could find then is one tantalizing scrap of information: the crossed Clasien condensation reaction of ethyl formate with ethyl acetate promoted by sodium ethoxide and subsequently quenched with acid.

Ethyl formate should be easy enough to make from formic acid and ethanol, same with ethyl acetate as it is just the ester of acetic acid and ethanol. Sodium ethoxide can be made by adding sodium hydroxide to ethanol.

And now the part from Vogel’s 5th edition, which is so much better than the old version I have: Experiment 8.29 DIETHYL 2,6-DIMETHYLPYRIDINE-3,5-DICARBOXYLATE AND 2,6-DIMETHYLPYRIDINE

Diethyl 1,4-dihydro-2,6-dimethylpyridine-3,5-dicarboxylate. Cool 52 g (51 mL, 0.4 mol) of ethyl acetoacetate to 0 degrees C and add 15 mL (0.2 mol) of 40% aqueous formaldehyde solution, followed by a few drops of diethylamine as a catalyst. Keep the mixture at 0 degrees C for 6 hours and then at room temp for 40 hours. Separate the lower organic layer, extract the aqueous phase with ether and dry the combined organic fractions over anhydrous calcium chloride. Remove the ether under reduced pressure (rotary evap) and transfer the residue together with an equal volume of ethanol to a stout reagent bottle cooled in an ice bath. Pass a steady stream of ammonia gas (from a cylinder) into the solution held at 0 degrees C for 1 hour, close the bottle with a bung securely attached with a wire and set the bottle and contents aside at room temp for 40 hours. Filter the resulting yellow solution to remove a small quantity of almost colorless material and heat the filtrate on a boiling water bath in an evaporating dish until most of the ethanol has been removed, and then cool and crystallize the residue from about 400 mL of rectified spirit. The yield of the pale yellow crystalline dihydropyridine derivative is 36 g (71%) mp 181-183 degrees C.

Diethyl 2,6-dimethylpyridine-3,5-dicarboxylate. Place 35.5 (0.14 mol) of the above dihydropyridine derivative in a 1-L round-bottomed flask and add carefully a cold mixture of 50 mL of water, 9 mL of concentrated nitric acid (d 1.42) and 7.5 mL of concentrated sulfuric acid. Swirl the mixture and heat it cautiously on a boiling water bath until a vigorous reaction, accompanied by much foaming, sets in. When the reaction has moderated continue to heat cautiously for 15 minutes until oxidation is complete and a deep red solution is obtained. Cool the solution, add 100 mL of water and 100 g of crushed ice, and make it distinctly alkaline with concentrated aqueous ammonia solution (d 0.88). Filter off the solid product, wash it with a little cold water and recrystallise it from aqueous ethanol. The yield of colorless crystals of the pyridine derivative is 22.5 g (64%) mp 71-72 degrees C.

2,6-dimethylpyridine. Place an intimate mixture of 10 g (0.04 mol) of the above pyridine diester and 60 g of soda-lime (10-14 mesh) in a 100-mL round-bottomed flask fitted with a still-head and condenser arranged for distillation. Heat the flask gradually in an oil bath to about 250 degrees C, and maintain this temperature until no further material distills below 105 degrees C (about 2 hours may be required). Remove the oil bath, clean the outside of the flask and continue to heat more strongly with a Bunsen burner held in the hand, keeping the flame moving over the surface of the flask. Collect the product which now distills, and continue to heat strongly until the flask reaches dull red heat and nu further distillate is obtained. Treat the distillate with potassium hydroxide pellets so that the pyridine separates, and isolate the latter by extraction with ether. Dry the ether extract over fresh potassium hydroxide pellets, and remove the ether and distill the residue at atmospheric pressure. Collect the dimethylpyridine as a fraction of bp 142-145 degrees C; the yield is 2.8 g (65%).

That’s all folks. This procedure differs significantly from the old method I have, and not just because it uses aldehyde ammonia instead of formaldehyde. It forms the aromatic ring by adding nitrous oxides formed by reacting nitric acid with arsenious oxide.

What do you think the chances of preparing pyridine from the variant of this method are, Marvin? One could prepare their formic acid from oxalic acid and glycerol, then it’s a matter of combining the necessary alcohols and acetic acids to form the esters, and sodium ethoxide of course. Then all the other miscellaneous reagents, none of which seem too far fetched. Some difference in solubilities of the intermediate product will have to be taken into account as we will no longer be preparing a dimethyl dicarboxylate, just a dicarboxylate. Eh… diethyl 1,4-dihydropyridine-3,5-dicarboxylate; and diethyl pyridine-3,5-dicarboxylate; and then pyridine of course. I would think the major contaminate would be 4H-pyran if the aldehyde sections do their intramolecular inbreeding and the reaction with mixed acids does nothing to those double bonds.

EDIT: Ahh, I see the rub here, if we use an aldehyde like ethyl formylaldehyde, what's to stop it from self condensing and forming some huge muto molecule rather than condensing with formaldehyde? We could in theory apply an impromptu acetal protecting group to the aldehyde end of ethyl formylaldehyde with ethylene glycol. This would allow the first part of the reaction to form. Then we add HCl to break the acetal, and subsequent addition of ammonia neutralizes any excess acid, and we continue where we left off. I know, I know, what about the other carbonyl of the ester, won't it form an acetal too? Well, acetal formation dosen't do so good with things other than aldehydes or ketones, but I don't know about esters. Hopefully it won't react.

[knowledgehungry]

You might be able to obtain pyridine from nicotine. For those of you who dont know nicotine is a pyridine ring with a 1-methyl azacyclopentane ring attached to it.

[Mumble]

You'd need a lot of nicotine to get an appreciable amount of Pyridine. This link (http://67.1911encyclopedia.org/P/PY/PYRIDINE.htm) shows several other ways to pyridine. None of them are too viable for a decent level of production according to the site at least. It would be uneconomical as the amount of pyridine actually produced is small. Perhaps a method or two, with further investigation, could yield a new more efficient proceedure.

[megalomania]

I have been looking into some vitamin B3 sources (niacin) and it looks like most of the brands selling niacin are primarily niacinamide, or nicotinic acid-niacinamide mixes. It looks like nicotinic acid is the stuff that causes "flush" when using niacin, and there are many flush free brands. Flush free probably means all niacinamide. The bulk niacin at Wal-Mart is flush free.

Is there anything that can be done with niacinamide? It is just an amide after all. It would seem a trivial workup to hydrolyze the niacinamide by boiling with either sulfuric acid or sodium hydroxide. In fact boiling with sodium hydroxide (or potassium hydroxide) leads to a sodium (or potassium) salt, and that is required anyway for the decarboxylation step if it proceeds like the step in collidine synthesis.

Hmm, I guess I answered my own question. Any OTC niacin should be boiled with a strong base just in case it has any niacinamide. In fact there is a general procedure in Vogel's... Amides may be hydrolysed by boiling with 10% sodium hydroxide solution to the corresponding acid (as the sodium salt). Lets see, it gives an example of another compound which is basicially boiled for 3 hours to complete the reaction, during which ammonia is given off. Since sodium nicotinate is quite soluable in water it won't precipitate, but we just boil off the water to obtain the crystals which will be mixed with NaOH, combine that with lime, mix, and heat to liberate pyridine. I am off to try that right now. To the vitamin store!

EDIT: 1 hour later... Well it turns out Wal-Mart doesn't sell either nicotinic acid or nicotinamide, it sells niacin as inositol hexanicotinate. That stuff as it appears is the predominant "no-flush" version. No matter, I went to another store and bought some niacin that is all nicotinic acid. A quick check of Merck shows inositol hexanicotinate is quite the funny molecule, but would probably be convertable to nicotinic acid via acid or base hydrolysis anyway since it is an ester. The only downside is inositol hexanicotinate is insoluable in water, so it may take awhile longer to react with an aqueous base. On the plus side digestion with water could remove some of the soluable components of the pill beforehand.

[thrall]

Sorry to bring up an old topic,but what were your results megalomania! Looks like pyridine IS OTC;).

[frogfot]

Darn, never noticed this thread..

I heard that ethoxide can be prepared by boiling ethanole with NaOH and destilling away water azeotropically with benzene. Any idea on other solvent than benzene? This would be a fun project to investigate.. :)

We could in theory apply an impromptu acetal protecting group to the aldehyde end of ethyl formylaldehyde with ethylene glycol. This would allow the first part of the reaction to form.

As far as I see, the carbonyl groups that will be protected suppose to take place in reaction by resonating.. When they're "fixed" by protective groups, enol will not form, therefore, aldehyde will not connect to it. Though I'm just caviling to the details..

Btw, any idea where diethyl amine could be ubtained? Or, maby there can be used some other catalyst..

[megalomania]

Due to the onset of colder weather I don't get much opportunity to do experimentation in my lab. We shall see in the spring. I am actually eager to attempt some of the old 19th century high temp catalytic reactions and see what kind of yields I can get.

[snuk5]

Marvin,

__________________________________________________ ______________________________________

The problem of a rigid insulator that will withstand attack by molten caustic soda rears its ugly head
__________________________________________________ ______________________________________

I've also looked into the problem.
What do you think about somthing ceramic, possibly porcelain?
Does it have to be a nickel membrane, could stainless steel not work
(for at least a few runs)

[Marvin]

Ceramics seem to be mostly silicondioxide/silicate materials. These tend to dissolve/disintegrate in caustic soda. An experienced chemist has suggested magnesia mixed with a tiny amount of sodium silicate solution, moulded and baked at about 1100C. This should only dissolve only very very slowly. I dont have a furnace so I wasnt able to try this, if you can regulate that pet dragon of yours this might be an option. Steel or iron cage will probably work fine. There is quite a lot of info in threads here and on sciencemadness about sodium manufacture youd do well to read as much as possible before doing a design. The major gotchas seem to be sodium metal shorting electrodes to eachother or the diaphram, and the number 1, hydroxide too hot (25 degrees over melting point and no sodium forms).

[FragmentedSanity]

If someone wanted to make pyridine using OTC B vitamins, would they need a propper lab - or would it be possible to improvise? Is the synth simple enough for your average Kitchen chemist to follow a "recipie". Id often wondered about such a process, so this topic got my attention when it first came up - but Ill freely admit that I dont have the knowledge to devise a process by which one could achieve it. I had hoped to see something that would allow one to produce small quantities themselves; but as yet that hasnt occured. Would anyone care to tackle the project of writing up a synth aimed at "kitchen chemists" - simplicity being the main objective along with ease of precursor acquisition. The economics of the synth would be of far less importance.
I for one would be forever grateful - Im sure others would like access to small amounts of pyridine without the hassle of buying a watched chemical.

[Marvin]

From niacin, yes, same basic reaction as the production of benzene from benzoic acid. Slightly harsher conditions required though I would think.