The theory of OTC tetrahydrofuran (THF) synthesis

[megalomania]

Some months ago I desired to use THF as a solvent for a Grignard reaction in the preparation of benzene from para-dichlorobenzene. Since then I have learned THF is not necessary, but I did begin a side journey in learning how this solvent could be made from all OTC materials. THF is still a very valuable solvent useful in numerous chemical reactions of interest to the improvisational chemist. This usefulness has also earned THF a spot on the DEA’s list of watched chemicals, so some may be wary of purchasing this chemical from reputable vendors.

THF is of interest to those who synthesize the drug GHB as its direct precursor 1,4-butanediol can be used to make butyrolactone, the precursor for GHB. There is an excellent article on the Hive describing how THF can be distilled from OTC pipe cement (for plastic PVC tubing). Some question as to the purity of this form of THF is still being debated, but for most having an easy OTC source is nice. Unfortunately, this particular source is quite expensive for obtaining THF in any but small quantities. I think it averages out to something like $100 a liter when procured in this manner. That’s way to expensive for the likes of me.

Making THF is easy enough once you have 1,4-butanediol, but getting that precursor is the tough part. I think they regulate that as well since it is considered a GHB precursor. My first instinct was to try some sort of chain extending reaction to create a butanol molecule. The tricky part is having the hydroxyl on either end of the butane chain. One method of extending a chain is through the Grignard reaction of an epoxide. This epoxide, for example ethylene oxide, can be added to a halogen containing molecule. The good part about this reaction is it turns the oxygen of the epoxide into an alcohol, and that is half the work. Now all we have to do is make sure the halogen containing molecule is an alcohol itself. That molecule being ethylene chlorohydrin. The ethylene chlorohydrin has an alcohol at one end and the ethylene oxide adds an alcohol to the other.

The good news here is that one can make ethylene oxide from ethylene chlorohydrin, and one can make ethylene chlorohydrin by electrolyzing a solution of ethyl alcohol and salt. The jury is still out on how much ethylene chlorohydrin can feasibly be made via electrolysis as they usually require lengthy periods of time to return any useful yields. The bad news is the hydroxyl hydrogen of ethylene chlorohydrin will interfere with a Grignard reaction. That means we need to append a protecting group, and protecting groups tend to be a bit on the sophisticated side. One might as well just buy THF if you have to resort to buying a protecting group for the reaction. I even researched how to make chlorotrimethylsilane, a protecting group for hydroxyls, but that turned out to be a rather complex reaction, although doable.

Subsequently I have found a possible alternative method that closely resembles one of the earlier industrial methods of THF preparation. This process satisfies the requirements of all OTC reagents and can potentially prepare large quantities of THF rather cheaply. The process begins by reacting acetylene gas with a 50% solution of formaldehyde to make 2-butyne-1,4-diol. This is a rather straightforward reaction that combines these two chemicals. Both are available OTC and do not require any special equipment. Then comes the hard part. The second step is to hydrogenate the 2-butyne-1,4-diol to 1,4-butanediol. Catalytic hydrogenation is a quite useful tool in chemistry, but is it as useful on an improvisational scale? I have been studying quite a bit about catalytic hydrogenations lately and I believe it can be done.

The problem with a catalytic hydrogenation lies in our choice of catalyst and our ability to sustain tremendous pressured. Industrially, or in a professional laboratory, they would use a special hydrogenation tank pressurized to 100-300 atmospheres and use expensive platinum oxide catalysts. These would be the optimal choices for the complete hydrogenation of an acetylene bond without hydrogenolysis (clevage) of the alcohols. On an improvised scale we lack the ability to pressurize a container any more than a few atmospheres, and platinum is just to damn expensive.

There are ways around these problems as I shall discuss. First of all we don’t have to use platinum. One excellent catalyst is something called Raney nickel (W-6). Raney nickel has about a tenth of the reactivity of platinum, but it is about a thousandth of the price. The trouble with Raney nickel is you have to make it by reacting a 1:1 alloy of nickel and aluminum metal with sodium hydroxide. The prospect of making such an alloy seems a bit daunting to me. There is an alternative; something called Urushibara nickel, which has a reactivity right on par with Raney nickel. This stuff is made by adding zinc or aluminum metal to a solution of nickel chloride and digesting with acid or base. Acid digestion makes a version called U-Ni-A, which is what we want because the hydrogenation of acetylinic bonds is promoted by the acid. With base you make U-Ni-B, which has an inhibiting effect on our type of hydrogenation. The acid used is 18% acetic acid (you cannot substitute any other acid like HCl or H2SO4, they have a poisoning effect on the catalyst).

I remember inquiring some months ago about the feasibility of acquiring nickel metal from nickels (US 5 cent coins). I think you can react the nickels in HCl to make nickel chloride. If anyone has any information on this I would like to hear about it. I have not looked into this lately. I am sure the Hive must have something because I remember reading about Urushibara catalysts there. I am sure there are ways to easily acquire nickel metal without the hassle of being put on The List. It’s a catalyst anyway, so you can recycle the stuff almost indefinitely once you have it.

Our biggest problem is the lack of our ability to produce high pressures in catalytic hydrogenation. The reason they use such high pressures is to increase the rate of reaction. A reaction that takes 24 hours at 1-3 atm can take 5 minutes at 300 atm! There are a few ways we can compensate for a lack of pressure. The first is time. Industrial people are always in a hurry because time is money. They want to produce product as quick as they can, and if that means a little pressure, so be it. We on the other hand have the luxury of time. Nobody likes the prospect of waiting around hours and hours for a reaction to go to completion, but that is the price of improvisation. The second thing we can do is use more catalyst. Catalysts use is not linear; double the amount of catalyst in a reaction and you can increase the rate by 4-5 times. Since we will be using cheap nickel catalyst we can pile it on. It is recommended one use from 10-20% by mass the weight of catalyst vs. the weight of reactant when using nickel catalysts. That means we need to use 200 g of nickel per 1 Kg of 2-butyne-1,4-diol. Using 400-600 g would greatly increase the rate of reaction. A third thing we can do to increase the rate of hydrogenation is to scale up the reaction. Maintaining the same ratio of catalyst to reactant, 1 Kg of reactant will hydrogenate faster than will 100 g. Scaling up the reaction has a similar effect to increasing the ratio of catalyst. Many people may not want to run this reaction on a multi liter scale, but it will give better results.

What about our choice of hydrogenation container? I have heard at the Hive such a thing as a “wine bottle hydrogenator” which is exactly what it sounds like, a wine bottle. Wine bottles offer large capacity, often 750-mL or more, and have a heavy wall construction capable of withstanding several atmospheres of pressure. The downside of course is glass+pressure= trouble. I would not want such a container exploding on me. These containers may be safer at lower temperatures, but we will have to heat our reaction to increase the reaction rate, and that can weaken the glass. If I had to choose the best container I would use a 20-L propane tank. These tanks offer many advantages including stainless steel construction designed for pressure allowing safe operation up to a few dozen atm, large capacity allowing for a scaled up reaction, and ubiquity since the things are everywhere.

I was researching Urushibara catalysts today in the library, and while reading Feiser’s Reagents for Organic Synthesis I just happened to notice the next page was about hydroxyl protecting groups. Specifically it was about how one can use boric acid to make a boron ester that is easily removed under mild acidic conditions. Imagine my happiness since boric acid is sold in just about every drug store for some purpose that escapes me. I have a ton of the stuff. This puts the Grignard epoxidation route back in business. I know the complexity of a Grignard reaction vs. a hydrogenation reaction is probably 6 of one and half a dozen of the other for improvisational chemists.

All of this may be daunting to some and seem laughable to others, but bear with me I am just theorizing. My theory is backed by hard fact though, and I believe this can be done. There are other issues to cover, like generating hydrogen and pressurizing it for the hydrogenation rxn, producing magnesium metal for a Grignard reaction, getting nickel from OTC sources, the exact conditions of the formaldehyde-acetylene rxn, practical and safe hydrogenator construction, and probably many more issues that escape me at the moment.

I hope my introduction to OTC THF production may serve as a springboard for others to start their own investigations. I know I will be conducting further research into this goal. I certainly don’t like the fact the government has closed the tap on this most useful of solvents and anything I can do to make impotent their misguided Gestapo tactics would be a great service to our community.

[John Ashcroft]

The synthesis involving acetylene seems promising, but the one involving Grignards seems too challenging for something so simple as THF.

As for nickel from coins: http://www.sciencemadness.org/talk/viewthread.php?tid=534

Keep in mind that nickel oxide and nickel carbonate are sold by pottery suppliers: www.clayartcenter.com

[frogfot]

IMO, protecting group on hydroxyl in ethylene chlorohydrin isn't neccessary, because Grignard part is more reactive than an alcohol group. However, the addition of alcohol to epoxide would be probably catalysed by present water, which leads to a problem.. how can an amateur experimenter find absolutely dry solvents?

Dunno about 300 atm, and such complex hyrdogenation, but I would first experiment with simpler things like formaldehyde production since it seems to appear as precursor for many things.

[Tuatara]

Nickel also is to be had from NiCd or NiMH batteries. Can't get better OTC than that!

FYI if pressurising a wine bottle, use a champagne bottle and check very carefully for chips and scratches. The normal champagne process usually results in a pressure of around 4 atm inside the bottle - so I think this pressure could be safely used for the hydrogenation, given the bottle should have been designed with a considerable saftey margin.

[Milamber]

frogfot, the problem isn't that the alcohol would react with the epoxide, the problem is that the alcohol will readily react with the grignard - an R-OH group reacts basically the same as H-OH, and we all know water stuffs things up (this is because grignard's are very strong bases - they react with anything that can be acidic, like water or alcohols, quite readily. This is because a grignard is a much stronger base than either of these types of compounds, easily taking the acidic hydrogen). These bases can act as nucleophiles, and happily react with the grignard reagent, in this case to give ethanol... we dont need that!

As fara as a wine bottle, perhaps you could wrap it with electrical tape to try and prevent flying glass in case of explosion - might not be totally effective, but it should help. But I was thinking, what about using a pressure cooker as a reaction vessel? You could sit a big beaker with your stuff in it in there, pressure it up... I'm just not sure what sort of pressures can they hold? Certainly better than the bottle, and safer, too!!

[frogfot]

Forgot about that one.. This will be too much work then. How do one protecs -OH when there are reactive -CH2-Cl present. For example, the cheapest protective group for an -OH I could find in Vogel is a ketone. But ketone will most likely react with -CH2-Cl. :(
Also, is it possible that HO-CH2-CH2-Cl will polymerise in conditions used to place protective group?..

How is bottle catalyst used, shouldn't this be a continious process?

[Marvin]

megalomania, I dont understand why you think the reduction would be anything other than first order with respect to catalyst. Particually in the case of a hydrogenation I would expect reaction rate to be directly proportional to the surface area of the nickel exposed.

On the subject of THF, a few thoughts occur. Furfural is reported to be prepaired by the dry distillation of sugar. Though if sucrose will do isnt specifically said. It can also be prepaired by acid dehydration? of pentoses, which is what makes me wonder about the exact meaning of the word 'sugar' in the older reference and by treating material with pentoses in with dilute acid and distilling, somewhere I have a method for making it from bran. Removing the CHO aldehyde group in 1 or 2 steps gets you furan, and reducing say with a dissolving metal gets you THF. I dont know the yield from bran but I suspect you'd need sizable quantities, maybe even semiindustrial for a litre of THF at the end.

1 part bran is mixed with 1 part sulphuric acid in 3 parts water.
Mixture is distilled and the distillate saturated with sodium carbonate, then mixed with salt, and redistilled. On adding salt to this distillate furfural seperates.

Suggested oxidising agent here is silver oxide, but I'm sure we can find something cheeper.
Orgsyn has a method for converting the acid into furan though its method for converting furan into THF is hardly applicable. It does suggest that its been done with raney nickel at 50C in butyl alcohol.

On another tack, we know we can make it by dehydration of 1,4-butylene glycol which we cant make easily, but we can make 1,3-butylene glycol easily from acetaldehyde, so this makes me wonder if it could be persuaded to rearange in the vapour phase dehydration to produce THF.

Ethylene chlorohydrin can be made easier than electrolysis by reacting ethene with hypochlorite though I dont have a specific synthesis handy.

Tell us more about the borate ester protecting groups please, they sound useful.

Edit, ethylene chlorohydrin can also be made from ethylene glycol and hydrochloric acid.

[Al Koholic]

Chlorohydrin Synth from lets seeeee....Journal of American Chemical Society....
I have here a paper detailing the results of a bunch of experiments on diols...
Without further ado,

- A mixture of 20.3g of the glycol and 271cc of conc. hydrochloric acid (one molecule of the glycol to 9 of the acid) was heated at its boiling point for 4.5 hours. The product was then fractionated with a Le Bel-Henninger tube. The fractions containing the chlorohydrin and the glycol were neutralized and extracted with ether. The ethereal solutions were dried over anhydrous K2CO3 and distilled. Yield of chlorohydrin was 12%; bp 127.6-128.6; d=1.194. The glycol recovered weighed 10.5g. -

Bad yield, and to get ethylene glycol you will -most- likely have to distil from antifreeze. This sounds trivial but really isn't as it boils at 760mm at 197 degrees and I believe forms an azeotrope with water (not totally sure about this). Either way, one can definetly obtain the pure glycol but using water diluted stuff will lower yield of chlorohydrin even further. The glycol is also verrrrrrrrrry hydroscopic. However, all my bitching aside....this is a cheeeeeeap source and worth the distiling.

Yes, furfural can be obtained by boiling cellulose in hydrochloric acid....orgsyn detailed a synth with corn husks or something...

[megalomania]

I too would expect hydrogenation to proceed in a linear fashion with increasing amounts of catalyst, but I have half a dozen books on hydrogenation sitting right next to me and they all mention the rate of reaction increases non-linearly with increasing catalyst. None of these books give any explanation as to why this is, only that it is an experimentally determined fact of hydrogenations.

I checked up on the procedure in Organic Synthesis for the production of furfural, and it is quite interesting. The starting material is ground corn cobs. I cannot imagine something easier to get. The only other ingredients are sulfuric acid, sodium hydroxide, and salt (this has the makings of a tasty dish). I am sure furfural could be hydrogenated easily, but how to remove the aldehyde? As luck would have it I am planting corn next year.

Ahh… there is an orgsyn procedure for preparing furan that uses furfural as a starting material, although technically the detailed procedure uses furancarboxylic acid. There is another orgsyn procedure that converts furfural into furancarboxylic acid although half is also converted to furfuryl alcohol. Again this procedure requires only sulfuric acid and sodium hydroxide.

There are also several references to using oats as a starting material for all this furfuryl/furan stuff (with patents). I wonder which material would be cheaper? One could go about collecting corn cobs from friends and family, although they would probably think you were a bit off your nut. Corn on the cob is also somewhat of a season thing. A big tub of Quaker oats could be purchased anytime of the year, and oats have the added benefit of already being ground up to bits. I don’t care to imagine how one would go about grinding down a bunch of corn cobs. For those of us in the country there are farms aplenty with corn just waiting to be taken. Actually some farmer may be nice enough to part with their corn cobs, I think the only commercial value they have nowadays is as a fuel or compost.

I found the bit about using boric acid on page 29 of volume 3 of Fieser and Fieser Reagents for Organic Synthesis. The information in the article is tantalizing thin on details, but promising all the same. And I quote:

Protection of hydroxyl groups. Although boric acid esters have been known for some time, these have not been used to any extent for protection of hydroxyl groups. Fanta and Erman1 have found this means of protection useful, especially in a synthesis of dihydro-b-santalol. The esters are prepared in quantitative yield by refluxing the alcohol with one third molar equivalent of boric acid in benzene with constant removal of the water formed. They are readily hydrolyzed in an aqueous medium, neutral, acidic, or basic, during workup.

1W.I. Fanta and W.F. Erman, Tetrahedron Letters, 4155 (1969)


It looks like the downside here is the ester is so unstable to water, which is produced in the formation of the ester, that not much can be made unless you remove the water. I am not sure how this could be done since ethylene chlorohydrin is miscible with water; how would you separate it? Maybe calcium oxide or anhydrous magnesium sulfate could be added to the mixture to absorb the water? I suppose the presence of benzene (although I think toluene would work just as well) may be able to separate the water. Then all one needs is an improvised Dean Stark trap.

There is a graphic that goes along with the boric acid ester; to sum up 1 molecule of boric acid can esterify 3 molecules of alcohol, so 3ROH + H3BO3 = (RO)3B + 3H2O.

I have read about the glycol methods of making ethylene chlorohydrin, but the reason I suggest the electrochemical method is because I believe it to be the most economical. The starting material is only ethanol and salt, and you don’t get much simpler than that. Of course in my experience electrochemical reactions give a rather small yield despite considerable lengths of time. A reasonably high current source would be needed to drive this reaction to give good yields in a short amount of time. I wonder if the improvised electric furnace setup with its salt water rectifier could do the trick? There is a document that was on the FTP, and it’s somewhere on the net as well that is the scan of an old science book that describes making an electric furnace from line current (sounds dangerous I know). You just yank a plug from an extension cord or something and drop the two ends in salt water connected to lead anchors to regulate the current (by moving them closer or farther apart). I know this sounds exceedingly dangerous, but this book was geared to school children (oh the good old days of 1950’s science books).

Anyway, this is just one example of how an electrolysis of alcohol could be conducted to give large yields in a minimum amount of time. I am sure the situation is far more complex than this. Such a large current load through water would cause it to heat up real quick, external cooling may be needed, or one could start with ice water. I don’t know how safe it would be to run this thing continuously because I would imagine it would still require quite a few hours of running the electrolysis to produce sufficient quantities of product.

[Al Koholic]

If you are really looking for the cheapest substitutes for corn cobs, I believe you could probably get away with using any kind of plant matter at all. Leaves, sticks, grass, etc, etc. If I'm not mistaken, the point of the corn cobs is the high lignin content which provides the 5 membered ring. Any cellulose based matter should be able to be treated similarly.

[Milamber]

Mega, as far as having to remove water for formation of the ester, you're gonna have to do it sooner or later anyway if you want to do a grignard... what about distilling the chlorohydrin/water azeotrope through a drying tube with a good dessicant, and having dessicant in the receiving flask. The success of this would probably depend on the composition of the azeotrope, of course - too much water and the tube would need to be too long and losses would be huge. I think the BP of the azeotrope is around 96oC, but I cant find the composition. Anyway, assuming it worked, you could then form the ester in the presence of dessicant, which should soak up the water that forms.

Hey Al, I think that you've just come up with the perfect incentive for mowing the lawn - I knew there had to be a reason to do it!! ;)

[Marvin]

Lignin is a sort of gum based on aromatic rings but lignin and cellulose generally occur together with sizable amounts of hemicellulose which is composed mostly of pentose monomers.

I think the corn cob method on orgsyn is very important becuase it gives a yeild, and its a damn good yeild. Are corn cobbs sold as animal feed at all?

I understand nbk has a synthesis somewhere of ethylene chlorohydrin by bubbling HCl gas into ethylene glycol, I would think this gets better yeilds, but I dont seem to be able to actually find the method in either the forum or his pdf. He has also listed a brand of antifreeze that is a mixture of only methanol and ethylene glycol, this should be fairly easy to seperate.

The borate ester method looks interesting, but not being acid or base stable is a problem, being water sensitive is also a problem and in this instance Id be worried the boric acid would attempt to esterify both OH and Cl.

On the speed of hydrogenation, this is very strange, but I'll defer to the people with the books.

On electrolysis, this seems like a bad method. The ethanol salt mixture probably does not conduct electricity very well, and as for using rectified line current, this is wasted energy, the high voltage is only overcoming the high resistance of the cell, over 95% of the power used is wasted.

Cannizzaro is certainly one way to go for producing the carboxylic acid from the aldehyde, but as you said this puts half of the furfural into a useless state. Much better would be to find a cheep, mild oxidising agent that will oxidise the aldehyde but not the double bonds. Silver oxide will do this and with recycling this would end up dilute nitric acid which is good or bad depending on your point of view.

[markgollum]

About synthesizing nickel chloride:
NiCl2 can be made by reacting HCl with nickel according to the following equation. Ni(s) + 2HCl(aq) > NiCl2(aq) + H2(g)
I think that this reaction occurs quite slowly, so if you wish to accelerate it, you could probably electrolyse HCl with a nickel anode.
The nickel chloride produced via this equation is green,
because of the hydrated nickel(2) cation Ni(H2O)6 2+.
Anhydrous NiCl2 is yellow.

[megalomania]

I received a large load of books from the library yesterday and as luck would have it I found a synthesis for ethylene glycol that is the best I have seen yet. I rather liked the electrolysis method because up until that point the only method I had uncovered was a mid 19th century process that forced hypochlorous acid into pressurized containers of alcohol that required some weeks. Compared to electrolysis they both seem exceedingly slow, except for the fact improvised chemists are probably far less likely to be able to conduct the pressurized reaction, nor are they likely to be able to make more than small amounts of hypochlorous acid. Electrolysis wins by default. Hmm, that’s not just a 19th century method, that’s the method they used until the mid 20th century when the catalytic oxidation of ethylene came into vogue.

Now of course I have found a chemical method to prepare ethylene glycol from (hopefully) all OTC materials. The problem with distilling ethylene glycol from antifreeze is the stuff is rather expensive. I have a slow leak in my radiator, believe me I hate buying antifreeze because it costs too damn much. I think the scheme I have here is doable and more importantly is quite economical. The process involves producing ethylene gas that is immediately bubbled into bromine and converted into dibromoethane. The dibromoethane is then reacted with a strong base to make ethylene glycol. The glycol is converted into ethylene chlorohydrin merely by adding conc HCl (for some reason only 1 of the hydroxyls is replaced by chlorine).

Ethylene can be generated by heating ethyl alcohol with sulfuric acid, or it can be made by passing heated ethyl alcohol vapors through a heated tube of aluminum sulfate. The ethylene gas immediately reacts with bromine by bubbling the gas into it. I have to check, but I believe bromine can be made from OTC bromide salts quite readily. Of course the bromine can be recycled in this process since upon reaction with sodium or potassium hydroxide the dibromoethane is rendered into a sodium or potassium bromide salt (and ethylene glycol of course).

Nbk2000 does have a method of converting ethylene glycol to ethylene chlorohydrin in his mustard gas document, but I don’t think that is in general circulation (nor does he want it released, so I won’t pass it around). His method involves heating ethylene glycol to 148 degrees C while passing in a continual stream of hydrogen chloride gas. The ethylene chlorohydrin and water are distilled off as they are formed. At the end of the operation the temperature is raised to 160 degrees C to distill off the last traces of chlorohydrin. According to his notes the processing of 100 g of glycol takes 16 hours. The distilled ethylene chlorohydrin is washed to sodium carbonate solution to neutralize and hydrochloric acid, and later washed with 2-3 times its volume of ethyl ether. The ether solution is removed and completely dried over freshly melted sodium carbonate. Yield is about 60%.

I have a note here in a book that says ethylene chlorohydrin can be made from ethylene glycol by heating with concentrated hydrochloric acid. Although that is the extent of it this sounds somewhat easier than passing hydrogen chloride gas into glycol. In fact one could probably just add a gallon of antifreeze directly to a jug of muriatic acid without bothering to distill the former first. The acid will eventually be neutralized as it forms chlorohydrin and water. Some antifreezes come watered, so maybe that isn’t such a good idea all the time.

The same book says ethylene oxide is made from the chlorohydrin by adding concentrated potassium hydroxide. I assume sodium hydroxide would work just as well. Ethylene oxide is soluble in water, but it is a gas at room temp (bp 10.7 degrees C). Since the ethylene oxide must be free of water for a Grignard reaction, the post reacted chlorohydrin and base could be boiled to liberate that gas, the vapors passed through a drying tube to remove moisture. It could also be extracted with ether by shaking it with the (salted out) aqueous solution.

The preparation of ethylene glycol proceeds from 1,2-dibromoethane through the intermediate glycoldiacetate. Glycoldiacetate: A mixture of 63 g (0.33 mol) of ethylene dibromide, 20 g of glacial acetic acid, and 60 g of freshly fused and finely powdered potassium acetate is placed in a short necked, round-bottomed flask (500-mL) provided with a reflux condenser and is vigorously boiled for 2 hours on a sand bath or on wire gauze over a large flame. The flask is then connected by means of a bent tube with a downward condenser and the reaction product is distilled over by means of a large luminous flame, which is kept in constant motion and is made progressively less luminous towards the end of the distillation. The distillate is then mixed with a further 60 g of ethylene dibromide and 80 g of potassium acetate; the mixture is vigorously boiled for 2 to 3 hours on a sand bath as described above and is again distilled. By means of a Hempel column of about 10 cm length the distillate is fractionated and the following fractions are collected separately: 1. up to 140 C, 2. 140-175 C, 3. 175 C to completion. Fractions 2 and 3 are then redistilled separately, when pure glycoldiacetate passes over between 180 and 190 degrees C (the bulk at 186 C). Yield is about 70 g.

Ethylene glycol: In order to obtain free glycol from the acetate ester, the latter is “transesterified” by boiling with a solution of hydrogen chloride in absolute methyl alcohol. This solution is prepared by passing hydrogen chloride gas into cooled absolute methyl alcohol, while excluding moisture, until the concentration is about 3%, as indicated by the increase in weight on a balance sensitive to 0.1 g. If too much hydrogen chloride is passed in more alcohol is added.
Glycoldiacetate (49 g, 0.33 mol) is boiled under reflux in a 200-mL round-bottomed flask with 60 mL of the hydrogen chloride/methyl alcohol solution for 30 minutesand then methyl acetate and part of the methyl alcohol are removed by distillation (slowly at first) through a downward condenser, and the rest under vacuum at about 50 degrees C. In order to separate small quantities of unchanged ester from the glycol remaining in the flask, the latter is closed by a rubber stopper, and the contents are shaken with 50 mL of absolute ether, in which ethylene glycol is insoluble. Adherent ether is then removed on a boiling water bath and the hot glycol is poured into a small distilling flask, fitted with an air condenser, and distilled. The main portion passes over at 195 degrees C. Yield is 17-18 g (80-90%).

The potassium acetate used above is prepared by heating the hydrated salt (made by neutralizing vinegar with potassium hydroxide or carbonate and boiling off the water) over a naked flame in a shallow iron or nickel dish. After the water of crystallization has evaporated the material solidifies. By careful heating the anhydrous salt is now also fused. After resolidification the still warm substance is powdered and immediately transferred to an air-tight container. Commercial anhydrous potassium acetate should also be fused before use.

The preparation of 1,2-dibromoethane (ethylene dibromide) proceeds directly from the synthesis of ethylene gas. A freshly prepared and preferably still warm mixture of 30 mL of ethyl alcohol and 90 mL of concentrated sulfuric acid is heated not too strongly to 160 degrees C with 60 g of fine sea sand (helps remove water) or an equal weight of anhydrous aluminum sulfate in a large round-bottomed flask (about 3-L) over a flame or sand bath. The flask is stoppered with a very tightly fitting 2-holed cork (or rubber). Through one hole is pushed a thermometer which dips into the liquid and through the other a T-tube. A dropping funnel with a long stem is inserted through the upper constricted opening of this tube and fixed with a small piece of rubber tubing, while the end of the side tube is also constricted, and is connected with some receiver bottles (see illistration, #49). Before the cork is pushed in, the stem of the funnel, which is constricted at the end by drawing out, is filled by suction from a mixture of 190 mL of ethyl alcohol and 170 mL of concentrated sulfuric acid.
As soon as the vigerous evolution of ethylene has set in, the alcohol/sulfuric acid mixture is dropped in at such a rate that no copious frothing occurs and a regular current of ethylene is evolved. The temperature is kept continually under control (small flame!).
In order to remove alcohol and ether the gas is passed through a wash-bottle containing concentrated sulfuric acid, and sulfurous acid is removed in a three-necked wash-bottle containing 4N sodium hydroxide and provided with a safety tube. The gas then passes into two moderately wide wash-bottles each containing 25 mL of bromine. In order to reduce loss by evaporation the bromine is covered in a layer of water 1 cm deep and the two bottles are kept cool in a vessel of cold water. If the pressure permits, the last receiver in the train is a Peligot tube containing 2N sodium hydroxide solution; if not, the bromine vapor which escapes is absorbed in sodium hydroxide solution in a corked Erlenmeyer flask (notch the side of the cork!) in which the delivery tube is above the surface of the liquid (shake from time to time). The first two wash-bottles are conveniently closed with rubber stoppers. See illustration #50 for an example of the glassware setup. As soon as the bromine has been decolorized or, at least, as soon as bromine vapor is no longer visible over the brownish-red reaction product (usually after 2-3 hours) the flask is disconnected from the receivers. Then the crude ethylene dibromide is shaken in a separating funnel with water and sodium hydroxide solution until decolorized and washed repeatedly with water. After drying with calcium chloride it is completely purified by distillation. BP 130 degrees C, yield 125-150 g.

Even if one does not appreciate the route to THF that uses ethylene glycol, it is still nice to have this information handy for mustard gas preperation.

[Milamber]

Hey Mega,

Just thinking... if you're going to make ethylene and convert it into the dibromide, then to the glycol then to the chlorohyrdin... sounds long and tedious and, since working with gases at at least some point, potentially very easy for losses to occur. I have two suggestions that may work - at least worth considering.

First, what about, instead of bubbling the ethylene in bromine, we instead bubble it into aqueous sodium hypochlorite - if memory serves, this should add OH and Cl across the double bond and thus give the chlorohydrin directly. Other advantage here is not having to deal with bromine... Disadvantage is the fact that you're doing it in water, of course, perhaps leading to purification problems.

Second, if this doesn't work, what about using only 1 equivalent of NaOH and, instead of converting the 1,2-dibromide into the glycol, we replace just one of the bromines and get the bromohydrin (which would be a better choice than the chlorohydrin anyways, as it would be more reactive, and higher boiling as well for easier handling). Of course this way you'd have to purify carefully, as you'd have a mix of unreacted dibromide, bromohydrin and glycol, but the bromohydrin should be the major product.

Worth thinking about, anyway!

[Marvin]

I do bilieve Ive accidentally stepped into the twilight zone.

Probably the fault of the last person to give me directions.

If we want a diol from an ene we only need oxidise. Only problem normally is stopping the reaction going furthur, ultimatly to cleaved acids. If memory serves cold dilute and neutral permanganate will do this.

If we are going the ethene route to chlorohydrin though, why are we not using the hypochlorite method?
Bubble the ethene into hyperchlorite/acid (free from chloride!) or chlorine in water (aparently this works too) for hypochlorous acid in situ and get ethylene chlorohydrin directly.

If you do try electrolysis for chlorohydrin, Id be very interested in details, such as electrode dimentions, cell resistance etc. I still think it will be more trouble than its worth, but the method might be useful for other things.

I think the chlorohydrin method of making mustard gas is suitably convoluted and best kept that way.

Edit,
Milamber has basically beaten me to the punch while trying to check a few things, but I'm not sure if alkaline hyochlorite will work or not. I am still clearly in the twilight zone however.

[snuk5]

Re:
__________________________________________________ ______________________________

“ethylene chlorohydrin by electrolyzing a solution of ethyl alcohol and salt”

“try electrolysis for chlorohydrin” & “such as electrode dimensions, cell resistance etc”

__________________________________________________ ______________________________


I have no experience with this particular electrolytic synthesis, but I had to device a method to make dimethoxyfuran from furan and methanol in an electrolytical cell.

The fact that dimethoxyfuran is not too stable and the whole thing had to be done in my refrigirator’s freezing compartment did not help the conductivity a lot either!

Because that is the whole rate-determining factor in syntheses like these,

I’ll give some recommendations to the problems at hand:

1) Consider some other electrolyte than salt. It will be very insoluble in EtOH. (Unless you feel that you can have quite some water present but I don’t think you want that…) Perhaps ammonium chloride could do the job. I used ammonium bromide at the time, but I could do so as the halogen derivative of my compound was only an intermediate, not a product that has to be isolated.
2) The cell resistance is inversely proportional to both the surface area of the smallest electrode and the distance between them.
Use a graphite rod for your anode; they stand up very well in non-aqueous electrolyses. (Get it from the people who make up carbon brushes and electrodes for EDM engineering.) If you can’t get this, break up some batteries and put the anodes in parallel.
3) For the cathode use stainless steel mesh; roll it into a cylinder around the anode. “Sew” this up with strands of the mesh and keep the distance small -use spacers at the bottom and the top. (Preventing shorting and allowing for adequate flow).
4) Provide some means of agitation, (magnetic stirrer) as local depletion of the electrolyte slows down things.
5) Depending on the current and surface area, nature of the reaction cell, cooling might be required. (We are pumping electrical current here) But if things don’t decompose, the higher the temp, the lower the resistance and thus the lower the voltage we can employ.
6) Electrolytic synths take time, but they proceed smoothly and just about quantitatively -the driving force is a current that we can modulate by fiddling with the voltage. (Start with higher V than for aqueous electrolyses.)
7) Use an ammeter, as well as a thermometer.
8) If you keep a log of the time and the current, you can take your pocket calculator and get a nice idea of how many Coulombs have gone down and thus how many moles of product you have made already!


If I were to do this thing and I meant business –scale wise, I’d use my DC welder as a current source and cool the reactants with coils of 8 or 10 mm stainless steel tube and run water through them. If I’d find that the voltage were much higher than required, I’d put two banks of electrodes in series.

Coincidentally, I was looking for THF myself some time back, if you don’t buy it by the truckload, but from your local lab supplier, it costs the earth.
My friendly supplier traced two 250 ml bottles of AR grade that went for a song.

I’ll save this page, just in case, as I reckon it’s totally feasible and should I have no other source, I’d whip up a going pilot cell in a matter of days, and then go the big Tupperware route. Without all the custom-made ground jointed stuff that I used at the time.

Come to think of it, I was looking for a hypothetical source of 1,4-butanediol (even harder to get than THF), for some hypothetical jet printer ink formulation; this OTC stuff is great!

[FlaAlchemist]

Did the grignard with the p-dichlorobenzene work?
Was it a traditional grignard rxn in ether solvent?
If so, what was the major product?
I'am curious because the p-dichlorobenzene is readily avalable as a starting material.
I have used p-dichlorobenzene by adding a NO2 in the ortho position to deactivate those chlorines on the ring.

[megalomania]

A little off topic, but I can't find a thread just for benzene, so... The Gringnard of p-dichlorobenzene did not proceed so well. After several trials is is damn hard to initiate the reaction. The only reason I am back in this thread is because I am getting ready to build myself a poor-man's hydrogenator that I hope will withstand a good deal of pressure. While refreshing myself on hydrogenations last night I came accross a chapter about hydrogenolysis of aryl-halogen bonds. Apparently hydrogenolysis of aryl halogens is quite easy under mild conditions (those conditions being a few atm's of pressure, nickel catalysts, and about 100 C).

This is good news for the production of benzene from p-dichlorobenzene. The conditions required to completly hydrogenate the benzene ring would require much higher pressures, or better still, precious metal catalysts. The order of reactivity is typical of halogens, iodine > bromine > chlorine > fluorine. Apparently hydrogenolysis can even cleav fluorine, which is tough to do most of the time (damn near impossible with a Grignard).

Halogen hydrogenolysis may also be a good way to make aniline from p-dichlorobenzene. p-DCB can be nitrated, and the nitro group will actually make it easier for hydrogenolysis to occur. Under the same hydrogenation conditions for hydrogenolysis of chlorine, the nitro group is hydrogenated to an amine, in this case aniline.

Experimantal conditions would consist of p-DCB dissolved in 95% ethanol with either sodium hydroxide or potassium acetate as an acid scavenger. The presence of hydrochloric acid would poison the hydrogenation catalyst, hence the need for a base. The reaction can be carried at room temperature and atmospheric pressure with precious metal catalysts. Since nickel catalysts have less activity it would probably be best to heat the reaction and raise the hydrogen pressure a few atmospheres. One should be careful because eventually you will hydrogenate the aromatic ring.

[FUTI]

Mega maybe you could try one Grignard modification I read few years back. Instead of lightly heating the reaction to onset, they just imerse the flask with reaction mixture in ultrasonic bath.

Maybe even electrolysing the benzoate solution or heating the sodium benzoate/sodium-carbonate mixture can get you some benzene. Sorry I must ask, why? Just food for thought, real problem or what. I can't believe that you can't find available benzene. Solution with hydrogenation is elegant you bring some of mine memories. Some parts of chemistry are acctually not so easy. Most of reaction in organic chemistry are understandable and self-explicable but there are few areas that defies this rule as catalytic hydrogenation and electroorganic reaction.

Rest of the thread is fine. For those who want to experiment little in toxicology I recomend to convert a 2-chloro-ethanol to epoxide(carefull gas not to mention toxic one) and then heat that one with KSCN to replace O with S, you are now very close to sulfur mustard gas...and man if this could work with epychlorohydrine it would make cheapest hell on earth ever.

[megalomania]

It seems the post I made about my experiment this past summer was deleted in the great crash. In particular nbk2000 described a method of making an improvised bubbler from Teflon tubing? Would you describe that process again nbk? I don't remember the specifics.

I am going to try an build a high pressure hydrogenator using a "pipe bomb" design. The burst pressure of common steel alloy is in excess of 7000-13000 psi for a 2” pipe, and the working pressure is half that. A max pressure of 1500 psi (100 atm) is quite sufficient. Apparently as the diameter of a steel vessel increases its max burst pressure drops. This explains why air compressor tanks have a max pressure of only 150 psi, whereas a ¼” pipe has a minimum burst pressure of 13500-17000 depending on wall thickness.

I was thinking of connecting two pipe sections together along with a U joint that includes a ball valve in between the parts. The hydrogen can be generated chemically in one section of the pipe, while the hydrogenation reaction is held in the other. If extra hydrogen is needed over the course of the reaction the valve can be closed, the hydrogen generator portion opened, and additional chemical reactants added. The valve will keep the hydrogenation section from being depressurized.

I was also thinking about molding a loose fitting concrete shell for the pipe just for safety reasons. The concrete can be reinforced with chicken wire so it does not crack and shatter should an explosion occur in the pipe. This may make it difficult to apply heating and stirring to the pipe though. I could rig up a rocker system like they have with Parr hydrogenators…

[nbk2000]

Take a length of teflon tubing, clamp one end shut with Vise-Grips, leaving about a 1/8th inch sticking out, and use a low flame on a lighter to melt the end. You may use a spoon to smooth the end while its still soft, and immerse it ASAP under cold water to set it. :)

A hot needle makes the holes for the gas to bubble out through.