Sciencemadness Discussion Board - thionyl chloride and alcohols

thionyl chloride and alcohols

woelen - 30-5-2008 at 13:03

I mixed some methanol with thionyl chloride and this gives a very violent reaction, in which copious amounts of HCl are evolved. I did an experiment, with 1 ml of methanol and 0.5 ml or so of thionyl chloride and mixing these. A lot of gas is produced and I immediately loosely stoppered the test tube. After the reaction stopped, I put the test tube under water and removed the stopper. AT ONCE, the test tube is filled with water for almost 100%, only a small bubble of gas remained. So, the gas must almost be 100% HCl. This is in agreement with the following:

http://www.lambdasyn.org/synfiles/dimethylsulfit.htm

I, however, had the impression that thionyl chloride also can be used for chlorinating compound with hydroxyl groups, e.g.

ROH + SOCl2 --> RCl + SO2 + HCl

Is this reaction possible? I would like to make some CH3Cl from methanol. Can SOCl2 be used for this purpose, under some different reaction conditions? I know of other methods of making CH3Cl, but they are not really easy.

The wiki on thionyl chloride mentions the chlorination of alcohols as a possible reaction:
http://en.wikipedia.org/wiki/Thionyl_chloride#Organic_chemis...

Is this wiki-page wrong?

[Edited on 30-5-08 by woelen]

Formatik - 30-5-2008 at 14:41

There's some information in Beil about making CH3Cl from MeOH.

Beilstein 1, 144
Beilstein supl. 1, 33
Beilstein references

Fleaker - 30-5-2008 at 15:01

Thionyl chloride will chlorinate alcohols into their alkyl halides. I don't think methanol is an exception--you will have methyl chloride, HCl and of course SO2 produced. I think the methyl chloride all boiled away during the initial violence of the reaction.

Klute - 30-5-2008 at 15:29

But he would have obtained some SO2 afetr dissolving the HCl in water, no?

I think alcohol which form less stable carbocations, like methanol, can not be deplaced by the Cl- anion once the thionyl chloride has added to the oxygen, without further heating...

I guess the formation of the dialkyl sulfite is limited to methanol and other short primary alcohols.

This is only speculation though.

Woelen, can you not substitute MeBr for MeCl? The former cna be esaily genearted with methanol, alkali bromides and H2SO4/H3PO4 as Painkilla demonstrated... Or do you specifically need MeCl?

chemrox - 30-5-2008 at 23:07

I need an alkyl chloride too. One not commercially available from the usual suspects. The bromide is resistant to the reaction being planned. I am considering two approaches; DMF/TCC complex, a la DeLuca, looks promising, @Sauron (see thread) and the classical, ZnCl2/HCl method. Both should work. The first more efficient than the second. SOCl2 would work too but I don't want to use it when the other methods are available and the SOCl2 might be needed for something else.

Sauron - 30-5-2008 at 23:58

Thionyl chloride and methanol in stoichiometric quantities will give MeCl, SO2 and HCl.

All of these are gases at STP. What did you expect?

Methanol was simply not a very good choice of substrate.

Try this with, say, cyclohexanol, and run the reaction with a stoichiometric equivalent of pyridine or TEA or doethylaniline to trap the HCl, and you will get cyclohexyl chloride in decent yield.

Vogel has some examples of alcohol to alkyl chloride preps with SOCl2.

Running the reaction in the cold will moderate reaction rate and make things tamer.

If you want to make MeCl, use methanol and TCT and reflux at about 15 C under the bp of the methanol. Use a Dewar condenser to collect the MeCl and store it in a stainless steel sample bomb (pressure vessel.)

Nicodem - 31-5-2008 at 02:26

In the Organikum, a typical practical chemistry book, the authors claim that: "With thionyl chloride there exist two different possibilities: in any case the ester (R-O-SO-Cl) forms first, which in presence of pyridine reacts with the chloride ion leading to an inversion of configuration, but in the absence of pyridine an 'internal nucleophilic substitution' occurs (SNi) which leads to a product retaining the original configuration." (my translation)
So chances are that it is possible to prepare MeCl with SOCl2 even without using pyridine or other bases. However, whenever I was using SOCl2 to form alkyl chlorides from primary alcohols I always added an equivalent of pyridine and left the reaction run at room temperature in CH2Cl2 until TLC indicated more or less complete conversion. That is also the most common procedure you can found in numerous papers, so I don't have any experience using any alternative method in absence of pyridine.

Woelen, your experiment with the absorption of reaction gases in water does not demonstrate that no MeCl formed since according to one of the Beilstein abstracts Schockwave provided above, methyl chloride forms a clathrate with water (though under pressure and lower temperature), so it could be relatively soluble in water. In fact its MSDS claims it is soluble up to 0.5g per 100ml water and surely there could have been no more than a few milligrams in your test tube.

PS: According to SciFinder, supposedly J. Chem. Soc., 1963, 1059-1064 (DOI: 10.1039/JR9630001059) describes the reaction of methanol with SOCl2 to form MeCl, but unfortunately my access to RSC is malfunctioning today so I can not check or post the paper.

Polverone - 31-5-2008 at 08:54

Here's JR9630001059, though it appears to not actually report results for methanol + thionyl chloride.

S.C. Wack - 31-5-2008 at 10:35

Dimethyl sulfite is not the only possible product, methyl chlorosulfite aka methyl chlorosulfinate is possible when there is less than a 2:1 ratio.

garage chemist - 1-6-2008 at 00:07

In Vanino, there is a preparation of MeCl that says that MeCl gas can be conveniently stored by dissolving it into methanol, in which it is very soluble.
When this solution is then dripped into water, the MeCl escapes under fizzing since its solubility in water is very low. This way, a steady stream of gaseous MeCl can be generated.

So woelen, if you want to find out whether your reaction makes any MeCl you have to add your reaction mix to water and see if there is fizzing from MeCl.

Nicodem - 1-6-2008 at 01:05

I suspected that SciFinder entry to be erroneous as there was also another abstract using SOCl2 that also lead to a paper where alcohols other than methanol were actually used. Until I get access to Beilstein I wont be able to find any useful references where this method is tried on methanol. However, I thought about it and considered what if MeOSOCl in methanol (without pyridine) reacts faster with methanol rather than the chloride ion? This would explain the procedure for preparing dimethylsulfite. On the contrary, in CH2Cl2 using stoichiometric amount of MeOH and SOCl2 with pyridine the concentration of pyridine hydrochloride is higher than the residues of unreacted MeOH so that all MeOSOCl reacts with the chloride ions rather than forming the dimethyl sulfite. At least that is the only way I can currently explain this anomaly.
So, Woelen, could you try again using the standard procedure (with or without pyridine) and dichloromethane as solvent? MeCl is surely highly soluble in dichloromethane, thus a washing with 1M NaOH(aq) should remove all unreacted SOCl2, SO2 and HCl, followed by 1M HCl(aq) to remove pyridine. However I'm unsure on how could you prove that MeCl formed without analytical instruments in an amateur setting or without derivatization. If it is present the solution will start boiling bellow the bp of dichoromethane, but this is no definitive proof.

woelen - 1-6-2008 at 04:21

I had a near-accident when mixing triethylamine and thionyl chloride

. These two react insanely violently with each other.

The idea was to make a dilute solution of SOCl2 in triethylamine and let this react with cyclohexanol. I added 0.5 ml of thionyl chloride to 2 ml of triethylamine and the stuff almost exploded

. The contents of the testtube became black and part of the triethylamine is charred. I also obtained a black plume of smoke (presumably carbon I think). Before I added the thionylchloride, I saw white smoke, but that was what I expected (HCl and SO2, formed in the humid air give white solid triethylamine hydrochloride and sulfite), but as soon as the liquid SOCl2 touched the liquid triethylamine the contents was blown out of the test tube, there was a lot of crackling noise and a black plume of smoke came out of the test tube. Luckily, the test tube was not pointed towards me.

I skipped the experiment with cyclohexanol, I first need a method of moderating the reaction. triethylamine apparently is not the moderator of choice.

Right now, I'll think about moderation of reactions with inert solvents. Could the method, proposed by Nicodem be done with ligroin (bp range 60 ... 90 C)? I expect CH3Cl to be very soluble in ligroin as well, and it is easy to see whether there is CH3Cl in the ligroin or not.

Nicodem - 1-6-2008 at 05:16

Quote:

Originally posted by woelen
I had a near-accident when mixing triethylamine and thionyl chloride

. These two react insanely violently with each other.

That's what you get if you play solventless with thionyl chloride!

Good you were not hurt, but rather be careful with this stuff. It reacts violently with a lot of stuff. Except for preparing acyl chlorides from carboxylic acids it is nearly always used in solutions, generaly in chloroform or dichloromethane, and with cooling.

As for using ligroin instead od CH2Cl2, I just set the reaction and while I was at it, I did not add any pyridine out of curiosity on what will happen. I do not have ligroin, so I mixed 20ml n-heptane with 30ml petroleum ether (bp 40-60°C). I guess this mixture is quite close to ligroin by properties. To 50ml of this solvent stirring in a flask was added 2.03ml methanol (50mmol). It took some time, but it finaly dissolved in a clear homogenous mixture. To this mixture stirring in a cold water bath was then added dropwise a solution of 4.4ml SOCl2 (60mmol) in 10ml petroleum ether. Not much happened. Some HCl is forming as expected, indicated by checking the exhaust gases with a moist pH paper. I will let it stir for a day or two and then check the presence and ratios of MeCl, MeOH, MeOSOCl, (MeO)2SO, etc. via HNMR (the alkanes have all signals bellow 1.5 ppm, so it should be possible to scan the reaction mixture in CDCl3).

Klute - 1-6-2008 at 05:27

Nice!

At least we will be sorted out as to what happens during the reaction with SOCl2 and methanol! Thank you for your time Nicodem.

Sauron - 2-6-2008 at 03:59

Well, all that about inversion vs retention of configuration only applies to secondary alcohols, and then only in cases where a chiral center can result (R1 does not equal R2),

There are other reagents that will give MeCl from methanol, most of them will also spit out HCl, but few also SO2. That makes for a somewhat more straightforward isolation process of the methyl chloride, as the HCl is easily scrubbed away, leaving the MeCl to be condensed.

TCT is one such. As I mentioned upthread somewhere, just refluxing TCT in MeOH at about 15 C under the bp of MeOH will work just fine. Excess MeOH serves as solvent. Yields are good, on order of 50-60%.

A better procedure is to employ the TCT/DMF complex and react that with methanol in DCM at room temperature, the yield is near quantitative. I believe I have posted papers on both of these reactions previously.

woelen - 2-6-2008 at 04:30

I have read quite a few reactions with TCT, but unfortunately I don't have it and I have still not found a source for this. I do have a lot of special chems, but this particular one is hard to find (at least for me). Maybe someone who reads this thread could tell me a source of TCT in the EU?

kmno4 - 2-6-2008 at 06:09

Something interesting to read:
A Study of the Reaction of Alcohols with Thionyl Chloride
William E. Bissinger, Frederick E. Kung
J. Am. Chem. Soc.; 1947; 69(9); 2158-2163.
and
Factors in the formation of isomerically and optically pure alkyl halides. Part X. Reactions of saturated aliphatic alcohols with thionyl chloride
J. Chem. Soc., Perkin Trans. 1, 1976, 104 - 108

Sauron - 2-6-2008 at 09:14

@woelen, I buy my TCT (cyanuric chloride) from Acros in Belgium. One of the main manufacturers is Lonza in Italy. The prime manufacturers are usually ion the HCN business because TCT is a downstream product via cyanogen chloride. The companies in question make things like malononitrile, chlorosulfonyl isocyanate etc. Lonza I believe sells TCT (CC) in 20 Kg sacks or drums. Acros, Merck etc. probably just repack it rather than produce it themselves.

Where I am I can import TCT no problem but thionyl chloride is a no no.

It is possible to prepare TCT in the lab by a modification of the industrial method but this is highly dangerous. The safer method is to chlorinate methyl thiocyanate. TCT crystallizes out of the mixture that forms in good yield. Continued chlorination of the supernatant liquor eventually gives CCl4. Do note that the intermediate mixture is toxic and highly obnoxious, containing perchloromethyl mercaptan and thiophosgene and sulfur chlorides. But this is still safer than mucking around with neat HCN and ClCN.

woelen - 2-6-2008 at 22:54

Can you buy from Acros, as a private individual? I personally have very bad experience with the big chemical companies, they only sell to companies and research labs from universities and so on. I have all my chems from 'less official' suppliers, e.g. from eBay, or from hobby supplies, who sell to individuals. But such sources usually have a limited range of chemicals (e.g. only pyro stuff, or only common organics).

Klute - 2-6-2008 at 23:38

Without wanting to go too offt opic, try localdistributors, they will ahev most acros, alfa aesar, Fluka etc etc products. I buy all my chems from my local glassblower, the cheapest are often from Panreac a spanish manufacturer, very close to where I live (south of France).
Try to see if there are any local companies that might reatisl chemicals, and contact them. Obviosuly, at first I would go and ask about other things that PCl5, chlorosulfonic acid or even TCT, and then work my way up once there's a good relathionship going on.
How did you get your thionyl chloride? Surely not on ebay or similar? pardon my curiosity.

Sauron - 3-6-2008 at 00:31

I also buy through the local Acros distributor and sometimes through another Thai company who buy from same distributor. I also deal with them rather than Merck's local office. There is a local agent for Fluka and Aldrich, although Aldrich is usually ridiculously costly due to Hazmat bullshit not imposed by European vendors. There are also local agents here for Panreac, Ajax and Carlo Erba.

No agent here for Alfa.

Klute - 3-6-2008 at 02:04

Localpanreac agent in Thailand? I'm impressed! i didn't think it was such abig company. I thought it didn't go further than western europe. Good news then. Although slightly overpriced in some instance, I have always been very satisfied by their purity and the speed in which the products are delivered from the plant in Spain to my distributor. It's just a pity that speciality reagents are not pacakged in less than 100-250g, sometimes 500-1000g.

But this is totally OT. Sorry.

EDIT: I just checked, Panreac doesn't seem to sell TCT. Maybe they cna have some on demand, though.

Woelen, are you trying to make MeCl with SOCl2 for academical purposes, or because you need the reagent for a subsequent reaction? I would rather try the MeBr route in the latter case.

[Edited on 3-6-2008 by Klute]

Nicodem - 3-6-2008 at 02:34

Quote:

Originally posted by Nicodem
As for using ligroin instead od CH2Cl2, I just set the reaction and while I was at it, I did not add any pyridine out of curiosity on what will happen. I do not have ligroin, so I mixed 20ml n-heptane with 30ml petroleum ether (bp 40-60°C). I guess this mixture is quite close to ligroin by properties. To 50ml of this solvent stirring in a flask was added 2.03ml methanol (50mmol). It took some time, but it finaly dissolved in a clear homogenous mixture. To this mixture stirring in a cold water bath was then added dropwise a solution of 4.4ml SOCl2 (60mmol) in 10ml petroleum ether. Not much happened. Some HCl is forming as expected, indicated by checking the exhaust gases with a moist pH paper. I will let it stir for a day or two and then check the presence and ratios of MeCl, MeOH, MeOSOCl, (MeO)2SO, etc. via HNMR (the alkanes have all signals bellow 1.5 ppm, so it should be possible to scan the reaction mixture in CDCl3).

Here are the results…

After letting the reaction mixture stand for two days, 0.2ml were taken for an H NMR analysis in CDCl3. There were only traces of methanol left and only very little methyl chloride formed. The main products are the sulfite esters. Unfortunately I was not able to obtain the H NMR spectra of methyl chlorosulfite (MeOSOCl) in the literature, but I would lean toward assigning the singlet at 4.05 ppm to this product. In such case this is the main product and the composition of products in the mixture is:

MeOSOCl (?): 100
(MeO)2SO : 33.4
MeCl : 3.7
MeOH : 0.5
 (experimental error not included)

So I would tend to conclude that this method does not work in the absence of pyridine, though there still is a remote chance that the reaction does perform differently in dichloromethane and that the absence of SNi reaction is due to the very hydrophobic environment of the solvent used. Or perhaps it is just this specific case of methanol simply not working in this reaction.

SOCl2_MeOH.gif - 14kB

Klute - 3-6-2008 at 02:40

Thank you for posting the spectrum, very interesting.

What could be worth a try is doing the reaction in DCM, taking a sample of H NMR, then adding pyridine/triethylamine and retaking a sample. Maybe the amine will promote direct decomposition of the sulfites, or maybe it needs to be present before introduction of the thionyl chloride for the chloride to be produced.

Nicodem - 3-6-2008 at 02:45

I agree that it would be interesting, but unfortunately I don't have the time for these experiments. I'm currently flooded with work and I did the above experiment just out of curiosity, as well as for its potential application since a solution of MeCl in an inert, nonpolar solvent could be useful to me (it is commercially available in gas bottles though).

woelen - 3-6-2008 at 04:39

Klute, for me this experiment with thionyl chloride and MeOH is purely an academic thing, and this discussion on its own already is a pleasure to read (also thnaks to Nicodems very nice contribution). But I certainly will try the MeBr thing as well.

I have the thionyl chloride from a German supplier, who is geared towards the hobby market and sells and ships many fantastic chems to private individuals in the EU (unfortunately no TCT, but he does sell all common ...yl chlorides like SOCl2, SO2Cl2, C6H5COCl, CH3COCl and even (COCl)2, I purchased all of these, except the oxalyl stuff, that is too hefty for me). For more info, please U2U, I don't post such sources here, because k3wls are reading this also.

chemrox - 3-6-2008 at 17:59

The reaction of alcohols and thionyl chloride in the presence of pyridine is a known (literature) procedure. It is even named. It is called, "Darzen's procedure." Why one would omit the pyridine puzzles me. Anyway, you've proved the necessity of the pyridine. Thanks for the spec. How did you manage to paste it like that and make it fit so well?

Sauron - 4-6-2008 at 16:42

I'm sure other appropriate tertiary amines would do just as well as pyridine. Diethylaniline is often mentioned in the lit. in this contest, because its hydrochloride (unlike that of pyridine) is easily filtered and is non-hygroscopic.

Another obvious alternative is TEA (triethylamine).

woelen - 26-6-2008 at 11:36

I just tried the reaction between triethylamine and thionyl chloride, moderated by diluting it 20 times with DCM. If I mix these two dilute solutions in DCM, then immediately after mixing, the liquid becomes yellow, and from that moment on, it fairly quickly turns darker and darker. One minute after mixing, the liquid is totally black and turbid and slightly warmer (35 C or so).

Mixing the pure triethylamine and thionyl chloride results in a very violent reaction with formation of a black solid (I described it earlier in the thread). But apparently the dilute material also produces this black compound at room temperature. So, the black material most likely is not carbon. What could this dark compound be?

If I mix methanol with DCM, and also add a little amount of triethylamine and add a solution of thionyl chloride in DCM to this, then the liquid remains clear and colorless, and there is occasional bumping, when gas is produced in large bubbles.

--------------------------

I also did a related experiment, adding a few drops of pure acetyl chloride to pure methanol. When this is done, then each time when a drop of acetyl chloride falls into the methanol, there is a little explosion. After a few drops, the test tube was cracked. Next time, I'll use a thick-walled test tube. A fruity smell was produced (methyl acetate I assume).

I am really impressed by the violence of the reaction of the acyl chlorides with methanol. The reaction is MUCH more violent than with water.

It is time for me to get more DCM, the reactions with the pure compounds simply are too violent to be of any practical use

kmno4 - 27-6-2008 at 05:54

Quote:

Originally posted by woelen
Mixing the pure triethylamine and thionyl chloride results in a very violent reaction with formation of a black solid (I described it earlier in the thread). But apparently the dilute material also produces this black compound at room temperature. So, the black material most likely is not carbon. What could this dark compound be?

It is can be some kind of plymer. Such polymers are mentioned for reaction SOCl2 with NH3 in paper 10.1021/ic50167a050 (and reference number 3[from Berichte der deutschen chemischen Gesellschaft]). Other amines propably give similar reactions.