[Ingeniosuccinimide]

Hello,

In one my previous topic I posted a question regarding synthesis of thiepane. It seems that the reaction worked, but from the NMR I would conclude that I have isolated a mixture of thiepane and 1,6-dibromohexane (starting material). Now I'm stuck with a colorless liquid and don't know how to separate these two. They seem to have quite simillar properties, they both freeze around 0°C and are soluble in similar solvents. I'm not sure about putting the liquid on hi-vac since thiepane has lower boling point. Any suggestions? I can append a spectrum if anyone is interested.

[kriggy]

You can try to distill one from the other.
You can try extract them - but since you said they are solubule in quite similar solvets it might not work.
Or you can run column chromatography

[Guitarmaniac86]

I havent read the previous thread, but I am assuming you are using sodium sulfide nonahydrate for this transformation? I have used it for synthesis of heterocycic sulfur containing compounds and I ran the reaction in dry DMF at 110 °C for 24 hours. Usually this would give me 100% conversion to product. Have you tried adding an excess of the reagent to force it to completion?

If you could provide the experimental procedure, that could be helpful.

Edit:

Also, just had a thought. Some of the heterocyclic sulfur compounds I have made have come out as an oil but I have managed to re-crystallise them from isopropyl alcohol (2-propanol). You could give this a go with your compound also.

[Ingeniosuccinimide]

Yes, I'm using Na₂S, according to this paper:

Mandolini, Vontor, Synthetic Communications, 9 (9), 857-861 (1979)

in which authors emphasize that by using mainstream polar aprotic solvents like DMSO and DMF they get very poor yields, even below 10%. People from this article did this reaction again in HMPA with much better yields. I'm following the same reaction but it seems I got something quite unclean (attachment) and have no idea how to purify it.

I have traces of pentane in the NMR, possibly also of HMPA and 1,6-dibromohexane but I'm quite uncomfortable to put the oil on hi-vac since the product (thiepane) is also volatile.

When mentioning excess of reagent what do you mean? A dihaloalkane or Na₂S?

One another thing came into my mind: how likely is it that these two might be separated by column or a short silica plug? Which of these two compounds would have larger R_f? They seem to be quite simillar, and I don't know if using a really long thin column and some mostly nonpolar system (e.g. chloroform:pentane=1:10) would work?

[Guitarmaniac86]

I did mean to use an excess of Na₂S after the reaction stalls. When I made theitanes (4m ring) the reaction sometimes stalled after 24 hrs so I added another half equivalent of Na₂S.

Unfortunately I do not have institutional access to the paper so I cant comment on it or make suggestions. However, have you tried looking at DMI or DMPU as solvents?

I just performed a scifinder search on thiepane and I found a source that used EtOH as a solvent and they got a yield of 52%. However, I would be apprehensive using an alcohol as a solvent for the reaction, I would be concerned with the Na₂S reacting with the ethanol. Not sure how likely this is.

I have looked at the spectra and from what you have provided there seems to be a lot of the dibromo compound in there. Isolating it might be hard, but what does your TLC look like? How do the authors say to purify the compound? If they have a solvent system for column chromatography you may need to set up a gradient.

As an example, for the thietane compounds I was talking about earlier in my post, the paper I was following stated I should use 6:1 hexane: ethyl acetate. On my first run, I got the compound with some impurities, but I re-ran the column starting with 95:1 hexane: EA for 2 column volumes, then ran it 9:1. then 6:1 (basically increasing the EA by 5% each time). By running a gradient I got the compound 99.5% pure by HPLC.

[salteen]

Have you considered a typical reactive extraction?  i.e. partition between DCM and mild aqueous acid to protonate the thiepane and force it into the aqueous layer, drain off the organic layer and dibromohexane, and then deprotonate the thiepane to return it to the organic layer.

EDIT: On second thought, you may need quite a bit of acid to force that reaction to take place...

[TheUnassuming]

If your compound is low boiling you should definitely try a distillation like Kriggy suggested earlier. 

If that doesn't work, play around with different mobile phases on TLC to see if you can find a system that gives you the desired separation.  The change in the structure is substantial, so I can't imagine they have identical Rf's in all mobile phases.  In general with difficult separations you don't want to go over the 3:1 (length:width) ratio.  This is the ratio we use in our lab for "snowmen" spots on TLC and get pure products with little co-elution.   A longer column does indeed give you better separation of the average/middle of each compound, but longer columns also spread out each compound as it travels down the column.  The effect is that the very first bit of the first compound off the column is pure, but then you get a giant smear of both compounds followed by a pure second compound at the tail end.

[Ingeniosuccinimide]

However, have you tried looking at DMI or DMPU as solvents?

Actually I did, and I found we have approx 100 mL of DMPU. The only clue here is that I don't know how do DMI and DMPU behave. The original procedure I mentioned above says that the reaction should be done in HMPA, then quenched with water, and mixture afterwards extracted repetitively with pentane. Does DMPU readily go into acqueous layer when the other layer is pentane? I checked the solvent miscibility chart, but these solvents are quite exotic and did not find any miscibility data..
 

I just performed a scifinder search on thiepane and I found a source that used EtOH as a solvent and they got a yield of 52%. However, I would be apprehensive using an alcohol as a solvent for the reaction, I would be concerned with the Na₂S reacting with the ethanol. Not sure how likely this is.

Yes, I checked all the options before deciding for the one with polar aprotic solvents, and the procedure with ethanol (I think we're considering the same) is insane since it is done on a large scale and requires quite huge dilutions and runs over 5 days with daily additions of fresh Na₂S solution. By the way, once previously when I was preparing solution of Na₂S in methanol for some other reaction, I noticed that a large portion of some white curdy precipitate forms on the bottom of the flask. What kind of process happens here?

I have looked at the spectra and from what you have provided there seems to be a lot of the dibromo compound in there. Isolating it might be hard, but what does your TLC look like? How do the authors say to purify the compound? If they have a solvent system for column chromatography you may need to set up a gradient.

The authors add water in the flask at the end of reaction and afterwards do a series of extractions with pentane. They also use silica (I didn't understand if it was an ordinary column or a short plug) and mixture of chloroform:petrolether=1:9 as mobile phase.

I don't have petrolether in the lab, I suppose n-pentane or hexanes would do?

The TLC done in chloroform:pentane=1:8 gives two spots which smear really a lot (i.e. have long "parabolic" tails) and are separated clearly, but only for about 3-4 mm. There is also one smal residual spot that never moves from the baseline. I guess that moving spots are the sulfide and dibromo, but I really don't have idea which is which and don't know how to distinguish them. They both stain brown in KMnO₄.

Have you considered a typical reactive extraction?  i.e. partition between DCM and mild aqueous acid to protonate the thiepane and force it into the aqueous layer, drain off the organic layer and dibromohexane, and then deprotonate the thiepane to return it to the organic layer.

EDIT: On second thought, you may need quite a bit of acid to force that reaction to take place...

Hmm, the issue in here is that most of HMPA goes into aqueous layer and I'm not sure how complicated it would be to return the sulfide back to pentane layer..

If your compound is low boiling you should definitely try a distillation like Kriggy suggested earlier. 

If that doesn't work, play around with different mobile phases on TLC to see if you can find a system that gives you the desired separation.  The change in the structure is substantial, so I can't imagine they have identical Rf's in all mobile phases.  In general with difficult separations you don't want to go over the 3:1 (length:width) ratio.  This is the ratio we use in our lab for "snowmen" spots on TLC and get pure products with little co-elution.   A longer column does indeed give you better separation of the average/middle of each compound, but longer columns also spread out each compound as it travels down the column.  The effect is that the very first bit of the first compound off the column is pure, but then you get a giant smear of both compounds followed by a pure second compound at the tail end.

That about smearing and spreading in longer column is the issue, I'm really not sure what to do since both compounds seem to be behaving quite similarly.

[MOTOBALL]

I would strongly suggest that you don't waste time trying a chromatographic/extractive/distillation purification of compounds with similar properties.  Since you have a lot of unreacted starting material, simply perform the reaction repetitively, on the reaction mixture, to bring the yield up.  When you get to around 90% desired product purification might be worthwhile, but monitor each repeat reaction by NMR.

[Ingeniosuccinimide]

I would strongly suggest that you don't waste time trying a chromatographic/extractive/distillation purification of compounds with similar properties.  Since you have a lot of unreacted starting material, simply perform the reaction repetitively, on the reaction mixture, to bring the yield up.  When you get to around 90% desired product purification might be worthwhile, but monitor each repeat reaction by NMR.

Maybe it's a good idea. However, when it comes to practical part, does anyone know how does DMPU behave as a solvent? Especially when in contact with water and pentane?

On another thing, when I try to dissolve Na₂Sx 9H₂O in methanol, quite a lot of white curdy precipitate forms on the bottom of the flask. What happens here and is it counterproductive? Can I avoid it somehow?