[curiouscat]

In conducting a halohydrin formation reaction I'm bottlenecked by a slow hydrolysis step (B  A):



Any tips as to how I might be able to speed up the acidic hydrolysis (B  A)? End goal being a commercial scale process.

Any catalysts that might work? Or a change of reaction conditions? Would using a non-Cl containing acid help (since HCl should be driving reaction backwards)? A temperature ramp?

These are results from a series of batch experiments using HCl + H2O2 (both 30%; as dosed) as my reagent. Currently reaction is being carried isothermally at ~90 C with almost stoichiometric amounts of reagents and an excess of H2O.

After 1 hour all R seems used up and B  A seems the only step going on. 5 hours of the hydrolysis boosts yield of A by about 15% so not a gain to ignore.

[zoork34]

Why dont you react your styrene with Cl2 in the presence of water?  Should get your right to your desired product. 

[curiouscat]

Why dont you react your styrene with Cl2 in the presence of water?  Should get your right to your desired product.

Yes, that's an option we've been considering. Thanks. Currently we lack the equipment to handle gaseous chlorine but I've some plans to get that rig going.

Besides, I'll probably get similar unwanted byproducts with that route too. This current route isn't too bad for yields.

Another point is the relative cost of Cl₂ versus H₂O₂ + HCl. I don't think Cl₂ is a definitely cheaper option. I could be wrong.

[discodermolide]

Using free Cl₂ will be a real safety problem. Stick with what you have, but you could throw in some LiCl, may help.
Knochel has had grey effects using LiCl in organometallic reactions. I know this is not organometallic but it can't harm anything to try.

[curiouscat]

Using free Cl₂ will be a real safety problem.

Agreed. Not insurmountable, but quite a bit of capital investment even if that route were the winner.

Stick with what you have, but you could throw in some LiCl, may help.
Knochel has had grey effects using LiCl in organometallic reactions. I know this is not organometallic but it can't harm anything to try.

Interesting tip. Thanks! Never knew.

I'll order some and try that.

Meanwhile, if any other similar ideas come to mind, please shout out! I'm stumped. Willing to give ideas a shot, even if improbable.

[opsomath]

Surely you can minimize formation of the 1,2-dichloro byproduct by keeping Cl- concentration very low?

[curiouscat]

Surely you can minimize formation of the 1,2-dichloro byproduct by keeping Cl- concentration very low?

I can try. But that means keeping my H2O2 conc. very low too. Right? Currently, H2O2:HCl is close to stoichiometric.

Currently I see ~20% dichloro product at the peak (~1hr) and this goes down to ~5% after 5 hours. Most of this reduction appears as corresponding gain in the desired product. Give or take. We are still plodding towards experimental precision here...

[opsomath]

Is there some reason they're stoichiometric? What you want is for HCl to react with peroxide, the product of that reaction to form a chloronium ion, and for that chloronium ion to react with water. (If that's actually the mechanism, which was a controversy on here earlier.) Seems like if that's how it works, what you want is to keep your HCl concentration as low as possible and still get a decent rate, because the more HCl, the more byproducts.

OTOH, if the mechanism is actually formation of HOCl and direct addition of that to the double bond, it's another story.

[curiouscat]

Is there some reason they're stoichiometric? What you want is for HCl to react with peroxide, the product of that reaction to form a chloronium ion, and for that chloronium ion to react with water. (If that's actually the mechanism, which was a controversy on here earlier.) Seems like if that's how it works, what you want is to keep your HCl concentration as low as possible and still get a decent rate, because the more HCl, the more byproducts.

That's a good idea. Thanks! No harm in trying. I will. Other than at some point the cost of excess H2O2. Unless I can recover H2O2 and recycle.

No other good reason as to why I use them stoichiometrically.

Yes, the mechanism remains a mystery. All I can say is the reaction does work, what I'm working on is the optimization.

[opsomath]

Well, I will be very interested to hear whether the low-HCl version does better or worse. At least then we will have some information about the mechanism. I am still carrying the "classic halohydrin formation from in situ Cl2" flag, Archer.

[curiouscat]

Well, I will be very interested to hear whether the low-HCl version does better or worse. At least then we will have some information about the mechanism. I am still carrying the "classic halohydrin formation from in situ Cl2" flag, Archer.

Ha! Just sit tight a few days. 

[DrCMS]

Would a phase transfer catalyst or a surfactant help as I assume the water solubility of B is not that great.

[curiouscat]

Would a phase transfer catalyst or a surfactant help as I assume the water solubility of B is not that great.

Probably would help (also help my overall rates since A has horrid solubility in water too).

Have any chemicals that you'd reccomend trying? The 90°C / oxidising environment / low pH make the choice a tad confusing. 

I have tried one surfactant (high MW polyacrylamide) but it didn't help. No PTC's yet but was planning on trying CTAB / CTAC & Sodium dodecyl sulfate.

The bewildering complexity & choice in surfactants has stupefied me into inaction a bit. I havn't figured out a good way to choose; a lot of them list their HLB / CMC values but I'm darned if I know what value I need.

[opsomath]

So, I did a little bit of Googling for catalysts that could improve the hydrolysis of your dichloro compound, using benzyl chloride as the model compound.

The good thing: I found one.

The bad thing: It's mercuric chloride. http://pubs.acs.org/doi/pdf/10.1021/ja01492a014

Soo, yeah, that's probably not going to happen. Perhaps a catalytic amount of one of the following:

Zn(II)
Ni(II)
Cu(II)
Fe(III)

in approximate order of Lewis acidity would help your hydrolysis.

[curiouscat]

@opsomath:

Thanks for digging that up. I'd missed that.

I'll start with ZnCl2 I think. Let's see how that goes.