[curiouscat]

As a Chemical Engineer I was called to troubleshoot a hydration reaction (details later) that's taking a long time (~ 3 hrs; batch process) during production. I investigated and it seems we are having a two phase immiscible system. My suspicion at this point is that this is what is causing slow rates because of mass transfer limitations.

This is a hydration reaction of a high boiling point (100+ C) aliphatic alkene using 60-70% H2SO4 in a stirred tank. This forms a sulfate intermediate which later we boil with H2O to give an alcohol.

Is there a co-solvent choice that would transform this into a miscible system? Or should I think of emulsifiers?

Any other suggestions?

[fledarmus]

A third suggestion would be to accept the immiscible system and to add a phase transfer catalyst

[curiouscat]

A third suggestion would be to accept the immiscible system and to add a phase transfer catalyst

Thanks! That's a good lead. I had never considered that option. Just might work if I found the right PTC.

I'm not too good at this aspect of the chemistry so if there are any suggestions for an agent I'd love to hear them.

PS. Is the reaction of a higher alkene with H2SO4 to give an alkyl hydrogen sulfate typically slow or fast? Anyone know? I couldn't get any good leads on the kinetics so I was somewhat assuming that the 3-hour time constant was mass-transfer limited. In hindsight, I'm wondering if the inherent reaction rate constant for this type of reaction can be that slow too (~3 hrs)? Would there be any heuristics or simulations to estimate this rate constant? Perhaps from an analogous reaction?

I wish I could give more direct chemical details, but my bosses are a bit touchy about confidentiality clauses. Please bear with me and apologies!

[BobfromNC]

You can absorb the H2SO4 onto silica gel or diatomateous earth to help increase the effective surface area of the acid.  You can also use an acid ionic exchange resin instead of the H2SO4, like sulfonic acid on resin, or a fluoronated version.   Lastly, using some TFA or Methanesulfonic acid instead will improve the organic solubility.   These are all areas that you could look up or test.

[milo2112]

Wait, secrecy in industrial processes? Say it ain't so! :-)

If it is only a hydration, MSA or PTSA might work. Maybe your stirring is not appropriately fast - I am assuming you have baffles in the tank, right?. The kinetics should depend on the interfacial surface area between the phases. Can you *Ignore me, I am impatient* up the temperature?

[curiouscat]

Thanks for the suggestions. Those are very useful; I'm trying out some now.

@milo: Upping temperature loses us selectivity. Yep baffling etc. all the standard goodies exist. OTOH these only get me macromixing: if it's an interfacial area / solubility issue agitation can only do so much.

Is there a test to know if the reaction happens in either phase or at the interface? How can one test.

PS. Glad you are impatient! I need more suggestions! 

[milo2112]

For testing interfacial chemistry: have three experiments, one not agitated, one at say X rpm and the last at say 4X rpm. Monitor concentration of product formed. The better the mixing, the better the yield for interface reaction. This is a quick and dirty way to probe the question.

[curiouscat]

Following up on Fledarmus' suggestion I tried adding a 0.5 w/w % of a non-ionic surfactant and it's working pretty well; reactions's now complete in ~2 hrs (instead of 3 hrs. before).  Problem is this makes the post-reaction phase separation a headache.
 
Tried RPM increase (milo's tip) too but with my current equipment only 2-settings were available ~350 and 700 rpm. Did get some improvement in rate but not as much as with a surfactant.

In  any case, milo's hunch about it being an inter-facial phase transfer issue seems valid. Question is what other options might I have to enhance mass transfer? Is it worth to beg / borrow / steal (not really!) a higher rpm stirrer to test? Not even sure how high I can go on rpm. Would some sort of liquid sparging system help?

[curiouscat]

You can absorb the H2SO4 onto silica gel or diatomateous earth to help increase the effective surface area of the acid. 

Any idea how much H2SO4 : diatomaceous earth ratio is typically used? Reason I ask is, my impregnation trials show that 1 gm of diatomite seems to adsorb only about 0.6 gm of H2SO4. That'd mean adding so much diatomite that my flask gets very hard to stir. (Am using stoichiometric amounts of acid and alkene)

Am I doing something wrong?

[Wastrel]

Ideally the mixing should churn the interface until the whole flask looks like an emulsion.  If you are getting laminar flow with a distinct interface this is not producing a high surface area. 

Baffles work well when they introduce currents that degenerate into turbulent mixing, with a two phase system they might be slowing down the currents and promoting settling even if they encourage mixing within the phases.  Rapid thin turbines and stators located at the interface might be better.  You might be able to do this with an old fashioned overhead stirrer.

Are the liquids of very different density?  Adding a co-solvent would reduce the rate of reaction for a given surface area, but if you can generate a mixture of similar density, that may result in a more stable emulsion and boost the overall rate.

[curiouscat]

Thank you @Wastrel. Typically I'm using ~700 rpm in an  1 L flask. Hard to know if I'm in the laminar range or not. Visually mixing doesn't look laminar. Would your experience say there's any merit in trying to go higher? (I'm already using an overhead stirrer and not a magnetic one. )

Density of the fluids are ~0.85 and ~1.5 gm/cc so yes they are pretty different.

The co-solvent is a good idea; but it would add a step to the commercial process to again separate it out. If I did add one, would it be in a minor catalytic amount or a almost stoichiometric quantity with my reactants?  Any suggestions for a co-solvent (needs to be inert to 60-70% H2SO4)?

[Wastrel]

My experience is pretty thin on the ground and my degree level education is in pure chem, EE and maths. 

I was wondering if you could throw in some carbon tet to balance the density.  I think this won't work.  My gut feeling is to try something much higher speed with thin blades and like milo says plot rate versus turbine speed.

[curiouscat]

My experience is pretty thin on the ground and my degree level education is in pure chem, EE and maths. 

I was wondering if you could throw in some carbon tet to balance the density.  I think this won't work.  My gut feeling is to try something much higher speed with thin blades and like milo says plot rate versus turbine speed.

Sorry for the delayed response. I somehow missed your post. I will try adding CCl4; if it works would be great.