Phase-vanishing reactions are triphasic reactions with a middle organofluorine phase through which reagent is slowly transported from the dense bottom-reagent phase to the phase on top containing substrate. As a result the reaction can be moderate and easier to control and at completion the reagent phase has vanished.
This concept as introduced in 2002 by Ryu et al. (DOI) can be demonstrated in a very simple experiment. Bromine (high density), perfluorohexane and a mixture of cyclohexane in cyclohexane (low density) are placed in a test tube forming three layers. After leaving the tube for two days in the dark the bromine layer has vanished and trans-dibromocyclohexane can be recovered from the cyclohexane layer in a 81% yield. Compare that to the 1932 OrgSynth procedure for the same reaction that requires slow addition of bromine at a low temperature (OS 12:26).
In a follow-up Ryu (DOI) reacted tin tetrachloride thiophene and the acid chloride of acetic acid in a three-phase Friedel-Crafts acylation. When the reagent phase is less dense than the fluorous phase as in thionyl chloride (for the chlorination of alcohols) a U-tube replaces the ordinary test tube (DOI). When additionally water is added as an acid scavenger for certain reactions, the system becomes quadraphasic (DOI)
At least one older concept exploiting U-tubes is known: Cram in 1979 used this setup with chiral crown ether shuttling one amine salt enantiomer from water phase A to B (DOI).
The concept is a nice detour from tried and tested biphasic reactions using a phase transfer catalyst and has been picked up by other chemical investigators. Iskra et al. react chlorine and alkenes through a u-tube setup (DOI). Jana et al. (DOI) mix in 1,2-dibromoethane when the reagent phase alone is less dense than the fluorous phase as an alternative to a u-tube solution But Curran et al (DOI) identifies an extractive mechanism for this particular setup.
Podgorsek et al. give benzylic brominations a try (DOI) and Windmon et al. extend the scope to halolactonization reactions(DOI)
A 4 layer stacked reactor is showcased with from bottom to top a layer of bromine, a perfluorohexane layer, an ionic liquid layer forming a new brominating reagent (tridecylmethylphosphonium tribromide) and a higher alkane layer containing 2-hexene (Ma et al. DOI).
In a most recent contribution Windmon et al. report on a solvent-free tandem phase vanishing reaction (DOI) with cyclopentadiene and dimethylfumarate reacting in a (fast) regular Diels-Alder reaction in the top phase in presence of a iodine monochloride bottom layer providing halogen for a (slow) iodolactonization.
In this particular setup care had to be taken to control the exothermic DA reaction. The quote from the report that the reaction inadvertently went out of control (stirring was set too high and iodine monochloride and Diels–Alder adduct came into a direct contact). The temperature was so high that all of the FC-72 evaporated. is one for the bad luck in science and being honest about it category. But are PV-reactions not supposed to prevent these things from happening? A solution to this problem must be near at hand. Did anyone say W-tube solution?