Tetrahydrofuran and water when mixed together form a single-phase whatever their relative proportions. Diethyl ether, which has two more hydrogens per molecule, forms two distinct phases when mixed with water. 2-Methyltetrahydrofuran forms two layers as well. Methyl ethyl ketone with the same molecular formula forms two layers.
Among four carbon alcohols, 1-butanol is only soluble between 6 and 9% by weight at 25℃. 2-Butanol is only soluble about 18% by weight at 25℃. 2-Methylpropanol is only soluble about 7-8% by weight at 25℃.
Therefore, each of these can be called immiscible with water; however, t-butanol with the same molecular formula is completely miscible with water.
Clearly at four carbons and one oxygen in a molecular formula we are getting close to some discontinuity in mixtures with water.
This is more than just curious. It is important because reactions conducted in these solvents are often quenched and worked up by adding water and it makes a difference whether they form one or two separate fluid layers.
The situation with regard to THF is particularly important because organolithium and Grignard reagents are so often necessarily or most often prepared in this solvent.
It is of considerable consequence that THF does not provide any azeotrope that can be used to dry THF.
A mixture of 18 grams of water and 64 grams of THF would contain 1 mole of each molecule. Here is my question: What is the smallest weight of any other common solvent that, added to this mixture, would give a clean interface between two distinct layers, and what is that third solvent? I do not have the answer. But the answer has a practical importance because such an addition would provide one simple element of a work-up for a reaction conducted in THF and quenched with water.
For simplicity and to inspire imaginative thinking I have chosen the moles of water and THF to be 1:1. My guess would be 2-methyl propanol. It is apparently poorly solvated by water alone but it would provide a hydrogen bond to donate to the electron pairs of THF. Another promising candidate would be t-butanol. If t-butanol caused separation into two discrete phases it would be truly remarkable since all three solvents are miscible as binary pairs! However, the hydrogen-bonded complex between a t-butanol molecule and a THF molecule might mutually satisfy their polarity needs and present a hydrophobic exterior to the water.
If I had a third guess, I would add a carbon and try 1-methyl-2-butanol (t-amyl alcohol). That would preserve the same hydrogen bonding but increase the overall hydrophobicity of any binary complex. t-Amyl alcohol has only an 11% solubility in water at 25℃.
I wish someone would do this last experiment:18 grams of water, 64 grams of THF, then add slowly t-amyl alcohol with stirring until 78 grams of the alcohol were added. Do the layers separate?