While preparing glycerol stocks of my transformed bacterial cells (DH5alpha) the protocol I had to follow was that I would make one 50% glycerol stock (500µl culture, 500µl glycerol 50%) and another 87% glycerol stock (850µl bacterial culture, 150µl glycerol 87%). 87 seems to be a very specific number. Any particular reason why 87% glycerol concentration is used?
Typically, a glycerol stock will have a final concentration of 10-20% of glycerol. Glycerol is highly viscous. Pipetting higher concentrations of glycerol is difficult (especially for smaller volumes). Therefore a lower concentration such as a 50% solution (works for me) is convenient: add 300µl to 700µl of bacterial culture.
Regarding 87%: some vendors provide 87% glycerol along with anhydrous (>99%) glycerol. The reason is probably the same: it is easier to pipette 87%.
From a vendor's website:
Glycerol is a component of several buffers, media or reagents in biological research. It is supplied as glycerol anhydrous or 87 %, the latter containing 13 % water. Both glycerols are from the same quality, the 87 % glycerol has the advantage of a lower viscosity. Anhydrous glycerol can be pipetted with difficulty only, especially at low temperatures.
But the question is why 87% and not 90 or 80? See Michal et al. (2017) for a relationship between viscosity and concentration of glycerol-water mixtures.
Another graph plotted from the data given in the document mentioned by iayork in the comments. Different coloured lines denote data at different temperatures (in °C). Original data from: The properties of ordinary water substance. Compiled by N. E. Dorsey, New York (Reinhold), 1940.
So I am guessing that 87% must have been optimal beyond which the viscosity just shoots up rapidly with a small increase in concentration. I wish to reiterate that this is just my guess. I could not find a authoritative (in fact any) reasoning on why precisely 87% glycerol is (or was historically) supplied.