Hodgkinson wrote:
The problem is that a density test on the distilled isopropyl alcohol showed 20.81g for 25cm3 of product, which, assuming that water is the impurity, would indicate about 20% water present :-?
How did you establish the density, and what are the uncertainties in the reported values? Wikipedia gives a density of ~0.79 g/cm^3 for IPA, and the values you give indicate a density of ~0.83 g/cm^3. If you are just 1 ml off (as in 26 ml), the density already drops to 0.80 g/cm^3!
Second, if these values are indeed highly accurate, and precise to four decimals each (i.e., 25.00 ml---very unusual if you manage that!) then you really want to be using partial molar volumes. In other words: if you add a volume of water to a volume of IPA, I'll guarantee you that the resulting volume is not the sum of the two bodies of liquid. It can be more, or it can be less. In other words yet again: the density of the mixture does not change linearly with water content; linearity is only an approximation which in case of very sensitive measurements fails to hold up.
Third, if you can find some, I'd recommend zeolite granulate as a drying agent rather than silica gel. The later is mechanically very weak; zeolite is stronger. The stuff looks like off-white sprinkles, and in order to activate them you just roast them in an oven at a high temperature.
