This place is for describing and discussing experiments by members with a focus on the practical side of things. Better methods make better science- and after 50 years I still learn new things about techniques and also the surprises that revisiting old methods reveal about science- be it physics, chemistry, or engineering. To kick it off, here's a little 'methods 101' on distillation of heavy water to remove impurities. This thread might be useful to those who never paid attention to science lessons in school, but now wish they had. Others might think some of it 'too elementary', but we have to start somewhere.
As you can see I have ended up with around three hundred grams of electrolyte and metal salts/hydroxide contaminated heavy water. It looks rather like horse-piss, but at around $1 a gram D2O is too precious to throw away.
The plan is to distill off the water leaving the impurities behind. The equipment is simple - here's a general view. The hotplate boils the D20 in the conical flask, the water-cooled condenser in the middle turns the D2O steam back into liquid, and the condensate is collected in the globe shaped flask in the left. There's a little downward-curving piece of silicon tubing on the swan-neck connector on the left btw - that is for pressure relief, and should be sitting in a beaker to catch any condensate that finds it's way out by that route.
The condenser water jacket is fed from a reservoir of cold water -flow is provided by a little aquarium pump submerged in the tank. These are very cheap - and this one has probably got 100+ hours of service so they are reliable too - which is important. If the cooling stops your precious distillate will just escape as steam.
Whisky and Gin distillers pay great attention to still-head temperatures, which they like to know to fractions of a degree. Since I am not worried about volatiles in my D2O a simpler method of control works. Although the hotplate is thermostatic, it is actually slaved to an external PID thermostat, so the thermostat on the hotplate is set to 'max' all the time, since the PID switches it on and off. Temperature sensing for the PID is via the thermocouple in the still head. The heat feedback loop is longer than I like, but I have used this crude method often enough to know it works well enough for this job. Finally, the conical flask does not sit directly on the hotplate, but on a PTFE mesh spacer. Direct contact causes -too often - cracking of the glass- even borosilicate 'oven glass'.
I'll follow up with the results -when available.