In a recent thread, the choice of flux for electronics work was being discussed, and as I mentioned there, I like to devise my own flux recipes. Since I didn't have any experience with water soluble fluxes, I decided, after some reading, to go to the lab and make some for testing.
I'll describe the process of creating the formulation, but you can just skip it to the final recipe.
Ok, first step, ingredient choice: prior art recommends an organic acid. I wanted excellent water solubility, so I thought about lactic acid, citric acid and glutaric acid. Settled for citric, since it will be easily available for any hobbyist, and it's non toxic (main component of lemon juice).
But, citric acid is a solid and I wanted a viscous liquid flux for easier handling, and that doesn't immediately run away from the soldering iron tip.
So, second ingredient: water soluble high boiling solvent. I had a bottle of PEG-300 (low molecular weight polyethylene glycol) nearby, so immediately decided to use it. I prepared a solution of 20% citric acid and 80% PEG-300.
Later, at home, I tested this flux, and it's pretty decent, but has low activity; clearly 20% citric acid is not enough, about as strong as unactivated rosin, or maybe even less. Moreover, PEG-300 sucks as solvent, because it can't handle the soldering temperature and quickly chars. On the bright side, it can be easily cleaned even with cold water, that's definitely very nice compared to rosin flux, which as everyone knows, is very difficult to clean.
Besides, after some more reading, I found that PEG can penetrate in the circuit board material at high temperatures and degrade its electrical insulation properties. This was all tested in the nineties, so I was making progress and almost arriving at the 21st century in terms of flux technology. Good.
This literature also mentioned glycerin, and I though: "of course, how didn't I think about it!". Glycerin will surely be an excellent solvent, and also much easier to obtain for an hobbyist than PEG.
However, there was a possible problem: glycerin and citric acid could react and form a polymer (like a plastic of sorts). I checked the literature and this indeed happens at high temperatures, forming a foam. Moreover, the polymer could form on storage and the flux could become semisolid over time.
To prevent this, I added water as a final ingredient; this will prevent the reaction during storage, but during soldering, the water will evaporate, so my hope was that the polymer didn't form immediately, so you have time to solder.
Glycerin, 2 parts
Citric Acid, 2 parts
Water, 1 part
Mix the glycerin and water until they form an homogeneous solution, then add the citric acid, heat with an heat gun until the citric acid dissolves, but don't let the water boil out (don't overheat).
I tested this flux and it works quite well, very active (once the water evaporates, that's 50% citric acid in glycerin), resists well to heat, allowing to work the solder with no hurries, can be easily cleaned with warm water or even cold water. The polymer only forms on the soldering iron, were the heat is very strong. It's a white foam that darkens after a while, but it's easy to clean.
In my opinion, it's much better than rosin flux, but I only used it for a couple days, maybe in time I will find my boards full of corrosion, who knows. For now I'm happy with the results.
Citric acid and glycerin can be purchased with no difficulties and are quite cheap.