When I was cleaning a heatsink from an older laptop, the aluminum, and I think also a bit of copper, started to disintegrate very quickly and sometimes violently. I used a cleaning solution that was made for removing thermal paste and cleaning CPUs and PCBs. After some time, the only thing left of the heatsink was some grey powder or ash. I have cleaned many heatsinks and CPUs with it before, and this never happened, and I can’t reproduce it.
you’re seeing elecrochemical corrosion. if you scratch aluminum in such a way that:
- oxide layer is removed, and
- finely divided copper is deposited, and
- oxide layer can’t reform
you’ll see rapid corrosion of aluminum. normally, alumnium doesn’t corrode because of very tight oxide layer. here, oxide layer is removed first by scratching and this also deposits copper in electrical contact with aluminum. then, citrate can bind aluminum removing some of oxide layer, making aluminum more exposed and so more reactive. if you used gallium as heat transfer compound (sp?) then gallium will also disturb oxide layer and make corrosion of aluminum faster. copper elements are in this case actually protected from corrosion (by dissolving aluminum)
It looks like you are using this one: ArctiClean 1
Which seems to be specifically made for this application and supposedly consists of “citrus and soy based solvents”. Aluminium is solved by highly acidic (like sulphuric acid) and highly basic (like sodium hydroxide) solutions. Which it really doesn’t sound like until it somehow broke down in a way it got more potent.
Was it maybe a specific thermal paste with a metal like mercury or gallium inside? Supposedly they can quickly dissolve aluminium. Of course, I don’t know what happened but maybe it was dissolved by the cleaning solution and then rapidly reacted with the Aluminium of the heat sink. If so, that would still make me wonder about the bubbling and foaming when the cleaning solution comes in contact with the left overs.