I An electrolytic capacitor charges by itself?

[Philip Koeck]

We've observed that an off the shelf electrolyte capacitor (330 μF) charges when it isn't connected to a current supply.
Depending on the surroundings we get something between 10 and 100 mV potential across the capacitor.

Does anybody know what's happening?

 

[Lord Jestocost]

To my mind, this effect is known as "dielectric absorption".

https://passive-components.eu/capacitors-capacitance-dipoles-and-dielectric-absorption/

 

[Philip Koeck]

To my mind, this effect is known as "dielectric absorption".

https://passive-components.eu/capacitors-capacitance-dipoles-and-dielectric-absorption/

That makes sense. I charged the capacitor in the lab, then shorted it and then measured the voltage, which rose slowly to about 100 mV.
Then I took it all home and measured the voltage during the weekend, never getting more than about 20 mV.
I thought it was due to a change in the environment (such as electromagnetic background), but I guess it was just because more time had passed since I charged it and I had shorted the capacitor several times in the mean time.

 

[kered rettop]

Most likely electrochemical. Electrolytic capacitors rely on the formation of an oxide layer on one plate. Hence you have one plate that is essentially metal and another that is highly oxidised, so it's not hard to imagine there will be some sort of battery action. Especially given that electrolytics become leaky and have to be re-formed after a period of storage, which implies that the oxide layer is vulnerable to attack by the electrolyte.

Anyway, the chemistry is bound to be a bit more complicated than just forming a simple oxide - for a start, the oxygen would have to come from water, leaving some hydrogen to be mopped up somewhere, or perhaps some other metal than aluminium being plated onto the "metal" plate. One way or anothert is very likely that highly-reduced species will be formed on the "metal" plate and hyper-oxidised ones on the "oxide" plate.