How can fluid evaporate from caps?

[fonograph]

The aluminum electrolytic capacitors are notorious for failing due to electrolyte fluid evaporation.I read that it happens to supercapacitors too.

I am like    Evaporate!? How on earth can it evaporate!? I mean the fluid is sealed inside,its not like soup boiling on pan.The entire thing is completly sealed off inside metal container and then wrapped in aditional PVC plastic,there is probably even more barriers inside,how is it possible that significant amount of that fluid evaporates?

When they say the cap went dry,that the fluid evaporated,I guess that means it moved somewhere else where it cant do its job anymore,but where did it went? Does it actualy escape the capacitor,or just move to different region inside? When it burst open due to internal pressure then obviously it went outside,but I saw caps where there was nothing outside and the metal can looked like it doesnt have any holes in it,yet it was described as dried out failed cap.

[Gyro]

Electrolytic caps are just very well sealed - but they're not hermetically sealed. You have a rubber 'bung' at the bottom where the leads come out.

If you think how little electrolyte there actually is in a small electrolytic cap (by the time you've accounted for the volume of foil and separator) it doesn't need to loose much. The capacitor doesn't need to visibly 'spill its guts', slow diffusion of water vapour through the seal itself, the leadthrough joints and seal to the can joint is enough over time, especially when hot.

If you check capacitor datasheets, you will normally find that they quote a longer life for larger can parts (higher volume to seal ratio) than the smaller package ones of the same model, even different aspect ratio can make a difference.

It's also the reason for being careful to avoid stress on the leads, overheating the joints, bending capacitors sideways on the PCB and this can cause premature failure. The datasheets normally specify limits.

[wraper]

Electrolytic caps are just very well sealed - but they're not hermetically sealed.

They are hermetically sealed, however seal may start leaking gasses because of poor manufacturing or/and  pressure build up in capacitor. If seal was good enough to hold the pressure, eventually capacitor top will bulge and safety vent will break releasing the gases and thus preventing explosion. Often you can puncture top of used capacitor, especially LOW ESR types and hear how gas leaks. That is a perfect sign that seal was hermetic as it was able to hold pressure inside.

[Gyro]

We have different definitions of hermetic. 

[Jeroen3]

Nothing humans can make is impermeable.

[Mr. Scram]

Nothing humans can make is impermeable.

This. The first rule of dealing with fluids or gasses is that they will get out, or in, depending on what's contrary to your goals.

[fonograph]

Thank you gyro for great post,I especialy like the seal to volume ratio tip

[David Hess]

With one exception that I know of below, aluminum electrolytic capacitors, unlike most wet tantalum electrolytic capacitors, are not hermetically sealed.  There is no bond between the leads, polymer/rubber stopper, and case in an aluminum electrolytic capacitor.  The linked Vishay document shows that the requirement for a hermetic seal includes a continuous ceramic/glass/metal envelope.

The exception are some hermetically sealed aluminum electrolytic capacitors made by Cornell Dubilier for military and aerospace applications.

[Vtile]

How many type of "Hermetics" there is as this is at least one type: MIL-STD-883 Method
1014.12

 

[coppice]

Electrolytic caps are just very well sealed - but they're not hermetically sealed.

They are hermetically sealed, however seal may start leaking gasses because of poor manufacturing or/and  pressure build up in capacitor. If seal was good enough to hold the pressure, eventually capacitor top will bulge and safety vent will break releasing the gases and thus preventing explosion. Often you can puncture top of used capacitor, especially LOW ESR types and hear how gas leaks. That is a perfect sign that seal was hermetic as it was able to hold pressure inside.

Try looking at the specs for aluminium electrolytics. They are not built with any expectation of achieving a hermetic seal. Quite the opposite. The specs tell you how long the cap it expected to last at various temperatures. Those times are based on how long it will take for the cap to dry out, so there is a planned venting rate. Typically this is just 1000 hours (about a month) at the maximum rated temperature. The drying out rate drops rapidly with temperature, so well below the rated temperature the lifetime is usually quite a few years.

Wet tantalum electrolytics are different. Most (probably all) of those are designed to be hermetically sealed, so any venting is a failure of the device.

[b_force]

In the old days some tubes were also made out of aluminium. As far as I can remember they changed most of them to glass for the same issues, they were leaky.
You can also see a big difference in caps, some are much more robust and solid than others and therefore have a bigger lifespan.
Just the fact that there plastic and metal components makes it already very hard to completely seal the whole thing.
It's not so difficult to pull the legs out as well, which also shows that there is no proper seal.
So there are always tiny microscopic little gaps. In combination with high temperature applications and your cap will be gone after some years.
Although most people make it a bigger problem than it really is.

[james_s]

Metal vacuum tubes are fairly common, IIRC the envelope is steel rather than aluminum. They're not leaky, I've seen quite a few of them in 60+ year old radios that still worked fine. They're much more durable than glass tubes and were used in a lot of military hardware. They're a lot less cool looking though, you don't get that vintage look of warm glowing glass, so metal tubes are generally less popular.

[b_force]

Metal vacuum tubes are fairly common, IIRC the envelope is steel rather than aluminum. They're not leaky, I've seen quite a few of them in 60+ year old radios that still worked fine. They're much more durable than glass tubes and were used in a lot of military hardware. They're a lot less cool looking though, you don't get that vintage look of warm glowing glass, so metal tubes are generally less popular.

Ok, well my bad, I thought I read some article about it.
Maybe that was only for one specific manufacturer.