I'm a glaciologist working in Patagonia where everything is always wet. Often water makes its way into instruments, radios and other devices and I feel that always the positive terminals of batteries (and the circuitry connected to it) corrodes more effectively than the negative side. After not finding a explanation for this, I thought it might only be my impression. But then I searched for some batteries and instruments that spent some time underwater and indeed the positive terminals are corroded and the negative ones are like new. The example below show some Li-Ion batteries that display that situation (of course the picture shows two batteries, but believe me that positive/negative terminals in both batteries looks pretty much the same).

I'm considering this question because I'm now designing an instrument that uses MOSFETs as switches for peripherals, and I'm struggling to find P-Channel MOSFETs to switch the positive side. As N-Channel MOSFETs are much more common and cheaper.

So, does really the positive side corrodes more? If so, why? Can that be avoided somehow if I decide to switch the negative side?

NOTE: One line of thought that I've explored is that the battery will produce water electrolysis. Although, my chemistry knowledge is very limited, I think in that case the highly reactive acid H+ (that I think is how hydrogen stays in aqueous solution) should go to the negative side, so I would expect corrosion there. The negative side gets O2, which could be the reason of the oxidation of the positive terminal, but as there is plenty of dissolved O2 in water and even more in air, I would still think that the negative side should get the worse corrosion, so this doesn't makes much sense to me either.

enter image description here

  • 1
    \$\begingroup\$ "The negative side gets O2" Is this a typo? \$\endgroup\$
    – DKNguyen
    Commented Oct 1, 2020 at 4:29
  • \$\begingroup\$ @jsotola I think OP Is referring to how acids dissociate into hydrogen ions. \$\endgroup\$
    – DKNguyen
    Commented Oct 1, 2020 at 4:32
  • \$\begingroup\$ The layman chemist in me says that the electrons are coming out the negative end and so nothing at the negative terminal gets reduced since there's enough electrons to go around. Whereas the positive end is electron deficient causing atoms to share the limited electrons which causes things to get reduced and corrosion componds to form. That makes sense, right? I think that makes sense. \$\endgroup\$
    – DKNguyen
    Commented Oct 1, 2020 at 4:33
  • \$\begingroup\$ electrolysis does not free dissolved oxygen ... it splits water molecules into oxygen and hydrogen \$\endgroup\$
    – jsotola
    Commented Oct 1, 2020 at 4:37
  • 1
    \$\begingroup\$ The "why" question is best answered on chemistry stack exchange, but is reflected in engineering practices such as sacrificial anodes on boats, and the choice of positive ground in some distribution systems. It's not immediate clear how this relates to FET's but you can use an N-FET as a high side switch if you have a gate drive which rides atop the load, that's pretty common at moderate to high power, and even some little USB downstream port power switch chips actually do this internally with an N-FET and driver. \$\endgroup\$ Commented Oct 1, 2020 at 5:10

2 Answers 2


Probably better asked on a chemistry site than electronics. At any rate, electrolysis does happen but it's mainly electrolysis of dissolved salts since pure water requires quite more energy to tear apart.

Now if you do electrolysis on, say, table salt, you get chlorine and sodium which are both aggressive substances. However, your electrodes are metal, and chlorine reacts aggressively with metal while sodium is a metal and react aggressively with water which does not affect the electrode: in fact, the sodium competes (and wins) in who gets corroded first.

Now the chlorine is produced at the positive electrode where \$\mathrm{Cl}^-\$ ions are turned into highly reactive \$\mathrm{Cl}\$ radicals or at least \$\mathrm{Cl}_2\$ gas. So that is where the corrosion happens.

It's not just sodium chloride in solution, of course, but the vast majority of salts are made from positively charged metallic ions and negatively charged substances that will, once you invest the energy to discharge them, act aggressively towards metals in order to regain the missing electrons, ripping out positively charged ions. Which the positive electrode does not mind all that much since it actually was short of electrons anyway.

How corrosion will pattern itself with non-metallic electrodes is a different question. It just depends then on which electrolysis product is more aggressive towards the electrode material.


Normally one would expect the negative electrode to get smaller and the positive to get bigger, as the technical defined direction of current from plus to negative is wrong. In reality its via versa, charge is phyically transported from minus to plus - so ions are transported from the negative towards the positive electrode.

"dry" Effects: https://en.wikipedia.org/wiki/Electromigration

"wet" Effects: https://en.wikipedia.org/wiki/Electrolysis

What you see should be caused by the reaction of the spring of the battery holder and the batteries electrode. Different metals cause a small voltage.


If then, like you say in your application it gets wet, you essentially build a small battery. (so you speed up the process a lot)

I guess you can improve like using a less reactive metal for the battery holder spring like e.g. a gold coated one.

And of cause using a (more) watertight housing would also help :-)

  • \$\begingroup\$ Good points, but this can not be the whole story. In particular the batteries of the picture where in a storage container when them accidentally got underwater (a storm turned our storage area in a pond), so there were not connected to anything. \$\endgroup\$ Commented Oct 1, 2020 at 15:26
  • \$\begingroup\$ hmm, high air moisture and/or condensation water might support creeping currents - saw once a Car ECU blown up by this... \$\endgroup\$
    – schnedan
    Commented Oct 1, 2020 at 16:04
  • \$\begingroup\$ PS: the Car ECU used a Gore-Tex membrane to get moisture out, but not in (which wasn't enough in that one case...) \$\endgroup\$
    – schnedan
    Commented Oct 1, 2020 at 16:05
  • \$\begingroup\$ You've got the migration issue backwards, metallic ions are positive ions so in an aqueous solution they migrate from anode to cathode. Unlike in a wire, the charge carriers in solution are not exclusively electrons, but rather also electron-deficient ions. In electroplating, the object to be plated is the cathode; in electrolytic protection, we speak of the "sacrificial" zinc anode. \$\endgroup\$ Commented Oct 1, 2020 at 16:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.