[Prev][Next][Index][Thread]

Alternate electrolytes

 From: Tim Chandler <tchand-at-slip-dot-net> 
> Also has anyone ever tried using a better conducting electrolyte in
salt water 
> capacitors, does it improve performance any?  Here are some equivalent 
> conductances of some electrolytes: 
> 
> NaCl              120.65 
> CaCl2             124.25 
> K3Fe(CN)6         150.7 
> K4Fe(CN)6         146.09 
> Ca(OH)2           232.9                 
> NaOH              240.8 
> HCl               415.80 
> 
> ....above in 0.005 gram equivalents per 1000 cubic centimeters 
> 
> Hydrochloric acid appears a good choice, that is if you could keep it from 
> destroying the electrode and wall of the vessel..:) 
> 
> Note that sodium chloride (table salt) is pretty low, I am sure their is a 
> better choice for the liquid-conductor plates in "salt-water" capacitors. 
> Question is does the conductivity of the liquid-plate electrolyte really 
> make that big of difference in overall capacitor performance, or maybe I 
> am just pissing in the wind...          


Tim; 
  I've been mostly a lurker to this list, having not posted  
anything for at least a year and a half, but I'd certainly like  
to add a comment to your very interesting suggestion. 
  Beyond conductivity, it would also be useful to look at the  
solubilities of the various materials. After all, the 
conductivity of the electrolyte will depend not only on  
the types of numbers you've given, but how much of the material 
you can actually dissolve in the water. I'm gonna go on memory 
here, because my CRC is at home, but NaCl is also a lousy  
choice because it has such a low solubility. I suspect it's  
been used so widely because it's so readily available. 
  While the Ca(OH)2 looks almost twice as good, it's so slightly 
soluble in water that it would make an incredibly poor  
electrolyte at the start, and then it would degrade by reacting 
with atmospheric CO2 to produce insoluble CaCO3. The NaOH is 
caustic enough that you wouldn't want even tiny spatters of the 
stuff, and if you do spatter it, it will absorb atmospheric  
moisture to the point that it will be constantly wet; something 
you don't need around high voltage. It also absorbs atmospheric 
CO2, though it produces a soluble product. A bigger problem 
with the HCl is that it's a gas dissolved in water, and will  
slowly evaporate out of the solution, corroding various items in 
your lab as it goes. Some problems might be solved by sealing  
the capacitors, but I'd personally feel uncomfortable with this, 
particularly when better alternatives are available. 
  Rather than starting with conductivity numbers, which seem to  
vary by a factor of 2 or 3, why not start with solubility numbers,  
which can vary by an order of magnitude? Sodium nitrate is far 
more soluble than table salt, and is a cheap fertilizer. (You'd 
have to go to a dealer that supplies farmers for this; the  
garden shop at the mall wouldn't carry it.) A solution like this 
would make materials like paper much more flammable, but it's  
not all that hazardous, not toxic (at least if you consider 
hot dogs non toxic) and is easy to dispose of if you find it 
unacceptable. (It _is_ sold as fertilizer!) Since it is an  
oxidizer and makes things more flammable, it might be best  
suited for feasability testing of the concept, before spending 
bucks for more expensive materials... 
  If you instead wish to go for a more inert 
(and more expensive) alternative, I beleive that the solubility 
of Potassium Iodide is absolutely fantastic. Again, the CRC  
handbook will provide lots of solubility numbers. Once you've  
found a really soluble material, then look for conductivity 
numbers. (or, get a small sample and measure it directly; I 
can provide some mail order chemical dealers who'll sell you 
quantities as low as an ounce.)  
  I personally think your idea is a good one. To me, the only  
obvious losses in the liquid capacitors are polarization  
losses in the dielectric (what folks here use the dissipation  
factor to describe), resistive losses in the electrolyte, and 
resistive losses due to bussing lots of these things together. 
I've no doubt that someone will add any others I may have missed, 
but if you can minimize the bussing problem, and if I haven't 
missed any other significant losses, you might just improve  
these things significantly. Best of luck if you decide to 
pursue it further. 

Wes B.