[NewCandle] Anodizable metals and essential conditions for glow discharge?

[Horace Heffner]

On Sep 14, 2008, at 10:41 AM, Keith Nagel wrote:

> Hey Nick + Horace,
>
> You got me to wondering about other metals. A quick survey of
> capacitor manufacturers yielded this list of candidates.
>
> aluminum, niobium, tantalum, titanium, tungsten, zirconium.
>
> I know we've tried the first and last, but how about the others?
> Anyone? Is glow unique to aluminum? Of course, disruptive discharge
> due to dielectric failure would be seen on all of them, as
> Nick described with zirconium. If it is unique to aluminum,
> perhaps the reason is that you need the porous
> structure to see glow. That would be a good question to answer
> with the scope. If you can get glow from Zr, is the oxide structure
> porous?
>
> K.
>


Some random thoughts.

I've written in my experiment notes that I haven't seen the glow  
occur at all until passivation occurs, i.e. until a clearly  
rectifying (i.e capacitive when on both anode and cathode) component  
shows up in the i vs V trace, an "eye".  I haven't seen the blue- 
green glow occur at all for metals where a significant capacitive  
layer does not form, like Cu, Pb, or Pt.  If memory serves, Ti works  
well to produce a glow.  I've also noted on various occasions that  
upon examination of (thin) electrodes from the side it *appears* the  
glow is in, or extends into, the electrolyte, and is not just on the  
surface, but even at the time I said that clearly could be an  
illusion.  If I had to make a bet, though I would still say the glow  
from some metals running at high voltage is in the electrolyte, and  
due to ionization of the electrolyte due to the high field strength  
present at a passified (conditioned, or rectifying)  surface.  I also  
think the glow is triggered by electrons tunneling through the  
conditioned surface, but once a tunneling occurs at the surface it  
can trigger a cascade back into the highly polarized electrolyte near  
the surface.  The higher the barrier strength, the further from the  
surface the glow extends.  The higher the current density at a given  
voltage, the more intense the glow.  If this is true then the  
geometry of structure on the anode surface should not have as much to  
do with the glow colors as much as the chemistry of the surface and  
the electrolyte, as well as the amount of active surface area  
visible.  Using differing electrolyte chemicals to get colors is  
similar to adding dyes.  In fact, if memory serves, some kinds of  
fluorescent dyes can be used to increase the brightness and change  
the color of the glow.  My experience with low or no electrolyte is  
the color is typically a blue-green glow, and I think this comes from  
the water itself.  Your use of a SEM may at last serve to prove or  
disprove some of this.

Though in my experience high enthalpy output occurred with electrodes  
operated at high voltage (typically 400 to 1600 VAC) it is possible  
to obtain a good glow using only 120 V and borax or baking soda.  See:

http://home.earthlink.net/%7Elenyr/borax.htm

Something notable on the SEM photos were little light patches in or  
slightly under the coating.  I have to wonder if they were "healed"  
electrospark locations.  I have often thought it would be interesting  
to know if the glow comes more from some spots than others, and  
whether it is in the electrolyte near the surface or actually on or  
even within the electrode surface.  I was going to try moving a fiber  
optic probe over an electrode at differing elevations and angles to  
try to find out, but never got around to it.

As an aside, something interesting to examine might be zinc.  Below  
are some old posts of mine here in that regard.  Zinc passivates at  
a  very low voltage.  Zinc passivation is an old and highly studied  
subject I think.  The bright flashes I observed with zinc and vinegar  
are not well known as far as I know, but I think maybe it is just due  
to vaporization of tendrils of a zinc compound released into the   
electrolyte. It is also notable that zinc is a hole conductor, so it  
may glow at much lower voltages than electron conductors due to  
electron-hole annihilation at the conductor surface, and this might  
be an entirely different mechanism than the glow from aluminum,  
titanium, or zirconium, which seems to me to depend on a high  
dielectric strength but thin electron tunneling barrier.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

> BTW,  as an aside, something cool (I thought was cool anyway) was
> pushing the voltage on Zinc.  I used some 2" wide Lilly Miller MOSS
> OUT! Roof Strip from Ace Hardware.  I didn't get the stuff to
> condition much above 70 V RMS AC in pure distilled white vinegar.
> However, when pushed into the electrospark mode hard there were
> periodically some much bigger sparks than usual - more like small
> flashes. About 1 a second.  I just *had* to think about some of the
> "events" that have showed up on cathodes and wonder if some kind of
> nuclear thing was happening with carbon or something fro the acetic
> acid.  At any rate it was cool.  And it was fairly silent. I fiured
> the flashes were big enough to cause electrolyte to fly, but no sign
> of that. Out in the shed I have some big 220 V variacs, so maybe one
> of these days I'll try to really fire up that baby and put a geiger
> counter on it.  I doubt anything will show up, but the sight just
> forces you to wonder...

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

                  Diode
                  Cathode
       -------|>|-----------(+)   (Cell anode) (no bubbles, but  
membrane)
                   stripe             (no normal oxygen evolution)


       stripe
       -------|<|-----------(-)   (Cell cathode) (bubbles)
                  Diode                (normal hydrogen evolution)
                  Anode

             Fig. 1 - Cell circuit diagram



When conditioning a zinc electrode cell, as shown in Fig. 1, with DC,
there were no bubbles coming off the cell anode, but lots coming off
the cell cathode.  The electrolyte was 350 ml with less than 0.5 g of
baking soda.  The electrodes were about 5 in long and 1/4 in wide
sheet metal (MOSS OUT! Roof Strip from Ace Hardware).

The diode cathode (electron acceptor) accepts electrons from the cell
anode. The diode anode (electron donor) gives electrons to the
cell.   The ring painted around the diode is on the diode cathode
end.  The diodes in Fig. 1 are merely diagrammatical.  There was
actually a bridge used.

Initially the cell was surge conditioned for one minute with about
127.6 VDC and 0.15 A.  There was no glow.  Lots of cathode (hydrogen)
bubbles.  No anode (oxygen) bubbles.  Dropped voltage to 25.3 V at
0.02 A for 10 minutes.  Over the next 10 minutes the cell voltage
dropped to 20.83 yet current remained at 0.02 A (I should have used
better current scale, but this is still weird because the variac was
not changed.  The drop in voltage across the cell indicates a reduced
cell resistance, yet the current remained roughly the same.  Possibly
due partially to the use of a long extension cord between the cell
and variac.)  The cell apparently was not actually in a conditioning
mode, unlike the prior case when a vinegar electrolyte was used.

At about 11 minutes I noticed a wispy cobweb like or silk like
rectangular membrane extending about 1 cm off the side of the anode
toward the cathode and from it extended a what looked like a cloudy
trail of smoke-like material flowing toward the cathode which was
about 4 cm away from the cathode.  I turned off the current out of
concern the structure might conduct and short out the cell. Maybe
that's what produced the earlier flashes when running with AC and
vinegar?

When I removed the electrode assembly some of the wispy stuff looked
like it disintegrated, turned to smoke, while some remaining small
membranes of the wispy stuff floated around the cell.  I've seen
stuff like that, colloidal things I think, in cells using Ni salt
electrolytes, but not Zn.  The colloidal like stuff must be taking
all the oxygen, possibly in the form of an oxide or hydroxide.

When using AC the membrane stuff must not be so clearly visible.  It
must conduct somewhat to flash like that - assuming it does cause the
flashes.

Horace Heffner
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -


Best regards,

Horace Heffner
http://www.mtaonline.net/~hheffner/