DIY Voltage cell type/behaviour explanation please

[old_codger]

hello PF community

i'm looking for help in understanding the classification and observed behaviour of some DIY voltaic (?) cells and their associated circuits with which I've been tinkering for a couple of years

the cell operation seems to belong somewhere between galvanic and the original 'contact tension' school of thinking because there appears to be no 'salt-bridge separating two half-cells' type of mechanism involved


i've used flat foil electrodes (approx 3x3"; zinc anode; copper cathode) with thin cotton or paper separator, impregnated only with a little (1cc?) starch-based glue-stick or honey

the open-circuit voltage, per cell, was in the range 0.7V-0.9V with a very high internal impedance (only able to support 1uA load or less)

i've been loading these cells (or batteries of two) using very low current drain circuits, continously flashing a single LED at 0.5Hz or less

the cells (along with their load circuits) have been contained within either mild-steel or aluminium cases

the first such DIY cell tested has been operating on-load since early 2011, the trend of the cell potential has continued to increase over the last 18 months approx.

comparison tests using commercial primary and secondary cells (Alkaline and NiMH AAA type) with similar load circuits have shown the trends of each cell potential to exhibit a slight but steady decrease of cell potential throughout their several-month test periods


firstly, is it possible to identify these DIY cells as being a particular type?

secondly, since there is no 'salt-bridge' type component, what 'mechanism' is enabling these cells to operate without some form of polarization starting to decrease the cell potential?

thanks in advance for any enlightenment!

 

[Bobbywhy]

old_codger, Welcome to Physics Forums!

Thanks for sharing your DIY experiment with voltaic cells. I am fairly certain you can identify your cells as a “Zamboni pile” or a “Duluc dry pile”. These cells do not oxidize or corrode their electrodes like a “salt –bridge” type of voltaic cell. I have included below a few references for you to compare your cells with.

One question: when you wrote: “the first such DIY cell tested has been operating on-load since early 2011, the trend of the cell potential has continued to increase over the last 18 months approx.” Do you mean the open circuit voltage increased over time? What voltage did you measure?

A Zamboni pile is an "electrostatic battery" and is constructed from discs of silver foil, zinc foil, and paper. Alternatively, discs of "silver paper" (paper with a thin layer of zinc on one side) gilded on one side or silver paper smeared with manganese oxide and honey might be used.
http://en.wikipedia.org/wiki/Zamboni_pile

A Zamboni battery or Duluc dry pile
http://tesladownunder.com/interesting_stuff.htm#ZamboniBattery

“De Luc's column.
the idea to eliminate chemical substances that deteriorated the metallic pair was not new; in fact it was tried by Volta himself, by J.W. Ritter, Biot (who used as conductor the potassium nitrate), by G.B. Behrens (with copper, zinc and gold paper), in 1803 by Hachette, Desormes (simple pair of zinc and copper separated by paste) and in 1809 De Luc (with zinc, silver and wetting paper).”
http://www.brera.unimi.it/sisfa/atti/1996/tinazzi.html

http://amasci.com/emotor/duluc.html

Cheers,
Bobbywhy

 

[NascentOxygen]

i've used flat foil electrodes (approx 3x3"; zinc anode; copper cathode) with thin cotton or paper separator, impregnated only with a little (1cc?) starch-based glue-stick or honey

the open-circuit voltage, per cell, was in the range 0.7V-0.9V with a very high internal impedance (only able to support 1uA load or less)

It's interesting to hear of your observations. On the basis of what you've written, I would venture that you have the ordinary galvanic cell. Both glue-sticks and honey are hygroscopic, absorbing moisture to remain sticky, so you do have an electrolyte contacting both. The observed high impedance would follow from the fact that the electrolyte is of scant amount. Can we presume that you found you needed a moist glue because the metals with a dry fabric separator showed no electrical activity?

But the telling point might be: how have the different metal surfaces and fabric changed over time?

Most importantly, congratulations on your investigative bent.

 

[old_codger]

hi Bobbywhy, thanks for your kind welcome - and for the comprehensive list of related info

any similarity of these DIY cells to the early-days cell construction of pioneers like Volta and Zamboni is no surprise; i believe that part of my original motivation to start experimenting with voltage cells, a few years back, arose from accounts of the protracted historical debate between the 'galvanic' and 'dry-pile' schools of thought in explaining the generation of electricity from dissimilar metals - but my motivation was also due in part to recent-day accounts of charge-transfer between conductive surfaces of differing work-function by electron tunneling; hence my use of a separating medium which was more dielectric (eg. sugar-based) than ionic (eg. electrolyte)


the '...tesladownunder...' link which you kindly provided had some interesting examples of voltage cells - however, the gel in the electrode pads contains electrolyte and the writer reported that the aluminium electrode had "almost disappeared" after a year

it's interesting that you say that the Zamboni and Duluc 'piles' do not oxidise their electrodes (in contrast to a 'salt-bridge' type galvanic cell)

Does this mean that they operate by some other method than a Redox reaction? If so, then i guess that 'other method' would be the answer to my first question - please tell me more!


wrt the voltage measurements over time, perhaps i should have clarified that the cell has remained on-load constantly (even during measurement)


my first question (of 'type') was intended, like the second, to be focussed more on the theory of operation, rather than the historical source-name

as an example, considering a voltage cell using zinc and copper as electrodes, each contained separately in a their own metal sulphate solution, with a porous link joining the two containers:

- from a historical PoV this is often named a 'Daniell' cell, after its originator;

- from a theory of operation PoV, this is an example of the basic 'two half-cells with a salt-bridge' type of galvanic cell

so my first question is attempting to identify the underlying 'mechanism', or operational type, of these DIY cells

my second question is attempting to identify what part of this underlying mechanism has prevented decrease of cell potential due to polarisation
(since this prevention is usually provided by some kind of 'salt-bridge, or semi-permeable membrane, between the electrodes)


apologies for any ambiguity in my post

 

[old_codger]

oops - apologies NascentOxygen - somehow i missed seeing your response until i just replied to Bobbywhy

thanks for your kind words, too - just trying to keep my brain active as long as possible!

yes, i agree that any hygroscopic activity in the separator will affect its ionic content - and that would bring us to my second question (ie. if it's a galvanic cell, what is preventing polarisation from decreasing cell potential with time, since it has no 'semi-permeable' membrane between electrodes?)

i've tried 3 variations of the cell construction which would relate to the hygroscopic issue:
- cells with all 4 exposed edges covered with plumbers 'self-sealing rubber' tape;
- cells with one electrode face drilled with pattern of 1/8" holes to expose impregnated separator to air;
- cells with no edge-sealing and no exposure holes.

not a definitive test by any means - but no broad indication of any difference in cell behaviour between these three methods

i've always used some kind of impregnation on the separator (i think this idea may have been prompted by descriptions of the original 'piles'); the glue-stick dries on exposure to air, but the honey remains sticky

i thought i'd taken photos of a dissembled cell, but I've just had a look on my PC and can't find any

i'm reluctant to disturb older cells, but i'll try to ease apart the layers of one corner of a cell which is at least 13 months old and report back; this is a cell with exposure holes in one electrode - there are no signs of corrosion visible around the edges of the holes where the separator is exposed, but i'll take a look inside anyway

thanks for the responses, folks!

 

[old_codger]

hi NascentOxygen

i'm hoping that i will have successfully attached some photos to this post!

i removed a cell from its load and peeled back the corner of the layers to take the photos

apologies for the quality - firstly that the focus is not great at close-up; secondly, it was difficult to get a good single photo of the zinc electrode, due to reflection from its surface

i've taken two shots of the zinc:
- one (more in focus) gives a reasonable idea of the condition of the separator layer and the honey residue on the zinc (but the zinc surface is reflecting the peeled back translucent separator above it which is coloured by the copper electrode behind);
- the other (less in focus) is taken more from above and hopefully gives a better impression of the surface colour of the zinc, although there is still a certain amount of reflection nearer the corner (of the background above the cell)

the cell was constructed early July 2011 and was connected to a load continuously for 7 months; it was then disconnected for 5 months and had been re-connected to the load continuously for approximately a month prior to these photos being taken today

i hope this provides an answer to your "telling point" regarding the condition of the metal surfaces and the fabric

i also hope that someone on the forum will be able to provide an answer to my second question!

thanks for reading

 

[NascentOxygen]

i'm hoping that i will have successfully attached some photos to this post!

The pics are interesting. It looks like you have enough zinc there for a couple of D cells, so it's not surprising they aren't showing much wear and tear. Did you polish the zinc foil to make it shiny when you bought it, or was it already shiny when purchased?

You see no difference from when the metals were new? Is there no pitting of the zinc, no white powder anywhere, not even any dulling of its lustre? With dissimilar metals, it's a case of the zinc dissolving to protect the copper (zinc providing a "sacrificial anode").

You say you've been blinking a LED for all that time. Can you post the circuit diagram of what you are using?

Did you originally get the idea for constructing these from something you read? Otherwise, how did you settle on using honey?

i also hope that someone on the forum will be able to provide an answer to my second question!

This is a good question for the chemists.

 

[old_codger]

thanks for your response, NascentOxygen

i'll try to address your latest questions & comments here, but i note what you say about directing my questions to the chemists instead of here in the General Physics forum>> "It looks like you have enough zinc there for a couple of D cells, so it's not surprising they aren't showing much wear and tear"

i opened up an old D cell this morning:

the zinc residue mass is approximately 1oz. greater than one of my zinc electrodes;
my zinc electrode mass is between 1/4 and 1/2 oz.

so, for reference, there is approximately 2-3 times more zinc material in a single D cell than in one of my cells>> "Did you polish the zinc foil to make it shiny when you bought it, or was it already shiny when purchased?"

i don't believe that i needed to polish the zinc material prior to assembling a cell; I've washed the electrodes with a little detergent and kitchen roll paper to remove any grease (to reduce possibility of subsequent corrosion caused by handling when i cut the foil)

i'm hoping to attach a photo of the zinc foil material, as received, so you can get an idea of its initial state

depending on the prior oxidation level of the copper foil i may have used a plastic scourer ('scotch bright'? type) to clean the surface, prior to washing

i notice from my photos posted for you last night that the copper electrode now looks shinier (and with a more even finish, eg., no indication of 'brushing') than that produced by my initial cleaning>> "You see no difference from when the metals were new? Is there no pitting of the zinc, no white powder anywhere, not even any dulling of its lustre?"

In the photos i posted last night (of a zinc electrode after 7 months continuous use, 8 months total use) the surface appears shiny, with no evidence of any white powder (the honey which I've used has a milky-white colouring, but the 'blobs' of it on the electrode don't appear to contain any 'sediment')

i couldn't detect any pitting; if anything, the surface looked slightly shinier than the material as delivered (but maybe no shinier than after I've washed it?)>> "Can you post the circuit diagram of what you are using?"

yes - an example hopefully attached to this post

the circuit doesn't appear to be at all critical; I've used a couple of different basic types, with a few variations of each

i'm attaching the circuit used with the cell shown in the photos i posted last night - this 'blocking oscillator' type of circuit can operate with supply voltages down to approx. 0.5V, so it's useful as a load for single voltage cell setups

circuit components aren't critical, they just need to be selected to provide sufficient pulse energy to flash an LED whilst minimising the on-going current drain on the cell

C1 0.22uF;
C2 200uF;
D1 1N4148 type;
Q1 BC556 type;
P1 piezo beeper; (optional - provides 'click' when LED flashes)
LED1 Hi-Brite Red;
V1 DIY voltage cell;
'T1', inductance tapped at Q1 emitter with toroidal ferrite transformer primary of approx 0.5mH to form approx. 1:4 step-up voltage ratio to Q1 base secondary winding (total emitter inductance approx. 2mH)

the test system detailed here flashes the LED approx. every 30 secs.

for reference, my oldest test system (operating now for >18 months, using 2 cells in series and an alternate circuit) flashes the LED approx every 2 secs.>> "Did you originally get the idea for constructing these from something you read? Otherwise, how did you settle on using honey?"

yes, i do a lot of background reading on the web (on this and other science and tech-related subjects) - my browser bookmarks are overflowing with interesting sites!

having looked at the links which Bobbywhy provided above, i recognise several which must have influenced me in starting out on this 'investigation' of DIY voltage cells

this would have been several years ago now (and a lot of electrons have flowed under the diode bridge since then!) but i seem to remember being intrigued by the historical debate between the two main opposing schools of thought: 'chemical action' and 'contact tension', as an explanation for the operation of so-called 'dry piles'

i guess that this seemed slightly reminiscent of the 'wave/particle' debate about light - there appeared to be some elements of support for both PoV

i think that i must have wanted to explore the 'long-lasting' characteristic which some of the dry pile structures exhibited - reducing the chemical action would have seemed like a good tactic to assist with such a study

somehow, shortly after reading around this subject, i found myself buying foils of various metals from craft shops and lab. suppliers and raiding the cupboards at home (don't tell my wife!) for possible materials to use in cell separators: starch-based glue, honey, linen and acid-free paper (some of which had been mentioned as components) were all quickly found and adopted for use in the DIY cells

i decided on using an LED flasher as a load so that the system provided its own quick indication that the voltage cell hadn't depleted completely - although i initially recorded the on-load cell terminal voltage each day, after a few months (and with several cell systems to monitor) i started to increase the gap between recordings, so it was useful between readings to see that a cell was still working

the oldest cell system (>18months) has an interesting inverse relationship between the changes in ambient temperature and on-load terminal voltage (graph now attached)anyway i see that, like a true old codger, I'm starting to ramble on, so i'd better call it a day!

thanks for reading

 

[old_codger]

thanks for your discussion, folks, and good to meet you both

i've taken up NascentOxygen's suggestion to direct the second question of my original post to the chemists and I've asked if the thread can be moved appropriately

all the best
o-c

 

[NascentOxygen]

i've taken up NascentOxygen's suggestion to direct the second question of my original post to the chemists and I've asked if the thread can be moved appropriately

No need to move this thread; you could just add a link back to here in any thread you start in the chem group.

 

[Bobbywhy]

Hi old codger,

Thank you for posting all the details of your votaic cell experiments! The "load" schematic and parts list helps me grasp the process. It is truly interesting to me to see your original research.
In your opening post you wrote "the trend of the cell potential has continued to increase over the last 18 months approx." Now, does that refer to the cell that you have plotted in the graph in post number eight? The "voltage of the pile" in that graph is indeed increasing. I'm not familiar with the term "power regression for Vpile", but it appears to be increasing as well.

Given those observations, it appears that power out of your pile is increasing with time. Is that correct?

 

[old_codger]

i appreciate your 'link back' suggestion, NascentOxygen - I've already asked an Admin if they can move the thread, but if they decide they would prefer to leave the thread here then i'll just use a link, as you suggest, thanks


i'm pleased to hear of your interest in my little 'project', Bobbywhy - calling it 'original research' is perhaps a bit grand for what is essentially messing about in the kitchen with a few bits of metal!

yes, the graph attached to post #8 shows data for the on-load terminal voltage and ambient temperature data recorded for my oldest battery using two Zn-Cu cells (aka. a pile); it was constructed in Feb. 2011 and has been operating on-load continuously, since then

the word 'power' in the regression trend curve is supplied by the spreadsheet chart function; it's one of the options, such as linear or exponential regression, and relates to the mathematical method used to calculate the trend of the voltage data - it doesn't relate to the power supplied by the cells

other types of measurements would be needed before we could confirm what is happening with regard to the power being supplied to the load

since the load is active, the oscillator pulse rate is quite likely to have a voltage dependency - whereas, if it was a passive load instead, such as a resistor, then we could directly infer that the output power was increasing with the rise in terminal voltage

it could be true that the output power is now increasing with the increase in terminal voltage - but we can't confirm that just from the terminal voltage data


for a couple of weeks after construction, the on-load terminal voltage initially decreased with time, as expected, although the voltage profile looked more like a capacitor discharge (ie. similar to an exponential decrease) rather than a steady decrease

after approx. 400 hours of continuous operation, the trend of the on-load terminal voltage levelled-out and then started to slowly increase - the graph in post #8 shows the data from around that point (March 2011) up to the last data recorded (as at the end of June 2012), prior to the graph image being produced

the increase in on-load terminal voltage in this type of DIY cells was unexpected (and remains unexplained)

i did anticipate a temperature dependence for the on-load terminal voltage , but i expected the voltage to rise with increase in temperature (as i'd seen previously with some types of commercial secondary cells); this particular battery of cells however has an inverse relationship between changes in temperature and on-load terminal voltage - the two graph traces mirror each other quite closely in many places
(this effect could be due in part, or in full, to the load circuit, of course)


i hope this fill in a few gaps for you

thanks
o-c

 

[old_codger]

no reply so far from Admin to my PM - i guess they're on vacation...