Introduction: Junk Battery - 2.65v Rechargable Aluminum-Ion

SAFETY FIRST: This project is not for the timid. Wear goggles and gloves, seriously. see my previous
instructable for further details


My goal, over the last four years, was to create a "good enough" battery for the billions who live in "light poverty" - 1.2-1.7 billion people who have no reliable source of light once the sun sets. Billions. So I decided that I can't help create a better LED, or necessarily in distribution, but I *COULD* help people build their own batteries for storage.

The first requirement was that the battery be cheap and easy to manufacture. It had to be an improvement over the toxicity of lead-acid, but could be as big and bulky as I might care to make it.

You can see some of my other instructables as steps toward this development.

What I've created is a wonderful 2.65v Aluminum-Ion battery due to an accidental discovery while exploring geopolymer-based electrodes for another cell. The aluminum silicates protect the aluminum from breaking down in the electrolyte, and allow the battery to work beautifully. I'm not 100% sure of the "final chemistry" of the cell, so I can't build one at 2.65 at the start, but the current cell (starting at 1.3-1.43 volts) can be conditioned to become the 2.65v cell over about 15 cycles. My "master" cell is still conditioning, so I'm not sure what the life cycle is at this point... but this is the best cell I've built in four years of working on it, and I want to make sure it's out there and ready.


0. Abundance - The battery should take advantage of abundant natural
resources, sticking within the top 10 or top 12 elements. (O, Si, Al, Fe, Ca, Na, K, Mg, Ti, H and possibly P, Mn)

1. A simple-to-make battery that could be constructed in primitive conditions

2. It should be easily constructed from recyclable/recycled materials "Junk"

3. It should provide enough power to light LED lighting systems through the night

4. It should be rechargeable

5. It should be easily deconstructed, "serviced", or recycled once it fails.

6. Open Source- the "recipe" will be freely available for anyone to use.

Acceptable Trade-Offs:

0. Carbon, because it's hard to make a cell without it, and is generally accessible. Sulfur as well, again, because it's generally accessible by consumers (battery acid is a common thing).

1. Size - I don't care if it takes up the size of a chair, up to 5 times equivalent size lead acid

2. Voltage - I'm not worried about a low voltage cell, if I can just stack several together

3. Graphite - While not ideal for primitive situations, it can be easily recycled from dead primary batteries

4. Time to condition - it might take a while to condition the battery to a useful state through repeated charge/discharge cycles

5. "Outgassing" - while a serious negative, it is understood that some outgassing may occur during conditioning. Gasses like Chlorine might require the battery to be stored outside.

6. Boost converter - Our goal is to create light, not worry about continuous voltage - so a pulsed boost converter is acceptable (a "Joule Thief") to get a lower voltage up to a pulsed signal capable of lighting an LED. I've driven "30 watt" LEDs on as little as 2 watts using this method.

Step 1: Bare Bones Cell and Theory

My original cell is actually the "Variant: Scrap Aluminum Tubing and Graphite stick" later in this instructable. This instructable is my first attempt at a flat cell. Below is the ingredient list, construction, and what I believe is the theory behind the cell.

The very basic ingredients:

Waterglass - Sodium Silicate - doesn't take a lot.



Salt (NaCl)


Epsom salt

Basic build process: (the aluminum tube variant is recommended)

1. Paint the aluminum surface "inside" with waterglass (sodium silicate). It will react and form a white gel, and a dark gel. It will produce hydrogen and heat up during this reaction.. so be careful.

2. You can place a paper seperator on top of this, as well as wave paper, and let it sit for 24-48 hours

3. Mix a slurry of urea, salt, borax, and epsom in equal parts. You may also add a small amount of waterglass to loosen it.

4. Spread the mixture on the paper seperator

5. insert/overlay with your graphite electrode. You may also try copper, but I haven't tried that yet. You will likely have a slighly lower voltage ~ 2.25v

6. The cell should have an initial voltage around 1.3-1.4v - I'm not 100% sure why... but that gives you a clue it's working.

7. charge >3.5v, discharge through a white LED. higher voltages (4v - 10v) are possible, but will create chlorine gas, which is bad news. This doesn't seem to ultimately affect the battery as far as I can tell. It may bubble over with what appears to be water with dissolved AlCl. Not good for you to touch or ingest.

8. Occasionally let it sit overnight without a charge.

9. repeat 6/7 - and the battery will slowly improve.

10. The cell can be smelted down into aluminum when you wish to recycle it.. you may wish to add lye to deactivate any salts in the electrolyte.

11. The first 10 or so charges can be a little disappointing. Keep it up, you'll see it gradually improve.

My working theory:

1. Waterglass (Sodium Silicate) reacts with the Aluminum to form (Aluminum Silicate - a relative of clay). This is the "white gel"

2. The "free" sodium will react with water to form lye (NaOH).

3. The lye reacts with the aluminum to form Aluminum hydroxide (and hydrogen)

4. The borax (Sodium tetraborate) and Salt (NaCL) already have sodium ions, so don't react

5. The magnesium sulfate I added since magnesium is far more reactive than aluminum, and I believe might help maintain the electrolyte. It may not be necessary

6. Overcharging the cell will change NaCl to NaOH and Chlorine.

7. The Chlorine reacts with AlOH and Al to form AlCl

8. The urea then helps for a Deep Eutectic Solvent with the AlCl, acting as our electrolyte.

9. Aluminum ions get intercalated in the graphite, but will rejoin with the Aluminum silicate on discharge

10. I believe the Aluminum silicate and Sodium silicate are the "merry go round" that causes the process to work, and that we, as we get to less and less water in the mix, find a perfect balance that allows the battery to work well.

11. I'm not sure the borax and epsom are needed, but in my experience they make batteries work better.

Step 2: Materials

Waterglass - Sodium Silicate - can be made with silica gel and NaOH.

Aluminum - the thicker the better so we get into a good balance. Foil may require "additions" until the reactions stop.

Aluminum Silicate - the reaction product of waterglass an aluminum... speeds the whole process along, paint your aluminum with already reacted silicate to cut down on conditioning time. Such a pretty blue-grey

Aluminum Chloride (AlCl) - the reaction product of the reacted NaCl and the aluminum electrode. another way to speed things along.

Urea - Ice melt is a great source. I'm sure you can think of others. Helps form an ionic liquid electrolyte with the AlCl

Salt (NaCl) - Table salt. I haven't tried iodized.. but I would guess the over potentials might just drive off the iodine.

Borax - Used to wash clothes. I believe it might be acting as a flux in the electrolyte.

Epsom salt - Not 100% sure, but both my batteries use it, and it *seems* to help.

Step 3: Paint Aluminum Surface With Sodium Silicate (Waterglass)

A thin layer is all that's needed

If you're using aluminum foil for your electrode, you'll need to pre-react some aluminum and waterglass before spreading - or the resulting reaction wil likely just eat through your foil. It may take 5 sheets of foil to completely react with the waterglass.

For this a thick piece of aluminum (or thick aluminum tubing) are definitely the way to go - they will have enough mass to allow the corrosion, but not suffer any problems.

Step 4: Seperator (Paper) and Pressing

For a flat battery, I lay a piece of paper on top to soak some of the electrolyte and acts as a seperator later. In the tube, this isn't necessary - as the electrolyte takes up the extra space.

Additionally, for contact, I added some wax paper and pressed the battery for 48 hours until it finished reacting.

Step 5: Additional Anode Material

After 48 hours, it was pretty stiff, with aluminum silicates forming, even around the edges.

You can see how it ate through the aluminum foil layers... so I added two more layers as additional anode material.

Step 6: Tape Up Edges and Paint With AlCl/Urea/Borax/Salt/Epsom Electrolyte

I taped the edges to help with current collection, and protect my output. Then painted with my electrolyte. In the last picture, you can see the edge of the packing tape resisted the liquid.

Step 7: Graphite Sheet/rod

Definitely not easy to get in the US. I had to order mine twice from china. You may try copper sheeting, or attempt to make your own with graphite by mixing polyurethane and graphite and activate carbon and painting a copper sheet. For tubes, use graphite rods. which can be harvested from old alkaline batteries or purchased directly. I got 10 12-inch graphite/carbon rods for ~$10 online.

Step 8: Repeat Painting and Pressing for Additional Anode Layer

The advantage of a flat cell is that you can make additional layers to surround the graphite (cathode +) layer for more capacity. Here you could continue to paint, press and build up the layers in your battery for more capacity.

Step 9: Variant: Scrap Aluminum Tubing and Graphite Stick

Same process, just a thick aluminum tube and graphite stick. So far, my best cell. It can bubble over when overcharged, so leave a few inches of "headspace" in there when you build it.

Step 10: Variant: Copper and Glued Graphite/activated Carbon (graphene?)

I haven't tried making a copper electrode, but would guess you'd only get 2.25v out of the cell, which still isn't bad. Plus the copper silicate is a very pretty blue color.

Graphene oxide, graphite, and activated carbon are all possiible electrode materials when mixed with a little polyurethane and glued onto a copper sheet.

I've also been experimenting with a geopolymer-based electrode, but will probably post about that at a later date.

Step 11: Conditioning and Voltage

The battery will exhibit 1.3-1.4 volts when initially constructed.

Conditioning requires repeated charge/discharge cycles. Run it up, run it down. It will start showing some capacity after 10-15 charge/discharge cycles.

I'm not 100% on the charge voltage, but I believe it is somewhere between 3 and 3.5 v - my guess is 3.25 or so. I'm still working on the reactions.. and from experience putting it right at 2.75 or 2.8v doesn't seem to get us much above the 2.5v "steady state"

I recommend a joule thief, or 100Ohm resistor for discharging. I prefer a joule thief (boost converter), because it runs a white LED, and is at least providing some "work" that I can watch.

It will tend to discharge to 2.5v overnight, which is something I'm still working on, but for a primitive setting, especially with a joule thief, we'll just abuse it as much as we want - the goal is light, and deep discharges don't seem to be a huge problem. Perhaps 2.5v is actually where the battery wants to live in daily use... only time will tell as we continue to experiment with it. It has a STRONG arc around 1.9-1.7v, but I don't know if we have to have the full 2.8v+ charge voltage to get that.

The cell doesn't handle high current well, which is why 100Ohms is a bit of a high drain for it. 1kOhm (about 2 mA) is a perfect drain for it. By using two cells in parallel, you could easily light a high brightness LED all night. Internal resistance is likely a problem with the electrolyte, or perhaps I should mix aluminum powder and graphite with the anode (-), or Manganese Dioxide in the Cathode(+) - but that puts me outside the parameters of the "junk battery"

Step 12: Additional Thoughts/Chemistries

Some preliminary experiments with copper cathodes look promising. While the voltage may ultimately be lower, it seems like a very strong battery can result from the formation of the copper silicates. I don't know if the liquid electrolyte is actually necessary, but I used it ( Paint the aluminum, allow to cure/dry, drip fresh aluminum silicate on the dried cake, and then lay a copper sheet on top, charge). So perhaps a simple copper/silicate/aluminum "solid state" cell is possible. Seems to exhibit a wonderful 1.7-1.5v chemistry - with no seperator, just the Al Silicate and Cu silicate, it's capable of a rapid charge at higher voltages, and a good 15mA discharge for nearly an hour. This might actually be the next stage in developing this battery... lower voltage, but intense durability.


BronsonMMiller made it!(author)2017-01-09

This is very interesting! I've been dabbling in diy batteries for a while now but it's hard to find info and I'm not good with chemistry. I've been trying to learn more but maybe you can help me. Would this be a feasible option for a small off grid set-up. Mainly as it pertains to the discharge current of the cells. I'm not familiar with the average discharge capabilities. I know lead acid is very high and lithium-ion as well but if I had several cells in series and parallel would it be doable without having to get insane with the number of cells? Only would need like 400 watts at any given time... Keep up the good work, hopefully I can contribute soon!

bigattichouse made it!(author)2017-01-10

"Pourbaix diagrams" get you in the neighborhood, they tell you what chemical is formed at what pH for a given voltage - in aqueous solutions. This will tell you what is happening on the +/- sides of a battery (generally).

BUT much of what is available online really doesn't help in the construction of a useful battery - the crappiest battery you might buy is going to be waaaay better than any you can build. I'm working towards one, but it's definitely not going to give you 400 watts. My record is something like 3 watt hours. It's been 4 years, and I can barely keep a white LED running all night.

But don't let that stop you. Start small with copper/zinc or copper/magnesium and work your way up and learn the chemistry - that's literally what I've been doing. In an emergency, this could be used to recharge a cell phone - VERY slowly.

for an off-grid setup:

Solar panel, charge controller, 12v lead acid batteries, inverter : something like

Edison cells will give you some leeway for discharging, they can go much deeper before they are considered "discharged".

24v will give you a much more robust system (2X12v in series, then all in parallel).

BronsonMMiller made it!(author)2017-01-10

Thank you! Yeah that's about what I've found but I keep looking for another way. Even making my own Edison or Lead-Acid batteries but Lead Oxide is so expensive and hard to make and I don't know how to turn Nickel into NiOH or whatever Edison batteries use but I'm going to keep looking into it. My goal is slightly different, I really like your goal a lot but for me, I'm ok with stuff that's a bit harder to get a hold of and everything I just want something that can be built for cheap. I'm not in quite the poverty level of your batteries goals but not too far away either haha. I don't want to be a slave to the system for any longer than I need to be so I'm trying to get setup on my own on the cheap so I can grow my own non gmo organic food and all that good hippy stuff haha.

Thanks again for all your help and info!


bigattichouse made it!(author)2017-01-10

Stay away from Lead, just buy them. They're messy and dangerous to create, and the factories do a much better job of keeping things environmentally friendly than you can at home.

One hint I can give you is Nickel Foam and potassium hydroxide. You can find nickel foam online, OR you can plate nickel on iron. If you made a nickel and iron cell, filled it with water, and drained it, it will make the oxides you need - no need to buy powder. Plating is even better than foam, because the active layer is very very thin on the surface - so you could nickel plate steel wool and have it work. Be careful with your plating solutions when done. Potassium hydroxide and water is the electrolyte.

I think within a few years I'll be production ready.. just gotta make that leap.

BronsonMMiller made it!(author)2017-01-10

Awesome man, Thank you very much! Keep it up!

tytower made it!(author)2016-10-15

Well I need to build something with a lot of storage capacity . Could it be scaled up to something waist high?

bigattichouse made it!(author)2016-10-15

You might take a look at the copper version (the last step).. you can then alternate layers of aluminum and copper to make something with ridiculous capacity... just make sure all the copper layers are connected to each other for your cathode (+) and the aluminum to each other for your anode (-). You overall voltage will be lower, but closer to AA-style voltages (1.3 or so). Starting with oxidized copper in the build will also help shorten conditioning time.

I just started building such a stack today, 5 interwoven layers of copper and aluminum. I have the graphite, but I realize that graphite sheets or solid electrodes just might not be readily accessible for many people. I was able to buy a 10' section of aluminum flashing ($6) and 10' of copper ($34) at my local hardware store.

RolandasM made it!(author)2016-10-15

very similar to Medelis Battery

bigattichouse made it!(author)2016-10-15

True, except he should be getting a much higher voltage with magnesium than he does. I haven't looked into magnesium yet - but using this technique he should be closer to 3.5-4 volts with Mg. I don't know about rechargability with MG either - once it becomes MgO, it might just stay there - giving you a "one way trip"

I'm able to get 1.7 down to 1.2v with a similar chemistry (see the added end chapter) with copper on the anode with aluminum.

Additionally, Aluminum is much easier to find as scrap than magnesium.

russ_hensel made it!(author)2016-10-11

I have the same scope, built it from a kit when I was 17. Think it cost a little over 100 bucks. How did you come by yours?

bigattichouse made it!(author)2016-10-11

My dad bought it in junior high (~1957) from a friend who built it from the kit a year or two before. Over the years he replaced most of the
capacitors and a few resistors. I received it last year as a gift from
him and had to replace a few bits of my own, works beautifully once it
warms up. As a reward for repairing it, he got me a brand new
oscilloscope that sits next to it, and we've taken bets on which one
will still be operable in another 60 years.