Sciencemadness Discussion Board - Water >> H2 and 1/2O2 electrolysis

Water >> H2 and 1/2O2 electrolysis

ElizabethGreene - 9-12-2015 at 20:46

Hi. long time no post.

I currently separate water into hydrogen and oxygen with electrolysis for a hydrogen torch. This works, but it provides a mix of gasses. This is unsafe, and I don't like trusting my hands to a $7 flash arrestor. I'd like to make another that provides separate feeds of the gasses so the hose isn't full of an explosive gas mixture. I'll also be able to use them for other fun stuff like fuel cell experiments, creating an oxygen scavenging environment, and making hydrogen balloons.

I know I need to separate the anode and cathode, providing separate gas collectors for each. I think I also need a membrane between them. Is this correct? If so, what materials are suitable for this?

Design advice much appreciated, as well as any information on diyable hydrogen fuel cells for electricity generation.

Thanks!
Greene

[Edited on 10-12-2015 by ElizabethGreene]

DraconicAcid - 9-12-2015 at 22:52

No, you don't need a membrane between them.

IrC - 13-12-2015 at 18:24

Easily done using a U-tube with electrodes properly designed and placed so each can only produce gas which rises up the corresponding arm. In the wiki pic I would raise the electrodes and have the lower portion insulated so that each gas can only rise in its respective side.

https://en.wikipedia.org/wiki/Electrolysis

Sulaiman - 14-12-2015 at 09:55

Not that way ... Sudden release of pressure of either the hydrogen or oxygen side will result in pressure from the other side forcing out the electrolyte.
This happened to one of our members who wrote about it recently, bad day in the lab OR when chemistry goes wrong.

IrC - 14-12-2015 at 10:07

The idea was to provide a method on how to produce the separate gasses from a mechanical viewpoint. Concerns such as yours are easily solved with either electrically permeable ceramic walls which block a pressure wave of water or gas, or two opposing one way valves normally open enough to allow current flow unimpeded yet able to slam shut if a sudden pressure differential occurs. Not difficult to think through.

Sulaiman - 14-12-2015 at 10:21

Oh, you mean a membrane per the OP ?

Or a complicated set of trap doors ?

or more space for gas above each electrode than the total volume of electrolyte

or ....

Consideration should also be given to pressure monitoring and/or over-pressure venting.
You could use an automotive oil pressure switch BUT they operate at 10 atmospheres pressure, I would not want to be near glassware at 10 bar.

[Edited on 14-12-2015 by Sulaiman]

IrC - 14-12-2015 at 10:36

Depends upon its ability to handle a sudden differential in pressure when one considers a membrane. A semi porous ceramic wall can handle several hundred PSI if properly designed, while not impeding current flow.

ElizabethGreene - 14-12-2015 at 16:32

Thank you for the kind replies.

I think I found a suitable design in Build A Solar Hydrogen Fuel Cell System by Phillip Hurley (Amazon Link.).

Ignoring the solar and fuel cell bits, the electrolysis cell is a PVC pipe with a pair of monel mesh electrodes separated by a felt cylinder. The unit needs to be mounted vertically to harvest the gas, and is gravity pressurized with a KOH/Water solution.

This seems doable, leaving only the relatively trivial problem of stepping my 5v supply down to ~1.8v.

The chap has a clever fix for the over/imbalanced pressure issue. I've cut out a few of the pictures and will attempt to link them here.

Does anything look horribly wrong with this?

IrC - 15-12-2015 at 23:00

Cannot tell much from these images but I would wonder about the possibility of gasses mixing by accident. I suppose it depends upon the rate of production which if not too high may not be an issue. I liked the idea of total separation where such a dangerous problem would be nearly impossible.

froot - 17-12-2015 at 06:29

Looking at H2 and O2 solubilities in water albeit relatively small this would ensure the presence of O2 in the H2 side of the cell and vice versa transported via the water being electrolysed. With that being a reasonable speculation the possibility that these gases diffuse into the O2 and H2 collection chambers from the water separating them seems quite plausible too. The question is if this would be cause for concern over a length of time of operation?

Magpie - 28-10-2016 at 11:00

I have been experimenting with the electrolysis of water using a 12vdc source. My goal is to produce and capture pure H2 at a reasonable and sustained rate.

My first try was with some carbon cathodes from a dead 6v dry cell. I used large glass test tubes to capture the gases. Since production was very low at first I thought more surface area for the electrodes was needed. So I made new ones from perforated, rolled sheet ss (from a vegetable scraper). This gave no improvement. After watching some YouTubes where people were making HHO I noticed that they were producing gas like gangbusters. So I took off the test tubes and then I too was producing gas like gangbusters. I learned form this experimentation that a clear path for ion movement must be provided.

Shown below are the electrodes and gas production in a saturated solution of NaHCO3. The amp draw is a little over 3 amps. At 3.5amps I calculated that the production rate of H2 is 1.5L/hr.

Chemetix - 28-10-2016 at 14:27

I'll put up some pics later on today, but I have had the same needs for H2 production. Which I mentioned in this thread earlier:

https://www.sciencemadness.org/whisper/viewthread.php?tid=68...

The problem was making a cell that would split the two gasses and supply H2 at line pressure for blending with another fuel gas. I didn't want to go down the path of using pumps and storage tanks and regulators and drying, cleaning ....you get the picture.

I first tried the U tube type arrangement with 0.5M approx H3PO4 as electrolyte, stainless steel anode and cathode. At 30VDC I was only managing about 0.15A. Gas production was infinitesimal. So the same electrodes were placed inside each other one was about 30mm Dia and the other 18mm Dia. and about 300 mm long. With the same electrolyte
12VDC draws 30A. The power supply can't handle that sort of current for very long before overheating. But it solved the geometry problem, and the rate is good, too good.
I've been converting a computer ATX supply rated to give 5VDC 20A just recently, which I expect to do the job nicely, and now I'm doing the plumbing for the cell.

The cell has a membrane of vidaflex [ http://www.cablecraft.co.uk/heat-resistant-glass-braided-sle... ] over the internal electrode which is sealed at the top by glass thread wound over the vidaflex, a cable tie gives it a bit more grip to be sure. which allows the internal electrode to become the anode and collect the oxygen through perforations around the top.
A silicon rubber bung plugs the outer tube and fits the inner tube with a little bit of an adapter- this pics will explain it.

The outer electrode was sealed at the bottom making the whole cell able to be pressurised. This meant that the cell can run completely filled with electrolyte and any gas rises out of the outlets to the reservoir tanks above the cell. The electrolyte drains to another external tank which is pressurised above the liquid to line pressure, where fresh water can be added. The pressurised system means no dead volume above the membrane, minimising the risk of H2 and O2 mixing, and the ability to draw off H2 without changing the pressure in the O2 reservoir.

Will have pics and more details soon.

Magpie - 28-10-2016 at 16:04

It is interesting that you talk about electrode spacing requirement. This seemed to have no effect in my cell: a 1L beaker.

My system will not require hydrogen pressure other than that needed to fill a collapsed Mylar party balloon. This balloon acts as a reservoir from which I will fill a regular heavy duty latex balloon using an aquarium pump. This scheme has been tested and works nicely.

The key problem is collecting the hydrogen while still allowing efficient travel of the ions. That's my conclusion anyway. The solution to this appears to be use of a semipermeable membrane.

I look forward to your pictures and further comments.

Chemetix - 28-10-2016 at 23:33

The following pics give you an idea of the construction and arrangement.

Here is the internal electrode with the vidaflex

Up close with the detail of the vidaflex held with glass fibre and cable tie

The current is 1.25A at about 5VDC

A gush of bubbles.

It seems to work ok except for a bit of bubble lock, the inlets to the tanks are a bit too small where the rising bubbles block the tubing and stop the tanks filling, I'm going to have to redesign the inlets to be bigger.

[Edited on 29-10-2016 by Chemetix]

metalresearcher - 29-10-2016 at 00:48

Where can I get such a device (Hoffmann apparatus) ?

Ebay has it but seller ships within US only (as many US sellers do).

Chemetix - 29-10-2016 at 01:52

I'm in OZ of course, but I can make a Hoffman electrolysis cell for you, if shipping is doable?

Its aliiiive!

Chemetix - 1-11-2016 at 02:46

Ok it looks a bit Dr Frankensteinish, but it works. The bubble lock was due to the liquid needing somewhere to go as the gas rises in the inlet tubes. This was solved by having a separate return path.

In this shot you can see that the left tank has hydrogen collecting in a big flow of bubbles while the oxygen just fizzes over more constantly.

Magpie - 1-11-2016 at 07:09

Very nice work.

Chemetix - 2-11-2016 at 01:14

@ Magpie. Thanks! That's high praise coming from one of the sites' more prolific and excellent experimentalists.

Magpie - 12-11-2016 at 12:37

Here's pictures of my progress on constructing the Phillip Hurley electrolysis unit. The smaller white cylinder is a piece of 1/8" polyester felt glued at the seam with a hot-melt gun. I'm currently awaiting receipt of 200 mesh 304 ss screen which will be used for the electrodes.

electrolysis caps internal view.jpg - 145kB

electrolysis caps outside view.jpg - 144kB

aga - 12-11-2016 at 13:08

Quote: Originally posted by Chemetix

Ok it looks a bit Dr Frankensteinish, but it works.

Nice ! (@ both)

Simple solutions tend to work best.

Is that a lamp stand being used there ?

[Edited on 12-11-2016 by aga]

Chemetix - 12-11-2016 at 13:39

Yes, lamp stand being used. That was a piece of obtanium, a la roadside. The whole arrangement has since been put in a wooden stand like an Orsat-Muenke apparatus.

On a further note I am now running 12V and the electrolyte has been diluted by at least half, and the overall performance goes something like:
12 V
4.2 A
1.2 L. hr^-1. H2.
H3PO4 @ 0.25 M

I'm interested to see how you collect your gasses Magpie.

@Elizabeth, any progress?

[Edited on 12-11-2016 by Chemetix]

aga - 12-11-2016 at 15:05

Cool ! Great improvisation with a readily-available piece of equipment !

Usually electrolysis (of any kind) rips up the electrodes really fast over about 6V.

Are your electrodes lasting any length of time at 12V ?

Chemetix - 13-11-2016 at 12:02

So far so good, can't see any oxides or scale in the electrolytye, will monitor and report.

Magpie - 9-1-2017 at 11:48

Here's an update on my Phillip Hurley design water electrolysis "electrolyzer." It's been completed for some time but it's too cold in my lab to try it out. I will load it with 29wt% KOH as recommended by Hurley when things warm up.

I made several modifications to save money. I'll wait to see if it works before disclosing those. Nobody wants hear about those if it fails.

Bezaleel - 4-2-2017 at 17:56

Magpie, do you have a link or reference to the Philip Hurley design? A Google search on "Philip Hurley electrolysis" does not seem to yield anything usable.

Magpie - 4-2-2017 at 20:51

Search Abebooks for "Build a Solar Hydrogen Fuel Cell System':

https://www.abebooks.com/servlet/SearchResults?an=Phillip+Hu...

Magpie - 6-4-2017 at 13:41

As noted earlier in this thread, I have constructed the Phillip Hurley electrolysis cell (electrolyzer) for producing H2 and O2 from an aqueous electrolyte. Today I was able to get back into the lab for the first time since December 2, 2016. My Hurley electrolyzer was put to its first test, other than leak testing with water earlier, which it passed.

The electrolyte recommended by Hurley is 29wt% aqueous KOH. As I'm not interested at this time in the most efficient or highest production I substituted sat. NaHCO3 for the KOH for safety reasons.

It fired right off, producing gasses at 5vdc, 2.5a. I tested it with both 5vdc and 12vdc. Production at 12vdc was too rapid and I quickly switched back to 5v.

I added some dish soap to the water where the H2 vents and tested the bubbles with a match. They are indeed flammable. All operations were conducted with my efficient hood fan on.

The pictures below show the test set-up. The small cup is where the H2 was being vented to form soap bubbles for the ignition test.

Below are some notes on my materials substitution to cut costs:

These comments may be mostly applicable to procurements in the US.

Hurley gives complete material procurement and construction directions in his book “BUILD A SOLAR HYDROGEN FUEL CELL SYSTEM,” 2013. Although simple in concept, fabrication of the parts and their assembly is tricky. I felt that to accomplish this successfully I needed to buy his book.

1. I used 3” thin wall PVC sewer pipe for the electrolyzer body. This was necessary to allow use of the flat end 3” caps that I wanted to use. I had to get the caps mail order and it may be wiser to just use 3” schedule 40 PVC. The domed caps for this pipe are available at most any hardware store.

2. I used a 1” PVC coupling instead of a 1” PVC pipe cap. I felt that this was easier to position and epoxy to the inside of the bottom 3” cap.

3. I used 0.2mm (0.008”) SS sheet metal instead of 0.010” nickel sheet metal. From China this was $7 + free shipping vs $61 + shipping for the nickel from McMaster-Carr. I later found out that I could get the 0.012”SS sheet metal for ~$7 at my local hobby shop.

4. I used 200mesh SS304 wire cloth from China (via Grainger) for $10 + free shipping vs 200mesh Monel wire cloth from McMaster-Carr for $27.72 + shipping.

5. I used 1/8” polyester felt for $1 from a local craft store vs 1/8” polypropylene felt from McMaster-Carr for $10.22 + shipping.

6. I used 1/8” EPDM rubber I had on hand vs ordering 1/8” silicone rubber from McMaster-Carr for $9.32 + shipping.

Hurley welded the polypropylene felt with a soldering iron. I had to use a hot melt gun to weld my polyester felt due to its higher melting point.

I like the design of this electrolyzer with one exception: it is hermetically sealed. There is no way to do any internal inspection, parts replacement, or other maintenance.

If you have any questions, comments, or suggestions please let me know.

electrolyzer in operation at 5v (2).jpg - 121kB

Chemetix - 6-4-2017 at 14:31

I cant help asking, what's the stability of the polyester like with alkaline conditions?

I can see the design flaw with not being able to open the unit, and how do you collect the gasses? I was also wondering, if brake and exhaust fitting businesses have bits of left over chrome moly tubing from exhaust fitting. If they do performance work then some scrap titanium tubing might be available too. This would be good for the tube in tube design, except it needs some metal working tools and equipment.

I guess the best designs are kitchen table construction for most mad scientists out there.

Magpie - 6-4-2017 at 15:21

Quote: Originally posted by Chemetix

I cant help asking, what's the stability of the polyester like with alkaline conditions?

The reference below says that the compatibility of polyester with potassium hydroxide is "limited." But it says the same for polypropylene which Hurley recommends for 29wt% KOH.

But it says also that compaibility with Na2CO3 is excellent so perhaps I will play it safe and just stay with my NaHCO3.

http://www.nfm-filter.com/nationalfilter/userfiles/file/Chem...

Magpie - 6-4-2017 at 16:09

Quote: Originally posted by Chemetix

...and how do you collect the gasses?

The gases are separated by the polyester fabric partition and led out of the body by the rubber hoses ... I'm guessing that you know that. The O2 is just spilled. The H2 is led to a collapsible bladder (Mylar He balloon). From there it will be pumped, using an aquarium air pump, into a rubber He party balloon. Hopefully this balloon will be adequate for some Pd/C hydrogenations at nominal atmospheric pressure.

Chemetix - 6-4-2017 at 21:07

"...rubber He party balloon. Hopefully this balloon will be adequate for some Pd/C hydrogenations at nominal atmospheric pressure"

Use it before you lose it I guess. Bog standard balloon rubber leaks H2 like a sieve, but you know that I'm sure...I wish I could find an aquarium pump that has an inlet port to speak of, anything I've seen has a hole somewhere on the housing but nothing to use as an inlet.

I have an old industrial refrigeration compressor mounted on a DIY frame, it's a bit of an overkill for this kind of work, small high efficiency compressors for H2 work aren't easy to find.

I used acetone as a hydrogenation test for atmospheric reaction conditions. Pd on either BaCO3 or C works, so does Raney Ni.

Off topic a bit buuuut....
I actually made raney from scratch many years ago, Ni and Al powder mixed and then poured into a hollowed out firebrick, I took it to a pottery place that agreed to fire it next time they were firing at 800C. The resulting mass was broken out of the firebrick and then crushed with a hammer on concrete. ( yes it was primitive) Lots of washing with NaOH and then demineralised water, and it worked very well indeed. At least as good as W3 or W5. It would ignite if left to dry on a tissue or cardboard.

[Edited on 7-4-2017 by Chemetix]

Magpie - 7-4-2017 at 07:55

Quote: Originally posted by Chemetix

...I wish I could find an aquarium pump that has an inlet port to speak of, anything I've seen has a hole somewhere on the housing but nothing to use as an inlet.

Yes, my pump was like that too. To solve this I took a 3/4"PVC coupling and sanded the end to conform to the curvature of the pump inlet. Then epoxy this to the inlet. This then takes a rubber stopper w/hose barb.

aquarium air pump inlet adapter 1.jpg - 201kB

aquarium air pump inlet adapter 2.jpg - 203kB

Quote: Originally posted by Chemetix

.
I actually made raney from scratch many years ago, Ni and Al powder mixed and then poured into a hollowed out firebrick, I took it to a pottery place that agreed to fire it next time they were firing at 800C. The resulting mass was broken out of the firebrick and then crushed with a hammer on concrete. ( yes it was primitive) Lots of washing with NaOH and then demineralised water, and it worked very well indeed. At least as good as W3 or W5. It would ignite if left to dry on a tissue or cardboard.

Nice!

Chemetix - 7-4-2017 at 20:44

Re. pump inlet
Masterful!

Magpie - 21-4-2017 at 17:35

My electrolysis unit is fully operational now. I first tried using a sat. soln of NaHCO3 as electrolyte but found this was unsatisfactory due to the generation of significant CO2 contaminating the product H2. A test bubbling the H2 into lime water definitely showed the presence of CO2. So I switched to Hurley's recommended 29 wt% KOH. After purging the system to eliminate air I successfully filled my first H2 balloon as shown below.

Edit: Warning: do not use PETE containers for 29wt%KOH. The small PETE cup I was using at the O2 outlet cracked overnight and leaked to my hood catch pan. The PP larger cup is OK. The milk container in the picture, probably also PETE, would not be useful for storing this liquid either.

[Edited on 22-4-2017 by Magpie]

RogueRose - 26-4-2017 at 07:30

I set up a trial electrolysis using a computer PSU and regular sheet steel. The plates are about 3" x 4" but only about 3" x 3" is in the water. The electrolyte is baking soda at a point near saturation. In this setup there are 7 plates but the last 2 are shorted so there are only 6 plates alternating.

I have not been able to test output but it is really amazing how much is made at 12v. I would guess it is well over 1L per minute, maybe 1.5-2. The PSU can only do about 100w on 5v (20a) and about 250w at 12v (22a) and there was no reaction with 3.3v - (that may have been b/c current was too low?).

The test was only run for about 30 seconds and the connecting wires got too hot as I didn't think it would draw so much current when I connected it.

The video shows clips of the 5v vs the 12v, switching back and forth to see the difference in production. I didn't realize when doing the test or editing the video that the system was probably being limited by the 5v current.

Electrolysis video

https://youtu.be/_PStp23sRE0

[Edited on 26-4-2017 by RogueRose]

RogueRose - 28-4-2017 at 00:38

I wanted to verify something before I proceeded with making a PSU for a cell. I've read that the main determination of how much gas an electrolysis cell produces (at least with H2 & O2 production) is based upon the current - higher current = more gas (for the most part); as long as the voltage is high enough to overcome the electrolyte resistance - in water that is about 1.2v or so.

So, if a 12v supply were used, 10x the needed voltage, will the same current be used in the setup and the extra electricity in a 12v setup lost as heat or erosion of the electrodes?

Would a cell, which works at 3v, use less current if 12v were applied?

The cell functions with the following voltage and corresponding current draw:
3v x 100A = 300w

Now, if 12v were applied, which of the following would be the more accurate representation of the power draw/needs for the cell?
12v x 25A = 300w
or
12v x 100A = 1200w

If the answer is the latter (or somewhere between the two), where does the extra energy go within the cell?

Twospoons - 28-4-2017 at 01:19

with 12V you would get about 4x the current as 3V.

12V x 400A = 4800W.

The extra energy is lost as heat in the resistance of the electrolyte . That's assuming your power supply can actually deliver 12V @ 400A. If its a halfway decent power supply it will limit its output current to its maximum rating by dropping its output voltage, - or it will decide there's a fault and shut down.

Better setup would be four cells in series, then each gets ~3V @100A, and the total power is 1200W.
'

[Edited on 28-4-2017 by Twospoons]

RogueRose - 28-4-2017 at 12:56

Quote: Originally posted by Twospoons

with 12V you would get about 4x the current as 3V.

12V x 400A = 4800W.

The extra energy is lost as heat in the resistance of the electrolyte . That's assuming your power supply can actually deliver 12V @ 400A. If its a halfway decent power supply it will limit its output current to its maximum rating by dropping its output voltage, - or it will decide there's a fault and shut down.

Better setup would be four cells in series, then each gets ~3V @100A, and the total power is 1200W.
'

[Edited on 28-4-2017 by Twospoons]

(smacks head in disbelief...) I can't beleive I didn't think about putting the cells in series! I was going about this trying to figure out how to get the V down to 1.5-3vdc and that was going to be a MAJOR PITA!!

When you said that 12v gets 4x the current, I don't think that is correct. You get 4x the power, if the current is the same (3v 100A & 12v 100A) - the 12v will put out 4x the power (wattage) but current is current.

I may be incorrect, but that is how I've learned things and as far as the power supply dropping the V to provide the needed current.

I'm going to look on an electronics sight and make sure..

Thanks for the response!

Twospoons - 29-4-2017 at 16:23

Ok, there are essentially two modes of power supply control: Voltage mode, where the output voltage is the controlled parameter, and Current mode where the output current is the controlled parameter.

In Voltage mode the resulting current is determined by the set voltage and the resistance of the electrical load. I=V/R
In Current mode the resulting voltage is determined by the set current and the resistance of the load: V=I*R

Your electrolysis cell sets what R is, through its dimensions, electrolyte and electrode composition. If that is fixed then the cell current is determined by the voltage applied (assuming a voltage mode supply). So if you apply 4x the voltage you will get 4x the current. If you have 4x the voltage and 4x the current, that's 16x the power.

Now just to confuse things, many power supplies will automatically switch from voltage mode to current mode if the current demand exceeds some predetermined threshold. This is usually to protect the power supply. In some supplies this threshold is adjustable - very useful if powering up something for the first time (you can set it low so faulty things don't go BANG).

BTW if you are making a series connection of cells, make them all as similar as possible, so the R's will be the same, and the voltage distribution will be even.

You probably know this already, but for the benefit of others reading this thread you can make R smaller (smaller = less heating = better efficiency) by:
* increasing electrode area
* reducing electrode spacing
* increasing the ion concentration in the electrolyte

[Edited on 30-4-2017 by Twospoons]

Vancoillie - 4-6-2017 at 07:00

[Edited on 4-6-2017 by Vancoillie]