Deuterium Generator

[rheil]

I am attaching some photos which should help to explain the details of my Deuterium Generator. Deuterium is generated by electrolysis of heavy water, with the oxygen being vented and the D2 gas collected for use in the fusor.

The heart of the generator is the electrolysis cell, the main components of which are shown in the first image. The platinum wire was a lucky find many years ago, I just knew I would need it for something someday. The body of the cell is made from acrylic plastic, purchased as scrap from a local plastics vendor.

The second image is a close-up of the bobbin and spindle I fabricated, wound with some of the precious platinum wire. The platinum wire is sealed into the lid of the cell where it passes through to the small terminal strip shown in the first photo.

The third photo was taken before assembling the cell. The D2O was purchased from Wilmad, and the sodium carbonate is left over from a very old chemistry set. [I heated the sodium carbonate prior to use to drive off any hydrogen (and water) accumulated over the years.] A full charge in the cell is 10 cc of D2O with 30 mg of the dried sodium carbonate.

The fourth photo shows the cell installed in a box with a meter to monitor current consumption and valves to close off both the D2 and O2 ports when not operating. The front panel is transparent so you can see the general arrangement and the first in-line gas drier. Since that photo was taken I added a second U-tube drier loaded with Calcium Chloride. This seemed to work better than just the little commercial drier shown in the photo.

In steady-state operation the O2 is vented to the air, and the D2 line builds up a very slight pressure of a couple of cm of water (heavy) leading to the needle valve I use to meter the D2 into the fusor. Having the deuterium manifold run at essentially atmospheric pressure greatly simplified the overall design. As the D2 pressure builds, it displaces the D2O out of the central column of the electrolysis cell, uncovering the central electrode and thus reducing the rate of generation.

I tried to make the design as fail-safe as I could, with the cell sealed with an o-ring so that any build-up of pressure (perhaps from operating with both valves closed) will result in the lid of the cell popping off and thus venting pressure. I also kept the D2O fill low enough so that excessive vacuum on the D2 port would simply draw the D2O up into the central tube with ambient air being drawn into the outer tube and then bubbling through the D2O and into the fusor. So far, I have not accidently tested these safety features.

Bob H.

[rheil]

During recent operation I have wondered if there was still too much moisture getting into the fusor from my D2 generator. This prompted me to make a chiller to further dry the D2.

The first image shows dewar (thermos bottle) I purchased from a local thrift shop for a dollar, and the chiller coil I made from 1/8 inch copper tubing.

The second image shows the chiller in operation. It is charged with a mixture of dry ice and alcohol. Note the ice forming on the copper tubing just above the stopper. There is plenty of clearance around the copper tube where it passes through the stopper so that no pressure builds up from the CO2. And the loose fitting stopper itself provides another safety feature should ice plug the openings in it.

Bob H.

[Steven Sesselmann]

Bob,

Looks like a well designed system, quite a project on its own.

Steven

[Dan Tibbets]

A few questions. The system you have built looks like it produces the D2 as needed. Couldn't the syringe be filled with heavy water, inverted in the bath and the the electrode inserted into its tip so that the gasous D2 displaces the water, filling the syringe, then the syringe could be caped and stored till needed so that all you would need at the fuser would be the syringe, valve, and water absorber. I have no idea how porous the syringe is to duterium gas, ie the shelf life. And how much gas is needed, even 10 cc would fill a fuser at ~ 10 microns many times over.
What flow rate is needed to maintain the duterium against the pump loses and absorption?
How dry does the gas need to be, any results?
And, does the electrode (anode?) have to be platinum?

Dan Tibbets

[Richard Hull]

The flow rate has been discussed in some detail in the past. It is truly a tiny fraction of a cc (stp) per minute. Thus, a reservoir of D2 at atmospheric pressure will supply a fusor volume at 100,000th of one atmosphere easily. Note: The vacuum pumps are always choked off once bottom pressure is achieved to allow for a slow bleed of D2 into the fusor, (using a bleeder valve), to maintain the approx.10 micron fusion operating level.

The D2 generator shown here is a superb example of amateur engineering and, with the chiller, should produce sufficiently dry D2 to a fusor.

Richard Hull

[nemesistech]

You do want to remove moisture from the system, I found that drierite works great, is cheap, and easy to set up in a system.

http://www.motionpicturetransportation.com/d2.html

[cdicken]

Hey guys,

I apologize to push this thread up again, but Dan asked whether or not one needs a platinum electrode. Does anyone know the answer to this? Why is platinum prefered ... I assume it is more resistent to corrosion ... Will graphite electrodes work too?

Thanks,
Chris

[Richard Hull]

Platinum is recommended due to its corrosion resistance and the fact that it will not plate out or chemically react with the electrolyte. Carbon is often a good second choice, but I can't recommend it here as I am unsure of its ultimate impact. platinum will provide years of superb, trouble free service. A tiny piece of rolled up foil on a wire is all that is needed.

Richard Hull

[Carl Willis]

Corrosion resistance is probably the only real issue behind the use of platinum in such a system. A corrosion-resistant substance is mostly needed at the anode, where most common metals will be oxidized into soluble ions that contaminate the water.

Graphite is indeed suitable for the anode, but some caveats are in order. Ideally a gas generator like this will remain sealed up and trouble-free for long periods of time, to protect the expensive heavy water. Poorly-bound graphite will disintegrate and make a mess in no time, so probably a few experiments are in order to determine how long graphite electrodes last.

-Carl

[cdicken]

Richard, thanks for your fast reply.

So, as any service means to disturb the system, I'll tend to use platinum wire.

I will try a small helix of 0.1mm diameter pt-wire, soldered to bell wire or something above D2O level. The junction one could separate from D2O using heat shrink tube so it wont matter if the heavy water level in any way reaches the junction.

Thanks also to Bob H and Andrew Seltzman for inspiration

Chris

[rheil]

Chris,

As it sounds like you may be considering building a deuterium generator, perhaps I could share some of the thinking which went into mine. I didn't select my approach because it was the best or the easiest. If you want easy and best, then perhaps you should consider compressed gas. I selected this approach because it seemed the best for me. My experience has been that gasses escape, metals corrode, schedules slip. I know that some of my projects can spread out in time and literally take decades (or a lifetime?) to reach completion – if ever. I wanted something which would not require an ongoing investment. I have had a couple of experiences with compressed gasses which 'got away' while I wasn't looking, so I wanted something easier to keep an eye on. I wasn't entirely successful (so far) since the drier (cold trap) I am employing requires me to buy some dry ice for its operation. But I can go down to the market and buy this as needed, and the cost is not great. In the future I hope to replace this with a small reservoir loaded with a drying agent, but at present I am generating the D2 'on demand' and thus the need for the cold trap. Mike M. has previously posted a link which shows a most impressive generation and storage system which uses drierite.

You may have noticed from my other posts that I enjoy the hands-on aspects of design. As an Electrical Engineer who started out to be a chemist, I had a little insight into some aspects of designing an electrolysis cell. None the less, I learned much more than expected by assembling a number of prototypes while developing the final system shown in the photographs. I would encourage you to do the same. I could work the equations and run the numbers, but I didn't really know how to build this device until after I ran some experiments.

I started with distilled water, salted with various chemicals to increase conductivity. Inserting a couple of copper wires let me run some voltage/current plots against concentration. I repeated the classic high school experiment and collected some gas in an inverted test tube to compare the calculated and observed production rates. The electrochemical reaction involving the copper wires was immediately visible. I expected this, but wanted to explore options. Tinned hookup wire replaced the copper, then replaced by nichrome alloy and finally stainless steel. A number of pretty colors were developed as the electrodes were eaten away, as did an understanding that if I wanted the cell to be able to operate for more than a few hours, I would have to use something more inert. I chose the platinum because I had it, and knew it would do the job. I thought about carbon, but the physical aspects (structure, mounting, porosity, purity, etc.) as well as the availability of the platinum wire drove my decision. In the end, it was my desire to build something which could work over a long period of time without requiring the sacrifice of the relatively expensive D2O during cleaning and rebuilding while also being simple and reliable which drove the final design.

Platinum is expensive, but so is a gas handling system. The wire I obtained was intended for use in replacing the corona wire in a very old copy machine. Perhaps similar wire was used in the original laser printers? Or perhaps some jewelry or jewelry making supplies will present themselves? I believe that Richard H. also sells some materials which may be of use. Surface area is important here, so don't waste money using thick electrodes.

I also have experience with the galvanic corrosion which can arise from connections of dissimilar metals in a moist environment. Since the interior of the cell would be at 100% humidity, with at times an applied voltage, I wanted to avoid any dissimilar junctions inside the cell. This is why I made the platinum electrodes a little longer so that I could bring them directly out through the top where I made my connections using a terminal strip. I wasn't worried about the platinum corroding, but I believe that if I had used copper wire to connect to the platinum inside the cell I would have eventually had a problem.

I estimate that the generator shown has been powered on for about 100 hours, with about 8 of them actually being used to actively source D2. No deposits or corrosion are evident inside the cell.

While operating, my fusor only consumes about .15 sccm of the D2, which is supplied through the deuterium manifold to a micrometer controlled needle valve at nominally 10 Torr above atmospheric pressure. In preparing for operation, I tend to flow the D2 at close to the maximum generation rate (about 0.5 SCCM with the small transformer in use) for an extended period to help sweep residual impurities from the system. I initially charged the cell with 10 cc of D2O and, from the looks of it, there is still more than 9 cc left in the cell. A quick calculation convinces me that I probably have only used less than 2% of the initial 10 cc for generating deuterium, with the rest of the loss being due to D2O vapor being lost to the driers and out of the O2 vent.

Operation with a higher voltage transformer can double the D2 production rate, but the cell gets warm, and my first fusor can't choke down D2 at this rate while still keeping the voltage reasonably high.

Changes implemented in the design since I posted the photos have been few. I added an LED behind the cell for illumination so that I could see the difference in height between the liquid columns. This visual indication of pressure is actually more useful to me in setting up operation than the meter. I also plan on replacing the existing tubing with stainless steel capillary tubing to reduce the volume of the deuterium manifold. I already mentioned my plan to try to eliminate the cold trap. If I have to take the cell apart, I may also rewind the electrodes with something approaching a log spacing as opposed to the current linear spacing. I think that this may help to give me more stable operation, at present the generation tends to swing quickly from nearly full off (central electrode full of D2 gas) to full on (central electrode fully covered by D2O). A log arrangement would let the generation rate rise and fall exponentially with gas flow, rather than linearly as in the present system. In reality, I have had so few problems with this device that I probably will defer any changes unless absolutely necessary.

Simpler designs than mine have been used and discussed in these forums. Even simpler might be a 'throw-away' design consisting of stainless steel or copper electrodes which were operated for a short period and then discarded. This would result in the waste of some D2O, but not much is needed for our purposes. 1 cc of D2O should produce about a liter of D2 gas at STP. How long would it take your system consume this much?

A final thought; while D2 is essential for fusion, it is not the most difficult part of building a fusor!

Good luck, stay safe, have fun,

Bob H.

[Wilfried Heil]

Bob,

let me add my appreciation on the deuterium generator. Nice design and the concise description will help others who want to build something similar.

How dry does the deuterium need to be, or how much D2O in the gas will be acceptable without deteriorating the fusion performance too much?

Wilfried

[Jon Rosenstiel]

Hi Bob,

I agree with Wilfried, an excellent and easy to read report.

Bob, being we don’t live that far from each other we should get together sometime and run a fusor ease-of-operation and neutron output comparison test between the deuterium from your generator and the deuterium from my lecture bottle.

Jon R

[cdicken]

Bob,

thank you many times for this great description! I will take most of you suggestions into account, especially the point of galvanic corrosion is of high interesst. For the sake of comparsion I will try the device with a short platinum helix connected to chopper wire or something. In any case I will try the logarithmic turn separation.

Hope to report about my own machine soon!

Chris

[Steven Sesselmann]

Carl,

How about trying titanium anodes?

I was in the electroplating business for a while, and in the gold and rhodium plating tanks, we used titanium mesh anodes. these lasted a long time and appeared passive.

Steven

[rheil]

Thanks for the kind words.

Wilfried, I am not sure how essential the cold trap is to operation, but it will be one of the first items I evaluate in F2, when I have the larger power supply and the improved chamber on line. I don't like the need to keep buying dry ice! I remember your post about injecting D2O into your the chamber and detecting fusion.

Jon, I would love to take you up on your offer. Do you have a portable neutron detector that we could employ? I am planning a 'paint can' style wax moderator which will employ either silver or indium and a GM tube to use as a simple neutron detector. I would love an opportunity to get some sort of a coarse calibration on it.

I edited my post to correctly indicate that the D2 delivery pressure to the needle valve is about 10 torr above atmospheric pressure.

Bob H.

[cdicken]

Steven,

after reading your post, I just ordered titanium wire to verify your assumptions. I will run several tests with ordinary destilled water and post my results here.

Chris

[Jon Rosenstiel]

Bob,

Yes, I have a battery operated Ludlum model 12-4 Bonner ball neutron detector. Should be no problem to do a rough transfer of calibration to your instrument.

Jon R

[AFW]

About anode corrosion
In an alkaline system like sodium carbonate solution most metal ions won't be soluble, but you may still find insoluble oxide gunk slowly accumulates. Aluminium becomes "passive" due to oxide film, but this may become insulating as in electrolytic capacitors. Chromium- plated anodes might do quite well, though. I'll try one, using the tip of an old telescopic antenna (in sodium carb.+ ordinary water) and report what happens.
Tony Webb

[Chris Trent]

I've attempted aluminum electrodes before in an unrelated experiment and they dissolved rather quickly regardless of the electrolyte I used.

Some formed a precipitate, some not.

[AFW]

Further to my last post, I've just been trying various anodes, with the following results:


The cell was set up in a sodium carbonate solution containing 5.2% w/v Na2CO3.
a copper wire cathode was used.The solvent was tapwater.

Anode material : Comments
Copper : Bulky green precipate formed rapidly. Metal surface
corroded.

Chromium : Green / brown precipitate formed rapidly. Surface
deeply pitted within 5 minutes.

Aluminium : Cell became non – conductive very quickly, even at 18V
(the maximum used). Metal remained bright.

Nickel : Similar to Al. but metal dulled and acquired green
surface.

Stainless Steel : Cell maintained a good current, with no obvious
polarisation
effects, for 10 minutes. No precipitate was seen, though
the metal surface was darkened slightly. The anode
current density was 0.24A / cm2, and the cell
voltage approx. 6V. (total current 1.2A, not much variation
from switch- on for 10 minutes.)

I’m not sure what type of S / steel I used, but it is a cheap “cutlery” grade, and is magnetic. It looks quite promising as an anode material.
Regards
Tony Webb

[cdicken]

Tony,

I share your observations concerning copper. Using brass wire as electrodes, greenish deposition formed at the anode and surronding water withing 5 minutes of operation. The cathode keeps rather clean, but shows black marks where it hits the water level. I expected this, because in an similar way, one produces probes for scanning tunneling microscopes ... the metal is eaten away much faster where it hits the water surface and thus a fine needle is formed. Mean voltage was probably around 7V ... I used 5% w/v NaHCO3 solution ..

I will do another run to observe current versus time for constant voltages. Maybe its possible to get the thickness of the electrodes using a laser beam and observing its diffractionpatterns. I did this with hairs some time ago and it worked well ...

Chris

[Richard Hull]

I am loath to use a non-deuterated chemical ion source added to a D2 cell.

Long ago, in these forums, I suggested the addition of a microscopic piece of pure sodium or lithium to about 10ml of D2O. This would create NaOD as an electrolyte. This new concentrated solution could then be added a drop or two at a time into the cell with D2O to make it conductive for electrolysis. Thus, the electrolyte ionic, itself, would be a deuterate/deuteride.

Richard Hull

[cdicken]

Thanks, Richard, for this hint ...

Another issue then would be the desiccant. I have dried calcium chloride in stock, but I am afraid it will gas out and mess up the whole system. I know that some people here use drierite, which is as far as I know dried calcium sulfate, but I cant see how this change things ...

Any suggestions?

Chris

[Richard Hull]

Drierite, Zeolite, phosphorus pentoxide.

Richard Hull