Target Material vs Fusion Yield for D-D, and D-T Reactions

[lutzhoffman]

I have been pondering what the relationship is if any between the target metal, and the fusion yield, for these common reactions.

Mo was for example recently shown at Berkeley to produce 50% of the N yield, of a fresh D saturated Ti on copper target. Even an Aluminum blanking flange is sometimes used in the laboratory as a N source when bombarded by medium energy deuterons (100kev - 500kev)

What I would like to know is: If there is a relationship between the fusion yield. and the target metal used. One which goes beyond just the amount of D implanted into the metal, by the D+ beam?

To help answer this question I have been looking for a study where they bombarded a bunch of different metals with D+ beams, and recorded the N / fusion yield, under similar conditions.

I am wondering if there is a possibility that some metals, or even alloys of them, can hold the D in a way which lines up the D atoms to increases the hit probability as the D+ ion comes along from a specific direction?

In dense high A# metals the D+ ion may retain enough energy for example to still initiate a reaction with another implanted D atom, this is not however what I am looking for. I am looking for the "ducks in a row" at the shooting gallery so to speak. Under the chance that someone has such a study in their library, then I would love to see it. If Mo can do 50%, then what about Ta, U, or even the infamous Pd, with its ability to turn H2 & D2, into H+ and D+ while in its matrix? Thanks......

PS: Here is some work on the subject, but it is not exactly what I am looking for:

[Carl Willis]

Hi Lutz,

I personally don't know of anything meeting the description of what you want, but I attached a paper about neutron generators that I think is a good overview of the design problems in general. These authors consider scandium as a target material, in addition to titanium.

An experimental comparison that aims to isolate effects other than "just the amount of D implanted in the metal" would seem to be enormously difficult to control, since this is the dominant issue and depends considerably on the metal in use. From the attachment:

>The neutron production efficiency of materials depends mainly on their capacity to retain deuterium and tritium and on their stopping power. The more deuterium and tritium they can retain, the more nuclear fusion reactions can occur between incoming ions and occluded gas. The lower the stopping power of the material, the less energy the ions will lose by interactions with it.

Provided for what it's worth.

-Carl

[Chris Bradley]

Just a 2¢ comment for your reflection: I would tend to think this is a job for a metallurgist with an interest in practical fusion!

Two chemically similar metals are not necessarily alike, and it may well not be due to the chemistry of the substance but rather the crystal structure phase and the interstitial characteristics therein.

Whether there is some phase of some doped or alloyed metal that is better than just that pure metal is a parameter space that you are on your own to experiment with, so I would suggest to approach this subject then you'd need a theoretical grounding in metallurgy to scope out the best candidates prior to experimentation.

Perhaps I might usefully suggest you begin by taking a look at the advances in those substances being developed to store hydrogen and understand why they are good at storing the stuff.

I cannot decide, thereafter, whether you would want to develop such a meta-material out of high Z elements (less nuclear energy transfer but more electrons) or low Z (more energy transfer to material, but fewer electrons). I'd hazard a guess you should want to aim for the highest [Z/density] value.

[Chris Bradley]

(EDIT: I posted these comments in error, without thinking it through. See below.)

So for a 120kW beam input, they are getting >2kW of neutrons out!

. >2% efficiency with just a beam into a hi-Z-target!

If ture, it rather makes the fusor's attempt to recirculate multiple-pass ions look silly, at 1E-6% efficiency, if you can do a one-pass beam at 2%!

OK, I've only read the abstract, but I find that to be a statement that needs to be teased apart for its veracity.

[Carl Willis]

>I find that to be a statement that needs to be teased apart for its veracity.

I find that your math needs to be teased apart for its veracity.

The abstract claims a neutron yield of 1E+14 / s with either 1A or 2A of beam current through 120-150 kV, depending on design. Many other projects have arrived at similar numbers. This is not an outlying report in that regard.

The fusion energy output from the DT reaction:

1E+14 / s * 17.5 MeV * 1.602E-13 J / MeV = 280W

Of this, the neutron yield is ~225W, an order of magnitude lower than what you figured. Of course this is better than you get from a fusor...it's the DT reaction, for godssakes. Number yield from the DD branches at similar conditions is down by a factor of 100 or so...energy yield from DD is down by a factor of five...if this were a DD neutron tube you might expect about half a watt of fusion.

Neutron generator tubes do their job very well. It has never been a secret that they outpace amateur fusors by a long shot, and it's also easy to figure out why. The fusor's advantage as an energy reactor remains purely speculative despite all the sycophancy on its behalf. They are attractive to amateurs probably because they are simple. Neutron generator tubes are more complicated to realize.

-Carl

[Carl Willis]

Lutz,

I found another reference of potential interest to you by Fiebinger, although I cannot obtain it at my library. The excerpt below is from an article by Coon ("Targets for Production of Neutrons") in Marion and Fowler, Fast Neutron Physics Part I, that is a good read in its own right:

>Fiebinger [Z. Naturforsch. 11a, 607 (1956)] measured the yield of neutrons from 16 different elements bombarded with 50 uA of 300-keV deuterons. He found that after bombarding for a time sufficient to reach saturation as measured by neutron yield, gold and gray electroplated chromium gave the highest yields. In each case the yield was about half that obtained from D2O ice. Materials with high hydrogen solubility such as tantalum, zirconium, and titanium gave very low yields in these tests. Saturation yield for gold was obtained after bombarding with about 0.6 coulomb of beam through limiting aperatures of 1.2 cm diameter. Calculation of the maximum yield obtainable with this amount of deuterium to serve as target, indicates most of the beam must have been concentrated in a much smaller area of the order of 0.1 cm^2.

>[...] Fiebinger observed that in all his targets the yields dropped with increase in temperature of the target, the yield going nearly to zero at 800 C for all elements investigated.

>Attempts at Los Alamos to confirm the high D(d,n) yield from gold have failed. A near-saturation yield of only 0.12 that for D2O ice was observed after 0.75 coulomb on an area estimated to be 0.2 cm^2. This yield is only one-fourth that observed by Fiebinger, and the low value in this observation may have been the result of condensation of residual vapors on the surface of the gold. Fiebinger employed extreme vapor trapping near the target whereas Los Alamos experiments did not. To observe yields which are characteristic of the metal employed, the surface of the metal must be kept clean. [...]

>Absorbed gases in the surface of a metal have marked influence on further gas absorption. The performance observed by Fiebinger for tantalum, zirconium, and titanium probably would have been markedly improved if the metal had been outgassed prior to bombardment. Observations at Los Alamos gave much higher yields for zirconium outgassed by melting in vacuum than for untreated sheet stock.

>Many experimenters [references omitted, let me know if you want them] who have employed the beam loading technique as a convenient means for generating D(d,n) neutrons have obtained saturation yields only 0.05 that of D2O ice or even less. If gold were employed and if there were a feasible way of keeping the metal surface clean, these yields could probably be increased by a factor of 10.

-Carl

[Chris Bradley]

Fair play. I think I punched in "10>exp>14" into the calculator.

OK, so out by x10, and DD is usually presumed to have a yield 1/200th of DT. Still, that'd still be 1W which is 1E-3%, still 3 oom better than a fusor, or have I played butterfingers on the calculator again?

[Carl Willis]

Chris,

That might be ballpark accurate for a lot of the amateur fusors that run at a few dozen kV.

But it may not be a fair comparison. I would expect closer competition for devices of similar scale. To be a fair comparison with this kind of neutron generator, you'd have to look for examples of fusors running 120-150 kV / 1-2A. There may be some that do this on a pulsed basis, such as at the University of Wisconsin.

I recall a comparison of actual devices sometime in the distant past on this board...a real experimental DD neutron generator, versus a fusor that used a similar amount of juice. I can't find it right now.

-Carl

[Chris Bradley]

Carl Willis wrote:
> To be a fair comparison with this kind of neutron generator, you'd have to look for examples of fusors running 120-150 kV / 1-2A.
That's true also, theoretically another oom closer still. (..I'm not sticking by my earlier comment, incidentally, happy to withdraw it... I was incorrect to ignore the cumulative totting up of each of these differences.)

Were you thinking of the NSD devices?: http://www.nsd-fusion.com/2.5mev.php

[lutzhoffman]

Thanks everyone, I have a lot to digest here now. I do realize that I am searching for something which may very well not exist. In this case I am looking for confirmation that it indeed does not exist, and that its all just a function of the amount of D trapped in the metal matrix.

What is interesting to me is that the crystal structure of metals can be influenced by so many things, and then it goes on and on when you get into alloys. Then there is also the posibility of bombarding with negative ions etc. My objective is not to make neutrons, rather it is to tweak the efficiency of the reaction for optimum yield. With D-D it just so happens that the neutrons produced provide a convenient way to measure the number of fusions occurring.

Maybe Doug's approach of using crystals is better, I just do not want to give up on metals just yet, because metals lend themselves to easy heat removal etc. and represent very durable targets.

I suppose the best approach would be to read as much as I can find on the subject, and then resort to good old experimentation. I like the idea of using low power uA level beams, at 50-500KV, this way the neutron hazard is kept down.

The discrepancy between the yield on Gold mentioned by Carl is also very interesting, it makes you wonder if there is some meta-stable products can be formed under bombardment conditions which boost the yield?

I agree with the point that to compare a fusor with a solid target, you have to be in the same ion energy range. I think that all approaches to fusion are equally valid when it comes to research. I mean if everyone only tried one approach, then the chances of a major breakthrough some day, would be much reduced. Either way it allows for so many potential experiments to be done, and that is after all the fun part, for many of us. We may not have multi million dollar labs, but we do have one advantage which is not having a board to consult for permission to try something : ) The hard part is in being scientific in the process, and in knowing when to move on, and when to try something else : )

[Richard Hull]

A lot of the better CF researchers have pushed into the very area you are discussing.

When results are seen and can't be duplicated, many fall back to a possible lucky metal lattice situation based on the way the metal was processed or handled.

It goes back to finding some expert metalurists willing to be on a team to study the D loading in controlled lattice structures. (if that is really possible) Another whole world in itself.

Richard Hull

[Doug Coulter]

This is similar thinking to the "smart target" idea I spread at the last HEAS. If you could use Rutherford scattering to make a flood beam channel down along lines where the desired target material was (D in this case) you could improve the interaction rates considerably. If that is, the scattering isn't swamped by energy losses to the electrons in the target before that happens.

I have been talking to the Chem dept at Va Tech about this, they are finding me a crystallographer to work with one of their better inorganic chemists on this for us. To get things like this done on the good 'ol boy network for free takes awhile, but it's been in progress for some time now, so I'm about to check back with them and see, and have been studying myself. Sadly, monovalent hydrogen doesn't often take on the right places in crystals to make this fly all that well, and simple hydrides both don't have the right crystal structure, and aren't too stable at high temps and vacuums. There are some hints that trinary compounds with the other two things multivalent could work, but there's not much public info on say, the chemistry of U beyond some very simple compounds used in refining. I doubt it's secret, it's just that crystallography is considered a boring field by many and has it's own journals read only by other crystallographers....

Obviously higher Z atoms would scatter better -- but also are larger which cuts the percent of the desired stuff down, and make the incoming ions lose more energy per interaction. So the really best thing might be a surprise when we find it. Could be you're really close with Ti already.
Of course, if you could make the target out of other things too that would take part in a reaction with the beam, well....but even at what you call medium energies are far past what most of this group is willing to play with....not many chances. I think Be and B get interesting at fairly high energies -- and interestingly will get into fairly complex chemical compounds that may be good for this.

Chris -- neutron tubes are far better than most fusors (although mine is catching up fast) but!
With them there is a zero chance of recirculation, with a fusor we can still hope. I am now making transit time measurements and building high power video amps so I'm doing more than just hoping on that topic. As of course you are with your interesting idea....I doubt a plain jane DC fusor will ever have much, though.

I absolutely don't buy the idea that neutron tubes are "more complex" because in the design I've put links up here to many times, it's patently obviously not so -- they are if anything, simpler, especially if you forget the "sealed off" part of the normal requirement. It's the same stuff, just arranged differently.
The old '50s Phillips design had one cool trick -- the target electrode shape to prevent back acceleration of electrons. Not hard to duplicate at all...

The thing with them is that that's it -- there is little chance, outside of a "smart target" of ever improving one much. So if you just want a ton of neutrons, make that tube. If you want fusion energy, try something that has at least a chance of doing better than one strike per ion.

[Richard Hull]

I'm with Carl on this. Neutron tubes are so good due to D-T's 100+ gain and if D-D, their 100kev + potentials. Amateurs, even the best here, work at 60kev max and use only D-D. Not much room for casual or inexpensive advance.

Of course, leave behind all hope ye who enter for power fusion; be it fusor or tube based.

Fusion beyond a certain number, maybe 10e7 or 10e8 fusions/sec, in amateur hands, demands complication and cash outlay that amateurs rarely have time, skills or money for. There might be rare exceptions, but the rule is in place and fixed rather firmly due to the physics, legal limits and costs involved.

Richard Hull

[Doug Coulter]

I left hope behind a long time ago Richard ;~} It'll follow you if you pull on the leash some, though.

And as usual you are right about a lot of things. Here at my lab we are looking at how to scale things *down* as our Q numbers get better -- too much of a good thing! There do appear to be some scale-dependent effects not in our favor for that, but we are looking at it. As you say, above a certain point all the radiation is a problem to deal with while we just want to learn how to do this "at all" -- we can learn how to do it big later if the results warrant that. The sheer amount of lead and borated wax is getting daunting around here.

Yes, the DT gives the first hundred or so gain factor in neutrons....now to account for the other few thousands difference in Q -- or just neutrons, since of course the DT has more energy output per reaction too.

Note these numbers are for neutrons per micro-coulomb, not per 10 milliamp-seconds...

I shake my head sometimes how people who get excited over a few percent error or unexpected small observed effect can't seem to handle the idea they may be off by many orders of magnitude in assumptions elsewhere.

In fancier systems (solid D or T targets , pulsed beams and such) people have done much better than this, this is from a 50's Phillips gas tube design handbook, and are numbers for a small sealed off tube running at 125 kv. Still nowhere close to net power gain, of course, but we have a looooonnnggg way to go with fusors to even get to here -- and again this is half a century old "state of the art" many of us could duplicate in an afternoon. Things haven't stood still in this field either, though as you say, at some point enough neutrons is enough, so they are working other areas (price and reliability for example).

I would further dispute that the pictured design is much harder to make than a fusor -- that's simply not so, this is very simple, especially if you don't want to make a sealed-off one. If you wanted to do that, you'd need some practice in making sealed off *anything*, which is indeed another step (I am still struggling for long life neon tubes here, but getting better), and swipe the D reservoir out of some old thyratron...The only cool trickery here is getting the focus just so -- not to a dot, but spread out to even out the heat load on the target. You will note the design of the target/electrode that prevents electrons from being back-accelerated much. Not hard to make at all.

They mention in the text here that a good ion source was one of the things foregone for simplicity and ruggedness, the one pictured reduces the yield a good bit due to not making very many monatomic ions. And it still kicks all of us around the block!

I plan to make one in a quartz tubing stubbed off one of the vacuum systems here and verify the numbers if I can at some point. But if you're after gain, this is a dead end -- stick a fork in this one, it's done already.

Here's the link again for those who RTFA, this is at the end of the chapter.
A drive in target is easy to make with the stuff you have already to have a fusor.
I note they only claim that T adds 3x, not 100x. I suppose this could be due to the non monatomic ions or something similar -- practical applications often deviate from theory that only covers some ideal case. But for us, that's *good* news -- it means we lose only 3x from these numbers for all D systems. If you just want a buncha neutrons, this is it. If you want gain someday, look elsewhere.

http://www.coultersmithing.com/OldStuff/pdfs/Pch8.pdf

[lutzhoffman]

Hello:

Thanks for posting the tube design, man it is pretty simple. One thing I wonder about is ref, to the ion source that they used. This looks to be an internal PIG source, which will give a broad range of species, when compared to a cleaner RF one, with 80% or more of d+1.

Doug: This has me wondering if in the one that you plan to build: If you plan to replace the stock ion source, with an RF/ Microwave based one? Construction wise it should be even simpler, and even cleaner from a vacuum perspective, just a closed off end of quartz, with a single pin electrode.

The pressure regulation could even be made via an induction / external heating based system, to avoid the extra glass / metal seals for the filament, etc. Kind of like they did in the old gas x-ray tubes with the side stub of glass, with a gas emitting contents inside. If you used Quartz this would be easy with an external heated stub.

The pin for the ion source tube end, could be made from an off the shelf Pyrex-Quartz graded seal, with a W pin in the Pyrex end. I am just mentioning this because W to Pyrex is the only HV seal which I was ever able to get to work right! So in other words it must be very easy : )

I would clean the sanded W in molten NaNo2 (Nitrite, not Nitrate), and melt a small bead onto it from a small piece of Pyrex tube, and then the rest was easy. I used the smallest Pur-Tung TIG welding electrodes, simply because I had them, but I suppose a more maleable form of W would be better. This would be a cool skill to have, to be able to do large D housekeeper seals etc. Take Care..... Lutz