Making heavy water

[Curix]

Hello, I have found me a new useless project: making heavy water.

Q:Why? A:I dont know.

Q:What are you going to do with it? A:Posess it, or I will get the D of it.

 

Well, I have some questions about it, are there other ways than Electrolysis of Water? They told me: try distillation, but isn't it cheaper just to let water vapourise and safe the last bits of water and vapourise that again and again and again? I don't mean to heat the wather, just let it stand there till there is only 1/10 left.

 

will it work?

[John Cuthber]

Do you know how little heavy water there is inn ordinary water?

Also, letting the water evaporate does concentrate the heavy water, but not very well. A lot of the heavy water evaporates along with the light water.

How will you tell if you have increased the heavy water concentration?

[Curix]

Do you know how little heavy water there is inn ordinary water?

 

 

6000 molecules of the one million, and I will know if it is heavy water by weighing it. Yes it will take a while, I think some months but after a while and lots of water, I should have around the 60% heavy water, isn't it?

[John Cuthber]

Where do you get the 60% figure from?

[Rocket Man]

heavy water is 17th heavier than ordinary water. i'd recomend electrolysing and putting the hydrogen gas in a centrifuge or chilling it to ridiculously low temperatures.

or exposing ordinary hydrogen to a neutron source. (does that work?)

 

six parts per thousand is a bit over the top by several orders of magnitude. 6ml of heavy out of a litre? i don't think so.

[John Cuthber]

"heavy water is 17th heavier than ordinary water"

In the very real senses that heavy water has a molecular weight of 20.04 where ordinary water is 18.053.

Of course, that quote might have meant its density is 1/17 greater which would have been roughly equally wrong since it's about 1/9 greater.

About 1 molecule in 6400 of ordinary water is HDO (a lot fewer ar DDO).

I'd have to sit and think about it but I have a feeling that the relatively large amount of oxygen present messes up the 6ml from 1L idea. If anyone wants to do the

1 litre of water is so many moles so it has so many molecules which gives 1/6400 molecules of D2 which would give so many moles of heavy water ie such and such a volume maths

it might be interesting.

 

I'd still like to know where the 60% figure came from.

[Rocket Man]

unless you're talking about tritium water(not naturally occuring) or DDO, heavy water has a molecular mass of very slightly less than 19. where did you get 18.053 for water?

[insane_alien]

O = roughly 16 AMU H = roughly 1 AMU

 

16 + 1 + 1 = 18

 

D= roughly 2

 

DDO 16 + 2 +2 =20. it'll be more than 19. HDO will be roughly 19.

[John Cuthber]

"unless you're talking about tritium water(not naturally occuring) or DDO, heavy water has a molecular mass of very slightly less than 19"

Rocket man, DDO is heavy water so, yes, I'm talking about D2O.

"where did you get 18.053 for water?"

Oops!, I'm not sure, probably by missreading/ mistyping from wiki

http://en.wikipedia.org/wiki/Vienna_Standard_Mean_Ocean_Water

gives 18.015268 grams per mole

The Merck index gives 18.016.

[pioneer]

One can make heavy water using only a couple of basic principles of science. First heavy water contains D, T and not just H. The extra neutrons implies these isotopes of hydrogen are heavier. This will make the heavy water heavier than regular water. So let us use this to our advantage. Since heavy water and regular water is so close in weight, and since water has binding forces that will prevent separation of density, we need to overcome this with a little trick.

 

The simpliest way is electrolysis to make H2, HD, D2, etc.. Now the percentage weight differences are much higher. Here is what you do, you take this gas mixture and let it diffuse down a long pipe. Since H2 is the lightest, it will get to the end first. The HD will be next and then the D2 will be last. The binding forces between hydrogen gas is very weak so they will diffuse more or independantly. Then you take some of the O2 from the electrolysis, and react it with the collected D2, to make D2O. The yield will be tiny and it may not be pure; it gives you plump water.

[Rocket Man]

tritium has a half life of about 12 years so you are not going to collect any of that.

if you take six parts per thousand as the conc of heavy water, the concentration of DDO is probably 36 parts per million.

if you want heavy water, you want to look for HDO and you'll need one hell of a centrifuge.

i still doubt 6:1K, i'll do some research

 

edit:

from wiki

Semiheavy water, HDO, occurs naturally in regular water at a proportion of about 1 molecule in 3,200 (each hydrogen has a probability of 1 in 6,400 of being D). Heavy water, D2O, by comparison, occurs naturally at a proportion of about 1 molecule in 41 million (i.e., 1 in 6,4002).

[Curix]

Where do you get the 60% figure from?

 

figure? I just mean that I am glad if I get a 60+% percentage of heavy water.

 

but I have calculated (sort of) the amount of water that I will need, and it needs to be very very much. so... Can somebody tell me where to buy some ml heavy water?

[John Cuthber]

I got some from ebay but the easy way is to buy it from a chemical supplier. It's not a restricted product.

[alan2here]

nothing is restricted

united nuclear sell uranium and ebay sells soduim

[grandzilla]

Anybody remember years ago a high school student making heavy water for a science project? NASA had spent millions trying to figure it out and this kid did it in his mom's kitchen. Maybe about 25 years ago. Any info?

[insane_alien]

grandzilla, there is a slight difference in scale.

 

on a small lab scale like the high school student, it is relatively easy to make heavy water. however making several thousand tonnes at a time has some very big issues, mostly in the cost and time department.

[noobrication]

Heavy Waters natural occurance is only 0.015%, so it may take a lot of filtration or distilation, to get any meaningful amounts

[insane_alien]

filtration won't get you anywhere.

 

distillation will only be slightly better, rough back of the envelope calculations tell me that to get a 50% mole fraction of heavy water your going to need a distillation column at least 25km high.

[alan2here]

your going to need a distillation column at least 25km high.

Could it be corkscrew shaped for efficiency of height to length?

[insane_alien]

not really, you could split the column up into more manageable chunks and just pump to top stream from the previous column into the bottom stream of the next.

 

but its still going to be impractical. you would need hundreds of columns and all just for one feed. and your product stream is going to be just a trickle compared to the torrent thats going into feed. not to mention distillation columns are expensive, expensive to build and expensive to run. especially when the relative volatilities are so close.

 

EDIT: whoops, mean to say, bottom stream of the previous into the next column.

Edited October 19, 2008 by insane_alien

[Gilded]

If someone is interested in how they do this industrially, one method is the Girdler sulfide process. AFAIK chemical processes in general are the way to go initially, followed by distillation.

[insane_alien]

i would have went for electrolysis followed by a seperation of normal hydrogen from deuterium.

[Gilded]

i would have went for electrolysis followed by a seperation of normal hydrogen from deuterium.

 

I guess it's possible, sure, but I'd imagine the Girdler process is more energy efficient (although this kind of enrichment processes always require ridiculous amounts of energy), and that it achieves a reasonable purity with a convenient chemical cascade setup. I'm not sure how desirable the process is for small scale production though.

[insane_alien]

if you're trying to separate two components that have the exact same chemistry with only a slight difference in reaction rates its going to be very energy expensive to do. there isn't much of a way round this.

[pioneer]

One way to make heavy water is first form H2, D2, T2 from the electrolysis of water. You will also get HD, HT, DT. If we took a long pipe and allowed a fixed amount of this gas to diffuse to the end, the H2 will exit first, then D2, then T2, due to diffusion rates being a function of molecular mass. D2 is twice as heavy as H2, while T2 is three times as heavy. This is called gaseous diffusion with the T2 slowest. A vertical column makes it easier since all three are less dense than air and will want to diffuse up.

 

 

This implies H2 will travel 1.414 times as far as D2, per given time. That is in a vacuum, but you get the gist.

 

You line up all the gas mixture at the beginning of the pipe, with a Y-valve at the exit end. After some basic calculations or a GC to do it empirically, time the y-valve to switch at the output to collect the D2, T2 in a separate area. Once that is done, add another gas pulse, etc. It is not perfect but one could begin to increase the ratio of heavy hydrogen. Then you react the gas collected with the oxygen, to form crude heavy water.