Is fusion safe ? and beryllium chemistry

Dear Reader,

I imagine that you have seen the suggestions by fusion experts that nuclear fusion will give us a cheap, safe, clean and green source of energy which will provide power for the world’s needs. I am currently thinking about how green is fusion, right now I have contacted a fusion expert who I know and I am awaiting his views on the matter.

While we are waiting I think it is important to ask the question of what was is the typical cause of a nuclear accident. Is it a issue with management, an “act of god“ or was it a technical failure ?

In the case of the Windscale fire I have seen suggestions that it was human error, poor design of the reactor or mismanagement of the project. I know that before the 1999 Tokaimura that a criticality accident at the JCO site was considered to be a was considered to “be an unrealistic scenario” according to the UN report on the accident.

I have to ask the question, did a failure in a regulatory body (either the external state regulator or the companies own internal regulation) cause the first step to be taken which lead to the accident in 1999.

One model of how accidents occur is the Swiss cheese model, the idea is that a weakness in a system is like a void in a lump of cheese. Due to some event a void can appear in the organisation, this void can grow in size, shrink, vanish or move around. As long as some solid cheese exists which prevents a path existing from one side of the block of cheese to the other then everyone is “safe”.

But when a series of holes align themselves to create a path through the cheese block then an accident occurs and then the airplane crashes, the core melts or some other horrible outcome occurs. In some ways the most important step is for the plant owner or the management is to recognise that a given type of accident is possible.

This first step of admitting that a given accident type is at least a theoretical possibility enables the company to start to take steps to prevent it occurring. For example the understanding that someone could get a body part caught in the moving parts of a machine lead to the idea of the 19th century UK law which requires where possible all moving parts of machines to be fenced off or guarded.

While it is impossible to fence off some moving parts such as the chain of a chain saw or all the parts of a handheld electric drill, this law does improve safety by greatly reducing the number of moving parts which can cause injury to factory workers. In the same way if a fusion reactor is going to be built we need a good understanding of the possible threats which it poses.

One is the beryllium used in the heat transfer fluid in some designs, I was reading recently about fusion reactor safety and I saw that a mixed lithium / beryllium fluoride has been proposed as a tritium breeding layer and as a heat transfer layer. I can tell you that beryllium is a very nasty element, in some ways it is worse than some of the radioactive elements. As a result special care will be needed if beryllium or its compounds are used in fusion reactors.

I have looked at the crystal structure of Li2BeF4 (J.H. Burns and E.K. Gordon, Acta Crystallographica, 1967, 1, 1948-1923), this is an interesting looking 3D network. But before we get stuck into it we should look at some organic salts of “H2BeF4″. L.A.Gerrard and M.T.Weller (Acta Crystallogr.,Sect.C:Cryst.Struct.Commun., 2002, 58, m407) report a nice and simple tetrahedral BeF4 unit which has protonated DABCO as the counterion. Those of you who know VSEPR should have predicted that one OK. Here is a picture of the anion in the solid.

The tetrahedral BeF4 dianion

If we have less fluorides per beryllium centre (to make the Be:F ratio 2:7) then we need to use one of the fluorides as a bridging ligand to give us four electron pairs (eight electrons) around all the metal centres. Then we get the following dinuclear complex. See S. Aleonard and M.-F. Gorius (C.R.Seances Acad.Sci.,Ser.II, 1989, 309, 683)

The BeF7 trianion

If we go a little further and have a Be:F ratio of 1:3 then we will end up with a dinuclear complex which has two bridging flourides. This is shown below. (B. Neumuller and K. Dehnicke, Z.Anorg.Allg.Chem., 2005,631, 2535)

The Be2F6 dianion, note the SiF6 dianion in the right of the picture

And now for something completely different (sorry no monty python for you today) if we mix lithium and beryllium fluorides with an salt of an amine fluoride to give us Li2Be4F14 in the unit cell (L.A. Gerrard and M.T. Weller, Chem.Commun., 2003, 716 ). This network will have a charge of -4 and it will form long strips of metal atoms which are in a 1D coordination polymer. Here is the picture for you of the metal atoms and flouride anions in the unit cell.

The metal and flourine atoms in one unit cell

How here are two strips of metal atoms side by side.

Two strips side by side, note that there are no interconnections between the strips

Now here is four strips viewed from a different angle.

Four strips viewed from a different angle, note that they do not touch each other

Now if we look at Li2BeF4 we will see it is a complex solid, I have looked and all the metal atoms have tetrahedral environments, here are a series of views of the unit cell to show you what the solid looks like. This is going to be hard, it is a 3D coordination polymer. These 3D coordination polymers can turn out to be what I call “atomic fog” but this one is not too bad, I have seen much worse in my time.

Side view of Li2BeF4 cell showing the bonds going in one dirrection

Now after turning by 10 degrees

Now the end view.

One last thing in case any of my readers are thinking of doing beryllium chemistry, my short answer is “do not do it !“. Beryllium is the most toxic non radioactive element, some forms of it are almost as bad gram for gram as Pu-239. In some ways I would like John Hunt (the voice of the UK’s AIDS advert) to dispense advice to you about beryllium chemistry using the voice of doom, but you just have me right now.

I would suggest that if any chemistry students do not want to turn back and do something else then I suggest they talk to your local friendly radiochemist  and learn how to work with gram amounts of plutonium. Then do the beryllium chemistry in the same way using negative pressure boxes and all the other safety precautions which you would use for large scale Pu work.

Molecular boxes

OK a bit more about prussian blue and what are known as coordination polymers. The idea of a coordination polymer is that it is a repeating network where bridging ligands link each metal centre to the next, thus making a polymer.

To understand them we need to build up to it, it is not a good idea to try to run before you walk. Lets start with a metal complex which has lone pairs poking out into space ready to bind to something else. Err Oh Err how about [CpFe(PR₃)(CN)₂]^- this is a 18 VE complex. For those of you who do not know what the 18 VE rule is then I suggest that you read Tony Hill‘s book which will explain the 18 VE rule. A copy of organotransition metal chemistry by Anthony Hill can be obtained from the RSC.

Here is a picture of the anion, you can see it has the two nitrile groups which each have a lone pair on the nitrogen atom.

The cyclopentadienyl iron triphenylphosphine dicyanide anion

Now here is a complex where the Cu(PCy₃) group is used, this group is a weak lewis acid so it has the ability to bind to lewis bases such as the lone pairs of the cyanide groups of the iron fragment.

You can now see the square-ish arrangement of two copper atoms (green) and two irons (yellow), what you are looking at may be a bit of an “atomic fog” but do not worry yourselves too much.

The diiron dicopper square complex

As I feel kind here is a picture of the inner core of the complex, I have removed the carbon groups attached to the phosphorus atoms.

Copper iron square complex without some of the groups to give you a better look

Now if we use three cyanides to link a metal to three other metals we can make a cube, well it is a bit of a distorted cube. I am sure that the more able minded persons reading this can understand how this is step towards a infinite solid with lots of cubes in it. Here is a picture of it. Before you ask the cage is anionic, and the atom at the centre of the cage is a potassium. This complex was published by T.B.Rauchfuss in the Journal of the American Chemical Society, 2007, 129, 1931.

The cube of eight cobalt atoms bridged by 12 cyanide ligands

The cube is capped with cyclopentadienyl and tetramethylcyclobutadiene ligands. For those of you who can not see through the atomic soup (or atomic fog) here is the cube without the capping ligands.

The cobalt cage without the capping groups

* If as a university teacher you do not get on with Tony Hill’s book then you can always replace this with another book.