8 February 2012
Ever since I first heard about the idea, I have loved inertial confinement fusion. The basic concept involves blowing stuff up with lasers to get some energy, then doing it again and again as fast as possible. What more could a 38-going-on-5-year-old want? Well, what I might also want is a fusion reaction that generates more energy than you put in to it.
One thing that lets me down about inertial confinement fusion is that the implosion that gets the fusion reaction going also acts to stop the fusion. One idea for improving the fusion reaction that has been floating around for a while is to use magnetic fields in place of lasers to increase the efficiency of the fusion burn. But until recently, no one could figure out how to make it work properly.
A crash course in inertial confinement fusion
Fusion is the process whereby the atomic nuclei of lighter elements are combined to make heavier elements. So sticking two deuterium atoms together (deuterium is a form of hydrogen with a neutron and a proton) will give you helium and 3MeV (480×10-15J) of energy. To put that in perspective, one gram of deuterium will provide 144 billion Joules of energy when it is completely burned into helium. One gram of benzene, a common hydrocarbon, releases just 48kJ when oxidized (burned in the normal sense).
But fusion is not so easy to achieve. Although atoms are electrically neutral, the parts that need to be stuck together—the atomic nuclei—are positively charged and repel each other. The external pressure needs to be high enough that it overcomes the Coulomb forces holding the nuclei apart.
In traditional inertial confinement fusion, the compression is driven by lasers (all the really cool stuff involves a laser somewhere). A perfectly spherical droplet of deuterium and tritium (tritium is hydrogen with two neutrons) ice is dropped through a target zone, where it is illuminated from many different directions by a very intense pulse of laser light. The photons are all either reflected or absorbed—either way, they give the deuterium and tritium atoms a kick toward the center of the target area. How hard a kick? The nuclei end up moving at about 30 million meters per second.
Right at the center of the target, the pressure is large enough to initiate fusion. Once that begins, the center of the pellet begins expanding, creating a compressed shell that also begins to fuse. Ideally, the chain reaction proceeds outward to completely burn away the deuterium-tritium pellet.
But a complete burn is usually prevented by the lasers that initiate the fusion process. There are two critical issues. First, the electrons are stripped away from the nuclei and leave the area. In doing so, they carry away vital energy, reducing the temperature and the initial pressure. This slows down the fusion process, allowing the pressure to drop and preventing the expanding shell of fusion from achieving a complete burn.
The second issue is more technical. The pressure that drives the initial compression needs to be evenly applied—the laser pulses all need to have exactly the same energy, same spatial beam profile, and arrive at the target at the same time. If they’re not, much of the deuterium and tritium sprays out of the pellet and never undergoes fusion.
A magnetic field makes everything better
For many years, fusion scientists had thought that if a magnetic field were used to compress the target, then a more complete fusion burn might be possible. The basic idea is that the role of the lasers changes. Instead of being responsible for compressing the pellet, it is only required to pre-heat the deuterium and tritium. Then, before the pellet explodes, the magnetic field is turned on, compressing it and initiating fusion.
The magnetic field acts on all charged particles, so it confines both electrons and nuclei, keeping the energy within the pellet. Furthermore, because everything is confined, the speed at which the nuclei need to be moving is reduced to just 1 million meters per second. If you think that isn’t significant, consider that energy is proportional to the square of speed, so we are talking about requiring a thousand times less energy to initiate fusion.
But the magnetic field itself uses energy, and early calculations showed that it might slow down the expansion of the burn shell, which would also result in an incomplete burn. It would help—the total gain in energy production from magnetically confined inertial fusion was predicted to be a factor of ten. But we need gains on the order of a factor of 50 to make fusion break even. So the entire idea seemed destined for the scrap heap.
This is where this latest bit of research comes in. Slutz and Vesey from Sandia National Laboratories have shown that, if you modify the structure of the pellet, then energy gains between 200 and 1,000 are possible. The major finding is that the pellet and initial heating stage need to be modified. Slutz and Vesy start with a fairly standard pellet: a cylindrical piece of cryogenically cooled deuterium/tritium, surrounded by either aluminum or beryllium (this is the conductor that the magnetic field acts on).
The pellet is fabricated so that the density of the ice is very high just inside the metal shell. And, it seems (though the authors never explicitly say) that the whole cylinder is large enough in diameter so that the only the center of the pellet is heated by the incoming laser beams. The laser beams themselves don’t hit it from every direction, but only along the axis of the cylinder.
Where the laser meets the hydrogen
The laser pulse heats the material at the very center of the pellet, creating a gas in that location. Before the outside of the pellet can heat up, the magnetic field is turned on, crushing the metal liner and compressing the gas. Fusion initiates, and the expanding shell of fusing material runs right into the layer of dense ice, slamming it into the shell before it can escape outwards. The result is a nearly complete burn.
The researchers calculated the amount of current and the duration of the current pulse required to produce the magnetic fields, and the numbers they came up with are not unreasonable (50-70MA for ~100ns). They also looked into the fabrication of the pellet. One critical issue is the smoothness of the inner shell of the surrounding metal layer. They show that they require the surface to be perfect to within about 20nm, while current technology routinely manages 30nm. The additional precision should be feasible with current technology.
Where I think the authors may have missed the mark is earlier in their calculations. It seems that they require the laser beam to create a gas with a relatively sharp boundary so that the shell of dense ice is left untouched, even as the interior is vaporized. It is unclear from the paper if they calculate the heating stage explicitly or not. I believe they do, but that leaves unanswered questions about how it’s done.
On a computer, it is very easy to create marvelous laser beams with very narrow effects. But in the laboratory, laser beams have strict limitations. Intensities change relatively smoothly, meaning that there is no sharp boundary between where the laser is heating material and where it is not. In addition, laser beams change diameter as they propagate, so the diameter of the heated zone compared to the unheated zone will change depending on where the pellet is hit by the laser.
From what is in the paper, it is hard to say if the initial conditions required for a good burn can be met with a laser. What this really calls for, of course, is a huge experiment where people like me get to blow stuff up.
Physical Review Letters, 2012, DOI: 10.1103/PhysRevLett.108.025003