Sonoluminescent Fusion Confirmed
According to a press release today, researchers have confirmed nuclear fusion resulting from sonoluminescence.
From a theoretical perspective, this is way cool. No, cooler than that. This is, like, totally cool.
The real question is whether it will be a practical route to fusion-based power generation. If so, then this could provide vast amounts of power using relatively inexpensive hardware in a compact space. Just what the world needs. No CO2 emissions, either (though waste tritium could become a problem, but tritium has a relatively short half-life compared to fission byproducts, and there's less of it, too).
In order to become a practical power source, this technique will have to overcome several hurdles:
1. Power Extraction: Generating fusion is one thing, but figuring out a way to harness the resulting energy production is something else entirely. As far as I know, there hasn't been any research into where the released energy goes, but my guess is mostly light (gamma rays) and neutron emissions, and some heat, too. The neutrons (unless absorbed) will effectively decay into hydrogen atoms in a matter of minutes, releasing more gamma radiation and neutrinos. Capturing energy from the neutrinos is pretty much hopeless, so forget that.
It isn't clear (to me) that enough of the released energy goes into heating the liquid to make a typical heat engine useful. But that might work.
The only other real alternative is capturing the gamma radiation and turning it into electricity. I don't know of any commercially useful process to do that, but it might work.
2. Excess Power: Once you've figured out a way to extract power from the fusion reaction, you have to get to the point where you're extracting more energy than you put in. Fortunately, sonoluminescence isn't a process which takes much power to begin with: just a couple of speakers and some circuitry to drive them. In this respect, the technology starts off with a huge advantage over traditional "hot fusion" approaches.
It isn't clear (and I doubt anyone has done the research, though it shouldn't be hard to calculate) if the current sonoluminescent reactors produce more energy from the fusion reaction than it takes to drive the reaction. Fortunately, fusion reactions release so much energy that it wouldn't take much to give excess energy production, and even a relatively inefficient power extraction mechanism may be able to produce excess power.
3. Scalability: The final hurdle is making a reactor which produces excess power and can scale to large sizes. Current reactors are the size of lab beakers, and may (at best) produce a few watts of power. A fusion reactor which gives you as much juice as a 9-volt battery is an interesting parlor trick, but not a useful energy source.
The unknowns here are just too great to make a guess if the technology is scalable, though there are some causes for optimism. First, nuclear fusion has been confirmed using a small apparatus and relatively crude methods. As with any new technology, we can expect that many orders of magnitude better performance is obtainable with enough refinement. Second, there are intuitive reasons to hope that the reaction rate might scale nonlinearly with the amplitude of the oscillations and/or the size of the container (why? Because at the low end of the curve, nuclear reaction rates grow very fast as the temperature and density increase). If this is true, then big reactors could produce prodigious amounts of power.
At the end of the day, this is very new and unexplored ground. But this is probably one of the most intriguing things to come out of pure physics research in a while.