LPP Fusion’s President and Chief Scientist Eric Lerner reported on June 21 new record ion energies of over 260 keV (equivalent to a temperature of over 2.8 billion degrees K) to 150 plasma scientists assembled in Prague, Czech Republic for the 27th International Symposium on Plasma Physics and Technology. The new results, obtained with the FF-1 plasma focus experimental device in Middlesex, NJ were a 50% advance over the previous record for a single shot, 170 keV, also achieved at FF-1 in 2011. Equally significantly, the mean ion energy for 10 shots at the same conditions also increased by 50% to 124 keV. Combined with other advances reported at the same conference (see next section) these results mean that FF-1 now has achieved the ion energy needed to ignite hydrogen-boron fuel in an average shot, not just in the best shots.
In addition, Lerner reported that in the same 10 shots, the variability in fusion yield from shot to shot was only about 14%, a factor of four reduction over previous results with FF-1. Researchers were impressed with this result, as the plasma focus device has consistently been hindered by large shot-to shot variability, especially at high peak currents. Mean yields were also 50% higher than in the best 10 shots with copper electrodes.
Lerner emphasized that these new results were possible only with the glow-discharge preionization used in the May-June experiments. This preionization, caused by a tiny, several-microampere current flowing in advance of each shot, smoothes the path for the main current, making breakdowns more symmetric and reducing or eliminating the vaporization of the anode material. “We see evidence of the reduction of vaporization from the reduction in the oscillations of the current,” Lerner explained (see Fig. 1). “This indicates that less energy is being drawn from the circuit to vaporize and then to ionize tungsten atoms.”
The more symmetric current sheath in turn leads to the elimination of the “early beam” phenomenon, when the current sheath splits in two during the compression of the plasma, robbing energy from the plasmoid (Fig. 2). Just moving to the monolithic tungsten electrode alone considerably reduced the early beam, which LPP Fusion researchers first identified as a problem back in 2010. This is likely due to the elimination of arcing between parts of the electrodes, since there are no such parts in the single-piece tungsten electrodes. But preionization completely eliminated the early beam.
This then leads to more energy in the plasmoid, as measured by a higher peak in the voltage at the time of the pinch—the formation of the plasmoid (Fig. 3). Higher plasmoid energy finally results in higher ion energies. As well, reduction of the asymmetries due to vaporization leads to reduced variability in yield.
Lerner pointed out that although a record yield of 0.25 J was possible just with the new monolithic electrodes (as reported in the May LPPFusion report), it took preionization to get the reduced variability and the record ion energy. The preionization success was truly a team effort. Research Physicist Dr. Syed Hassan suggested switching to the more even glow discharge. The team found that stabilizing the glow discharge was not possible using a current coming from a shunt resistor tied to the charging of the main capacitors. Chief Information Officer Ivy Karamitsos had earlier suggested separating the preionization and charging processes. When a separate high voltage power supply was used for the preionization, Electrical Engineer Fred van Roessel devised the circuit needed to protect the power supply from the current spike when the device fires.
Despite the progress reported, Lerner emphasized that much remains to be done. Oxides are still present in the device due to the introduction of water by a leaky valve and, unlike in the first 30 shots, are now declining very slowly, preventing further gains in yield. Impurities overall have only been reduced by about one third compared with last year’s experiments, so yield is still far below where it would be theoretically, with no impurities. In addition, there is no evidence yet of increases in the density of the plasmoids, nor of improved fusion performance with the deuterium-nitrogen mix (although 5% nitrogen is needed to stabilize the preionization discharge).
The next step is to use an ultrafast ICCD camera to get images of the area near the insulator where erosion has occurred, to see if vaporization has been eliminated or merely reduced, and to see the details of the process. A new reassembly of the device will almost certainly be needed to really eliminate oxides. Silver plating can be used to avoid tungsten’s affinity for oxygen (oxygen is bound very weakly to silver). In addition, by September, new beryllium anodes will be delivered. While beryllium lacks tungsten’s high melting and boiling points, for a given amount of energy, 15 times less beryllium than tungsten will be evaporated and each microgram of beryllium will have 17 times less effect on the plasma, due to beryllium’s far lower atomic charge. So, one way or the other, the impurity problem will be overcome. (Lerner’s full presentation is available here.)
“This was an extremely productive conference,” commented Lerner, “and I learned a lot from my colleagues. A plasma chemist offered to help interpret our optical spectra. Presentations explained how evaporation of electrodes, a widespread problem, is affected by surface acoustic waves—intense heat making the electrodes bounce—and by very thin plasma sheath layers that can re-accelerate even slow electrons. The problem of defeating erosion is complex—but it can be done. Another researcher explained how, in automotive tungsten arc lamps, lifetimes of 10,000 hours were achieved with current densities of 10 GA/cm2. That’s at least ten times the minimum lifetime and 10 times the current density that focus fusion electrodes will require for a working generator. But it may take the engineering phase of our project to get there.”