As the research team had suspected, an oxide coating on the tungsten electrodes prevented the achievement in the first shots in June of the very low impurity levels needed for increased fusion yield. This oxide coating has now been substantially reduced or eliminated and the electrodes are ready for a new round of experiments.
When the team opened up the vacuum chamber to inspect the electrodes in mid-July, they immediately noticed a widespread coating of a gold-colored material, mixed with a purple color. (see Figure 1) Research Associate Clifton Whittaker rapidly found in the literature that this material was tungsten bronze, a compound of tungsten, oxygen and hydrogen. X-ray analysis of the coating on a steel plate confirmed that it was mostly tungsten. The bronze formed when a layer of tungsten oxide on the electrodes reacted with the deuterium fill gas. Deuterium is an isotope of hydrogen, with an extra neutron in the nucleus, but it is chemically identical to ordinary hydrogen. The reaction occurs at only 400 C and causes the release of tungsten to the plasma. In contrast, metallic tungsten does not react with hydrogen and must be vaporized at 5500 C to enter the plasma.
After consultation with the electrode manufacturer and further literature search, the team learned that tungsten oxide forms at high temperature in two separate layers—one thick and easily removed and the other a few microns thick, tightly bound and basically invisible. While we had carefully removed the outer layer, the second layer remained and created the tungsten bronze—and the tungsten impurities in the plasmas.
The tungsten oxide was also a possible cause of the breakage of the ceramic insulator that separates the anode and cathode. This suffered a puncture after only a dozen shots. This appears to have been due to a particle of tungsten oxide on the insulator suddenly reacting with the deuterium, releasing tungsten and concentrating the electric field to a level that could break down the insulator.
A second problem was uncovered by our inspection. The electron beam generated by the plasmoids was heavily vaporizing the end of the hole at the tip of the anode. Vaporized material—mostly tungsten oxide again—hit the bottom of the of the vacuum chamber and bounced off, blowing back material onto the anode and cathode. This material added to the irregular distributed impurities entering the plasma, further interfering with a strong, symmetrical compression.
We’ve taken a number of steps to solve the problems we found. First, Whittaker carefully diamond polished by hand the entire anode and cathode surface to remove the oxides and minor roughness caused when the insulator broke. We’re confident that removed the oxides from the anode, but with the cathode’s complex surface the research team decided, through consultation with chemists, to put the cathode in a six-hour long ammonia bath at 80 C (which we had to do outdoors!). This chemical reaction should have removed the last of the oxides for the cathode. (see Figure 2)
We replaced the insulator with a new one, and ordered two more of higher purity 99.9% alumina, which increases by 50% their resistance to breakdown. In addition, we are going to closely observe the insulator and electrodes with a long distance magnifier and we will test the insulator for intactness electrically before each shot.
To reduce the blow-back, we have widened the hole at the bottom of the chamber that leads to the meter-long drift tube, so that any reaming debris will travel far from the electrodes. In addition, we are working out theoretically ways to transfer more of the energy from the electron beam to the heating of plasmoid, leaving less available to damage the anode. This work involves mixing in heavier gases and is still under way. We’ll report more on it next month. Reassembly of the cleaned electrodes is now almost complete, so we expect new experiments in early September.
