George Janes

In 1969, Richard Levy and I were co-leaders of an AEC-funded project whose objective was to produce highly stripped heavy ions for nuclear accelerators. This would have provided a way of increasing accelerator output energies at lower cost, and was of interest to the academic accelerator community. Through our contacts with Washington, we became aware that our contract would be terminating due to funding cutbacks. In the course of thinking about other directions to take, Levy and I decided that isotope separation by photochemical or photo physical means might, in the age of lasers, finally be possible AVCO Everett Research Laboratory resembles academia in that groups generate their own proposal ideas which they then try to sell to government or other organizations. The principal difference is that the research is more goals oriented. There is also more support and more critical discussion among peers in order to help cut down on unpromising projects.

Our Laboratory had already been making dye lasers at this time. These were of the “Hansch” variety, with pulsed nitrogen lasers as pump sources. With outputs of only a few millijoules, these were research-scale devices, good for chemical or physical experiments or for monitoring processes, but not for industrial processing. As Levy and I worked up our ideas on uranium isotope separation, we were in close communication with Arthur Kantrowitz. Initially, the three of us were very worried about the risk of nuclear proliferation, and we restricted access to our ideas rather tightly by requiring a kind of internal “need to know” although this kind of internal secrecy was not generally favored as laboratory policy. As time went on, however, we decided that this was not a garage technology, but rather one with a lot of tricky aspects.

When we first submitted our patent application to the Patent Office we stamped it “SECRET” because of our proliferation concern. However, the Patent Office questioned this and these markings were removed. The patent was eventually issued as an open document. We set for ourselves the goal of producing 3% enriched uranium in order to provide cheap reactor fuel.

This was a time of worry about energy shortages and sharply rising electricity use. Gaseous diffusion plants of the day took uranium with 0.7% concentration of U235 and enriched it to 3%, but left a tail stream with .2-.3%. We saw this as a waste. There was a large accumulation of such waste) counting in the waste from weapons grade uranium production, and we proposed to turn all this into light water reactor fuel. It was an exciting goal and one which we saw as responsible. At this laboratory, we not only had the liberty to mark out R&D goals but also to think creatively about correct technological policy, within the realities marked out by the demands of society.

Even as we worked out the details of our invention, the government set up its own study committee which decided that laser isotope separation (LIS) was not technically feasible. We wondered, however, about their objectivity, considering the enormous capital which the government had previously committed to gaseous diffusion plants. (These diffusion plants consume 10% of the electricity one gets back from the fuel they produce, and thus are very inefficient.)

In our collaboration, Levy was the theorist, with a strong background in aerodynamics and mathematics. He was also the one who took the lead with business plans He was a disaster in the laboratory, however. I was the experimentalist and inventor. In 1969-‘70-‘71, primarily through Levy’s efforts, we identified Jersey Nuclear (now Exxon Nuclear) as an appropriate partner for this effort. Ray Dickeman, then President, was willing to put up funding.

We started experimental work in March 1971 on the strength of a handshake between Dickeman and Kantrowitz, a simple verbal agreement. We started from scratch in March, and on July 14, 1971, we succeeded in proving enrichment on a laboratory scale at low atomic densities. We choose the simplest possible small scale experimental set-up for this initial proof-of-principle activity.

The laboratory team was also small. At first it was only Irving Itzkan and I with one or two technicians. Later Charles Pike and I worked at it and Larry Levin joined us in the summer. Jersey Nuclear initially funded us at about 1/2 million. The project was unclassified, but highly proprietary. We were mindful that the results were the property of Jersey Nuclear, and we did not publish, even though we were getting highly publishable results. We knew, then, by July 14, that we could use lasers with a two-color scheme to enrich. But we were working at particle densities of 10¹⁰ per cc. This was too low for industrial work, for which densities of 10¹² to 10¹³ are needed.

We now started to work simultaneously on the scientific questions that still needed elucidation, and on transforming our laboratory results into an industrial process. All this time, Dickeman was pushing us very hard. Some of our technical problems were these. In the two-step photoionization process, the cross section for exciting U235 was adequate, but that for ionizing into the continuum was very small. This step necessitated a very large flux of photons, because the flux required is 3 or 4 times the inverse cross section. In order to use these photons efficiently, we needed a total number of atoms comparable to the total number of photons. But since the density would be limited by collision phenomena, this dictated a length of many meters (the diameter. of the tube was of the order of a few centimeters.)

These requirements on photon flux and beam length were factors which indicated to us that LIS was a relatively low risk proliferation technology. We were convinced early on (and still believe) that the gas centrifuge was much more of a proliferation risk. In order to deal with these low ionization cross sections, we attempted to find better color combinations.

We primarily worked on 3 photon color schemes. This path was indicated because it was hard to obtain good efficiencies with lasers which gave out the blue light needed for the 2 photon color scheme. As we got into 3 photon 3 and 4 color photoionization schemes, we found ourselves in a regime where the uranium spectroscopy was not well known.

Another factor here was that there was a low-lying (620 cm1) atomic level near the ground level that was well populated. We wanted to access this level’s atomic population, and this dictated the use of at least one more color, hence the need for 4-colors. One technique which we developed depended on the fact that there are bound levels close to the ionization level with large cross sections. We could populate these atomic levels with visible photons and then use electrons or infrared radiation to carry the atom up into the ionized state. However, we found that it was still better to make use of a phenomenon known as “auto-ionization.”

According to this process, we gave the atoms enough energy to excite two electrons whereby, after a very short delay, the internally excited electron would relax and the total energy would go into an ionization process. Another of our technical problems was that of maintaining the requisite high optical beam quality, because the heat generated in the lasing medium distorted the wave fronts. A third problem was devising ways of collecting the ionized U235 atoms before they were lost through charge exchange or ordinary collisions.

Finding a practical collection process was also a major problem. This second phase of the program began in late 1971 and we were able to demonstrate enrichment at industrial process density in mid-1973. Jersey Nuclear AVCO Isotopes (JNAI) was formed in 1972. It still exists and holds a large patent portfolio.

I am, incidentally, Vice President for Research and a member of the Board of Directors. Scientific work continued to be done here at AVCO, but a group was formed at Exxon Nuclear in Richland, Washington, to do the engineering work. From the initial figure of less than 1 million per year, the budget rose to 3 million a year, and, by 1981, a total of 77 million had been spent (including money spent on engineering development). In 1974, Levy went to Exxon Nuclear in Washington and I took over the entire AVCO team. (Previously, I had headed the experimental work at AVCO.) Harold Forsen at Exxon Nuclear was overall program manager.

In order to conduct experiments at the scale and densities required for industry, we had to develop better lasers. In 1972 we launched an ongoing research effort under the leadership of Horace Furumoto, Mike Mack, and Henry Aldag. We developed two kinds of flash-lamp pumped dye lasers. One was a vortex type which used a flash lamp similar to a plasma discharge, and the other used a sealed-off lamp and reflectors. Both were successful but in 1976, we made the painful decision to discontinue work on the vortex lamp program for economic reasons.

One very vexing problem involved controlling the quality of the flash lamps which were made for us by EG&G, Xenon, and ILC. This problem was eventually brought under control through an “in-house” flash lamp program. Another problem was the degradation of the dye due to ultra-violet light.

This degradation causes the dye to absorb at the laser frequency and so kills the gain. It was eventually controlled with a small on-line reprocessing plant which filtered the solvent by reverse osmosis and added fresh dye. In the mid 1970’s, the center of gravity of our effort shifted to engineering. At this point, I had 12 PhD’s and an appropriate number of technicians and other supporting personnel. Our group did not shrink significantly, but the engineering group on the West Coast grew.

Unfortunately, by 1980, the business outlook for nuclear fuel was dramatically different (e.g., depressed) and Exxon Nuclear decided, therefore, not to continue to support the effort as a private project. At that time, we were actually ready to build a pilot plant and Exxon approached the government and suggested a joint venture whereby the government would pay for future work in return for royalties on the patents. In the summer of 1980, DOE gave us an initial 3-month contract. When this ran out, they encouraged us to continue on an exposure in anticipation of a further contract and we, therefore, kept the effort going, so as not to disperse our groups. We actually put $3 million into this work in the first 3 months of 1981.

Meanwhile, DOE directed its Energy Research Advisory Board to evaluate JNAI and the 3 other extant projects with the objective of selecting the most desirable means of attaining a reliable, less costly, National enrichment capability. These projects included Livermore’s atomic vapor laser isotope separation project, Los Alamos’ molecular process, and a process TRW was investigating that did not use lasers, as well as JNAI. The ERAB Committee did quite a thorough study and eventually recommended the atomic vapor process and further recommended that JNAI be the lead laboratory. However, this ruffled the feathers of Livermore and DOE Headquarters then generated a report internally which contradicted the ERAB conclusion.

The politics of this was very questionable and neither we nor ERAB were allowed to see this report. At this point (March 1981), Exxon pulled out, having concluded that it was not possible to do business with DOE because of the politics in effect. JNAI continued in existence in part to recoup the $3 million spent in anticipation of a government contract on top of its original investment of $77 million.

The Washington state Exxon group was disbanded entirely, and we at AVCO picked up their lasers. My group turned to the task of identifying new applications for the big flash lamp pumped lasers (mainly military) and that is what we are pursuing today.