As was previously mentioned, the actual work on continuous fluorination was surprising small during the Molten-Salt Reactor Program (MSRP) that ran from 1957 until 1976.

Fuel reconstitution was mentioned on pages 199-201 of ORNL-3936 in June 1966. Continuous fluorination was mentioned on page 202. They described an experiment using a 1-inch diameter nickel column with a salt depth of 48 inches. 15 cc/min of NaF-LiF-ZrF4 salt containing 0.5 wt% UF4 was contacted with 70 cc/min of F2 during 1 hour of continuous operation. Material balances were complicated by corrosion of the nickel vessel.

In ORNL-4037 (January 1967) continuous fluorination was again described on pages 232-233. A 1-inch nickel fluorination column, 72 inches long, was used to contact molten salt flowing at rates of 3-50 cc/min with salt depths of 12 to 54 inches. The offgas from the fluorinator flowed through a NaF bed where UF6 was absorbed and then onto a soda-lime bed for F2 disposal. They report that in their best run a molten salt flow rate of 20.7 cc/min and an F2 feed rate of 250 cc/min were maintained for 2.5 hours. The salt had a uranium concentration of 0.32 wt% and the fluorinator was operated at a temperature of 650C. The uranium discharged from the fluorinator at a uranium concentration of 0.0020 wt%, with 99.4% of uranium removed by fluorination.

In ORNL-4119 (July 1967), previous results were recapitulated on pages 204-206. No salt containing beryllium had been tested up until that time due to a lack of suitable facilities. Additional tests had been conducted, achieving efficiencies of up to 99.89%. In all tests, however, F2 utilization rates were low, ranging from 15 to 30%. A “protected” system was described as under construction that would be 5 inches in diameter and have a frozen wall section six feet high. It would operate on LiF-ZrF4 salt.

ORNL-4254 (August 1968) reported on progress in continuous fluorination on pages 252-254. Frozen wall thickness of 0.3 to 0.8 inches had been achieved and the thermal conductivity of the frozen wall had been measured. The introduction of F2 was mentioned as a problem, with a 45 degree inlet pipe as a potential solution. Argon gas was substituted for fluorine gas.

ORNL-4344 (February 1969) reported on continuous fluorination on pages 302-305. Most of the work reported concerned the frozen wall layer.

Advancements in frozen-wall formation were reported on in ORNL-4548 (August 1970) on pages 309-311. Radiofrequency (RF) induction heating appeared to provide a suitable heat source for experimental work on fluorinator development. The molten salt would be heated by eddy currents generated by an alternating magnetic field, which would be generated by a coil not in contact with the molten salt. But the metal walls of the fluorination column would also be heated as well, which was a disadvantage. Configurations where the induction coils were inside the frozen salt film or inside a frozen salt tube at the center of the column were considered. This work was again described in ORNL-4622 (January 1971) on pages 219-220. More work was described in ORNL-4676 (August 1971) on pages 262-266. This also described work on the estimated corrosion rates in continuous fluorinators and the axial dispersion coefficient in continuous fluorinators.

In ORNL-4728 (February 1972) further development of the induction-heated experiment was reported on pages 214-217. The predicted performance of continuous fluorinators was described on pages 217-220.

ORNL-4812 (August 1972) was a status report of technology development. Chapter 11 was about Fuel Processing. The status of continuous fluorination and fuel reconstitution is discussed on pages 350-351. They reference the studies done in the 1-in-diam, 72-in-long nickel fluorinator using NaF-LiF-ZrF4 containing UF4. Removal fractions were 97.5 to 99.9%. The 5-in-diam, 96-in-long simulated fluorinator using LiF-ZrF4 and argon was mentioned in connection with frozen-wall development. A future facility is mentioned.

In ORNL-4782 (October 1972) reconstitution chemistry was examined on pages 212-215. A container that was inert to UF5 was sought and initially gold and graphite were considered suitable, but under further examination graphite did not appear suitable, since CF4 and C2F6 were detected. Gold appeared suitable and less than 200 wt ppm was detected in the salt. Induction heating of salt for fluorinator testing was described on pages 230-234. Frozen salt layers were formed but they were unlike the smooth layers previously noted. They had dendrites up to 3/4″ long.

In ORNL-4832 (March 1973) on pages 180-181 a final set of experiments were described before design of a facility for studying continuous fluorination of molten-salt in a frozen-wall system was to be initiated.

In ORNL-5018 (December 1974) a programmatic plan for the future development of MSR technology is laid out. Section 3 concerns fuel processing. Continuous fluorinator development occupies pages 3-25 through 3-34. Initial work is in the development of autoresistance heating for simulating FP heating. Later work is for facility design for further experiments. Both a “Continuous Fluorinator Experimental Facility” and a “Fluorination-Reconstitution Engineering Facility” are mentioned. The CFEF would be installed in bldg 4505, and the FREF would be installed in bldg 7503. The CFEF would operate at flow rates half that of an MSBR. The FREF would also operate at flow rates half of an MSBR. Funding requested ranged from $1.0-2.0M/yr in inflation adjusted funds. Most would be spent on these facilities. Fuel reconstitution occupies pages 3-34 through 3-40. They requested funding of $700K – $1200K/yr for this task, inflation adjusted.

In ORNL-5011 (June 1975) the chemistry of fluorination and fuel reconstitution was described on pages 100-101. More heating tests were described on pages 122-127. The design of the fuel reconstitution experiment was on pages 127-129. On pages 129-132, the performance of open bubble columns, like fluorinators and reconstitutors, was described.

In ORNL-5047 (September 1975) the chemistry of fluorination and fuel reconstitution was described on pages 150-151. LiF-BeF2-ThF4 carrier salt containing 1 to 4 wt% UF4 was mixed with the stoichiometric amount of UF6 to convert all uranium to UF5. Then that salt was sparged with hydrogen to reduce the UF5 to UF4. The reducing gas was either pure hydrogen or an argon-hydrogen mixture. In all tests the hydrogen utilization was low (<1%) and most of the gas simply bubbled through the melt without reacting. 500 mg of platinum-black catalyst was added and hydrogen utilization jumped to 90 to 100%. To insure that the catalyst was not promoting the reduction of UF5 to UF4 the experiment was repeated without hydrogen, but no UF5 reduction took place. Hydrogen was reintroduced and the reduction was complete in 30 minutes. A gold lining had been used on the reaction vessel and it appeared that some of the platinum had alloyed with or sintered to the liner. This material appears to exhibit the same catalytic activity as the platinum black. Pages 162-165 talked about continuous fluorinators. Pages 165-174 discussed experiment design. Pages 174-182 went through mass transfer in open bubble columns. In ORNL-5078 (February 1976), continuous fluorinator development was discussed on pages 152-157 and experiment designs were discussed on pages 157-161. The final MSRP semi-annual report, ORNL-5132 (August 1976), reported again on the chemistry of fuel salt reconstitution on pages 170-171. Smooth platinum sheet was investigated for its catalytic ability and found to be sufficient. This led to modifications of a test that they planned to conduct. Since several species of fission products would also be removed from the fuel salt by fluorination and travel with the UF6 to the reconstitution step, the effect of these fission products was investigated. Niobium in particular was examined and its reduction proceeded very slowly. They hypothesized that the NbF4 behaved very similar with respect to reduction as UF5. Then they investigated reduction again using the platinum catalyst. Results were initially improved but then fell to a value similar to the uncatalyzed condition. They feared that niobium might have poisoned the platinum catalyst, which could have dangerous implications for the practical use of platinum-catalyzed reconstitution. The work done during the MSRP laid a very strong foundation for the investigation of continuous fluorination and reconstitution, but it was by no means exhaustive nor it reach its natural conclusion. The work ended simply because the funding ended, with many unanswered questions. Continuous fluorination in LiF-BeF2 salt, for instance, was never undertaken. There was no investigation of alternative fluorination agents, such as NF3 or BrF5. The aggressive nature of F2 caused them to expend a great deal of resources on the development of a frozen-wall fluorinator, but that same device was very very difficult to test in a non-radioactive environment. It relied on the heating effect of fresh fission products that simply weren’t present in the test fluids, and whose effect had to be simulated by external heating. That external heating also turned out to be very challenging, if one wanted an accurate simulation. Fuel reconstitution was investigated even less than continuous fluorination and even the small amount of work that was done uncovered some real surprises. It appeared that some type of catalyst was necessary to promote reduction of UF5 to UF4. Platinum appeared to work very well but there were indication that fission-product niobium might poison the catalyst. More tests were needed to get to the bottom of these problems but these tests never happened because the funding was cut off. And unfortunately, that is where things remained for over forty years. But now things are about to change…