Патент USA US3041146код для вставки
June 26, 1962 W, H, HEDLEY ETAL 3,041,136 FLAME DENITRATION AND REDUCTION OF URANIUM NITRATE TO URANIUM DIOXIDE Filed April 8, 1960 ' 4 l 2 Sheets-Sheet l June 26, 1962- w.'H. HEDLr-:Y ETAL FLAME DENITRATION AND REDUCTION OF URANIUM , Filed April 8, 1960 3,041,136 NITRATE TO URANIUM DIOXIDE 2 Sheets-»Sheet 2 ` INVENTORS ¿Z50/avg United States Patent O "ice 2 1 ~ 3,041,136 FLAME DENITRATIQN AND REDUCTION 0F URANIUM NHTRATE T0 URANIUM DÍOXIDE William H. Hedley, Kirkwood, and Robert I. Roehrs, St. Louis, Mo., and Courtland MJHenderson, Xenia, Ohio, assignors, by mesne assignments, to the United States of America as represented by the United States Atomic Energy Commission Filed Apr. 8, 1960, Ser. No. 21,071 5 Claims. (Cl. 23-14.5) The present invention `deals with a process for con 3,4l,i36 Patented .lune 26, 19562 FIGURE 1 is a diagrammatic illustration of the ap paratus; and FIGURE 2 is a vertical sectional view of the reaction chamber proper. Referring to FIGURE 1, the reaction chamber in which the dehydration, denitrification and reduction takes place is designated 101. A storage vessel for the aqueous uranyl nitrate is shown as 103. Positioned just below the reaction chamber 101 is the receiver 105. . Cyclones 107 10 and 109, to remove solid products from the gases, are connected to the receiver 105 by pipes 106 and 108. A verting aqueous uranyl nitrate to uranium `dioxide and filter 111 is connected to cyclone 109 by pipeline 110. hydrates and their solutions. atmosphere by pipe 116. A gas cooler 113 is connected to the outlet of the filter an apparatus therefor. This process will hereafter be 111. The outlet of the gas cooler 113 is connected to referred to as the “flame process.” The term “aqueous uranyl nitrate” will -be used to cover both uranyl nitrate 15 an absorber column 115 which communicates also to the It is an object of the invention to accomplish the denitrification and reduction of aqueous uranyl nitrate in a single step, decreasing the costs of the operation sub stantially. A rotary seal 119 communicates between the receiver 105 and a hopper 117. Similar rotary seals 121 and 123 connect the solid outlets of cyclones 107 and 109 to hop 20 per 117. A boil down tank 125 is connected to transfer pump 127, thence to storage vessel 103. In operation, a reducing flame is produced in the re action chamber 101 by the incomplete combustion of a uranium fluoride in that it reacts rapidly and completely hydrocarbon gas such as propane. To obtain the reduc with hydrogen fluoride to produce uranium tetrafluoride of very high purity. This tetrailuoride is a very important 25 ing flame propane is burned in a deficiency of air. A flame in which there is supplied less than 70% of the intermediate material for the production of uranium theoretically required air to burn the propane is Well metal by reaction with calcium or magnesium. It is also suited to the process. Aqueous uranyl nitrate is admitted a very useful intermediate in the production of uranium to the reaction chamber ‘101 from storage vessel 103. hexafluoride. This is the- compound of uranium utilized in the diffusion process of enrichment in the isotope 30 Some of the U02 produced drops into receiver 105; fur ther amounts are separated from combustion gases by Uz35_ cyclones 107 and 109. The last traces of U02 are re It is also an object of this invention to produce uranium moved by the ñlter 111. The combustion gas then is dioxide in a form that can be stored for long periods of cooled by the cooler 113 and nitric oxides absorbed in time with substantially no deterioration by conversion to higher oxides through mere contact with air at ordinary 35 the absorber column 115 before the combustion gas ex hausts to the atmosphere. temperatures. This occurrence would require further Referring to FIG. 2, a tubular shell 201 is 6 feet long reduction before conversion could be made to the fluo and 34 inches in diameter. At the top of this shell an ride. Otherwise there would be contamination with oxyupper flange 203 is welded extending radially outward. fluoride and other undesirable compounds. Finally, it is an object of the invention to furnish an 40 Lower flange 205 is welded to the bottom of shell 201, extending radially outward and inward therefrom. The apparatus capable of converting aqueous uranyl nitrate to outer periphery of shell 201 is covered with pipe insula uranium dioxide smoothly, completely and continuously. tion 207. Inside of shell 201 are, in order, a first insula The present practice for converting aqueous uranyl nitrate to uranium dioxide normally requires three steps. 45 tor layer 209, consisting of a layer of a few inches of a material of extra low thermal conductivity, such as high> The usual starting material is an aqueous solution, so the temperature service mineral wool Iblocks, a second insula first step involves driving off sufficient water to form the tor layer 211 consisting of a few inches of a material of molten hydrate. Next, sufficient heat is supplied to drive low thermal conductivity such as ñre brick or alumina off the water of hydration and to decompose the uranium bubbles, and an inner refractory layer 213, made up of 6 nitrate and obtain uranium trioxide. Finally, reduction rings of silicon carbide, one on top of the other in longi t0 uranium dioxide is accomplished by passing either tudinal alignment, each ring having one edge concavely hydrogen or cracked ammonia through the heated tri rounded and one edge convexly rounded. This construc oxide. It is usual to perform each step in different equip~ tion is employed to retain alignment. All insulating ment. e layers and the pipe insulation are supported by lower The denitrification step may take place in a fluidized 55 flange 205. 'I'he inner layer 213 defines a reaction bed or in a batch reaction vessel. Further, the hydrated chamber 213a which is open at its bottom to and com salt may be passed onto a bed of trioxide previously It is a further object of the invention to produce a unanium dioxide very well suited for conversion to formed by a screw or other mechanical device. The re municates with receiver 105 (see FIG. 1). A cylindrical flame throat 301 is located within the upper portion of duction also may be carried out either in a fluidized bed, chamber 213a and is provided with a ñange 301a extend in a mechanically agitated bed, or in a moving bed. It 60 ing outwardly therefrom near the top of the throat. is also possible to carry out the reaction lbatchwise under Throat 301 is coaxial with shell 201 and is formed from static conditions using thin layers of the trioxide in a graphite. The top of the flange 301e is flush with the reaction vessel. top of the shell 201. This throat 301 is secured to and The invention herein described can accomplish all three supported from an annular plate 303 which rests on and steps in a single operation, but for economic reasons 65 is fastened to upper flange 203 by bolts 305 and extends it is preferable to concentrate a uranium solution to the from throat piece 301 to the edge of insulation 207. A, molten hydrate stage before the process of this invention smaller annular plate 307 rests on the larger plate 303 is applied. Therefore, the preferred embodiment of this invention replaces only two steps of the usual process. and is flush with the top of throat piece 301. A burner assembly 309 comprises a double-walled The >equipment for one workable means for effecting 70 metal cylinder 313 which` has a lower flange 315 extend this one step reaction and the flow sheet are shown in the ing outwardly from the base thereof, and has an air inlet line 317 leading into the volume between the walls of following drawings wherein: ` 3,041,136 - onrls` lf the llame is not established within that time, the solenoid valve closes and can not be reopened for ap proximately 30 seconds. - However, if the flame is estab shell 313; and a gas inlet line 31.9 passing through both l walls of shell 3ll3.l The burner assembly`3ll9 .is secured to plates 303 and 307 `by boltsl 311 passing through ñange‘ lished the detection system is engagedl and holds lthe, feed 315. Both walls of shell 323 curve inwardly at the top. The outer wall of shell 313 .is closed by cover 321. A 5 gas solenoid open as long as it detects a flame. An ex cess of propane is fed .to the unit in order to simul :llange 32la extends inwardly froml lthe inner wall of taneously denitrate and reduce the uranyl nitrate to U02.A shell 3213 a short distance below the point at which it The flow of combustion gases is then increased to a starts curving inwardly. Shell 313 is coaxial with shell desired level and the llame is maintained automatically. 201. The feed system is. constructed so'thateither water or A cylindrical body 323 having an annular recirculation aqeuous uranyl nitrate can be forced through the spray chamber‘325 therein and .surrounding a flame channel nozzle. Common practice is to »ilow water through the 341 conforms closely to the inner‘wall of shell 313 he system prior to introducing aqueous uranyl nitrate. tween flange 321a and the bottom of shell 313. Ports` There are two reasons for doing so. 327 pierce the top of body 323 and ports 329 pierce body 323 at its bottom.' lA funnel tube 331 has its top aligned with and con nected to the incurved end of the inner wall of shell 313 and forms a venturi 333 for the entrance of air from air inlet 317.l The funnel. tube' 331 forms with the upper incurving end of the inner‘wall'of the shell 313 and the . increase toa flow rate equal to the anticipated flow of the aqueous uranylnitrate, the unit will approach an ' equilibriuml condition and therefore the temperature in the unit will not drop when the nitrate is introduced. Aqueous uranyl nitrateis not allowed to .flow until the flange 32M a gas inlet'chamber 333e` which communi» . temperature of the exit gases at the bottom of the re cates with gas inletl line 3119. A conical ring 335 aligned with and connected to flange 321e forms with funnel actor‘has reached approximately 1800° F. This tem perature -is' experimentally,determined as being sufficient for denitration and reduction of :uranyl nitrate to U02. tube 331 an orifice 337 for the entrance lof gas from. linlet 319; Air‘ignition plug 339 passes through both It is measured with a-sheathed chromel-alumel thermo walls of chamber member 323 into ñame channel 341 i - and when activated ignites the air-gas mixture. . The couple. . . In this process, denitration .must take place almost'in- . continuance of the llame in channel. 341 is monitoredby ' stantly, .but reduction to U02 of the powder formed takes ' flame detector 343 essentially an ultraviolet sensitive de - . ï vice which penetrates into chamber 341. First, the Water .keeps the nozzle from becoming overheated during the preheat cycle. Second, if the flow of water is allowed to 1 place only as the particles travel down the .length of the The function of chamber 325 is to allowthecirculation of. the. hot com 1 reactor. This was demonstrated by some early pilot-scale `bustion products to heat the combustion gases fed into chamber 34d and thus facilitate their combustion. A runs with inadequate-reduction in which the product con-` sisted predominantly of nitrate free UaOß. `It is believed that there :are'four main factors which affect the con passes throughl cover 321 and ilame channel 341 to 35 version of‘aqueous uranyl nitrate to U02 bythe flame process. They are droplet size, gas temperature, rel enter the chamber'213a. Nozzle 3fl5‘is water cooled and ducing power of gases and residence time. Early ex insulated (details not shown). The burner .is therefore built around the spray nozzle and the feed is introduced \ periments` resulted in 'USOS as a product `when using. a lliquid droplet lsize of about 4t() microns or less, a gen along the burner axis. ‘ p ' l l l ' pneumatic atomizing liquid spray nozzle assembly 345 erous excess of propane, an exit gas temperature of 1800° . Propane, vaporized‘at 50 p.s.i.g.- enters the apparatus through inlet tube 319; air enters through its inlet 317, F. and a reaction chamber four inches in diameter by one foot long. Since the use of small particle size, ex cess propane and high gas temperature alone did not also at 50 p.s.i.g. The gases ignite and pass into the chamber 2‘13a through the area between chamber mem ber 323 and the spray nozzle assembly 345. Meanwhile flame detector 343 detects the flame and releases a liquid through spray 345. At first the liquid is ordinary water, produce U02, it was decided that increased residence time in the reactor would be tried. Another reaction `chamber three feet long and live inches in diameter was until a temperature in excess of 1500° F. is detected in therefore constructed. Operating with this longer cham the reaction chamber by a thermocouple (not shown). Then aqueous uranyl nitrate is admitted under pressure through a control valve (not shown). The droplets of ber at the same gas and the liquid flow rates used pre viously, it was possible to produce U02 routinely with exit gas temperatures of approximately l800° F. In the apparatus used introduction of the feed parallel to the aqueous uranyl nitrate enter the zone with the com bustion gases, and are dried, denitrated and converted to burner axis is also important. Trial runs were made em ploying nozzles which directed the spray into the flame U02 by the time they reach the bottom of the chamber. in a direction perpendicular to the flame. In these runs It has «been found that with the pressures used, a length 55 the water was not all evaporated before it hit the opposite of about 6 feet is sufficient to complete the reaction. wall. The above-described unit is capable of converting In order to prevent freezing of the aqueous urtanyl aqueous uranyl nitrate to the dioxide in a single step, nitrate in the lines, Ia dilute liquid containing only 4.3 combining dehydration, denitriñcation and reduction. It pounds of U/gallon (approximately 50 w/o uranyl ni is also capable of converting the molten hydrate to the dixoide in a single step. The properties of the uranium 60 trate) has been used as the -feed for most of the runs. However, three runs have been made using concentrated dioxide so produced are superior to those of the ordinary liquid containing 12.0 to 13.0 pounds of uranium per product, as will be shown later in the specification. The gallon (uranyl nitrate hexiahydrate contains 9.75 lbs. operation of the apparatus will now be described. Prior to establishing a llame, nitrogen or other inert gas is used to purge the reaction chamber. The com 65 U/gal.). This equipment and process has been found capable of using solutions of uranium nitrate containing up to 13 pounds of uranium per gallon. The preferred con button on the flame detection unit. This sends current centration is the highest concentration which can be used to a spark plug-type ignition rod located in the com without plugging the nozzle, because this lowers the pro bustion area and also energizes two solenoid valves, one in the propane line and one in the air line. Opening of 70 pane consumption per pound of uranium processed, and also the olf-gases from the process will be licher in by~ these valves allows a predetermined mixture of propane product nitrogen oxides which will make their recovery and air to ñow through the burner and ignite by means easier. The `apparatus described is able to handle 160 of the spark from the ignition rod. bustion mixture of gases is ignited -by depressing a start During ignition, the electronic network of the llame detection system is `bypassed for approximately 15 sec gal/hr. of the l0 lbs. of U/gal. liquid. Heating and 75 reduction of this quantity requires 380 pounds of pro» 3,041,136 5 6 pane »and 6880 cu. ift. of 1‘00 p.s.i.g. fair per hour. This TABLE III produces a product U02 having unexpectedly desirable Spectrographic Analysis for Trace Impimties in U02 properties. In evaluating the product obtained by the iiarne process, both chemical and physical properties have been rather Flame proc- -Fluid bed exhaustively investigated. ess (p.p.m.) (p.p.m.) (Run D-15) EXAMPLE A series of runs were made in the pilot scale apparatus with 4a 3 «foot chamber length and an exit gas tempera ture approximately 1800" P. During these runs aqueous ur-anyl nitrate was used in ia concentration ranging from U02 made by other processes. TABLE I ee NO3 H20 <0.1 <0. 1 <0. 03 <0. 6 U02 85.6 80. 9 15 eo <0.5 <0. 5 so 2o <10 <10 <10 <10 4 20 1o 2o 2o <2 <1 <2 2o 7c 3 <20 <1 <20 70 <50 (such as Cu and Zn) lare probably 'derived from com 95. 7 88. 1 Pponents `of the experimental iapparatus that require fur ther engineering design. 30 The llame process ‘U02 contains considerably less than the 100 ppm. iron and 75 ppm. nickel which are the lonly limits for green salt. Trouble is not anticipated with any of the trace elements in this process. In both cases efforts were made to cool the sample to room temperature before exposing it to ai-r. Because Reactivity with hydrogen fluoride-'liable IV shows the HF concentrations, reaction completion times, and reaction temperatures used for llaboratory hyidroñuorina tion evaluation of various `oxide samples. All data listed it is difficult to iget a sample lanalyzed without oxidation occurring, 95% free U02 is among the highest percent »ages ever analyzed. <1 15 <1 The differences in concentration of impurities in these 'I'hose elements that appear to be higher in the llame process material Percent Percent Percent Percent Flame process (batch D-22) ........ __ Mall‘mckrodt fluid bed product ____ ._ <10 <o.10 <0.] <1 r two samples lare small in most cases. [Chemical Analysis] U-H <10 <0.1o <o.10 <1 - nium. Several batches were selected for tests. Since a large number of tests were to lbe made and the tests were lengthy, the tests were not repeated on every batch. Comparisons were made in each case with available « <o.1 <10 4 <1 l5 4.2 to 13.0 lbs. uranium per gallon. Aqueous uranyl nitrate containing 12.9 lbs. U/gal. was smoothly con verted to \U02 iat the rate of 35.0 pounds of uranium'per hour, using 0.225 pound of propane per pound of ura Oxide source 4 10 U02 used to make UF4 in ia com in this section were obtained from runs in a thermo pletely enclosed, leakproof system would probably under b-alance. go less reoxidation. TABLE IV Stability of flame process 'U02 against atmospheric oxi dation-_Samples of llame process U02, in layers less Reactivity With Hydrogen Fluoride than 1/8 inch thick, were placed in open bottles and left exposed to the atmosphere for varying lengths of time. The results are shown in Table II. Hydroñuorination reaction 45 Oxide HF, w/o Comple- Reaction TABLE II Y Stability of Flame Process U02 Against Atmospheric Oxidation Percent Percent at room U+4 free U02 temp. ° C. 50 Flame process U02 ___________________ ._ Time in Weeks exposed to air tion time, mins. [Oxide Source: Flame Process Run 12] temperature Concen tration 55 t0 U02) ____________________________ __ 100 7 396~398 67 32 100 100 100 22 55 3G <0. 7 7 390-392 398 402-405 495 504-510 100 22 498-501 100 33 506-509 Fluid bed U03 (laboratory reduced to 2 5 11 83.3 83.2 81. 9 91.9 91. 7 89.6 U03) _______________________________ ._ As shown in Table IV, the reactivity of llame process 60 material with either anhydrous HF or 70% is ex ceptionally good. Times vfor complete reaction for sev eral runs with this material have ranged from 7 to ‘13 minutes using anhydrous HF at 400 C. The reactivity of the fluid bed U02 is considered to be an average value paratus Iduring original cooling. The oxidation which 65 for plant material run at 400° C. took place on standing was not excessive even under these The extreme reactivity of flame process material is very adverse conditions. In plant practice it would not 4greatly emphasized by noting its lower completion time be necessary to expose U02 -as thoroughly or for such with 70% HF when compared with that of fluid bed U02 long periods as this, and thus the amount yot” oxidation that was treated with «anhydrous HF. The 32 w/o HF which would take place under these circumstances should 70 run with llame process U02 shows that this material can be considerably less. utilize HF in almost any concentration, »and that it should U02 trace impurities-Samples of llame process U02 be compatible 'with the low excess HF process change vThe lamount of free »U02 in the samples at time zero` is uncertain, but is known to have been less than 95%. This is believed to fall short of a 100% U02 product by reason of incomplete protection in the experimental »ap and Weldon Spring plant fluid bed U02 were «analyzed spectrographically. The results of this analysis are listed below: ' being installed in the Weldon Spring plant, thus insuring the llowest possible raw materials cost for hydrotluorina 75 tion. 3,041,136 Green salt properties-_Samples of ñame process U02 TABLE VII and were hydroiiuorinated . `U02 from standard pot U03 . P article . . . Si. e Distribution b with 100 w/o HF at several diiïerent temperatures. The analyses of the green salt produced are listed in Table V. TABLE V Z 5 . Mzcromero ra h y t Oxide source g p Diameter equal to Orlcssthan l Green Salt Properties . Hydrolluorination temp. ° C. Flame Process AOI Peak Level U02 DCI‘CCIlÉ uns# percent UO from std. ot UO 2 p 3 (Run D 22) , AOI, WS., ““““ " In i1 fred---------- _' UH, 50% 5% Microns Micrims llíicrmis w » _ 15) """""""" ' 10 Fluid heli UH, percent 95% g1 ’ _ >100' 40' 2'5 _ >100 44 4.5 “““““““““““““““““““ " ' percent percent percent l This signifies, for example, that 50% of Run D-15 U02 has a particle 458 422 (l O8 3. 71 73“ 0 0. 83 2“ 77 73' s 535 510 530 5(7)5 0. 05 0.06 0. 00 3. 57 73.1 74. 0 73.2 0. 3s 0‘08 0. 2ï 2. 70 4_37 4.59 73.8 72. 5 2.1i size equal to or smaller than 2.3M. 6Go 647 0.21 L 35 7¿1_ 8 0.21 3. g5 i218 „ ' _ The above data'show that ñamehprocess~U02 has v_irtu ally the same particle size distribution as high fired micro» nized material. The fluid bed U02 and the high fired U02 730 722 1. 50 1.00 75.1 2.07 4. 20 12.9 have very nearly the same particle size distributions, but 755 755 T50 1' 50 ‘5-5 5-12 /4‘ ‘ both are much larger than those of micronized or flame 558 g 2.38 3.18 17-7 15 1 W.S. means water soluble and indicates UOgFs; AOI 20 process U02' The particle Slze dlstnbutlon of flame means ammonium oxalate insoluble and indicates unconvei-ted PYOCCSS U02 has been nearly Constant for au funs Checked, oxide. with approximately 99 W/o of the material having a di. Uranium tetratliioride is often called green salt. The analyses show that low AOl green salt can be proamcter between ‘1/2 and l() microns. duced on a laboratory scale from both materials at temThe average particle 4sizes of the same types of powder peratures up to 660° C. In the hydroiluorinations at tem- 25 as determined by Fisher Sub-Sieve Sizer are listed below: peratures of 72.2° C. and above, the AOI’s increased due to thermal damage of the oxide. The llame process U02, TABLE VIII however, was less sensitive to thermal damage than was Average Particle Size as Determined by Fisher Sub-Sieve the standard material used for comparison. This seems Sizer ' s . Avei‘a o article r@a so n able becau se powders compo S ed of small particle 30 Onde Source: me 1% 11121 icmns tend to be less susceptible to thermal damage. The inaterial chosen for comparison is standard pot U03 which has been laboratory reduced to U02. The indicated lower ` D 22 o 93 F‘É‘me Proces? (Run T ) ‘‘‘‘‘‘‘‘ ‘- 0 60 1'08 susceptibility of iiame process U02 to thermal damage Fluld b‘ed ------------ n could ‘be of real advantage in production by reducing the 35 hlgh ñled ----------------------- “ Mlcfûmzed hlgh med -------------- _“ number of lots of green salt rejected for high AOI. ' _2‘43 4'05 ’ The results of these particle size measurements substan . The U02 samples used in these runs contained rela- tiate the conclusions drawn from Micromerograph meas >tively large amounts of U2O2, probably due to partial re- memems, oxidation of U02 which had been exposed to the 'atmos- Densiti'es of cold pressed and síntered U02 compacts Phol'o before oomploto Cooling» The Presollœ of 11h15 Usos 40 Samples of several types of U02 were cold pressed to form accounts for the relatively high water soluble contents of the low temperature runs as compared with plant green pellets which were ñred under a hydrogen atmosphere. The results listed in Table IX compare the densities of Salt. ` the pellets produced with the maximum theoretical density v _ _ . Physical PFOPÚI'UGS--Tho Physlcaltpfopßrties of the of UO2 (10.97 grains per cubic centimeter) and are the iiame process product to be discussed include U02 X-ray 45 highest attained by this method for each powder, at the analyses, U02 particle size, densities of cold pressed and conditions listed. sintered U02 compacts, U02 surface area, and tap density. The llame process, .the micronized high-fired and the U02 X-l'fïy ílf1f1_lyS@S~*X-fay analysis has been Used t0 standard Weldon Spring Huid-bed pellets were all fabri determine the unit cell size, the strain, and the crystallite cated and ñred by the same procedure; therefore, their size for iiame process U02 and for lluid bed U02. These 50 densities should be directly comparable. values are listed in Table VI. The high-iired `screwereactor powder which had not TABLE VI been micronized was not iired as long nor at as high a temperature as the other samples. A longer, higher~ ~ U02 X-Rfly Analysis temperature ñring of the material would be expected to 55 increase the density of the product slightly, but not to the Flam@ process U 02 (Run . extent that it would equal the micronized screw-reactor Fluid _ Bed U02 "22) Unit ceu size A ______________________________ __ Strain, diinensionless_-_ Crystaiiite size A ____________________________ -_ 5.4084 Nono 1,150 . material, or llame-process material. The pellets evaluated were cyiindiical in shape (0.4” in diameter by 0.4" high). 5. 40s G0 0. 055 TABLE IX _ ~ 800 ` Denszties of cold pressed and szntered U02 c0mpacts.-- The unit cell size of both «flame process U02 and fluid commu Percent bed U02 are the same within the limits of measurement. _ U02 The composition can definitelyofvbethe established main phase as being of thebetween flame process UOMO 65 Firing Finne Onde Source mangi' tion Ofmfix. máis/m sq. in. density and U02_01, probably closer to U02'20. A statistical analysis of the X-ray data shows that the ä/iëilln@ igrßâßïlëîIí-ï--d-«î ------ »strain trapped in the lattice of llame process U02 is not * different from zero within 95% confidence limits. Based 70 Stälìlâlard Weldon Sprintr Iluid~ on this saine analysis, the crystallite size is 1150i10() angstrorn units within 95% confidence limits. 20 20 2O 1y 700 17700 1 0 16 m2352222.“ggg;"r2-„55657155“ mlßl‘OlllZlDs) ----------------- -- 50 50 96.1 95.1 ’7 0 50 92'2 1,635 100 91.7 U02 particle size-«The Microinerograph particle size distribution of U02 producgd by Various methods is Shown . in Table VII. 1 The 'temperatures listed were obtained by means of an optical pyrom» eter which had been sighted on the molybdenum bout in which the 75 pellets were fired. ‘ 3,041,136 10 by this flame process. Analytical results are not yet avail Pellets produced yfrom lflame process U02 are slightly more dense than those produced from micronized high tired U02 and yet `did not require the further expensive particle size reduction. able, but X-ray analysis indicates that the product from thorium nitrate is Th02 with no other phases detected. It will be understood that this invention is not to be limited to the details given herein, but that it may be modiiied Within the scope of the appended claims. In particular the apparatus portion of the invention is not limited to the device described in detail. For example, it may be desirable to use multiple spray nozzles operat ing parallel to each other; it might also, under some cir cumstances, be advantageous to spray countercurrent to The cost of micronizing one ton of U02 at a rate of 25 pounds per hour has been estimated at $1,466.26. While this cost might be lowered considerably if the rate could be increased »to 50 or 75 pounds of U02 per hour, the cost of micronizing would still be many times larger than the flame process cost of converting uranyl nitrate l-iquor to U02. This cost could be completely saved in cases Where flame process U02 could be successfully substituted for micronized U02 in ceramic applications. U02 surface area-Table X shows .the surface area of several U02 powders. v TABLE X the hot combustion gases. with a pressure atomizing nozzle, but a pneumatic atom 15 izing nozzle could eliminate this need for high pressure. In general an apparatus for the process must include, for satisfactory operation: U02 Surface Area BET surface area OXlde SOUICSÍ sq. meters/ gram Flame process (Run D-22) ______________ __ 2.6 Fluid bed ____ High pressure would be re quired to form liquor droplets of sufficiently small size _ 3.14 Normal screw _________________________ _.- 1.38 High tired _____________________ __ _____ __ 0.57 (1) A feed system; (2) An insulated tubular reactor; 20 (3) A spray nozzle positioned parallel to the axis of the reactor; (4) A burner for continuously burning a fuel gas-air mixture with less than the stoichiometric amount of Micronized high fired ___________________ __ 2.19 air to produce hot reducing gases; The surface area (as determined by lthe BET krypton 25 (5) An ignition system; (6) A llame detection system; method) is approximately 2.6 -square meters per gram lfor (7) Devices for collecting the products. the three samples of flame process U02 evaluated thus far. e value determined for the fluid ybed U02 is slightly It is intended that the invention include all variations in higher than that of ñame process U02 even though its methods of assembling the apparatus. agglomerates are much larger than those of the llame 30 What is claimed is: process material. This is probably due to a large amount 1. A process for converting uranyl nitrate solution to of surface area inside the ñuid ybecl U02 agglomerates, but uranium dioxide comprising spraying fine droplets of the reason why ñuid bed U02 reacts so much more slowly aqueous uranyl nitrate solution into a high temperature than flame process U02 is unknown. 35 hydrocarbon llame, said ñame being deficient in oxygen Tap Densiiy-(a)-U02.-The tap densities of several approximately 30%, retaining the feed in the llame for a types of U02 are listed in Table XI. siu‘îicient length of time to reduce the nitrate to the diox ide, and recovering uranium dioxide. TABLE XI U02 Tap Density Oxide source: 2. A process according to claim 1 wherein the reduc 40 ing flame is formed by air and an excess of propane. Tap density, gJcc. 3. A process according to claim 2 wherein the feed is introduced axially of the llame. 4. A process according to claim 3 wherein the flame is maintained at such a temperature that the temperature Flame process ______________________ __ 1.5-3.6 Fluid bed _____ __ ~4.5 Micronized high fired ______________ __ 2.74-3.50 oi the exit gases is at least 1800° F. 5. A process according to claim 4 wherein the droplets of aqueous uranyl nitrate solution are no greater than 40 microns in diameter. High fired ________________________ __ 4.3-4.42 These numbers express the range of values which have been obtained for each material. The tap densities of the recently produced ilame process U02 have all been at the high end of the range listed. In recent experi- . ments, it has been possible to raise the tap density of 50 flame process U02 by changing operating conditions, and it is believed -that the upper limit has not yet been reached for this material. (b)-UF4.--Green salt was prepared by hydrolluori 55 nating llame process U02 with 100i w/o HF batchwise in a laboratory furnace. The range of tap densities for this material and for plant UF4 are listed below: TABLE XII Green Salt Source: 2,267,720 2,613,137 2,735,745 2,737,445 2,757,072 2,761,767 2,903,334 60 Tap Density of Green Salt ' References Cited in the rile of this patent UNITED STATES PATENTS -Cyr ________________ __ Dec. 30, Hellwig _____________ __ Oct. 7, Flook et al. __________ __ Feb. 21, Nassen ______________ __ Mar. 6, Kapp et al. __________ __ July 31, Perieres _____________ __ Sept. 4, Buckingham __________ __ Sept. 8, 1941 1952 1956 1956 1956 1956 1959 FOREIGN PATENTS density of 661,685 Great Britain _______ ___ Nov. 28, 1951 U 4, gms/cc. 707,389 Great Britain ________ __ Apr. 14, 1954 'l‘a Flame Process U02 (laboratory hydrofluori nated to U'F4) ___________________ -_ 1.7-2.3 Screw Reactors _____________________ __ 3.1-3.8 65 OTHER REFERENCES Ser No. 379,872, Ebner (A.P.C.), published Apr. 27, 1943. _ The recent values of tap density for llame process ma Katz: “The Chemistry of Uranium,” 1st edition, pages terial have been at the high end of the range quoted. 303, 304, 307, McGraw-Hill Book Co., New York, N.Y. However, as in the case of tap density for iiame process U02, this is not considered to be the maximum attainable 70 (1951). Johnson et al.: “Ceramic Bulletin,” vol. 36, No. 3, page value. Further experiments have been made in which thorium nitrate, thorium-uranium (5 weight percent uranium) ni trate and aluminum nitrate solutions have been treated 116 (1957). MCW-1429, pages 57-75, May 1, 1959. Y V Hedley et al.: MCW-l45‘1, pages 19-24, Aug. l, 1960.