The centrifugal contactor utilizes centrifugal force to achieve more rapid phase separation after mixing, thereby reducing the equipment size and holdup of process solutions below that required for the mixer-settler described in the previous section. A centrifugal contactor developed by Webster et al. [C2, Dl, W2] for use in reprocessing irradiated fuel is illustrated schematically in Fig. 4.26. Organic and aqueous phases to be contacted enter at the centerline of the pump-mix chamber, which creates dispersion for interphase mass transfer and also supplies pumping action for interstage flow. The mixture flows upward through a perforated rotating plate into the rotating settling chamber, which contains radial vanes to keep the liquid

Heavy Phase
Out

Impeller and Mixing Chamber

rotating at the same speed as the bowl. Under the influence of the centrifugal force, the heavier aqueous phase collects at the outer periphery of the bowl and flows through holes in the bowl into a collecting trough located between the bowl and the stationary outer housing. The lighter organic phase collects near the rotating shaft, flows upward through holes in a top baffle, and out through holes into a collecting trough. The contactor requires no submerged bearings or external seals, but a seal at the bottom of the rotating bowl is necessary to prevent liquid in the bowl from leaking into the annular space between the rotating bowl and the stationary housing.

Dimensions and operating properties of critically safe centrifugal contactors are given in Table 4.14, adapted from Davis and Jennings [Dl]. The holdup times for these centrifugal contactors are more than 20-fold less than the holdup times for the pump-mix mixer-settlers listed in Table 4.13. Contacting efficiencies approaching 75 percent or greater of the mass transfer obtainable with a single equilibrium stage have been reported [Dl].

Because uranium is present in sulfuric acid leach liquors as a mixture of the species U022+, U02S04, U02(S04)22′, and U02(S04)34′, it is possible to extract uranium either on cation-exchange resins or anion-exchange resins. However, the concentrations of other metal cations such as Fe2+, Fe3+, Ca2+, Mg2+, Al3+, and Na+ are so much higher than that of U022+ that uranium extracted by a cation-exchange resin is heavily contaminated by other metals. On the other hand, only a few metals other than uranium, such as Fe(III), V(V), and Mo(VI), form anions in sulfuric acid, and their concentration can be made small by control of oxidation-reduction potential. Consequently, anion exchange has been recognized as a selective means for recovering uranium from sulfuric acid leach liquors since pioneering research sponsored by the U. S. Atomic Energy Commission in 1948 to 1952 [С1]. The first reported commercial use of anion exchange in the processing of uranium ore was that of the West Rand Consolidated Mines, Ltd., mill in the Republic of South Africa, in 1952 [M4]. Anion exchange is also effective and selective in extracting uranium from carbonate leach liquor, where uranium is present mainly as the anion U02(C03)34-.

For a general discussion of ion exchange, reference may be made to standard texts, such as Helfferich [H5].

Anion-exchange resins. The type of anion-exchange resin first used in uranium mills and still most commonly used is the quaternary ammonium, strong base type, in which the active nitrogen atom is attached to four carbon atoms. Figure 5.10 shows the molecular structure of a typical quaternary ammonium resin. This is made by copolymerizing styrene and divinyl benzene, chloromethylating the polymer, and reacting the chlorinated polymer with trimethyl — amine. The principal reaction by which the resin adsorbs uranium in the form of a complex anion from solution is

4RCH2(CH3)3NX + U02(S04)34- ^ [RCH2(CH3)3N]4U02(S04)3 + 4X~

Here R is the styrene-divinyl benzene radical, and X" is an inorganic anion such as СГ, N03“,

or |S042 ". Some uranium in the form of the U02(S04)22" ion is also adsorbed, together with S042" and HS04" ions and other anions possibly present in the solution.

Resins are used in the form of small spherical beads. For fixed-bed operation on clear solutions, the standard particle size range is —16 to +50 mesh (1.0 to 0.6 mm). For operation on unclarified solutions, in the resin-in-pulp type of operation to be described later, larger particles in the size range —14 to +20 mesh (1.1 to 0.9 mm) are preferred.

The density of the material of the beads is around 1.15 g/ml, so they sink in leach liquors. They are light enough, however, to be readily transported by upflowing liquids, as required in moving-bed operations.

Resin consumption and cost. Resin deteriorates in use. Volume changes during adsorption and elution and mechanical wear cause breakage and attrition. Chemical poisoning and fouling cause gradual loss of adsorptive capacity. In three South African plants, useful resin life was estimated ([М3], p. 143) at from 19,000 to 27,000 volumes of solution treated per volume of resin. This represents about 3 years’ usage in normal service.

In 1978, resins used in uranium mills cost $95 to $110/ft3. For leach liquors containing 0.5 g U308/liter or 0.031 lb/ft3, the replacement cost of resin at the highest price and lowest life would be $110/(19,000X0.031), or $0.19/lb U308.

The anion-adsorption capacity of some commercially available anion-exchange resins is given in Table 5.21. Merritt [М3] uses 1.25 meq/ml resin as a representative number. For uranium adsorbed as quadruply charged U02(S04)34", this is equivalent to

or 5.5 lb U308/ft3. Partial saturation of the resin with S042", HS04", and other anions decreases resin capacity for uranium; some adsorption of U02(S04)22" increases it.

Adsorption equilibria. The upper curve of Fig. 5.11 shows how the U308 concentration of Amberlite IRA-400 resin measured in laboratory experiments [PI] varies with U308 content of a U02S04 solution containing 5 g free H2S04/liter and enough MgS04 to bring total S042" to 30 g/liter. This approximates conditions in a typical leach liquor. Because of the presence of other anions and deterioration of resin in service, the actual uranium content of resin iri

uranium mills may be lower. In the Dawn Mining Company’s uranium mill to be described later, a resin concentration of 3 lb U308/ft3 was reported [HI] for a leach liquor concentration of 0.5 g U308/liter. Values at other mills range from 2 to 5 lb U308/ft3, depending on the uranium content of the solution and other conditions. The lower dashed curve of Fig. 5.11 will be used in subsequent predictions of ion-exchanger performance.

Ion-exchange equipment. Three types of ion-exchange equipment used in U. S. uranium mills are the fixed-bed type, the moving-bed type, and the continuous resin-in-pulp (REP) type. These will be described in turn.

Adsorption in fixed-bed ion exchange. A fixed-bed ion-exchanger is a column of ion-exchange resin through which the solution whose ions are to be adsorbed flows downward. Ions collect in the bed starting at the top and work downward as the resin becomes saturated. When the height of resin remaining unsaturated becomes too short to remove the ions completely, they begin to appear in the effluent. This is known as “break-through.” With further flow, the ion concentration of the effluent steadily increases and reaches the feed concentration when the resin bed is fully saturated. Curve A of Fig. 5.12 shows a typical curve of uranium effluent concentration versus effluent volume for a condition in which the resin contained no uranium when flow began. Effluent volume is expressed as the ratio of effluent volume to volume of the resin bed. At point B, the break-through volume, uranium begins to appear in the effluent. Operation between points В and S deposits additional uranium in the bed, because the effluent concentration is less than the feed, but uranium is being lost in the effluent. At point S, effluent concentration has reached feed concentration, and the bed is saturated with uranium.

To prevent loss of uranium while still saturating the bed with uranium, it is customary to resort to cyclic operation of two or more beds in series. Figure 5.13 is an example of cyclic operation of four beds, with three beds in series adsorbing uranium while the fourth bed is having its uranium eluted. In cycle 1, feed solution is charged to bed 1, which has been on line for two previous cycles. Solution then flows through bed 2, which has been on line for one previous cycle. Solution finally flows through bed 3, which was freshly eluted and free of uranium at the beginning of the cycle. At the end of cycle 1, the feed point is moved to bed 2, freshly eluted bed 4 is put in series after bed 3, and bed 1 is taken out of adsorption and put on elution. This progression continues through four cycles, after which the sequence is repeated. This is sometimes called “merry-go-round” operation.

Figure 5.14 shows how the uranium concentration in the resin beds of Fig. 5.13 might be distributed along the beds at different times. To make the illustration concrete, it has been assumed that the beds and flow rates have approximately the characteristics of the ion-exchange beds of the Dawn Mining Company’s uranium mill [HI]. Specifically, the beds have a cross

Figure 5.13 Four cycles of fixed — bed ion-exchange operation. F = feed; В = barren solution.

section of 50 ft2 (4.64 mJ), a depth of 5.4 ft (1.65 m), and a volume of 270 ft3 (7646 liters). To load a bed fully with uranium at 3 lb U3Og/ft3 (48.04 g U308/liter) with a solution containing 0.5 g U308/liter would involve feeding 48.04/0.5 = 96.1 bed volumes of solution. To provide a cycle time of 10 h for this example, the feed flow rate should be 9.61 bed volumes/h, or 9.61 X 7646 = 73478 liters/h, or 323 gal/min. The flow rate reported [HI] for the Dawn plant was 300 gal/min.

At the start of cycle 1, at 100 h, bed 1 has been partially loaded with uranium, distributed over the top two-thirds of the bed. Beds 2 and 3 contain no uranium. At 102 h, uranium distribution in bed 1 has increased and extends over the entire bed, so that some is being deposited at the top of bed 2. This continues until at the end of the cycle, at 110 h, bed 1 is
fully loaded with uranium and the distribution in bed 2 has increased until it is the same as in bed 1 at the beginning of the cycle. Concentration distribution in bed 3 tracks that of bed 2, 10 h later.

The change in effluent solution concentration with time is determined from the uranium concentration at the bottom of the bed at 5.4 ft, the equilibrium relation between bed and solution concentrations (Fig. 5.11), and mass transfer rates in ion exchange. Curve C of Fig. 5.12 is an estimate of how the effluent concentration from the first bed would change with time in cyclic operation.

Elution. In the elution step, an aqueous solution about 1 M in nitrate, chloride, or sulfate ion is passed through the bed to reverse the reaction given previously, transfer the uranium back from the resin to an aqueous solution, and regenerate the resin so that it can adsorb more uranium in a later cycle. The feed solution is called the eluant; the uranium-bearing product solution, the eluate. The eluant should be at least 0.1 N in free acid to prevent precipitation of uranium.

Figure 5.15 shows how the uranium concentration of the eluate changes with bed volumes of eluate for three commonly used eluants, as reported by Greer et al. [Gl], The area under each curve represents the original, uniform uranium concentration of the bed, apparently around 80 g U308/liter in this example.

Disadvantages of sulfuric acid are (1) the low uranium concentration of eluate, (2) the high volume of eluant needed for complete elution, and (3) the high acid concentration to be neutralized if uranium is to be recovered from eluant by precipitation. The advantages of sulfuric acid are that it leaves the resin in the sulfate form, which adsorbs uranium more readily than the nitrate or chloride form, and it introduces no extraneous anions that must later be purged from the system. To facilitate use of sulfuric acid, the Eluex process, to be described later, was developed; in this process uranium is removed from sulfuric acid eluate by solvent extraction rather than by neutralization and precipitation, and the acid is recycled to a subsequent elution cycle.

Sodium chloride has the advantage of lowest reagent cost. Nitrate eluant has the advantage of providing the highest uranium concentration in the eluate. However, chloride and nitrate eluants require conversion of resin back to sulfate form with sulfuric acid to improve uranium recovery in the next adsorption cycle.

Elution in fixed-bed ion exchange. As an example of elution of uranium with nitrate ion from a fixed-bed ion-exchange system, a brief description will be given of the elution cycle in the Dawn Mining Company’s uranium mill [HI], Figure 5.16 shows the eight steps of the elution cycle. Table 5.22 gives the composition and flow rate of inflow in each step. Figure 5.17 shows the change with time of the effluent uranium, nitrate ion and total acid concentration throughout the elution cycle.

In step 1 leach liquor remaining in the column at the end of an adsorption cycle is flushed from the column by fresh water into the next column on adsorption.

In step 2 the column is flushed with water from the bottom to the top to remove fines that settle at the top of the bed during adsorption from leach liquor. Washings, which may contain traces of uranium, are used to wash filters in the leaching section of the mill.

In step 3 approximately one bed volume of recycle eluate displaces the solution remaining in the bed into the feed tank, for recovery of any uranium that may have started to appear in the effluent. Recycle eluate contains approximately 30 g N03 "/liter and 25 g free acid/liter, expressed as H2S04, and a low concentration of U3Og.

In step 4 approximately three volumes of recycle eluate removes most of the uranium from the bed, and transfers it to the product precipitation tank. At the start of this step, high concentrations of uranium and acid appear in the effluent while the bed is becoming converted to nitrate; toward the end of the step the nitrate concentration in the effluent approaches feed concentration and the uranium concentration declines.

In step 5 one volume of fresh eluant is used to flush more of the uranium from the bed into the product precipitation tank. Fresh eluant contains about 50 g NO3- and 25 g acid as H2 S04 per liter.

In step 6 fresh eluant continues to purge uranium, now into the recycle eluate tank.

In step 7 the resin is converted from nitrate form to sulfate with 5% H2S04. This effluent, with high concentrations of nitrate ion and acid, is added to recycle eluate.

In step 8 the acid remaining in the bed is flushed with water into the eluant makeup tank. The bed is now in the sulfate form, the liquid in the bed is water, and the bed is ready to be used as the third bed in series in the adsorption cycle. The eight elution steps take 462 min, which is less than the 10-h duration of an adsorption cycle.

Reagent consumption in the elution section of the Dawn mill was 10 lb H2S04 and 4 lb

Table 5.22 Elution cycle in Dawn uranium mill

Duration, Total inflow,

Step min Inflow bed volumes Disposition of effluent

HN03 per short ton of ore. For ore containing 0.2 w/o U308 this would represent 2.5 lb H2S04 and 1 lb HN03 per pound U308.

Precipitation of uranium from eiuate. In the Dawn mill, uranium was recovered from eluate by two-stage precipitation with alkali. In the first stage, pH is increased to 3.3 to 3.6 by addition of CaO slurry. This precipitates most of the ferric iron that may have been adsorbed with the uranium and removes most of the sulfate as CaS04. Because the filter cake contains 1 to 2 percent uranium, it is returned to the leaching circuit.

After filtration, in the second stage, ammonia is added to bring the pH to 7.0, which precipitates the uranium as ammonium diuranate. The precipitate is filtered, washed with water, and dried at 160°C. Filtrate is added to eluant makeup, to conserve nitrate.

Moving-bed ion exchange. Ion-exchange resin particles are small spherical beads. These can be readily moved out of an ion-exchange column with up-flowing liquid and pumped with the liquid to another tank or column. Transport of resin is used in two types of uranium extraction processes, (1) the moving-bed type and (2) the continuous RIP type.

Among U. S. mills, the moving-bed system is used in the mill of the Lucky Me Uranium Company at Gas Hills, Wyoming [D2a]. The moving-bed process is a modified batch-operated, fixed-bed ion-exchange system. Adsorption of uranium from leach liquor is carried out in two parallel sets of three columns in series. Each set is operated cyclically as in the fixed-bed system described earlier, with the feed point moved progressively around the cycle as the last bed in flow sequence becomes saturated. The novelty of the moving-bed system is in the physical transfer of loaded resin from an adsorption column to one of three elution columns, also operated cyclically in series. After elution is complete, the stripped resin is transferred back to one of the two adsorption sets, where it is placed last in flow sequence.

Physical transfer in the moving-bed system involves the following operations. The resin is first washed with down-flowing water in its original column. It is then transported with up-flowing water to a resin transfer and backwash tank. There it is given another wash and then transferred by up-flowing water to the column where it is next to be used.

Advantages of this system are simplification of fluid valving and piping: The adsorption columns need be connected only to leach liquor supply, to barren solution disposal, and to resin transfer tank; and the elution columns need be connected only to eluant and wash supply, to eluate and recycle storage, and to resin transfer tank. This reduces the chance of cross contamination of solutions between the adsorption and elution systems.

Continuous RIP ion exchange. As an illustration of the continuous RIP ion-exchange process used in several U. S. uranium mills, a brief summary will be given of Merritt’s [М3] description of the principal steps in the uranium mill of Federal-American Partners at Gas Hills, Wyoming, with reference to Fig. 5.18.

Ore containing around 0.15 percent U308 is crushed dry to particles smaller than 1 in, then ground wet with heated water and dilute sulfuric acid recycled from subsequent ion-exchange operations at B. Grinding is done in closed circuit with classifiers. Classifier overflow is a slurry, or pulp, 55 percent solids, 95 percent finer than 28 mesh. In a leaching period of 12 to 13 h, the pulp passes through six leaching tanks in series, to which are added enough sulfuric acid and sodium chlorate to bring the effluent solution to a free sulfuric acid concentration of 10 g/liter and the oxidation-reduction potential relative to the calomel electrode to —390 to —400 mV. The pulp passes through a system of cyclones and several classifiers. There it is separated into a sand fraction coarser than 325 mesh, which is washed with water and sent to tailings, and a fine slurry, or pulp, fraction, which contains 11 to 12 percent of slimes (particles finer than 325 mesh) and about 0.8 g dissolved U308/liter. The pH of pulp is raised to 1.65 by addition of ammonia to increase adsorption of U02(S04)34” while decreasing that of S042~ and HS04~.

The pulp flows through seven continuous ion-exchange stages countercurrent to ion — exchange resin beads in the size range 20 to 50 mesh. In these seven stages, dissolved uranium is transferred from the pulp to the resin leaving stage #1, while the slime tailings leaving stage #7 are substantially free of dissolved uranium.

Each adsorption stage consists of a mechanically stirred tank 11 ft (3.3 m) in diameter by 10 ft (3.0 m) deep, from which the mixture of resin and pulp flows to an airlift, which in turn discharges to a set of screens SI. These separate the coarser resin particles from the finer slimes. The resin drops by gravity into the next lower numbered tank, toward the feed end of the cascade, and the slurry flows to the next higher numbered tank, toward the tailings end. This counterflow is made possible by the absence of particles coarser than 325 mesh in the pulp and the absence of particles finer than 50 mesh in the resin. Residence time in each adsorbing stage is 18 to 20 min.

Resin leaving adsorbing stage #1 contains 2 to 3 lb U308/ft3 (32 to 48 g/liter). This resin is washed partially free of entrained slurry, with the washings and some dissolved uranium returned to the grinding circuit. Washed resin then passes countercurrent to eluant in 12 eluting

stages. Each stage consists of a mechanically stirred tank 6 ft (1.8 m) in diameter by 6 ft (1.8 m) deep. Interstage flow is driven by airlifts, with separation of resin from eluate in settling cones S2. This simpler separator is feasible because the slime content of the liquid in the eluting stages is low. Residence time in each eluting stage is 25 to 30 min.

Eluant containing 100 g H2S04/liter and about the same content of (NH^SC^ is charged to stage #12. In its flow through the 12 eluting stages it picks up uranium and water from the counterflowing resin and leaves with a concentration of around 10 g U308/liter and 80 g H2S04/liter. This rich eluate is sent to solvent extraction, Fig. 5.19, for recovery of uranium.

Stripped, barren resin leaving eluting stage #12 is washed with water in a thirteenth stage and charged to adsorbing stage #7. Barren resin washing is used to wash loading resin and then returned to grinding at B.

The total resin inventory is 825 ft3; the resin circulation rate is 1320 ft3/day or 26 liters/min. Despite the continuous flow of resin, attrition rate has been very low, averaging only

0. 076 percent per day. Absence of abrasive sand particles coarser than 325 mesh is important in keeping the attrition rate low.

Eluex process. Although it would be possible to recover uranium from the eluate leaving Fig.

5.18 by neutralization with ammonia, this would be costly because of its high H2S04 content. The Eluex process was developed to concentrate the uranium by solvent extraction before precipitation, thus reducing the ammonia requirement and simultaneously purifying the uranium further.

Figure 5.19 is a schematic flow sheet showing the Eluex process as used at the Federal-American Partners’ uranium mil, and the final steps for precipitating and calcining its uranium product. The Eluex process is a variant of the amine extraction (Amex) process described in Sec. 8.6.

Eluate containing 10 g U3 08 /liter flowing at the rate of 90 liters/min is extracted in four countercurrent mixer-settler stages with 150 liters/min of a solution of 6 v/o tertiary amine and 3 v/o isodecanol in kerosene. Organic extract contains about 6 g U3Os/liter. Uranium in aqueous raffinate is reduced to under 0.01 g U308/liter, and H2S04 is reduced from 80 to 65 g/liter. Raffinate passes through a settling tank to recover entrained solvent, is reacidified, and is returned to serve as ion-exchange eluant.

Organic extract is stripped of uranium with 23 liters/min of an aqueous ammonium sulfate solution in four additional countercurrent mixer-settler stages, to produce an aqueous product containing about 39 g U308/liter. To drive uranium into the aqueous product, pH in the first mixer stage is brought to 4.1 to 4.3 by addition of ammonia gas to react with most of the 9 g H2S04/liter carried by the extract.

Uranium in the aqueous product is precipitated by enough additional ammonia gas to bring the pH to 7.0. The precipitated (NH^UjO, is separated from the ammonium sulfate solution in a thickening tank and centrifuge and is washed with water. Ammonium sulfate filtrate is returned to the stripping section, with excess bled off to the IX circuit. (NH4)2U207 precipitate is dried and converted to U308 in a roaster at 600°C. Product contains 95 to 96 percent U308 and 3 to 4 percent S04.

Attempts to separate thorium and uranium from sulfuric acid solution of monazite by solvent extraction with TBP were unsuccessful because distribution coefficients of uranium and thorium from monazite solutions were too low, as these elements are complexed by phosphate ion. Development of extractants with higher distribution coefficients for these metals has made solvent extraction a practical process for recovering uranium and thorium from monazite sulfate solutions and from sulfuric acid solutions of other thorium ores. This section describes processes tested on a pilot-plant scale by Oak Ridge National Laboratory [С5].

Table 6.18 summarizes distribution coefficients of hexavalent uranium, thorium, and trivalent cerium (representative of rare earths) for four different types of long-chain amines, in sulfate solution with phosphate ion absent. Primary amines have the highest coefficient for thorium and the lowest for uranium, with the converse true of tertiary amines such as were cited for uranium extraction in Chap. 5. Secondary amines extract both metals, with thorium extraction favored by branching distant from the nitrogen. Either primary or secondary amines provide good separation of thorium from cerium. [19] [20]

Because of the high acidity and high sulfate and phosphate content of sulfuric acid monazite leach solutions, distribution coefficients with primary and secondary amines are lower than in Table 6.18. In monazite sulfate solutions, thorium distribution coefficients with the primary amines of Table 6.18 are still greater than 500, however. The coefficient with di(tridecyl)amine is 4.6. These are still high enough for practical processes [С5].

Table 6.19 lists distribution coefficients for amines considered [C5] for extracting uranium from monazite sulfate solutions after removal of thorium. Except for Primene JM, all coefficients were judged [C5] to be large enough and sufficiently greater than those of the rare earths to provide efficient solvent extraction separation of uranium.

Crouse and Brown [C5] give a number of alternative flow sheets for separating quite pure thorium, uranium, and rare earth products from monazite sulfate solution by solvent extraction, using several alternative organic amines and with different orders of separation. They named this type of process using amine extractants the Amex process. Figure 6.7 is a composite of several of their flow sheets showing one possible arrangement for separating these three components by solvent extraction.

Monazite sulfate solution containing 5.9 g thorium/liter, 0.2 g uranium/liter, and 34 g rare-earth oxides/liter and about 3 TV in sulfuric acid is extracted with a 0.1 Af solution of the primary amine Primene JM at an organic/aqueous flow ratio of 80:59. The solvent is 97% kerosine, 3% tridecanol. The flow ratio and solvent composition are so chosen that the solvent leaving the extracting section is effectively saturated with thorium (3 g/liter), to minimize extraction of uranium and rare earths. To ensure high thorium loading of solvent, the solvent-to-feed ratio is set below that which would extract all thorium in the feed. To prevent thorium loss, 25% of the aqueous stream is withdrawn from the second stage of the extracting contactor and recycled to feed. This permits reduction of thorium content of the raffinate leaving the fourth stage to <0.01 g/liter.

Rich solvent leaving the extracting section is scrubbed with 0.2 M H2 S04 in four scrubbing stages to remove traces of uranium and rare earths. Finally, the thorium is stripped into the aqueous phase by 0.75 M Na2C03.

A similar sequence of operations in the uranium separation section separates uranium from rare earths. The difference here is use of triisoocytylamine as solvent because of its high selectivity for uranium.

Finally, rare earths are extracted from the raffinate leaving the uranium separation section

by another solvent extraction with Primene JM. Because the rare-earth distribution coefficient is lower than thorium’s and the rare-earth concentration higher, the Primene concentration and the organic-to-aqueous flow ratio are higher than in the thorium extraction section. Rare earths may be stripped from this solvent by З M sulfuric acid. Evaporation of the sulfuric acid strip solution to 8 M H2S04 precipitates the rare earths as sulfates and, after filtration, provides acid for recycle (not shown).

An alternative method for separating rare earths from the raffinate from uranium extraction is salting out the sodium double sulfate with sodium chloride or sodium sulfate [С4].

Although there has been no reported use of solvent extraction for commercial processing of monazite sulfate solutions, it seems likely that this efficient method would be used if the demand for thorium increased sufficiently to require construction of new extraction plants.

1.3 Actinide Radioactivity in Uranium and Uranium-Plutonium Fuel

The important actinides in irradiated uranium fuel are uranium, neptunium, plutonium, americium, and curium, which are produced according to the reactions of Fig. 8.5. 236U,

^The terminology “radioactivity concentration limit” is that used in the U. S. Federal Regulations. In the publications of the International Committee in Radiation Protection [II], a similar concentration limit is referred to as the “maximum permissible concentration.”

4PWR, pressurized-water reactor; HTGR, high-temperature gas-cooled reactor; LMFBR, liquid — metal-cooled fast-breeder reactor. Data are calculated for 150 days after discharge. Calculated from data in [B2].

* Additional 3H produced by neutron activation is shown in Table 8.11.

§ Ruthenium may also form volatile compounds.

4 Total includes radionuclides not listed here.

produced by (n, y) reactions in 235 U, is important because of its neutron absorption. If uranium containing 236U is recycled, a slightly greater fissile concentration in the fresh fuel to the reactor is required. Neutron capture in 236U and the (n, In) reaction in 238U lead to 6.75-day 237U, which dominates the uranium radioactivity during the first several months that irradiated fuel is stored after discharge. Because of its relatively short half-life, 239U disappears rapidly after the fuel is discharged.

Decay of 237U forms 237Np, which is important because its (n, y) and (n, 2л) reactions lead to 238Pu and 236Pu. Also, 237Np is an important long-term constituent of radioactive wastes, particularly because its transport through some geologic media is not as delayed as that of other actinides and because of the toxicity of radionuclides in its decay chain, especially 233 U, 229 Th, and 225 Ra.

Although only small quantities of 238Pu are formed, its half-life of 86 years is long enough that 238Pu persists in plutonium recovered for recycle and is short enough that 238Pu is the greatest contributor to the alpha activity of plutonium in irradiated fuel. Although the quantities and activities of 2.85-year 236Pu are relatively small, its decay daughter 232U can build up when recovered plutonium is stored prior to fuel refabrication. As discussed in Sec.

2.3, the 232U decay daughters emit high-energy gammas and may contribute to the shielding requirements for handling recycled plutonium. The largest material quantities of plutonium are produced by neutron capture in 238U, leading through short-lived 239U and 239Np to fissile 239Pu, with a half-life of 24,400 years. Nonfission capture of neutrons in 239Pu results in ^Pu, and its neutron capture results in fissile 241 Pu. Because of its half-life of 6580 years, ^Pu is a strong and persistent alpha source in reactor plutonium, and 13.2-year 241 Pu is an extremely intense beta source. Because of the long half-life of 242Pu its radioactivity is not important compared to the other plutonium isotopes. Its neutron-capture daughter 243Pu is short-lived and decays away within a few days after plutonium is removed from the neutron environment of the reactor.

Radioactive decay of 241 Pu and 243 Pu form 241 Am and 243 Am, which are also important and persistent sources of alpha radioactivity in discharge fuel. Another persistent americium radioisotope is 152-year 2421,1 Am, formed by neutron capture in 241 Am. Its isomeric decay and the beta decay of its short-lived daughter result in 163-day 242 Cm, which is the most intense source of alpha activity in discharged uranium fuel. Successive neutron captures lead to 243Cm, 244Cm, and 245Cm. Higher-mass curium nuclides are usually not important in power reactor fuel.

M2Cm alpha decays to 238Pu, 343 Cm to 239Pu, шСт to “‘’Pu, and “5Ст to 241 Pu. Also, the alpha decay of 243 Am results in 239Np, which decays quickly to 239Pu. The decay of 242Cm prior to fuel reprocessing adds to the quantity of 238 Pu in recovered plutonium. Also, these decay reactions are the most significant sources of plutonium in the high-level wastes resulting from reprocessing uranium fuel. Although the 242Cm decay daughter 238Pu is not an important contributor to the alpha activity of high level wastes, the subsequent decay daughter 226 Ra is one of the most important contributors to the long-term ingestion toxicity of these wastes.

Material quantities and activities of the actinides in the discharge fuel can be calculated from the equations in Chap. 2. If the irradiation is at constant neutron flux, Eqs. (2.104) and (2.113) can be applied directly, as in the example of Sec. 6.5 of Chap. 2. However, power reactors usually operate at constant power, and because of the changing inventory and composition of the fissile material the neutron flux usually increases between refueling intervals. Equations (2.104) and (2.113) can still be applied to calculate the amount of a nuclide in an actinide chain by assuming constant neutron flux during a small but finite time increment, solving the nuclide equations for that time increment, recalculating the flux, and proceeding stepwise through subsequent time steps. This is the calculational method of the ORIGEN code [B2], which was used to calculate [PI] the quantities of actinides in discharge fuel for the pressurized-water and fast-breeder reactors. The results appear in Table 8.4.

The data in Table 8.4 show that curium is the strongest alpha source during fuel reprocessing, assuming that fuel is reprocessed 150 days after discharge from the reactor. The 246 kg of plutonium to be recovered yearly from the discharge fuel contains 1.2 X 10s Ci of alpha activity and 2.8 X 106 Ci of beta activity. The remaining actinide activity is associated with americium and curium, which will normally follow the high-level reprocessing wastes, along with the fission products.

The effect of plutonium recycle is to increase the production of higher-mass isotopes of plutonium and of americium and curium, because the recycled plutonium is exposed to neutrons throughout the entire irradiation cycle. The actinide quantities calculated [PI] for the same 1000-MWe reactor operating on an equilibrium fuel cycle with self-generated plutonium recycle are shown in Table 8.5. The alpha activity of the plutonium processed yearly is increased by a factor of 14 by plutonium recycle, the americium activity is increased by a factor of 5, and the curium activity by a factor of 7.

Also shown in Table 8.5 are the actinide quantities of a 1000-MWe fast-breeder reactor operating on an equilibrium fuel cycle with recycle of plutonium and uranium [PI]. The quantity of plutonium to be recovered and fabricated into recycled breeder fuel is greater than

for the LWR operating with plutonium recycle, because of the higher fissile concentration required for fast-breeder fuel. However, the breeder produces much less 238Pu, so the total alpha activity in the breeder plutonium is almost 10-fold less than in the water-reactor plutonium. Also, the breeder does not build up such large concentrations of 241 Pu and 242 Pu, and the yearly production of americium and curium is less [PI].

As in the case of uranium metal, production of pure plutonium metal presents many difficulties. It forms very stable compounds with oxygen and carbon, it oxidizes rapidly in air when in the form of powder, it cannot be deposited electrolytically from aqueous solution, and it boils at too high a temperature to be purified by distillation. Additionally, the extreme radiotoxicity of plutonium and neutron production from (a, n) reactions require that production operations be carried out in airtight and shielded enclosures. Nuclear criticality limits the amount of plu­tonium produced in any production operation to batch sizes of no more than a few kilograms.

Methods that have been used to produce plutonium metal are

1. Reduction of PuF4

2. Reduction of PuF3

3. Reduction of PuF4 — Pu02

4. Reduction of Pu02

5. Reduction of the double salt CaF2 -PuF4

6. Electrolysis of fused salts

7. Thermal decomposition of plutonium hydride

Thermochemical reduction. Elements that might be considered for reducing Pu02 or the plu­tonium halides are hydrogen, sodium, or calcium. Carbon is impractical because of carbide forma­tion, and magnesium or aluminum are undesirable because they form intermetallic compounds with plutonium. Excess reductant must be used to ensure high yields, and it is important that the remaining reductant not dissolve in or react with the molten plutonium. Lithium and potassium could also be considered, although the volatility of potassium and the chemical reactivity of lithium would make handling difficult.

The free-energy changes in reducing Pu02, PuF3, PuF4, or PuCl3 by hydrogen, sodium, or calcium are shown in Table 9.21. Magnesium is included in the table because it can be vacuum dis­tilled from the plutonium product, and its thermochemistry is relevant to the use of MgO as a crucible in the reduction operation. The free-energy data are evaluated at 1500 К (1227°C), which is high enough for rapid reactions, is above the melting point of plutonium 640°C, and is within the temperature range required to obtain molten slags formed by reduction of the plutonium halides. A molten slag is important in aiding the coalescence of the molten plutonium metal that is produced. The melting points of the slags produced are indicated by the melting tempera­tures of the reductant compounds shown in Table 9.22. For reduction to be complete without requiring a large excess in reductant, the free-energy change at 1500 К should be more negative

than about 10 kcal/g-mol of plutonium. The positive free-energy changes for hydrogen reduction, shown in Table 9.21, indicate that hydrogen is impractical. Sodium results in a suitably negative free-energy change with PuF4 or PuCl3, and magnesium is thermodynamically suitable for the reduction of Pu02, PuF3, and PuF4. Calcium is the only feasible reductant for all four plutonium compounds, including Pu02 and PuF3. Negative free-energy changes also result for the calcium reduction of PuBr3 or Pul3, but these plutonium salts are both hygroscopic and corrosive [C3].

Because of the small scale and correspondingly high heat loss in plutonium reduction, it is important that the reaction be sufficiently exothermic so that the reacting mixture will self-heat to the temperature range of 1500 K, after the reaction is initiated at lower temperature by an external heat source. Data on heats of formation and available heat of reaction are given in Table 9.22. The greatest heat of reaction, 161 kcal/g-mol of plutonium, results from the calcium reduc­tion of PuF4 . PuF4 is also easy to reduce because it is not hygroscopic, so it does not absorb moisture which could create excessive pressures during reduction operations.

Even the highly exothermic calcium reduction of PuF4 may be insufficient to result in slag temperatures above the CaF2 melting temperature of 1418°C, particularly in smaller-scale reduc­tions because of the high heat loss. To furnish additional heat, elemental iodine is added to the reactants as a booster, to react with excess calcium. The heat of formation of the resultant Cal2 at 298 К is —143.4 kcal/g-mol [B3]. No iodine booster is needed for larger-scale reductions of PuF4, in the range of about 1 kg of plutonium.

The thermodynamically favorable reduction of Pu02 with calcium has the disadvantage that the CaO coproduct is not molten, so that the resulting plutonium metal and unreacted calcium metal remain finely dispersed throughout the slag. However, the dispersed plutonium can be recovered as a massive metal by preferentially extracting the calcium oxide and unieacted calcium with molten calcium chloride at a temperature above the melting points of plutonium and calcium, leaving consolidated plutonium metal with yield efficiencies in excess of 99.9 percent [W1 ].

Electrolytic processes. Because of the positive oxidation potential for plutonium metal to displace hydrogen from aqueous solution, as shown in Table 9.7, nonaqueous solutions such as fused salts must be used for the electrodeposition of plutonium metal. One process involves the electrolysis of a molten equimolar mixture of LiCl-KCl containing 30 w/o PuCl3. The melt is contained in a MgO-Ti02 crucible heated to 950°C, with an anode through which chlorine gas can be introduced

The nitric acid evaporated from the high-level waste concentrator is too dilute and contains too much entrained radioactivity to be recycled without additional treatment. This acid, together with dilute acid waste streams from the uranium and plutonium purification solvent extraction systems, is decontaminated in the nitric acid evaporator. Entrainment can be suppressed by providing partial reflux through a few bubble-plate or perforated-plate trays, backed up by wire-mesh mist eliminators.

Decontaminated acid is separated in the acid fractionator into 15 M acid, water, and acid of intermediate concentration as needed in the reprocessing plant. The 15 M upper limit is the concentration of the nitric acid-water azeotrope. The acid evaporator and fractionator are made of stainless steel and usually run at an absolute pressure of 70 to 200 Torr to reduce corrosion and reaction of nitric acid with traces of TBP dissolved or entrained in the acid feed. At 200 Torr the azeotrope boils at 86.5°C.

The techniques of waste management depend largely on the type of waste to be dealt with. The criteria are the level of the radioactivity concentration in the waste, the nature of the radionuclides present in the waste, and the properties of the material that carries the radioactivity.

With respect to the level of radioactivity concentration, several waste classifications are in use appropriate for particular handling schemes. More basic distinctions are between waste that requires radiation shielding and that which does not and between waste that needs to be cooled and that which does not.

The radionuclides associated with the waste have to be considered in terms of the type of radiation, the half-life, and possibly the chemical nature. Long lived alpha-emitting actinides call for particular attention because of the high and long-lasting radiotoxicity typical of those alpha-emitters.

As for the material, the primary distinction is between solid, liquid, and gaseous waste. Solid waste includes any kind of contaminated or activated plant components, tools, filters, and protective clothing. Most of the liquid wastes are aqueous solutions or sludges. The term gaseous waste will be used for radioactive gases recovered from off-gas streams and contained in an appropriate form.

From these criteria a number of waste categories may be derived according to further treatment and final disposal requirements.

0. High-level waste (HLW): alpha, beta, and gamma emitters; shielding required, cooling may be required.

(a) Liquid HLW concentrate

(b) Solid HLW

1. Non-high-level waste (medium-level waste, MLW; low-level waste waste, LLW): shielding may be required.

(a) Alpha waste (liquid and solid): alpha (beta, gamma) emitters, alpha activity dominating

(b) Non-alpha waste (liquid and solid): beta, gamma emitters

2. Special radionuclide wasted

(a) 85Kr (gaseous beta emitter; 10-year half-life)

(b) Tritiated water (weak beta emitter; 12-year half-life)

(c) IJ9I (beta, gamma emitter; 107-year half-life)

Because 238 U undergoes 14 successive reactions while decaying into stable 206 Pb, the activity of pure 238 U is only one-fourteenth as great as that of undisturbed uranium ore. Freshly extracted uranium consists of a mixture of 238 U and an equal activity of 234 U, and has one-seventh the activity of the ore from which it was extracted. Fresh uranium emits mainly alpha activity. As uranium ages, beta and gamma activity develop, owing to the growth of the first two decay products of 238 U:

24.1-day 234 Th (UX-1) and 1.17-min 234Pa (UX-2)

After a month these approach saturation activity, so that uranium older than this has four-fourteenths the activity of the uranium ore from which it came. The activity of uranium then remains constant for hundreds of years because of the long half-life of 230Th, the first daughter of 234 U.

Removal of the elements responsible for the other ten-fourteenths of the activity of uranium ores is one of the important aspects of uranium concentration and purification.

Purified uranium, freed from the decay products of 234 U, is much less toxic than uranium ores. For the same reason, the tailings of uranium mills are much more toxic than purified uranium. The most dangerous impurities among the decay products of 234 U are 230 Th and radium, relatively long-lived alpha emitters; radium’s gaseous daughter radon, which disperses radio­activity in uranium mines and near mill tailing piles; and 210Po, a very toxic alpha emitter. These, more than uranium itself, are responsible for the radioactive hazards of uranium mines and concentrating plants.

The enthalpy data of Table 5.31 may be used to show that the reaction of UF4 and calcium metal initially at 25°C (298 K) provides sufficient heat to melt the reaction products (uranium metal and CaF2) at the melting point of CaF2 (1691 K), with an excess of 35.3 kcal/g-mol uranium available to offset heat losses. However, with magnesium metal, the same table shows that its heat of reaction with UF4 initially at 25°C is insufficient to melt the reaction products completely at the melting point of MgF2 (1536 K), and that 6.8 kcal of heat must be supplied per gram-mole uranium produced.

The lowest temperature to which the stoichiometric mixture of UF4 and magnesium must be preheated to just melt the reaction products without heat loss may be found by reference to Fig. 5.24, in which the enthalpies of 1 mol of UF4, 2 mol of magnesium, and their mixture are plotted against temperature. The temperature at which the enthalpy change of the mixture from 25°C is 6.8 kcal, or 192°C, is the desired minimum preheat temperature. In practice, it is necessary to preheat the charge to a higher temperature to compensate for heat losses and because a slight excess of magnesium is used, as illustrated by Prob. 5.6.

Zirconium dioxide, zirconia, is the only oxide of zirconium stable chemically at temperatures below 2000 K. At higher temperatures some dissociation into ZrO and oxygen takes place. The phases of ZrOj, their densities, and phase-transition temperatures are listed in Table 7.5. Zirconia stabilized in the high-temperature cubic phase by addition of 3 to 5 percent calcium oxide is used as a refractory at temperatures up to 2200° C. Zr02 has been used to dilute 235 U02 in fuel elements.

Zirconium dioxide can be partially reduced to zirconium metal by reactive metals such as calcium or magnesium, but the product is contaminated by some unreduced oxide, owing to the solubility of oxide in the metal.

Zirconium dioxide reacts with carbon to form the carbide ZrC. Zirconium dioxide is inert to the halogens, but when mixed with carbon reacts to form tetrahalides at high temperatures. This is the basis of one process for extracting zirconium from zircon.

Zirconium dioxide is unattacked by all mineral acids except concentrated HF and H2S04, which slowly dissolve it. Zr02 can be converted to ZrF4 by reaction with gaseous HF at 550°C. Fusion with a fluosilicate converts Zr02 to fluozirconate:

K2SiF6 + Zr02 -*■ KjZrF6 + SiOj

Fusion with alkali hydroxides converts Zr02 to a zirconate:

2NaOH + Zr02 -+ Na2Zr03 + H20

As alkali fluozirconates are soluble in water, and alkali zirconates are soluble in strong acid, fusion with fluosilicates or with alkali hydroxides are useful steps for getting Zr02 into aqueous solution.