Until recently, most U. S. thorium production has been as a by-product of monazite processing from placer deposits in Florida, Georgia, and South Carolina. Some bastnaesite has been mined at Mountain Pass, San Bernardino, California.

The U. S. Geological Survey was scheduled to publish a revised study of U. S. thorium resources in August 1979. Partial results of this study, which cover most of these resources but do not include the beach placers of Florida, Georgia, and the Carolinas, were presented orally by Staatz [S5] of the U. S. Geological Survey in 1978. Table 6.14 lists the types of deposit, the principal districts in which potentially economic thorium-bearing deposits have been found, the principal thorium minerals, and estimates of thorium reserves and resources. Thorium from the vein deposits, the first type, could be produced for less than $30/lb. Thorium is the principal salable product in these deposits. Thorium could be coproduced with other elements from disseminated deposits, massive carbonatites, and placers; the amount of thorium that might be produced from them, and its cost, depends on the marketability of the other minerals that occur with the thorium.

Table 6.13 Thorium resources of non-Communist world

Processes for making ductile zirconium by electrolysis of K2ZrF6 dissolved in molten chlorides have been described by Steinberg et al. in the United States [S4, R2] and by Ogarev et al. [01] in the Soviet Union. An advantage over the Kroll process is that a coarsely crystalline product is obtained from which coproduced halides can be removed by leaching with acidified water without undue contamination of zirconium by oxygen. The washed crystals are then vacuum dried and consolidated by arc melting.

Steinberg et al. [S4], of Horizons, Inc., developed a process for making ductile zirconium by electrolysis of K2ZrF6 dissolved in fused sodium chloride. The essential features of the process were as follows:

1. A gastight cell filled with a purified argon atmosphere

2. A carbon crucible to hold the fused-salt bath and act as anode

3. Electrolyte consisting of high-purity K2ZrF6 and NaCl, free from oxygen and water

4. Steel cathode

5. 850°C bath temperature

6. Initial concentration, 30 to 35 w/o K3ZrF6 in NaCl

7. 3.5 to 4.0 V

8. Current density, 250 to 400 A/dm:

The current efficiency was about 60 percent and the current yield around 0.5 g zirconium/A-h.

The largest cell developed [R2] produced zirconium at a rate of from 4 to 6 lb/h, and 30 to 40 lb of zirconium metal per run. A cell could be used for a minimum of six runs before the operation was terminated by buildup of NaF and KF formed in the overall reaction

K2ZrF6 + 4NaCl ->■ Zr + 2C12 + 2KF + 4NaF

A somewhat similar process has been described by Ogarev et al. [01], the principal differences being the use of KC1-K2ZtF6 as electrolyte and a steel cell with frozen salt wall to contain it.

9 ALTERNATIVES FOR PRODUCING HAFNIUM-FREE ZIRCONIUM FROM ZIRCON

The foregoing processes for extracting zirconium from zircon, separating zirconium from hafnium, and reducing zirconium compounds to metal can be combined in a number of ways, some of which are shown in Fig. 7.15. The combination of processes used in the United States in 1978 is at the left of this figure.

Neptunium oxides. Keller [K2J reports five binary oxides or oxide hydrates of neptunium: Np02, Np2Os, Np3Og, Np03’2H20, and NpOu’H^O. Anhydrous Np(VI) oxide has not been prepared.

Neptunium dioxide, Np02, is the most stable of the neptunium oxides. It crystallizes with the fluorite structure of all the actinide dioxides, with a crystalline density of 11.14 g/cm3. It can be formed from the thermal decomposition of other neptunium compounds, such as the hy­droxide, the nitrate, or the oxalate, in the temperature range of 600 to 1000°C. High-fired Np02 can be dissolved in hot concentrated nitric acid containing small amounts of fluoride.

The mixed oxide Np308 is structurally analogous to U308. Above 500°C it decomposes to Np02.

Neptunium halides. Neptunium forms binary halides in the oxidation states of Np (ПІ, TV, and VI). The trifluoride is prepared by hydrofluorination of the dioxide, hydroxide, carbonate, oxa­late, or nitrate in the presence of hydrogen:

NpOj + |Hj + 3HF = NpF3 + 2H20 (935)

The trifluoride can be carried to the tetrafluoride by hydrofluorination in the presence of oxygen:

NpF3 + *02 + HF = NpF4 + jH2 0 (936)

The above hydrofluorination reactions are carried out at 500°C [F4].

Fluorination of Np02 or NpF4 by F2, BrF3, or BrFs at 300 to 500°C results in NpF6 [М2], which has a triple point of 55.759°C and 758 Torr [04, W2]. Neptunium hexafluoride, like PuF6, is decomposed in the presence of light [K2]. The hexafluoride is readily hydrolyzed by moisture to form the Np(VI) oxyfluoride Np02 F2, which can be reduced by hydrogen to form the Np(V) oxyfluoride NpOF3.

Dissolution, described in Sec. 4.4, produces an aqueous solution of uranyl nitrate, plutonium(IV) nitrate, nitric acid, small concentrations of neptunium, americium, and curium nitrates, and almost all of the nonvolatile fission products in the fuel. With fuel cooled 150 days after bumup of 33,000 MWd/MT, the fission-product concentration is around 1700 Сі/liter. The first step in the solvent extraction portion of the Purex process is primary decontamination, in which from 99 to 99.9 percent of these fission products are separated from the uranium and plutonium. Early removal of the fission products reduces the amount of required shielding, simplifies maintenance, and facilitates later process operations by reducing solvent degradation from radiolysis.

The decontamination system consists of an extracting section, often designated HA, and a scrubbing section, HS. In the extracting section, uranium and plutonium are extracted from the dissolver solution by multistage countercurrent contacts with 30 v/o TBP in a normal-paraffin diluent. Fission products, which have much lower distribution coefficients than uranium and plutonium, largely remain in the aqueous phase and leave the extracting section in the aqueous raffinate. Americium and curium are predominantly trivalent, have low distribution coefficients like the rare-earth fission products, and also leave in the raffinate. Neptunium in the feed is partly in the extractable hexavalent state and partly in the inextractable pentavalent state and divides between aqueous raffinate and organic extract.’*’

^Neptunium behavior in Purex systems is discussed further in Sec. 7.

In the scrubbing section, most of the small amounts of fission products from the feed carried by the solvent leaving the extracting section are moved from the solvent and returned to the extracting section by countercurrent washing with aqueous nitric acid, about З M.

Contacting equipment used in the extracting section must have low holdup to minimize solvent degradation from the intense fission-product radioactivity. Here, centrifugal contactors or pulse columns are preferred to mixer-settlers. In the scrubbing section and in the balance of the solvent extraction plant, mixer-settlers are often used.

The extracting section is usually run at or near room temperature, to reduce solvent degradation and because the uranium distribution coefficient is higher the lower the tempera­ture. It has been found advantageous to operate the scrubbing section at higher temperature, around 60°C, primarily because decontamination of ruthenium is more complete at higher temperature.

Figure 10.8 illustrates the effect of nitric acid concentration on distribution coefficients in the Purex process, for 80 percent saturation of the solvent with uranium, a condition that obtains near the feed point. Increasing acid concentration improves separation of ruthenium

from uranium and plutonium, but impairs slightly separation of zirconium. An acid concentra­tion around 2.5 to 3.0 M is a practical optimum for these contaminants.

Figure 10.9 illustrates the effect of uranium saturation of solvent on distribution coefficients, at nitric acid concentrations approximately those in the extracting and scrubbing sections. The ratio of plutonium distribution coefficient to fission products is improved at high uranium loadings, a condition sought at the feed point.

More quantitative data on distribution coefficients for uranium, plutonium, and HN03 are given in Sec. 4.15.

Process selection. The processes just described recovered neptunium only partially and in variable yield because of the difficulty in controlling the distribution of neptunium valence between 5 and 6 in the primary extraction step with nitrite-catalyzed HN03 and the incomplete reduction of neptunium from valence 5 to 4 in the partitioning step with ferrous ion. This section describes a modified Purex process that could be used if more complete recovery of neptunium were required. It is based on process design studies by Tajik [Tl]. The principal process steps are shown in the material flow sheet Fig. 10.32. In the primary decontamination step, pentavalent vanadium oxidizes neptunium to the extractable hexavalent state. In the partitioning step, tetravalent uranium reduces plutonium to the inextractable trivalent state while converting neptunium to the still-extractable tetravalent state.

Decontamination. Prior to decontamination, nitrites in the aqueous feed must be decomposed by air sparging, to prevent them from reducing vanadium in the HA contactor. Extraction of uranium and Pu(IV) in the HA contactor and scrubbing of fission products in the HS contactor are carried out substantially as in the conventional Purex process described in Sec. 4. The neptunium oxidant, 0.054 M pentavalent vanadium in З M nitric acid, is fed to the extracting section two theoretical stages below the aqueous feed point. This point is selected so that most of the uranium and plutonium will have been already extracted from the aqueous stream. As the aqueous stream flows through the remaining five extracting stages, vanadium oxidizes neptunium to Np(VI) which is extracted by counterflowing solvent. Experiments by Srinivasan [S20] and Koch [K4] suggest that 87 to 97 percent of the neptunium can be recovered in this way.

Uranium, plutonium, and neptunium in the extract are returned to the aqueous phase in the HC stripping unit.

Reduction. To convert plutonium to inextractable Pu(III) and neptunium to still extractable Np(IV), 0.5 M U(IV) reductant is added to the aqueous stream from the HC unit. It is necessary to hold the reacting mixture for half an hour or more to obtain nearly complete reduction of neptunium. This is best done batchwise in a set of reactors, some of which would be reducing while others are receiving feed to be reduced. Reduced fuel is then concentrated to 1.172 M uranium in a set of batch evaporators.

Figure 10.32 Principal steps in Purex process modified for neptunium recovery. Circles indicate relative volume flow rate;——— aqueous;

—— organic.

Partitioning of plutonium. Evaporator product is made 2 M in nitric acid and extracted with four volumes of 30 v/o TBP in the plutonium partitioning unit. This leaves plutonium in the aqueous raffinate and extracts the uranium and neptunium.

Partitioning of neptunium. Uranium and tetravalent neptunium in the extract are separated by fractional extraction with 0.5 M HN03. The less extractable Np(IV) is returned to the aqueous phase while uranium remains in the solvent, from which it can be stripped with 0.01 3fHN03 (not shown).

The process just described has the advantage of providing nearly complete recovery of neptunium. Its principal disadvantages are addition of vanadium to first-cycle wastes and the need to recycle some uranium as U(IV).

The pulse column is an outgrowth of a contactor patented by van Dijk [VI] in which perforated plates extending across the column were oscillated up and down to disperse the phases. Van Dijk also suggested that dispersion might be effected by leaving the plates stationary and pulsating the liquid contents of the column. Pulsating flow of liquid through a column bridged by perforated plates is the principle of the pulse columns in use today.

Figure 4.30 illustrates how a pulse column works. The column consists of a vertical tube packed for the major part of its length with perforated plates. At each end are calming, or disengaging, sections that may be enlarged in diameter, to prevent the carryover of entrainment. The principal interface may be carried at the top of the column, at the bottom, or anywhere in between by adjusting the back pressure applied by the gravity leg, as is done in any extraction column. The pulsating action is supplied by leading a pulse leg from the bottom of the column to the pulse generator. This may be simply a reciprocating pump with check valves removed, or it may be a bellows actuated by a cam. In a small installation, air pressure can be used to actuate the bellows. Larger columns require either hydraulic or mechanical operation.

There are at least two distinct modes of operation of a pulse column. One way is to pulse very gently. This action forces the discontinuous phase through the holes in the plate, forming bubbles. They rise (or fall) to the next plate, where they coalesce to await the next pulse. This is referred to as mixer-settler operation. If the pulsing is more vigorous, the bubbles never coalesce, but are repeatedly forced through holes in various plates as they work their way up and down the column. The passage of a bubble through a hole in a plate deforms it considerably, and the internal agitation thus produced improves the extraction. Also, passage through a hole tends to strip off any stagnant film of the continuous phase. Most pulse columns give optimum performance in this latter region, wherein the bubbles do not coalesce completely between pulses.

Since each cycle of pulsing pushes light phase upward through the holes and then pushes heavy phase downward, the net throughput through the column depends on the pulsing action. In the limit of low frequency and amplitude of pulsing, such that mixer-settler operation occurs between plates on each pulse cycle, the total volumetric flow rate of the light and heavy phases is equal to the product of the volumetric displacement of the pulse generator and the pulse frequency. Thus, throughput increases linearly with pulse frequency until, at higher frequencies, incomplete phase separation occurs between pulses and some back flow of an individual phase occurs in each pulse stroke. Finally, at a higher pulse frequency sufficient emulsification persists throughout the pulse cycle to cause flooding. Further increases in pulse frequency decrease the throughput. Pulse columns usually operate with pulse frequencies in the range of 30 to 120 cycles/min and amplitudes of 1,3 to 5 cm [L2].

Pulse columns provide more efficient phase dispersion and mass transfer than do packed columns, and pulse columns provide more uniform distribution of individual phases across the column cross section, with less tendency toward flow channeling. Pulse columns with phase-contacting sections as large as 0.86 m in diameter [13, L2] and with heights as great as 10 m[Gl, L2] have been used in recovering plutonium from irradiated uranium. A compre­hensive study of the use of pulse columns 4 cm in diameter for the extraction of uranium by TBP has been reported by Durandet et al. [D3]. Operating conditions that resulted in the lowest values of the column height equivalent to a theoretical stage (HETS) of equilibrium contacting are listed in Table 4.16.

Packed columns can also be pulsed to improve their mass-transfer performance, with efficiency increases of as much as 300 percent reported. The flooding limits are lower than for a packed column without pulsing. The pulse amplitude and frequency must not be large enough to float the packing during a pulse, otherwise the packing will lose its random orientation, resulting in flow channeling and loss in mass-transfer efficiency [Dl].

A disadvantage of pulsed columns is the need for a pulsing pump. However, similar pumps are needed in any case for transfer of liquids. In fact, the pulsing can also be used for transfer by installing check valves in the piping between the columns, although with some sacrifice in operating flexibility.

The great improvement in performance of pulse columns over other column-type con­tactors, and the simple and reliable equipment involved, have led to the widespread use of pulse columns in many solvent extraction operations separating and purifying nuclear materials. In addition to their use in some fuel reprocessing operations, as mentioned above, pulse columns have been used in uranium purification plants at Femald, Ohio [Cl], and Gore, Oklahoma (cf. Chap. 5).

NOMENCLATURE

In U. S. plants the aqueous solution of UNH is converted to U03 in two steps, concentration and denitration. In the concentration step the UNH solution is evaporated in a boil-down tank to a syrupy liquid with the approximate composition of the hexahydrate. Three types of equipment have been used for denitration: a heated pot; a fluidized bed; and a stirred, heated

trough. The last was used at Weldon Springs. This consists of an enclosed, heated trough in which a horizontal, rotating agitator keeps the bed thoroughly mixed. The U03 product overflows an adjustable weir and is cooled and collected for subsequent grinding. Water, nitric add, and oxides of nitrogen given off in these steps are recovered and recycled to the dissolver.

In European refineries a different process is used to convert UNH to U03. The uranyl nitrate solution from solvent extraction is neutralized with gaseous ammonia to precipitate (NH4У2 U2 07. This ammonium diuranate is filtered off, dried, and calcined to drive off ammonia and form U03.

Production of pure thorium metal is beset by ah the difficulties cited for uranium metal in Chap. 5, Sec. 10.1, complicated further by the higher melting point of thorium, 1750°C. Table 6.21 lists the principal processes that have been used on a semiindustrial scale to produce thorium metal.

Electrolysis of fused salts. The first electrolytic processes used thorium fluorides, as these are less hygroscopic than thorium chloride. In a process developed for the U. S. AEC [S3], a solution of dried 15 to 20% KThFs in molten sodium chloride was electrolyzed at 800 to 900°C in a graphite anodic cell with a molybdenum cathode under an inert argon atmosphere. As only chlorine was produced at the anode, fluorides accumulated in the electrolyte and required its periodic renewal. A somewhat similar process, involving electrolysis of ThF4 in molten KCl/NaCl has been used in the Soviet Union [Kl].

To permit continuous operation, workers in both the United States [FI] and the United Kingdom [G2] have developed processes for electrolysis of ThCl4 in molten NaCl [Rl] or KCl/NaCl [W4]. By rigorous exclusion of moisture and oxygen, coarsely crystalline thorium of high purity was produced through electrolysis at 800°C in an argon atmosphere.

Reduction with reactive metals. As with uranium, processes for producing thorium by reduction with reactive metals have been developed starting with thorium oxide, chloride, or fluoride. To show which combinations of thorium compound and reactive metal are thermodynamically favorable, the free-energy changes in reducing Th02, ТЬСЦ, or ThF4 by sodium, calcium, or magnesium have been evaluated in Table 6.22 at the temperature used in practice for the respective thorium compound. This table shows that calcium is the only metal capable of reducing ThOj or ThF4, but that any one of sodium, calcium, or magnesium could reduce ТЬСЦ. Magnesium has been preferred because it costs less and can be handled in air without picking up oxygen.

Reduction of Th02. Because the heat of reaction of Th02 with calcium is small, it is incapable of melting the thorium product. However, fmely divided thorium metal powder has been prepared by reacting Th02 and calcium metal in an argon atmosphere. In a U. S. process [F2], calcium chloride was added to promote thorium particle growth. A mixture of Th02, CaCl2, and calcium chips in weight ratio 1:0.4:0.45 was heated to 950°C to start the reaction and held there for 2 to 5 h. In a U. K. process [B8], no CaCl2 flux was used. The reaction was carried out in a CaO-lined, argon-purged metal cylinder at 1200°C. In both cases, the reaction products were cooled and the calcium compounds were leached from the thorium metal by dilute acid.

Reduction of ThCl». Of possible reductants for ThCl4, magnesium is the most convenient. Magnesium alloys with thorium metal and reduces its melting point so that it is possible to produce massive metal instead of a fine, reactive powder. After reaction, magnesium and adherent MgCl2 are removed from the alloy by vacuum distillation. This process is a variant of the Kroll process, which is used commercially for production of titanium and zirconium, with the latter application described in Chap. 7. Its use on a pilot-plant scale for production of thorium at the Albany Station of the U. S. Bureau of Mines has been described by Cuthbert [C6], pp. 182-184. Because of the comparatively small free-energy change with magnesium, it was necessary to use 100% percent excess, and yields of acceptably pure thorium metal were less than 50% of the ThCl4 fed.

Reduction of ThF4. Reduction of ThF4 by calcium is the process used to produce most of the nuclear-grade thorium metal in the United States. The process was developed by workers at the

Ames Laboratory of the U. S. AEC at Iowa State College under the direction of F. H. Spedding, H. A. Wilhelm, and W. H. Keller [S4]. Details of the process have been described by Wilhelm [W2], pp. 78-103, and Cuthbert [C6], pp. 175-180.

Because of the high melting point of thorium and the high heat of formation of ThF4, the metallothermic reduction process used for making massive uranium metal (Chap. 5, Sec. 10.4) will not liberate enough heat to melt thorium, even when calcium is used as reductant. To get around this difficulty, the Iowa workers added ZnCl2 and additional calcium to the charge, to act as a “booster.” The reaction

ZnCl2 + Ca -*• Zn + CaClj

is more exothermic than reduction of ThF4 by calcium, and the extra heat brings the reactants to higher temperature. Use of the booster has two other advantages:

1. Zinc alloys with thorium, reducing its melting point.

2. Calcium chloride reduces the melting point of the CaF2 slag.

The reactor was a flanged steel cylinder lined with dolomitic lime similar to the one used for producing metallic uranium (Chap. 5, Sec. 10.4). It was charged with a mixture of 75.3 kg ThF4, 27.22 kg granular calcium metal, and 7.26 kg anhydrous ZnCl2. This amount of ZnCl2 produced an alloy containing 18 m/o (mole percent) Zn. This amount of calcium was 25 percent more than needed for stoichiometric reduction of the ThF4 and ZnCl2. The excess was needed to drive the reduction of ThF4 to completion.

The charge was covered with a graphite disk and a layer of lime, and a steel cover was bolted to the steel flange. The reactor was placed in a furnace preheated to 660°C. After about 40 min when the charge reached an average temperature around 475°C, the highly exothermic reactions between calcium and ZnCl2 and ThF4 took place, the molten zinc-thorium alloy settled to the bottom, and the molten CaF2-CaCl2 slag rose to the top.

After the reactor was cooled to room temperature, it was opened and the mass of metal was mechanically freed of frozen slag. Ninety percent of the zinc in the alloy was removed by distillation in a retort heated to 1150°C at a vacuum lower than 0.2 Torr. The retort was then filled with argon or helium to prevent oxidation of the spongy thorium and cooled to room temperature. The thorium was transferred to a beryllia crucible in an induction-heated vacuum furnace for melting, evaporation of the residual zinc, and casting into a graphite mold. Thorium metal yield was 94 to 96 percent.

Heat balances for Iowa process. To show the need for addition of the “booster” ZnCl2 to the charge, Table 6.23 shows that when 1 mol ThF4 and 2 mol CaCl2 at 475°C (748 K) react to produce 1 mol liquid Th and 2 mol liquid CaF2 at the melting point of thorium (1750°C or 2023 K), there is an enthalpy deficiency of 12.84 kcal/g-mol thorium. It would thus be impossible to melt the thorium product and obtain massive thorium metal free of CaF2 from these reactants preheated to 475°C.

Table 6.24 shows that 49.57 kcal of heat is available when 1 mol ZnCl2 and calcium at 475°C (748 K) react to form liquid zinc and CaCl2 at 1750°C (2023 K). Thus, simultaneous reduction of 12.84/49.57 = 0.26 mol ZnCl2 and 1 mol ThF4 initially at 475°C with stoichiometric amounts of calcium would bring the products to 1750°C and melt all the thorium.

In the actual Iowa process it was sufficient to heat the products only to around 1360°C (1633 K) because the thorium-zinc alloy produced has a lower melting point than pure thorium. This permitted use of only 0.218 mol ZnCl2/mol ThF4 and provided enough heat to melt the alloy and slag, bring the 25 percent excess calcium to reaction temperature, and still allowed for heat losses.

Table 6.25 gives the material balance for the Iowa process as described by Wilhelm [W2], Table 6.26 gives the heat balance for this process, with reactants at 475°C (748 K) and products molten at 1360°C (1633 K). An excess of 17.7 kcal/mol thorium product was available to compensate for heat losses.

Thermal dissociation of Thl4. Veigel et al. [VI] have prepared massive thorium metal of high purity in lots of several hundred grams each by the Van Arkel-de Boer “hot-wire” process, which has been used for semicommercial production of zirconium as described in Sec. 8.4 of Chap. 7. The process is less suitable for thorium because the thorium metal product is less coherent, so that batch sizes are small. In this process, Thl4 is evaporated at 455 to 480°C in an evacuated vessel containing a metal filament heated to 900 to 1700°C. The iodide dissociates at the higher temperature,

Thl4 -* Th + 2I2

and deposits thorium metal on the heated wire. Temperatures in the range 500 to 900°C must be avoided to prevent formation of nonvolatile Thl2 and Thl3, which are stable in this range [SI].

When fresh thorium is irradiated, 231 Th builds up quickly to equilibrium because of its relatively short half-life of 25.5 h. After a time TR of irradiation, the amount Nn of 231 Pa is obtained by applying Eq. (2.101). For simplicity, we shall assume an essentially constant amount N02 of 232Th during the irradiation and will assume no 230Th in the thorium:

(8.17)

where On >s the effective absorption cross section of 231Pa and a„y2n ‘s the (n, 2n) cross section for 232Th. Even though the 232Th(n, 2n) reaction occurs for neutrons at energies above 6.37 MeV, we may define an effective (n, 2n) cross section such that when multiplied by the thermal flux, the proper (n, 2n) reaction rate is obtained. The effective (n, In) cross section will depend, in part, on the reactor core composition.

Because of its relatively short half-life, 1.31-day 232Pa will be in secular equilibrium with 231 Pa, so that the concentration N22(TR) of 232U as a function of irradiation time TR is given by an extension of Eq. (2.101):

11000-MWe uranium-thorium-fueled HTGR. 95 MWd/kg heavy metal, 38.7% thermal efficiency, 80% capacity factor, 150-day cooling, equilibrium fuel cycle.

* Natural thorium is assumed to contain 100 ppm 230 Th. Discharge thorium is not recycled.

® Includes 59.0 kg/year of second-cycle uranium, from initial makeup 135 U, which is not to be recycled. Composition of discharged second-cycle uranium: 0.8% 234U, 3.6% 23SU, 75.5% 236U, 20.1% 238 U.

^ Plutonium is not recycled.

When uranium is recycled, the initial amount of 232 U for fuel generation n is related to the final concentration from generation n — 1 by

Nb, n=Nn, n-i(TR) (8.21)

where process losses and decay of 232 U in the external fuel-cycle operation have been neglected. For the equilibrium fuel cycle,

=N2i(TR) = N£

and Eq. (8.19) becomes

N22 _ Оп. щОп Г _ /1 — е-°ч9 £гЛ Nm ои(оц-а22) |_ 1 — е-а“в / а„J

where в is the flux time at the end of the irradiation.

For the first “generation” of thorium-uranium fuel, for which 1V®2 =0, Eqs. (8.19) and

(8.23) show that the 232 U content N2 at the end of the first cycle is related to the equilibrium content Nri by

?k = l-e-°„e (8.24)

N22

which assumes the same flux time for all cycles. Equation (8.24) is also valid if 230Th is present as an additional source of 232 U.

In the case of equilibrium recycle, the concentration of 232U in the discharged thorium is the same as that in the makeup thorium containing the recycled uranium. In Fig. 8.12 this concentration is shown as a function of the total flux time of the fuel irradiation. However, during irradiation the 232 U in the fuel decreases below its initial concentration and then recovers as 231 Pa is formed.

5.2 Curium Isotopes

Table 9.25 lists the isotopes of curium important in nuclear technology and some of their im­portant nuclear properties.

242 Cm. The isotope 242 Cm is the largest contributor to the alpha activity of irradiated uranium fuel from power reactors. It is an important source of the 2n + 2 decay chain in the high-level wastes from fuel reprocessing. The alpha activity of 242 Cm results in an internal heat-generation rate of 120 W/g of pure 242Cm. Separated 242Cm, prepared by the neutron irradiation of241 Am, provides a useful alternative for a thermoelectric source and for radionuclide batteries when relatively high outputs are desired over short periods of the order of its half-life of 163 days. For example, a space power generator denoted as SNAP-11 contained 7.5 g of 242Cm and produced 20 W of thermoelectric power.242 Cm is also the decay source of 238 Pu, which is used as a longer — lived radioisotope heat source.

To optimize isotopic purity when producing 242 Cm it is desirable that the irradiation of the 241 Am target be carried out at low neutron flux. At higher fluxes the increased chain branching by fission of 242 Am and the increased neutron capture in 242 Cm result in greater con­tamination by 243Cm per unit amount of 242Cm produced. The actual production rate of 242Cm optimizes at a neutron flux of about 8 X 1014 n/(cm2 — s). At higher fluxes the increased chain branching from 242 Am fission more than offsets the increased rate of neutron absorption in 241 Am [K2].

243 Cm. The isotope 243 Cm is an alpha emitter with a half-life of 32 years. About 0.3 percent of its decays occur by beta emission, and the accompanying gammas contribute to shielding problems when 242 Cm heat sources are contaminated with 243 Cm.

Reaction with 2200 m/s
neutrons

249 Cm. The isotope 249 Cm is the 64-min beta emitter that terminates the curium chain formed by neutron irradiation. It decays to 249 Bk.