Reprocessing Technologies

Of particular interest for sustainability in nuclear energy is reprocessing of spent (or depleted) nuclear fuel. There are three main reasons to reprocess spent nuclear fuel: (1) to recycle the plutonium and uranium and reuse them in a nuclear reactor that can handle this type of mixed oxide fuel (MOX), thus increasing the efficiency of fuel consumption, (2) to decrease the volume of high-level waste, and (3) to sepa­rate out the minor actinides. As noted above, the minor actinides Np-237, Am-241, and Cm-244 are the most hazardous in terms of long-term radiotoxicity. Using the radiotoxicity of naturally occurring uranium as a baseline, it takes around 350,000 years for spent fuel to return to the same level as natural uranium ore. On the other hand, if the plutonium and uranium isotopes can be removed from the spent fuel waste stream, that same level of radiotoxicity can be reached in roughly 5000 years. If plutonium, uranium, and the minor actinides are removed, the radiotoxicity drops relatively quickly to match that of natural uranium in a relatively short period of 500 years (The Royal Society 2011). Furthermore, if the minor actinides can be separated and concentrated as their oxides, they can be converted in a separate nuclear process into less harmful nuclei, a process referred to as partitioning and transmutation (P & T). Thus, reprocessing of spent nuclear fuel is a very important goal—but it is replete with political repercussions due to the presence of Pu-239. Any reprocessing scheme could, potentially, separate out the Pu-239 for recovery and use in nuclear weapons. It is primarily for this reason that the United States no longer has any reprocessing facilities and does not reprocess spent nuclear fuel, unlike several other countries around the world. (For example, the French reprocessing technology is able to extract over 99% of both uranium and plutonium (Bodansky 1996).)

Reprocessing spent nuclear fuel is basically fractionating it in such a way as to have separate waste streams that can be reused, treated, or disposed of. An overview of spent nuclear fuel separation schemes is given in Figure 9.6. There are huge technological difficulties associated with these separations. Some of the waste components are volatile, some are neutron emitters, plutonium-239 pres­ents a proliferation risk, and the minor actinides are tremendous radiotoxicity haz­ards. As can be imagined, the technologies that accomplish these separations are

not trivial and it is far beyond the scope of this text to go through all of them. Instead, we will selectively focus on (1) the historical process, PUREX, which has been used for decades and (2) two small-scale processes for the separation of lanthanides from actinides (TALSPEAK and SANEX) and some recent research developments in this context.

PUREX stands for plutonium uranium refining by extraction and begins with the chopping and dissolution of the used reactor fuel in nitric acid to form a witch’s brew of dissolved ions, including actinides (UO2+ and Pu4+, among others, includ­ing the minor actinides) and lanthanides and various fission products (iodine-129, technetium-99, cesium-135, strontium-90, etc.). Some volatile elements (iodine, krypton, and xenon) are given off and captured during the dissolution process. The resulting acidic and highly radioactive solution is then extracted with a mixture of 30% tributyl phosphate (TBP, Figure 9.7) in a nonpolar organic solvent (kerosene or dodecane) to selectively extract the plutonium and uranium compounds into the organic phase as neutral complexes with tributyl phosphate [UO2(TBP)2(NO3)2] and [Pu(TBP)2(NO3)4] (Whittaker et al. 2013). The remaining fission products (the lan­thanides and the minor actinides) stay in the acidic aqueous phase (Whittaker et al.

2013) . The subsequent separation of U from Pu may or may not be then undertaken depending upon the country’s policy toward nuclear proliferation risk (WPFC Expert Group on Chemical Partitioning of the NEA Nuclear Science Committee 2012).

Improvements in reprocessing technology have developed over the years and continue to be an active area of investigation. Several reprocessing options have been developed to address the fundamental issues associated with recycling of spent nuclear fuel and are identified by a plethora of acronyms:

• SANEX—selective actinide extraction

• TRUEX—transuranic extraction (the transuranic elements are those of atomic number >92)

• DIAMEX—diamide extraction (removal of nonlanthanide fission products)

• TALSPEAK—trivalent actinide-lanthanide separation by phosphorus reagent extraction from aqueous komplexes

Others include UREX+, NUEX, COEX, GANEX, and so on. These processes basi­cally vary in regard to exactly which chemicals are separated out: all of the minor actinides, only uranium, uranium plus plutonium, and so on.

The separation of lanthanides from actinides in the waste stream is necessary so that transmutation can be carried out to reduce the radiotoxicity of the waste stream. The high neutron absorption capability of some of the lanthanides makes transmutation less

O

CH3CH2CH2CH2O————————- P OCH2CH2CH2CH3

OCH2CH2CH2CH3

feasible if the lanthanides cannot be separated out (Hudson et al. 2012). Many of the processes listed above have been developed to address this challenge to remove enough of the actinides (>99%) in high purity to outweigh the costs of this separation process.

The lanthanide-actinide separation is particularly difficult because of the simi­larity of the An(III) (actinide3+) and Ln(III) (lanthanide3+) behaviors. The trivalent actinides are very slightly more covalent (presumably due to the presence of 5f instead of 4f orbitals although the origin of this behavior is not well understood) and therefore interact more favorably with soft ligands like nitrogen and sulfur. The TALSPEAK process separates Ln(III) from An(III) as follows: a lipophilic ligand, bis-2-ethyl(hexyl) phosphoric acid (HDEHP, see Figure 9.5), is used to extract Ln(III) ions into an organic phase while diethylenetriamine-AWN’,M’,M’-pentaacetic acid (DTPA) complexes with the An(III) ions and retains them in the aqueous phase (Figure 9.8) (Braley et al. 2011). The TALSPEAK process is buffered with lactic acid and run at a pH of 3.5-3.6. The competition between the lipophilic extraction of the lanthanide ions and the aqueous retention of the actinide ions is revealed in the reaction and equilibrium coefficients for the aqueous phase (Equation 9.15) and organic phase (Equation 9.16). In this example L stands for the ligand DTPA and the acid, HA, is HDEHP with AHA representing the HDEHP dimer. An example of a M(AHA)3(org) complex for americium is given in Figure 9.9.

M3+ + H3L2- ^ ML2- + 3H+

K [ML2- ] x [H+]3 (9.15)

eq “ [M% x [H3L2-]

M3a; + 3(HA)2(org) ^ M(AHA)3(OTg) + 3H+

K _ [M(AHAU, g x [H%+ СШ)

* [M3+]a? x [(HAUrg

CH2CO2H

HO2CCH^.

CH2CO2H CH2CO2H

DTPA

Retains An3+ in aqueous phase

Combining these relationships gives the overall equilibrium for conventional TALSPEAK (Equation 9.17):

MLaq + 3(HA)2(org) ^ M(AHA)3(org) + H3Laq

The TALSPEAK process, while effective, is highly complex and has room for improvement; thus, research to tailor the chemical behavior of the lipophilic phos­phoric acid agent and the hydrophilic complexing agent is ongoing. For example, substitution of triethylenetetramine-N, N,N,,N"N"’,N"’-hexaacetic acid (TTHA, Figure 9.8) for the pentaacetic acid reagent DTPA resulted in a slight improvement in the separation of the actinides from heavy lanthanides but poorer efficiency in separation of the lighter lanthanides (Braley et al. 2011).

The SANEX process is a related solvent extraction process that is implemented after the removal of Pu and U by PUREX and removal of the non-lanthanide fission products by the DIAMEX process. In the SANEX process, a bis-(1,2,4-triazine) ligand (e. g., 6,6,-bis(5,5,8,8-tetramethyl-5A7,8-tetrahydro-1,2,4-benzotriazin-3-yl)2,2′- bipyridine (thankfully abbreviated as CyMe4-BTBP, Figure 9.10)) is used in con­junction with phase-transfer agent NA’-dimethyl-NA’-dioctyl-2,(2′-hexyloxyethyl) malonamide (DMDOHEMA, Figure 9.10) to selectively extract Ln(III) into an organic solvent. SANEX ligands must be able to work well in solutions of low pH for which the 1,2,4-triazines are particularly well suited due to their weak basic­ity. Further, they must be lipophilic and stable to both radiolysis and hydrolysis. To that end, the presence of the four methyl groups of CyMe4-BTBP is not arbitrary; it was found that the benzyl positions must be blocked to prevent decomposition of this ligand. Research efforts to improve the selectivity and effectiveness of this ligand has led to the development of many different possibilities (Hudson et al. 2012) with the more rigid phenanthroline triazine CyMe4-BTBPhen (Figure 9.10) being particularly effective. An ORTEP drawing of the europium(III) ion complexed with CyMe4-BTBPhen is shown in Figure 9.11 (Hudson et al. 2012). The logical next step in further development of these reprocessing technologies is attachment of the appropriate complexing agents to a solid support to simplify the technical aspects of

the separation, research that is already in progress (see, e. g., Raju and Subramanian (2007)). New developments in nuclear reactor technology (Generation IV, the topic of the next section) will require new methods of fuel reprocessing, such as electro­lytic reprocessing (also known as pyroprocessing), a topic left to the interested reader for additional investigation.

Updated: September 26, 2015 — 1:06 pm