Partitioning and Transmutation of Minor Actinides and Fission Products


Partitioning is a complex series of chemical and/or metallurgical operations, intended to separate selected radiotoxic nuclides or groups of nuclides occluded in the spent fuel elements. The separated fractions can be stored or transformed into irradiation targets. Partitioning of a specified element is an indispensable requisite for its transmutation. The partitioning may limit the potential of transmutation, and thus is of great importance.

The concept of P&T comes originally from the idea that highly toxic nuclides should be transformed into stable or shorter-lived ones. On the other hand, a positive concept has also been proposed. That is the utilisation of several elements. For example Ru and Rh might be used as catalysts after sufficient cooling period. Strontium (90Sr) and caesium (137Cs) are strong gamma-ray emitters and can be used as gamma-ray sources.

The partitioning aims at separating minor actinides (MA: Np, Am and Cm) and several fission products (FPs: Tc, I, Sr etc.). One of the difficulties of the partitioning technique can be ascribed to the similar chemical properties of Am, Cm and lanthanides (Lns) such as Ce, Nd, Eu etc. In particular the fractions of Am and Cm in spent nuclear fuels are smaller than those of the lanthanides which are part of FPs, although precise figures depend on neutron spectra. In the transmutation process, the lanthanides act as poisons in thermal neutron spectra and influence reactivity coefficients limiting the operationial and safety aspects of transmutors LWRs, FRs, ADSs).

For transmutation purposes in LWRs, a high selectivity and high separation yield are required for the partitioning of the MAs. This high selectivity is mainly based on the combined effects of fuel cycle impacts (fuel fabrication, shielding) and the request for small Pu-contents in the final waste going to geological disposal.

At present the PUREX (Plutonium and Uranium Reduction and Oxidation) process is employed on industrial basis. The advanced aqueous reprocessing method could separate Am and Cm in future.

Pyrometallurgical process may be more feasible for high burnup fuels or fast reactor spent fuels.

Isotopic separation is in some cases needed in order to achieve effective transmutation. However, isotopic separation is not handled here.

The PUREX Process

The PUREX process is a solvent extraction process based on the use of tributyl phosphate (TBP), which extracts actinide cations in even oxidation states IV and VI. The complex formed is as follows;


where M is the metallic cation and A an anion, generally nitrate ion.

In nictric acid media, U and Pu exist most stable in the oxidation state of respectively VI and IV. Thus, U and Pu are co-extracted by TBP and separated from FPs which remain in the aqueous phase. Uranium and plutonium are recovered with an industrial yield close to 99.9 %. On the other hand, Am and Cm are stable in the oxidation state III in acidic media, and cannot be extracted by TBP. At present Am and Cm are conditioned in a glass matrix together with the FPs. Concerning Np, the stable oxidation state is V. However, Np can be oxidized into VI in the presence of nitrous acid, and accordingly extracted in the organic phase.

The separation of FPs is not intended at present in the existing reprocessing plants. However, 129I and 99Tc could be separated during the PUREX process. Iodine-129 is released into the ocean as medium-level liquid waste, but the scrubbing could result in a stable matrix. Technetium-99 occurs both as insoluble residue from the dissolver process and as soluble TcO4- in HNO3. The latter could be eliminated quantitatively by scrubbing the TBP stream with concentrated HNO3. This fraction of 99Tc could be combined with the insoluble one.

Advanced aqueous reprocessing

Within the PUREX process only U and Pu are recovered at present, and Np, I and Tc could be separated by the modified PUREX process. The other elements cannot be retrieved by the PUREX process anyway.

In the frame of P&T technology, the targets of the partitioning should satisfy at least one of the following two conditions:

In other words the partitioning would be meaningless without benefit.

In this context, the targets are MA, particularly Am, I and Tc. At present the most efforts are devoted to the development of novel methods for the recovery of Am. The removal of the strong gamma-ray emitters, Sr and Cs, has also been studied.

Modification of PUREX process

As described above, the PUREX process could separate Np, I and Tc if some improvements were installed. The separation of the other elements from HLLW requires special off-line techniques. For a decade R&D on this subject has been performed in the US-DOE laboratories (ANL, SRP and LANL), in Japan (JAERI and JNC), in France (CEA-FAR & Marcoule), in Russia (Sint-Petersburg), in China and within the national nuclear laboratories in Europe and the laboratories of the European Commission (TUI).
At present the advanced separation methods aim prevalently at the recovery of MAs. Fission products separation has been studied only for I, Tc, platinum group metals (PGM) and g-ray emitters such as Sr and Cs.

TRUEX process

In the TRUEX (TRansUranium EXtraction) process developed at ANL (USA), a powerful extractant, n-octyl-phenyl-di-isoburyl-carbamoylmethyl-phosphine-oxide (CMPO), is employed. The TRUEX solvent is a mixture of 0.2 M CMPO and 1 M TBP. CMPO desplays high and low affinities for An(III) and Ln(III) nitrates at high and low nitric acid concentration, respectively. Consequently, an extraction-stripping cycle can easily be designed.

The TRUEX process is applicable to HLLW solutions with HNO3 concentration of 0.7 to 5 M, which results from the conventional PUREX process. It was demonstrated at JNC (Japan) that the process was efficient for real HLW solutions.

There are several problems on the TRUEX process. The first drawback is the effect of solvent degradation products, some of which are cation exchangers and prevent the stripping of An(III)+Ln(III). The second is the difficulty in stripping U(VI) and Pu(IV) from the solvent because of their high affinity with CMPO. The other problems are the separation of Lns from MAs and the volume of secondary wastes.

DIDPA process

At JAERI (Japan) a process using di-isodecylphosphoric acid (DIDPA) was developed. The solvent is a mixture of 0.5M DIDPA and 0.1M TBP. The MAs and Lns are extracted by DIDPA, and An(III)/Ln(III) sepration is achieved by selective stripping of An(III). The advantage of the DIDPA process is the incorporation of Np into the Am-Cm component. In the presence of hydrogen peroxide, Np(V) is reduced to Np(IV) and thus Np is extracted together with other Ans and Lns. On the other hand, the shortcoming is that it cannot be applied to concentrated liquid HLW without denitration which leads to precipitation of Ans and reduced separation yield.
More than 99.95 % recovery of all actinides was demonstrated with a simulated HLLW, and 99.99 % recovery of Am and Cm with real HLLW.

TALSPEAK process

The TALSPEAK process is similar to the DIDPA process. In the TALSPEAK process the extractant is di-2-ethyl-hexyl-phosphoric acid (HDEHP).

TRPO process

The TRPO process was developed in Tsinghua University (China). The solvent is a mixture of trialkyl phosphine oxide (TRPO) soluble in aliphatic hydrocarbon diluent. The affinity of the TRPO extractant for An(III) and Ln(III) is high and low in low (1M) and high (5M) nitric acid concentration, repectively. Thus, the neutralization of nitric acid is necessary. The process was successfully tested at the ITU of Karlsruhe with diluted HLLW solution. The advantage is the reversibility of extraction and stripping, the miscibility with TBP and loading capacity. The drawbacks are the interference of the separation by some FPs such as Zr, Mo, Ru and Tc, and the solution of An+Ln mixture in high nitric acid concentration.

CYANEX301 process

CYANEX301 consists chiefly of bis (2,4,4-trimethylphentyl) dithiophosphinic acid. The commercial product, which contains many impurities, does not show An/Ln separation. But, after purification, CYANEX301 shows greatly high efficiency for An(III)/Ln(III) separation. One of the drawbacks is that it works well only in a rather high pH from 3.5 to 4.

CYANEX301 shows the synergistic effect when it is used together with TBP: higher separation factor at pH < 2. The successful separation of Am(III)/Ln(III) using real waste has been reported. Modolo et al. synthesized a new extractant by replacing 2,4,4-trimethylpentyl group of CYANEX301 with halogenated phenyl group. An extractant (Cl-C6H4)2PS(SH) showed good extraction properties and chemical stability when it was used with trioctylphosphine as synergist.

DIAMEX process

The DIAMEX (DIAMide EXtraction) process was first developed by Musikas et al. at CEA Fontenay-aux-Roses Research Centre (France) and by C. Madic and M. J. Hudson in a joint European research programme involving the CEA (Fontenay-aux-Roses) and the University of Reading (UK). This process is based on the use of malonamide extractants, which are completely incinerable. A reagent di-methyl-di-butyltetradecylmalonamide (DMDBTDMA), which is used in solution in an aliphatic diluent, has extracting properties similar to CMPO. Optimization of the diamide formula is underway. The DIAMEX process was tested successfully in 1993 with real waste.

SESAME process

The SESAME process developed at the CEA aimes at selective separation of Am. Americium can be electrochemically oxidized to the oxidation state IV or VI in aqueous media, and then madke complex with phosphtungstate, P2W17O6110-. The Am could be separated, (i) by organophosphorus solvents or amides for Am(VI), (ii) by amides for Am(IV) in complexed form, or (iii) by nanofiltration of the complexed species of Am(VI).

The following three applications are planned:

Separation of FPs

Separation of Sr and Cs

The adsorption method with inorganic exchangers, titanic acid and zeolite, has been developed for the separation of heat generating nuclides such as 90Sr and 137Cs. More than 99.9 % recovery from real HLLW was successfully demonstrated. The inorganic exchangers loaded with Sr and Cs can be solidified into a stable form by direct calcination at high temperature.
Cobalt dicarbollides were first synthesized in Czechoslovakia. Since the dilution by toxic nitrobenzene was originally needed, the modified dicarbollide was sought. Promising results were obtained with bis-ylene cosan (BISPHECOSAN) diluted in nitrophenyl-alkyl-ether (NPHE or NPOE) or in solubilizers such as diethylpropanesulfonamide (DEPSAM) or dibutylmethanesulfonamide (DIBUMESAM).

Separation of Cs

In the separation of Cs, the most difficult problem is the similar chemical property of Cs+ to Na+. But the ionic radii are different. Hence, the concept of the Cs separation is based on the use of cyclic molecules named calix-crowns, which have proper sites adjusted to the dimension of Cs+ ion. It was reported that Cs was sufficiently separated with the calix-crowns from a real high-level effluent.

Separation of Sr

The SREX (StRontium EXtraction) process complemented the TRUEX process for the strontium removal from acidic HA liquid waste. Horwitz proposed the lipophilic di-t-butylcyclohexano 18-C-6 (0.2M) diluted in hydrocarbon mixtures.

Separation of Tc and PGM

The two methods have been developed at JAERI for the Tc and PGM separation from the DIDPA raffinate: the precipitation by denitration with formic acid, and the adsorption method by active carbon. The Tc recovery yield is higher in the latter while the secondary waste production is smaller in the former.

Pyrometallurgical reprocessing

The pyrometallurgical reprocessing is promising in particular for high burnup fuels or FR fuels, since molten salts have higher resistivity to radiation exposure than aqueous media. The shortcoming is lower recovery yields than aqueous methods. The corrosion of materials induced by high temperature molten salts is severe, and should be solved in future.

Pyrometallurgical reprocessing of oxide fuels for oxide fuel cycle

The Research Institute of Atomic Reactor (RIAR) in Russia has been developing a pyrometallurgical reprocessing method based on the electrorefining of oxides combined with vibro-pack fuel fabrication. There are two methods:

The DOVITA (Dry reprocessing, Oxide fuel, Vibropac, Integral, Transmutation of Actinides) Programme is planned. The purpose is incorporation of the separation process of MAs. In molten chloride salts Np behaves like Pu, and thus is deposited together with UO2 and PuO2. Americium and curium act similarly to the rare-earth elements (REs), thus it is difficult to separate Am and Cm from REs. Under high oxgen partial pressure, AmO+ is generated, which is utilized for the co-deposition of Am with UO2-PuO2.

Pyrometallurgical reprocessing of metal and oxide fuels

The basic pyrometallurgical process developed at ANL consists of anodic dissolution of spent fuel, partial recovery of U on a solid cathode, and electrolysis of P with the remaining U into a liquid cadmium cathode, using LiCl-KCl media above 500 °C. Oxide fuels to be treated by this process must be first reduced to metals. For this purpose the lithium process is promising.

Pyrometallurgical reprocessing of TRU from HLLW

The process investigated at CRIEPI to recover TRU elements from high-level PUREX waste consists of denitration to oxides, chlorination, reduction extraction and electrorefining in an LiCl-KCl/Cd or LiCl-KCl/Bi system. A small-scale experiment showed that over 99 % of each actinide could be recovered from a simulated waste. The waste produced throughout the process is expected to be minimal.

Related links

P&T projects and links

P&T publications and documents

P&T contacts

P&T expert group