The first nitrogen based systems investigated in France and in Europe to selectively extract An(III) from nitric acid solutions consisted of synergistic binary mixtures of tridentate polyazine ligands and organophilic acids. They played the role of cation exchangers by extracting the cationic complexes formed between the An(III) and the tridentate polyazine ligands, through proton exchange. As illustrated in Fig. 11.11 and more deeply detailed in

11.11 Examples of nitrogen-donor ligands used in synergistic mixtures with cation exchangers.

the comprehensive reviews of Kolarik (2008) and Ekberg et al. (2008), the tridentate polyazine ligands studied belonged to the families of the:

• terpyridines (Drew et al., 2000a);

• monopyridyl-1,2,4-triazines (Hudson et al., 2003a) and dipyridyl-1,2,4- triazines (Boubals et al., 2002, Hudson et al., 2003a, b, Drew et al., 2006);

• dipyridyl-1,3,5-triazines (Drew et al., 2000b) and tripyridyl-1,3,5-tri — azines (Chan et al., 1996, Cordier et al., 1998, Hagstrom et al., 1999);

• dioxa-/benzoxa-/binzimida-/zol-pyridines (Kolarik and Mullich, 1997, Andersson et al., 2003, Weigl et al., 2003, Drew et al., 2004b);

• amino-dipyridyl-1,3,5-triazines (ADPTZ).

These tridentate nitrogen-donor ligands all provided moderate selectivity of complexation toward An(III), with selectivity factors versus Ln(III), SFAn/Ln, ranging from 5 to 20 at most, as demonstrated by Miguirditchian et al. (2005, 2006), who investigated the thermodynamics of M(III): ADPTZ complex formation (M = An or Ln) and pointed out the more exothermic complexation reactions of An(III) than Ln(III), thus highlighting the assumed greater covalent character of the N-An(III) bonds, which is sup­ported by quantum chemistry calculations. The X-ray crystal structures of metal : terpyridine solvates resolved by Berthet et al. (2002a) also revealed that the U-N(central pyridine) distances are shorter than the U-N(distal pyridines) distances, while the reverse is true for lanthanide compounds. It has been suggested that these differences reflect the presence of a n back — bonding interaction between U and terpyridine.

Unfortunately, the use of carboxylic acids (e. g., octanoic or alpha-bromo — capric acids) and even dinonylnaphthalene-sulphonic acid (HDDNS), as sources of lipophilic anions in synergistic mixtures, was of little help in discriminating between An(III) and Ln(III) from aqueous feeds whose acidity exceeded 0.1 mol. L-1, as is the case for DIAMEX An(III)-product solutions. Therefore, the only two reported counter-current tests imple­mented in laboratory-scale mixer-settlers used surrogate spiked feeds of low acidity. The first one was carried out by Vitart et al. in 1986 and made use of the synergistic mixture composed of TPTZ (Fig. 11.11, R = — H) and HDNNS, dissolved in carbon tetrachloride. The second one was carried out in 2000 with the binary mixture composed of 2-(3,5,5-trimethylhexanoyl- amino)-4,6-di(pyridin-2-yl)-1,3,5-triazine (TMHADPTZ, Fig. 11.11) and octanoic acid (Madic et al., 2002). Although the experimental profiles were in good agreement with the calculated flowsheet, the latter system has not been further developed because of its high sensitivity to pH variations and the necessity to buffer the feed acidity (pH > 3).

Within the NEWPART collaborative project (Madic et al., 2000), the discovery of the bis-(1,2,4-triazinyl)-pyridines (BTP) synthesized by Zdenek Kolarik appeared as a real breakthrough in the design of nitrogen-donor extractants selective toward An(III). Surprisingly, BTP ligands firstly pre­cipitated when combined with organophilic carboxylic acids (Kolarik et al., 1999a). Even more unexpectedly, these tridentate compounds, consisting of two dialkylated 1,2,4-triazines attached sideways to a central pyridine (Fig. 11.12), proved to extract Am(III) from acidic aqueous solutions ([HNO3] > 1 mol. L-1) when dissolved in polar diluents, such as mixtures of HTP and alcohols. Furthermore, their selectivity was shown to be amazingly high as compared to any of the aforementioned nitrogen-donor ligands: SFAm/Eu > 100 (Kolarik et al., 1999b).

Various BTP ligands have been further synthesized in the framework of NEWPART and PARTNEW collaborative projects (Madic et al., 2000, 2004). The parametric and thermodynamic studies carried out on these molecules revealed that the nature (e. g., normal or branched alkyl chains, aromatic rings) and the length, as well as the position of the substituting groups, not only influence the complexing and extracting properties of BTP compounds (An(III)/Ln(III) selectivity and kinetics of extraction), but also their chemical stability (Iveson et al., 2001, Colette et al., 2002, Hill et al., 2002, Hudson et al., 2003a, Weigl et al., 2006).

Several reasons could explain the peculiarity of BTP’s extracting behaviour:

• The unsymmetrical positions of the nitrogen atoms within the lateral triazine rings, which give these compounds low and great affinities respectively for protons and An(III) (Petit et al., 2006). As opposed to bis-(1,2,4-triazinyl)-pyridines, symmetrical bis(1,3,5-triazinyl)-pyridines present bad extraction properties (Drew et al., 2004c).

• The M :L3 (metal : ligand) stoichiometry of the extracted complexes pointed out by different structural studies (Fig. 11.13, Drew et al., 2001a, b, Berthet et al. 2002b, Colette et al., 2003), which results in the fully saturated inner coordination spheres of the extracted trivalent f cations, wrapped by three triply bonded BTP ligands, hence leaving no spare space for extra water or nitrate coordinating molecules (Colette et al., 2004, Deneke et al., 2005, 2007).

• The chemical reactivity of the carbon atoms located on the a-positions of the alkyl groups grafted onto the lateral triazines. Nitrous acid and oxygen tend to oxidize non fully substituted carbon atoms, leading to hydrophilic alcohol and ketone derivates, which degenerate the intrinsic extraction properties of BTP ligands, as experienced during the demon­stration hot tests carried out with 2,6-bis-(5,6-di-n-propyl-1,2,4-triazine — 3-yl)-pyridine (nPr-BTP, dissolved at 0.04 mol. L-1 in a mixture of HTP and n-octanol (70/30 vol.%)) on genuine DIAMEX An(III)-product solutions, both at the CEA Marcoule (France) and the ITU (Karlsruhe, Germany) in 1999.

FL .N. J-L — Y. Ns .R Y Y N T Y rAn"n N’fA 2,6-Bis-(5,6-dialkyl-1,2,4-Triazin-3-yl)-Pyridines (BTP) JL Ns. YY Y n n Y/ n’nYU Bis-Annulated-Triazine-Pyridine (BATP): CyMe4-BTP Bis-Annulated-Triazine-Pyridine (ВАТР): фСуМе4-ВТР ) Vn n- N=( Yn r4 ,n _>r /—N N Y R R YY ) Yn n4 ,,N 6,6’-bis(5,6-dialkyl-[1,2,4]-triazin-3-yl)-[2,2’]-bipyridines (BTBP) Bis-Annulated-Triazine-Bis-Pyridine (BATBP): CyMe4-BTBP 77.72 Polypyridine-triazine extractants developed for An(lll)/Ln(lll) separation.

• The longer and bulkier are the alkyl chains grafted onto the lateral triazines of the BTP compounds, the slower are their kinetics of mass transfer, probably due to adsorption difficulties at the interface (BTP molecules are not assumed to aggregate). Therefore, the formulation of BTP based solvents most often requires the addition of a mass transfer catalyst (such as a diamide) to aid phase transfer and thus enhance kinetics of extraction and stripping (Hill et al., 2002). However, this additive tends to decrease the selectivity of the BTP toward An(III).