English

• Nuclear power as a clean energy source is now a strategically important option for many countries, which calls for the advanced nuclear fuel cycle models to achieve efficient utilization of nuclear fuel and reduce environmental hazards [1,2]. Although the majority of plutonium and uranium in spent nuclear fuels can be extracted through the well-established plutonium uranium reduction extraction process, its radioactive residue still contains long-term radiotoxic minor actinides (MAs) [3]. The partitioning and transmutation (P&T) strategy is widely accepted as an effective method to convert MAs into shorter-lived or stable nuclides by neutron irradiation [4]. The co-existing lanthanides, having large neutron absorption cross sections, can significantly reduce the efficiency of the P&T and therefore need to be separated from the MAs [5]. The trivalent actinide (An(Ⅲ)) and lanthanide (Ln(Ⅲ)) ions, however, have similar physicochemical properties, making their separation extremely challenging [6,7].

  In general, solvent extraction is a popular f-element extraction method with scalability and robustness up to now [8–11]. An(Ⅲ) and Ln(Ⅲ) ions are "hard" Lewis acids based on the Hard-Soft Acid-Base theory (HSAB) [12], the former is slightly softer than the latter. Many soft S- or N-donor ligands have been studied for their selectivity for An(Ⅲ)/Ln(Ⅲ) ions through solvent extraction [13–15]. According to the CHON principle [16], N-donor ligands have received more attention, e.g., 2,6-bis(5,6-dialkyl-1,2,4-triazin-3-yl)pyridines derivatives (BTPs, Scheme 1) [17], and 2,9-bis(1,2,4-triazin-3-yl)-1,10-phenanthroline derivatives (BTPhens) [18]. Bis-triazinyl ligands have achieved excellent extraction ability, but there are some practical drawbacks, such as poor chemical and radiological stability, slow extraction kinetics, and difficulty in synthesis, which still limit their wide application [13]. In recent years, polyazole based ligands, especially those with 1,2,4-triazol [17,19], pyrazol [20–24], and 1,2,3-triazol [25–31] side chains, have received a lot of attention owing to their relatively easy synthesis [13] and high extraction kinetics [26]. Kolarik et al. reported that 2,6-bis(5–butyl–1,2,4-triazol-3-yl)pyridine (DBTZP, Scheme 1) in the presence of 2-bromohexanoic acid could extract Am(Ⅲ) with Am(Ⅲ)/Eu(Ⅲ) separation factors (SF_Am/Eu) up to 150 [17]. Bremer et al. synthesized 2,6-bis(5-(2,2-dimethylpropyl)-1H-pyrazol-3-yl)pyridine (C5-BPP, Scheme 1), which could achieve SF_Am/Eu about 100 under 1 mol/L HNO₃ with the existence of 2-bromohexanoic acid [20]. Ding's group has done extensive works on the synthesis of pyrazol based ligands [21–23], in which 2,9-bis(5-alkyl-1H-pyrazol-3-yl)-1,10-phenanthroline (BPPhen, Scheme 1) [23] exhibits SF_Am/Eu value of about 100. Kiefer et al. reported that the lipophilic 2,6-bis(1-(ptolyl)-1H-1,2,3-triazol-4-yl)pyridine (BTTP, Scheme 1) exhibits a high complexation selectivity with Cm(Ⅲ) than Eu(Ⅲ) in CH₃CN [25]. Subsequently, Macerata's group synthesized hydrophilic pyridine-2,6-bis(1H-1,2,3-triazol-4-yl) (PyTri, Scheme 1) with high extraction kinetics and chemical/radiolytic stability, which has the SF_Eu/Am values of about 60–150 [26]. In addition, Ye's group proposed a process for An(Ⅲ)/Ln(Ⅲ) extraction separation, with N, N, N', N'-tetraoctyl diglycolamide (TODGA) as the extractant, tri-n–butyl phosphate (TBP) as the phase modifier and PyTri as hydrophilic stripping agent, the 'hot test' could achieve 99.92% removal of Am(Ⅲ) with SF_Eu/Am of 3.8 × 10³ [31]. The first 1,10-phenanthroline-derived hydrophilic ligand 2,9-bis-(1H-1,2,3-triazol-4-yl)1,10-phenanthroline was reported to exhibit SF_Eu/Am ≈ 36–47 and SF_Cm/Am ≈ 2.5 [27]. Ossola's group synthesized five novel lipophilic derivatives of PyTri with different alkyl groups [28], among which 2,6-bis(1-(2-ethylhexyl)-1H-1,2,3-triazol-4-yl)pyridine (L², Scheme 2) with ethylhexyl alkyl chain showed the best solubility and separation performance. Despite the great progress in the research of polyazole based ligands in this area, the understanding of how polyazole side chains affect separation performance is still very limited.

  Scheme 1

  

  Scheme 1.  Structures of N-donor ligands.

  DownLoad: Full-Size Img PowerPoint

  Scheme 2

  

  Scheme 2.  Structures of target pyridine-derived polyazole ligands.

  DownLoad: Full-Size Img PowerPoint

  With the rapid development of actinide computational chemistry, theoretical calculations can provide reliable results to elucidate the solvent extraction of An(Ⅲ)/Ln(Ⅲ) ions [32–35], which further expand our understanding of the electronic structure, bonding nature and spectra of the complex actinide systems [36–39]. We have theoretically explored a series of studies on the Am(Ⅲ)/Eu(Ⅲ) separation by N-donor and N, O-mixed donor ligands [40–48] by considering the factors of hard-soft donor combined strategy [40], the variations of the central heterocyclic moieties [41], preorganization [42,43], and substitution effects [44,45]. In addition, we have also investigated the Am(Ⅲ)/Eu(Ⅲ) separation by hydrophilic ligands [46,47]. Recent work [48] reported the effect of different side chains containing imidazol, tetrazol, and 1,2,4-triazol, which showed that the phenanthroline derivative with 1,2,4-triazole side chains has the best Am(Ⅲ)/Eu(Ⅲ) selectivity. Experimentally, pyrazole, 1,2,3-triazol and 1,2,4-triazole have better Am(Ⅲ)/Eu(Ⅲ) separation than imidazol and tetrazol [49]. To further explore the effect of pyrazole, 1,2,3-triazole and 1,2,4-triazole side chains on the Am(Ⅲ)/Eu(Ⅲ) separation, we attempts to design polyazole-based ligands based on above works. Therefore, the Am(Ⅲ)/Eu(Ⅲ) selectivity of three similar pyridine-derived polyazole ligands, 2,6-bis(5-ethylhexyl-1,2,4-triazol-3-yl)pyridine (L¹), L², and 2,6-bis(5-(2,2-ethylhexyl)-1H-pyrazol-3-yl)pyridine (L³) (Scheme 2) were investigated with scalar relativistic density functional theory (DFT). L¹ and L³ have similar structures to C5-BPP and DBTZP (Scheme 1), while their alkyl chains were replaced by ethylhexyl, the same as L². We theoretically investigated the electronic structures of L¹, L², and L³, and evaluated their Am(Ⅲ)/Eu(Ⅲ) selectivity. Based on the results of the complexation and selectivity ability for Am(Ⅲ)/Eu(Ⅲ) ions of the three ligands, we designed a new ligand, 2,9-bis(5-(2-ethylhexyl)-1H-1,2,4-triazol-3-yl)-1,10-phenanthroline (L⁴) with 1,2,4-triazole side chains and a preorganized phenanthroline backbone. We first elucidated the selectivity and binding nature of three similar pyridine-derived polyazole ligands and then designed new preorganized phenanthroline derivatives. This work helps to understand how the polyazole side chains affect the Am(Ⅲ)/Eu(Ⅲ) selectivity of ligands and provides design strategies for highly efficient polyazole based ligands in the future.

  The optimizations of all the structures were performed in the gas phase with B3LYP functional [50], applying Gaussian 16 software [51]. The scalar-relativistic effective core potentials (RECPs), which replace 28 and 60 core electrons for Eu [52] and Am [53] and the corresponding ECP28MWB-SEG [54] and ECP60MWB-SEG [55,56] valence basis sets were used, respectively. The 6–31G(d) basis set was used for the other atoms. Previous studies suggest that the structure is not sensitive to solvation [57,58], so the optimizations of the Am(Ⅲ)/Eu(Ⅲ) complexes were carried out at the B3LYP/RECP/6–31G(d) level of theory in the gas phase. The septet state is ground state for Am(Ⅲ)/Eu(Ⅲ) complexes and the spin contamination is negligible based on the results reported in previous works [40,45,59,60]. Spin-orbit coupling effects are not included considering the precision and computational cost [40,61]. To improve the accuracy of the energies, we performed single-point calculations for each species in solution using the B3LYP function with the larger 6-311G(d, p) basis set. The conductor-like polarizable continuum model (CPCM) [62,63] was applied for calculating the thermodynamic properties with Klamt's radii in cyclohexanone (ε = 15.619), nitrobenzene (ε = 34.809) and water (ε = 78.3553) solutions. The Gibbs free energy of each species in the solution is the sum of the solvation energy and thermal correction to the Gibbs free energy obtained at the B3LYP/RECP/6-31G(d) level of theory in the gas phase [64].

  The molecular electrostatic potential (ESP) maps of ligands were carried out using Multiwfn 3.8 software [65]. The molecular orbitals (MOs) were obtained with the GaussView program [51]. Mulliken charge and Mayer bond order (MBO) [66] were performed by natural bond orbital (NBO) [67,68]. The topological analysis of electron density was performed through the quantum theory of atoms in molecules (QTAIM) [65,69,70]. Energy decomposition analysis (EDA) [71] and the extended transition state method with the natural orbital for the chemical valence (NOCV) theory were performed at the BP86/TZ2P/ZORA level of theory with Amsterdam Modeling Suite 2022 (AMS 2022) [72–74] without frozen core.

  The ESP map can visualize the variably charged region of the ligand. The red and blue areas indicate the negative and positive values of ESP, respectively, as shown in Fig. 1. The negative minima of the ESP surface for L¹, L² and L³ are located in the cavities created by the N atoms of pyridine skeleton and polyazole side chains, indicating that the metal ion can be bound in the cavity. The more negative ESP for L¹ and L³ implies that these two ligands are more able to bind with metal ions than L².

  Figure 1

  

  Figure 1.  ESP maps and negative minimum ESP values (kcal/mol) for ligands L¹, L² and L³.

  DownLoad: Full-Size Img PowerPoint

  Fig. 2 shows the highest occupied MO (HOMO), the lowest unoccupied MO (LUMO), and HOMO-LUMO gap for ligands. The HOMOs have delocalized π bonding character with electron density distributed over pyridine and polyazole rings, reflecting that pyridine and polyazole ring are good electron donors. LUMOs for the three ligands have the anti-bonding nature. L¹ has the lowest HOMO-LUMO gap (5.05 eV), suggesting that L¹ is the most active ligand. According to the HSAB principle [12], the hardness (η) of the ligand is an important indicator for An(Ⅲ)/Ln(Ⅲ) separation. η is calculated by the equation η = (IP-EA)/2, where IP and EA represent the vertical ionization potential and electron affinity of the neutral ligand, respectively [75]. Table S1 (Supporting information) shows that the three ligands have almost the same η with the value of about 4.0, which indicating that the three ligands have potential selectivity for Am(Ⅲ)/Eu(Ⅲ) ions.

  Figure 2

  

  Figure 2.  LUMOs and HOMOs for ligands and the corresponding MO energies (eV) as well as the HOMO-LUMO gap (eV) with the isosurface value of 0.02 a.u.

  DownLoad: Full-Size Img PowerPoint

  The neutral ML(NO₃)₃ (M = Am/Eu) complexes were considered, as shown in Fig. 3. The Am(Ⅲ)/Eu(Ⅲ) ions were coordinated to ligand and nitrate ions in a tridentate and bidentate modes, respectively. The M-N and M-O bond lengths of the complexes are shown in Table 1. The bond length of M-N bonds follows the order of M-N₂ < M-N₁ for each complex, suggesting that the N₂ atoms of polyazoles binds metal ion more strongly than the N₁ on pyridines. And the M-N₂ bond lengths of ML(NO₃)₃ follow the order of M(L¹)(NO₃)₃ > M(L²)(NO₃)₃ > M(L³)(NO₃)₃, which suggests that the coordinating ability of N₂ atom and metal ions is in the order of 1,2,4-triazole < 1,2,3-triazole < pyrazole. Moreover, all Eu-N bonds are a little shorter than the corresponding Am-N ones, owing to the smaller ionic radius of Eu(Ⅲ) compared to Am(Ⅲ) [76]. Besides, the M-O bonds are significantly shorter than the M-N bonds, suggesting that the nitrate ions are more tightly coordinated to metal ions.

  Figure 3

  

  Figure 3.  Structures of ML(NO₃)₃ complexes. C, H, O, N and metal atoms are denoted as green, white, red, blue and pink, respectively.

  DownLoad: Full-Size Img PowerPoint

  Table 1

  

  Table 1.  Average M-N and M-O bond lengths (Å) in ML(NO₃)₃ complexes.

  DownLoad: CSV

  The extraction process was carried out in acidic environment and therefore the protonation ability of ligands was evaluated here. Polyazole and pyridine are easily protonated due to the lone pair of electrons on their N atom. The protonation capacity of N atom were calculated based on the reactions L + H⁺ → [LH]⁺ and L + [H₃O]⁺ → [LH]⁺ + H₂O. Table S2 (Supporting information) shows that the N₁ atom on pyridine is more likely to be protonated compared to the N₂ atom on polyazole.

  The interactions between ligand and metal ion can be evaluated by MOs. The α-spin valence MOs of ML(NO₃)₃ complexes are given in Fig. 4, Figs. S1 and S2 (Supporting information). Similar bonding interactions between the ligands and Am(Ⅲ)/Eu(Ⅲ) ions were found, the lower MO energy levels of AmL(NO₃)₃ complexes compared to the corresponding EuL(NO₃)₃ ones indicates that the former has a stronger interaction. Moreover, the contribution of Am and Eu atoms for the MOs are dominated by their d and f atomic orbitals according to the composition listed in Table S3 (Supporting information).

  Figure 4

  

  Figure 4.  Energy level diagrams (eV) of MOs for M(L³)(NO₃)₃ with the isosurface value of 0.02 a.u.

  DownLoad: Full-Size Img PowerPoint

  Table 2 shows the calculated Mulliken charges (Q) and charge transfer (ΔQ) in ML(NO₃)₃. It is obvious that electrons transferred from ligand and nitrate ion to Am and Eu, leaving the charge on them less than +3. In addition, ΔQ(Am) is greater than ΔQ(Eu) in each ML(NO₃)₃ complex, suggesting a tendency for the ligand to select Am(Ⅲ) ions. In addition, the charge transfer difference ΔΔQ(L) follows the trend of L¹ > L² > L³, which is related to the Am(Ⅲ)/Eu(Ⅲ) selectivity of the three ligands.

  Table 2

  

  Table 2.  Calculated Mulliken charges (Q) and charge transfer (ΔQ) in complexes.^a

  DownLoad: CSV

  To investigate the bonding properties of ligand and metal, multiple theoretical approaches were applied. The Mayer bond order (MBO) could indicate the covalency of M-L interactions. Table 3 shows the M-N and M-O MBOs of complexes. The M-N₂ MBOs are greater compared to the corresponding M-N₁ ones, suggesting that the N₂ atoms have significant interactions with metal ions. Moreover, all Am-N MBOs are greater than the corresponding Eu-N MBOs, reflecting a greater covalency interaction between ligand with Am(Ⅲ) ions. In addition, M-N MBOs are smaller than M-O MBOs because O atoms have stronger coordination ability with metal ions than that of N atoms, which is in accordance with the bond length results.

  Table 3

  

  Table 3.  The average M-N and M-O MBOs in complexes.

  DownLoad: CSV

  In QTAIM approach, the electron density (ρ), Laplacian density (∇²ρ), kinetic energy density (G) and energy density (H) at the bond critical point (BCP) are calculated to evaluate the bonding properties [77,78]. Generally, a value of ρ < 0.10 a.u. with ∇²ρ > 0 a.u. at a BCP refers to an ionic bond, while ρ > 0.20 a.u. with ∇²ρ < 0 a.u. reflects a typical covalent bond [77]. And usually, a ratio G/ρ > 1 and G/ρ < 1 indicates ionic interaction and covalent interaction, respectively. Table 4 shows that the ρ and ∇²ρ indicate that the M-N₁, M-N₂ and M-O bonds are ionic bond with weak covalency interaction, which is also supported by the G/ρ values. The covalency of the Am-N bond is larger than that of the Eu-N bond, and the M-N₂ bond has more covalency compared to the M-N₁ bond, which may explain the selectivity of Am(Ⅲ) ions compared to Eu(Ⅲ) ions and the more important role of N₂ atoms than N₁ atoms for the Am(Ⅲ)/Eu(Ⅲ) selectivity. What's more, ρ values at the M-N₂ BCPs are in order of M(L¹)(NO₃)₃ < M(L²)(NO₃)₃ < M(L³)(NO₃)₃, which reflects the polyazole side chains affect the covalent interaction of M-N₂ bonds.

  Table 4

  

  Table 4.  Properties (a.u.) at M-N and M-O BCPs in complexes.

  DownLoad: CSV

  EDA was performed to study the bonding nature between the ligand and metal fragments. Each ML(NO₃)₃ complex was split into the [L] and [M(NO₃)₃] fragments. Table 5 lists the interaction energy (ΔE_int) between the two fragments in the complex and gives three main terms [79], Pauli repulsive interaction (ΔE_Pauli), electrostatic interaction (ΔE_elec), and orbital interaction (ΔE_orb). The values of ΔE_orb range from −46.00 kcal/mol to −65.46 kcal/mol and ΔE_elec range from −85.32 kcal/mol to −116.27 kcal/mol, the latter being more negative than the former, reflecting a greater contribution of electrostatic interaction to the interaction between [L] and [M(NO₃)₃] fragments of each complex. The ΔE_orb of each AmL(NO₃)₃ complex is more negative than the corresponding Eu-ones, which also suggests the greater covalent interactions between the [Am(NO₃)₃] and [L] fragments.

  Table 5

  

  Table 5.  The Interaction energy (kcal/mol) between [L] and [M(NO₃)₃] (M = Am/Eu) fragments in complexes.

  DownLoad: CSV

  NOCV can provide intuitive and detailed information about the interactions between ligand and metal ion. In this method, ΔE_orb is decomposed into pairwise contributions of interacting orbitals. Figs. S3 and S4 (Supporting information) show the NOCVs contributing significantly to the ΔE_orb of complexes. The red to blue regions indicate the direction of electron flow. The electrons transfer from the N₁ and N₂ atoms of ligand to the metal ions. In addition, most Am-complex NOCVs have more negative ΔE_orb values than the corresponding Eu-complex NOCVs, which reflects that there are more electron transfer from ligand to [Am(NO₃)₃] fragment than that to [Eu(NO₃)₃] fragment. Fig. S5 (Supporting information) shows the donor-acceptor interactions of the main pairwise interacting orbitals in AmL¹(NO₃)₃ complex. To compare the possible difference in orbital interaction between ligands L¹, L², and L³ and ML(NO₃)₃ fragments, the major donor-acceptor interactions of ΔE_orb1 in complexes are given in Fig. 5. Noticeably, the acceptor orbitals for the [M(NO₃)₃] fragments are dominated by Am 5f, 6d orbitals and Eu 4f, 5d orbitals. Moreover, Am 5f contributes significantly more to the [Am(NO₃)₃] fragments than the corresponding Eu 4f to the [Eu(NO₃)₃] fragments, which may explain the ligands' Am-preference over Eu. The HOMO and HOMO-4 of [L] for AmL¹(NO₃)₃ act as electron donor orbitals, and only HOMO for EuL¹(NO₃)₃ provide electron to [Eu(NO₃)₃] fragment, reflecting that ligand L¹ has a greater difference in electron donating ability for Am and Eu ions. The HOMO of [L] acted as donor in ML²(NO₃)₃ complexes, and both HOMO and HOMO-6 of [L] acted as donors in ML³(NO₃)₃ complexes, suggesting that ligand L³ has a stronger electron-donating ability.

  Figure 5

  

  Figure 5.  The leading donor-acceptor interactions of ΔE_orb1 in complexes. The metal composition in [M(NO₃)₃] acceptor orbital is also shown.

  DownLoad: Full-Size Img PowerPoint

  The actual extraction process is very complex, TPH (tetrapropylhydrogen) [17], tert–butylbenzene (TBB) [21,23] and meta-nitrobenzotrifluoride [22] were usually used as organic solvents for the polyazole based ligands to extract Am(Ⅲ)/Eu(Ⅲ) ions. Here, we used nitrobenzene as an organic solvent and also tested cyclohexanone solvent. According to previous study [40], [M(NO₃)(H₂O)₈]_aq²⁺ + L_org + 2NO_3aq⁻ = [ML(NO₃)₃]_org + 8H₂O_aq is the most probable reaction that occurs during the Am(Ⅲ)/Eu(Ⅲ) ions extraction process with ligands. Table 6 shows the Gibbs free energy (ΔG) and their differences (ΔΔG = ∆G_Am - ∆G_Eu) for extraction reactions in cyclohexanone/water and nitrobenzene/water solutions at the B3LYP/RECP/6-311G(d, p) level of theory. All negative ΔG values for the reactions suggests that the three ligands can spontaneously complex with Am(Ⅲ) and Eu(Ⅲ) ions. The ΔG values of Am-extraction reactions are more negative than the corresponding Eu-ones, reflecting the extraction preference of the three ligands for Am(Ⅲ) ion. In addition, the ΔG values follows the order of L³ > L² > L¹ in both cyclohexanone/water and nitrobenzene/water solutions, reflecting that L³ has the strongest extraction ability for Am(Ⅲ) and Eu(Ⅲ) ions. However, the ΔΔG values show the opposite trend L¹ > L² > L³. To conclude, L¹ has the highest selectivity for Am(Ⅲ)/Eu(Ⅲ) ions, although it has the lowest complexation ability with them. The coordination number of the Eu(Ⅲ) and Am(Ⅲ) complexes in the solution is generally 6–12, and the preferable coordination number is 8 and 9 based on the result of previous work [80]. So we also investigated the possible reactions with initial reactants [M(H₂O)₉]³⁺ and [M(NO₃)(H₂O)₇]²⁺. The corresponding ΔG and ΔΔG values are provided in Table S4 (Supporting information), it shows that the trend of ΔG and ΔΔG values for the three ligands in Table 6 and Table S4 is similar, which is also consistent with our previous work [40]. Considering the protonation ability of ligands, N₁ atom on pyridine is more likely to be protonated than N₂ atoms on polyazole in highly acidic media. Thus, the reactions of the protonated ligands complexing with metal ions were also investigated in Table 6. We found that the ΔΔG values are the same with those of the unprotonated ones. However, the complexation abilities of the protonated ligands with Am(Ⅲ)/Eu(Ⅲ) ions are significantly lower compared to the unprotonated ligands, which suggests that the complexation ability of the polyazole ligands towards metal ions is sensitive to the acidic environment.

  Table 6

  

  Table 6.  ΔG (kcal/mol) and ΔΔG (kcal/mol) for extraction reactions in cyclohexanone/water and nitrobenzene/water solutions using the B3LYP functional and 6–311G(d, p) basis set.^a

  DownLoad: CSV

  According to the above results, it reveals that the polyazole side chains do affect the electronic structure and Am(Ⅲ)/Eu(Ⅲ) extraction performance of the three pyridine-derived polyazole ligands. Among them, L¹ with 1,2,4-triazole side chains has the largest separation factor for Am(Ⅲ)/Eu(Ⅲ) ions according to the ΔΔG values for the reactions. Therefore, we conclude that the 1,2,4-triazole side chains obviously enhance the Am(Ⅲ)/Eu(Ⅲ) selectivity of ligand. As we know, the preorganized phenanthroline moiety could significantly increase the complexation ability of ligand and its selectivity of Am(Ⅲ)/Eu(Ⅲ) than the bipyridine backbone [18,40,43]. For example, phenanthroline-based ligand CyMe₄-BTPhen could extract Cm(Ⅲ) and Am(Ⅲ) from lanthanide ions with significantly faster kinetics, efficiency and higher selectivity than CyMe₄-BTBP [18]. Thus, we predict that if the side chain 1,2,4-triazol is grafted onto the phenanthroline moiety, a new ligand, 2,9-bis(5-(2-ethylhexyl)-1H-1,2,4-triazol-3-yl)-1,10-phenanthroline (L⁴, Fig. 6) would show great An(Ⅲ)/Ln(Ⅲ) selectivity. To verify our expectation, the ESP map as well as the HOMO and LUMO orbitals for ligand L⁴ are presented in Fig. 6. The negative minimum value of ESP for L⁴ (−80.78 kcal/mol) is obviously more negative than that of L¹ (−70.47 kcal/mol), which reflects that L⁴ has stronger complexation ability with metal ions than L¹. Furthermore, the higher energy level of HOMO (−5.64 eV) with smaller HOMO-LUMO energy gap (4.23 eV) for L⁴ compared to that of L¹ indicate that L⁴ is a better electron donor. These results are consistent with earlier findings that tetradentate ligand with the same side chain has a stronger complexation ability compared to the tridentate one [40]. In addition, the η value of L⁴ is 3.37 eV, which is 0.65 eV smaller than that of L¹, suggesting that L⁴ is relatively softer than L¹.

  Figure 6

  

  Figure 6.  The structure (a), ESP (b), HOMO (c) and LUMO (d) of L⁴.

  DownLoad: Full-Size Img PowerPoint

  The optimized structures of AmL⁴(NO₃)₃ and EuL⁴(NO₃)₃ complexes with the bond lengths and MBOs of M-N and M-O bonds are shown in Fig. 7. The average M-N₁ bonds in ML⁴(NO₃)₃ complex is shorter than that in ML¹(NO₃)₃ due to the larger coordinated cavity of the former. By contrast, the M-N₂ bond is longer in ML⁴(NO₃)₃ than that in ML¹(NO₃)₃ thanks to the steric hindrance in the former. The Am-N₁ and Am-O bonds in AmL⁴(NO₃)₃ are slightly longer than the Eu-N₁ and Eu-O bonds in EuL⁴(NO₃)₃ complex, whereas the Am-N₂ bond is shorter than the Eu-N₂ bond, which suggests that L⁴ may have a higher Am(Ⅲ)/Eu(Ⅲ) selectivity than L¹. The Am-N and Am-O MBOs in AmL⁴(NO₃)₃ are greater than the corresponding Eu-ones in EuL⁴(NO₃)₃, demonstrating that L⁴ has a complexation preference for Am(Ⅲ) ions. Fig. S6 (Supporting information) displayed the α-spin valence MOs of AmL⁴(NO₃)₃ and EuL⁴(NO₃)₃ complexes. It can be seen that ligand L⁴ has similar bonding interactions with Am(Ⅲ) and Eu(Ⅲ) ions, while the former has somewhat lower MO energy levels than the latter, suggesting the higher stability of AmL⁴(NO₃)₃. The values of ΔG and ΔΔG for the reactions with ligand L⁴ in Table 7 reflect significant complexing ability and high selectivity of Am(Ⅲ)/Eu(Ⅲ). For example, the ∆∆G value for reaction [M(NO₃)(H₂O)₈]²⁺ _aq + L⁴ _org + 2NO₃⁻ _aq = ML⁴(NO₃)_{3 org} + 8H₂O _aq in cyclohexanone/water solutions is −8.09 kcal/mol, which is more negative compared to that of L¹ (−5.78 kcal/mol). Thus, we theoretically confirm that L⁴ has better Am(Ⅲ)/Eu(Ⅲ) selectivity than L¹.

  Figure 7

  

  Figure 7.  The optimized structures, bond length (BL) and MBOs of Am(L⁴)(NO₃)₃ and Eu(L⁴)(NO₃)₃.

  DownLoad: Full-Size Img PowerPoint

  Table 7

  

  Table 7.  ΔG (kcal/mol) and ΔΔG (kcal/mol) for the reactions of ligand L⁴ with Am(Ⅲ) and Eu(Ⅲ) complexes in cyclohexanone/water and nitrobenzene/water solutions using the B3LYP functional and 6-311G(d, p) basis set.

  DownLoad: CSV

  In summary, we first investigated the effects of polyazole side chains on Am(Ⅲ)/Eu(Ⅲ) separation performance of three pyridine-derived ligands with side chains 1,2,4-triazole (L¹), 1,2,3-triazole (L²) and pyrazole (L³) by scalar relativistic DFT methods. The results of M-L bond length and bonding nature suggest that the polyazole side chains play an important role in the complexation with Am(Ⅲ)/Eu(Ⅲ) ions and also reveal stronger complexation and higher covalency with Am(Ⅲ) than Eu(Ⅲ) ions of the three ligands. L¹ with 1,2,4-triazole side chains has the strongest separation ability while the lowest complexation ability for Am(Ⅲ)/Eu(Ⅲ) among the three pyridine-derived polyazole ligands based on the ΔG and ΔΔG values for the reactions. Motivated by the above results, we suggested that the 1,2,4-triazole side chains could enhance the Am(Ⅲ)/Eu(Ⅲ) selectivity of ligand and designed a new ligand (L⁴) bearing 1,2,4-triazole side chains and a preorganized phenanthroline backbone. The results of ESP, molecular orbitals, bonding nature, and thermodynamic properties reflected that L⁴ has significant separation properties for Am(Ⅲ)/Eu(Ⅲ). We firstly clarify the effect of polyazole side chains on the complexation and selectivity ability for Am(Ⅲ)/Eu(Ⅲ) ions and then design a new ligand with potential Am(Ⅲ)/Eu(Ⅲ) separation performance. This work provides a new strategy for designing effective N-donor ligands with excellent actinides and lanthanides separation performance in the future.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (Nos. U2067212, 22076188, 11875058, U20B2019), the National Science Fund for Distinguished Young Scholars (No. 21925603).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2023.109402.

  
  
• 1. [1]

     R.C. Ewing, Nat. Mater. 14 (2015) 252–257. doi: 10.1038/nmat4226

  2. [2]

     J. Veliscek-Carolan, J. Hazard. 318 (2016) 266–281. doi: 10.1016/j.jhazmat.2016.07.027

  3. [3]

     M. Salvatores, G. Palmiotti, Prog. Part. Nucl. Phys. 66 (2011) 144–166. doi: 10.1016/j.ppnp.2010.10.001

  4. [4]

     A. Salvatores, Nucl. Eng. Des. 235 (2005) 805–816. doi: 10.1016/j.nucengdes.2004.10.009

  5. [5]

     J.N. Mathur, M.S. Murali, K.L. Nash, Solvent Extr. Ion. Exch. 19 (2001) 357–390. doi: 10.1081/SEI-100103276

  6. [6]

     K.L. Nash, Handbook on the Physics and Chemistry of Rare Earths, Elsevier, 1994, pp. 197–238.

  7. [7]

     A. Khayambashi, L. Chen, X. Dong, et al., Chin. Chem. Lett. 33 (2022) 3429–3434. doi: 10.1016/j.cclet.2022.02.011

  8. [8]

     L. Xu, N. Pu, G. Y, et al., Inorg. Chem. Front. 7 (2020) 1726–1740. doi: 10.1039/d0qi00200c

  9. [9]

     X. Yang, S.H. Wang, L. Xu, et al., Inorg. Chem. Front. 9 (2022) 4671–4684. doi: 10.1039/d2qi01153k

  10. [10]

      T.Y. Xiu, S.M. Zhang, P. Ren, et al., Chin. Chem. Lett. 34 (2023) 108440. doi: 10.1016/j.cclet.2023.108440

  11. [11]

      D.S. Tian, Y.Y. Liu, Y. Kang, et al., ACS Cent. Sci. 9 (2023) 1642–1649. doi: 10.1021/acscentsci.3c00504

  12. [12]

      R.G. Pearson, J. Am. Chem. Soc. 85 (1963) 3533–3539. doi: 10.1021/ja00905a001

  13. [13]

      A. Leoncini, J. Huskens, W. Verboom, Chem. Soc. Rev. 46 (2017) 7229–7273. doi: 10.1039/C7CS00574A

  14. [14]

      Z.R. Ye, Q.Y. Wu, C.Z. Wang, et al., Inorg. Chem. 60 (2021) 16409–16419. doi: 10.1021/acs.inorgchem.1c02256

  15. [15]

      A. Geist, P.J. Panak, Solvent Extr. Ion. Exch. 39 (2021) 128–151. doi: 10.1080/07366299.2020.1831235

  16. [16]

      C. Madic, M.J. Hudson, Report EUR 18038EN. Nuclear Science and Technology, 1998.

  17. [17]

      Z. Kolarik, U. Müllich, F. Gassner, Solvent Extr. Ion. Exch. 17 (1999) 23–32. doi: 10.1080/07360299908934598

  18. [18]

      F.W. Lewis, L.M. Harwood, M.J. Hudson, et al., J. Am. Chem. Soc. 133 (2011) 13093–13102. doi: 10.1021/ja203378m

  19. [19]

      M.G.B. Drew, M.J. Hudson, P.B. Iveson, et al., Dalton Trans. (1999) 2433–2440.

  20. [20]

      A. Bremer, C.M. Ruff, D. Girnt, et al., Inorg. Chem. 51 (2012) 5199–5207. doi: 10.1021/ic3000526

  21. [21]

      D.P. Su, Y. Liu, S.M. Li, et al., Eur. J. Inorg. Chem. (2017) 651–658 2017.

  22. [22]

      Y. Liu, X.Y. Yang, S.D. Ding, et al., Inorg. Chem. 57 (2018) 5782–5790. doi: 10.1021/acs.inorgchem.8b00074

  23. [23]

      J.R. Wang, D.P. Su, D.Q. Wang, et al., Inorg. Chem. 54 (2015) 10648–10655. doi: 10.1021/acs.inorgchem.5b01452

  24. [24]

      X.H. Kong, Q.Y. Wu, C.Z. Wang, et al., J. Phys. Chem. A 122 (2018) 4499–4507. doi: 10.1021/acs.jpca.8b00177

  25. [25]

      C. Kiefer, A.T. Wagner, B.B. Beele, et al., Inorg. Chem. 54 (2015) 7301–7308. doi: 10.1021/acs.inorgchem.5b00803

  26. [26]

      E. Macerata, E. Mossini, S. Scaravaggi, et al., J. Am. Chem. Soc. 138 (2016) 7232–7235. doi: 10.1021/jacs.6b03106

  27. [27]

      A.C. Edwards, P. Mocilac, A. Geist, et al., Chem. Commun. 53 (2017) 5001–5004. doi: 10.1039/C7CC01855J

  28. [28]

      A. Ossola, E. Macerata, E. Mossini, et al., J. Radioanal. Nucl. Ch. 318 (2018) 2013–2022. doi: 10.1007/s10967-018-6253-y

  29. [29]

      P. Weßling, M. Maag, G. Baruth, et al., Inorg. Chem. 61 (2022) 17719–17729. doi: 10.1021/acs.inorgchem.2c02871

  30. [30]

      F. Galluccio, E. Macerata, P. Weßling, et al., Inorg. Chem. 61 (2022) 18400–18411. doi: 10.1021/acs.inorgchem.2c02332

  31. [31]

      L.K. Liu, S.B. Xie, H.B. Lv, et al., Chin. Chem. Lett. 33 (2022) 3439–3443. doi: 10.1016/j.cclet.2022.04.001

  32. [32]

      Y.A. Ustynyuk, M.Y. Alyapyshev, V.A. Babain, et al., Russ. Chem. Rev. 85 (2016) 917–942. doi: 10.1070/RCR4588

  33. [33]

      X. Zhang, S.L. Adelman, B.T. Arko, et al., Inorg. Chem. 61 (2022) 11556–11570. doi: 10.1021/acs.inorgchem.2c00534

  34. [34]

      C. Ebenezer, R. Vijay Solomon, Inorg. Chem. Front. 8 (2021) 3012–3024. doi: 10.1039/d1qi00097g

  35. [35]

      R.C. Chapleski Jr., A.S. Ivanov, K.A. Peterson, et al., Phys. Chem. Chem. Phys. 23 (2021) 19558–19570. doi: 10.1039/d1cp02466c

  36. [36]

      J.P. Yu, K. Liu, Q.Y. Wu, et al., Chinese J. Chem. 39 (2021) 2125–2131. doi: 10.1002/cjoc.202100149

  37. [37]

      G.J.P. Deblonde, M.P. Kelley, J. Su, et al., Angew. Chem. Int. Ed. 57 (2018) 4521–4526. doi: 10.1002/anie.201709183

  38. [38]

      J.N. Cross, J. Su, E.R. Batista, et al., J. Am. Chem. Soc. 139 (2017) 8667–8677. doi: 10.1021/jacs.7b03755

  39. [39]

      J. Su, E.R. Batista, K.S. Boland, et al., J. Am. Chem. Soc. 140 (2018) 17977–17984. doi: 10.1021/jacs.8b09436

  40. [40]

      Q.Y. Wu, Y.T. Song, L. Ji, et al., Phys. Chem. Chem. Phys. 19 (2017) 26969–26979. doi: 10.1039/C7CP04625A

  41. [41]

      C. Wang, Q.Y. Wu, X.H. Kong, et al., Inorg. Chem. 58 (2019) 10047–10056. doi: 10.1021/acs.inorgchem.9b01200

  42. [42]

      X.H. Kong, Q.Y. Wu, J.H. Lan, et al., Inorg. Chem. 57 (2018) 14810–14820. doi: 10.1021/acs.inorgchem.8b02550

  43. [43]

      X.P. Lei, Q.Y. Wu, C.Z. Wang, et al., Dalton Trans. 51 (2022) 16659–16667. doi: 10.1039/d2dt02474h

  44. [44]

      X.P. Lei, Q.Y. Wu, C.Z. Wang, et al., Inorg. Chem. 62 (2023) 2705–2714. doi: 10.1021/acs.inorgchem.2c03823

  45. [45]

      Y.M. Chen, C.Z. Wang, Q.Y. Wu, et al., Inorg. Chem. 59 (2020) 3221–3231. doi: 10.1021/acs.inorgchem.9b03604

  46. [46]

      Z.R. Ye, Q.Y. Wu, C.Z. Wang, et al., Inorg. Chem. 61 (2022) 6110–6119. doi: 10.1021/acs.inorgchem.2c00232

  47. [47]

      Y. Zou, J.H. Lan, L.Y. Yuan, et al., Inorg. Chem. 62 (2023) 4581–4589. doi: 10.1021/acs.inorgchem.2c04476

  48. [48]

      Y. Zou, J.H. Lan, L.Y. Yuan, et al., Inorg. Chem. 61 (2022) 15423–15431. doi: 10.1021/acs.inorgchem.2c01952

  49. [49]

      A. Bhattacharyya, T. Gadly, A.S. Kanekar, et al., Inorg. Chem. 57 (2018) 5096–5107. doi: 10.1021/acs.inorgchem.8b00142

  50. [50]

      C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785–789. doi: 10.1103/PhysRevB.37.785

  51. [51]

      M.J. Frisch, G.W. Trucks, H.B. Schlegel, et al., Gaussian 16, Revision B, Gaussian, Inc., Wallingford, CT, 2016.

  52. [52]

      M. Dolg, H. Stoll, H. Preuss, J. Chem. Phys. 90 (1989) 1730–1734. doi: 10.1063/1.456066

  53. [53]

      W. Küchle, M. Dolg, H. Stoll, J. Chem. Phys. 100 (1994) 7535–7542. doi: 10.1063/1.466847

  54. [54]

      X.Y. Cao, M. Dolg, J. Mol. Struct. 581 (2002) 139–147. doi: 10.1016/S0166-1280(01)00751-5

  55. [55]

      X.Y. Cao, M. Dolg, H. Stoll, J. Chem. Phys. 118 (2002) 487–496.

  56. [56]

      X.Y. Cao, M. Dolg, J. Mol. Struct. 673 (2004) 203–209. doi: 10.1016/j.theochem.2003.12.015

  57. [57]

      G.A. Shamov, G. Schreckenbach, R.L. Martin, et al., Inorg. Chem. 47 (2008) 1465–1475. doi: 10.1021/ic7015403

  58. [58]

      G.A. Shamov, G. Schreckenbach, J. Phys. Chem. A 109 (2005) 10961–10974. doi: 10.1021/jp053522f

  59. [59]

      J. Narbutt, A. Wodyński, M. Pecul, Dalton Trans. 44 (2015) 2657–2666. doi: 10.1039/C4DT02657H

  60. [60]

      A. Zaiter, B. Amine, Y. Bouzidi, et al., Inorg. Chem. 53 (2014) 4687–4697. doi: 10.1021/ic500361b

  61. [61]

      C. Clavaguéra-Sarrio, V. Vallet, D. Maynau, et al., J. Chem. Phys. 121 (2004) 5312–5321. doi: 10.1063/1.1784412

  62. [62]

      M. Cossi, N. Rega, G. Scalmani, et al., J. Comput. Chem. 24 (2003) 669–681. doi: 10.1002/jcc.10189

  63. [63]

      V. Barone, M. Cossi, J. Phys. Chem. A 102 (1998) 1995–2001. doi: 10.1021/jp9716997

  64. [64]

      J. Ho, A. Klamt, M.L. Coote, J. Phys. Chem. A 114 (2010) 13442–13444. doi: 10.1021/jp107136j

  65. [65]

      T. Lu, F. Chen, J. Comput. Chem. 33 (2012) 580–592. doi: 10.1002/jcc.22885

  66. [66]

      I. Mayer, Chem. Phys. Lett. 97 (1983) 270–274. doi: 10.1016/0009-2614(83)80005-0

  67. [67]

      A.E. Reed, F. Weinhold, J. Chem. Phys. 78 (1983) 4066–4073. doi: 10.1063/1.445134

  68. [68]

      A.E. Reed, R.B. Weinstock, F. Weinhold, J. Chem. Phys. 83 (1985) 735–746. doi: 10.1063/1.449486

  69. [69]

      R.F.W. Bader, C.F. Matta, Inorg. Chem. 40 (2001) 5603–5611. doi: 10.1021/ic010165o

  70. [70]

      E. Espinosa, I. Alkorta, J. Elguero, et al., J. Chem. Phys. 117 (2002) 5529–5542. doi: 10.1063/1.1501133

  71. [71]

      T. Ziegler, A. Rauk, Theor. Chem. Acc. 46 (1977) 1–10. doi: 10.1007/BF02401406

  72. [72]

      G. te Velde, F.M. Bickelhaupt, E.J. Baerends, et al., J. Comput. Chem. 22 (2001) 931–967. doi: 10.1002/jcc.1056

  73. [73]

      C. Fonseca Guerra, J.G. Snijders, G. te Velde, et al., Theor. Chem. Acc. 99 (1998) 391–403.

  74. [74]

      T.Z.E.J. Baerends, J. Autschbach, D. Bashford, et al., ADF2022, SCM, Theoretical Chemistry, Vrije Universiteit, Amsterdam, The Netherlands, URL: http://www.scm.com.

  75. [75]

      R.G. Parr, R.G. Pearson, J. Am. Chem. Soc. 105 (1983) 7512–7516. doi: 10.1021/ja00364a005

  76. [76]

      R. Shannon, Acta Crystallogr. A 32 (1976) 751–767. doi: 10.1107/S0567739476001551

  77. [77]

      M.P. Jensen, A.H. Bond, J. Am. Chem. Soc. 124 (2002) 9870–9877. doi: 10.1021/ja0178620

  78. [78]

      D. Cremer, E. Kraka, J. Am. Chem. Soc. 107 (1985) 3811–3819. doi: 10.1021/ja00299a010

  79. [79]

      W.H. Tu, S.J. Zeng, Y. Bai, et al., Ind. Chem. Mater. 1 (2023) 262–270. doi: 10.1039/d2im00041e

  80. [80]

      P. Thakur, J.L. Conca, G.R. Choppin, J. Solution Chem. 41 (2012) 599–615. doi: 10.1007/s10953-012-9826-3

• 

  Scheme 1  Structures of N-donor ligands.

  下载: 全尺寸图片 幻灯片
  

  Scheme 2  Structures of target pyridine-derived polyazole ligands.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  ESP maps and negative minimum ESP values (kcal/mol) for ligands L¹, L² and L³.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  LUMOs and HOMOs for ligands and the corresponding MO energies (eV) as well as the HOMO-LUMO gap (eV) with the isosurface value of 0.02 a.u.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Structures of ML(NO₃)₃ complexes. C, H, O, N and metal atoms are denoted as green, white, red, blue and pink, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Energy level diagrams (eV) of MOs for M(L³)(NO₃)₃ with the isosurface value of 0.02 a.u.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  The leading donor-acceptor interactions of ΔE_orb1 in complexes. The metal composition in [M(NO₃)₃] acceptor orbital is also shown.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  The structure (a), ESP (b), HOMO (c) and LUMO (d) of L⁴.

  下载: 全尺寸图片 幻灯片
  

  Figure 7  The optimized structures, bond length (BL) and MBOs of Am(L⁴)(NO₃)₃ and Eu(L⁴)(NO₃)₃.

  下载: 全尺寸图片 幻灯片

  Table 1.  Average M-N and M-O bond lengths (Å) in ML(NO₃)₃ complexes.

  下载: 导出CSV

  Table 2.  Calculated Mulliken charges (Q) and charge transfer (ΔQ) in complexes.^a

  下载: 导出CSV

  Table 3.  The average M-N and M-O MBOs in complexes.

  下载: 导出CSV

  Table 4.  Properties (a.u.) at M-N and M-O BCPs in complexes.

  下载: 导出CSV

  Table 5.  The Interaction energy (kcal/mol) between [L] and [M(NO₃)₃] (M = Am/Eu) fragments in complexes.

  下载: 导出CSV

  Table 6.  ΔG (kcal/mol) and ΔΔG (kcal/mol) for extraction reactions in cyclohexanone/water and nitrobenzene/water solutions using the B3LYP functional and 6–311G(d, p) basis set.^a

  下载: 导出CSV

  Table 7.  ΔG (kcal/mol) and ΔΔG (kcal/mol) for the reactions of ligand L⁴ with Am(Ⅲ) and Eu(Ⅲ) complexes in cyclohexanone/water and nitrobenzene/water solutions using the B3LYP functional and 6-311G(d, p) basis set.

  下载: 导出CSV
•