Hindered rotation in a novel 1,2,4-triazinyl phenanthroline (t-phen) ligand leading to improved separation of Am³⁺ and Eu³⁺vis-à-vis 1,2,4-triazinyl bipyridine (t-bipy): a computational validation of the experimental results

Abstract

A novel nitrogen donor ligand having a more pre-organized structure, 1,2,4-triazinyl phenanthroline (t-phen), is synthesized and evaluated for lanthanide–actinide separation in the present work. The extraction and selectivity for Am³⁺ over Eu³⁺ was found to be improved with t-phen as compared to the analogous 1,2,4-triazinyl bipyridine (t-bipy), which was explained by analyzing their conformational energies and energy differences between the frontier orbitals (highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO)) with the help of density functional theoretical calculations. Higher covalence in Am³⁺ complexes as compared to the Eu³⁺ complexes was indicated by more shared electrons between Am³⁺ and bonded ‘N’ atoms, shorter ‘Am–N’ bonds and higher overlap between the orbitals of Am³⁺ and ligands in the frontier orbitals of the complexes in the case of both the ligands. Natural bond order analysis also supports these observations.

---

Introduction

The high level waste (HLW) generated from the PUREX process contains a number of fission products, including several lanthanide elements along with unextracted uranium, plutonium and very small amount of minor actinides, i.e. Am and Cm. The major concern for the safe management of HLW is due to the presence of minor actinides because of their long half lives and high radiotoxicity. The safe management of these minor actinides is possible either by vitrifying them in a proper solid matrix or by their transmutation into short lived or stable nuclides. For this purpose, the minor actinides need to be separated from the bulk of the metal ions present in the HLW. In the process called “Actinide Partitioning” the minor actinides are extracted along with the lanthanides leaving behind most of the other elements in the raffinate. The co-extraction of the trivalent lanthanides and actinides is a consequence of their similar chemical behaviour.¹ However, the transmutation process using high flux reactors necessitates the separation of trivalent actinides from the trivalent lanthanides which act as neutron poisons. The current strategy on lanthanide–actinide separation is based on the use of soft donor extractants for preferential actinide extraction. The higher spatial distribution of ‘5f’ orbitals of the actinides as compared to the ‘4f’ orbitals of the lanthanides leads to the formation of preferential covalent bonding of actinides with soft donor ligands as compared to that of lanthanides and this property is being exploited for the separation of trivalent actinides and lanthanides. A number of ‘N’ donor heteropolycyclic ligands starting from terpyridine (terpy) (Fig. 1a) to bis-1,2,4-triazinyl pyridine (BTP) (Fig. 1b) and bis-1,2,4-triazinyl bipyridine (BTBP) (Fig. 1c) have been evaluated for this purpose.^2,3 The complexes formed with Am³⁺ or Ln³⁺ are very different for different ‘N’ donor ligands, viz. the terpy complexes of Ln³⁺ or An³⁺ contain only one or two terpy molecules^4,5 whereas three molecules of BTP are present^6,7 in An³⁺/Ln³⁺ BTP complexes. Hudson et al., therefore, explored a class of ligands, i.e. 6-(5,6-dialkyl-1,2,4-triazine-3-yl) bipyridine (R₂(t-bipy)) (Fig. 1d) derivatives which fall in between the terpy and BTP ligands.⁸ They found that only one t-bipy molecule is present in the extractable complexes of Ln³⁺ and An³⁺, which resembles the behaviour of terpy. Poor extraction and selectivity for trivalent actinides over the lanthanides was observed employing the t-bipy derivatives as the extractant. Hudson et al. analyzed the energies of different conformations of free t-bipy molecules and observed that the rotation around the two pyridine rings is much more hindered (ΔE ∼ 9.0 kcal mol⁻¹) as compared to the pyridine-triazine rings (ΔE ∼ 0.78 kcal mol⁻¹), because of the presence of ortho-hydrogen atoms in the two pyridine rings in the cis conformation. The requirement of this 9.0 kcal mol⁻¹ energy can be avoided if we can preorganize the two ‘N’ atoms of the pyridine rings in a cis conformation resulting in the formation of a stronger complex with the metal ion as compared to t-bipy derivatives. An attempt was, therefore, made in the present work to synthesize a ligand where the 1,2,4-triazine ring is attached to the 1,10-phenanthroline ring (2-(5,6-dialkyl-1,2,4-triazine-3-yl)1,10-phenanthroline (R₂(t-phen))) (Fig. 1e) instead of the bipyridine ring in R₂(t-bipy). The extraction and separation behaviour of Am³⁺ and Eu³⁺ with Me₂(t-bipy) and Me₂(t-phen) was compared at different pH values of the aqueous phase. Computational studies were carried out to corroborate the higher Am³⁺ extraction and selectivity of t-phen as compared to that of t-bipy.

Fig. 1 Various ‘N’ donor heteropolycyclic ligands for the separation of trivalent actinides and lanthanides; (a) terpy, (b) BTP, (c) BTBP (d) t-bipy and (e) t-phen.

Experimental

Synthesis of Me₂(t-phen)

a) 1,10-Phenanthroline-1-oxide.
Hydrogen peroxide (4 ml, 30% solution in water) was added to a solution of 1,10-phenanthroline monohydrate (5 g, 25.25 mmol) in glacial acetic acid (30 ml). The reaction mixture was stirred at 75 °C for 4 h and another portion of hydrogen peroxide (4 ml, 30% solution in water) was added to the mixture. After 3 h at 75 °C, the reaction mixture was cooled to room temperature and acetic acid was neutralized slowly by adding solid sodium carbonate into the reaction mixture. The resultant pasty mass was extracted multiple times with chloroform. The combined extract was concentrated to give a yellow green colored solid that was crystallized from chlorobenzene to give 1,10-phenanthroline-1-oxide (4.16 g, 84%).⁹ M.P. 184–185 °C, (reported M.P. 180–181 °C) IR (KBr): 3065, 1626, 1598, 1577, 1517, 1412, 1327, 1256, 1203, 1069, 1016, 842, 815, 711, 679 cm⁻¹; ¹H NMR (200 MHz, CDCl₃): δ 7.39 (dd, 1 H, J = 6.4, 8.0 Hz), 7.57–7.761 (m, 4 H), 8.17 (dd, 1 H, J = 1.8, 8 Hz), 8.68 (d, 1 H, J = 6.4 Hz), 9.25 (dd, 1 H, J = 1.8, 4.3 Hz); ¹³C NMR (50 MHz, CDCl₃): δ 122.5, 122.8, 123.2, 124.2, 126.1, 128.5, 128.6, 132.9, 135.5, 140.3, 142.2, 149.4.

b) 2-Cyano-1,10-phenanthroline.
Freshly distilled benzoyl chloride (4.75 ml, 40.9 mmol) was added slowly (about 15 min) to a vigorously stirred solution of 1,10-phenanthroline-1-oxide (5 g, 25.51 mmol) and potassium cyanide (4.72 g, 72.42 mmol) in distilled water (80 ml) at room temperature. After 15 min, the solid formed was collected by filtration, washed with water and dried in vacuuo. Light tan needles were obtained after crystallization from ethanol (3.92 g, 75%).⁹

M.P. 234–235 °C, lit.¹ M.P. 233–234 °C. IR (KBr): 3068, 2229, 1505, 1486, 1390 cm⁻¹. ¹H NMR (300 MHz, CDCl₃): δ 7.76 (dd, 1 H, J = 4.5, 8.1 Hz), 7.87 (d, 1 H, J = 9 Hz), 7.98 (d, 1 H, J = 9 Hz), 7.99 (d, 1 H, J = 8.1 Hz), 8.33 (dd, 1 H, J = 1.5, 8.1 Hz), 8.42 (d, 1 H, J = 8.1 Hz), 9.3 (dd, 1 H, J = 1.8 Hz, 4.2 Hz). ¹³C NMR (50 MHz, CDCl₃): δ 117.2, 123.9, 125.5, 126.0, 129.0, 129.6, 133.2, 136.1, 137.0, 145.0, 146.5, 151.0. MS (ESI) m/z: 205 (M⁺, 100%), 178 (15), 152 (12), 125 (9), 102 (9), 89 (3), 75 (3). GC (260 °C isothermal): t_R 21.6 min (100%).

c) 3-{2-(1,10-Phenanthrolyl) -(5,6-dimethyl)}-1,2,4-triazine (Me₂(t-phen)).
Hydrazine hydrate (7 ml, 224 mmol) was added to a solution of 2-cyano-1,10-phenanthroline (2.5 g, 12.2 mmol) in ethanol (85 ml) and stirred overnight. The reaction mixture was concentrated under vacuum and the residue was crystallised from water to give the hydrazone (2.38 g, 82%). A mixture of monocarboxamide hydrazone (2.0 g, 8.44 mmol) and butan-2,3-dione (0.85 ml, 9.54 mmol) in benzene (13 ml) was heated under reflux. After 2.5 h, the solvent was evaporated and the residue crystallized from ethanol to give the hemi triazinyl monohydrate derivative (2.1 g, 82%).¹⁰ M.P. 211–212 °C, lit.¹⁰ M.P. 210–211 °C. ¹H NMR (200 MHz, CDCl₃): δ 2.79 (s, 3 H), 2.80 (s, 3 H), 3.90 (s, broad, 2 H, H₂O), 7.69(dd, 1 H, J = 4.4, 8 Hz), 7.87 (s, 2 H), 8.30 (dd, 1 H, J = 1.4, 8 Hz), 8.45 (d, 1 H, J = 8.4 Hz), 8.98 (d, 1 H, J = 8.4 Hz), 9.22 (dd, 1 H, J = 1.4, 4.4 Hz). ¹³C NMR (50 MHz, CDCl₃): δ 19.5, 21.8, 122.7, 123.2, 126.2, 127.8, 128.9, 129.4, 136.1, 137.0, 146.1, 146.2, 150.3, 152.8, 157.2, 159.8, 161.6.

Synthesis of Me₂(t-bipy)

a) 2,2′-Bipyridine-N-oxide.
A solution of 3-chloroperoxybenzoic acid (3.2 g, 70%, 12.9 mmol) in dichloromethane (22.8 ml) was added drop wise to a solution of bipyridine (2 g, 12.6 mmol) in dichloromethane (13.4 ml) at 5–10 °C. After 16 h at room temperature, the reaction mixture was cooled on an ice bath and sodium hydroxide (2 M solution in water, 14.5 ml, 29 mmol) was added into it. After 15 min, the reaction mixture was diluted with chloroform (25 ml). The organic layer was separated and washed with 2 M NaOH solution, dried over sodium carbonate and evaporated under reduced pressure. The residue was purified by silica gel column chromatography to give the mono N-oxide (1.7 g, 78%).¹¹ M.P. 54–55 °C (reported M.P. 55–56 °C), ¹H NMR (200 MHz, CDCl₃): δ 7.27–7.42 (m, 3 H), 7.84 (dt, 1 H, J = 1.8, 8 Hz), 8.19 (dd, 1 H, J = 2, 8 Hz), 8.33 (d, 1 H, J = 6.2), 8.71 (d, 1 H, J = 4.7 Hz), 8.87 (d, 1 H, J = 8 Hz). ¹³C NMR (50 MHz, CDCl₃): δ 124.0, 125.0, 125.2, 127.6, 136.0, 140.4, 147.0, 149.0, 149.4.

b) 2-Cyano-2,2′-bipyridine.
A solution of 2,2′-bipyridine-N-oxide (0.3 g, 1.7 mmol) and trimethylsilyl cyanide (0.27 ml, 1.88 mmol) in dichloromethane (3.5 ml) was stirred for 5 min at room temperature. N,N′- dimethylcarbamoyl chloride (0.17 ml, 1.88 mmol) was added to the reaction mixture and stirred for 7 days under an argon atmosphere. The reaction mixture was diluted with diethyl ether (30 ml) and washed successively with 5% NaHCO₃ solution and brine and dried over solid Na₂CO₃. The solvent was evaporated and the residue was purified by silica gel column chromatography to give the monocyano compound (0.29 g, 93%).¹¹

M.P. 132–133 °C (reported M.P. 132–133 °C), ¹H NMR (200 MHz, CDCl₃): δ 7.37 (ddd, 1 H, J = 1.1, 4.8, 7.5 Hz), 7.70 (dd, 1 H, J = 1, 7.5 Hz), 7.85 (dt, 1 H, J = 1.8, 8 Hz), 7.94 (t, 1 H, J = 8 Hz), 8.46 (d, 1 H, J = 8.0 Hz), 8.64–8.69 (m, 2 H). ¹³C NMR (50 MHz, CDCl₃): δ 117.3, 121.5, 124.2, 124.7, 128.0, 133.2, 137.2, 137.8, 149.2, 153.9, 157.6.

c) 6-(5,6-Dimethyl-1,2,4-triazin-3yl)-2,2′-bipyridine (Me₂(t-bipy)).
Hydrazine hydrate (0.19 ml, 6 mmol) was added to a solution of 6-cyano-2,2′-bipyridine (0.108 g, 0.6 mmol) dissolved in ethanol (7.2 ml) and stirred for 2 days. The solvent was removed under reduced pressure to give the monocarboxamide hydrazone (0.133 g, 0.6 mmol, 100%). Butan-2,3-dione (0.06 g, 0.7 mmol) and benzene (2 ml) were added to the monocarboxamide hydrazone and the mixture was refluxed for 5 h. The solvent was evaporated and the residue was purified by silica gel column chromatography to give the Me₂(t-bipy)⁸ (0.117 g, 74%).⁸ M.P. 148–149 °C. ¹H NMR (200 MHz, CDCl₃): δ 2.68 (s, 3 H), 2.77 (s, 3 H), 7.38 (t, 1 H, J = 5.4 Hz), 7.91 (t, 1 H, J = 7.6 Hz), 8.03 (t, 1H, J = 7.9 Hz), 8.56–8.73 (m, 4 H). ¹³C NMR (50 MHz, CD₃OD): δ 19.5, 21.9, 123.3, 123.7, 124.8, 125.5, 138.8, 139.3, 149.8, 153.6, 156.6, 157.3,158.9, 161.9, 162.2. Calcd. for C₁₅H₁₃N₅; C, 68.42; H, 4.98; N, 26.60%; Found C, 67.93; H, 5.27; N, 26.36%.

Other reagents and chemicals

2-bromo-octanoic acid (>97%) was procured from Fluka chemie, Germany. ²⁴¹Am was purified as reported earlier¹² and the purity was checked by alpha spectrometry. ^152,154Eu was procured from the Board of Radiation and Isotope Technology (BRIT), Mumbai, India. Suprapur nitric acid (Merck) was used for preparing the tracer solutions. All other reagents were of AR grade.

Distribution studies

Distribution studies were carried out using ²⁴¹Am and ^152,154Eu tracers under varying experimental conditions. A mixture of 0.03 M Me₂(t-bipy) or Me₂(t-phen) and 1 M 2-bromo-octanoic acid in n-dodecane was used as the organic phase while the aqueous phase was the dilute nitric acid in the pH range of 1.2 to 2.2. Equal volumes (1.0 mL) of the organic and aqueous phases containing the required tracer were kept for equilibration in a thermostated water bath at 25 ± 0.1 °C for 60 min. The two phases were then centrifuged and assayed by taking suitable aliquots (100–200 μL) from both the phases. The gamma activities were measured using a high purity germenium detector procured from Baltic Scientific Instruments. The distribution ratio (D_M) was calculated as the ratio of counts per minute per unit volume in the organic phase to that in the aqueous phase. The mass balance was within the experimental error limits (± 5%). The separation factor (S.F.) was calculated as the ratio of D_Am to D_Eu.

Computational studies

The alkyl substituents show insignificant changes in the charge distributions and the energy gap between the frontier orbitals.¹³ A number of studies were also carried out, where H-form of BTP was considered as a model for the alkyl BTP derivatives.^14,15 Most of the calculations of the free ligands and their Am³⁺ and Eu³⁺ complexes were, therefore, performed for hydrogen form of both the ligands (t-bipy and t-phen). The presence of alkyl groups can, however, alter the conformational energies require to rotate the rings. The relative energies of different conformations of Me₂(t-bipy) and Me₂(t-phen) were determined varying the angle of rotation around the two adjacent rings. All the calculations of free and protonated ligands were performed at the density functional (DFT) level of theory using the hybrid exchange correlation functional (B3LYP) with the GAMESS molecular orbital package¹⁶ using the valence double zeta Gaussian type of basis set 6-31G(d,p). Molecular orbital pictures were drawn using the visualizing program MOLDEN.¹⁷ The calculations on the Am³⁺ and Eu³⁺ complexation were carried out with TURBOMOLE program package¹⁸ at the DFT level using Becke′s exchange functional¹⁹ in conjunction with Perdew′s correlation functional²⁰ (BP86) where for the heavy atoms relativistic energy consistent ab initio f-in-valence 28 (Eu) and 60 (Am) electron core pseudo potentials (ECP) along with the corresponding def-SV(P) basis set were selected. All other lighter atoms were treated at the all electron (AE) level and the standard def-SV(P) basis sets as implemented in the TURBOMOLE program was used. It may be noted that for the Am and Eu atoms the def-SV(P) basis set as present in the TURBOMOLE basis set library is quite large and consists of (14s13p10d8f1g) functions contracted to [10s9p5d4f1g].^21–27

Results and discussion

Distribution studies of Am³⁺ and Eu³⁺

The distribution behaviour of Am³⁺ and Eu³⁺ by Me₂(t-bipy) and Me₂(t-phen) was compared in the presence of 2-bromo-octanoic acid in an n-dodecane medium at different aqueous phase pHs. With increasing the pH of the aqueous phase, the distribution ratio (D_M) of both Am³⁺ and Eu³⁺ was found to increase with similar slopes due to more deprotonation of 2-bromo-octanoic acid with increasing pH of the aqueous phase (Fig. 2). When the extraction efficiency of these two ligands for Am³⁺ over Eu³⁺ was compared, it was interesting to note that the D_Am values were an order of magnitude higher in case of Me₂(t-phen) as compared to Me₂(t-bipy) at all pHs (Fig. 2). Beside the D_Am value, selectivity for Am³⁺ over Eu³⁺ was also found to be higher with Me₂(t-phen) as compared to Me₂(t-bipy) (Fig. 3). An attempt was, therefore made in order to rationalize these observations with the help of computational studies.

Fig. 2 The effect of aqueous phase pH on the extraction of Am³⁺ and Eu³⁺ with Me₂(t-phen) and Me₂(t-bipy); Org. Phase: 0.03 M ligand + 1 M 2-bromo-octanoic acid in n-dodecane; Aq. Phase: dilute HNO₃ with varying pH.

Fig. 3 A comparison of the selectivity of Am³⁺ over Eu³⁺ with the two ligands (Me₂(t-phen) and Me₂(t-bipy)); Org. Phase: 0.03 M ligand + 1 M 2-bromo-octanoic acid in n-dodecane; Aq. Phase: dilute HNO₃ with varying pH.

Computational studies of free ligands

Analysis of conformational energies of Me₂(t-bipy) and Me₂(t-phen). During the geometry optimization of Me₂(t-bipy) and Me₂(t-phen), the lowest energy (global minimum) was achieved in the trans–trans (tt) and trans (t) conformations, respectively. A number of local minima are, however, possible for this class of molecules. A local minimum was obtained with non co-planar geometry of the aromatic rings when the rings are in cis-conformation. In the case of Me₂(t-bipy), the dihedral angle is as high as 37.8° between the two pyridine rings and 10° between the central pyridine-triazine rings at the local minimum. For the free 2,2′bipyridine molecule, similarly, the global minimum at a dihedral angle of 180° (the trans planar conformation) and a local minimum at around 45° were reported.²⁸ The higher dihedral angle between the two pyridine rings is due to the unfavourable interaction between the ortho-hydrogen atoms of the pyridine rings which is much lower in the case of pyridine-triazine rings as the triazine ring does not contain any ortho-hydrogen atoms. There are four possible low energy conformations (minima) for Me₂(t-bipy) (tt, tc, ct and cc (Fig. 4)), which is described in detail in the literature .⁸ The lowest energy conformation is ‘tt’, but the conformation ‘tc’ is only 0.7 kcal mol⁻¹ higher in energy. The relative energies of the conformations ‘ct’ (7.4 kcal mol⁻¹) and ‘cc’ (8.9 kcal mol⁻¹) are higher in energy as compared to the ‘tt’ and ‘tc’ conformations. The reason for the higher conformational energies of the ‘cc’ and ‘ct’ conformations is the presence of ortho-hydrogen atoms in the two pyridine rings of t-bipy in the ‘cc’ and ‘ct’ conformations, which makes the two pyridine rings non coplanar. The ligand needs to be in the ‘cc’ conformation for the complexation with metal ions. There are two possible ways to convert Me₂(t-bipy) from the most stable ‘tt’ to the complexable ‘cc’ conformation, either through ‘ct’ or through ‘tc’ conformations (Fig. 4). The route through the ‘tc’ conformation is energetically more favourable because of the lower energy of the intermediate ‘tc’ as compared to the ‘ct’ conformation. The transition states (TS) between various minima were also found out. The lesser heights of the two energy barriers in the ‘tc’ route (relative energies of TS are 5.6 and 10.2 kcal mol⁻¹) as compared to that in the ‘ct’ route (relative energies of TS are 8.8 and 12.6 kcal mol⁻¹) may also favour the ‘tc’ route. Me₂(t-bipy), however, has to cross two energy barriers through any of the two routes (Fig. 4). In the case of the ligand Me₂(t-phen), only two low energy conformations (cis and trans) are possible (Fig. 5). The ‘cis’ conformation is higher in energy than its most stable trans-conformation by only 0.5 kcal mol⁻¹ and they are interconvertable through a single barrier of much lower energy (relative energy of TS is 3.2 kcal mol⁻¹) as compared to that in the case of Me₂(t-bipy). The energy required to preorganize the ligand for complexation is, therefore, lower in the case of Me₂(t-phen) as compared to Me₂(t-bipy), resulting in more favourable metal ion complexation in the case of Me₂(t-phen) as compared to Me₂(t-bipy). This is also reflected from the solvent extraction studies, where the distribution ratios of both Am³⁺ and Eu³⁺ were found to be higher in the case of Me₂(t-phen) as compared to Me₂(t-bipy) (Fig. 2).

Fig. 4 The relative energies of different conformations of Me₂(t-bipy) (minima and transition states) varying the angle of rotation around two rings with respect to the most stable ‘tt’ conformation

Fig. 5 The relative energies of different conformations of Me₂(t-phen) (minima and transition states) varying the angle of rotation around two rings with respect to the most stable ‘trans’ conformation

Basicity of the ligands. Protonation energies (E_prot) of the ligands were considered as a measurement of their basicity. The protonation energies were calculated from the difference in the energies of the most stable trans conformation of the ligand and its protonated form in cis-conformation as reported by Benay et al. in case of BTBPs.²⁹ Protonation on the ‘N’ atom of the central pyridine ring was found to be most favourable in case of both the ligands and the protonation energy was found to be more negative in the case of t-phen (−263.2 kcal mol⁻¹) as compared to that in the case of t-bipy (−254.9 kcal mol⁻¹) resulting in higher basicity of t-phen. Benay et al. also, similarly, found a higher basicity of BT-phen (where 2,2′-bipyridine in BTBP was replaced by 1,10-phenanthroline) as compared to BTBP and they attributed this to the preorganized structure of BT-phen in the cis conformation.²⁹

Analysis of frontier orbitals. The energies of the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) were calculated for t-bipy and t-phen at the DFT level of theory. The HOMO–LUMO energy gap was found to be lower in the case of t-phen (92.1 kcal mol⁻¹) as compared to t-bipy (98.3 kcal mol⁻¹) (Fig. 6). The HOMO–LUMO gap is considered to be proportional to the absolute hardness.^30,31 The lower HOMO–LUMO gap of t-phen, therefore, makes it softer than t-bipy. The higher selectivity of t-phen for Am³⁺ over Eu³⁺ as compared to t-bipy, as observed from the solvent extraction studies (Fig. 3), can be explained from the higher softness of t-phen. The higher selectivity of bis-1,2,4-triazinylpyridines (BTP) as compared to terpyridine (terpy) was similarly explained on the basis of lower energy gaps between the frontier orbitals in case of BTP.³¹ In the case of both the ligands, the triazine ring makes very little contribution in the π-donor HOMO and the major contributors here are the central and outer pyridine rings. The triazine ring, however, makes a significant contribution in the ligand′s LUMO, which can accept electrons from the metal centered occupied orbitals through back bonding.

Fig. 6 A Walsh diagram showing the HOMO and LUMO and their energies (kcal mol⁻¹) for Me₂(t-bipy) and Me₂(t-phen)

Computational studies on Am³⁺ and Eu³⁺ complexes of t-bipy and t-phen

Stronger ‘Am–N’ bonding was indicated by the higher sharing electron number (SEN) between the ‘Am’ and ‘N’ atoms in Am-complexes as compared to that between ‘Eu’ and ‘N’ atoms in Eu-complexes (Table 1). If we analyse the occupied valence orbitals of α spin in same cut off value, it is clear that the metal–ligand overlap is higher in the Am³⁺ complexes as compared to the Eu³⁺ complexes for both the ligands (Fig. 9 and ESI†). This indicates higher covalence in the Am³⁺ complexes as compared to the corresponding Eu³⁺ complexes resulting in selectivity of these classes of ligands for Am³⁺ over Eu³⁺. In the case of Am³⁺ complexes of the two ligands, t-bipy and t-phen, the HOMO of the complexes indicates that the overlap between the metal and ligand orbitals was higher in the t-phen complex and hence Am³⁺ complexation is further improved with t-phen as compared to t-bipy resulting in higher selectivity for Am³⁺ over Eu³⁺ using t-phen, and between the two ligands studied, t-phen shows higher covalence with the Am³⁺ ion. The triazine ring was reported to be a better electron acceptor as compared to the pyridine ring when their interaction with different trivalent actinide ions is considered.³² The major part of back bonding is, therefore, expected in the ‘M–N₃’ bond because of the less negative charge on the ‘N₃’ atom in the free ligands and significant contribution of the triazine ring in the LUMO of the ligands (Fig. 6). The shared electrons between the ‘M–N₃’ bond, therefore, estimates the extent of metal–ligand bond strength or bond order in the complexes. For both the ligands, more electrons are shared between Am and N₃ as compared to Eu and N₃ atoms, resulting in stronger bonding in the Am-complexes, and hence selectivity of these ligands for Am³⁺ over Eu³⁺. A higher bond order or stronger metal–ligand bond in the Am-complex is also indicated by the shortening of the ‘Am–N₃’ bond as compared to the ‘Eu-N₃’ bond in the Eu-complexes (Table 1), in spite of marginally higher ionic radius of the Am³⁺ ion (1.230 Å) as compared to the Eu³⁺ ion (1.206 Å) for the same coordination number.³³ If we compare the two ligands, t-bipy and t-phen, more electrons are shared between ‘M’ and ‘N₃’ atoms in the case of t-phen for both the metal ions, indicating stronger bonding in the case of t-phen for both the metal ions. The difference in shared electrons in the ‘Am–N₃’ bond and ‘Eu–N₃’ bond is higher in the case of t-phen (0.082) as compared to the case of t-bipy (0.064), resulting in the higher selectivity of t-phen for Am³⁺ over Eu³⁺ as observed from the distribution studies. In the case of t-phen, all the three ‘M–N’ bonds are shorter in the Am-complex as compared to the Eu-complex, whereas in the case of t-bipy, the ‘M–N₂’ bond is marginally longer in the Am-complex as compared to that in the Eu-complex. This also explains the higher selectivity of t-phen for Am³⁺ over Eu³⁺ as compared to t-bipy. Natural bond order (NBO) analysis or natural population analysis (NPA) is the technique where the orbitals used are the natural orbitals (eigenfunctions of the first order reduced density matrix) in spite of using the molecular orbitals directly for obtaining the occupancies and charges. This results in less basis set dependence as compared to other population analysis schemes, viz. Mulliken population analysis (MPA).³⁴ The charges (Table 1) on the metal and three coordinating ‘N’ atoms in the Am³⁺ and Eu³⁺ complexes of t-bipy (Fig. 7(a)) and t-phen (Fig. 7(b)) were calculated using natural population analysis (NPA) as implemented in TURBOMOLE at their optimized geometries (Fig. 8). The more positive charge on Am as compared to Eu also supports higher metal to ligand back donation in the case of the Am-complex resulting in an increasing electronic charge on the coordinating nitrogen atoms in the Am-complex as compared to that in the Eu-complex.

Fig. 7 The structures of the metal complexes of t-bipy (a) and t-phen (b) used for the DFT calculation (R=H).

Fig. 8 The optimized structures of Am³⁺ and Eu³⁺ complexes of t-bipy and t-phen.

Fig. 9 The frontier orbitals of Am³⁺ and Eu³⁺ complexes of t-BTP and t-phen with same cut off value.

Table 1 The NPA charges on the directly bonded ‘N’ atoms and the metal ions (M), ‘M–N’ bond distances and shared electrons between these two bonded atoms in the Am³⁺ and Eu³⁺ complexes of t-bipy and t-phen (Fig. 8)

	NPA Charges				Bond Length			Shared Electrons
	M	N1	N2	N3	M–N1	M–N2	M–N3	M–N1	M–N2	M–N3
t-bipy	–	−0.43	−0.41	−0.22	–	–	–	–	–	–
Am-t-bipy	2.09	−0.70	−0.64	−0.43	2.381	2.504	2.390	0.710	0.155	0.117
Eu-t-bipy	1.96	−0.69	−0.64	−0.39	2.456	2.500	2.509	0.706	0.034	0.053
t-phen	–	−0.43	−0.40	−0.22	–	–	–	–	–	–
Am-t-phen	2.05	−0.69	−0.64	−0.44	2.469	2.483	2.401	0.470	0.254	0.175
Eu-t-phen	1.94	−0.67	−0.63	−0.42	2.515	2.505	2.480	0.532	0.160	0.093

---

Conclusions

A novel ligand, Me₂(t-phen), with increased rigidity as compared to Me₂(t-bipy), was synthesized and studied for the first time for the separation of trivalent actinides and lanthanides. The selectivity of these ‘N’ donor based ligands was attributed to the higher covalence in their Am-complexes as compared to their Eu-complexes, as indicated by the higher metal–ligand overlap in the case of Am³⁺-complexes as compared to the Eu³⁺-complexes. The stronger ‘Am–N’ bond as compared to the ‘Eu-N’ bond was also reflected in shorter ‘Am–N’ bonds in spite of the bigger size of the Am³⁺ ion, the higher number of shared electrons between ‘Am’ and ‘N’ atoms in the Am³⁺-complexes as compared to ‘Eu’ and ‘N’ atoms in the Eu³⁺-complexes and higher metal to ligand back donation in the Am-complex resulting in a higher residual positive charge on Am as compared to Eu. Improved extraction and separation of Am³⁺ over Eu³⁺ was experimentally observed using the ligand Me₂(t-phen) as compared to Me₂(t-bipy). The higher extraction of Am³⁺ with Me₂(t-phen) was explained on the basis of its lower conformational energy requirement for complexation as compared to Me₂t-bipy and the higher metal–ligand orbital overlap in the Am³⁺–t-phen complex as observed by analyzing the frontier orbitals of the Am³⁺ and Eu³⁺-complexes. The higher selectivity of t-phen for Am³⁺ over Eu³⁺ as compared to t-bipy was because of the higher softness of t-phen as compared to t-bipy as a result of the lower energy gap between the frontier orbitals and the higher value in the difference in shared electrons in the ‘Am–N₃’ bond and ‘Eu-N₃’ bond in case of t-phen as compared to that in case of t-bipy, resulting in a stronger Am³⁺-complex with respect to the Eu³⁺-complex in the case of t-phen as compared to that in the case of t-bipy.

Acknowledgements

DM and TKG would like to thank Dr S. K. Ghosh and Dr T. Mukherjee for their keen interest and constant encouragement.

References

1. J.N. Mathur, M.S. Murali and K.L. Nash, Solvent Extr. Ion Exch., 2001, 19, 357.
2. C. Madic, M. J. Hudson, J. O. Liljenzin, J. P. Glatz, R. Nannicini, A. Facchini, Z. Kolarik and R. Odoj, Report EUR 19149EN, 2000.
3. Z. Kolarik, U. Müllich and F. Gassener, Solvent Extr. Ion Exch., 1999, 17, 23.
4. M.G.B. Drew, M.J. Hudson, P.B. Iveson, M.L. Russell, J.-O. Liljenzin, M. Skalberk, L. Spjuth and C. Madic, J. Chem. Soc., Dalton Trans., 1998, 2973.
5. M.G.B. Drew, M.J. Hudson, P.B. Iveson, J.-O. Liljenzin, L. Spjuth, P.-Y. Cordier, A. Enarsson, C. Hill and C. Madic, J. Chem. Soc., Dalton Trans., 2000, 821.
6. Z. Kolarik, U. Müllich and F. Gassener, Solvent Extr. Ion Exch., 1999, 17, 23.
7. M. J. Hudson, M. G. B. Drew, D. Guillaneux, M. L. Russell, P. B. Iveson and C. Madic, Inorg. Chem. Commun., 2001, 4, 12.
8. M. J. Hudson, M. G. B. Drew, M. R. S. Foreman, C. Hill, N. Huet, C. Madic and T. G. A. Youngs, Dalton Trans., 2003, 1675.
9. E. J. Corey, A. L. Borror and F. Thomas, J. Org. Chem., 1965, 30, 288.
10. F. H. Case, J. Heterocycl. Chem., 1970, 7, 1001.
11. T. Norby, A. Borje, L. Zhang and B. Akermark, Acta Chem. Scand., 1998, 52, 77.
12. P. K. Mohapatra, Ph.D. Thesis, University of Bombay, 1993..
13. L. Petit, C. Daul, C. Adamo and P. Maldivi, New J. Chem., 2007, 31, 1738.
14. L. Petit, C. Adamo and P. Maldivi, Inorg. Chem., 2006, 45, 8517.
15. D. Guillaumont, THEOCHEM, 2006, 771, 105.
16. M. W. Schmidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert, M. S. Gordon, J. H. Jensen, K. Koseki, N. Matsunaga, K. A. Nguyen, S. Su, T. L. Windus, M. Dupuis and J. A. Montgomery Jr., J. Comput. Chem., 1993, 14, 1347.
17. MOLDEN visualization software is developed by CMBI, The Netherlands; http://www.cmbi.ru.nl/molden/molden.html.
18. TURBOMOLE is program package developed by the Quantum Chemistry Group at the University of Karlsruhe, Germany, 1988: R. Ahlrichs, M. Bär, M. Häser, H. Horn and C. Kölmel, Chem. Phys. Lett., 1989, 162, 165.
19. A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38, 3098.
20. J. P. Perdew, Phys. Rev. B, 1986, 33, 8822.
21. W. Kuechle, M. Dolg, H. Stoll and H. J. Preuss, Chem. Phys., 1994, 100, 7535.
22. X. Cao, M. Dolg and H. Stoll, J. Chem. Phys., 2003, 118, 487.
23. X. Cao and M. Dolg, THEOCHEM, 2004, 673, 203.
24. M. Dolg, H. Stoll and H. Preuss, J. Chem. Phys., 1989, 90, 1730.
25. X. Cao and M. Dolg, J. Chem. Phys., 2001, 115, 7348.
26. X. Cao and M. Dolg, THEOCHEM, 2002, 581, 139.
27. M. Dolg, H. Stoll and H. Preuss, Theor. Chim. Acta, 1993, 85, 441.
28. L. Oresmaa, M. Haukka, P. Vainiotalo and T.A. Pakkanen, J. Org. Chem., 2002, 67, 8216–8219.
29. G. Benay, R. Schurhammer and G. Wipff, Phys. Chem. Chem. Phys., 2011, 13, 2922.
30. R.G. Pearson, Proc. Natl. Acad. Sci. U. S. A., 1986, 83, 8440.
31. L. Petit, C. Daul, C. Adamo and P. Maldivi, New J. Chem., 2007, 31, 1738.
32. P. Maldivi, P. Petit, C. Adamo and V. C. R. Vetere, Chemie, 2007, 10, 888.
33. R.D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1976, A32, 751.
34. D. Young, Computational Chemistry: A practical guide for applying techniques to real world problems, John Wiley & Sons, Inc.Canada, 2001, 100.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c2ra00445c

---