A Novel Type of Tetradentate Dipyridyl-Derived Bis(pyrazole) Ligands for Highly Efficient and Selective Extraction of Am3+ Over Eu3+ From HNO3 Solution

ABSTRACT The extraction of Am3+ and Eu3+ with 6,6’-bis(5-alkyl-1H-pyrazol-3-yl)-2,2’-bipyridine (BPzBPy, alkyl = i-Bu, n-Bu, n-Oct) from HNO3 medium and the complexation of BPzBPy with Eu3+ were investigated. By using meta-nitrobenzotrifluoride (F-3) as a diluent, iBu-BPzBPy ligand in combination with 2-bromohexanoic acid was able to extract Am3+ over Eu3+ with the nature of significantly high efficiency, relatively fast extraction kinetics, and easy stripping. Slope analysis showed the formation of a 1:2 metal/ligand extraction species. The analyses of electrospray ionization mass spectrometry (ESI-MS), Fourier transform infrared (FT-IR) and time-resolved laser fluorescence spectrum (TRLFS) revealed that the composition of the extracted complex was [Eu(H2O)L2]A3 (L = ligand; HA = 2-bromohexanoic acid). The complexation of iBu-BPzBPy ligand with Eu3+ was an enthalpy-driven spontaneous, exothermic, and entropy-decreasing process. Besides, the complex stability constants were also obtained via UV–vis spectrophotometric titration. Combining the results of solvent extraction and complexation studies, a cation exchange extraction model was also proposed. GRAPHICAL ABSTRACT


Introduction
Nuclear energy as a clean, efficient, and low-carbon source of electricity is being widely used in many countries worldwide. [1] However, a large amount of strongly radioactive and highly toxic spent nuclear fuel will inevitably be generated in the production of nuclear energy. Thereinto, the long-term radiotoxicity of spent fuel mainly comes from the plutonium and minor actinides (MAs), such as Np, Am, and Cm. [2] Although Pu can be recovered via the PUREX (Plutonium Uranium Reduction EXtraction) process, [3] and be reused as mixed oxide fuel, the remaining high-level liquid waste (HLLW) still contains the MAs that are responsible for the major long-term radiotoxicity. [4] To relieve the burden of these nuclear wastes on the environment, a partitioning and transmutation (P&T) strategy has been presented, i.e., the MAs are partitioned from the fission products and transmuted into short-lived or stable nuclides in an advanced nuclear reactor or accelerator-driven system. [5,6] Nevertheless, the fission product lanthanides (Lns) in HLLW must be removed before MAs transmutation due to their high neutron capture cross-sections. [7,8] Thus, it is essential to separate MAs from Lns for P&T strategy.
The separation of trivalent MAs from Lns has been recognized as one of the most challenging tasks in spent nuclear fuel reprocessing because of the very similar physical and chemical properties between them. [9,10] However, it is considered that the greater availability of the valence orbitals in MAs suggests a more covalent contribution to metal-ligand bonding with MAs than with Lns. [11] Based on this small but significant difference, a selective extraction separation of MAs from Lns may be accomplished by utilizing a ligand with soft S-or N-donor atoms as an extractant. [12] In contrast to the S-donor ligands, the N-donor ligands comprising only C, H, O, and N not only have better chemical and radiolytic stabilities but also can be totally incinerated to gas products without any sulfur-containing residual waste. [13] As a result, the research and development of N-donor ligands get increasing attention.
So far, the most representative N-donor ligands used for the study of the separation of trivalent MAs from Lns are the triazinyl-based ligands such as BTP, BTBP, and BTPhen. [14,15] These triazinyl-based ligands showed good separation efficiencies for trivalent MAs over Lns. However, due to issues with hydrolytic and radiolytic instabilities, [16,17] or slow extraction kinetics, [18,19] none of these ligands were suitable for an industrial-scale application. To address the problem of hydrolytic and radiolytic instabilities, a kind of new tetradentate nitrogen-based extractant EH-BTzBP [20] and EH-BTzPhen [21] was developed by substituting a triazole ring for a triazine ring in BTBP or BTPhen ligands. They demonstrated fast extraction kinetics, high solubility in non-polar solvents, and resistance to acid hydrolysis and radiolytic conditions. However, for the EH-BTzPhen ligand, a white precipitate was easily formed by using the mixture of 1-octanol and Aliquat-336 as a diluent when the nitric acid solution was above 0.50 mol/L.
In recent years, a novel kind of pyrazole-pyridine-based ligands such as BPPhen, [22] Cn-PypzH, [23] Dbnpp, [24] BisPypzH, [25] and C5-BPP [26] has received a lot of attention because of their excellent MAs/Lns separation effect, high hydrolysis and radiolysis stability, [22,27,28] as well as fast extraction kinetics. For example, BPPhen ligand was found to be able to selectively extract Am 3+ from Eu 3+ with high distribution ratio (D Am ) and separation factor (SF Am/Eu ) values, both of which were in the magnitude of 1.0 × 10 2 . Meanwhile, BPPhen also displayed excellent acid tolerance and hydrolytic stability on long-term exposure to a highly acidic solution (5.0 mol/L HNO 3 for 100 days). [22] Particularly, the C5-BPP ligand showed remarkable irradiation stability in contact with nitric acid. Only a 30% reduction in the extractant concentration was found after the irradiation dose of 150 KGy.
For a solvent extraction system with a promising application, it is required not only strong extractability but also a good stripping performance. Unfortunately, it is difficult for the abovementioned pyrazole-pyridine-based ligands to be stripped because of the strong affinity between the extracted metal ions and them. For example, for the BPPhen ligand, even if a high HNO 3 concentration of 3.0 mol/L is used as a stripping agent, a back extraction at 3-4 stages are still needed to obtain a complete stripping of Am 3+ and Eu 3+. [22] To achieve efficient stripping at low acidities, less polar diluents such as hydrogenated tetrapropene (TPH) or toluene as co-diluents have to be added to reduce the D values. For instance, using n-octanol and toluene as diluent and co-diluent, respectively, the D Am value for 2,9-bis (5,5,8,8-tetramethyl-5,6,7,8-tetrahydro-1,2,4-benzotriazin-3-yl)-1,10-phenanthroline (CyMe 4 -BTPhen) ligand could decrease from 1.57 in 80/20 (v/v) n-octanol/toluene to 0.11 in 20/80 (v/v) n-octanol/toluene with 0.0010 mol/L HNO 3 solution as the aqueous phase. [29] However, the introduction of co-diluents makes the extraction system more complicated and also greatly restricts their practical applications. This necessitated modifying the structure of phenanthroline-derived ligands. As a result, 6,6'-bis (5,5,8,8-tetramethyl-5,6,7,8-tetrahydro-benzo [1,2,4]triazin-3-yl)-2,2'-bipyridine (CyMe 4 -BTBP) has been optimally chosen to be a prime candidate for SANEX (Selective ActiNide EXtraction) process due to its appropriate extractability and easy stripping, [30][31][32] where the bipyridine fragment was substituted for the more rigid phenanthroline fragment. In this way, the extraction efficiency for CyMe 4 -BTBP towards Am 3+ was lower than that for CyMe 4 -BTPhen but without obviously compromising its selectivity. CyMe 4 -BTBP consisting of a bipyridine structure has stronger practicality and applicability. Thus, along the same lines, it would be reasonable to expect that the ligands obtained by substituting bipyridine fragment for phenanthroline fragment in BPPhen ligands may be able to efficiently separate Am 3+ over Eu 3+ not only with good extractability but also with easy stripping at low acidity.

Materials and instruments
All the reagents were procured from Adamas (Shanghai Adamas Reagents Co., Ltd., China) with analytical grade unless otherwise stated. The stock solution of Eu 3+ ions was obtained by dissolving Eu(NO 3 ) 3 ·6H 2 O purchased from Energy-Chemical (Saan Chemical Technology Co., Ltd., China) in HNO 3 solution. 241 Am and 152 Eu radiotracers with 99.9% radiochemical purity dissolved in 0.010 mol/L HNO 3 solution were provided by the China Institute of Atomic Energy. 1 H NMR spectra were measured using dimethyl sulfoxide-d6 as a solvent on a Bruker 400 MHz NMR spectrometer (Bruker Inc., Switzerland). ESI-MS spectra were recorded on a Bruker Amazon SL spectrometer (Bruker Inc., Switzerland). FT-IR spectra were performed on a Nicolet Nexus 670 model instrument (Thermo Fisher Scientific Inc., USA). TRLFS was measured by a HORIBA TEMPRO-01 transient fluorescence spectrometer (HORIBA Ltd., Japan) with a laser wavelength of 370 nm. The UV-vis spectra were collected with a UV-3600 I PLUS spectrophotometer (Shimadzu Scientific Instruments, LLC., Japan).
To prepare the iBu-BPzBPy-Eu complex, the iBu-BPzBPy ligand (2.5 mmol, 1.0 g) dissolved in dry methanol (20 mL) was added to a solution of Eu(NO 3 ) 3 ·6H 2 O (1.25 mmol; 0.56 g) in dry methanol (20 mL). The resulting solution was stirred in a 100 mL round bottom flask at room temperature for 24 h. During the stirring process, the precipitate was formed gradually in the flask and then was collected by filtration after the reaction finished, followed by washing with the methanol three times and drying under a vacuum to yield a white solid complex.

Solvent extraction
The solvent extraction was conducted by using 25 mmol/L BPzBPy plus 1.0 mol/L 2-bromohexanoic acid in F-3 as the organic phase and 1.0 mmol/L Eu(NO 3 ) 3 spiked with a trace amount of 241 Am or 152 Eu dissolved in HNO 3 solution as the aqueous phase. Prior to the extraction of metal ions, the organic phase was preequilibrated with an equal volume of the corresponding concentration of HNO 3 solution to keep the acidity constant during the extraction process. The organic and aqueous phases (1.0 mL) were then contacted in a 5.0 mL tube for 40 min using a magnetic stirrer at 500 rpm under a water bath of 25 ± 0.5°C. After the samples were centrifuged to separate phases, about 0.50 mL of both phases were taken out for measurement. 241 Am and 152 Eu activities were assayed using a well-type NaI(Tl) scintillation counter, and the reported results were the average values of three measurements. separately. They are simply obtained from the organic and aqueous phase count rates of the respective radionuclides. The selectivity for Am 3+ over Eu 3+ is defined as the separation factor SF Am/Eu = D Am /D Eu .

FT-IR
The solid sample was mixed with dried KBr powder and sufficiently ground. The KBr/sample mixture forms a clear disc when pressed under high pressure using a hydraulic press. And then the FT-IR spectra of the clear sample were recorded on a Nicolet Nexus 670 spectrometer in a mode of KBr transmission. The FT-IR spectra were obtained in the 4000-400 cm −1 region for a total of 32 scans at 2 cm −1 resolution.

TRLFS
TRLFS study on the complexation of iBu-BPzBPy ligand with Eu 3+ for an F-3/HNO 3 or F-3/HClO 4 biphase system was performed. The organic phase and the aqueous phase were prepared by dissolving 25 mmol/L iBu-BPzBPy plus 1.0 mol/L 2-bromohexanoic acid into F-3 solution, and by dissolving 12.5 mmol/L Eu(NO 3 ) 3 in 0.50 mol/L HNO 3 , or Eu(ClO 4 ) 3 in HClO 4 solution, respectively. Each 3.0 mL of the two phases was mixed in a 10 mL tube and shaken at 25.0 ± 0.5 °C for 40 min, followed by centrifuging to separate phases. And then, the organic phase (2.0 mL) was sampled and taken into a 4.0 mL cuvette. The samples were excited at 370 nm, and emission spectra were collected between 550 and 750 nm. Finally, the emission decay curves were fitted into the exponential function to obtain the lifetimes/decay rates of the excited states using FelixGX Software from the PTI QuantaMaste.

UV-vis spectrophotometric titration
UV-vis spectrophotometric titration of BPzBPy ligand with Eu 3+ was conducted in chromatographic-grade methanol by recording the absorption spectra in the region 210 − 390 nm at 15 °C, 20 °C, 25 °C, 30 °C, and 35 °C, respectively, to give the complexation thermodynamic parameters. A solution of 2.0 × 10 −5 mol/L BPzBPy was kept in a 4.0 mL cuvette and titrated with 4.0 × 10 −4 mol/L Eu(NO 3 ) 3 . A certain amount of tetraethylammonium nitrate (Et 4 NNO 3 ) was added into the cuvette as the background electrolyte to keep the ionic strength of all the stock solutions at 0.10 mol/L. For each titration, about 10 μL aliquot of Eu(NO 3 ) 3 solution was added into the cuvette and mixed for about 5.0 min. And then UV-vis spectra were recorded in the wavelength region of 220 − 350 nm. The titration was carried on until further spectral variations in absorbance were barely detectable. The overall constant of the Eu 3+ /L complexes was obtained by employing the HypSpec program from the obtained spectral data. [35][36][37] When dealing with the data, the spectroscopic data was imported into the HypSpec software first. Instrument absorbance data files are ASCII data files typically produced by a spectrophotometer, that is, they contain an absorbance spectrum. And then a complete spectrum in HYPERQUAD format was recorded and saved after entering relevant experimental conditions, such as the volume and concentration of the titrant. At last, the wavelength in the HYPERQUAD formatted file, which is most suitable for the calculation, was chosen and fit to calculate the stability constant using the program.

Solvent extraction
Preliminary screening tests had revealed that pyrazole-pyridine-based ligands alone in the commonly used diluents for N-donor heterocyclic ligands, such as F-3, tert-butyl benzene, and octanol scarcely extracted Am 3+ and Eu 3+. [38,39] Fortunately, their extraction power can be significantly enhanced by introducing 2-bromohexanoic acid into the organic phase. [40] Thus, the influence of diluent on the extraction was investigated in the presence of a co-extractant of 2-bromohexanoic acid as shown in Table S1. It can be seen that the iBu-BPzBPy ligand dissolved in F-3 solvent displayed excellent extractability and selectivity for Am 3+ over Eu 3+ . Meanwhile, it was also observed that 2-bromohexanoic acid itself alone hardly extracted these two ions (Table S2). Consequently, F-3 was chosen as a diluent for the following study of solvent extraction using 2-bromohexanoic acid as a co-extractant.

Influence of contact time
For ensuring a full solvent extraction equilibrium, the influence of contact time on the extraction of Am 3+ and Eu 3+ by BPzBPy in a mixed solvent of F-3 and 2-bromohexanoic acid was first examined and shown in Figure 1. It is obvious that the extraction equilibrium could be attained within only 30 min of phase contact time, demonstrating relatively fast extraction kinetics. Thus, the phase contact time for a complete equilibrium was determined as 40 min in the following extraction experiments.

Influence of HNO 3 concentration in aqueous phase
The influence of HNO 3 concentration on the extraction was shown in Figure 2. In the examined HNO 3 concentration range of 0.10 − 3.0 mol/L, both the D Am and D Eu values for three BPzBPy ligands decreased gradually with increasing acidity. There may be two main factors contributing to this trend. On the one hand, due to their high affinity between N-donor atoms and H + ions, BPzBPy ligands tend to be protonated in contact with increasing acidity in the aqueous phase, decreasing the extraction of metal ions. [41,42] On the other hand, under high acidity conditions, the dissociation of 2-bromohexanoic acid was inhibited, reducing the synergistic extraction effect of 2-bromohexanoic acid. [20] Meanwhile, the plot of log D vs. log [HNO 3 ] gives two straight lines with slope values close to 3.0 both for three ligands, suggesting that three H + or NO 3 − ions take part in the extraction process. However, which one is involved in the extraction needs to be further confirmed by the subsequent extraction experiment. However, the slope was slightly lower than the expected value of 3.0. This deviation may be caused by the partition of 2-bromohexanoic acid in the aqueous phase due to its deprotonation at lower acidity. [20] Besides, it can also be observed that the extractability for the three BPzBPy ligands changed in the order of iBu-> nOct-> nBu-BPzBPy at the same experimental conditions. Among them, the extractability of iBu-and nOct-BPzBPy was quite similar, with the former being slightly stronger than  the latter. This phenomenon can be attributed to two opposing effects: longer alkyl chains with a larger steric hindrance for the metal ion coordination with the ligand decreased the extraction ability of the ligand while increasing the solubility of the extracted metal complex in the organic phase. [43] The SF Am/Eu values for these ligands are listed in Table 1. As shown in the table, the BPzBPy ligands demonstrated significantly higher selectivity for Am 3+ over Eu 3+ . Especially, for iBu-BPzBPy, at around 0.50 mol/L HNO 3 , the SF Am/Eu value can reach as high as 276, which is superior to other N-heterocyclic ligands such as BTP, [44] EH-BTzBP, [20] and C5-BPP. [26] This suggests a promising application in the An/Ln separation.
With respect to the An/Ln separation, the acidity of the feed liquid issued from the back-extraction following the co-extraction of Ans and Lns by trialkyl phosphine oxides (TRPO), N,N'-dimethyl,N,N'-dioctylhexylethoxymalonamide (DMDOHEMA), or n-octyl(phenyl)-N,N-diisobutylcarbamoylmethylphosphine oxide (CMPO), etc., is usually between 0.10 and 0.50 mol/L HNO 3 . [45] Thus, further extraction investigation was conducted from 0.50 mol/L HNO 3 solution by using iBu-BPzBPy as a representative of the BPzBPy ligands.

Influence of the ligand and 2-bromohexanoic acid concentration in organic phase
In our previous work, it was found that the pyrazole-pyridine-based ligand alone was incapable of extracting Am 3+ and Eu 3+ from HNO 3 solution and required combining with 2-bromohexanoic acid as a co-extractant to stimulate extraction. [22-25,] Consequently, the investigation on the influence of the ligand and 2-bromohexanoic acid concentration was carried out, by which the number of iBu-BPzBPy ligand and 2-bromohexanoic acid molecules involved in the extraction process can be determined via slope analysis. The results are shown in Figure 3. Both D Am and D Eu values increase with increasing the concentration of iBu-BPzBPy and 2-bromohexanoic acid. Furthermore, the plots of log [iBu-BPzBPy] (org.) vs. log D revealed two straight lines with slope values of 2.02 and 1.98 for Am 3+ and Eu 3+ in Figure 3a, severally, indicating the formation of a 1:2 metal/ligand extracted complex. Similarly, the plot of log [2-bromohexanoic acid] (org,) vs. log D also gave two straight lines with slopes close to 3.0 in Figure 3b, revealing that the extraction process involved three 2-bromohexanoic acid molecules. It is worth noting that the acidity of the aqueous phase was maintained at 0.010 mol/L for the experiment of 2-bromohexanoic acid concentration dependence to ensure the dissociation of 2-bromohexanoic acid (pK a ≈ 2.97) was not inhibited. [ where M, L, and HA represent metal ion, iBu-BPzBPy ligand, and 2-bromohexanoic acid, severally. The subscripts aq. and org. denote the aqueous phase and organic phase, separately. The extraction obeys a cation-exchange model. For a solvating extraction system involving N-donor ligand, the electroneutrality of an extracted species is usually acquired through the coordination of counter-ions to the metal center, such as NO 3 − ions in HNO 3 or NaNO 3 media. [20] As a result, the D values would increase with increasing NO 3 − concentration. Whereas, for the aforementioned cation exchange extraction system involving 2-bromohexanoic acid, the metal center seems to be surrounded by three 2-bromohexanoic acid counter-ions. This suggests that the NO 3 − ions might not involve in the extraction process. To further confirm the proposed extraction model, the influence of NO 3 − concentration was performed. In Figure 4, the D values for both Am 3+ and Eu 3+ hardly changed with the variation of NO 3 − concentration, suggesting independence of extraction equilibrium on NO 3 − ions. This result is also consistent with the cationexchange model.

Loading capacity
The loading capacity of an extractant in the organic phase denotes its maximum extraction capacity to the extracted metal ions and is also an important selection criterion for extractants in solvent extraction. [46] Thus, the loading test for iBu-BPzBPy ligand to Eu 3+ was conducted by maintaining the initial concentrations of iBu-BPzBPy and 2-bromohexanoic acid constant at 2.5 mmol/L and 1.0 mol/L in F-3 diluent, severally, and changing the concentration of Eu(NO 3 ) 3 in 0.50 mol/L HNO 3 . Under the experimental conditions, the emulsification or the formation of two organic phases was not observed. Plotting the ratio of the original ligand concentration (c L ) to the extracted Eu 3+ concentration (c Eu,org. ) in the organic phase versus the original Eu 3+ concentration in the aqueous phase (c Eu,aq. ), the loading curve could be gotten as shown in Figure 5. It is observed that the ratio of iBu-BPzBPy to the extracted Eu 3+ in the organic phase was close to 2.0 at the c Eu,aq. value exceeding 12.5 mmol/L, meaning that each mole of iBu-BPzBPy extractant can load a half mole of Eu 3+ . This result also coincided with that obtained by the slope analysis of solvent extraction mentioned above.

Stripping
The stripping performance is one of the important indexes to evaluate whether a solvent extraction system has a good application prospect and has to be examined. As a reverse process of extraction, stripping would be promoted by the factors that negatively affect extraction. As shown in Figure 2, both D Am and D Eu values are very low as aqueous phase acidity was above 1.0 mol/L, suggesting that high acid conditions may be favorable for their stripping. Thus, the stripping of metal ions from the loaded organic phase was studied by changing the HNO 3 concentration. As shown in Table 2, HNO 3 shows a significantly superior stripping efficiency than oxalic acid. Meanwhile, the accumulated stripping percentage (E) gradually increases with the increase of the HNO 3 concentration, which was well matched to the influence of HNO 3 concentration on extraction as depicted in Figure 2. Furthermore, Am 3+ and Eu 3+ are stripped almost quantitatively by 1.0 mol/L HNO 3 solution at only three and one stages, respectively. Thus, 1.0 mol/L HNO 3 solution was chosen as the stripping agent for Am 3+ and Eu 3+ .

Reusability of the extractant
In terms of application, the extractants used in the extraction process must be regenerated and reused for further extraction without marked loss of efficiency. To assess the reusability of the extractant, 25 mmol/L iBu-BPzBPy plus 1.0 mol/L 2-bromohexanoic acid dissolved in F-3 was contacted with 1.0 mmol/L Eu(NO 3 ) 3 with a trace quantity of 241 Am or 152 Eu in 0.50 mol/L HNO 3 solution and then stripped by 1.0 mol/L HNO 3 in ten successive cycles. For each step of the stripping, a fresh stripping agent was used. The D values were determined to evaluate the stability of the resultant lean organic phase. It was observed in Figure 6 that even after 10 successive loading and stripping cycles, the D value for Am 3+ and Eu 3+ was almost constant at the initial value of about 30 and 0.12, respectively, indicating that the extractant had excellent reusability. Additionally, the metal ions accumulated recovery percentage (E) was still performed by stripping the organic phase with successive cycles. The E values for both Am 3+ and Eu 3+ were almost 100 %, which further suggests the excellent reusability of the iBu-BPzBPy extractant.

Complexation of Eu 3+ with BPzBPy ligand
To obtain more information about the complex composition and structure, the complexation behavior of BPzBPy ligand with metal ions was studied in biphasic and monophasic systems, severally, contributing to the deep understanding of the extraction model mentioned above. At present, we were not able to run the experiments with radioactive material due to the fact that there was not enough americium available in our laboratory and just used the nonradioactive Eu 3+ for the following complexation studies.

ESI-MS analysis
Over the past decade, mass spectrometry (MS) has been increasingly employed in composition analyses of the complexes during the solvent extraction process. ESI-MS is particularly useful for this purpose. [14] In particular, the in-situ ESI-MS analysis of the organic phase loaded with metal ions after extraction can reflect the actual situation of solvent extraction, which is good for understanding the extraction model. Thus, the ESI-MS investigation for the Eu-loaded organic phase was carried out in an HNO 3 medium, and the positive ESI-MS are shown in Figure 7a. The strongest peak at m/z 951.3700 could be attributed to the 1:2 Standard deviation calculated from three independent measurements is lower than 5.0%. b No metal ions were detected in the organic phase. complex [Eu(iBu-BPzBPy-H) 2 ] + . It implies that one Eu 3+ ion was surrounded by two ligand molecules in the extracted species, which was consistent with the result obtained by slope analysis. Meanwhile, the peak at m/z 401.2403 could correspond to the adduct of [iBu-BPzBPy+H] + , revealing the presence of free ligands in the organic phase. Besides, the NO 3 − was not observed in the extracted complex from Figure 7a, meaning that it probably did not participate in the inner coordination sphere of the Eu 3+ . To confirm this view, the ESI-MS analysis for the organic phase after contacting with the aqueous phase containing Eu(ClO 4 ) 3 in the HClO 4 medium was also performed. It is well known that ClO 4 − has almost no affinity for Eu 3+ and cannot enter the inner sphere of Eu 3+ in an aqueous solution. [47] Thus, it can be determined whether NO 3 − ion participated in the coordination through the comparison between HNO 3 and HClO 4 systems. In Figure 7b, the mass spectra for the two systems were almost the same. This provides further evidence that NO 3 − ion did not involve the formation of an extracted complex. Besides, the negative ESI-MS was shown in Figure S10. The mass spectra for the two systems are also the same, and there is only one peak of m/z = 192.9865 attributed to [A] − (HA = 2-bromohexanoic acid), indicating that 2-bromohexanoic acid was involved in the extraction process. Although there are a few minor differences between extraction and complexation environments, the ESI-MS results could also provide some indirect evidence for solvent extraction about the influence of ligand concentration, i.e., the formation of 1:2 metal/ligand extracted complex.

FT-IR and TRLFS analyses
Infrared spectroscopy can provide information about the interaction between the metal ion and ligand according to the positions and shifts of characteristic vibration peaks. [48] For deep insight into the complexation of Eu 3+ with iBu-BPzBPy ligand, the FT-IR analyses for free Eu(NO 3 ) 3 ·9H 2 O, iBu-BPzBPy ligand, and iBu-BPzBPy-Eu complex were carried out. As shown in Figure 8, the most characteristic bands for iBu-BPzBPy ligand are those assignable to the pyrazole and dipyridine group: [ν(C=C), ν(C=N)] at 1559 cm −1 and [δ(C=C), δ(C=N)] at 1465 cm −1 , respectively. [49] Whilst in the spectrum of iBu-BPzBPy-Eu complex, these two bands obviously shifted to higher wavenumber at about 1581 cm −1 and 1478 cm −1 , separately, revealing a strong affinity interaction between the four N atoms in the dipyridyl framework and Eu 3+. [50] Besides, the broadband at 3376 cm −1 in the free Eu(NO) 3 ·9H 2 O spectrum could be attributed to the H 2 O molecule. This band was also observed in the spectrum of iBu-BPzBPy-Eu complex at 3384 cm −1 , indicating the possible presence of coordinated H 2 O molecules. [51] In order to further probe the presence of water molecules in the inner sphere of the extracted complex, TRLFS analysis for the Eu-loaded organic phase following contact with HNO 3 or HClO 4 was performed. Based on the determined fluorescence lifetime of the extracted complex, the number of water molecules present in the inner coordination sphere of Eu 3+ can be obtained from the following equation: where τ denotes the lifetime (ms). [52,53] The fluorescence decay curves are shown in Figure 9. and Eu(ClO 4 ) 3 in HClO 4 solution (Figure 9c) was around 112 µs. The corresponding N H 2 O value was calculated to be nearly 9.0, indicating that the initial Eu 3+ species in the aqueous phase are identical, Eu(H 2 O) 9 3+ in both cases. [54] After the two aqueous solutions were contacted with the organic phase containing the ligand, both of the two decay curves become less steep, and the lifetime increased to 0.64 ms and 0.62 ms for HNO 3 (Figure 9b) and HClO 4 system (Figure 9d), respectively, indicating one H 2 O molecule being present in the inner sphere of the central Eu 3+ in the extracted complexes. It should be noted that the same results are achieved for HNO 3 and HClO 4 systems. Generally, the ClO 4 − ion hardly participates in the inner sphere of the metal ion. So, this means that the NO 3 − ion, just like the ClO 4 − ion, did not involve in the coordination of Eu 3+ , which was also consistent with the results obtained by solvent extraction and ESI-MS analysis.
In general, the number of coordination for trivalent f-elements in the complexes is about 8, 9, or 10. [55,56] According to the results of solvent extraction and complexation study, it could be deduced that two iBu-BPzBPy ligands are bonded to the metal ion via four N atoms on bipyridine rings and four N atoms on pyrazole rings. Additionally, one O atom in an H 2 O molecule is also involved in the inner coordination. This satisfies the coordination number of the metal ion. As a result, the extraction model for Am 3+ and Eu 3+ with iBu-BPzBPy ligand in HNO 3 solution can be described via a cation exchange model. During the extraction process, 2-bromohexanoic acid is primarily responsible for transporting the metal ions from the aqueous phase to the organic phase. It releases its proton to neutralize the charge of the metal ion. Once the complex of 2-bromohexanoic acid with metal ion is generated, iBu-BPzBPy can form an adduct with them by replacing the remaining H 2 O molecules, thus enhancing the lipophilic nature of the extracted species. [57,58]

UV-vis spectrophotometric titration analysis
The extraction power of a ligand depends on its affinity for metal ion that can be described by a complex stability constant. Utilizing the analysis of UV-vis spectrophotometric titration spectra, the complex stability constant of the ligand with metal ion can be obtained. Moreover, due to the fact that the complexation equilibrium strongly depends on the temperature, the thermodynamic parameters for the complexation can also be acquired from the temperature dependence of the complex stability constant. [59,60] where R denotes the gas constant. The absorption spectra of iBu-BPzBPy titrated with Eu(NO 3 ) 3 at various temperatures are shown in Figure S11 (a), and the molar absorbance spectra obtained by the HypSpec program are shown in Figure S11(b). Three absorbing species, i.e. the free iBu-BPzBPy and two Eu 3+ /iBu-BPzBPy complexes (1:1 and 1:2 types) were observed, revealing the presence of two kinds of complexes during the titration. The measured stability constants of the complexes as a function of reciprocal temperature are displayed in Figure 10. As can be observed, the β value decreases with increasing temperature, implying that the complex formation is exothermic and favorable with decreasing temperature. [61] Meanwhile, the plot of log β vs. 1/T also yields two straight lines. Based on the values of slope and intercept, the apparent thermodynamic parameters for the complexation were obtained. In Table 3, it is obvious that the step complexation of both EuL 3+ and EuL 2 3+ species is driven by the favorable enthalpic changes, accompanying entropy reduction. [62] The exothermic effect can be related to the strength of the cation-ligand interaction. [63] The negative ΔS o value suggests an increased order degree and a decreased freedom because of the displacement of solvent molecules by the ligands. [64] Besides, the negative ΔG o values mean that the complexation is a spontaneous process. Overall, from a thermodynamic perspective, the EuL 2 3+ species is formed more easily than the EuL 3+ species, which is reflected in the more negative ΔG o value. It also agrees with the results from slope analysis and ESI-MS studies.
Additionally, the UV-vis spectrophotometric titration spectra for the other two BPzBPy ligands, viz. nBu-BPzBPy and nOct-BPzBPy, titrated with Eu(NO 3 ) 3 in CH 3 OH at 25 °C were also obtained and displayed in Figure S12 and Figure S13, separately. The determined stability constants for these three ligands are listed in Table 4. The log β value varies in the order of iBu-BPzBPy > nOct-BPzBPy > nBu-BPzBPy, which is in close conformity with that of their extractability mentioned above.

Conclusions
Three dipyridyl-derived bis(pyrazole) ligands were synthesized and developed for the extraction of Am 3+ and Eu 3+ from the HNO 3 medium. It has been shown that the iBu-BPzBPy ligand had excellent selectivity toward Am 3+ than Eu 3+ with the SF Am/Eu value ranging from 70 to 276. Slope analysis revealed that 1:2 metal/ligand extracted species were formed. Stripping was easy by using 1.0 mol/L HNO 3 solution as a stripping agent. The complexation of Eu 3+ with ligand was a spontaneous, exothermic, and entropy-decreasing process. There were two tetradentate ligand molecules (L) and an H 2 O molecule was present in the inner coordination of Eu 3+ , along with three 2-bromohexanoic acid anions (A) in the outer sphere coordination as counter ions to equilibrium the charge of the coordinated metal ion. The extracted complex composition was inferred as [Eu(H 2 O)L 2 ]A 3 . The Am 3+ and Eu 3+ extraction followed a cation exchange extraction model. Although the radiolytic and hydrolytic stability experiments were not carried out in this study, other pyrazole-pyridine-based ligands such as BPPhen and C5-BPP have displayed excellent high hydrolysis and radiolysis stability. Overall, the iBu-BPzBPy ligands demonstrate a promising application in the separation of trivalent Ans from Lns due to strong extractability, high selectivity, and easy stripping.

Disclosure statement
No potential conflict of interest was reported by the author(s).