Effect of radiolysis of TODGA on the extraction of TODGA/n-dodecane toward Eu(III): an experimental and DFT study

Hang Zhang ·Yin-Yong Ao ·Yue Wang ·Shang-Jie Zhao ·Jia-Yang Sun ·Mao-Lin Zhai ·Jiu-Qiang Li ·Jing Peng ·Hui-Bo Li

Abstract N,N,N′,N′-Tetraoctyl diglycolamide (TODGA) is one of the most promising extractants tailored for high-level liquid radioactive waste treatment during nuclear fuel reprocessing.The γ-radiolysis of TODGA (0.2 mol/L) in n-dodecane (nDD) solution with and without pre-equilibrated 3.0 mol/L HNO3 was investigated using HPLC and UPLC-QTOF-MS and compared with the γ-radiolysis of neat TODGA in this study.With increased absorbed doses, the concentration of TODGA decreased exponentially for the studied systems.Moreover, pre-equilibration with HNO3 (3.0 mol/L) slightly influenced the γ-radiolysis of TODGA in nDD.Seven radiolytic products generated from the rupture of the C-C, C-O, and C-N bonds in TODGA were identified in the studied extraction system.The influence of γ-radiation on TODGA/nDD for the extraction of Eu(III) was evaluated using the first combination of extraction experiments and density functional theory (DFT) calculations, in which the complexations of Eu(III) with TODGA and its radiolytic products were systematically compared.Based on the radiolysis kinetic model of TODGA, the slope curve of the distribution ratio of Eu(III) ( DEu ) and the absorbed dose, and fluorescence titration analysis, the empirical equation of the absorbed dose and DEu was obtained successfully.Below 300 kGy, the experimental DEu agreed well with the obtained empirical equation for TODGA/nDD.Conversely, at a high absorbed dose,the experimental DEu was higher than the theoretical DEu based on the empirical equation because the radiolytic products of TODGA with similar coordination structures still possessed partial complexation toward Eu(III), which was confirmed by DFT calculations.This work provides a method to predict the extraction distribution ratio of an irradiated extractant system and to understand the complex extraction process.

Keywords TODGA ·Radiolytic products ·Extraction ·Density functional theory ·Complexation

1 Introduction

Spent nuclear fuel (SNF) reprocessing generates high-level liquid radioactive waste (HLLW), which is composed of unextracted U and Pu, long-lived fission products (129I,99Tc,135Cs, and93Zr), short-lived fission products (90Sr and137Cs), minor actinides (MAs) (Np, Am, and Cm), and stable lanthanide isotopes (Eu, Nd, La, Tb, Pr, Gd, Sm, Ce,and Pm) [1-10].To reduce the radiotoxicity and volume of HLLW and improve resource utilization, partitioning and transmutation strategies have been explored to shorten the half-life of long-lived radioisotopes.However, the transmutation efficiency of MAs is limited by the high thermal neutron capture cross sections of several lanthanide isotopes owing to competition for available neutrons [11, 12].Thus,separating MAs from lanthanides (Lns) before transmutation is crucial.

During the liquid-liquid extraction process, N,N,N′,N′-tetraoctyl diglycolamide (TODGA) has proven to be effi-cient in extracting actinides (Ans) and lanthanides [2,13-16].Furthermore, TODGA is a tridentate ligand that shows high distribution ratios for Lns(III)/Ans(III) because the hydrophilic groups exhibit stable metal coordination in the aqueous phase and the hydrophobic groups ensure its solubility in the organic phase [17, 18].It has been used in some processes, such as the innovative Selective ActiNide EXtraction (i-SANEX) process [19, 20] and the EUROpean Grouped Actinide EXtraction (EURO-GANEX) process [21,22].The i-SANEX process was first arranged utilizing an organic solution with TODGA (0.2 mol/L) and 1-octanol(5% v/v) to enable co-extraction of trivalent Lns and Ans from simulated HLLW.At the same time, Pd and Zr were masked by cyclohexane diamine tetraacetic acid (CDTA).Second, the co-extracted Mo and Sr were stripped using 3.0 mol/L HNO3aqueous solution with CDTA (0.05 mol/L)and oxalic acid (0.2 mol/L) as scrubbing solvents, followed by a back-extraction step using an aqueous solution containing 2,6-bis(5,6-di(sulfophenyl)-1,2,4-triazin-3-yl)pyridine(SO3-Ph-BTP) (1.8 × 10-2mol/L) in HNO3(0.35 mol/L) for selective stripping of Ans(III).The final step was the stripping of Lns(III) using a citrate buffer solution.The experimental results obtained by Wilden et al.[19] demonstrated the effectiveness of this process.In the Ans product fraction, Cm(III) and Am(III) were efficiently stripped (more than 99.8%) and regained within six stages.The purity of the Cm(III) and Am(III) products was found to be very good, with only Ru (0.4%), Sr (0.3%), and Lns(III) (0.1%)contamination.Lns(III) stripping is efficient (more than 99.5%) within only four stages using a stripping solution containing citric acid with a pH of 3.The EURO-GANEX process consists of two cycles.In the first cycle, hydrogenated tetrapropene (TPH) containing N,N-di (2-ethylhexyl)isobutyramide (DEHiBA) (1.0 mol/L) was applied to extract U(VI) selectively.In the second cycle, the extraction of Cm(III), Am(III), Pu(IV), Np(VI), and Lns(III) from the first cycle raffinate was achieved using N,N’-dimethyl-N,N’-dioctyl-2(2-hexyloxyethyl)-malonamide (DMDOHEMA)(0.5 mol/L) and TODGA (0.2 mol/L) in a kerosene diluent, while CDTA (0.055 mol/L) was applied as a masking agent, to prevent the extraction of Zr(IV) and Pd(II).The Sr and Fe were extracted using HNO3(0.5 mol/L) from the loaded solvent, and trans-uranium elements (TRU) were back-extracted by applying an aqueous solution containing aceto-hydroxamic acid (AHA) and SO3-Ph-BTP, while Lns(III) was retained in the organic solution.The EUROGANEX process was successfully demonstrated by the hot test performed by Malmbeck et al.[23].Moreover, Pu(IV),Np(VI), Am(III), and Lns(III) were efficiently co-extracted from the raffinate of the EURO-GANEX process during the first cycle; Mo and Zr were also co-extracted, whereas the other fission products were successfully refused.Within six strip stages, 99.91% Np, 99.85% Pu, 99.90% Am, and co-extracted Mo and Zr were well back-extracted, whereas 99.9% Lns(III) was retained in the organic solution within four strip stages using an aqueous solution containing AHA and SO3-Ph-BTP.

Previous studies have shown that TODGA provides significantly higher distribution ratios in alkanes such asn-dodecane (nDD) than in other studied diluents, such as secondary ketones, primary alcohols, and secondary alcohols [24, 25].The extraction system of TODGA/nDD can be applied to separate MAs and Lns from other fission products.Once separated, the extraction of individual elements can be simplified with additional processing, culminating in the separation of MAs from Lns [26-28].

Because the extraction system is exposed to the radioactive environment formed by the radionuclides, its radiation stability is an important factor to be examined before its practical application [29].In the solvent extraction process,the γ-radiolysis of TODGA is problematic because of the highly radioactive HLLW.Because the loss of TODGA and the radiolytic products of TODGA formed during γ-radiation can adversely affect the extraction performance in the liquid-liquid extraction process toward metal ions, it is essential to measure the γ-radiolysis yield and radiolytic products of TODGA and to study their effects on the performance of extracting metal ions.Radiolysis of TODGA innDD under different conditions has been investigated in some studies[30, 31].It was found that the magnitude of α-radiolysis of TODGA is lower than that of γ-radiolysis, and they have similar radiolytic products [32].This is attributed to the fact that the linear energy transference of α particles is higher than that of γ particles, and the radical yield ofnDD decreases by recombination in α-radiolysis.Although the γ-radiolysis and extraction of TODGA have been investigated, the effect of γ-radiolysis and radiolytic TOGDA products on extraction in the TODGA/nDD system has rarely been reported.In addition, if we could predict the extraction performance of the irradiated extractant system using equations, it would be very beneficial and important in controlling the partition process.To the best of our knowledge,the mathematical correlation between the radiolysis of the extractant and the distribution ratio has seldom been studied.

In this work, because α and γ rays result in similar radiolysis products of TODGA and γ-radiation is easier to carry out in the laboratory, the radiolytic behavior and γ-radiation stability of TODGA in the 0.2 mol/L TODGA/nDD system (TODGA/nDD) and the 0.2 mol/L TODGA/nDD that was pre-equilibrated with 3.0 mol/L HNO3system(HNO3-TODGA/nDD) were studied qualitatively and quantitatively.They were also compared with radiolysis of neat TODGA.The content and radiolysis products of TODGA were determined using HPLC and UPLC-QTOF-MS,respectively.Considering the chemical similarity between trivalent lanthanides and minor actinides, and to avoid the dangers of handling radioactive isotopes in standard laboratories, Eu(III) was selected as a model metal ion in our extraction experiment to assess the extraction performance of TODGA/nDD after γ-radiation.The density functional theory (DFT) calculation, which can estimate the coordination abilities of TODGA and its radiolytic products, was applied to explain the effect of the radiolytic products on our extraction of Eu(III).For the first time, a combination of theoretical calculations and extraction experiments was used to demonstrate that the radiolytic products maintain partial complexation, which capably explains the extraction behavior of irradiated samples.For the TODGA/nDD system, an in-depth understanding of the effects of γ-irradiation on the extraction behavior was obtained from this study.

2 Experiments

2.1 Materials

Sichuan University synthesized TODGA (≥ 98%); its purity was confirmed by FT-MS (ESI+) and1H NMR (Fig.S1-S2).Moreover, Eu(NO3)3·6H2O (99.99%) was obtained from MREDA Technology Co.Ltd.Formic acid, liquid chromatography-mass spectrometry (LC-MS) grade, was purchased from Thermo Fisher Scientific.Methanol (LC-MS grade)was purchased from Honeywell Trading (Shanghai) Co.,Ltd.The aqueous dilution was performed using ultrapure water (18.2 MΩ·cm).All other chemicals were of analytical grade, and the chemicals were not further purified.

2.2 Gamma irradiation

TODGA and TODGA/nDD solutions equilibrated with and without HNO3(3.0 mol/L) before γ-radiation were irradiated in the air at room temperature (25 ± 5 °C) using a60Co source (Department of Applied Chemistry in the College of Chemistry, Peking University, China) with an absorbed dose rate of 6.6 kGy/h, as determined by a Fricke dosimeter.

2.3 Quantitative analysis of TODGA after irradiation

An HPLC method with UV detection (SPD-16, Shimadzu,Japan) was used to quantitatively analyze the TODGA concentration in the irradiated samples.Additionally, HPLC was carried out at 25 ± 5 °C using an Agilent HC-C18 liquid chromatography column (5 μm, 4.6 mm × 150 mm) to achieve chromatographic separation with a UV wavelength fixed at 210 nm.The aqueous component was ultrapure water with formic acid (1‰ v/v) (A1); the organic component was methanol (B1).The elution program in our liquid chromatography was listed hereafter: 0-10 min, 30-100%B1; 10-20 min, 100% B1; 20-22 min, 100-30% B1.The error values in the quantitative analysis experiments were within 5%.The flow rate was 1.0 mL/min, and the injection volume was 20 μL.

2.4 Identification and semiquantification of radiolytic products

Ultra-high-pressure liquid chromatography (UPLC)(ACQUITY I-Class, Waters, USA) was carried out using an Agilent HC-C18 liquid chromatography column (5 μm,4.6 mm × 150 mm) at 40 °C to achieve chromatographic separation.The aqueous component was ultrapure water with formic acid (1‰ v/v) (A2), and the organic component was methanol with formic acid (1‰ v/v) (B2).The liquid chromatography elution program was listed hereafter: 0-4 min:30-100% B2; 4-13 min: 100% B2; 13-13.1 min: 100-30%B2; 13.1-15 min: 30% B2.The flow rate was 0.8 mL/min with a 1 μL injection volume.The mass spectrometer conditions of the quadrupole time-of-flight mass spectrometry(QTOF-MS) (Vion IMS QToF, Waters, USA) are listed here: desolvation temperature: 300 °C; source temperature:120 °C; desolvation gas flow: 800 L/h; capillary voltage:2.8 kV; cone gas flow: 50 L/h; positive mode.The data were acquired and processed using a Waters UNIFI Scientific Information System with a mass target match tolerance of less than 5 ppm.

2.5 Extraction experiment

The TODGA was dissolved withnDD to obtain a 0.7 ml organic phase with a concentration of 0.2 mol/L.The Eu(NO3)3·6H2O was dissolved in 3.0 mol/L HNO3to obtain a 0.7 ml aqueous phase with a concentration of 1000 mg/L.Pre-equilibration of the organic extraction phase with 3.0 mol/L HNO3was performed before extraction.A vortex mixer at a speed of 2500 rpm (LPD2500, Leopard Scientific Instrument Co., Ltd, Beijing) was used to perform the extraction experiments at 298 ± 1 K.To achieve complete separation between the two phases, the mixtures were centrifuged for 1 min at 5000 rpm in a centrifuge (TGL-16 M,Cence, Hunan).After dilution with ultrapure water, the aqueous phase was detected using an ICP-MS (ELEMENT XR, Thermo Scientific, USA) to estimate the concentration of Eu(III).

The distribution ratio of Eu(III) (DEu) was calculated using Eq.(1).The phase ratio equals 1:1 in the extraction experiment.

whereCfandCirepresent the final and initial Eu (III) concentrations in the diluted aqueous phase, respectively.The extraction had an error rate of less than 5%.

2.6 Fluorescence spectra titration

The fluorescence emission spectra of the titration experiments were investigated at 298 ± 1 K in a cuvette cell with a 1 cm path length using a fluorescence spectrophotometer (F7000,Hitachi, Japan).The concentration of TODGA was 0.1 mol/L in acetonitrile and the initial concentration of Eu(III) was 0.001 mol/L in 2.0 mL acetonitrile.Next, 4 μL of TODGA solution was added at a time.The solution was mixed using a vortexer (VORTEX 3, IKA, Germany) for 5 min after the addition of each ligand.The titration experiments used an excitation wavelength of 395 nm; the bandwidth of the excitation and emission was 2.5 nm.The interval was 1 nm and the spectra were recorded between 550 and 700 nm.Single-component spectra of the metal-ligand complexes and metal solvent species were obtained using the HyperSpec program.

2.7 Theoretical calculations

DFT calculations incorporating electron correlation effects were performed at the level of B3LYP using the Gaussian 09 package [33-35].The relativistic effects of Eu atoms were investigated using the quasi-relativistic effective core potentials (RECPs) and associated valence basis sets, which were improved by the Stuttgart and Dresden groups [36-40].The large-core RECPs used in the structural optimization of Eu(III)contained 52 electrons [39, 40].All other C, N, O, and H atoms utilized the 6-31G (d) basis set.Geometrical optimization and the electronic calculations of all structures were first performed at the level of B3LYP/6-31G(d)/RECP under the gas phase conditions.The Gibbs free energy (Gg), entropy(Sg), and enthalpy (Hg) were obtained under the gas phase conditions at 298.15 K using the same theory level.To obtain the Gibbs free energy (Gaq), entropy (Saq), and enthalpy (Haq)of the species in thenDD phase at 298.15 K, all species were optimized at the B3LYP/6-31 G (d)/RECP level innDD to better predict their solvation energy [41].This theory is based on the SMD universal continuum solvation model[42].

3 Results anddiscussion

3.1 Quantitative analysis of TODGA under different conditions after γ-ray irradiation

TODGA/nDD, HNO3-TODGA/nDD, and neat TODGA were subjected to γ-irradiation with absorbed doses ranging from 20 to 500 kGy.The irradiated samples were measured quantitatively using HPLC-UV, and their radiolytic stability was analyzed.Figure 1 shows that the TODGA concentration decreased exponentially with increasing absorbed doses, suggesting pseudo-first-order degradation kinetics for TODGA/nDD, HNO3-TODGA/nDD, and neat TODGA.By comparing TODGA/nDD and HNO3-TODGA/nDD, it was found that HNO3had little influence on the radiolysis of TODGA.Moreover, in thenDD solution, the radiolysis rate of TODGA was higher than that of the neat TODGA.A dodecane-induced "sensitization effect" is responsible for the sensitization of TODGA to radiolysis [43] due to the fact that the radiolysis of TODGA is accelerated bynDD solvents because of the efficiency ofnDD in transferring positive charges [44-46].

Because the radiolysis behavior of TODGA depends on its concentration in pseudo-first-order kinetics, the rate of radiolysis causes an exponential drop in the TODGA concentration.The radiolysis rate equation is defined by Eq.(2):

The solution to the rate equation is presented as Eq.(3):

whereCTODGA(mol/L) is the concentration of TODGA,(mol/L) is the original concentration of TODGA before γ-radiation,R(kGy) is the absorbed dose, and the dose constantk(kGy-1) is the coefficient obtained by fitting the exponential curve.

Fig.1 (Color online) CTODGA∕C0TODGA of irradiated TODGA/nDD,HNO3-TODGA/nDD, and neat TODGA as a function of the absorbed dose ( R ) (y: CTODGA∕C0TODGA , x: R)

The dose constantkcan be used to scale the absorbed dose when exponential behavior occurs [47].Using all the information acquired during the irradiation experiment provides more accurate results, and standard statistical methods can be used to assess the accuracy of the results.The application ofkto characterize an exponential curve in terms of the absorbed dose is very similar to that of traditional timedependent kinetics.As expressed by exponential Eq.(3), the reciprocal absorbed dose is governed by the first-order rate law instead of the reciprocal time.

If the dose constantskof different systems can be obtained, Eq.(4) can be used to calculate theGvalue(μmol/J), which is known as the radiation chemical yield:

whereρis the solution density (kg/L) (ρTODGA= 0.91 andρ0.2mol∕LTODGA∕nDD= 0.77 (25 °C).The measurement results are shown in Table S1.

Table 1 shows the dose constantskand radiation chemical yieldsGof irradiated TODGA/nDD, HNO3-TODGA/nDD,and neat TODGA.The dose constantkof TODGA innDD is higher than that of neat TODGA because of the “sensitization effect,”while the radiation chemical yieldsGof TODGA/nDD and HNO3-TODGA/nDD are lower than that of neat TODGA because of the different initial concentrations of irradiated TODGA.These results agree with the literature in which Sugo et al.reported that thekvalue of 0.2 mol/L TODGA/nDD was 2.2 × 10-3kGy-1[43, 48].By using thekvalue, we can predict that the absorbed doses for the half-loss of TODGA concentration (R0.5) in irradiated TODGA/nDD, HNO3-TODGA/nDD, and neat TODGA will be approximately 330, 315, and 1060 kGy, respectively.

3.2 Identification and semiquantitative analysis of radiolytic products

Ultra-high-performance liquid chromatography with quadrupole time-of-flight mass spectrometry (UPLC-QTOF-MS)can be used to identify and quantify complex mixtures of unknown compounds owing to its high resolution and mass accuracy.

The major radiolytic products of TODGA were identified using QTOF-MS; these are presented in Table 2, and their UPLC-QTOF-MS spectra are shown in Figures S3-9.Seven mass signals were identified and attributed to the possible radiolytic products of TODGA (P1-P7, Table 2).These radiolytic products were formed by the rupture of the C-N,C-O, and C-C bonds of TODGA.Four radiolysis routes after irradiation are shown in Scheme 1.

Figure 2 shows the semiquantitative analysis of the radiolytic products P1-P7in the irradiated HNO3-TODGA/nDD and neat TODGA.The absorbed doses ranged from 20 to 500 kGy.Seven radiolytic products were found to exist in the two systems.The detector counts of the radiolytic products P1-P7roughly increased with increasing absorbed dose.However, some radiolytic products, such as P1in irradiated neat TODGA, showed a decrease in detector counts with absorbed doses from 400 to 500 kGy.This is probably because the contents of the radiolytic products during irradiation increased from the radiolysis of the TODGA and decreased by self-radiolysis.The concentration’s rate of increase was less than the rate of decrease with absorbed doses from 400 to 500 kGy for P1in irradiated neat TODGA.The detector counts of all radiolytic products in irradiated TODGA/nDD were higher than those in neat TODGA (Figure S10), indicating the “sensitization effect”ofnDD.This agreed well with the above result of the radiolysis kinetics of the TODGA with or withoutnDD.However, the relative detector counts of some radiolytic products differed between the two irradiation systems.For example, the detector count of P6was less than that of P4in irradiated neat TODGA; however, the converse was true for irradiated HNO3-TODGA/nDD.This can be attributed to the different reaction mechanisms, wherein irradiated neat TODGA produces radical cations by direct ionization reaction.Conversely, in irradiated TODGA/nDD,nDD radical cations may transfer their charge to TODGA molecules, leading to radiolysis of TODGA [43, 49].

3.3 Influence of γ-ray irradiation on the extraction of TODGA system toward Eu(III)

HNO3-TODGA/nDD and neat TODGA were subjected to γ-ray irradiation at absorbed doses ranging from 50 to 500 kGy.The irradiated neat TODGA was diluted withnDD to 0.2 mol/L.These organic phases were used to extractEu(III) from the 3.0 mol/L HNO3aqueous phase to evaluate the extraction performance.

Table 1 Dose constants k , radiation chemical yields G , and absorbed dose for half-loss of concentration of TODGA R0.5 of irradiated TODGA/nDD, HNO3-TODGA/nDD, and neat TODGA

Table 2 Radiolytic products of TODGA determined by the QTOF-MS

As shown in Fig.3,DEudecreased as the absorbed dose increased, but it was slower for irradiated TODGA than for irradiated HNO3-TODGA/nDD, indicating that γ-ray irradiation of the ligand significantly affected the extraction of Eu(III).This further proves that the presence ofnDD increased the radiolysis of TODGA.However, the HNO3-TODGA/nDD system after γ-ray irradiation still possesses a highDEu(2.2 × 103), even at 500 kGy.

The loss of extractants was the main cause of the decline in extractability after irradiation.Owing to the presence of TODGA and its liquid radiolysis products in the irradiated organic extraction phase, the complexation of radiolytic products on Eu(III) was studied using DFT calculations.This determined the effect if any of the radiolytic products of TODGA on the extraction behavior of the irradiated extraction system.

3.4 Slope analysis and fluorescence titration

Before DFT calculations were conducted to investigate the complexation of TODGA and its radiolytic products with Eu(III), we assessed the stoichiometry of Eu ions and TODGA ligands in the extraction reaction, which could be obtained by the slope analysis in the curve of logDM- log[L](org.).

The extraction reaction of Eu(III) by TODGA is expressed by Eq.(5):

Scheme 1 Radiolysis routes of TODGA after γ-irradiation

The extraction equilibrium concentration constant,Kex,is described as Eq.(6):

The distribution ratio,DEu, is defined as Eq.(7):

Equation (7) was substituted in Eq.(6), and Eq.(6) was then transformed into a logarithmic form to obtain Eq.(8).

Thenvalue can be calculated from Eq.(8), which represents the average number of TODGA coordinated to one metal ion.The plot in Fig.4 of logDEu-log[TODGA](org.)at a constant aqueous solution of 3.0 mol/L HNO3shows that the slope is 3.10 ± 0.11, which indicates that Eu(III) would form complexes with three TODGA molecules.

Complexation studies were also performed with TODGA and Eu(III) using fluorescence titration.Figure 5 displays the normalized fluorescence emission spectra with different ratios of Eu(III) and TODGA resulting from the transitions of 5D0→ 7F1and 5D0→ 7F2.The transition of 5D0→ 7F1at 593 nm, owing to the magnetic-dipole transition, was considered independent of the ligand field.The transition of 5D0→ 7F2at 617 nm owing to the electric-dipole transition is considered to be hypersensitive to the ligand field, and its intensity depends on the coordination symmetry around the metal ions [50].With the addition of the TODGA ligand,the change in the ligand field led to a gradual transformation of the single peak at 617 nm into a double peak, which was finally formed atM:L= 1:3.This indicated that the inner coordination sphere of Eu(III) was gradually occupied by TODGA with the addition of the ligand, which is similar to some results from the literature [51, 52].

Fig.2 (Color online) Semiquantitative analysis of radiolytic products P1-P7 in: (a) irradiated neat TODGA and (b) HNO3-TODGA/nDD with absorbed doses ranging from 20 to 500 kGy

Fig.3 (Color online) DEu of irradiated neat TODGA and HNO3-TODGA/nDD for extraction of Eu(III) versus the absorbed dose

Fig.4 Plot of logDEu versus log[TODGA](org.) at a constant aqueous 3.0 mol/L HNO3 solution (x: log[TODGA](org.) , y: logDEu)

The spectra of each component were obtained from the fluorescence emission spectra (Fig.6).The [Eu(TODGA)n]3+(n= 1-3) complexes and Eu(III) solvent species were also discovered.No complexes with TODGA values greater than or equal to 4 were found in the fluorescence emission spectra, proving that the ratio of Eu(III) to TODGA equals 1:3.

Fig.5 (Color online) Normalized Eu(III) emission spectra as a result of the transitions of 5D0 → 7F1 and 5D0 → 7F2 for the complexation of Eu(III) with increasing TODGA in acetonitrile

Fig.6 (Color online) Single-component spectra of the[Eu(TODGA)n]3+ complexes (n = 1-3) and Eu(III) solvent species

The slope value was disputable, although many studies showed that the Ln(III): DGA structures are 1:3 complexes.Antonio et al.elucidated the inner coordination sphere of Eu(III) and TODGA using X-ray absorption spectroscopy(XAS) [53].One ether O atom and two carbonyl O atoms of a TODGA molecule make up each ligand’s tridentate to Eu(III).The coordination number is nine, with twelve distant carbon neighbors, six closest ones from carbonyl carbon atoms, and six slightly farther ones from the ether carbon atoms.Turanov et al.[54] found that the calculated slope value of Eu(III) extracted by TODGA in n-decane containing 0.002 mol/L dinonylnaphtalene sulfonic acid(HDNNS) was 3.28 ± 0.04.This was estimated by the plot of logDEu-log[TODGA] and the existence of EuLn(NO3)3species by the slope of logDEu-log[HNO3], which was equal to 2.86 ± 0.08.Pathak et al.[55] investigated the influence of solvent type, acidity of extraction, and concentration of TODGA on luminescence lifetime and found that various ratios of TODGA to Eu ions resulted in the formation of Eu(TODGA)33+species when Eu(III) was complexed with TODGA.According to the findings of Sasaki et al.[56],extracted Eu(III) complexes require three or four TODGA molecules and three nitrate ions to maintain stability in nonpolar diluents by slope analysis.Zhu et al.[57] performed a slope analysis of TODGA to extract some Lns(III) and found that the slope was 3.9 for Eu(III), suggesting that the M(TODGA)4(NO3)3species finally formed.However, in this work, a 1:3 complexation ratio for TODGA with Eu(III) was determined.

The theoreticalDEuvalues at different concentrations of TODGA can also be obtained using Eq.(8).Equation (9) is obtained by linear fitting, as shown in Fig.4.

Equation (10) can be obtained from Fig.1 for irradiated HNO3-TODGA/nDD.It is an exponential radiolysis equation for the concentration of TODGA and the absorbed dose.

We can obtain Eq.(11) forDEuand the absorbed dose by substituting Eq.(10) in Eq.(9).This empirical equation enables us to construct a mathematical relationship between theDEuand absorbed dose in this system.

The theoreticalDEuwith absorbed doses ranging from 50 to 500 kGy was obtained using Eq.(11).As shown in Fig.7,the experimentalDEuwas higher than the theoreticalDEuafter irradiation, particularly at a high absorbed dose above 300 kGy for the HNO3-TODGA/nDD system.The experimentalDEuis close to the theoreticalDEubelow 300 kGy,indicating that the decrease inDEuis mainly related to the loss of TODGA concentration, with a slight influence from radiolytic products.

Because some radiolytic products remained in coordination structures similar to those of TODGA, such as P1and P2, which were identified by UPLC-QTOF, the complexation of radiolytic products on Eu(III) was investigated further using DFT calculations.

3.5 Theoretical calculations of complexation of TODGA and radiolytic products with Eu(III)

According to the slope and fluorescence titration analyses,the Eu(III): TODGA stoichiometry was 1:3.In addition,Kimberlin et al.[58] reported that several radiolytic products with the TODGA skeleton are involved in heteroleptic complexes with TODGA with a stoichiometry of 1:3 from the ESI-MS spectra.The time-resolved laser fluorescence results showed no change in the peak shape of the fluorescence spectra before and after irradiation (Fig.S11).However, the fluorescence lifetime gradually decreased with increasing absorbed dose (Fig.S12), indicating that γ radiation did not change theM:Lratio of 1:3 but caused a slight partial change in the original fluorescent species.

Fig.7 Experimental logDEu and theoretical logDEu after irradiation for HNO3-TODGA/nDD

DFT calculations were employed to investigate the coordination abilities of TODGA and its radiolytic products with TODGA skeletons (P1and P2).Figure 8 shows the optimized structures of the complexes formed by Eu(III) and the ligands.Theoretical calculations were performed for the [Eu(TODGA)2(Pn)]3+(n= 1-2) species and[Eu(TODGA)3]3+complexes.It is realistic that one TODGA molecule in the initial complex is replaced by one radiolytic product to form mixed Ln-radiolytic product-TODGA complexes during γ irradiation, which was already found in the ESI-MS spectra by Kimberlin et al.[58].In addition, we investigated the optimized structures and coordination abilities of [Eu(Pn)3]3+(n= 1-2), which were formed using three radiolytic products to represent a more extreme coordination environment of Eu(III) after γ irradiation.

The changes in the Gibbs free energy, entropy, and enthalpy for the complexes formed by the ligands (TODGA,P1and P2) and Eu(III) are shown in Table 3.It can be seen that the formation of [Eu(TODGA)3]3+complex has more negative ΔG in the gas phase (- 2527.3 kJ/mol)and in thenDD phase (- 1115.0 kJ/mol) than that of the[Eu(TODGA)2(Pn)]3+and [Eu(Pn)3]3+(n= 1 - 2) complexes in the two phases.Moreover, the coordination abilities of the radiolytic products of TODGA decreased, as reflected by the changes in the Eu-O bond length (Table 4).The average Eu-O bond length of the complexes changed from 2.4911 Å([Eu(TODGA)3]3+) to 2.5000 Å ([Eu(P2)3]3+), indicating a decline in the coordination ability of the complex.Theseresults indicate that TODGA shows better coordination ability than the radiolytic products of TODGA; the complexation order is TODGA ＞ P1＞ P2.However, the formation of the [Eu(TODGA)2(Pn)]3+and [Eu(Pn)3]3+complexes (n= 1 - 2) had a high negative ΔG in the two phases.Theoretical calculations also indicated that the radiolytic products of TODGA, such as P1and P2, still maintain a good coordination ability with Eu atoms, proving that the radiolytic products retain partial complexation for Eu(III).Consequently, the experimentalDEuof TODGA was higherthan the theoreticalDEufrom the loss of TODGA with the increase in radiolytic products at a high absorbed dose.Furthermore, if the contents of P1and P2are determined exactly,by combining the TODGA contents, we can predict theDEuof the irradiated TODGA/nDD system more precisely.This study will be conducted in future works.In our previous experiments, we found that the radiolytic products P1and P2were difficult to synthesize, and the radiolytic product P1was not sufficiently stable.Therefore, we used DFT calculations to evaluate the coordination abilities of these radiolytic products, thereby justifying the experimental result that the experimentalDEuwas higher than the theoreticalDEudue to the high dose TODGA losses in this study.

Fig.8 (Color online) Optimized structures of the ligands(TODGA, P1 and P2) and complexes formed by Eu(III);the ligands (light pink, blue,green, red, and white spheres represent Eu(III), N, C, O, and H, respectively)

Table 3 Changes in Gibbs free energy, entropy, and enthalpy (298.15 K, kJ/mol) for the complexes formed by ligands (TODGA, P1 and P2) and Eu(III) were acquired separately in (a) gas phase, as well as (b) nDD at the level of B3LYP/6-31G(d)/RECP

Table 4 Changes in Eu-O bond length of complexes formed by Eu(III) and ligands (unit: Å)

In conclusion, by combining the exponential equation for the radiolysis kinetics of TODGA innDD and the equation for slope analysis of the concentration of TODGA andDEu, we can derive a mathematical equation that includes the absorbed dose andDEu, which could be used to predict the extraction performance at low absorbed doses.This method can be used to study other extractant systems.This work provides in-depth insights into the effect of γ irradiation on the extraction behavior of the TODGA/nDD system and provides useful information on the use of nuclear fuel reprocessing.

4 Conclusion

In summary, γ-radiolysis of TODGA/nDD and HNO3-TODGA/nDD was investigated and compared with γ-radiolysis of neat TODGA by analyzing the loss of TODGA and liquid radiolytic products.When the absorbed dose was increased, the TODGA concentration decreased exponentially, and the dose constant of TODGA innDD was higher than that of neat TODGA because of the “sensitization effect”ofnDD.However, the radiolysis of TODGA innDD was only slightly affected by the pre-equilibration with HNO3.Seven radiolytic products of TODGA were identified and semiquantified using UPLC-QTOF-MS, and four radiolysis routes for TODGA were proposed.TheDEuof the irradiated systems decreased after irradiation, but the HNO3-TODGA/nDD system irradiated with an absorbed dose of 500 kGy maintained a highDEuof 2.2 × 103.For the irradiated HNO3-TODGA/nDD, we obtained an empirical equation betweenDEuand the absorbed dose by combining the radiolysis kinetics of TODGA and the complexation equation of TODGA with Eu(III), which fits well with the experimental results with absorbed doses below 300 kGy.DFT calculations demonstrated that the radiolytic products of TODGA with similar coordination structures possessed good coordination abilities with Eu atoms, which led to the experimentalDEubeing higher than the theoreticalDEubased on the ligand content at high absorbed doses.

Supplementary InformationThe online version contains supplementary material available at https:// doi.org/ 10.1007/ s41365- 023- 01198-z.AcknowledgementsThe measurements of UPLC-QTOF and ICP-MS were performed at the Analytical Instrumentation Center of Peking University.We acknowledge the help from PKUAIC (Dr.Jiang Zhou,Dr.Li Zhang, and Dr.Jia-Hui Liu).The authors also wish to acknowledge Prof.Song-Dong Ding (Sichuan University) for his assistance in providing TODGA.

Author contributionsAll authors contributed to the study conception and design.Material preparation, data collection and analysis were performed by Hang Zhang, Yin-Yong Ao, Hui-Bo Li and Jing Peng.The first draft of the manuscript was written by Hang Zhang, and all authors commented on previous versions of the manuscript.All authors read and approved the final manuscript.

Data availabilityThe data that support the findings of this study are openly available in Science Data Bank at https:// www.doi.org/ 10.57760/ scien cedb.j00186.00037 and http:// resol ve.pid21.cn/ 31253.11.scien cedb.j00186.00037.