Astrid
Barkleit
*ab,
Jerome
Kretzschmar
a,
Satoru
Tsushima
a and
Margret
Acker
c
aInstitute of Resource Ecology, Helmholtz-Zentrum Dresden – Rossendorf, P.O. Box 510119, 01314 Dresden, Germany. E-mail: a.barkleit@hzdr.de
bRadiochemistry, Department of Chemistry and Food Chemistry, Technische Universität Dresden, 01062 Dresden, Germany
cCentral Radionuclide Laboratory, Technische Universität Dresden, 01062 Dresden, Germany
First published on 6th May 2014
Thermodynamic parameters for the complex formation of Am(III) and Eu(III) with lactate were determined with UV-vis and time-resolved laser-induced fluorescence spectroscopy (TRLFS) in a temperature range between 25 and 70 °C. The reaction enthalpy decreased with increasing ionic strength. ATR FT-IR and NMR spectroscopy in combination with density functional theory (DFT) calculations revealed structural details of the Eu(III) lactate 1:1 complex: a chelating coordination mode of the lactate with a monodentate binding carboxylate group and the hydroxyl group being deprotonated.
The understanding of the complex formation behavior of radionuclides with such small organic molecules and the thermodynamic quantification of the interaction is of great importance to simulate and predict their migration behavior in the environment. Especially data at elevated temperatures are crucial, because not only in various organisms but particularly in the near field of nuclear waste disposals higher temperatures are prevailing.1
We investigated the complex formation and thermodynamic data of Am(III) and its non-radioactive analogue lanthanide Eu(III) with lactate. Lactate was selected as a representative ubiquitous small organic molecule that exists as metabolite in all organisms and also in significant amounts in clay rock formations.2 Experiments were performed at ambient and elevated temperatures with time-resolved laser-induced fluorescence spectroscopy (TRLFS), and for Am(III) additionally with UV-vis spectrometry. Furthermore, spectroscopic investigations concerning structural features have been carried out for Eu(III) lactate with attenuated total reflection Fourier transform infrared (ATR FT-IR) spectroscopy and nuclear magnetic resonance (NMR) spectroscopy, supported by calculations with density functional theory (DFT).
For the Eu(III) as well as the Am(III) lactate system, several investigations have been published (Eu(III),3–12 Am(III)3,10,13–15). But studies about the complex formation behavior at trace metal concentration, lower ionic strength and higher temperatures, which are important parameters influencing the migration behavior of radionuclides in the environment, are still missing. The proposed combination of methods is highly suitable to fill this gap. TRLFS as a sensitive and selective technique has been extensively used to analyze actinide and lanthanide complex formation with inorganic and organic ligands at trace metal concentrations.16,17 The application of TRLFS onto Am(III) complexation was up to now limited because of its much lower luminescence intensity and much shorter lifetime in comparison to Cm(III) or Eu(III). Some publications about TRLFS with Am(III) at ambient18–26 or low temperatures22 exist, but no studies at elevated temperatures have been published until now.
Some structural suggestions for the Eu(III) lactate, which do exist are only assumptions from indirect methods.4,5,12 In this work, we want to provide direct structural information. ATR FT-IR spectroscopy combined with calculations of structure and spectroscopic data using DFT gives useful information about structural features as it has been shown previously for the Eu(III) complexes with pyromellitic and citric acid.27,28 Lanthanide induced shifts in NMR spectroscopy as caused by the interaction of nuclear spins with electronic unpaired spins can be used as a helpful tool for signal separation, probing the potential binding sites and structure including geometries and distances.29
The combination of all these methods should offer new insights concerning the structure of the Eu(III) lactate complex thereby resolving contradicting suggestions in the previous works whether the hydroxyl group is protonated or not.4,5,12,30
The TRLFS measurements for Am(III) were carried out with a pulsed Nd:YAG-MOPO laser system from Spectra Physics (Mountain View, USA), combined with a Spectrograph M270 and an ICCD camera system Spectrum One from Horiba-Jobin Yvon. The time difference between the trigger of the laser system and the start of the camera was adjusted by a delay generator from Spectrum One. The excitation wavelength of the laser source was varied between 503 and 508 nm with pulse energies of 10 mJ. Emission spectra were recorded between 625 and 773 nm, averaging 10 spectra with accumulating 80 laser pulses for each spectrum. The gate width of the camera was set to be 1 μs. The step width between two spectra in time-resolved mode was 2 ns, 50 to 60 delay steps (up to 120 ns) were measured for every sample. The spectrograph and the camera system were controlled by Spectramax from Horiba-Jobin Yvon.
The TRLFS measurements for Eu(III) were carried out with a pulsed flash lamp pumped Nd:YAG-OPO laser system from Continuum as described31 at an excitation wavelength of 394 nm and a gate width of 1 ms for all measurements. Static and time-resolved luminescence spectra of Eu(III) were recorded in the range of 565–650 nm (1200 lines mm−1 grating, 0.2 nm resolution, 2000 accumulations) and 440–780 nm (300 lines mm−1 grating, 0.7 nm resolution, 200 accumulations), respectively. For time-resolved measurements, 41 spectra were recorded with 20–50 μs separation.
The spin-orbit effect and multiconfigurational character of the system were neglected. The first coordination sphere around Eu was saturated with water molecules fixing the coordination number to 8 or 9. The rest of the solvation shells were also considered through the use of the PCM model.
Static luminescence spectra of Eu(III) have been normalized to the peak area of the 5D0→7F1 transition, which is a magnetic dipole and therefore not influenced by complexation.
The fluorescence decay lifetimes were calculated by fitting the integrated luminescence signal to a sum of exponential decay functions:
(1) |
E(t) is the total luminescence intensity at time t, Ei0 the luminescence intensity of the species i at the time t = 0, and τi the corresponding decay lifetime.
The number of water molecules in the first coordination shell was determined from the luminescence lifetimes τ (in ms). For Am(III), the empirical formula from Kimura and Kato19 (eqn (2)), and for Eu(III), the linear relationship developed by Horrocks and Sudnick39 and the resultant empirical formula from Kimura40 (eqn (3)), were used:
n(H2O) ± 0.5 = 2.56 × 10−4 × τ−1 − 1.43 for Am(III) | (2) |
n(H2O) ± 0.5 = 1.07 × τ−1 − 0.62 for Eu(III) | (3) |
The complex stability constants were determined from the absorption or luminescence spectra by using the factor analysis program specfit.41 Input parameters for the data fitting were the total concentrations of the metal ion and the ligand, the pH, and the pKa of lactate from literature (pKa1 = 3.69,42 pKa2 = 11.20,30 recalculated to I = 0.1 M). A brief description of the operation mode of this program43 and of the fitting procedure is given elsewhere.44
Thermodynamic data were calculated with the modified linear form of the van't Hoff equation:
(4) |
The extrapolation of the constants to infinite dilution, I = 0, was done applying the Specific Interaction Theory (SIT) using the IUPAC software for Ionic Strength Corrections.45 The ion interaction parameters ε were taken from ref. 46 (based on ref. 47,48) for Eu3+,ClO4−, from ref. 48 for Am3+,ClO4− and from ref. 48,49 for Na+,ClO4−, and Na+,CH3COO− (acetate as analog for lactate as it is proposed in ref. 50), whereas that for Eu3+,Lac− and Am3+,Lac− were calculated using the guidelines given in ref. 50. The temperature dependencies of ε and the Debye–Hückel parameter B can be neglected.48,50 Values for the Debye–Hückel parameter A as function of temperature have been calculated from literature data51 within the program.45
Fig. 1 Absorption spectra of 5 μM Am(III) in dependence of the lactate concentration (50 μM to 0.1 M), pH = 6.0, I = 0.1 M, T = 25 °C (left), deconvoluted spectra of the single species (right). |
The quantitative analysis of the spectra clearly shows the formation of three Am(III) lactate complexes with a lactate concentration up to 0.1 M, Am(Lac)2+, Am(Lac)2+ and Am(Lac)3, as it was expected from literature.3,13Fig. 1B shows the deconvoluted single spectra of each individual species. They are very similar to those recently determined in trifluoromethansulfonate media.13 The simultaneous determination of all three complex formation constants results in relatively large uncertainties. Only for the 1:1 complex, the quality could be increased by using only the spectra up to a lactate concentration of 0.01 M. In this concentration range the 1:1 complex should be the dominating species. The complex formation constant was determined to be log β11 = 2.22 ± 0.11, providing that only the carboxylic group of lactate is deprotonated. This is in accordance with literature3,10,13,15 (see Table 1). The quality of the formation constants of the 1:2 and 1:3 complexes, respectively, could not be increased properly; possibly the spectral changes are too small to get more precise results than log β12 = 4.5 ± 0.3 for Am(Lac)2+, and log β13 = 6.3 ± 0.3 for Am(Lac)3. Nevertheless, these values are in the range of published data3,10,13,15 (see Table 1).
T/°C | I/M (NaClO4) | Am(Lac)2+ | Am(Lac)2+ | Am(Lac)3 | Ref. |
---|---|---|---|---|---|
log β11(1) | log β12(2) | log β13(3) | Method | ||
log βML(H) are the stability constants with protonated hydroxyl group(s), not considering the pKa2 of lactic acid.a Electrophoresis.b Solvent extraction.c UV-vis.d TRLFS. NaTf = Na-trifluoromethansulfonate; p.w. = present work. | |||||
10 | 1.5 | 2.57 | 4.21 | 10 | |
25 | 2 | 2.52 ± 0.04 | 4.77 ± 0.05 | 5.98 ± 0.08 | 15 |
1 | 2.43 ± 0.09 | 4.23 ± 0.27 | 5.65 ± 0.15 | 3 | |
1 (NaTf) | 2.60 ± 0.06 | 4.7 ± 0.1 | 6.2 ± 0.2 | 13 | |
0.1 | 2.27 ± 0.05 | 4.5 ± 0.3 | 6.3 ± 0.3 | p.w.c | |
0.1 | 2.22 ± 0.11 | p.w.d | |||
0 | 2.87 ± 0.26 | 5.5 ± 0.4 | 7.5 ± 0.4 | p.w. | |
45 | 0.1 | 2.17 ± 0.19 | p.w.d | ||
0 | 2.82 ± 0.31 | p.w. | |||
65 | 0.1 | 2.35 ± 0.31 | p.w.d | ||
0 | 3.03 ± 0.39 | p.w. |
Fig. 2 Emission spectra of (A) 5 μM Am(III) and (B) 5 μM Am(III) + 0.1 M lactate in dependence of the excitation wavelength. |
The Am(III) aqua ion shows at pH 6.0 and different temperatures a luminescence lifetime of 23.8 ± 2.4 ns (25 °C), 22.8 ± 0.9 ns (40 °C) or 23.3 ± 1.5 ns (65 °C), corresponding to approximately 9 coordinating water molecules. This agrees with previous measurements.18,25 Complex formation with lactate causes a strong increase of the luminescence intensity and a red shift of the luminescence maximum of about 5 nm (Fig. 3, left). The luminescence decay is always mono-exponential, irrespective of the number of expected different Am(III) species (see Fig. S1 and S2, ESI†). This is caused by an exchange of the Am(III) coordination environments, which is faster than the luminescence decay rate of the excited state and results in a concentration-weighted average number of water molecules of all Am(III) species.52 The luminescence lifetime is prolonged up to 38.3 ± 0.4 ns (25 °C, 0.1 M lactate, pH 6.0). This value corresponds to 5 remaining water molecules, indicating an exchange of 4 water molecules with ligand molecules’ coordination sites. It implies the formation of not only a 1:1 complex but also a certain amount of complexes with higher stoichiometry like 1:2 or 1:3 complexes.
Fig. 3 Emission spectra of 5 μM Am(III) (left) and 10 μM Eu(III) (right) in dependence of lactate concentration (10 μM to 0.1 M each, pH = 6.0, I = 0.1 M, T = 25 °C). |
The quantitative deconvolution of the luminescence spectra in order to determine complex stability constants for all three complexes failed. Possibly the spectral changes are too small to discriminate all three complexes. A reasonable stability constant could only be determined for the 1:1 complex. For determination of the formation constant of the 1:1 complex, the spectra with a lactate concentration up to 0.01 M were considered in analogy to the UV-vis measurements, yielding a log β11 = 2.27 ± 0.05 (25 °C). This is in very good accordance to the value from UV-vis spectroscopy and to literature values (see Table 1).
TRLFS measurements were also done at elevated temperatures (45 °C and 65 °C). The stability constant of the 1:1 complex shows no relevant tendency with rising temperature, indicating that the complex formation reaction causes only a very small enthalpy change. The van't Hoff plot (see Fig. 4) results in an enthalpy change closed to zero within the error bars (see Table 3). Other studies calculated a negative reaction enthalpy, corresponding to an exothermic reaction3,13 (Table 3). This discrepancy is possibly due to different ionic strengths and ionic media. A detailed discussion to this effect is provided in the next section.
Complex stability constants were determined for all three complexes and various temperatures between 25 °C and 70 °C, provided that only the carboxylic group of lactate is deprotonated (see Table 2). The stability constants show no significant trend with rising temperature, however, the van't Hoff plot (Fig. 4) results in a small positive reaction enthalpy change for the 1:1 and 1:2 complexes and a very small negative reaction enthalpy for the 1:3 complex (see Table 3), equivalent to an endothermic reaction for the former and an exothermic reaction for the latter complex. Previous investigations from Tian4 gave small negative reaction enthalpy changes for all three complexes. However, these measurements were done at an ionic strength of 1 M. The reaction enthalpies determined from Aziz7 and Choppin6 at an ionic strength of 2 M are even more negative (see Table 3). This gives the following ionic strength dependency of the reaction enthalpy: the higher the ionic strength the lower the reaction enthalpy. A comparison of literature data in data collections has shown a decrease of reaction enthalpy with increasing ionic strength for several metal–ligand systems; even changes from positive to negative enthalpy can be observed.55 This is caused by the ionic strength dependent variation in the activity coefficients of the reagents.56 In general, two contrary effects contribute to the reaction enthalpy: (1) the partially dehydration of the reactants which is usually endothermic, and (2) the complex formation which is expected to be exothermic.57 At low ionic strength the solvation spheres are tightly bound, equivalent to lowering the activity coefficients of the ions. This results in a larger endothermic dehydration enthalpy and the effect (1) mainly contributes to the reaction enthalpy. With higher ionic strength the solvation spheres loosen up because of a higher amount of competition ions. This causes increased activity coefficients of the ions and a smaller endothermic dehydration enthalpy term. In consequence, the contribution of the effect (2) might become dominant and even cause a change of sign of the reaction enthalpy.58
T/°C | I/M (NaClO4) | Eu(Lac)2+ | Eu(Lac)2+ | Eu(Lac)3 | Ref. |
---|---|---|---|---|---|
log β11(1) | log β12(2) | log β13(3) | Method | ||
log βML(H) are the stability constants with protonated hydroxyl group(s), not considering the pKa2 of lactic acid.a Electrophoresis.b Potentiometry.c TRLFS.d Solvent extraction. p.w. = present work. | |||||
10 | 1.5 | 2.62 | 4.22 | 10 | |
1.0 | 2.90 ± 0.36 | 4.90 ± 0.37 | 6.24 ± 0.30 | 4 | |
1.0 | 2.91 ± 0.24 | 5.02 ± 0.22 | 6.03 ± 0.28 | 4 | |
25 | 2.0 | 2.53 | 4.60 | 5.88 | 9 |
1.0 | 2.80 ± 0.02 | 4.76 ± 0.02 | 6.33 ± 0.02 | 4 | |
1.0 | 2.99 ± 0.17 | 5.09 ± 0.23 | 6.09 ± 0.27 | 4 | |
1.0 | 2.46 ± 0.09 | 4.28 ± 0.25 | 5.87 ± 0.10 | 3 | |
1.0 (NaCl) | 2.95 | 4.40 | 5.47 | 11 | |
0.2 | 2.55 ± 0.05 | 4.67 ± 0.06 | 5.55 ± 0.18 | 8 | |
0.1 | 2.51 ± 0.13 | 4.45 ± 0.12 | 5.83 ± 0.18 | p.w.c | |
0 | 3.14 ± 0.28 | 5.49 ± 0.28 | 7.07 ± 0.31 | p.w. | |
30 | 0.1 | 2.43 ± 0.14 | 4.58 ± 0.16 | 6.07 ± 0.17 | p.w.c |
0 | 3.06 ± 0.29 | 5.63 ± 0.30 | 7.32 ± 0.30 | p.w. | |
37 | 0.1 | 2.59 ± 0.19 | 4.78 ± 0.13 | 6.15 ± 0.16 | p.w.c |
0 | 3.24 ± 0.31 | 5.85 ± 0.28 | 7.42 ± 0.30 | p.w. | |
40 | 1.0 | 2.78 ± 0.15 | 4.57 ± 0.11 | 6.25 ± 0.21 | 4 |
1.0 | 2.91 ± 0.17 | 5.04 ± 0.27 | 6.17 ± 0.25 | 4 | |
45 | 0.1 | 2.77 ± 0.11 | 4.61 ± 0.21 | 6.01 ± 0.14 | p.w.c |
0 | 3.42 ± 0.27 | 5.69 ± 0.33 | 7.30 ± 0.29 | p.w. | |
55 | 1.0 | 2.70 ± 0.11 | 4.43 ± 0.15 | 6.28 ± 0.11 | 4 |
1.0 | 3.04 ± 0.24 | 5.00 ± 0.27 | 5.95 ± 0.26 | 4 | |
0.1 | 2.78 ± 0.14 | 4.21 ± 0.16 | 6.48 ± 0.20 | p.w.c | |
0 | 3.45 ± 0.29 | 5.32 ± 0.30 | 7.79 ± 0.32 | p.w. | |
65 | 0.1 | 2.87 ± 0.19 | 4.76 ± 0.17 | 6.11 ± 0.23 | p.w.c |
0 | 3.55 ± 0.31 | 5.89 ± 0.30 | 7.46 ± 0.34 | p.w. | |
70 | 1.0 | 2.81 ± 0.19 | 4.49 ± 0.21 | 6.33 ± 0.20 | 4 |
1.0 | 2.99 ± 0.23 | 4.88 ± 0.25 | 5.98 ± 0.26 | 4 | |
0.1 | 2.37 ± 0.09 | 4.67 ± 0.11 | 5.58 ± 0.28 | p.w.c | |
0 | 3.06 ± 0.27 | 5.82 ± 0.27 | 6.94 ± 0.38 | p.w. |
I/M (NaClO4) | Am(Lac)2+ | Eu(Lac)2+ | Eu(Lac)2+ | Eu(Lac)3 | |
---|---|---|---|---|---|
a Solvent extraction. b Calorimetry. c TRLFS. d Potentiometry. NaTf = Na-trifluoromethansulfonate; p.w. = present work. | |||||
ΔrH*/kJ mol−1 | 2.0 | −4.3 ± 0.87a | −8.6 ± 1.67a | −23 ± 47a | |
2.0 | −8.17 ± 1.056b | −4.8 ± 2.56b | −23.3 ± 5.96b | ||
1.0 | −163a | −2.14 ± 0.774a | −4.31 ± 0.424a | −12.37 ± 0.674a | |
1.0 (NaTf) | −5.38 ± 0.0713a | ||||
0.1 (p.w.) | 3.7 ± 4.6c | 6.7 ± 6.2c | 3.3 ± 6.6c | −2.1± 6.9c | |
ΔrH0/kJ mol−1 | 0 (p.w.) | 6.2 ± 4.6 | 9.3 ± 6.2 | 8.0 ± 6.6 | 3.1 ± 6.9 |
ΔrS*/J mol−1 K−1 | 2.0 | 33 ± 37a | 58 ± 67a | 34 ± 127a | |
2.0 | 21 ± 36b | 72 ± 96b | 36 ± 206b | ||
1.0 | −63a | 46 ± 34a | 76 ± 24a | 78 ± 294a | |
1.0 (NaTf) | 32 ± 213a | ||||
0.1 (p.w.) | 55 ± 15c | 71 ± 29c | 100 ± 31c | 109 ± 34c | |
ΔrS0/J mol−1 K−1 | 0 (p.w.) | 75 ± 15 | 92 ± 29 | 133 ± 31 | 150 ± 34 |
ΔrG*/kJ mol−1 (25 °C) | 2.0 | −14.15 ± 0.127a | −26.0 ± 0.27a | −33.2 ± 0.47a | |
2.0 | −14.57 ± 0.049d | −26.23 ± 0.129d | −34.06 ± 0.209d | ||
1.0 (NaTf) | −14.813a | ||||
0.1 (p.w.) | −12.5 ± 4.6c | −1.6 ± 6.2c | −29.2 ± 6.6c | −53.36 ± 6.9c | |
ΔrG0/kJ mol−1 | 0 (p.w.) | −16.2 ± 4.6 | −18.2 ± 6.2 | −31.7 ± 6.6 | −41.62 ± 6.9 |
It is worth mentioning that the entropy changes are quite high (see Table 3), so we can assume that the complex formation reaction is predominantly entropy driven. The number of water molecules in the first coordination shell of Eu(III) (deduced from the luminescence lifetimes) helps to explain this. As written earlier, up to 5 water molecules are replaced by 3 ligand molecules which increases the entropy in the system.
The challenge is to receive information about the coordination type of Eu(III) lactate. Possible structures for the 1:1 complex are depicted in Fig. 5. Is it monodentate coordination of the carboxylate group with a high sterical requirement of the lactate (A) or bidentate coordination with carboxylate and hydroxyl group, protonated (B) or deprotonated (C) or bidentate coordination of the carboxylate group (D)? Spectroscopic (FT-IR, NMR) and computational (DFT) techniques were carried out to get an idea about the coordination behavior of the Eu(III) lactate 1:1 complex.
The difference spectrum (Fig. 6c) shows additionally significant changes of the spectral modes at around 1120 cm−1 and 1040 cm−1 (strong negative bands in the difference spectrum). According to DFT calculations (see Fig. 7 and Fig. S7, ESI†), these modes can be assigned to the C–O stretching vibration of the hydroxyl group and the subsequent C–C stretching vibration of the C–CH3 unit, respectively.
Fig. 7 Experimental (ATR FT-IR) and calculated (DFT) spectra of Eu(III) lactate. Calculations were done for structures A, B, and C (from Fig. 5). |
Fig. 7 shows the DFT calculated vibrational spectra for the models A, B, and C from Fig. 5 in comparison with the measured ATR FT-IR spectrum of Eu(III) lactate. The best accordance to the measured spectrum is given by model C. Especially the peak at around 1120 cm−1 (measured spectrum) finds its equivalent only in the calculated IR spectrum of model C (1145 cm−1). In the calculated IR spectra of models A and B, this peak is missing or shifted strongly to 1224 cm−1 (model A) and 1238 cm−1 (model B), respectively. This mode is caused by the C–O stretching vibration of the hydroxo group which is protonated in models A and B. In model C this functionality is deprotonated resulting in a covalent binding to the Eu(III) cation. Due to the position of this stretching vibration mode in the measured spectrum compared to the calculated spectra it is assumed that model C reflects best the binding behavior of the Eu(III) lactate complex. The coordinating hydroxyl group seems to become deprotonated under complex formation with Eu(III).
Fig. 9 13C-NMR spectra of 100 mM lactate, (A) without metal, (B) containing 120 mM La(III), and (C) 5, (D) 10, (E) 50, (F) 100 mM Eu(III). |
Fig. 10 Plot of 13C chemical shift differences vs. Eu(III) concentration. ■CH, ▲COO, ●CH3, lines drawn for better visualization. |
Using La(III) (4f0 configuration, closed shell) as a diamagnetic analogue of Eu(III), the chemical shift changes induced by interaction of the unpaired f-electrons (LIS) can be separated from the pure charge and complex formation induced shifts. As expected, the positive charge of the trivalent metal ion causes a de-shielding, i.e. reduction of electron density of the nuclei at or near the binding site, cf.Fig. 9B. Therefore, the carbon signals are shifted to higher chemical shift values. Interestingly, also in this case the CH carbon is affected most, pointing towards a strong participation of the hydroxyl oxygen in the complex formation. In the case of Eu(III), magnitude and direction of the shift are related to the distribution of f-electron-density at the nuclei of interest, overcompensating the pure charge induced effects.
The LIS has two contributions: (1) contact term, i.e., interaction via bonds and (2) pseudo-contact term, i.e., through space. The contact term depends on type and number of bonds between the (open shell) metal center and the atom of interest. The pseudo-contact term is mediated through dipolar interaction and strongly distance dependent. Both terms can contribute to different extent, depending on, e.g., the electronic configuration and the energy of the ground state or the ligand field splitting.61 Neither for 1H nor for 13C do the shifts of signals of adjacent atoms show alternating signs, indicating that the contact contribution can be neglected.62 Thus, the observed differences in LIS can be fully attributed to spatial europium distances.
Interestingly, the CH carbon atom shows the strongest LIS, indicating that this carbon is affected mostly by the europium's unpaired electrons. This can be explained only by participation of the hydroxyl oxygen in Eu(III) coordination. Therefore, model A (Fig. 5) can be excluded.
Model B (Fig. 5) contains the coordination by the hydroxyl oxygen, but being protonated. Due to this hydrogen, the distance between Eu(III) and this particular oxygen as well as the adjacent carbon is bigger than for both the carboxylic oxygen and carbon (distances calculated with DFT, see Fig. S7, ESI†). This is in contradiction to the 13C-NMR results.
Model C (Fig. 5), however, reflects perfectly the NMR findings: same distance between the carbons of interest and Eu(III) (3.24 Å, calculated with DFT, cf. Fig. S7, ESI†), resulting in similar magnitude of LIS. The small discrepancy in the chemical shift differences of these two particular carbons (cf.Fig. 10) is probably related to the angle between the crystal field axis of the complex and the radius vector from Eu(III) to the respective carbon. The NMR findings strongly support the results obtained from ATR FT-IR measurements in combination with DFT calculations.
With respect to this new findings, the complex formation constants of the Eu(III) lactate 1:1 complex have to be recalculated considering the pKa of the hydroxyl group.30 The resultant constants and thermodynamic data are listed in Table 4. Interestingly, the recalculation process of the complex formation constants shows that only the 1:1 complex seems to exist with deprotonated hydroxyl group. The calculation of stability constants with further fully deprotonated lactate ligands failed. At pH 3, the hydroxyl group even of the first lactate remains protonated.
T/°C | log β*110 | log β°110a |
---|---|---|
a Uncertainty ± 0.25. | ||
25 | 7.52 ± 0.07 | 8.80 |
30 | 7.28 ± 0.07 | 8.57 |
37 | 7.33 ± 0.06 | 8.64 |
45 | 7.66 ± 0.04 | 8.99 |
55 | 7.73 ± 0.05 | 9.08 |
65 | 7.53 ± 0.05 | 8.92 |
70 | 7.23 ± 0.12 | 8.63 |
ΔrH/kJ mol−1 | 7.7 ± 3.0 | 12.9 ± 3.0 |
ΔrS/J mol−1 K−1 | 169 ± 10 | 210 ± 10 |
ΔrG/kJ mol−1 (25 °C) | −42.7 ± 3.0 | −49.7 ± 3.0 |
The results from ATR FT-IR and NMR measurements combined with DFT calculations provided detailed structural information for the Eu(III) lactate 1:1 complex. The finding that the hydroxyl group seems to be deprotonated under complex formation (model C, Fig. 5) contradicts former structure suggestions, which suppose a coordination of the trivalent metal ion with the protonated hydroxyl group (model B, Fig. 5).4,5,12,63 Both experimental methods, ATR FT-IR and NMR, as well as the DFT calculations yielded an impressively homogeneous structural explanation of the investigated Eu(III) lactate 1:1 species.
The thermodynamic results indicate that the complex formation of trivalent actinides and lanthanides with organic matter is strongly influenced by different parameters like temperature and ionic strength. This makes it difficult to simulate and predict the migration behavior of the metal ions in the environment. Insights in the structural behavior of the complexes in aqueous solution (like it is provided with this study) improve understanding and may result in a more reliable prediction of such migration processes.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4dt00440j |
This journal is © The Royal Society of Chemistry 2014 |