Elizabeth C. Lewandowski
,
Colin B. Arban
,
Morgan P. Deal
,
Andrea L. Batchev
and
Matthew J. Allen
*
Department of Chemistry, Wayne State University, 5101 Cass Avenue, Detroit, Michigan 48202, USA. E-mail: mallen@chem.wayne.edu
First published on 4th September 2024
Europium is an f-block metal with two easily accessible oxidation states (+2 and +3) that have vastly different magnetic and optical properties from each other. These properties are tunable using coordination chemistry and are useful in a variety of applications, including magnetic resonance imaging, luminescence, and catalysis. This review describes important aspects of coordination chemistry of Eu from the Allen Research Group and others, how ligand design has tuned the properties of Eu ions, and how those properties are relevant to specific applications. The review begins with an introduction to the coordination chemistry of divalent and trivalent Eu followed by examples of how the coordination chemistry of Eu has made contributions to magnetic resonance imaging, luminescence, catalysis, and separations. The article concludes with a brief outlook on future opportunities in the field.
![]() | ||
Fig. 1 Energy levels for EuIII and EuII. All levels for each ion are shown directly over the ion label. For clarity, subsets of levels are projected to the right of the full diagram and labeled with term symbols. Projected subsets with black lines represent 4f states, and red lines represent 4f5d states. Values for the levels were obtained from previous reports.1,2 |
With respect to divalent europium, the ion largely adopts three coordination numbers: eight, nine, and ten, with eight being the most common. Trivalent europium typically adopts coordination numbers of eight and nine because of the smaller atomic radius of EuIII, 1.09 Å, compared to divalent europium, 1.31 Å.6 With lanthanides, coordination number and ligand bulk tend to drive geometry. Readers are referred elsewhere to learn about the possible geometry of Eu ions with coordination numbers 8–10.7–9 EuII complexes adopt geometries like bicapped trigonal antiprisms, hula hoops, and tetracapped trigonal prisms.10–18 Within the hula hoop and tetracapped trigonal prism geometries, there is often a combination of a multidentate ligand and one or more monodentate coordinated solvent molecules. Overall, the shape of the europium complex depends on the ligand and the space available on the metal coordination sphere.
Beyond coordination number and geometry, hard–soft acid–base properties of ligands are another way to tune the properties of europium. Hard–soft acid–base theory is a way to predict and explain the general types of atoms that bond together. In essence, classifying an atom as either hard or soft is dependent on its polarizability, or its response to an electric field.19 Trivalent europium, and other trivalent lanthanides, tend to be thought of as hard acids, but divalent europium is a much softer ion. Consequently, the types of donor atoms used for the two ions tend to be harder donors like oxygen, used more for the trivalent ion, and softer donors like nitrogen and sulfur, used for the divalent ion. The hard–soft properties of donor atoms affect the stability of Eu-containing complexes. For example, sulfur donor atoms shift the oxidation potentials of EuII-containing complexes more positive.20 The ability to tune the electrochemical properties of Eu-containing complexes is an important aspect of ligand design for applications in MRI and separation science. However, substantial overlap exists in the identities of ligand donor atoms that will bind to divalent and trivalent ions.
Solvent selection is another important factor to consider when studying Eu-containing complexes. As previously mentioned, solvent molecules coordinate with the Eu ion, influencing electrochemical potentials. For example, the formal potential of EuII/III is −0.26 V versus the calomel electrode in acetonitrile, a weakly coordinating solvent, but the formal potential of EuII/III is −0.95 V versus the calomel electrode in strongly coordinating solvents like hexamethylphosphoramide.21 Therefore, when performing studies in solution, selection of solvent is an important consideration.
Finally, when working with complexes of EuII, it is critical to prevent inadvertent oxidation to EuIII by atmospheric O2. This prevention involves standard inert atmosphere methods including glovebox and Schlenk techniques and the use of degassed solvents. For applications relevant to MRI, wet gloveboxes are often used for sample preparation in which water is allowed but not molecular oxygen. For many other applications, dry gloveboxes are often used. When preparing samples for solution-phase characterization, including MRI, NMR spectroscopy, and fluorescence or UV-visible spectroscopy, samples need to be sealed to prevent oxygen from entering the sample. Depending on the longevity of the study the samples can be sealed with tape or wax or placed in flame-sealed tubes. For a thorough review of methods to handle divalent europium, readers are referred elsewhere.22
Studying aqueous coordination chemistry is critically important for applications involving aqueous solutions of europium-containing complexes. An important factor of a contrast agent is that it needs to be water soluble and coordinate with rapidly exchanging water molecules. For divalent europium, short metal–water distances, multiple coordinated water molecules, and optimal water exchange rates increase contrast enhancement. However, slow dissociation rates are also important to avoid europium dissociation from complexes. Ultimately, a balance of properties is required to tune all of these interconnected aspects of coordination chemistry. Ligands described in this section that influence coordination chemistry relevant to MRI are depicted in Fig. 2. This focus is shared by the Allen Research Group in the study of europium complexes.
One useful technique for determining dissociation constants is UV-visible spectroscopy; however, the absorption spectra of EuII-containing complexes and the EuII aqua ions often overlap, limiting the ability to monitor dissociation. Inspired by UV-visible studies of murexide with other metals,37,38 murexide was studied as an indicator to aid in the measurement of dissociation rates of Eu. Using EuII1 and murexide, absorption measurements at 485 nm were recorded as a function of time, and the absorption values were used to determine the concentration of EuII.15 The natural log of the concentration versus time indicated a first order relationship with the slope of the plot revealing the dissociation constant. Results from the absorption spectra analysis were consistent with the results determined from electrochemical techniques. When comparing EuII1 and EuII3, dissociation rates increased upon addition of a benzo group to the complex. Differentiation between functional group additions to benzo group complexes is seen when comparing EuII4 and EuII5. The results show that the addition of a variety of electron-withdrawing groups can be differentiated when using murexide with UV-visible spectroscopy to determine dissociation rates. Overall, this study demonstrated that murexide can be used to measure dissociation rates of EuII-complexes. The murexide indicator adds to the toolbox of methods for determining dissociation rates of EuII-containing complexes. Additionally, the inertness of EuII-containing cryptands was studied with respect to transmetallation by endogenous ions such as CaII, MgII, and ZnII based on previous studies.39 Along these lines, EuII1, EuII3, EuII6, and EuII7 were studied in the presence of CaII, MgII, and ZnII and found that EuII-containing cryptates without amide groups are kinetically stable in the presence of the ions.40
Another important aspect of Eu chemistry relevant to contrast agents is the electrochemical potential, which needs to be more positive than the electrochemical potential of water to avoid reduction of water by EuII. EuII-containing complexes in aqueous solutions were studied to report the complexes that are most oxidatively stable.20 Results demonstrated that ligand modifications to the hard–soft acid–base properties of donor atoms shifted the oxidative potential of EuII. Additionally, the electrochemical behavior of Eu-cyclen-based complexes was studied with various ratios of glycinamide to acetate arms.41 It was found that the substitution of an amide for a carboxylate donor in the coordination environment led to more positive oxidation potentials of EuII. The results from this investigation showed the significance of the ligand design in the tuning of the electrochemical potentials of the Eu-containing complexes that are used for redox-responsive contrast agents.
Understanding the thermodynamics and kinetics of oxidation is important to aid in the rational design of new complexes for the potential use of oxidatively responsive EuII-containing complexes in vivo. Toward this goal, the oxidation of EuII-containing complexes, EuII4, EuII8 and EuCl2 was studied.42 The oxidation of EuII by molecular oxygen was observed in varying pH conditions because more acidic media results in slowed oxidation rate.34 Interestingly, the largest oxidant studied, glutathione disulfide, was unable to oxidize EuII, hinting that it is possible to kinetically control oxidation. This kinetic control would later be demonstrated by the Allen Research Group with ligand selection instead of oxidant selection.43
Other groups have demonstrated that the EuII aqua ion and cryptate have outstanding water-exchange properties for potential use in contrast agents for MRI.34,44,45 One report illustrates the relationship between relaxivity and the molecular weight of EuII-containing cryptates, EuII1, EuII3, and EuII9.46 The relaxivity of EuII1, EuII3, and EuII9 increases as a function of molecular weight at all field strengths, consistent with complexes of GdIII.47 Further, the correlation between slowing molecular tumbling rates and relaxivity through covalent and noncovalent interactions with macromolecules were investigated.15,48 In this study, cryptate EuII9 was expected to form noncovalent inclusion complexes with cyclodextrins and albumin to increase relaxivity by slowing the molecular tumbling rate. It was found that the relaxivity of EuII9 in the presence of the macromolecules increased relaxivity as a function of the molecular weight of the macromolecules; however, the increase in relaxivity was smaller than expected based purely on molecular weight due to internal rotations.
To address the disparity between molecular weight and rotation, a rigid linker was incorporated to bridge two ions of EuII to study the relationship between rotational dynamics and relaxivity.18 Comparison of the rotational dynamics with the relaxivity of the contrast agents was done by comparing EuII10, a rigid complex, with EuII11, a EuII-containing complex that can rotate freely around the single bonds of the linker between two EuII ions. The nuclear magnetic relaxation dispersion data shows local rotation of EuII11 stemming from the rotation of about the flexible linker between the two cryptates. The more rigid linker in EuII10 correlated to longer global rotational correlation time, resulting in an increase in relaxivity. These results demonstrate that rigidity in linking EuII to other molecules is beneficial for increasing the relaxivity of redox-active, EuII-based contrast agents.
The knowledge of EuII relaxivity and oxidation kinetics was integrated with coordination chemistry to enable screening of new ligands.49 The rational design of ligands to complex both oxidation states of Eu is important to prevent dissociation upon oxidation of the metal; this design requires a method to detect ligands that can bind both the soft, large EuII ion, as well as the harder and smaller EuIII ion. The method involves complexation of ligands to both EuII and GdIII, in which GdIII is used as an MRI-active surrogate for EuIII. In the method, metals are mixed with ligands, and uncomplexed metal is precipitated with phosphate buffer. The resulting mixtures are filtered into a multiwell plate that is imaged using MRI, with bright spots in the image indicating that a contrast-enhancing complex was formed. Additionally, for EuII-containing complexes, a bright spot indicates that a complex was formed that does not reduce water. Dark spots in the images indicate lack of coordination with the ligand or the formation of a EuII-containing complex that reduced water to form EuIII. To test the screening procedure, a set of eight ligands, 1–3, 8, 12–15, were selected due to the range of ability of the ligands to coordinate EuII and GdIII. The screening procedure was able to provide binding information (Fig. 3),49 and was subsequentially used in future studies.43
![]() | ||
Fig. 3 T1-weighted MR image of a multiwell plate of solutions of EuII1–3, EuII8, EuII12–15 and GdIII1–3, GdIII8, GdIII12–15 at ambient temperature. Label a represents phosphate-buffered saline, b denotes GdCl3, and c represents EuCl2. Adapted from Corbin et al.,49 copyright 2018, with permission from Elsevier. |
The first example of persistence of EuII in oxygenated solution stemmed from the study of 21 ligands with varying denticities, sizes, donor atom identities, and cavity sizes that were screened for ability to coordinate both EuII and EuIII.43 From this screening process, a tetraamide phosphonate-containing ligand, 16, was identified as being able to coordinate to both EuII and EuIII,43 and was further investigated. Interestingly, previous studies with 16 demonstrated pH-dependent relaxivity with GdIII.51–54 The pH-dependency was of interest to study based on other EuII studies of tetraamine ligands showing temperature-dependent cage formation, contrast enhancement, and slowed oxidation of EuII.55 If similar cages are formed with the phosphonate groups to trap coordinated water molecules, those cages could sterically interfere with the approach of oxygen to slow oxidation, similar to how glutathione could not oxidize EuII8 despite being thermodynamically possible.42 Studies of EuII16 were performed to understand if the phosphonate groups could slow oxidation of EuII as a function of pH.43 UV-visible measurements were performed from pH 7 to 10 to understand EuII persistence within EuII16. As pH increased, the persistence of EuII increased, with EuII half-lives of 6.4 and 12.4 minutes at pH 7 and 10, respectively. Those pH values surround one of the pKa values of phosphonates.56,57 Consequently, it was hypothesized that the longer persistence of EuII in basic solution is due to a cage-like formation of 16, trapping an innersphere molecule of water as a function of the protonation state of the phosphonates. Because of the oxygen resistance observed in the UV-visible measurements, EuII16 was used for systemic delivery studies in healthy mice that were monitored using dynamic contrast-enhanced MRI.43 EuII16 was injected into the tail vein of mice, and contrast-enhanced signal was monitored over time using MRI. The half-life of the contrast enhanced signal is roughly seven minutes, which is consistent with in vitro studies at neutral pH. Overall, this study introduces a new EuII-containing complex that showed persistence in oxygenated solutions during in vitro and in vivo measurements, which is an important step toward systemic delivery of EuII contrast agents for detection of hypoxia.
In addition to increasing persistence using coordination chemistry of phosphonate-containing ligands, an outersphere approach toward increasing persistence was pursued using highly fluorinated ligands in perfluorocarbon nanoemulsions.58 Within this study, EuII was complexed to a perfluorinated, cyclen-based ligand, 17, and subsequently dispersed in an N2-saturated perfluorocarbon/lecithin nanoemulsion to slow oxygen diffusion toward EuII (Fig. 5). The multiple interfaces between water and lecithin and lecithin and N2 surrounding perfluorocarbons were hypothesized to slow oxygen diffusion to EuII based on reports of gas diffusion across interfaces.59,60 Additionally, saturation of the perfluorocarbon emulsion with N2 was also hypothesized to slow diffusion of oxygen in the vicinity of Eu. The perfluorinated ligand and nanoemulsion are useful for detection of EuII concentration via 19F-MRI measurements because 19F-signal increases as EuII concentration decreases due to EuII line broadening effects.58 To understand the relationship between 19F-signal and EuII, 19F-NMR measurements of EuII17 and EuIII17 in equimolar concentrations were acquired in degassed and oxygenated perfluorocarbon mixtures. These studies revealed that the oxidation state of EuII is responsible for almost all the signal change in 19F-NMR measurements. Contrast enhanced signal of solutions of EuII17 within a perfluorinated nanoemulsion and a solution of EuII15 were measured within thigh muscles of healthy mice using MRI measurements. The perfluorinated nanoemulsions of EuII17 revealed persistence of EuII in detectable amounts for at least 30 minutes (compared to <5 minutes for a control complex), indicating that the nanoemulsion is able to slow the diffusion of oxygen to EuII. This study represents an important step toward increasing the persistence of EuII signal through an outersphere approach. Overall, these studies show how the surrounding environment of EuII can influence EuII persistence in oxygenated solutions, which is an important step toward hypoxia detection using MRI via systemic delivery of contrast agents.
![]() | ||
Fig. 5 Cartoon representation of EuII17 in lecithin/perfluorocarbon emulsion. The orange represents a perfluorocarbon mixture surrounding EuIII17, the grey circle, and EuII17, the red circle. |
One study involved encapsulation of a EuII-containing complex within liposomes to create a dual-mode contrast agent that is oxidation-responsive, and upon oxidation of EuII to EuIII would result in a probe for chemical exchange saturation transfer (CEST) MRI.62 CEST is a method of using exchangeable protons to create images using MRI.63,64 Liposomes were selected because the inner cavity can hold water-soluble contrast agents, increasing the ratio of water protons associated with the liposome to bulk water protons. The designed liposome system uses T1 enhancement that visualizes the presence of the EuII oxidation state along with CEST that visualizes both EuII and EuIII.61 From observation of the potential outcomes from the CEST and T1 measurements, the oxidation state of the Eu ion can be determined without knowledge of the concentration of Eu. The liposome encapsulated EuII-containing complex enabled the observation of the first oxidation-responsive dual-mode contrast agent designed around the oxidation state of Eu. Moving from liposomes to small molecules as CEST probes, the use of EuII/III-cyclen based derivatives as oxygen-sensitive MRI contrast agents was explored.65 In this study, EuII8 provides T1-weighted contrast enhancement but no CEST signal, and EuIII8 produces CEST signal but no T1-weighted signal enhancement (Fig. 6). The small molecule probes were characterized by a variety of methods to characterize both oxidation states of the complex. Ultimately, the ability to image before and after oxidation of EuII was demonstrated.
Another system studied as potential multimodal contrast agent integrated MRI and photoacoustic imaging techniques.66 In this system, MRI has the ability to scan large areas and photoacoustic imaging has the ability to create images of areas that are accessible by light. To determine the potential of having a redox-responsive multimodal contrast agent for both MRI and photoacoustic imaging, Eu8 was selected based on previous studies demonstrating that the divalent complex is yellow in color, to enable absorption for photoacoustic imaging, and that it is an effective contrast agent for MRI.65 Eu8 was effectively able to function as a multimodal contrast agent; however, because both modalities could only be detected with EuII and not EuIII, the system was not ideal for overcoming the challenge of detecting EuIII.
In addition to CEST MRI, the use of 19F signals can be combined with 1H MRI in multimodal probes, because 19F MRI has been used in vivo in various applications.67–70 A major concern of using 19F signal is finding the amount of fluorine that enables solubility of EuII-containing complexes in water and results in detectable signal. One study aimed to address this concern by studying changes based on the number and placement of fluorine atoms in a set of ligands.55,71 One of the ligands, 15, contains four p-trifluoromethylbenzyl groups that when complexed with Eu form a pocket where a coordinated water molecule is caged in the solid state.55 The trapping of the water molecule is unique compared to other tetraamide complexes that have arms that point away from each other instead of toward each other.72–74 Oxidation of EuII to EuIII enables the presence of an observable 19F signal and the detection of the EuIII-containing complex.55 The ratio of 19F and 1H signals could also be potentially used for ratiometric imaging beyond just observing EuIII after oxidation of EuII.71 Ratiometric imaging is important because of its ability to lead to quantification of medically relevant biomarkers like oxygen concentration in hypoxic environments. Quantification of hypoxia is difficult because the MRI signal produced using EuII-based contrast agents is dependent on the concentrations of oxygen and EuII. Based on previous studies involving multimodal imaging and the relationship of 19F signal and EuII signal, levels of hypoxia can be determined ratiometrically using a fluorinated EuII-based contrast agent.71 and comparing the 1H and 19F-MRI signals produced from a dual-mode redox responsive contrast agent.64 In this study, hypoxia was detected by plotting the T1-weighted 1H-MRI signal with 19F-MRI signal. These results enable quantification without the knowledge of the concentration of Eu.
![]() | ||
Fig. 8 Calculated d-orbital splitting of EuII1 and EuII12. Adapted from Corbin et al.,81 copyright 2018, with permission from Elsevier. |
Because of the seemingly important role of the chloride ligand in the luminescence of EuII12, one subsequent study of EuII12 investigated how the identity of counterions influences excitation and emission peaks.10 Ligand 12 was used to occupy eight coordination sites of EuII so only one coordination site was left for the following ions: Cl−, Br−, I−, and PF6−. Initial luminescence measurements were performed in methanol and showed similar absorbance and emission values for each counterion in solution. Acetonitrile was also used because it is a weaker ligand than methanol, decreasing competition between solvent and counterion for binding to EuII.84 In acetonitrile, a blue shift in emission spectra is observed with Cl− and Br− counterions in EuII12 relative to the same complexes in methanol, whereas I− and PF6− counterions do not result in changes in emission from methanolic solutions.10 The blue shift arising from Cl− and Br− is attributed to those ions binding more preferentially to EuII than acetonitrile. The I− and PF6− ions are less likely to coordinate,84 in which case acetonitrile coordinates to EuII12 to complete a nine-coordinate geometry or no monodentate ligand coordinates, resulting in an eight-coordinate geometry. This study demonstrates the importance of counterions and solvent selection for EuII luminescence measurements.
Another avenue of study of the luminescence of EuII explored the effect of changing the secondary amines in 12 to tertiary amines.11 In this study, the secondary amines of 12 were methylated to yield 18 that has increased steric bulk and slightly stronger donors compared to 12. The initial expectation with 18 was that the tertiary amine donors would split the d-orbitals of the EuII complex more than the secondary amines within 12, resulting in more red-shifted emissions with EuII18 than with EuII12. Additionally, the lack of NH oscillators on the ligand and steric bulk was expected to increase luminescence quantum yield by minimizing pathways for nonradiative quenching. The crystal structure of EuII18 revealed that the methyl groups blocked all inner sphere coordination sites for monodentate ligands, resulting in EuII adopting an eight-coordinate geometry in the ligand. Upon studying the luminescence of EuII18, emission wavelengths are more shifted toward the UV region compared to the emission wavelengths of EuII12. A computational study suggested that the change in geometry was the dominant factor that regulated the emission wavelength. Despite the unexpected luminescence wavelength, the quantum yield of the EuII18 was 47%. This quantum yield was the largest reported for EuII in aqueous solution at the time and is due to the methyl groups on the tertiary amines shielding EuII from luminescence quenchers in solution. This study highlights the importance of geometry and the effect of quenchers on the luminescence of EuII.
In an attempt to understand what components of 12 were needed to control the luminescence of EuII, several ligands of smaller denticity that represent parts of 12 were studied with EuII.85 Within this study, absorbance measurements were compared of EuII in the presence of nineteen ligands of varying degree of functionalization, denticity, and structure. When comparing degree of functionalization of amines between ligands 19–25, the secondary amines on linear and macrocyclic ligands bathochromically shift the absorbance of EuII more than primary and tertiary amines. The absorbance of EuII did not shift greatly when comparing linear and macrocyclic ligands 20 and 23–25, except for an additional shoulder reaching into the blue-light region for 23. Finally, increasing denticity of both macrocyclic and linear ligands further blue-shifted absorbance spectra of EuII, likely due to the chelating effect. Overall, secondary amines, macrocyclic ligands, and increasing denticity shifted the emission maxima from the UV toward the blue region. This study compared ligand characteristics necessary to understand and control the luminescence of EuII, and it is an important contribution toward rational ligand design.
Outside of the Allen Research Group, others have investigated how varying donor atoms within ligands affect emission maxima, luminescence lifetime, and redox potential of EuII. Liu, Bian and coworkers, compared 1 against ligands with fewer oxygen donors, resulting in ligands 26 and 27.86 The photophysical properties in the solid state and in solution were studied of each EuII complex. Absorption and emission maxima of EuII-complexes in methanol shifted toward the blue-light region when ligands had fewer oxygen donors and more nitrogen donors. Specifically for emission maxima, there was a shift of 19 nm toward the blue region when the number of N donors was increased, consistent with a greater d-orbital splitting of EuII. Luminescence lifetime of EuII also increased with more N donors than O donors when in methanol. Luminescence lifetime also increased with more N donors by ∼175 ns because as the number of N donors increase, the excited state of EuII is stabilized. Conversely, emission maxima and luminescence lifetime of EuII with ligands containing more O donors decreased. Overall, an increasing, linear trend was found between the number of O and N donors and the emission maxima and luminescence lifetime of EuII complexes. Within this study, the different effects of N and O donors for EuII luminescence were compared to understand the tunability of EuII properties that are important for light-emitting diodes.
Following studies with 12 and 18, further studies to tune EuII electronic properties were undertaken using another ligand, tris[2-(2-methoxyethoxy)ethyl]amine, 28.87 Ligand 28 is structurally similar to 1, but lacks one tertiary amine, making it acyclic and more flexible than 1. EuII28 was of interest to investigate how flexible acyclic ligands influence the properties of EuII with respect to coordination number, geometry, bond length, and luminescence spectra. Within this study, 1 and 28 were complexed with EuII and studied initially in the solid state. Complexes of EuII1 contain EuII with coordination numbers of ten and square antiprism geometries, and complexes of EuII28 contain EuII with coordination numbers of nine with muffin geometries.75,87 The solid-state structures show slight differences in bond lengths and bond angles of EuII1 and EuII28, which were likely due to differences in counterion and solvent molecule binding between each complex. To further understand coordination environment with different counterions, absorption and luminescence spectra of each complex in solution were studied (Fig. 9). The UV-visible spectra for EuII1 and EuII28 show almost no dependence on counterion identity for most complexes of EuII1 and EuII28, with absorbance peaks centered around 320 nm.87 For complexes of EuII1, the emission and excitation spectra did not vary greatly between different coordinating counterions, indicating that different counterions do not influence the luminescence of EuII1 significantly in acetonitrile. Despite the lack of shifts seen in excitation and emission spectra for EuII1, complexes of EuII28 with different counterions have a wide variety of emission and excitation wavelengths, spanning an emission maxima range of 355 to 525 nm. The wide range of wavelengths recorded implies that the flexibility of 28 enables innersphere interactions of counterions with EuII, splits d-orbitals, and influences geometry in solution. This study highlights the difference between flexible, linear ligands, and macrocyclic ligands, and how that difference alters the properties of EuII. The ligands studied for luminescence with EuII can also be used with other lanthanides to elicit similar trends in luminescence properties.88,89 Further, the luminescence studies of EuII have influenced ligand design for applications in catalysis, lighting, and imaging.
To further understand the proposed transition state of EuIII-containing complexes and the aldehyde substrates, a variety of aldehyde and silyl enol ether substrates were used in different reactions to monitor enantioselectivity.97 The yields of the various substrates using EuIII-containing complexes as catalysts had some of the largest stereoselective yields reported in aqueous solution. Further support for the proposed transition state was obtained by demonstrating that as bulkiness of the aldehyde substituent increased, the yield decreased. One study focused on ligand design for EuIII-containing precatalysts, specifically how changing functional groups from esters to alcohols, carboxylic acids, and amides influences catalytic yields.100 From these studies, catalytic yields varied between 6 and 20% for ligands with alcohols, carboxylic acids, or amides due to the lack of steric bulk compared to previously studied ester groups. Overall, the results of these studies demonstrate how adapting ligand motifs from MRI enable the generation of highly enantioselective catalysts in aqueous conditions, where lanthanide complexes tend to dissociate, resulting in outstanding enantioselectivities of the aqueous Mukaiyama aldol reaction. These studies further demonstrate the utility of luminescence-decay measurements in elucidating catalytic mechanisms of EuIII-containing complexes.
Photoredox catalysis uses a redox-active catalyst to transfer an electron to a substrate to form products during organic reactions, typically using transition metals as catalysts.101,102 Previous luminescence studies of EuII12 showed that the excitation and emission maxima were blue shifted compared to EuII1, with luminescence lifetimes of 0.98 μs.75 Both of these qualities are desirable in photoredox catalysts, which led to studies using EuII12 as a potential photoredox catalyst. First, the excited-state potential was calculated using cyclic voltammetry. The excited state potential of EuII12 is −3.0 V vs. Ag/AgCl,82 which is more reducing than SmI2, a commonly used metal for photoredox catalysis.103–105 The reactivity of EuII12 was tested using reductive coupling reactions of benzyl chloride to form 1,2-diphenylethane.83 An in situ mixture of one equivalent of EuII, 12, and benzyl chloride were reacted in the presence of 460 nm blue light in methanol, which gave a 85% yield of 1,2-diphenylethane and 4.7% of toluene in 30 minutes. Control reactions revealed that no product formed when blue light, EuII, or 12, were omitted, indicating all three components are needed to form product. To investigate why there were differences in quenching and product yields for the substrates, cyclic voltammograms were acquired of each substrate to find the cathodic peak to compare to the excited-state potential. The cathodic peak, indicating reduction, of (CH3)3CCl, C6H5Cl, CH2CHCH2Cl, and C6H5CH2Cl were −3.05, −2.93, −2.35, and −2.34 V, respectively vs. silver/silver chloride. Once the electron transfer mechanism was studied, catalytic reactions of Eu and 12 were investigated using benzyl chloride reductive coupling reactions. Within these reactions, EuCl3, 12, and Zn0 were used as catalytic precursors, with Zn as a sacrificial reducing agent. UV-visible absorbance and fluorescence measurements of EuCl3, 12, and Zn0 show that Zn0 reduces EuIII to EuII but is not able to perform reductive coupling. Benzyl chloride reactions at 10 mol % of EuCl3 and 12 in blue light over six hours gave 80% yield of 1,2-diphenylethane and 11% of toluene. Catalyst loading was also studied by varying the amount of EuIII and 12. At 5, 1, and 0.5% loading of EuIII and 12, the yield of 1,2-diphenylethane was 71, 70, and 60%, respectively, with increasing catalytic turnovers at lower catalyst loading. As the catalytic loading was decreased, toluene yield increased. All reactions in this study were performed under anhydrous conditions, so to understand how small concentrations of water affected these reactions, EuCl3·6H2O was used in a wet glovebox, allowing water but no molecular oxygen within the atmosphere. The yield for the wet reaction was 80%, indicating that water does not affect the precatalyst. The proposed catalytic mechanism for this reaction is EuII12 is excited by blue light to form EuII12*. EuII12* transfers an electron to benzyl chloride that ultimately leads to a reductively coupled product and oxidation of EuII to EuIII and 12. Zn0 reduces EuIII to restart the cycle (Fig. 12).77 This study showed the first visible-light promoted EuII-based photoredox catalysis and was an important step in using Eu-containing complex for photoredox catalysis.
![]() | ||
Fig. 12 Proposed photoredox catalytic mechanism of EuII12 for reductive coupling reactions of benzyl chloride. |
Beyond ligand 12, previously studied complexes of EuII that shifted the absorbance spectrum of EuII into the visible light region from the UV based on various ligand structures were studied to understand their ability as photoredox catalysts.85 To understand the excited-state potentials, cyclic voltammograms of EuII19, EuII21, EuII22, and EuII23 were obtained. The excited-state potentials of the complexes were between −2.3 and −3.2 V versus natural hydrogen electrode. More negative excited-state potentials are useful because they enable the reduction of a wide range of substrates. To study the catalysis-promoting ability of each complex, benzyl chloride coupling reactions were performed in blue light with 10:
1 ligand-to-metal ratios to enable complex formation in solution. At large metal-to-ligand ratios, EuII22 produced low yields and precipitated. The primary amine complex EuII21 had the largest yields compared to secondary and tertiary amine-containing ligands. For EuII21, moderate and large ligand-to-metal ratios produced good yields, but large ligand-to-metal ratios produced multiple alkylated byproducts. Comparisons of yields of EuII12 and EuII21 show that cryptates like EuII12 are less likely to form ZnII-containing complexes due to the macrocyclic effect that improves overall product yield by increasing catalytic turnover. This study demonstrated the importance of ligand design for photoredox catalysts and how ligand structures and coordination chemistry can influence catalysis.
EuII-based photoredox catalysis was reported by Borbas and coworkers using chromophore-based ligands.106 They investigated how chromophores excited by blue light transfer energy to reduce EuIII to EuII, to perform a variety of reductive couplings and functional group transformations. A series of three ligands with varying denticity and donor atoms were studied, with 35 and 36 having 6,7-oxycoumarin as a chromophore, and 37 having 7-aminocarbostyril as a chromophore. The excited-state potentials of these chromophores are more negative than −1.88 V vs. ferrocene/ferrocenium. These potentials are negative enough to reduce EuIII to EuII. To test the reactivity of EuIII35–EuIII37, several reductive coupling reactions with benzyl halide-based substrates and reduction reactions for sulfoxide, diazo, imine, azo, alkyl, phosphonate, aldehyde, ketone, and nitrile groups were performed. In most of these, EuIII35–EuIII37 was successfully reduced to EuII35–EuII37 resulting in excellent yields when in the presence of blue light and a sacrificial reductant (Zn0 or diisopropylethylamine). Control experiments showed that no product formed without EuIII, 35–37, or blue light. The proposed mechanism for this reaction is a photoinduced electron transfer from the chromophore to EuIII results in the formation of EuII that transfers an electron to a substrate (Fig. 13). This mechanism was confirmed using electron paramagnetic spectroscopy. The results of these studies showed that Eu-based photoredox catalysts using a chromophore was capable of multiple reductive coupling and reduction reactions through photoinduced electron transfer.106
Solid–liquid extraction is another method that uses coordination chemistry to separate rare-earth elements from each other.114 Solid–liquid extraction involves a solid support, which is often chemically modified with ligands, that collects metals from solution. After a metal ion of interest coordinates with the ligand on the support, and other metal ions are removed by washing, the metal ion of interest is removed from the ligand by washing with a different solvent. Solid–liquid extraction is often tuned through the ligand attached at the solid support, because ligands can bind rare-earth elements with a range of affinities. The interaction between the solid support and the ligand can also be tuned through covalent or noncovalent interactions. Solid–liquid extraction has gained interest because of its specificity and the small amount of waste it produces, and will be further discussed within this section.
Some methods of separation specific to the isolation of Eu from other elements involve reducing EuIII to EuII,115–122 but those methods often require many steps to obtain high purity Eu. One collaboration with the Dittrich Research Group involved the hypothesis that 1 covalently bound to a solid support to form 38 would enable separation of Eu from Gd.123 This hypothesis is based on these two elements having similar coordination chemistries when they are both in the +3 oxidation state but having large differences in coordination chemistry when Eu is in the +2 oxidation state. The change in coordination chemistry between EuII and GdIII arise from differences in atomic radii and charge density. Further inspiration for this hypothesis stems from reports that show 1 selectively binds divalent lanthanides over trivalent lanthanides.36 The atomic radius of EuII enables the ion to fit within 1 and coordinate more effectively with faster rates of association and slower rates of dissociation than EuIII.36 The resulting thermodynamic and kinetic differences between EuII and trivalent lanthanides with 1 increase the efficiency of separations.123 To separate Eu and Gd using solid-supported cryptands, a mixture of EuII and GdIII was passed through solid NovaPEG resin that was covalently functionalized with 1 to form 38 (Fig. 15) which had a greater affinity for EuII than GdIII. After GdIII was eluted, the EuII-loaded support was exposed to air to oxidize EuII to EuIII, causing the atomic radius to change, releasing it from the solid support. The purity of the recovered EuII ranged from 86.7 to 99.3% after a single pass through the solid support. Solid support 38 is most effective between pH 3 and 5.5. Overall, the cryptand-modified solid support enabled isolation of high-purity Eu from a mixture of Eu and Gd.
Following studies using 38 for solid–liquid separations of EuII, another study reported noncovalently solid-associated diesters of diethylaminetriaminepenta-acetic acid, 39, to separate rare-earth elements based on the affinity of 39 for the heavy rare-earth elements at an acidic pH.124 Ligand 39 was selected for this separation because it binds strongly to lanthanides and is reusable. Ligand 13 was functionalized with ethylhexylamine to form 39 that differs from 13 by the incorporation of two hydrophobic amides. To test the separation efficiency of 39, a solution containing 16 rare-earth elements (only Pm is excluded), 5 ppm each, was prepared. The results of the separation showed that 39 elutes the elements in the same order as the affinity 13 for the elements (Fig. 16). Solid support 39 is 130 times more effective at separating rare-earth elements at pH 3.3 than unmodified media.
![]() | ||
Fig. 16 Metal sorbed onto 39 as plotted as a function of complexation constants of 13 for rare-earth elements at pH 3.3. Reprinted with permissions from Hovey et al.,124 copyright 2021, with permission from Elsevier. |
Solid support 39 shows a tendency to bind to heavy rare-earth elements during separations using coal-fly ash that contains many different elements. Unlike other systems, 39 has shown the ability to selectively bind heavy rare-earth elements over Ce and La (Fig. 17), two common elements that can make separations more challenging. This extraction of heavy rare-earth elements using 39 is achieved without the use of organic solvents during the separation, an issue with modern separation methods. This system has demonstrated effectiveness at recovering rare-earth elements from waste products with a preference for the rare-earth elements over Fe and Al, despite the concentration of metals. Further studies of this system showed improved enrichment of rare-earth elements in coal fly-ash from 0.024 to 10 wt%.125 The ligand involved in the system was also used to electrochemically remove Gd from samples with potential applications in hospital effluents.126
![]() | ||
Fig. 17 Sorption at pH 0.9 and 3.3 using 39. Reprinted with permissions from Hovey et al.,124 copyright 2021, with permission from Elsevier. |
Another system designed to separate Eu from Yb uses inorganic tetrathiotungstate (WS4−) anions to induce electron transfers to the metal upon heating or exposure to light.127 Within this study, WS4− reduces EuIII to EuII. When EuIII is mixed with WS4− in acetonitrile under ambient light, a EuII-containing precipitate forms. The proposed mechanism of this precipitation is that EuIII binds to WS4−, and once bound, light exposure induces an electron transfer from WS4− to reduce EuIII to EuII, with the resulting EuII-containing complex precipitating. The WS4− separation technique was used on a complex phosphor mixture, resulting in 98.9% separation efficiency of Eu from the mixture. This study is an important step toward separation of rare-earth metals, showing the ability to separate metals from complex lanthanide mixtures like phosphors.127
Overall, solid support systems designed using the principles of coordination chemistry can selectively separate europium from other lanthanides using differences in oxidation state and pH, and these techniques can be extended to other lanthanides.125,128
This journal is © The Royal Society of Chemistry 2024 |