On the Origin of Covalent Bonding in Heavy ... - ACS Publications


On the Origin of Covalent Bonding in Heavy...

14 downloads 152 Views 2MB Size

Article pubs.acs.org/JACS

On the Origin of Covalent Bonding in Heavy Actinides Morgan P. Kelley,† Jing Su,† Matthew Urban,‡ Morgan Luckey,‡ Enrique R. Batista,† Ping Yang,*,† and Jenifer C. Shafer*,‡ †

Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, United States Department of Chemistry, Colorado School of Mines, Golden, Colorado 80401, United States



S Supporting Information *

ABSTRACT: Recent reports have suggested the late actinides participate in more covalent interactions than the earlier actinides, yet the origin of this shift in chemistry is not understood. This report considers the chemistry of actinide dipicolinate complexes to identify why covalent interactions become more prominent for heavy actinides. A modest increase in measured actinide:dipicolinate stability constants is coincident with a significant increase in An 5f energy degeneracy with the dipicolinate molecular orbitals for Bk and Cf relative to Am and Cm. While the interactions in the actinide−dipicolinate complex are largely ionic, the decrease in 5f orbital energy across the series manifests in orbital-mixing and, hence, covalency driven by energy degeneracy. This observation suggests the origin of covalency in heavy actinide interactions stems from the degeneracy of 5f orbitals with ligand molecular orbitals rather than spatial orbital overlap. These findings suggest that the limiting radial extension of the 5f orbitals later in the actinide series could make the heavy actinides ideal elements to probe and tune effects of energy degeneracy driven covalency.



INTRODUCTION

Recent reports have noted the ability for californium borates and dipicolinates to participate in more significant covalent interactions than comparable systems of the earlier actinides, plutonium, americium, and curium.13−17 A follow-on study examining berkelium borates and dipicolinates noted berkelium seemed to be a unique transition point on the actinide series, where more covalent interactions were observed relative to the earlier actinides, but not as pronounced as those observed with californium.17 To date, the origins of these chemical patterns across the series remain unresolved. Unraveling origins of covalency in heavy actinide interactions may require considering if covalency arises predominately from orbital energy degeneracy or spatial orbital overlap.4,5,9,18 Orbital overlap covalency would be more favorable earlier in the actinide series when the 6d and 5f orbitals are less core-like. Energy degeneracy driven covalency has been reported since the 1960s and has recently been observed in actinide chemistry for the 5f electrons.8,19,20 The decrease in energy of the 5f orbitals across the actinide series could lead to degeneracy of the actinide 5f orbitals with the molecular orbitals of a given complexant.5 Due to the steady ingrowth of energy degeneracy driven covalency, decreased favorability of orbital overlap covalency, and otherwise consistent chemistry across these elements, the mid to late actinide elements are the ideal candidates to contrast the roles of orbital overlap versus energy degeneracy driven covalency. By comparing complexation

The actinide series has been both symbolic of discovery and destruction since the late 1930s. A visual “foundation” of the periodic table, actinides served as the initial building blocks for the discovery of new, “man-made” elements, but their accessibility has frequently limited their investigation. For the heaviest transuranic elements (americium through lawrencium) the increased infrastructure, radiological hazard, and relative chemical similarity to the heavy lanthanides has further impeded investigations regarding their chemistry. This report highlights chemistry unique to the heavy actinides that could drive fundamental interest for further investigations. The chemistry of the early actinide series has, to a general extent, been well described by the actinide hypothesis originally developed by Glenn Seaborg in the 1950s.1 Per Seaborg’s hypothesis, the early actinides (actinium through plutonium) mimic certain aspects of transition metal chemistry regarding oxidation state variability and the ability to interact covalently. When traversing from the light to heavy actinides, the contraction of the 5f orbitals stabilizes the trivalent oxidation state, and to date, stabilization of the trivalent oxidation state has been thought to limit covalent interactions for actinides heavier than plutonium. Synthetic, spectroscopic, and computational studies of the actinides since Seaborg’s time have better characterized orbital participation (5f versus 6d versus 7s) in early actinide binding and quantified the amount of covalency in a given actinide−ligand interaction for many systems,2−11 but the limited studies on heavy actinides have continued to suggest lanthanide-like behavior.12 © 2017 American Chemical Society

Received: April 10, 2017 Published: June 28, 2017 9901

DOI: 10.1021/jacs.7b03251 J. Am. Chem. Soc. 2017, 139, 9901−9908

Article

Journal of the American Chemical Society thermodynamics with covalency trends, the role of covalency in heavy actinide interactions can be more rigorously defined. Only limited reports exist regarding the solution complexation thermodynamics of actinides heavier than curium, and the heaviest actinide with experimentally determined solution phase stability constants is nobelium.21 Most investigations regarding heavy actinide complexation thermodynamics have assessed stability constants but not the entropic and enthalpic contributions giving rise to the actinide−ligand interaction.12 Until our recent publication, no enthalpic or entropic complexation information had been reported for berkelium with any ligand.17 Consequentially, the role of covalency in heavy actinide complexation thermodynamics has remained an open question. For the first time, heavy actinide complexation thermodynamic measurements are coupled with electronic structure calculations to elucidate the role and origin of covalency in heavy actinide complexation by examining Am, Cm, Bk, and Cf dipicolinate (DPA) complexes. A discussion regarding the hydration and coordination numbers of heavy actinides is also provided, as these parameters were necessary to reconcile computational and experimental thermodynamic investigations. A modest increase in the 1:2 and 1:3 actinide:dipicolinate stability constants values are observed for Bk and Cf relative to Cm and Am. This jump in binding strength late in the actinide series is coincident with increased 5f actinide energy degeneracy with the dipicolinate molecular orbital and suggests the origin of covalency in heavy actinide interactions stems from 5f energy degeneracy with ligand molecular orbitals. Therefore, due to the increasing localization the 5f orbitals later in the series, the heavy actinides (Bk through No) are in a unique position in the periodic table to probe the effects of energy degeneracy driven covalency.

Figure 1. Average ion−water and ion−DPA distances for the aqueous, 1:1, 1:2, and 1:3 An3+:DPA2− complexes. Black and gray points indicate CN = 9 and CN = 8 structures, respectively. Given distances are An3+−O (water) (◆), An3+−N (■), and An3+−O (DPA) (●). Open points represent experimental data; EXAFS for the aqueous complexes22−25 (◊), and the crystal structure of the An(HDPA)3 complex for the 1:3 structure17 (□ for An3+−N, ○ for An3+−O). Distances are given in angstroms.

evidence has suggested these methods may be more suitable for actinide systems.27 All three methods show the Bk(H2O)8· (H2O)3+ structure as favorable at the same order of energy difference (Supporting Information, Tables 4 and 5). The SOCASPT2 calculations show the single-determinant method is suitable for these systems and the spin−orbit coupling correction is negligible; this is also the case in spin−orbit PBE calculations, which show the same energy trends as the scalar PBE calculations presented here (Supporting Information, Table 6). Observations from this study are consistent with previous abinitio molecular dynamics simulations of the solvated Cm3+ ion that showed thermal effects in solution overcoming the small energy differences of the optimized solvated ions.28 Electronic structure calculations on these systems using solvation corrections have determined coordination numbers of either eight or nine, depending on the method used.29,30 Those studies, however, compared the energies of the An3+(H2O)8(H2O) and An3+(H2O)9 structures and did not include a standard state correction.31 Experimental evidence, including crystal structures, high-energy X-ray scattering, and extended Xray absorption fine structure measurements on ions in aqueous solutions, likewise suggests these ions exist in an equilibrium between eight and nine coordinate in aqueous solution.22−25,32,33 The literature presented to date suggests some uncertainty still exists regarding the hydration number of Cm, Bk, and Cf actinides and both eight- and nine-coordinate species should be accounted for during data interpretation. While the An3+ ion likely starts as nine coordinate, computational Gibbs free energy trends show the best agreement with experimental data if the actinides are eight coordinate in the 1:1, An(DPA)+, and 1:2, An(DPA)2−, complexes. The eight-coordinate structures of 1:1 Am and Cm DPA complexes are favored over nine coordinate by less than 1.6 kcal/mol, which is potentially overcome by thermal effects in solution. This ambiguity is removed for Bk and Cf, where the eight-coordinate structure is favored by more than 3.3 kcal/mol. All 1:2 complexes likewise favor the eightcoordinate structure. Such decreases in coordination number



RESULTS AND DISCUSSION Actinide−Dipicolinic Acid Complexation. Computed An3+:(DPA2−)n (An = Am, Cm, Bk, and Cf; n = 1−3) structures compare well with experimental studies assessing actinide hydration numbers and distances using extended X-ray absorption fine structure (EXAFS)22−25 and An(HDPA)3 crystal structures.17 Ion−water and ion−DPA distances are shown in Figure 1 and decrease from Am to Cf, as do the ionic radii of the ions.26 Coordinating water molecules are approximately 0.15 Å further out from the ions than DPA2− oxygen atoms and 0.05 Å further out than DPA2− nitrogen atoms within the same structures. As higher order DPA2− complexes are formed, all coordinating molecules move out from the ions due to greater electronic repulsion (each DPA2− ligand carries two negative charges). Complexation of DPA2− does not affect the oxidation state of the actinide ions considered in this study, as illustrated by the spin density at the metal center in the Supporting Information, Figure 6. Calculations showed the formation of the nine-coordinate aqueous ion is favorable for Am3+, Cm3+, and Bk3+ but not for Cf3+. The free energies of water addition to form the An(H2O)93+ species, however, are less than ±1 kcal/mol in every case at the DFT/PBE level (Supporting Information, Table 1). This free energy difference between the eight- and nine-coordinate ions is very small, comparable to room temperature kBT, and indicates a coexistence of the two species with a very labile ninth coordinated water molecule. High level ab-initio CCSD(T) and spin−orbit CASPT2 (SO-CASPT2) calculations were used to verify the DFT results, as recent 9902

DOI: 10.1021/jacs.7b03251 J. Am. Chem. Soc. 2017, 139, 9901−9908

Article

Journal of the American Chemical Society

Figure 2. Partitioning of americium, curium, berkelium, and californium at 25 °C between aqueous phases containing dipicolinic acid and organic phases containing bis-2-ethyl hexyl phosphoric acid. Lines presented are data fits based on mass balance relationships and are grouped based on experiments using a common HDEHP concentration. Uncertainties are reported at 1σ.

upon complexation have been previously observed.34 The 1:3 complexes are all nine coordinate. Direct comparison of calculations with experimental values was precluded by the approximate treatment of solvation energies of the free water and DPA2− molecules using continuum solvation models; qualitative trends across the An(III) ions studied here were therefore compared. Thermodynamics of Complexation. The experimental partitioning data for actinides at 25 °C and van’t Hoff relationships for Cf−DPA complexation are in Figures 2 and 3, respectively. This information, and other data collected and fit, is in Table 1 to allow comparison of all experimentally derived actinide−dipicolinate complexation thermodynamics. The 1:1 metal:dipicolinate values are comparable for all actinides assessed in this study. A significant increase in stability constants occurs between the Cm and Bk 1:2 and 1:3 values, with differences of 0.73 and 0.47 in the logarithmic values, respectively. Differences in logarithmic stability constants of less than 0.3 or 0.2 exist between adjacent lanthanides heavier than neodymium for the 1:3 lanthanide:dipicolinate complexes.35,36 The lack of notable covalency in lanthanide interactions suggests that larger stability constant differences observed for heavier actinides could be related to an increase in covalent interactions. The enthalpic data presented in Table 1 indicates Cf may be participating in significantly different chemistry than the earlier actinides (especially Am and Cm). All reactions with DPA and Cf are less exothermic than the comparable reactions with americium and curium. Also noteworthy is the decrease in exothermicity of Cf−HDEHP extraction (Supporting Information, Table 2) relative to the other actinides. The origin of the decreased exothermicity for californium interactions with HDEHP and dipicolinate may be a signature of weak covalency.

Figure 3. Observed van’t Hoff relationships for californium with dipicolinic acid in 1 M HClO4.

9903

DOI: 10.1021/jacs.7b03251 J. Am. Chem. Soc. 2017, 139, 9901−9908

Article

Journal of the American Chemical Society Table 1. Complexation Thermodynamics for Select Trivalent Actinides with Dipicolinic Acid in 1 M HClO4a β101 ΔH ΔS β102 ΔH ΔS β103 ΔH ΔS a b

Am3+

Cm3+

Bk3+b

Cf3+

9.14 ± 0.07 −5.3 ± 0.7 24.1 ± 0.7 16.4 ± 0.1 −10.0 ± 0.2 41.1 ± 0.5 22.65 ± 0.1 −15.1 ± 0.2 52.8 ± 0.7

9.1 ± 0.3 −4.8 ± 0.5 26 ± 2 16.5 ± 0.1 −11.0 ± 0.7 38 ± 2 22.67 ± 0.05 −15.8 ± 0.5 51 ± 2

9.07 ± 0.07 −4.5 ± 0.2 26 ± 1 17.2 ± 0.2 −8.6 ± 0.2 50 ± 0.7 23.14 ± 0.01 −15.3 ± 0.7 55 ± 2

9.1 ± 0.1 −3.8 ± 0.5 28 ± 2 17.3 ± 0.4 −8.6 ± 0.2 50.4 ± 0.7 23.29 ± 0.03 −12.7 ± 0.5 64 ± 1

Overall stability constants are reported for 25 °C, whereas ΔH (kcal mol−1) and ΔS (cal mol−1 K−1) terms are assessed over a temperature range. Reference 17.

Electronic Structure of An(III)−Dipicolinic Acid Complexes. Density functional theory (DFT) calculations on the An(DPA)33− (An = Am, Cm, Bk, and Cf) complexes show changes in the electronic structure across the series (Figure 4).

character bonding interactions with ligands. In Am, six occupied MOs have majority 5f character. For Cm, seven occupied MOs have majority 5f character with an additional eight MOs containing at least 5% 5f character. Bk has only three MOs with majority 5f character and 28 with at least 5%, while Cf has only a single MO with majority 5f but 27 MOs with at least 5% f character (see Figure 8 in the Supporting Information). This represents a significant increase in orbital mixing relative to the aqua ions, in which the number of MOs with majority 5f character is 6 at Am and then constant at 7 across the series, while only 6, 12, 15, and 18 MOs have greater than 5% 5f character in Am, Cm, Bk, and Cf, respectively. Data on the molecular orbitals of the aqueous ions are given in the Supporting Information, Figures 15 and 16, with a direct MO comparison of the DPA2− complex and aqueous ion in the Supporting Information, Figure 17. Therefore, the orbital mixing of a complex is ligand dependent. While recent calculations completed on the Bk(DPA)33− complex show significant participation of the 6d orbitals in the metal−ligand bonds, those calculations used f-in-core basis sets to explicitly prevent the 5f orbitals from participating in the bonding.17 The calculations presented here with 5f electrons in the valence space, in contrast, show significant orbital-mixing of the metal 5f orbitals in the An(DPA)33− complexes increasing as the series is traversed. This covalent aspect complements the dominant ionic interactions, and contributes to the observed An:DPA thermodynamic trends. The participation of the 5f orbitals in the Bk and Cf DPA complexes is a significant difference from the earlier actinides, which have been described as having the 6d orbitals dominate the metal−ligand interactions while the 5f orbitals remain unaffected.37,38 Unlike the 5f orbitals, the virtual orbitals with metal 6d character remain relatively unchanged across the series. As the 5f orbital energy drops and the energy of LUMO remains constant, the HOMO/LUMO gap increases with increasing Z, from 0.15, 2.5, 2.9, and 3.0 eV in Am, Cm, Bk, and Cf, respectively. Given these trends, UV−vis measurements of 5f to DPA ligand transitions (MLCT) should increase in energy as the series is traversed. Care must be taken when generalizing these observations, however, as they are highly ligand dependent. For example, An(H2O)93+ complexes have six MOs with greater than 50% 5f character, while Am, Cm, Bk, and Cf respectively have 6, 12, 15, and 18 MOs with greater than 5% 5f character (Supporting Information, Figure 16). These results are consistent with those observed in the DFT study of solid AnO2 species, where, from CmO2 onward, significant orbital overlap exists between the O 2p orbitals and the An 5f orbitals.5,39

Figure 4. Molecular orbital diagram of the An(DPA)33‑ complexes as well as ligand DPA2‑. Blue and red colors represent An 5f and 6d atomic orbital contribution to the canonical molecular orbital, respectively. The length of the red or blue line is proportional to their contribution (percentage) in the canonical molecular orbital. Dashed lines represent unoccupied orbitals, and solid lines represent occupied orbitals.

The americium complex shows localized 5f characteristics, with the highest occupied molecular orbital (HOMO) and subsequent five lower lying orbitals having only small contributions (less than 5%) from the DPA ligands. These occupied orbitals are approximately 1.5 eV higher in energy than a block of orbitals derived predominantly from the DPA ligands, consisting primarily of oxygen lone pairs and π bonds in the pyridine ring. Occupied molecular orbitals (MO’s) containing metal 6d character are deep in energy (approximately 7−8 eV below the HOMO) with negligible mixing with the ligand (no more than 3%). In Am, the HOMO/LUMO gap is very small, only 0.15 eV, as Am has only six 5f electrons and the LUMO is primarily 5f in character. Above the LUMO sits a block of unoccupied orbitals with largely π* character of DPA2− ligands, and above these are a number of unoccupied orbitals with 6d character from the metal ion. Moving across the series, the energies of An 5f orbitals drop and become degenerate in energy with the DPA ligand orbitals, resulting in increased orbital mixing and more delocalized 5f 9904

DOI: 10.1021/jacs.7b03251 J. Am. Chem. Soc. 2017, 139, 9901−9908

Article

Journal of the American Chemical Society

Figure 5. Gibbs free energies of DPA complexation to form the 1:1−3 DPA complexes. Reactions with differing coordination states for the ions are given in the Theory panels: (◆) CN 9 to CN 9, (▲) CN 8 to CN 9, (■) CN 9 to CN 8, and (●) CN 8 to CN 8. Energies are plotted as ΔΔG relative to the complexation free energy of Am.

Comparisons with Recent Observations. Recent discussions regarding the chemistry of the heaviest actinides have centered on the unique chemistry of californium compared to other actinides in the series.13−17 The formation of individually unique borate lattices for Pu, Am, Cm, and Cf, as well as the observation of charge transfer bands and suppressed magnetic moments in californium complexes, have encouraged suggestions that californium may participate in fundamentally different chemistry than previously thought, and this chemistry could not be predicted by extrapolating information from the earlier actinides.15,16,40 Hypotheses concerning the “break” in californium’s chemistry relative to the lighter actinides have centered on the thermodynamic accessibility of a metastable divalent state resulting from a ligand-to-metal charge transfer band starting at 1.86 eV.3,40 These dialogues have generated interest regarding the origin of increased covalency for the heavy actinides. The chemistry of berkelium is particularly interesting for comparison with Cf for two reasons: (1) their adjacent proximity on the periodic table can resolve questions about the precise location of a second actinide “break” and (2) Bk does not have an accessible divalent state but does have an accessible tetravalent state. If the available redox chemistry of a particular actinide drives the accessibility of covalency in

actinide dipicolinate and borate interactions, then Bk and Cf should have inverted chemical properties. However, observations comparing Bk or Cf seem unrelated to the thermodynamic accessibility of their respective tetravalent and divalent states. The formation of the An[B6O8(OH)5] borate lattice was unique to Cf and has been observed for Bk. The suppression of the magnetic moment for dipicolinate and borate structures, while more pronounced for Cf, exists for Bk as well. This suggests the acquired borate lattice structure and suppressed magnetic moment are not uniquely related to the accessibility of californium’s divalent state. Other signatures regarding californium’s status as the “break” point in the actinide series are more ambiguous. Broad band photoluminescence and vibronic coupling signatures do suggest covalency in californium, but a photoluminescent charge transfer band with vibronic coupling is observed in both berkelium and californium dipicolinate samples.17 The photoluminescence spectrum of the berkelium sample is provided in the Supporting Information of this recent berkelium report. The fair possibility exists that the charge transfer band in the Bk sample arises from Bk decay to Cf and the charge transfer arises from the Cf dipicolinate structure, but the possibility of a Bk charge transfer cannot currently be disregarded. 9905

DOI: 10.1021/jacs.7b03251 J. Am. Chem. Soc. 2017, 139, 9901−9908

Article

Journal of the American Chemical Society

actinides, through nobelium, and a variety of ligands would be appropriate to test the validity of the hypothesis presented here. Due to the limited amounts of material available, experiments and theory need to work in tandem to constructively iterate on the individual methodologies. The lack of other chemical effects in this part of the periodic table could allow the heavy actinides to serve as a unique playground to probe the competing effects of orbital overlap versus energy degeneracy driven covalency. These recent observations should continue to reinvigorate the field of heavy actinide science beyond californium and encourage the use of heavy actinides to understand fundamental chemical interactions throughout the periodic table.

The molecular orbital diagram presented in Figure 4, and comparison with thermodynamic data in Table 1 and Figure 5, provides a framework to deconvolute factors driving heavy actinide covalency. The molecular orbital diagram suggests for the earliest considered actinide, Am, the 5f orbitals are too high in energy to interact with the dipicolinic acid molecular orbitals and complex is largely ionic. The significantly more exothermic thermochemical signature for americium, relative to the lanthanide dipicolinates, suggests the very small amount of orbital overlap covalency drives group actinide/lanthanide separations. The core-like nature of the 4f lanthanide orbitals significantly limits their ability to participate in orbital overlap covalency. Only an energy difference of 0.12 eV is necessary to generate a separation factor on the order of 100.41 Previous studies comparing americium and europium dipicolinate stability constants only have an energy difference of 0.014 eV.36 When traversing the actinide series, the 5f orbitals contract and the orbital energetics decrease. This contraction decreases the probability of orbital overlap covalency; however, the decrease in orbital energetics encourages degeneracy with the DPA2− MO. This increased degeneracy promotes energy degeneracy driven covalency. The complexation thermodynamics and electronic structures both show effects of energy degeneracy driven covalency. The significant contribution of berkelium and californium 5f orbitals to the actinide− dipicolinate molecular orbital complex seems to encourage the jump in complexation strength for berkelium and californium relative to americium and curium. If the interaction was strongly orbital overlap, much more significant jumps in stability constants might be observed. The relatively modest increase in separation factor between curium and berkelium, with a difference in 1:3 stability constants of only 0.47, and the decrease in exothermicity, is consistent with a relatively weak interaction from the 5f orbital energy degeneracy driven covalency as opposed to 6d-orbital driven covalency. Additionally, the computational panels of Figure 5 suggest these observations are not the result of a changing coordination number across the series, as all considered reactions show a clear difference in the complexation energy between Cm and Bk, corresponding to that observed experimentally, regardless of coordination number. Amending a Classic Hypothesis. The interactions of borate and dipicolinate ligands with berkelium and californium suggests the opportunity for coaxing energy degeneracy driven bonding phenomena across the heavier part of the actinide series is possible and, courtesy observations from previous studies,15,17 may be a more prominent aspect of their chemistry than previously recognized. Energy degeneracy driven binding is relevant in this part of the periodic table since decrease in 5f orbital energetics across the series can bring these atomic orbitals into energy degeneracy with many common complexants. Meanwhile, orbital overlap covalency decreases due to the contraction the 5f orbitals relative to the limited radial extension of complexants. These observations allow for an amendment to the original Seaborg actinide hypothesis, which would have predicted less covalency for heavier actinides due to contraction of the f orbitals. The origination of heavy actinide covalency from 5f energy degeneracy affirms the current arrangement of the periodic table. Since the chemistry of these elements is relatively underexamined,12 knowledge about what complexants can induce these interactions, and how significantly this chemistry pervades though the rest of the series, is limited. Work with heavier



METHODS

Caution! The isotopes used here (241Am, 244Cm, 249Bk, and 249Cf) are hazardous radionuclides with high specific activities. Any work with these or other transuranic elements should be carried out in specialized laboratory spaces with appropriate controls for radiological hazards. Acid Association Constants. The acid association constants of dipicolinic acid were assessed at appropriate temperatures to allow the calculation of free DPA2− available for complexation. Titrations were completed manually using a Thermo/Ross semimicro combination pH electrode. To allow work in NaClO4 media, the filling solution of the electrode, originally potassium chloride (KCl) was replaced with 3.0 M sodium chloride (NaCl) to prevent potassium perchlorate (KClO4) precipitation at the glass frit. Titrations were also maintained at 1.0 M ionic strength and completed under a nitrogen blanket to prevent carbon dioxide contamination. The measured mV data were converted to hydrogen ion concentrations using an mV vs pcH calibration curve generated in a strong acid-strong base titration of HClO4 with NaOH at 1.0 M total ionic strength. All titrations were repeated in at least triplicate, and data fitting was completed using the fitting program Hyperquad. The values for the acid dissociation constants have been previously reported.17 Extraction Measurements. All complexation thermodynamics were assessed by using competitive solvent extraction investigations with bis-2-ethyl phosphoric acid (HDEHP) dissolved in o-xylene (berkelium) or n-dodecane (americium, curium, and californium). The distribution ratios are calculated as D = [M]org/[M]aq. The extraction constant (KEx) was assessed for all actinides at various temperatures. To assess β101, β102, and β103 metal−DPA stability constants, metal partitioning between an HDEHP organic phase and an aqueous phase with increasing DPA concentration was measured. The 249Bk had a 2% contamination of 249Cf at the time of the Bk extraction experiments. Since the Bk concentration was low (10−11 M) relative to the concentration of the DPA2− (10−9 to 10−6) corrections to the 249Bk HDEHP extraction and DPA stability constant measurements were not made. The ionic strength of the aqueous phase was maintained at 1.0 M using NaClO4. All phases were pre-equilibrated with an appropriate aqueous or organic phase prior to use in the distribution study at the temperature of a given study. Pre-equilibration contact times were 5 min and contact times for thermodynamic measurements were 15 min. Contacts were completed using a Labteck shaker with aluminum temperature block fabricated in house. The pcH of the aqueous phase was measured after contact by using a series of standardized acid solutions at 1.0 M NaClO4. Conversions molarity were afforded by density determinations at 22 °C. Partitioning of 241 Am, 244Cm, and 249Bk was monitored by the use of a Packard 2500 Liquid Scintillation counter. Partitioning of 249Cf was monitored using Packard Cobra II Autogamma counter. Radionuclides were provided by Eckert & Ziegler (244Cm), the Department of Energy (249Bk), and available stocks from Florida State University (241Am and 249Cf). All thermodynamic constants were fit in QtiPlot using nonlinear regression model weighting of the distribution data using w = 1/σ2 weighting. Distribution values are determined based on D = [M]org/ [M]aq. Metal KEx values were fit assuming equilibria and mass balance relationships previously established in the literature.36 9906

DOI: 10.1021/jacs.7b03251 J. Am. Chem. Soc. 2017, 139, 9901−9908

Article

Journal of the American Chemical Society Electronic Structure Calculations. Structures of the aqueous An(III) ions and their 1:1−3 deprotonated DPA complexes at both CN = 8 and CN = 9 (due to water addition) were optimized using density functional theory (DFT) with the PBE functional,42 relativistic ZORA Hamiltonian, and triple-ζ plus one polarization function (TZP) basis sets with the frozen core approximation applied to the inner shells [1s2-4f14] for actinide atoms and [1s2] for C, N and O atoms.43,44 Frequency calculations were performed to determine thermodynamic properties of each structure. All calculations were performed using the COSMO solvation model using a radius of 2.223 Å for the An ions and the default van der Waals radius from the MM3 method divided by 1.2 for other atoms.45−47 At standard state conditions (1 atm), reactant species, particularly reactant water and DPA2− molecules, have more freedom prior to complexation than they do after, which is untrue in solution. Therefore, frequency calculations were completed at a pressure of 1354 atm, determined from the density of liquid water at the standard state.31 Similar corrections have been applied only to reactant water molecules in studies of the hydration lanthanides,48,49 but in this study the corrections were applied to all species. All calculations were performed using ADF 2016.50,51 DFT results were verified using ab initio wave function theory methods, including CCSD(T), SR-CASPT2, and SO-CASPT2. Spin−orbit effects were found to be negligible for these systems, as expected due to the unchanging oxidation state of the An ions. DFT/ PBE was found to reasonably reproduce these high level results. Additional details are available in the Supporting Information.



operated by Los Alamos National Security, LLC, for the National Nuclear Security Administration of U.S. Department of Energy (Contract DE-AC52-06NA25396). The Bk-249 and Cf-249 used in this and prior research by this PI was supplied by the United States Department of Energy Office of Science by the Isotope Program in the Office of Nuclear Physics. Electronic structure calculations were completed using the Molecular Science Computing Facilities in the William R. Wiley Environmental Molecular Sciences Laboratory, a national scientific user facility sponsored by the U.S. DOE BER and located at Pacific Northwest National Laboratory.



(1) Seaborg, G. T. Nucleonics 1949, 5, 16−36. (2) Choppin, G. R. J. Alloys Compd. 2002, 344, 55−59. (3) Jensen, M. P.; Bond, A. H. J. Am. Chem. Soc. 2002, 124, 9870− 9877. (4) Neidig, M. L.; Clark, D. L.; Martin, R. L. Coord. Chem. Rev. 2013, 257, 394−406. (5) Prodan, I. D.; Scuseria, G. E.; Martin, R. L. Phys. Rev. B, 2007, 76. (6) Burns, C. J. Science 2005, 309, 1823−1824. (7) Kozimor, S. A.; Yang, P.; Batista, E. R.; Boland, K. S.; Burns, C. J.; Clark, D. L. J. Am. Chem. Soc. 2009, 131, 12125−12136. (8) Lukens, W. W.; Edelstein, N. M.; Magnani, N.; Hayton, T. W.; Fortier, S.; Seaman, L. A. J. Am. Chem. Soc. 2013, 135, 10742−10754. (9) Minasian, S. G.; Keith, J. M.; Batista, E. R.; Boland, K. S.; Clark, D. L.; Conradson, S. D.; Kozimor, S. A.; Martin, R. L.; Schwarz, D. E.; Shuh, D. K.; Wagner, G. L.; Wilkerson, M. P.; Wolfsberg, L. E.; Yang, P. J. Am. Chem. Soc. 2012, 134, 5586−97. (10) Lukens, W. W.; Speldrich, M.; Yang, P.; Duignan, T. J.; Autschbach, J.; Kögerler, P. Dalt. Trans 2016, 45, 11508−11521. (11) Bradley, J. A.; Yang, P.; Batista, E. R.; Boland, K. S.; Burns, C. J.; Clark, D. L.; Conradson, S. D.; Kozimor, S. A.; Martin, R. L.; Seidler, G. T.; Scott, B. L.; Shuh, D. K.; Tyliszczak, T.; Wilkerson, M. P.; Wolfsberg, L. E. J. Am. Chem. Soc. 2010, 132, 13914−13921. (12) Drader, J. A.; Luckey, M.; Braley, J. C. Solvent Extr. Ion Exch. 2016, 34, 114−125. (13) Polinski, M. J.; Wang, S.; Alekseev, E. V.; Depmeier, W.; Albrecht-Schmitt, T. E. Angew. Chem., Int. Ed. 2011, 50, 8891−8894. (14) Polinski, M. J.; Wang, S.; Alekseev, E. V.; Depmeier, W.; Liu, G.; Haire, R. G.; Albrecht-Schmitt, T. E. Angew. Chem., Int. Ed. 2012, 51, 1869−1872. (15) Polinski, M. J.; Garner, E. B.; Maurice, R.; Planas, N.; Stritzinger, J. T.; Parker, T. G.; Cross, J. N.; Green, T. D.; Alekseev, E. V.; Van Cleve, S. M.; Depmeier, W.; Gagliardi, L.; Shatruk, M.; Knappenberger, K. L.; Liu, G.; Skanthakumar, S.; Soderholm, L.; Dixon, D. A.; Albrecht-Schmitt, T. E. Nat. Chem. 2014, 6, 387−392. (16) Cary, S. K.; Vasiliu, M.; Baumbach, R. E.; Stritzinger, J. T.; Green, T. D.; Diefenbach, K.; Cross, J. N.; Knappenberger, K. L.; Liu, G.; Silver, M. A.; DePrince, A. E.; Polinski, M. J.; Van Cleve, S. M.; House, J. H.; Kikugawa, N.; Gallagher, A.; Arico, A. A.; Dixon, D. A.; Albrecht-Schmitt, T. E. Nat. Commun. 2015, 6, 6827. (17) Silver, M. A.; Cary, S. K.; Johnson, J. A.; Baumbach, R. E.; Arico, A. A.; Luckey, M.; Urban, M.; Wang, J. C.; Polinski, M. J.; Chemey, A.; Liu, G.; Chen, K.-W.; Van Cleve, S. M.; Marsh, M. L.; Eaton, T. M.; van de Burgt, L. J.; Gray, A. L.; Hobart, D. E.; Hanson, K.; Maron, L.; Gendron, F.; Autschbach, J.; Speldrich, M.; Kogerler, P.; Yang, P.; Braley, J.; Albrecht-Schmitt, T. E. Science 2016, 353, aaf3762. (18) Kaltsoyannis, N. Inorg. Chem. 2013, 52, 3407−3413. (19) Hubbard, J.; Rimmer, D. E.; Hopgood, F. R. A. Proc. Phys. Soc., London 1966, 88, 13−36. (20) Barros, N.; Maynau, D.; Maron, L.; Eisenstein, O.; Zi, G.; Andersen, R. A. Organometallics 2007, 26, 5059−5065. (21) Pettit, G.; Powell, K. J. The IUPAC Stability Constants Database, ver. 5.76; IUPAC: Otley, U.K., 2001. (22) Antonio, M. R.; Williams, C. W.; Soderholm, L. Radiochim. Acta 2002, 90, 851−856.

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b03251. Calculated energy of water addition to aqueous and DPA complexes; HDEHP extraction constants; fitting of stability constants; thermodynamic parameters of HDEHP extraction; spin density on the actinide ions; mulliken charge on actinide ions; 6d and 5f orbital populations; orbital mixing of actinide 5f; isosurface view of molecular orbitals; molecular orbital diagrams; ab initio calculations; Cartesian coordinates of selected structures. (PDF)



REFERENCES

AUTHOR INFORMATION

Corresponding Authors

*[email protected] *[email protected] ORCID

Enrique R. Batista: 0000-0002-3074-4022 Ping Yang: 0000-0003-4726-2860 Jenifer C. Shafer: 0000-0001-9702-1534 Author Contributions

J.C.S., M.U., and M.L. carried out the experiments and analyzed the data. P.Y., M.P.K., J.S., and E.R.B. completed computational experiments and data interpretation. J.C.S. and P.Y. designed the experiments and computations. J.C.S., P.Y., and M.K. wrote the manuscript with input from all authors. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, Heavy Elements Chemistry Program at Colorado School of Mines (under Award Number DE-SC0012039) and at Los Alamos National Laboratory. Los Alamos National Laboratory is 9907

DOI: 10.1021/jacs.7b03251 J. Am. Chem. Soc. 2017, 139, 9901−9908

Article

Journal of the American Chemical Society (23) Skanthakumar, S.; Antonio, M. R.; Wilson, R. E.; Soderholm, L. Inorg. Chem. 2007, 46, 3485−91. (24) Revel, R.; Den Auwer, C.; Madic, C.; David, F.; Fourest, B.; Hubert, S.; Le Du, J. F.; Morss, L. R. Inorg. Chem. 1999, 38, 4139− 4141. (25) Allen, P. G.; Bucher, J. J.; Shuh, D. K.; Edelstein, N. M.; Craig, I. Inorg. Chem. 2000, 39, 595−601. (26) D’Angelo, P.; Martelli, F.; Spezia, R.; Filipponi, A.; Denecke, M. A. Inorg. Chem. 2013, 52, 10318−10324. (27) Banik, N. L.; Vallet, V.; Réal, F.; Belmecheri, R. M.; Schimmelpfennig, B.; Rothe, J.; Marsac, R.; Lindqvist-Reis, P.; Walther, C.; Denecke, M. A.; Marquardt, C. M. Dalton Trans. 2016, 45, 453−457. (28) Kelley, M. P.; Yang, P.; Clark, S. B.; Clark, A. E. Inorg. Chem. 2016, 55, 4992−4999. (29) Yang, T.; Bursten, B. E. Inorg. Chem. 2006, 45, 5291−5301. (30) Wiebke, J.; Moritz, A.; Cao, X.; Dolg, M. Phys. Chem. Chem. Phys. 2007, 9, 459−465. (31) Martin, R. L.; Hay, P. J.; Pratt, L. R. J. Phys. Chem. A 1998, 102, 3565−3573. (32) Lindqvist-Reis, P.; Apostolidis, C.; Rebizant, J.; Morgenstern, A.; Klenze, R.; Walter, O.; Fanghänel, T.; Haire, R. G. Angew. Chem., Int. Ed. 2007, 46, 919−922. (33) Lindqvist-Reis, P.; Klenze, R.; Schubert, G.; Fanghänel, T. J. Phys. Chem. B 2005, 109, 3077−83. (34) Jensen, M. P.; Beitz, J. V.; Rogers, R. D.; Nash, K. L. J. Chem. Soc. Dalt. Trans. 2000, 3058−3064. (35) Grenthe, I. J. Am. Chem. Soc. 1961, 83, 360−364. (36) Heathman, C.; Nash, K. Sep. Sci. Technol. 2012, 47, 2029−2037. (37) Burns, C. J.; Bursten, B. E. Comments Inorg. Chem. 1989, 9, 61− 93. (38) Pepper, M.; Bursten, B. E. Chem. Rev. 1991, 91, 719−741. (39) Wen, X.-D.; Martin, R. L.; Henderson, T. M.; Scuseria, G. E. Chem. Rev. 2013, 113, 1063−96. (40) Liu, G.; Cary, S. K.; Albrecht-Schmitt, T. E. Phys. Chem. Chem. Phys. 2015, 17, 16151−16157. (41) Fryer-Kanssen, I.; Austin, J.; Kerridge, A. Inorg. Chem. 2016, 55, 10034−10042. (42) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865−3868. (43) Cao, X.; Dolg, M.; Stoll, H. J. Chem. Phys. 2003, 118, 487−496. (44) Kuchle, W.; Dolg, M.; Stoll, H.; Preuss, H. J. Chem. Phys. 1994, 100, 7535−7542. (45) Klamt, A.; Schuurmann, G. J. Chem. Soc., Perkin Trans. 2 1993, 2, 799. (46) Pye, C. C.; Ziegler, T. Theor. Chem. Acc. 1999, 101, 396−408. (47) Allinger, N. L.; Zhou, X.; Bergsma, J. J. Mol. Struct.: THEOCHEM 1994, 312, 69−83. (48) Dinescu, A.; Clark, A. E. J. Phys. Chem. A 2008, 112, 11198− 206. (49) Gutowski, K. E.; Dixon, D. A. J. Phys. Chem. A 2006, 110, 8840− 8856. (50) Te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca Guerra, C.; van Gisbergen, S. J. a.; Snijders, J. G.; Ziegler, T.; Velde, G. T. E.; Guerra, C. F.; Gisbergen, S. J. a V. a N. J. Comput. Chem. 2001, 22, 931−967. (51) Fonseca Guerra, C.; Snijders, J. G.; Te Velde, G.; Baerends, E. J. Theor. Chem. Acc. 1998, 99, 391−403.

9908

DOI: 10.1021/jacs.7b03251 J. Am. Chem. Soc. 2017, 139, 9901−9908