Investigations of the Cobalt Hexamine Uranyl Carbonate System: Understanding the Influence of Charge and Hydrogen Bonding on the Modification of Vibrational Modes in Uranyl Compounds.

Hydrogen bonding networks within hexavalent uranium materials are complex and may influence the overall physical and chemical properties of the system. This is particularly true if hydrogen bonding takes places between the donor and the oxo group associated with the uranyl cation (UO22+). In the current study, we evaluate the impact of charge-assisted hydrogen bonding on the vibrational modes of the uranyl cation using uranyl tricarbonate [UO2(CO3)3]4- interactions with [Co(NH3)6]3+ as the model system. Herein, we report the synthesis and structural characterization of five novel compounds, [Co(NH3)6]Cl(CO3) (Co_Cl_CO3), [Co(NH3)6]4[UO2(CO3)3]3(H2O)11.67 (Co4U3), [Co(NH3)6]3[UO2(CO3)3]2Cl (H2O)7.5 (Co3U2_Cl), [Co(NH3)6]2[UO2(CO3)3]Cl2 (Co2U_Cl), and [Co(NH3)6]2[UO2(CO3)3]CO3 (Co2U_CO3), which contain differences in the crystalline packing and extended hydrogen bonding networks. We show that these slight changes in the supramolecular assembly and hydrogen bonding networks result in the modification of modes as observed by infrared and Raman spectroscopy. We use density functional theory calculations to assign the vibrational modes and provide an understanding about how uranyl bond perturbation and changes in hydrogen bonding interactions can impact the resulting spectroscopic signals.

Introduction

Hydrogen bonding represents an important interaction in chemical systems, and the formation of hydrogen bond networks can directly influence chemical and physical properties of solid-state materials.[1−6,5−19] The extent to which hydrogen bonding impacts the properties of high-valent actinide materials is of interest because of the unique nature of bonding within these complexes.[7] Uranium is one of the most naturally abundant actinide elements and is commonly found in the hexavalent oxidation state in aqueous solutions and oxidizing conditions.[8] Typically, U(VI) engages in covalent interactions with two oxygen atoms to create the linear triatomic uranyl cation [O=U(VI)=O]2+ that further coordinates to four, five, or six equatorial ligands to create a square, pentagonal, or hexagonal coordination geometry.[9,10] Given the strong bonding within the actinyl unit, the trans-oxo groups are considered weak Lewis bases that do not readily engage in intermolecular interactions, including hydrogen bonding.[11] For example, Watson and Hay utilized density functional theory (DFT) calculations to evaluate the geometries and energetics of the uranyl oxo group as a hydrogen bond acceptor and found that traditional hydrogen bond donors are actually repelled by the oxo groups in [UO2(H2O)5]2+.[12]

In a review compiled by Fortier and Hayton, instances where uranyl oxo groups interact with Lewis acids, including hydrogen atoms, were highlighted, yet this interaction does not seem to disrupt the uranyl bond to any extent as evidenced by U=O bond distances.[13] This suggests that the intermolecular forces are quite weak and do not activate the oxo in any significant way. However, there are instances, such as in Pacman pyrrole-imine macrocycles, where the uranyl bond is perturbed by the presence of a hydrogen bond because of the specific ligand architecture.[14−16] In addition, Watson and Hay observed that the identity of the equatorial ligand seems to play a role as the repellent nature of the uranyl oxo within [UO2(H2O)5]2+ can become attractive with the addition of nitrate groups to form [UO2(NO3)2(H2O)2]0.[12] Therefore, it is important to consider what factors can increase the hydrogen bonding interaction to the uranyl oxo and how this impacts the observable properties of the material.

In the current study, we explore the influence of charge-assisted hydrogen bonding on the uranyl oxo by investigating uranyl tricarbonate coordination complexes [UO2(CO3)3]4– crystallized with [Co(NH3)6]3+ (Figure ). We hypothesized that a stronger hydrogen bond network, specifically charge-assisted hydrogen bonds, would more readily interact with the oxo atom, impact the overall bond strength within the uranyl cation, and influence the related vibrational spectroscopy of the solid-state material. Charge-assisted hydrogen bonds are unique because of the ionic character of the acceptor and donor atoms, which strengthen the electrostatic interaction of hydrogen bonds and have the potential to weaken and elongate the uranyl bond.[17] Uranyl coordination compounds also have characteristic and identifiable vibrational spectra, where the symmetric (v1) and the asymmetric (v3) stretching bands of the uranyl are Raman- and IR-active, respectively. Positional changes or shifts in vibrational signals for characteristic UO22+ bands are commonly attributed to the identity of the equatorial ligands, but additional activation of bands may occur in the presence of hydrogen bond networks.[18−20] We hypothesized that the hydrogen bonding networks would lead to distortion of the uranyl bond and result in modification of the asymmetric and symmetric stretching band within the vibrational spectra. Uranyl carbonate compounds are good model systems because they have been studied thoroughly experimentally and computationally because of their importance in geologic environments, aqueous systems, and the nuclear fuel cycle.[8,21] Cobalt hexamine represents an excellent hydrogen donor group because it possesses a high charge density at the metal center, and multiple hydrogen atoms are available for bonding interactions. Cobalt(III) hexamine has also been shown to engage in charge-assisted hydrogen bonding within biological systems and crystalline materials, including U(VI) compounds.[22−24] In the current study, we report the structural characterization of five novel compounds with varied hydrogen bonding networks, [Co(NH3)6]Cl(CO3) (Co_Cl_CO), [Co(NH3)6]4[UO2(CO3)3]3(H2O)11.67 (Co4U3), [Co(NH3)6]3[UO2(CO3)3]2Cl(H2O)7.5 (Co3U2_Cl), [Co(NH3)6]2[UO2(CO3)3]Cl2 (Co2U_Cl), and [Co(NH3)6]2[UO2(CO3)3]CO3 (Co2U_CO). We utilize solid-state Raman and IR spectroscopy to then evaluate the influence of the structurally characterized hydrogen bonding network on the uranyl bond. In addition, we employ DFT calculations to perform geometric and vibrational analysis and calculate force constants which provide further insights into the impact of charge-assisted hydrogen bonding in solid-state U(VI) compounds.

Figure 1

Molecular units of (A) uranyl tricarbonate and (B) cobalt hexamine that are utilized in this study to evaluate the impacts of charge-assisted hydrogen bonding on the U=O bond.

Experimental Methods

Synthesis of Materials

Uranyl nitrate hexahydrate (UO2(NO3)2·6H2O) was purchased from Flinn Scientific Inc. Caution! Uranyl nitrate hexahydrate [UO2(NO3)2]·6H2O contains U-238, a naturally radioactive element. All uranium-bearing materials should be handled with standard precautions and by trained personnel. Tetramethyl ammonium hydroxide (TMAOH) 25% in water, cobalt(III) hexamine chloride ([Co(NH3)6]Cl3), and potassium carbonate (K2CO3) and sodium carbonate (Na2CO3) were purchased from Acros Organics, TCI, and Sigma-Aldrich, respectively. All chemicals were used as received, and stock solutions were prepared with millipure (18 MΩ) water.

[Co(NH3)6]Cl(CO3) (Co_Cl_CO)

Equivalent amounts (1.0 mL) of a 0.18 M [Co(NH3)6]Cl3 solution were added to 0.20 M K2CO3 and millipure water in a 20 mL glass scintillation vial. The vial was left uncapped to encourage slow evaporation and reddish-brown crystals with a prismatic morphology formed over 12 h with a 60% yield based upon Co.

[Co(NH3)6]4[UO2(CO3)3]3H2O11.67 (Co4U3)

In a 20 mL glass scintillation vial, 0.18 M [Co(NH3)6]Cl3 (0.5 mL) was added to 1 mL of 0.2 M UO3 dissolved within the K2CO3 stock solution and 1 mL of millipure water. The solution slowly evaporated at room temperature, and orange blocks formed after 1 day. Percent yield of the Co4U3 synthesis is 60% based upon Co.

[Co(NH3)6]3[UO2(CO3)3]2Cl H2O7.5 (Co3U2_Cl)

Aliquots of a 0.1 M uranyl nitrate stock solution (0.50 mL), 0.1 M TMAOH (0.15 mL), 0.46 M [Co(NH3)6]Cl3 (0.45 mL), and 1.0 mL of 1.0 M K2CO3 were combined in a 10 mL glass vial and placed in a refrigerator uncapped for 4 days. Orange crystals formed with a plate morphology with yields of <10% based upon Co.

[Co(NH3)6]2[UO2(CO3)3]Cl2 (Co2U1_Cl)

A 0.18 M [Co(NH3)6]Cl3 solution (1 mL) was combined with 1 mL of 0.2 M UO3 in 0.2 M K2CO3 and 3 mL of millipure water. The vial was left uncapped, and light orange columnar crystals formed after 1 day in yields of 48% based upon Co.

[Co(NH3)6]2[UO2(CO3)3]CO3H2O3 (Co2U1_CO)

Aliquots of a 0.1 M uranyl nitrate stock solution (1 mL), 0.1 M TMAOH (0.3 mL), 0.18 M [Co(NH3)6]Cl3 (1 mL), and 1 mL of 1.0 M K2CO3 were combined in a 10 mL glass vial and left uncapped for 5 days. Orange crystals with a columnar morphology were produced in yields of 30% based upon Co.

Compounds Co3U2_Cl and Co2U1_CO could also be synthesized without the addition of the carbonate anion if the solutions were left exposed to standard atmospheric conditions in the laboratory. For Co3U2_Cl, aliquots of the 0.1 M uranyl nitrate stock solution (1 mL), 0.1 M TMAOH (0.3 mL), and 0.01 M [Co(NH3)6]Cl3 (1 mL) were combined in a 20 mL glass scintillation vial. The pH of the solution was initially 12, and a solid yellow precipitate formed on the bottom of the vial. This initial solid dissolved after 18 h, and the clear orange solution was allowed to slowly evaporate for 3 days in an uncapped vial to produce orange plates of Co3U2_Cl. Co2U1_CO formed from a similar synthetic condition where the 0.1 M uranyl nitrate stock solution (1.0 mL) and 0.1 M TMAOH (0.3 mL) were added to a 20 mL glass scintillation vial. In the case of Co2U1_CO, a higher concentration of [Co(NH3)6]Cl3 (1.0 mL of 0.18 M) was also added to the solution. An initial yellow precipitate again formed and then re-dissolved after 24 h. After 3 days of slow evaporation in an uncapped vial, orange rods of Co2U1_CO crystallized from the mother liquor.

This process fits well with what is currently understood for U(VI) chemistry under basic conditions (pH 12). Initially increasing the pH will result in U(VI) hydrolysis, which will cause precipitation of kinetically stable oxyhydroxide phase values.[25] When water is in equilibrium with the atmosphere, it will contain dissolved CO2 at concentrations that are controlled by Henry’s law. This means that at pH 12, with 400 ppm CO2 in the air, we will reach levels greater than 10–1 moles of dissolved CO2 per liter (where dissolved CO2 is equal to H2CO3 + HCO3– + CO32–).[26] At these concentrations, the uranyl tricarbonate phase is the only species present for U(VI) at these pH values.[25]

Single-Crystal X-ray Diffraction

Single crystals of each coordination compound were visually identified on a polarized microscope, harvested from their respective mother liquors, and mounted on a MiTeGen MicroMount using NVH immersion oil (Cargille Labs). Structural information was collected on a Bruker D8 Quest single-crystal X-ray diffractometer equipped with a microfocus beam (Mo Kα; λ = 0.71073 Å) and an Oxford Systems low temperature cryosystem. Data were collected with the Bruker APEX3 software package,[27] and peak intensities were corrected for Lorentz, polarization, background effects, and absorption. The structure solution was determined by intrinsic phasing methods and refined on the basis of F2 for all unique data using the SHELXTL version 5 series of programs.[28] Metal atoms (U, Co) were located by direct methods, and the C, O, N, and Cl atoms were identified and modeled from the difference Fourier maps after partial refinement.

Many of the compounds contained positional disorder, which was accounted for by considering partial occupancy and split sites. Co4U3 displayed disorder associated with the [UO2(CO3)]4– complex that resulted in unreasonable U–U distances if the complex was fully occupied. Additional unit cell parameters were evaluated using CELLNOW, and doubling of the axes or lowering the symmetry of the space group did not rectify the positional disorder. The complex was successfully modeled using partial occupancy (50%) as evidenced by reasonable displacement parameters and bond distances/angles. Co3U2_Cl was originally solved in the hexagonal, P-3 space group, but significant disorder and unreasonable displacement parameters suggested that a lower symmetry space group was more appropriate. The lowest Rint value and most reasonable displacement values were achieved in an orthorhombic space group (Cmma) and resulted in the most agreeable thermal displacement parameters. Co2U1_Cl was placed in a triclinic P-1 space group, after multiple attempts to model the lattice Cl over several crystallographic positions did not provide proper thermal parameters or charge neutrality. Co2U1_CO solved in the hexagonal space group P-3 and displayed disorder associated with one cobalt hexamine cation. One disordered water was also modeled as partially occupied over three positions in the lattice, while the interstitial carbonate anion required a DFIX constraint to enable reasonable C–O bond distances. Carbonate anions coordinated to the U(VI) atom in Co2U1_CO also contained displayed disorder for one of the O atoms and were modeled as 50% occupied over two crystallographic positions.

Hydrogen atoms were included on all well-ordered NH3 and H2O molecules for the CoU compounds. A riding model was used to place the hydrogen atoms on the amine groups of the [Co(NH3)6]3+ cation, except for the Co(2) atom in Co2U1_CO because there was significant disorder of the amine groups about the metal center. Water molecules within the lattice were also modeled with H atoms when possible, and the crystallographic positions of these atoms were in the difference Fourier map after modeling electron density of all the heavier atoms in the lattice. The bond distances and angles of the H atoms associated with the water molecules were restrained using DFIX and DANG commands.

Selected crystallographic parameters can be found in Table and additional bonding information regarding the Co and interstitial anions/molecules can be found in the Supporting Information in Tables S1–S5. Images depicting the asymmetric unit with thermal ellipsoids for each of the compounds can also be found in the Supporting Information (Figures S1–S5). Crystallographic information files can be found on the Cambridge Structural Database by requesting numbers 2177491–2177495.

Table 1

Select Crystallographic Parameters for [Co(NH3)6]Cl(CO3) (Co_Cl_CO), [Co(NH3)6]4[UO2(CO3)3]3(H2O)11.67 (Co4U3), [Co(NH3)6]3[UO2(CO3)3]2Cl (H2O)7.5 (Co3U2_Cl), [Co(NH3)6]2[UO2(CO3)3]Cl2 (Co2U1_Cl), and [Co(NH3)6]2[UO2(CO3)3]CO3(H2O)3 (Co2U1_CO)

	Co_Cl_CO3	Co4U3	Co3U2_Cl	Co2U1_Cl	Co2U1_CO3
empirical formula	CN6H18O3CoCl	C9N24H69O44.67U3Co4	C6N18H69O29.5U2Co3Cl	C3N12H36O14UCo2Cl2	C4N12H42O17Uco2
formula weight	256.56	2178.14	1553.73	891.13	886.16
space group	P213	P21/n	Cmma	P-1	P-3
a (Å)	9.9014(5)	16.978(5)	25.2392(19)	6.8093(3)	15.5979(5)
b (Å)	9.9014(5)	7.780(3)	15.1709(13)	12.5621(7)	15.5979(5)
c (Å)	9.9014(5)	23.796(7)	12.9936(13)	14.2362(7)	6.5340(2)
α (°)	90	90	90	92.979(2)	90
β (°)	90	95.835	90	91.030(2)	90
γ (°)	90	90	90	103.331(2)	90
V (Å3)	970.7(1)	3127.0(2)	4975.3(8)	1182.77(10)	1376.7(1)
Z	4	2	4	2	1
ρ (g/cm3)	1.756	8.892	7.613	8.515	7.143
μ (mm–1)	2.030	2.334	1.982	2.413	2.102
F(000)	536	2103	2720	828	834
θ range (°)	2.909–25.983	2.412–26.143	2.250–26.369	2.136–26.110	3.464–26.839
limiting indices	–12 ≤ h ≤ 12	–21 ≤ h ≤ 21	–31 ≤ h ≤ 31	–8 ≤ h ≤ 8	–19 ≤ h ≤ 19
	–12 ≤ k ≤ 12	–9 ≤ k ≤ 9	–18 ≤ k ≤ 18	–15 ≤ k ≤ 15	–19 ≤ k ≤ 19
	–12 ≤ l ≤ 12	–29 ≤ l ≤ 29	–16 ≤ l ≤ 16	–17 ≤ l ≤ 17	–8 ≤ l ≤ 8
refl. collected/unique	34,609/637	87,272/6222	123,734/2707	32,224/4535	20,496/1969
Rint	0.0521	0.0658	0.0513	0.0384	0.0568
data/restraints/parameters	650/6/39	6222/12/479	2707/148/0	4708/0/331	1969/3/122
GOF on F2	1.190	1.059	1.089	1.115	1.253
final R indices	R1 = 0.0182	R1 = 0.0227	R1 = 0.0543	R1 = 0.0242	R1 = 0.0345
[I > 2σ(I)]	wR2 = 0.0477	wR2 = 0.0555	wR2 = 0.1635	wR2 = 0.0579	wR2 = 0.1007
R indices (all data)	R1 = 0.0192	R1 = 0.0273	R1 = 0.0568	R1 = 0.0257	R1 = 0.0361
	wR2 = 0.0488	wR2 = 0.0576	wR2 = 0.1671	wR2 = 0.0579	wR2 = 0.1014
largest peak and hole	0.220 to −0.454	1.286 to −0.782	2.965 to −5.237	2.209 to −2.058	2.538 to −1.154

Powder X-ray Diffraction

Purity of the bulk crystalline material was confirmed by powder X-ray diffraction using a Bruker D-5000 Advanced Powder Diffractometer equipped with Cu Kα radiation (λ = 1.5418 Å) and a LynxEye solid-state detector. Data were collected from 2 to 40° 2θ with a step size of 0.02° 2θ and a count time of 0.5 s/step. Predicted X-ray diffraction patterns were plotted using the Mercury Software version 3.1 and compared to the experimental data. Diffractograms of the experimental and calculated patterns can be found in the Supporting Information (Figures S6–S10).

Vibrational Spectroscopy

Solid-state compounds were analyzed by Fourier transform infrared (FTIR) and Raman spectroscopy. After confirming the bulk purity of the material, approximately 5 mg of the sample was mixed with KBr and pressed into translucent disks for analysis on a Nicolet Nexus 760 FTIR Spectrometer. Infrared spectra were collected from 500 to 4000 cm–1 with a resolution of 2 cm–1. Solid-state Raman spectra were collected on a SnRI High-Resolution Sierra 2.0 Raman spectrometer equipped with 785 nm laser energy and 2048 pixels TE-cooled CCD. Laser power was set to the maximum output value of 15 mW, and the system was configured to acquire data by the Orbital Raster Scanning mode, giving the highest achievable spectral resolution of 2 cm–1. Each sample was irradiated for an integration time of 60 s and automatically reiterated six times in multiacquisition mode. The average of the six Raman spectra collected for a sample is reported as the final Raman spectrum. Because of smaller yields, the solid-state Raman spectra for Co4U3 and Co2U1 were collected on a Renishaw inVia confocal Raman microscope with a Leica DM2700 series microscope using a 785 nm laser and a CCD detector. Each sample was isolated, mounted to a glass slide using double-sided tape, and loaded onto the sample stage. Laser focusing was performed by utilizing the confocal microscope, and the spectra were collected from 200 to 3000 cm–1. To accurately process the vibrational spectra, the background was subtracted, multiple peaks were fit using the peak analysis protocol with Gaussian functions, and all the fitting parameters converged with a chi-squared tolerance value of 10–14 in the OriginPro 9.1.0 (OriginLab, Northampton, MA) 64-bit software.[29]

DFT Methods

DFT calculations were used to gain a deeper understanding of the hydrogen bonding interactions that occur within these systems. Initial geometries for the DFT calculations were isolated molecular models generated based on the experimental crystal information files (CIFs) obtained from the structural analysis of the CoU compounds. Because of the disorder present in Co3U2_Cl, only the distances and geometries of central atoms were considered. This molecular approach allows us to systematically induce subtle structural changes in the coordination environment of the uranyl cation and incrementally increase the H-bonding present, which affords for a methodical analysis of the roles these features have on the vibrational spectroscopy. Full geometry optimization and vibrational analysis calculations were conducted using the Becke 3-parameter Lee-Yang-Par (B3-LYP) hybrid functional within the TURBOMOLE 7.2 software package and the default triple-zeta valence polarized (def-TZVP) basis set for U, Co, C, O, and N atoms.[30−33] SCF energy converged to at least 0.3 meV, and forces were converged to a minimum of 5 meV Å–1. The potential between system electros and U is accounted for using the small-core (60 core electrons) relativistic effective core potential (RECP) by Dolg and co-workers.[34] The isolated molecular models were embedded in the continuum solvent model COSMO with a dielectric constant (ε) of 78.54 to simulate aqueous solvent contributions to the electrostatics.[35]

To systematically explore how the ν1 and ν3 uranyl stretching modes are influenced by perturbations to the bonding environment, two DFT studies utilized fixed geometries paired with vibrational analysis. Both the free UO22+ cation and then the [UO2(CO3)3]4– complex were isolated from the Uco compounds and then allowed to relax to the energy minimized form. Then the U=O bonds in both complexes were varied in step sizes of 0.02 Å, and the vibrational analysis was performed to explore the impact on the position of the stretching modes. A second series of calculations evaluated the effects of counter-cation interactions on the U=O stretching bands by fixing the [Co(NH3)6]3+ positions relative to the [UO2(CO3)3]4–. Jaquet and Haeuseler reported similar methodologies as a means to evaluate simulate coordination environments that were associated with the crystallographic positions.[36] Unconstrained [UO2(CO3)3]4– + [Co(NH3)6]3+ calculations were first fully optimized, and subsequent molecular models fixed the positions of the [Co(NH3)6]3+ cations according to data obtained from the structural characterization of the solid UCo compounds. Vibrational modes were calculated in these specific environments to evaluate changes in the expected spectral features. For all calculations, SCF energy was converged to at least 0.3 meV, and forces were converged to at least 5 meV Å–1.

Results and Discussion

Structural Analysis

The first reported compound (Co_Cl_CO) does not contain U(VI) within the crystalline lattice but serves as a model compound for the spectral signals associated with [Co(NH3)6]3+ and CO32– ions within the solid phase (Figure SI1). This cobalt hexamine complex contains six Co–N bonds at distances of 1.963(2) to 1.964(2) Å, and the crystalline lattice contains a single chloride anion and a carbonate anion with C–O bond distances of 1.287(2) Å. Hydrogen bonding occurs between the H atoms on the amine groups and oxygen atoms on the carbonate anion with donor to acceptor (D-H···A) distances ranging from 2.848 to 3.004 Å.

Each of the other four compounds reported in this study contains the [UO2(CO3)3]4– coordination complex and exhibits subtle differences in U=O bond distances (Table ). In all cases, the U(VI) cation is strongly bound to two oxygen atoms to create the nearly linear dioxo cation (UO22+) with bond lengths ranging from 1.757(11) to 1.801(4) Å. Three of the four compounds contain symmetric U=O bond lengths within the uranyl moiety, and only in the case of Co3U2_Cl do we notice a slight asymmetry in the uranyl bonds, with a difference of 0.02 Å. Therefore, we did not observe significant asymmetry in the uranyl bond within any of the compounds presented herein. In all cases, three carbonate anions surround the uranyl cation through the equatorial plane in a bidentate coordination mode. Equatorial bond distances within this metal complex range from 2.393(4) to 2.461(6) Å among the four compounds. This leads to an overall hexagonal bipyramidal coordination geometry and results in the [UO2(CO3)3]4– species.

Table 2

Summary of Bond Distances for the [UO2(CO3)3]4– Complex in Co4U3, Co3U2_Cl, Co2U1_Cl, and Co2U1_CO and Literature Values for Bond Distances and Reported Vibrational Bands for the Uranyl Cation within Coordination Compounds Containing [UO2(CO3)3]4–

compound	U=O axial (Å)	U–O equatorial (Å)	reported UO22+ spectral modes (cm–1)	reference
Co4U3	1.796(3), 1.798(3)	2.412(3)–2.434(3)	805a	this work
	1.781(5), 1.793(5)	2.409(6)–2.461(6)
Co3U2_Cl	1.76(1), 1.78(1)	2.415(9)–2.423(7)	809a	this work
Co2U1_Cl	1.798(4), 1.801(4)	2.393(4)- 2.438(4)	807a	this work
Co2U1_CO3	1.771(8), 1.776(8)	2.435(5)–2.439(5)	806a	this work
NH4[(UO3)(CO3)3]	1.79(1)	2.44(1)–2.46(1)	831, 883	Graziani et al.,[37] Čejka, Novitskiy et al.[38,48]
[C(NH2)3]4[(UO2)(CO3)3]	1.78(1), 1.80(2)	2.440 (6)–2.451(2)	831(ν1), 892(ν3)	Fedoseev et al.,[58] Allen et al.[39]
[N(CH3)4]4[(UO3)(CO3)3] (H2O)8	1.803(3), 1.814(3)	2.418(3)–2.450(3)		Reed et al.[40]
Na4[UO2(CO3)3]	1.807(5), 1.814(5)	2.385(4)–2.427(3)	810, 816(ν1), 843(ν3)	Li et al.,[79] Čejka[38,41]
Na2Ca[(UO2)(CO3)3](H2O)6	1.81(2), 1.78(2)	2.41(1)–2.46(1)	833(ν1), 919 (ν3)	Coda et al.,[43] Driscoll et al.[80]
Na2Rb2[UO2(CO3)3]	1.779(8)	2.418(8)–2.433(6)		Kubatko and Burns[44]
Na6Mg[UO2(CO3)3]2(H2O)6	1.792(6)	2.392(7)–2.486(7)		Olds et al.[45]
Mg2[(UO2)(CO3)3](H2O)18	1.788(4), 1.785(4)	2.419(4)–2.457(4)	822, 875	Mayer and Mereiter,[42] Colmenero et al.,[46] Amayri et al.[47]
K4[(UO2)(CO3)3]	1.802	2.425(4)–2.434 (4)	815, 881	Anderson et al.,[81] Novitskiy et al.[48]
K2Ca3[(UO2)(CO3)3]2(H2O)8	1.781(4), 1.769(4)	2.423(4)–2.444(4)		Plášil et al.[49]
	1.791(4)	2.400(4)–2.454(4)
	1.775(4), 1.798(4)	2.404(4)–2.441(4)
Rb4[(UO2)(CO3)3]	1.79(1)	2.43(1)–2.45(1)	828, 877	Chernorukov et al.,[50] Gorbenko-Germanov and Zenkova[51]
Cs4[(UO2)(CO3)3]	1.806(4)	2.420(4) -2.435(4)	808, 877	Krivovichev, Burns, Gorbenko-Germanov and Zenkova[51,52]
Ca2[(UO2)(CO3)3](H2O)11	1.784(7), 1.774(7)	2.417(6)–2.448(7)	822, 902, 885, 883	Mereiter[53]
Ca9[(UO2)(CO3)3]4(CO3)(H2O)28	1.773(9), 1.779(9)	2.411(5)–2.481(5)		Kampf et al.[54]
	1.76(9), 1.773(9)	2.416(6)–2.457(5)
CaMg[(UO2)(CO3)3](H2O)12	1.777(3), 1.788(3)	2.412(3)–2.457(3)		Mereiter[55]

See Table for additional information on the uranyl stretching modes associated with these compounds.

Table 3

Observed Raman Frequencies for Solid-State Raman Spectra and the IR Frequencies of [Co(NH3)6]Cl(CO3) (Co_Cl_CO), [Co(NH3)6]4[UO2(CO3)3]3(H2O)11.67 (Co4U3), [Co(NH3)6]3[UO2(CO3)3]2Cl(H2O)7.5 (Co3U2_Cl), Co(NH3)6]2[UO2(CO3)3]Cl2 (Co2U1_Cl), and [Co(NH3)6]2[UO2(CO3)3]CO3(H2O)3 (Co2U1_CO) within the Spectral Window of Interest (500–1100 cm–1)

Co_Cl_CO3		Co4U3		Co3U2_Cl		Co2U1_Cl		Co2U1_CO3		assignment
R	IR	R	IR	R	IR	R	IR	R	IR
				1070						NH3 breathing
1053		1058		1065		1057		1056		ν4CO32– breathing
		1052	1054		1049	1051				ν4CO32– breathing + NH3 twist
		1041			1008	1041	1044		1043	ν4CO32– breathing + NH3 twist
					957					ν3UO22+ + NH3 twist
				950	948					ν3UO22+ + NH3 twist
			940							NH3 twist + ν3UO22+
			922							NH3 twist + ν3UO22+
			888	896	889			881	888	NH3 twist + ν3UO22+
			869		876		871			NH3 twist + ν3UO22+
			854	856					859	NH3 rocking
	832		845		832		833			NH3 rocking
									827	CO32–
		805		809		807		806		ν1UO22+ + CO3 wag
				753						NH3 breathing + ν1UO22+
		727		734		728		726		ν1UO22+ + ν2CO3
		719			721	720	719	720	722	ν2CO32–
					704					H2O libration
		684	690		692		696	693	691	ν2CO32– + NH3 twist
							685		668	H2O libration

Notable differences in the structural arrangement within the CoU compounds are variations in the molar ratio of the cobalt hexamine cation and the uranyl tricarbonate anion. In the case of Co4U3, we observe a Co:U ratio of 1.33 to give a formula of [Co(NH3)6]4[UO2(CO3)3]3, and no additional charge balancing anions or cations are necessary to neutralize the overall charge of the compound. Water molecules are located throughout the crystalline lattice of Co4U3 that results in the overall structural formula of [Co(NH3)6]4[UO2(CO3)3]3H2O11.67. Increasing the Co:U ratio to 1.5 in Co3U2_Cl results in the need to include additional charge balancing anions within the lattice, and we determined that the structure contained an additional Cl– anion located within four partially occupied sites. The Cl– anion is present in significant quantities because of the addition of the cobalt hexamine chloride reagent. Additional water molecules are again present, and the overall formula for Co3U2_Cl is [Co(NH3)6]3[UO2(CO3)3]2Cl(H2O)7.5. Both Co2U1_Cl and Co2U1_CO possessed a Co:U ratio of 2:1 and required an additional −2 charge compensation to create neutrality. The negative charge is achieved through the incorporation of either two Cl– (Co2U1_Cl) or one CO32– anion (Co2U1_CO) and results in overall formulas of [Co(NH3)6]2[UO2(CO3)3]Cl2 and [Co(NH3)6]2[UO2(CO3)3]CO3(H2O)3, respectively.

Comparisons between the CoU compounds and other uranyl tricarbonate phases reported in the literature indicated similarities in relative ratios of charge balancing constituents and bond distances (Table ). The uranyl (U=O) bond lengths in this class of compounds ranged from 1.73(4) to 1.85(4) Å, and the U–O equatorial distances occurred between 2.38(4) to 2.48(4) Å. All CoU compounds exhibit bond distances within this range. A majority of the uranyl tricarbonate compounds exhibited symmetric U=O bond lengths within the uranyl moiety, except in the case of the mineral Paddlewheelite (MgCa5Cu2[(UO2)(CO3)3]4(H2O)33).[45] In this case, there is significant asymmetry in the uranyl bond that ranges from 0.02 to 0.07 Å. The largest asymmetry in the U=O bond (0.07 Å) within Paddlewheelite occurs in the region where there are significant differences in the intermolecular interactions that occur between the oxo groups and neighboring cations and hydrogen bond donors. However, most of the previously reported uranyl bond asymmetry is similar to the value observed within the Co3U2_Cl coordination compound (0.02 Å).

Intermolecular interactions within the CoU compounds occur through charge-assisted hydrogen bonding that takes place between the cobalt hexamine donors and the uranyl oxo acceptor groups. An extensive hydrogen bonding network is noted within the solid-state compounds, and significant differences are observed based upon the arrangement of the [Co(NH3)6]3+ cations (Figure ). Co4U3 displays symmetric bifurcated hydrogen bonding between the uranyl oxo groups and the [Co(NH3)6]3+ cations (Figure a). Hydrogen bonding distances are relatively long, with distances ranging from 2.96 to 3.01 Å. Stronger interactions occur between the carbonate anions and water molecules located in the interstitial region (2.734–2.809 Å). The uranyl oxo groups in compound Co3U2_Cl engage in asymmetric H-bonding interactions because of the arrangement of the counterions within the layers. We note that in this compound, the O1 atom acts as a H bond acceptor to two different donors (N6) at a donor to acceptor distance (D–H···A) of 2.96 Å (Figure b). Donor (N5) to acceptor (O2) distances for the hydrogen bonding interactions occurring at the second oxo group are similar in distance (2.94 Å), but again the uranyl bond distance is asymmetric. Hydrogen bonding in compound Co2U1_CO follows the symmetric nature of the [Co(NH3)6]3+ cations and exhibits interactions to oxygen acceptors on the uranyl moiety and carbonate anion (Figure d). Each oxo group (O1 and O2) interacts in a symmetric fashion to hydrogen atoms on the cobalt hexamine cation with D–H···A distances of 2.97 and 3.01 Å, respectively. The arrangement of the [Co(NH3)6]3+ cations around the oxo groups leads to trifurcated H-bonding to each layer of [UO2(CO3)3]4–. In addition, the carbonate anions also participate in H-bonding with the oxygen atoms (O3, O4, and O5) linked to the U(VI) metal center. Again, the uranyl bond distance is symmetric, and each O atom can interact with H atoms located above and below the uranyl tricarbonate complex with D–H···A distances ranging from 2.95 to 3.00 Å. Co2U1_Cl (Figure c) shows a similar hydrogen bonding network to Co2U1_CO, with slight differences in the donor to acceptor distances (2.893–3.28 Å).

Figure 2

Hydrogen bonding for the cobalt hexamine to the uranyl tricarbonate compounds, showing differences in the arrangement of the [Co(NH3)6]3+ counterions and hydrogen bonding networks for Co4U3, Co3U2_Cl, Co2U1_Cl, and Co2U1_CO3. The U, Co, Cl, O, N, and C atoms are depicted as yellow, dark blue, green, red, light blue, and black spheres, respectively. The H atoms have been removed for clarity. Hydrogen bonding is illustrated using dashed red lines.

Synthetic compounds containing the uranyl tricarbonate anion and hydrogen bond donors have been previously reported, but there is little evidence of this type of interaction occurring with the uranyl oxo groups in these materials. The ammonium cation can cocrystallize with the [UO2(CO3)3]4– complex and engages in medium strength H-bonding based upon Jeffrey’s classification (D–H···A = 2.85–3.22 Å).[37,56] All the NH4+ cations were located either below and above the equatorial plane of the uranyl cation but only participate in H-bonding with the O atoms of the bound CO3– anions. A H-bonding network within the tetramethylammonium uranyl tricarbonate compound has also been delineated, and again the uranyl oxo groups do not participate in additional intermolecular interactions with the hydrogen bond donors.[57] Within the guanidinium system, the trimeric species [(UO2)3(CO3)6]6– was the major species isolated, but [C(NH2)3]4[(UO2)(CO3)3] has also been reported by Fedosseev and co-workers.[58] In both cases, no hydrogen bonds were observed between the guanidinium cations and the uranyl oxo groups.

Evaluation of the uranyl carbonate literature suggests that the CoU compounds are unique in that the uranyl oxo groups do participate in the hydrogen bonding network created by the cobalt hexamine cation. As mentioned in the Introduction section, the cobalt hexamine cation was chosen specifically to engage the uranyl oxo groups because of the high charge density associated with the complex. We observe this to occur within the compounds presented herein, and the H-bonds can all be classified using the categories delineated by Jeffrey as medium strength.[59,60] The [Co(NH3)6]3+ cation has also been previously reported to crystallize other uranyl coordination complexes, including substituted malonato and tetrahydroxide complexes.[61−63] Extensive hydrogen bonding networks occur within both systems that include interactions between the amine and uranyl oxo groups; however, the hydrogen bonding strength is much weaker for the malonato and tetrahydroxide complexes (average D–H···A distances = 3.2(1) Å). Clark et al. discussed the H-bonding interactions in relation to the uranyl oxo distances, pointing out that the shortest bond (1.802(6) Å) showed only one interaction to the [Co(NH3)6]3+ unit and the longest U=O bond at 1.835(5) Å possessed multiple hydrogen bonds.[63]

Within the evaluation of [Co(NH3)6]2[UO2(OH)4)]3H2O, Clark et al. also noted that there was a 10 cm–1 difference between uranyl symmetric stretching (ν1) mode of the solid and that of the related solution phase. It was suggested that this difference could be due to variability in the number of hydro ligands attached the uranyl cation or the impact of hydrogen bonding within the solid-state material. Additional experimental and computational analysis has indicated that the uranyl tetrahydroxide is the dominant species under alkaline conditions, so the impact of hydrogen bonding is the likely explanation of this spectral variability.[64,65] Thus, we turn to vibrational analysis to further identify the impact of hydrogen bonding within the CoU materials.

For this work, we will focus specifically on the uranyl symmetric stretch (ν1) and asymmetric stretch (ν3) associated with the uranyl cation. If one considers the uranyl point group symmetry to be D∞, then the symmetric and asymmetric stretches are predicted to be Raman- and IR-active, respectively. However, U=O bond perturbation can result in lower symmetry of the uranyl cation through either bond asymmetry (C∞) or bending (C2) that would result in activation of both the ν1 and ν3 bands in the Raman and IR spectra. Additional combination modes with the carbonate ligands and the cooperative nature of the hydrogen bonding network can also influence the spectral signals.[46] To focus specifically on these issues, we evaluated the spectral window 700–1100 cm–1 to capture major features of the uranyl, carbonate, and cobalt hexamine components (Figure and Table ). Assignments were determined based upon the DFT spectral band analysis. Additional spectral features associated with the [Co(NH3)6]3+ are observed between 300 and 500 cm–1, and full spectra are provided in the Supporting Information (Figure S8).

Figure 3

Solid-state vibrational spectra of (a) Co4U3, (b) Co3U2_Cl, (c) Co, and (d) Co2U1_CO. IR spectra are on top, while Raman spectra are located on the bottom for each sample.

Variability in the spectral signals is observed in the CoU compounds that are associated with differences in the hydrogen bonding network. Co_Cl_CO only exhibits one band in the spectral window of interest (1053 cm–1) that corresponds to the ν4 CO32– breathing mode. In the presence of UO22+, multiple bands are observed which correspond to concerted motion. Some of these bands (726–734 cm–1) are associated with concerted motions between the uranyl cation and the bound carbonate anion. In addition, the hydrogen bonding interactions between the uranyl oxo groups and the cobalt hexamine cation lead to a concerted uranyl stretching with NH3 twisting motions (753, ∼805, 856, 881, and 891 cm–1). Similar hydrogen bonding interactions the amine group and the carbonate anion also exist and lead to multiple bands associated with the CO32– breathing modes between 1041 and 1065 cm–1.

One notable band in Co3U2_Cl is located at 950 cm–1 and could be assigned to the activated ν3 asymmetric stretching vibration for the uranyl cation, twisting of the amine group associated with the [Co(NH3)6]3+ cation, or a combination of the two modes together. This modified vibrational signal does pair with the subtle asymmetry of the uranyl bond length noted in this compound that may lower the overall point group symmetry and lead to an observable peak in the Raman spectra. However, a difference of only 0.02 Å is quite small, and this asymmetry alone may not account for this specific band. Thus, it is more likely that it is associated with the concerted NH3 twisting and asymmetric stretching of the uranyl that gives rise to the band in the spectra.

Infrared spectroscopy was also performed for all compounds and allows us to confirm the position of the ν3 bands. Multiple bands are present in the 870–960 cm–1 region that correspond to concerted motions between the amine and the uranyl oxo groups. Both Co2U1 compounds contain fewer bands in this region and may be related to the trifurcated, symmetric bonding that exists between the cobalt hexamine and uranyl oxo groups. Co4U3 and Co3U2_Cl contain six and five modes within that region, respectively, that are related to NH3 twisting and the asymmetric stretch of the uranyl cation. It is notable that there is a band within the IR spectra of Co3U2_Cl at 948 cm–1 that corresponds to the band at 950 cm–1 within the Raman spectrum; however, there is no evidence of activation of the ν1 band within the IR spectrum at 809 cm–1. This suggests that it is not bond asymmetry driving the resulting spectral bands, but the interaction between the hydrogen donor and the uranyl oxo acceptor.

Comparing these values to previous literature results is difficult because the ν1 and ν3 bands reported may not be assigned correctly. The uranyl symmetric stretching bands for Cs4[UO2(CO3)3] and Na4[UO2(CO3)3] possess similar values to the CoU compounds, but other compounds range from 815 to 831 cm–1 (Table ).[38,41,51,52] Similarly, the ν3 band has been reported with values ranging from 843 to 912 cm–1 for uranyl tricarbonate species. Colmenero et al. performed DFT calculations to assess the infrared active modes of the mineral Bayerite (Mg2[UO2(CO3)3] • 18 H2O) and found that the band at 872 cm–1 could be assigned to a combination of the uranyl antisymmetric stretching vibration and water rocking modes.[46] In addition, theoretical bands at 837 and 827 cm–1 are ascribed to ν3 stretching vibrations, carbonate out of plane bending vibrations, and twisting motions. Thus, even the presence of water within the tricarbonate system can lead to difficulties in identifying the spectral modes in these materials.

Evidence of vibrational coupling or combination modes, including the uranyl O=U=O stretch, is not without precedent, particularly with solids that contain strong intermolecular interactions. Cahill and co-workers suggested a combination mode of anharmonic resonance coupling between the benzoate ligands and the uranyl “yl” stretch of their halogenated benzoic acid and uranyl crystalline materials.[66] In addition, Anderson and co-workers evaluated the impact on interstitial water content within the schoepite mineral phases (UO3•nH2O) on the resulting spectral features.[67] Hydrogen bonding effects were found to strongly influence the symmetric stretch of each unique uranyl moiety enough to give rise to multiple stretching modes in the Raman spectra within a relatively large spectral window (810–880 cm–1). As noted earlier, Colmenero et al. observed that multiple uranyl features in Bayerite are coupled with water librations, water twists, and carbonate bending modes.[46]

Because our systems display significant differences in the vibrational spectra, we considered many approaches in evaluating the signals. When we base our vibrational analysis on the simple D analysis of the uranyl cation, this change in the spectra could be related to lowering of the symmetry and inducing activation. We can also consider coupled vibrational motions that can occur with the specific hydrogen bonding networks in the material as a source of the varied spectral signals. In a reductionist approach to understand the real system, we can first evaluate the simpler models and then build up the complexity to include the additional interactions, DFT calculations are well suited for this approach, and in the next section, we utilize this methodology to explore bond asymmetry without additional structural contributions and in varied coordination environments to evaluate the impacts on the vibrational modes. The first simplified set of calculations is used to delineate the impact of uranyl bond asymmetry on the vibrational features for a free uranyl cation and then for a uranyl tricarbonate species. Following those studies, we further add hydrogen bonding to the system to compare the influence of these intermolecular interactions on the resulting spectral features.

DFT Analysis

Forced UO2 Bond Asymmetry

DFT calculations are used to further evaluate the extent that uranyl bond asymmetry can impact the position of the symmetric and asymmetric stretching of the uranyl cation. To begin, we optimized the structure of a single UO22+ cation in the absence of additional counterions. This resulted in two equivalent U=O bond lengths of 1.76 Å and ν1 and ν3 modes of 876 and 937 cm–1, respectively, which are within range of the previously reported computational results for the uranyl cation.[68,69] We then varied the length of one U=O bond by 0.02 Å increments from 1.66 to 1.86 Å and fixed the second U=O bond to the optimized length of 1.76 Å. For each of the UO22+ structures described above, a set of single-point energy calculations are performed in which the vibrational modes were calculated. The calculated values for the resulting symmetric and asymmetric stretching vibrations are listed in Table .

Table 4

DFT-Computed Vibrational Modes for the UO22+ Unit, Where One U=O Bond Length Is Systematically Increased by 0.02 Å from 1.66 to 1.86 Å, While the Other Is Held Constant at 1.76 Åa

ΔU=O length (Å)	ν1 (cm–1)	ν3 (cm–1)	ν1/ν3	kF (mdyn/Å)	k12 (mdyn/Å)
–0.10	908	1177	1.30	9.64	–1.89
–0.08	906	1120	1.24	9.08	–1.34
–0.06	903	1066	1.18	8.56	–0.88
–0.04	900	1017	1.13	8.11	–0.48
–0.02	893	973	1.09	7.69	–0.17
0 (1.76 Å)	876	937	1.07	7.26	–0.03
+0.02	844	922	1.09	6.89	–0.17
+0.04	814	948	1.16	6.85	–0.61
+0.06	761	913	1.19	6.19	–0.73
+0.08	720	911	1.27	5.89	–1.00
+0.10	681	910	1.33	5.62	–1.25

Boldface is used to highlight the structure where the U=O bond lengths are equal at 1.76 Å.

We compare how the computed vibrational modes change as a function of bond elongation and contraction, bringing the uranyl oxo atoms closer together or further apart. When one U=O bond is elongated by 0.10 Å, the change in the ν3 (+195 cm–1) is much greater than the change in the ν1 (+32 cm–1). For both vibrational modes, there is an observed red shift. Alternatively, the contraction of one U=O bond results in a more significant change in the ν1 (−195 cm–1) compared to the ν3 (−27 cm–1); both vibrational modes exhibit a red shift as a result of asymmetric bond contraction. In general, asymmetric bond elongation results in an increase in the value of the vibrational frequencies, while bond contraction results in a decrease in the value of the vibrational frequency. When the U=O bond lengths are equivalent at 1.76 Å, the difference between the ν1 and ν3 vibrational frequencies is at a minimum. As the U=O bond length difference increases, the difference between the ν1 and ν3 vibrational frequencies increases.

The ν1/ν3 ratio is reported and compared to previous results where it is used to evaluate the impact of the interaction force constant within the uranyl bond.[19,20,38,70] Vibrational modes associated with the uranyl cation are also related to the force constant (k1) and the interaction force constant (k12). If we consider a simple valence force field and assume harmonic vibrational for the linear ion, then the interaction force constant can be omitted and the relationship between ν1 and ν3 can be written as:where MO and MU represent the mass of the O and U atoms, respectively. This leads to a ν3/ν1 of 1.065, which is identical to that calculated for our symmetric uranyl bond (1.07). When the interaction force constant is included, then the ν3/ν1 ratio will increase or decrease depending on the overall sign of the k12. In the case of our bond asymmetry, we note that ν3/ν1 increases, which indicates that the interaction force constant decreases. This can be observed in the calculated k12 values, where more negative values are obtained when one bond is either lengthened or shorted to induce asymmetry. The trend is different for k1, where the value is dependent on the length of the bond, with shorter distances related to stronger force constants.

Schnaars and Wilson evaluated force constants for a series of uranyl tetrachloride compounds, and these compare well to the results associated with our computed uranyl cation.[71,72] At a symmetric bond distance of 1.76 Å, the theoretical k1 was calculated at 7.26 mdyn/Å and decreased to 5.62 mdyn/Å with a bond elongation of 0.1 Å. This is well within the range that has been experimentally observed within the tetrachloride system (6.39–6.74 mdyn/Å).[73] Additionally, the k12 was observed between −0.10 and −0.53 mdyn/Å, and this matches well with a small negative value that was obtained from DFT analysis. It is interesting to note that the ν3/ν1 ratio for the uranyl tetrachloride compounds ranges from 1.08 to 1.10, which is slightly higher than the value assumed for minimal contribution of the k12 (1.065). This suggests that the small contribution from the interaction force can be observed by utilizing the vibrational band ratios.

A similar approach was followed for the [UO2(CO3)3]4– structure. The initial coordinates were obtained from the experimental crystal structure. The [UO2(CO3)3]4– was first subjected to geometry optimization, where the U=O bonds optimized to equivalent lengths of 1.82 Å. Visualization of the vibrational modes for the optimized [UO2(CO3)3]4– structure displayed two ν1 modes at 789 (ν1a) and 717 (ν1b) cm–1 and a ν3 mode at 830 cm–1. The two ν1 modes display coupling of the UO2 symmetric stretch and the ν2 wagging motion of the bound CO32– group. The ν1a band displays symmetric uranyl contraction coupled to an inward ν2 CO2 wag, whereas ν1b consists of a uranyl contraction coinciding with an outward ν2 wag motion of the bound CO32– ligands.

To investigate the effects of U=O bond asymmetry on the vibrational modes in [UO2(CO3)3]4–, a series of calculations at fixed geometry was carried out. The interatomic separation of one of the U=O was varied by 0.02 Å from 1.82 to 1.72 Å. We chose 1.82 Å as the longest distance that is observed for the [UO2(CO3)3]4– with symmetric U=O bond lengths. These calculations allow for the comparison of the change in the vibrational modes with the presence of ligands in the equatorial plane, but without the interaction of additional species.

The ν1 and ν3 vibrational frequencies were monitored as the extent of the bond asymmetry increased along the series (Table ). Comparing the ν1a and ν1b modes, we observe an overall red shift of 9 or 11 cm–1, respectively, when the U=O lengths differ by 0.1 Å. When the asymmetric U=O bond contraction for the [UO2(CO3)3]4– complex differs by 0.10 Å, there is a more significant red shift in the ν3 (+188 cm–1) than for the ν1 (+9 cm–1), which is similar to the free UO22+ system.

We evaluated the ratio of the symmetric and asymmetric bands to provide further insight into the system (Table ). Inducing U=O bond asymmetry of the uranyl tricarbonate complex does not change the ν1a/ν1b ratio, which remains constant at 1.10 throughout the entire range of tested asymmetry values. Both the ν3/ν1a and ν3/ν1b ratios increase with increasing U=O bond asymmetry. For ν3/ν1a, the ratio is similar to the harmonic model (1.05) when the bonds are both at 1.82 Å and increase to 1.26 when the bond difference is −0.1 Å. For the ν3/ν1b ratio, it begins with a larger value (1.16) because of a larger energy difference between the modes and increases to 1.40 with induced asymmetry.

Table 5

DFT-Computed Vibrational Modes for the [UO2(CO3)3]4– Unit, Where One U=O Bond Length Is Systematically Decreased by 0.02 Å and the Other Is Held Constant at 1.82 Åa

ΔU=O length (Å)	ν1-a, ν1-b (cm–1)	Δν1	ν1-a/ν1-b	ν3 (cm–1)	ν3/ν1-a	ν3/ν1-b
–0.10	802, 726	76	1.10	1018	1.26	1.40
–0.08	801, 725	76	1.10	973	1.21	1.34
–0.06	800, 724	76	1.11	930	1.16	1.28
–0.04	803, 733	70	1.10	914	1.14	1.24
–0.02	795, 720	75	1.10	855	1.08	1.19
0 (1.82 Å)	789, 717	72	1.10	830	1.05	1.16

Boldface is used to highlight the structure where the U=O bond lengths are equal at 1.82 Å.

Counter-Cation Interactions

The next series of DFT calculations were performed on systems that varied the position and number of the [Co(NH3)6]3+ cation around the uranyl carbonate complex. Isolated uranyl carbonate (UC) was optimized, and then either one (UC1-A, UC1-B, or UC1-C) or two (UC2-D and UC2-E) [Co(NH3)6]3+ cations were placed around the uranyl tricarbonate anions in locations obtained from the crystallographic information files (Figure ). The geometry of the isolated [UO2(CO3)3]4– complex (denoted as UC) was optimized and included here for comparison to the models that contained the cobalt hexamine counterion. As previously mentioned, bond lengths in the isolated [UO2(CO3)3]4– molecular complex are symmetric, with U=O bonds of 1.82 Å. Bonding to the carbonate anions leads to U–Oc equatorial distances of 2.45 Å and U–C interatomic values of 2.91 Å. These theoretical bond distance values agree well with other values observed by Reeder et al. and Ikeda et al.[74,75]Figure also depicts the interactions between the [Co(NH3)6]3+ cation with asymmetric uranyl bonds. In UC1-A, the cobalt hexamine interacts with the O atoms associated with the shorter U=O bond (1.75 Å) with a D–H···A distance of 2.91 Å. Alternatively, UC-B shows the interaction of the [Co(NH3)6]3+ cation with the oxo group of the longer U=O distance (1.80 Å) and D–H···A distance of 3.51 Å. Geometry optimization of either UC1-A or UC1-B results in the structure shown in UC1-C, where the cobalt hexamine has moved from its position near the axial oxo groups to the equatorial plane, where it can engage in H-bonding interactions with the carbonate anions. The optimized structure has identical bond U=O bond lengths (1.82 Å) to that observed for the isolated uranyl tricarbonate anion. UC2-D and UC2-E depict the interaction between the uranyl tricarbonate and two [Co(NH3)6]3+ cations at different positions around the metal complex. In both cases, the uranyl bond is modeled as asymmetric to further understand the impact of the intermolecular H-bonding on the vibrational modes within these complexes. UC2-D is modeled with the two cobalt hexamine cations engaged in interactions with the longer U=O (1.80 Å), with D–H···A distances of 3.63 Å. UC2-E is modeled with the one cobalt hexamine cations engaged in interactions with each uranyl oxo, with D–H···A distances of 3.63 Å to the longer U=O (1.80 Å) and D–H···A of 2.91 Å to the shorter U=O (1.75 Å).

Figure 4

Ball and stick representations of the molecular models used in the DFT calculations of UO2(CO3)34– interacting with [Co(NH3)6]3+ cations, where the number and position of the cation are altered. Models UC1-A, -B, and -C contain one [Co(NH3)6]3+ cation, whereas UC2-D and -E contain two counter cations. Uranium, oxygen, carbon, nitrogen, cobalt, and hydrogen are depicted as yellow, red, gray, dark blue, light blue, and white spheres, respectively.

The symmetric and asymmetric uranyl stretching modes were identified in all three structures, and associated frequency values are reported in Table . Only the symmetric/asymmetric modes obtained from structures that exhibit atomic displacements greater than 0.01 Å for the uranyl are reported herein. As mentioned previously, the isolated [UO2(CO3)3]4– unit (UC) has two ν1 modes at 717 and 789 cm–1 and a ν3 mode at 830 cm–1, where the presence of multiple ν1 modes is related to coupling between the uranyl stretching modes and the carbonate motions. Each mode will be described in detail for a given structure to show how the uranyl modes shift as a function of counter cation position and number.

Table 6

DFT-Calculated Uranyl Active Vibrational Modes for UC Structures Containing Cobalt Hexamine Cation(s) and an Isolated Molecular Complex for Comparisona

structure	U=O lengths (Å)	ν1-a, ν1-b (cm–1)	Δν1	ν3 (cm–1)	ν1-a/ν1-b	ν3/ν1-a	ν3/ν1-b
isolated UC	1.82, 1.82	789, 717	72	830	1.10	1.05	1.15
isolated UC/asymmetric	1.82, 1.78	800, 724	76	930	1.11	1.16	1.28
UC1-A	1.80, 1.75(O···H)	861, 738	123	960	1.17	1.15	1.30
UC1-B	1.80(O···H), 1.75	868, 716	145	986	1.21	1.13	1.37
UC1-C	1.82, 1.82	803, 724	79	835	1.11	1.04	1.15
UC2-D	1.80(O···H), 1.75	871, 738	133	960	1.18	1.10	1.30
UC2-E	1.80(O···H), 1.75(O···H)	864, 736	128	961	1.17	1.11	1.30

Uranyl bond lengths engaged in a H-bond are denoted with (O···H) following the value.

With the presence of hydrogen bonding, all ν1 and ν3 uranyl symmetric stretching features are shifted to higher wavenumbers compared to the isolated forms. The UC1-A interaction geometry has ν1a and ν1b modes located at 738 and 861 cm–1 and one major ν3 mode at 960 cm–1. With the presence of the [Co(NH3)6]3+ cation, additional concerted motions are noted as the mode at 861 cm–1 shows coupling between the uranyl stretching mode, inward ν2 CO32– wag, and the twisting of the cobalt hexamine cation. Similarly, the band at 960 cm–1 exhibits coupled uranyl and cobalt hexamine motions. Structure UC1-B indicates that the ν1b mode is predicted to be at 716 cm–1 and ν1a band located at 868 cm–1 is coupled to the cobalt hexamine through hydrogen bonding.

Structure UC1-C has equivalent U=O bond lengths, and the vibrational modes can be compared to the [(UO2)(CO3)3]4– complex. Like the isolated molecule, the symmetric ν1 modes in UC1-C occur at 724 and 803 cm–1. Both modes exhibit identical motions of coupled uranyl ν1 and carbonate ν2 modes with minimal hydrogen-bonding contributions between the Oyl atoms and the cobalt hexamine that are now located along the equatorial plane. The asymmetric ν3 uranyl mode occurs at 835 cm–1 and contributions from the carbonate ν3 mode.

Comparatively, structures UC1-A and UC1-B have more similar vibrational modes to one another than to either structure UC1-C or the isolated UC. These two structures share the same unequal uranyl axial bond lengths (1.75 and 1.80 Å) and partake in hydrogen-bonding interactions with the counter cation. Structures UC1-A and UC1-B have ν1 modes at 861 and 868 cm–1, respectively, which are not present in the Isolated UC or Structure UC1-C. Alternatively, both optimized geometries of structure UC1-C and the isolated UC model have the same uranyl axial bond lengths (1.82 Å) and similar vibrational modes. For example, both structure UC1-C and the isolated UC models have a symmetric mode near or at 717 cm–1. Additionally, the isolated UC complex has an asymmetric ν3 mode at 830 cm–1, which resemble the mode for structure UC1-C at 835 cm–1. The asymmetric modes for structures UC1-A and UC1-B are blue-shifted to frequencies greater than 960 cm–1 and are not present in the 830–850 cm–1 range.

The ν1 modes vibrational modes for structures UC2-D and E are analogous to the bands associated with the UC1-A. Both the ν 1b modes (738 and 736 cm–1) can be linked to uranyl bond displacements coupled to carbonate ν2 motions, with minor hydrogen breathing motions from the cobalt hexamine cations. UC2-D has a second ν1 mode at 871 cm–1, which is blue-shifted relative to the same band in UC2-E at 864 cm–1. Both structures display ν3 modes at 960/961 cm–1, in which the uranyl oxo engages in hydrogen-bonding interactions with cobalt hexamine groups.

From the vibrational bands, we can evaluate the impact of hydrogen bonding on the ν1 and ν3 ratios. Unlike the isolated system where the ν1a/ν1b ratio was constant at 1.10, the series including the cobalt hexamine shows a range of higher values 1.17–1.21. This may be caused by an inequivalent hydrogen bonding network that contributes to the combination mode (oxo groups and carbonate anions), which impacts these bands differently, manifesting in nonmonatomic changes in the ratio. The ν3/ν1a and the ν3/ν1b ratios are larger than those observed for the isolated system with symmetric U=O bonds, but these values are within the region for isolated [UO2(CO3)]4– complexes with induced bond asymmetry. This suggests that the impact of the hydrogen bonding network combined with the asymmetry can lead to larger values for this ratio which would need to be considered if using ratios to confirm vibrational assignment of uranyl modes.

Comparison of Theoretical and Experimental Results

If we compare the DFT vibrational analysis to the spectroscopic measurements of the CoU compounds, we find that we can utilize band ratios to help confirm our spectral assignment. Previous combined experimental studies with computational efforts found similar success in assignments of challenging vibrational spectra.[76−78] The ν1-a/ν1-b ratios for all CoU compounds were calculated to be 1.10–1.11, which is the same as what is expected for a symmetric uranyl bond. There are multiple concerted motions that contain the ν1-a or ν3 bands, so determining the accurate ratio is not straightforward. This speaks to the impact that the hydrogen bonding network has on the asymmetric stretch (ν3). DFT calculations demonstrated that hydrogen bonding between the uranyl oxo and a hydrogen donor can cause a blue shift in the ν1 and ν3 uranyl stretching bands. We do not see evidence of this perturbation for the symmetric ν1 stretching bands but do see significant concerted motions of the NH3 and uranyl oxo groups to create multiple ν3 asymmetric modes that are shifted to higher wavenumbers, with respect to the isolated complex. This demonstrates that the Raman spectra are less impacted by the hydrogen bonding network than the infrared spectra in our system.

Conclusions

The uranyl tricarbonate anion was crystallized with the cobalt hexamine cation to form four different solid-state materials ([Co(NH3)6]4[UO2(CO3)3]3H2O11.67 (Co4U3), [Co(NH3)6]3[UO2(CO3)3]2Cl H2O7.5 (Co3U2_Cl), [Co(NH3)6]2[UO2(CO3)3]Cl2 (Co2U_Cl), and [Co(NH3)6]2[UO2(CO3)3]CO3 (Co2U_CO). Structural analysis of the compounds revealed no significant perturbation of the uranyl bonds but displayed differences in the hydrogen bonding network within these compounds. Raman and infrared spectroscopy revealed different spectral features that were related to combination modes associated with the hydrogen bonding networks. DFT calculations were performed to evaluate the impacts of bond perturbation compared to changes in the hydrogen bonding network. Overall, bond perturbation led to an increased ν3/ν1 ratio, indicating that the interaction force constant (k12) should be considered for this system. Addition of hydrogen bonding to the uranyl oxo groups led to a blue shift in the vibrational features, and these interactions impact the ν3 band more significantly than the ν1.

Understanding the complex spectral features related to hydrogen bonding networks in solid and solution U(VI) phases will provide insights into uranyl bond modification, improve our knowledge of uranyl speciation in aqueous solutions, and enhance the use of these methodologies in sensing and nuclear forensics capabilities. This study demonstrates that combination bands within uranyl solids can lead to significant complexity within the vibrational spectra and should be further evaluated to provide a detailed understanding of these features. Additional studies should also focus on providing relationships between spectral bands that can be further utilized to identify specific U(VI) phases and a more descriptive understanding of the interactions that take place between the actinyl oxo groups and neighboring molecules and ions.