Infrared Spectroscopy of Discrete Uranyl Anion Complexes
Gary S. Groenewold,* Anita K. Gianotto, and Michael E. McIlwain
Idaho National Laboratory, Idaho Falls, Idaho
Michael J. Van Stipdonk* and Michael Kullman
Wichita State UniVersity, Wichita, Kansas
David T. Moore, Nick Polfer, and Jos Oomens
FOM Instituut Voor Plasmafysica Rijnhuizen, Nieuwegein, The Netherlands
Ivan Infante and Lucas Visscher
Vrije UniVersiteit Amsterdam, The Netherlands
DEN/DRCP/SCPS, CEA Marcoule, 30207, Bagnols-sur-Ce `ze cedex, France
Wibe A. de Jong
Pacific Northwest National Laboratory, Richland, Washington
ReceiVed: September 11, 2007; In Final Form: October 25, 2007
The Free-Electron Laser for Infrared Experiments (FELIX) was used to study the wavelength-resolved multiple
photon photodissociation of discrete, gas-phase uranyl (UO22+) complexes containing a single anionic ligand
(A), with or without ligated solvent molecules (S). The uranyl antisymmetric and symmetric stretching
frequencies were measured for complexes with general formula [UO2A(S)n]+, where A was hydroxide,
methoxide, or acetate; S was water, ammonia, acetone, or acetonitrile; and n ) 0-3. The values for the
antisymmetric stretching frequency for uranyl ligated with only an anion ([UO2A]+) were as low or lower
than measurements for [UO2]2+ligated with as many as five strong neutral donor ligands and are comparable
to solution-phase values. This result was surprising because initial DFT calculations predicted values that
were 30-40 cm-1higher, consistent with intuition but not with the data. Modification of the basis sets and
use of alternative functionals improved computational accuracy for the methoxide and acetate complexes,
but calculated values for the hydroxide were greater than the measurement regardless of the computational
method used. Attachment of a neutral donor ligand S to [UO2A]+produced [UO2AS]+, which produced only
very modest changes to the uranyl antisymmetric stretch frequency, and did not universally shift the frequency
to lower values. DFT calculations for [UO2AS]+were in accord with trends in the data and showed that
attachment of the solvent was accommodated by weakening of the U-anion bond as well as the uranyl. When
uranyl frequencies were compared for [UO2AS]+species having different solvent neutrals, values decreased
with increasing neutral nucleophilicity.
The chemical behavior of uranium in general, and the linear
uranyl dication [UO2]2+in particular, are diverse on account of
the relative ease of redox processes,1and the availability of f
and d orbitals2-4for complex formation. The latter has a
profound affect on the solubility of the element.5,6In solution,
[UO2]2+is the dominant species,6-8where it plays an important
role in heavy element separations9and in mobility of the element
in the environment.10At low pH, [UO2]2+exists as the solvated
dication in solution with weakly complexing anions.10Hydroly-
sis11,12at higher solution pH values, or the presence of more
strongly coordinating anions, produces uranyl complexes co-
ordinated by one or more anionic ligands.8,13
The chemical diversity of species has motivated research in
vibrational spectroscopy and computational chemistry to un-
derstand the coordination and nature of bonding in uranyl
complexes containing different ligands because these factors
have reactivity and stability implications.14,15Infrared and
Raman spectroscopy studies of [UO2]2+have shown that the
respective antisymmetric (ν3) and symmetric (ν1) stretching
frequencies16act as convenient “thermometers” for gauging the
electron-donating capability of the equatorial ligand field,
because the frequencies are strongly correlated with the
coordination environment. Nucleophilic ligands in the coordina-
tion sphere donate electron density to the cationic metal center,
and this spills over into the π*-antibonding orbitals of the uranyl
ion to cause a concomitant decrease in the associated ν1and ν3
* Corresponding authors. E-mail for G.S.G.: email@example.com.
J. Phys. Chem. A 2008, 112, 508-521
10.1021/jp077309q CCC: $40.75© 2008 American Chemical Society
Published on Web 12/29/2007
frequencies. Increased electron density at the uranium metal
center can be effected by attachment of more donor ligands,17
or by increasing the nucleophilicity of the ligands.18,19Generally,
for a modestly complexing solution environment, ν3values near
960 cm-1are typical,18as originally reported by Jones and
Penneman in 1953.20However, when more strongly basic
ligands like hydroxide7,12,21,22are present, the resulting com-
plexes exhibit much lower ν3values, which have been noted in
both solutions23,24and solids.25-28Similar trends for the
symmetric ν1stretch are seen in Raman spectra12,15,29and strong
correlations between ν1 and ν3 frequencies have been estab-
lished.30Increasing the local electron density at the metal center
in other ways, such as by formal reduction (to UO2+)24,31or
substitution of a more electron-rich metal (i.e., Np, Pu, Am),
produces a similar effect.29,32
Computational chemistry2,3helps provide a quantitative
understanding of structure and bonding in uranyl complexes.
Impressive progress has been made using density functional
theory (DFT),33-36which is remarkable given the theoretical
difficulty of accounting for the large number of electrons, spin-
orbit coupling, and relativistic effects encountered in modeling
uranyl molecules.33,35The computational results are strongly
influenced by the choice of functional, basis set, and effective
core potential employed.37,38Vibrational frequencies generated
using DFT37,39,40are invaluable because they provide a basis
for the interpretation of spectroscopy experiments. However,
comparisons between theory and, for example, solution-phase
species may be present in solution8,41as a result of rapid
ligand exchange, ion pair formation, redox reactions, and
solvent effects.1,7,11,12,17,42Because vibrational spectra collected
from solution-phase experiments potentially contain contribu-
tions from multiple species, comparisons to results produced
by DFT calculations (which are generated for discrete, well-
defined species) are difficult. Longer-range interactions with
the second solvation sphere also influence the spectroscopy of
the complexes and further complicate comparisons. An elegant
way around this is to compare DFT results to structures
determined using X-ray crystallography; however, the effect of
neighboring molecules in the crystal lattice is a complicating
An alternative approach for converging vibrational spectros-
copy and computational chemistry is to measure the infrared
spectra of discrete species isolated in the gas phase,44which
can be accomplished using a trapped ion mass spectrometer
(MS) (e.g., a Fourier transform ion cyclotron resonance [FT-
ICR] or quadrupole ion trap instrument) interfaced to a high-
intensity, tunable infrared source that is provided by a free
electron laser.41,44-47Using electrospray ionization (ESI),48-51
a wide range of UO22+species52-55can be formed and isolated
in the FT-ICR-MS. Normally, ion concentrations in the gas
phase are too low to enable direct absorption measurements,
but by rapid absorption of tens to hundreds of photons, the
vibrational energy of a discrete species may be raised to the
point where bond cleavage occurs.44,56In this case, photon
absorption is signaled by a change in ion mass, and plotting
ion intensities as a function of wavelength produces infrared
multiple photon dissociation (IRMPD) spectra which bear strong
similarity to those measured using conventional absorption
In prior research campaigns, the IRMPD strategy was used
to produce spectra of discrete uranyl-solvent complexes
[UO2Sn)2-5]2+where S ) acetone (ACO) and or acetonitrile
(ACN).60The uranyl ν3 frequency underwent systematic red
shifts with serial addition of donor ligands, and with substitution
of a stronger nucleophilic ligand for a weaker one (e.g., ACO
for ACN). Interestingly, the uranyl ν3 frequencies measured
using IRMPD were never as low as the value for UO22+in
solution,20despite the fact that the ligands in the gas-phase
experiments (ACO and ACN) were stronger nucleophiles than
H2O (the dominant ligand in solution). This observation led to
conjecture that additional interactions may be contributing to
the observed uranyl shift in solution-phase experiments, an
expectation that was recently substantiated using computational
The subject of this report is the IRMPD spectroscopy of gas-
phase [UO2A]+species (where anion A ) OH, OCH3, and
acetate (OAc)), and complexes in which [UO2A]+is modified
by the attachment of a single neutral donor solvent S, to form
[UO2AS1-2]+(where S ) H2O, NH3, ACN, or ACO). The
hydroxide and acetate anions are representative of those
commonly encountered in solution-phase studies of UO22+
speciation,61and acetate and methoxide are models for functional
groups expected to interact with UO22+in biological and
geochemical environments. The primary focus of this work is
to explore and understand the trends in the antisymmetric uranyl
stretching frequency (ν3), as a function of the number and
binding strength of the various anionic and neutral ligands, by
comparing the experimental IRMPD results with predictions
from electronic structure calculations employing several different
computational methods. In general, the measured ν3frequencies
for the bare anion complexes were significantly lower than the
predicted computational values and approached those measured
in solution for coordinatively saturated UO22+. Addition of a
neutral donor to form [UO2AS]+did not substantially alter the
ν3values compared to [UO2A]+, which was surprising because
prior studies showed that the antisymmetric stretch is systemati-
cally red-shifted upon attachment of a donor neutral ligand.
Comparison of the ν3values for different [UO2AS]+complexes
showed a systematic decrease with increasing nucleophilicity
of the neutral donor S. DFT calculations also suggested that
when the neutral is added, bonding is accommodated by
weakening both the uranyl-anion bond, as well as the uranyl
IRMPD spectra were collected at the Free Electron Laser for
Infrared eXperiments (FELIX) facility, located at the FOM
Instituut voor Plasmafysica “Rijnhuizen” (Nieuwegein, The
Netherlands).47The free electron laser is interfaced to a custom-
built Fourier transform ion cyclotron resonance (FT-ICR) mass
Generation of Uranyl Complexes by Electrospray Ioniza-
tion (ESI). ESI was used to generate singly and doubly charged
uranyl complexes.52,54A one millimolar solution of uranyl nitrate
was generated by dissolving the hexahydrate salt (Fluka/Sigma-
Aldrich, St. Louis, MO) in water to produce uranyl complexes
that were introduced into the hexapole ion accumulation
chamber. The ESI source (Micromass, Manchester, U.K.) was
operated at 3 kV with respect to ground. Ions were generated
at atmospheric pressure and were extracted into vacuum using
ion optics oriented orthogonally with respect to the spray axis
and then gated into a hexapole ion accumulation chamber where
they were stored for 0.5-1.0 s prior to being transmitted into
the FT-ICR-MS. The mass spectra observed were sensitive to
various temperatures, voltages, and carrier-gas and solution flow
rates of the ESI source. Particularly important were the
desolvation temperature54(which was controlled by a heater and
Discrete Uranyl Anion Complexes
J. Phys. Chem. A, Vol. 112, No. 3, 2008 509
thermocouple on the block through which the spray capillary
passed) and the temperature of the desolvation gas, which were
maintained at 29 and 52 °C, respectively. The flow rate of the
spray solution was 25 µL min-1, and the desolvation gas (N2,
which ensheathed the solution spray) flow rate was maintained
at 30 L min-1. Attempts to make hydrated [UO2NO3]+were
not successful because traces of methanol, acetone, acetonitrile,
acetic acid, and ammonia in the spray chamber resulted in
production of hydroxide, methoxide, and acetate complexes. By
increasing the radio frequency power on the ion accumulation
hexapole, most of the ion population was converted to [UO2-
OH]+, [UO2OCH3]+, and [UO2OAc]+; these species were also
formed as complexes with a single solvent molecule (see below),
which provided the ensemble of species for infrared spectros-
Fourier Transform Ion Cyclotron Resonance Mass Spec-
trometry (FT-ICR-MS) and Infrared Multiphoton Dissocia-
tion (IRMPD).46,57,59Ions accumulated in the external hexapole
were gated into the ICR cell, where complexes identified for
IRMPD were isolated using a stored waveform inverse Fourier
transform (SWIFT) pulse.63This ejected all species except those
having the desired mass. Isolated ionic complexes were irradi-
ated using two FELIX macropulses, which induced elimination
of a solvent molecule, a radical, or a rearrangement product
(depending on the complex) when the incident wavelength
matched an absorption band. The IRMPD mechanism has been
described in detail elsewhere.45,62,64Briefly, it involves sequen-
tial, non-coherent absorption of many (tens to hundreds) infrared
photons, with each photon being “relaxed” by intramolecular
vibrational redistribution (IVR) before the next one is absorbed.
In this way, the internal vibrational energy of the molecule can
be resonantly increased above the dissociation threshold, result-
ing in fragmentation. It has been shown that the infrared spectra
obtained are comparable to those obtained using linear absorp-
tion techniques.56,65FELIX (60 mJ per macropulse, 5 µs pulse
duration, bandwidth 0.2-0.5% of central λ) was scanned
primarily through the spectral region of interest around 10 µm,
in increments e0.04 µm, after which IRMPD product ions and
un-dissociated precursor ions were measured using the excite/
detect sequence of the FT-ICR-MS.66,67The IRMPD efficiency
was then expressed as -log(1 - [summed fragment ion yield]),
corrected for the width of the acquisition channels and linearly
normalized to correct for variations in FELIX power over the
spectral range. Peak centers were chosen by fitting a Gaussian
peak to the data using Origin plotting software (version 7.5,
OriginLab, Northampton, MA). Precision was not evaluated,
because of the time required for repetitive acquisition of the
peak profiles, and the precious nature of beam time at FELIX.
Nevertheless, the precision of measurement of peak position is
probably on the order of a couple of cm-1, based on earlier
examination of the position(s) of the antisymmetric uranyl
stretch and carbonyl CdO stretch in doubly charged com-
Because some of the complexes were difficult to fragment,
signal-to-noise was less than desired, and so the isolated
complexes were subjected to multiple irradiation/acquisition
sequences at each wavelength across the scanned region. This
lengthened acquisition time, together with the need to strictly
economize beam time at the FELIX FT-ICR-MS beamline,
constrained data acquisition for most complexes to the diagnostic
OdUdO antisymmetric stretch (ν3).
Density Functional Theory (DFT) Structure and Fre-
quency Calculations. DFT calculations of structures and
harmonic frequencies were performed with the DMol3,68,69
NWChem,70,71Gaussian,72and ADF2006.0173suite of programs.
Different combinations of functionals, basis sets, and relativistic
treatments were employed in efforts to derive a consistent view
of the IRMPD phenomena measured in the context of complex
structure and dissociation behavior. Functionals used were local
density approximation (LDA) with the Vosko, Wilk and Nusair
(VWN) parametrization,74,75the hybrid B3LYP functional,76,77
and the PW91 functional.78
(1) For calculations with DMol3the so-called density
functional semicore pseudopotential (DSPP)79was used, which
accounts for scalar relativistic effects and includes 60 electrons
in the uranium core (comparable to Stuttgart/Dresden small core
ECPs), combined with the use of polarized numerical basis sets
(DNP) for the active electrons. A fine (10-8) energy convergence
criterion was employed to ensure optimal geometries and
representative vibrational frequencies.
(2) For calculations with NWChem and Gaussian, the uranium
was described by an effective core potential and its associated
basis set: either the LANL2dz ECP and orbital basis set80or
the MWB60/Stuttgart RSC ECP and basis set (SDD),81-84which
features Stuttgart/Dresden effective core potentials. Other atoms
in the complexes (O, C, H, and N) were described using the
aug-cc-pVDZ orbital basis sets,85the small D95V basis set,81-84
the all-electron 3-21g* basis sets, or the 6-31+g(d) basis set.
The relatively small basis sets (3-21g (which includes single
first polarization functions on row 2 atoms) and D95V) are
generally considered to be too small for use in modeling actinide
molecules; however, the experimental results and higher-level
basis sets used in this study provide a good platform to evaluate
the use of the SDD/3-21g* and SDD/D95V general basis set
approaches for quick, on-the-spot interpretation of gas-phase
(3) For calculations with ADF, the scalar relativistic ZORA86
Hamiltonian combined with the TZ2P type basis set was applied
within the framework of unrestricted DFT, and a small frozen
core that takes 78 electrons in the uranium core and also freezes
the 1s electrons of the first row atoms.
NWChem, ADF, and Gaussian were used to compute binding
energies and the energetic requirements for different dissociation
channels. Prior experience indicated reliable thermodynamic
accuracy and thus motivated this approach. For NWChem, the
B3LYP energies were obtained using the geometry obtained
from LDA geometry optimizations. The energies computed with
ADF included spin-orbit coupling terms. The energy differ-
ences refer to the fully dissociated limit and hence do not include
the additional energy that may be needed to overcome a possible
barrier in a transition state (which may be relevant for the loss
of H2O from [UO2OH(ACO)]+and photofragmentation of [UO2-
In addition, NWChem was used to perform correlation cor-
rected vibrational self-consistent field (cc-VSCF) calculations87-89
using the LANL2dz basis for uranium, the aug-cc-pVDZ basis
sets for the first-row atoms, and the B3LYP functional. These
calculations provide an estimate of the effect of anharmonicity
and mode-coupling for the fundamental vibrational states.
Starting with the VSCF method, degenerate perturbation theory
is used to correct for effects of correlation between different
vibrational modes, enabling calculation of anharmonic vibra-
tional states for polyatomic molecules.
Finally, we also employed ADF to quantitatively assess the
donation of the ligands to the uranyl moiety. The charge transfer
between ligands and the uranyl was analyzed using both
Hirshfeld charge analysis90and Voronoi Deformation Density
510 J. Phys. Chem. A, Vol. 112, No. 3, 2008
Groenewold et al.
Results and Discussion
IRMPD of Uranyl-Hydroxide Complexes [UO2OH]+.
Electrospray ionization mass spectrometry of aqueous uranyl
nitrate solutions modified with organic solvents had previously
been shown to produce dications ligated with neutral donors,
provided the capillary temperature was kept close to ambient.54,60
However, by modestly increasing the capillary temperature and
the RF power of the ion accumulation hexapole, uranyl ion pairs
were formed that enabled examination of their IRMPD spectra,
as a complement to prior measurements made for the uranyl
dication bound with neutral ligands.60A prominent [UO2OH]+
ion was produced at m/z 287, and lower-abundance complexes
were observed at m/z 345, 328, 305, and 304 that correspond
to [UO2OH(ACO)]+, [UO2OH(ACN)]+, [UO2OH(H2O)]+, and
[UO2OH(NH3)]+, respectively. The ammonia, acetone, and
acetonitrile were present in the hexapole accumulation region
of the ESI/FT-ICR instrument from prior experiments that
involved the use of the solvents and ammonium acetate buffer
solutions. The five different hydroxide complexes that were
furnished by manipulation of the electrospray conditions were
isolated using a SWIFT sequence,67and then photofragmented
by scanning the free electron laser over the spectral region
frequency (∼1000 cm-1); the resulting IRMPD spectra are
shown in Figure 1.
Photofragmentation of [UO2OH]+resulted in reductive
elimination of a hydroxyl radical, and the antisymmetric uranyl
ν3stretch for this species appeared as a broadened absorption
centered at 971 cm-1(Figure 1). The low abundance and profile
of the peak reflected inefficient photofragmentation and high
energetic requirements:44calculations using ZORA/PW91/TZ2P
(vide infra) indicated that the minimum energy to dissociate
the complex to [UO2]+and a hydroxyl radical was 93.9 kcal
mol-1. A somewhat smaller dissociation energy of 74.5 kcal/
mol was obtained from calculations with the Stuttgart RSC/
B3LYP/TZVP combination. The ν3value for [UO2OH]+was
lower than that for the most red-shifted dication complexes
[UO2(ACO)4]2+(988 cm-1) and [UO2(ACN)5]2+(995 cm-1),
which suggested at first glance that a single hydroxide transfers
as much or more electron density to the uranium center as
does four or five strong donors in a fully coordinated uranyl
complex. VDD analyses indicated a charge transfer of 0.52 e
from the hydroxide to the uranyl, and the charge transfer
from four acetones accounted for 0.62 e, suggesting at first
glance that the ν3 values should follow the opposite trend.
However, charge transfer would be expected to be more closely
correlated with the ν1 shift, which is influenced purely by
electronic factors, and our preliminary estimates indicate that
ν1 would be very similar for both the OH and (ACO)4
The value for the [UO2OH]+uranyl antisymmetric stretching
frequency was lower than anticipated on the basis of prior DFT
calculations, in which a red shift of 183 cm-1was predicted
for [UO2(OH)2] by Marsden and co-workers;40subtraction of
this value from 1140 cm-1(the value calculated for unligated
[UO2]2+)38produces a frequency of 957 cm-1for the bis-
hydroxy complex. The measured value of 971 cm-1for the
monohydroxy cannot be compared directly, because the modeled
and measured complexes are different. But the values are
reasonably close to each other, which suggests that most of the
red shift results from attachment of the first OH-ligand, and
that attachment of the second ligand produces a much more
modest change in frequency. This trend is in qualitative
agreement with very small shifts produced by attachment of
neutral donors to [UO2A]+complexes (vide infra).
The [UO2OH]+ν3value measured in the gas phase is very
close to that measured in aqueous solution for [UO2]2+(960-
965 cm-1), which is considered to have five inner sphere aquo
ligands.17,18,20,92,93Lower values have been measured for
hydroxide complexes in solution, but these have been attributed
to species having multiple uranyl moieties; e.g., a ν3measure-
ment at ∼940 cm-1has been assigned to [(UO2)2(OH)2]2+,17,21,94
and an even lower ν3value of 923 cm-1to [(UO2)3(OH)5]+.17,94
These ν3 measurements indicate that the frequency is
decreased by the presence of more than one U atom in the
complexes but is also certainly influenced by coordinated solvent
The appearance of [UO2OH]+complexed with one or more
solvent molecules enabled the effect of neutral donor ligands
on the antisymmetric stretching frequency to be examined. The
prior study of [UO2]2+complexes with neutral ligands60showed
that the antisymmetric stretching frequency was sequentially
red-shifted by the serial attachment of additional neutral donor
ligands, for example in the acetone (ACO) complexes [UO2-
(ACO)n]2+, the frequency decreased from 1017 to 1000 to 988
cm-1as n went from 2 to 3 to 4 (respectively).60The trend
measured for a series of acetonitrile (ACN) complexes was
similar, as was the magnitude of the red shift caused by an
additional donor neutral. These observations led to the expecta-
tion that attachment of a neutral donor to [UO2OH]+would
result in a similar red shift.
antisymmetric OdUdO stretching region for [UO2OH]+and ligand
complexes containing (clockwise) a single ACN, H2O, NH3, and
ACO. The yield values for the ACN complex were multiplied by a
factor of 2, to visually distinguish it from the unmodified hydroxide
Infrared multiple photon dissociation spectra of the
and Photofragmentation of [UO2OH(ACO)]+
Parallel Elimination Reactions of Isolation
Discrete Uranyl Anion Complexes
J. Phys. Chem. A, Vol. 112, No. 3, 2008 511
Isolation and photofragmentation of [UO2OH(ACO)]+re-
sulted in parallel elimination reactions: loss of intact ACO, and
loss of H2O (Scheme 1, Table 1). Computationally, the two
pathways were found to have very similar reaction energies:
using ZORA/PW91/TZ2P both channels were endothermic by
41.4 kcal/mol, and a similar conclusion was derived using
different basis sets and functionals. The loss of H2O involves
transfer of a proton from a methyl carbon on ACO to the
hydroxide, leaving behind the acetone enolate which calculations
show remains coordinated through the oxygen atom. No
difference in frequencies was observed in the two photofragment
channels, which had maxima at 972 cm-1. This value was
effectively equal to the measurement for the unmodified [UO2-
OH]+, which initially would seem to indicate that addition of
the strong ACO donor had no further effect on the uranyl
moiety. However, this conclusion is inconsistent with the fairly
strong binding of ACO predicted by PW91 and B3LYP
calculations, which should shift the uranyl ν3value further to
the red. Furthermore, the carbonyl stretching region for the
[UO2OH(ACO)]+complex was scanned and an absorption with
a value of 1633 cm-1was found. In our previous study of [UO2-
(ACO)n]2+complexes, we observed that the ligand CO stretch
was strongly red-shifted to 1515 cm-1in the n ) 2 complex,
and that this shift decreased with increasing cluster size, to 1583
for n ) 3 and 1630 for n ) 4, as the binding energy per ligand
was reduced.60,95Thus, on the basis of this comparison of
carbonyl stretching data, one would expect the binding of the
ACO ligand in the [UO2OH(ACO)]+complex to be comparable
to binding in the [UO2(ACO)4]2+complex. Interestingly,
addition of a second and third ACO ligand to [UO2OH(ACO)]+
produced red shifts consistent with the prior experiments.
Photofragmentation of [UO2OH(ACO)n)2,3]+complexes resulted
in the loss of an ACO, with ν3values of 961 and 948 cm-1for
n ) 2 and 3, respectively. Therefore, a red shift of 11 cm-1
was observed for n ) 1 f 2, and 13 cm-1for n ) 2 f 3; these
shifts were in good agreement with the magnitude of the red
shifts resulting from ACO addition to the [UO2(ACO)n]2+
complexes.60Taken together, these results suggest that it may
be the uranyl antisymmetric stretching frequency in the bare
[UO2OH]+complex that is anomalously shifted, which is
certainly possible given the particular susceptibility of this very
small system to anharmonicity effects arising from the IRMPD
mechanism. The possible anharmonicity effects are discussed
in more detail below.
Isolation of the [UO2OH(ACN)]+complex followed by
photofragmentation produced elimination of ACN. The anti-
symmetric UO2 stretch was measured at 972 cm-1, nearly
identical to that for [UO2OH(ACO)]+and to [UO2OH]+. The
hydroxide complex with ammonia [UO2OH(NH3)]+underwent
photofragmentation solely by loss of NH3, with a ν3value at
976 cm-1, which is slightly blue-shifted compared to the
unmodified uranyl hydroxide. The final hydroxide complex
examined was [UO2OH(H2O)]+, which eliminated H2O upon
irradiation that maximized at 983 cm-1, which was 12 cm-1
higher than the [UO2OH]+value.
The trend for the uranyl ν3frequencies for the [UO2OH(S)]+
complexes were internally self-consistent; i.e., they decreased
in the order H2O > NH3> ACN ∼ ACO > (ACO)2(Figure
2). These frequency values are inversely correlated with the
calculated coordination energies of different S molecules40and
are in accord with comparisons of ACO and ACN uranyl
complexes.60The observed ordering again highlights the
remarkably low value measured for unmodified [UO2OH]+,
which would be expected to be higher than 983 cm-1(i.e., the
value for the complex with the most weakly bound neutral,
[UO2OH(H2O)]+). Because it does not seem reasonable to
expect that the addition of weakly donating species actually
strengthens the uranyl UdO bonds, we must seek other
explanations for the blue-shifted bands for the NH3and H2O
One factor likely contributing to the low ν3value for [UO2-
(OH)]+is vibrational anharmonicity derived from the multiple
photon absorption process. Red shifts in the spectra of the para-
aminobenzoic acid58and [CeOH(ACO)3]2+cations95have been
attributed to IRMPD anharmonicity, and the same phenomenon
may contribute to the low frequency measured for [UO2OH]+.
These studies demonstrate that when molecules or complexes
attain very high internal energies via the IRMPD process, their
vibrational bands exhibit non-negligible red shifts. This can be
particularly dramatic for systems with low densities of states,44
such as the [UO2OH]+complex studied here, which only has
nine vibrational modes. The energy required to dissociate [UO2-
OH]+to [UO2]+and a hydroxyl radical was evaluated compu-
tationally, and both the Stuttgart RSC ECP and the ADF TZ2P
basis sets produced high values (Table 2). This was consistent
with the fact that hydroxide had the highest coordination
energy of any ligand in the extensive compilation calculated
by Marsden and co-workers.40These considerations support the
TABLE 1: Dissociation Energies for IRMPD Reactions of [UO2OH(ACO)]+Calculated Using Different Basis Sets
binding energy (kcal/mol)
[UO2OH(ACO)]+f [UO2OH]++ ACO
[UO2OH(ACO)]+f [UO2(OC3H5)]++ H2O
Figure 2. Variation in the antisymmetric OdUdO stretching frequen-
cies for [UO2(OH)]+and [UO2(OH)(neutral)]+complexes. Black
squares are IRMPD measurements, and filled red dots are ν3 values
calculated using PW91, both evaluated using the left y axis. Open red
circles represent dissociation energies for the complexes, calculated
using PW91 and evaluated with the right y axis.
512 J. Phys. Chem. A, Vol. 112, No. 3, 2008
Groenewold et al.
attribution of the anomalously low [UO2OH]+ν3 value to
anharmonicity effects. Another possible explanation is that the
dissociation proceeds via a low-lying electronically excited state.
The lowest triplet excited state, which should be accessible via
spin-orbit coupling, lies ∼31 kcal mol-1(calculated using
B3LYP and the Stuttgart RSC/TZVP basis) below the dissocia-
In contrast to the small [UO2OH]+molecule, more complex
systems containing a neutral ligand tend to have reduced
dissociation energies and a significantly higher vibrational
density of states. For example, water binds to the [UO2OH]+
complex by only 30 kcal/mol, while doubling the number of
vibrational modes, three of which are low-frequency intermo-
lecular modes that may contribute disproportionately to the
vibrational density of states. This may explain why the ν3
frequencies of the [UO2OH(S)]+complexes are not apparently
affected by anharmonicity, compared to the bare [UO2OH]+
which is strongly shifted by anharmonicity induced by the
IRMPD process. A ν3value unshifted by anharmonicity would
be expected to be ∼990 cm-1for [UO2OH]+to be consistent
with the trend in the ν3bands of the [UO2OH(S)]+complexes
Correlation corrected vibrational SCF (cc-VSCF) calculations
can provide a rough estimate of whether or not differential red-
shifting would be expected in comparing [UO2OH]+and [UO2-
OH(S)]+complexes. The cc-VSCF estimates the effect of
anharmonicity by including coupling between the lower vibra-
tional modes calculated in the harmonic approximation. Factor-
ing in an influence by anharmonicity, the calculated ν3value
for [UO2OH]+decreased by 10 cm-1(from 987 cm-1in the
harmonic approximation to 977 cm-1with the anharmonic
corrections), which would account for part of the expected red
shift for the hydroxide seen in the IRMPD data. However, the
shift calculated for [UO2OH(H2O)]+was very nearly the same
at 8 cm-1(going from 968 to 960 cm-1with anharmonic
corrections), so the differential anharmonic shift calculated for
[UO2OH]+and [UO2OH(H2O)]+does not explain the low ν3
value measured for [UO2OH]+. It should be noted that the cc-
VSCF calculations only consider coupling of the lowest
∼10 vibrational levels, whereas coupling and population of the
higher excitation levels would certainly be expected to contribute
at the high excitation energies achieved in the IRMPD experi-
As mentioned above, participation of an electronically excited
state for [UO2OH]+could also contribute to the apparently low
ν3frequency. The energy calculated for promotion of an electron
into the lowest excited triplet state calculated using both
LANL2dz/B3LYP/aug-cc-pVDZ and scalar ZORA/PW91/TZ2P
using unrestricted DFT was ∼45 kcal/mol, which is less than
that calculated to dissociate [UO2OH]+(except when the
LANL2dz basis set was used). If vibrational-to-electronic
transitions are indeed occurring in the multiple photon photo-
dissociation experiments, then a lowered frequency would be
expected for the electronically excited molecule.96The ν3value
calculated for [UO2OH]+in this lowest excited state was quite
a bit lower, at 908 cm-1, and the broadening of the hydroxide
profile might thus be the result of overlap of absorptions of
ground-state and excited-state molecules. It should also be noted
that this triplet state can be considered a complex of reduced
uranyl [UO2]+and neutral hydroxyl radical and is hence directly
related to the observed dissociation pathway.
IRMPD of Uranyl-Acetate Complexes [UO2OAc]+. ESI
produced a relatively abundant ion at m/z 329, which was
attributed to uranyl acetate [UO2OAc]+, that was formed from
residual acetic acid that had been used to enhance the protonated
ion formation from peptide and protein solutions in previous
experiments at FELIX. Because acetate is a stronger conjugate
base, it replaces nitrate in the ion accumulation chamber prior
to injection into the FT-ICR-MS. The composition was con-
firmed by accurate mass measurement and the photofragmen-
tation pathway observed in the IRMPD experiment, in which a
neutral loss of 42 mass units (presumably as ketene) furnished
[UO2OH]+as the product ion. For IRMPD of [UO2OAc]+the
maximum for the antisymmetric uranyl stretch was 995 cm-1
(Figure 3), higher than any of the hydroxide complexes
measured. This is consistent with the fact that acetate is a weaker
gas-phase base97than either hydroxide or methoxide (vide infra)
and consequently is also likely to be a weaker uranophile.
Despite the presumed lower basicity, the antisymmetric uranyl
stretching frequency for [UO2OAc]+appeared at a value lower
than that for nearly all of the uranyl dication complexes ligated
with multiple neutral donor ligands reported earlier.60In solution,
acetate complexes have been the subject of several infrared
studies, and the most appropriate value for the antisymmetric
stretch to use in a comparison is 954 cm-1, which was measured
by Quiles17for [UO2OAc]+. This value is significantly lower
than the IRMPD measurement, which reflects the attachment
of additional solvent ligands to the [UO2OAc]+metal center.
Other studies have produced values that ranged as low as 919
cm-1,18,23but these measurements probably contain contributions
from species that contain more than one acetate ligand, and the
possibility of variable acetate coordination.17Recently, DFT
using the LDA functional was used by de Jong and co-workers
to calculate uranyl ν3value for [UO2OAc]+at 1025 cm-1.98A
careful examination of the carbonyl stretching frequencies could
provide insight into this and will be investigated in further
Also observed in the ESI spectrum were low-abundance ions
at m/z 346 and 347 that corresponded to ammonia and water
TABLE 2: Dissociation Energies Calculated for IRMPD of
[UO2OH]+f [UO2]++ OH
binding energy (kcal/mol-1)
dissociation from the singlet ground state
dissociation from a triplet excited state
antisymmetric OdUdO stretching region for [UO2OAc]+and ligand
complexes containing a single NH3and H2O.
Infrared multiple photon dissociation spectra of the
Discrete Uranyl Anion Complexes
J. Phys. Chem. A, Vol. 112, No. 3, 2008 513
complexes (respectively), having compositions [UO2OAc-
(NH3)]+and [UO2OAc(H2O)]+. Photofragmentation of these
complexes involved elimination of either NH3or H2O, producing
[UO2OAc]+in each case. Consistent with prior studies of
donors, the antisymmetric UO2stretching frequency was red-
shifted for both H2O and NH3complexes relative to [UO2OAc]+,
although the magnitude of the shifts were small: the value for
the H2O complex at 993 cm-1was lower by 2 cm-1, and a
shift of 7 to 988 cm-1was observed for the NH3complex. The
trend in the measured frequencies indicate that both H2O and
NH3are donating electron density, and that NH3is a stronger
uranophile than is H2O, consistent with Marsden’s prior DFT
results,40the relative order of proton affinities,97and bonding
to other metal cations.99The fact that the frequency of the
unmodified acetate is very close to that of both ligand
complexes suggests that the frequency for the [UO2OAc]+
complex may also be red shifted as a result of anharmonicity.
However, the effect is less pronounced than in the case of the
hydroxide complex, as indicated by the fact that the ν3of the
unmodified [UO2OAc]+complex is not lower than the values
of the H2O and NH3 adducts. Compared to the hydroxide
complex, a smaller anharmonic red shift for OAc is consistent
with a higher density of states,44and with a lower-energy
requirement for fragmentation of the acetate complex, which
involves rearrangement rather than direct bond-cleavage and
elimination of a radical.
IRMPD of Uranyl-Methoxide Complexes [UO2OCH3]+.
The accurate mass measurement of the ion at m/z 301 confirmed
the composition of uranyl methoxide, which was formed by
reaction of uranyl ion with methanol that was present in the
ESI solution. Photofragmentation of [UO2OCH3]+produced four
different product ions corresponding to the elimination of the
OCH3and H radicals, H2, and H2CO (Scheme 2).
The IRMPD spectrum of the methoxide complex contained
two bands, with maxima at 975 and 887 cm-1(Figure 4, orange
trace). DFT calculations (B3LYP/Stuttgart RSC/3-21G*) indi-
cated that the higher frequency corresponded to overlapped
antisymmetric uranyl ν3 and C-O stretching bands, and the
lower frequency to the symmetric uranyl ν1 band normally
observed in the Raman spectrum.7,12,15,22,29The appearance of
the symmetric stretch indicates a lowered symmetry in the
complex, resulting from strong binding of the methoxide that
perturbs the linearity of the uranyl functional group. This was
supported by the lowest-energy structures and bond angles
produced by B3LYP calculations (vide infra).
Striking differences were observed when comparing the IR
spectra produced using the different photodissociation channels.
The spectrum generated by monitoring the loss of the OCH3
radical contained a single sharply defined peak with a maximum
at 967 cm-1, a frequency slightly lower than that measured for
the uranyl antisymmetric stretch for the unmodified hydroxide
complex, and consistent with the fact that methoxide is a
stronger base than is hydroxide. The peak centered at 967 cm-1
was not observed in the spectra generated by the other three
fragmentation channels, either because it is not occurring in these
channels or because it was overlapped with the O-C stretch
(see below). If the latter is true, then it suggests that the ν3
frequency in the spectrum of the OCH3 loss channel is red-
shifted by about 20 cm-1due to the higher energetic require-
ments for that channel; shifts of this magnitude have been
The IR spectra generated from the photodissociation channels
corresponding to the loss of either H or H2 bore strong
similarities to one another in that they contained a peak with a
maximum at 987 cm-1and a second peak at about 887 cm-1.
The higher-frequency peak probably contains components from
the unresolved uranyl asymmetric stretch and C-O stretching,
and the lower-frequency peak corresponds to the symmetric UO2
stretch. The IR spectrum generated by the H2CO elimination
was similar to the H-loss spectra but lacked the band for the
symmetric stretch. The appearance of very dissimilar IR spectra
in the different photodissociation channels was remarkable,
because IRMPD spectra generated from competing mass chan-
nels are normally identical or are very similar, with the
fragmentation channels having the higher energetic requirements
being modestly red-shifted as a result of anharmonicity that
results from population of higher vibrational levels when
multiple photons are serially absorbed.58,95Fast intramolecular
vibrational redistribution randomizes the deposited energy
regardless of the frequency of initial deposition, and thus the
competing fragmentation channels display similar if not identical
A hypothetical interpretation of these observations is that the
order of reaction endothermicities for the four reactions is -H2
∼ -H < -OCH2< -OCH3. In the IR spectra generated by
losses of H and H2, peaks are seen in all three absorption modes,
symmetric uranyl, asymmetric uranyl, and C-O (assuming that
the asymmetric uranyl and C-O are overlapping). The sym-
metric uranyl and C-O are weakly absorbing modes and hence
are only seen in those eliminations having low-energy require-
ments. The spectrum generated by loss of OCH2contains the
asymmetric uranyl and perhaps the C-O, but the energetic
requirement for this channel is too high to enable observation
symmetric and antisymmetric OdUdO ν3stretching region for [UO2-
OCH3]+. The orange trace represents the spectrum generated by the
summed photofragment abundance, red represents the OCH3 radical
elimination, violet represents the H radical elimination, blue represents
the H2elimination, green represents the H2CO elimination, and black
represents the sum of the H-loss/rearrangement related channels.
Infrared multiple photon dissociation spectra of the
SCHEME 2: Photofragmentation of [UO2OCH3]+
Yielding Four Different Product Ions Corresponding to
the Elimination of the OCH3and H Radicals, H2, and
514 J. Phys. Chem. A, Vol. 112, No. 3, 2008
Groenewold et al.
of the symmetric uranyl stretch. The higher energetic require-
ments are probably related to the fact that the product ion has
to be reduced, forming either a U(IV) species [UOOH]+or a
uranyl hydride [UO2(H)]+. Finally, the spectrum generated by
loss of OCH3 contains only the antisymmetric uranyl peak,
because the energetics for this reaction are higher, which means
that it can only be accessed via the high-intensity ν3 uranyl
absorption. This peak is substantially red-shifted as a conse-
quence of the large number of photons that must be deposited
for the reaction to occur. Further, the presumed rapid rate for
OCH3 radical loss reduces observation of the slower, lower-
Enthalpy changes calculated for the different fragmentation
channels displayed some agreement with this hypothesis.
B3LYP and PW91 calculations using different core approxima-
tions, functionals, and basis sets (Table 3) showed that elimina-
tion of H2was the lowest. The losses of the radical species H
and OCH3were next and appeared to be roughly energetically
competitive, depending on computational approach; the PW91
calculations suggest that loss of OCH3requires more energy
than loss of H radical, and B3LYP gives somewhat smaller
dissociation energies and has the H elimination channel slightly
above the OCH3 elimination channel. Enthalpy requirements
for the loss of H2CO vary depending on the nature of the product
ion. Both B3LYP calculations indicate that formation of a linear,
triplet [UO(OH)]+is lowest, within about 7 kcal mol-1of H2
loss, but when a singlet is formed, or when the product is uranyl
coordinated with an equatorial hydride, fragmentation enthalpies
were higher. In summary, the PW91 energetics supports the
idea that loss of OCH3radical is higher in energy and thus more
susceptible to anharmonic shifting. However, the variations in
the reaction energies seen when comparing the computational
approaches do not allow us to state the energetic order of the
elimination reactions unequivocally.
An alternative explanation would be the existence of two or
more isomers of [UO2OCH3]+; however, DFT calculations did
not support the existence of energetically competitive isomers,
although rearrangement may be occurring during the IRMPD
process. An alternative structure that was considered contained
an H atom bound to uranium, with formaldehyde equatorially
coordinated: for such a structure an absorption corresponding
to carbonyl group should be observed, but a survey of the 1500-
1700 cm-1wavelength region did not show an additional peak.
Thus, a structure involving a bound formaldehyde ligand is
unlikely, as our prior studies60,95showed that the CdO stretch
can be readily detected in complexes with carbonyl-containing
Involvement of an excited state for the uranyl methoxide can
also be argued, which would be expected to have energetic
requirements similar to the 45 kcal/mol required for the
hydroxide complex. Intuitively, this is an attractive explanation
because a higher-spin species would be expected to have a
higher propensity for rearrangement and elimination of H and
OCH3radicals. When the energetic requirement for conversion
to a triplet excited state was calculated using PW91, it was also
found to be 45 kcal mol-1, in a range that would be accessible
during the IRMPD photofragmentation. However, as in the two
previously offered rationalizations, this too remains speculative
at the present time, and hence an unequivocal identification of
the origin of the differences in the spectra of the different mass
channels is still elusive.
The assignment of the higher frequency to a C-O stretch
drew support from the spectra acquired for the [UO2OCH3-
(H2O)]+and [UO2OCH3(NH3)]+adducts (Figure 5). The three
peaks in the spectra of these complexes had frequencies
consistent with the spectra of unmodified [UO2OCH3]+. In the
adduct ions, photofragmentation of the methoxide ligand did
not occur; instead, only the energetically favored losses of H2O
or NH3were observed.
The frequencies measured for the antisymmetric UO2stretch
for the H2O and NH3 complexes were modestly red-shifted
compared to the maximum value for the summed photofragment
channels of the unmodified [UO2OCH3]+, and the trend
TABLE 3: Calculated Enthalpies for the Dissociation Reactions of [UO2OCH3]+(Scheme 2)
binding energy (kcal/mol-1)
[UO2OCH3]+f H2+ [UO2(OCH)]+
[UO2OCH3]+f H + [UO2(OCH2)]+
[UO2OCH3]+f OCH3+ [UO2]+
[UO2OCH3]+f OCH2+ [UO2(H)]+
(H linear, triplet spin state)
[UO2OCH3]+f OCH2+ [UO2(H)]+
(H linear, singlet spin state)
[UO2-OCH3]+f OCH2+ [UO2(H)]+
aZORA numbers include spin-orbit interaction.
antisymmetric OdUdO stretching region for [UO2OCH3]+(black trace,
sum of all photofragment channels), [UO2OCH3(H2O)]+(blue trace,
scaled by 0.55), and [UO2OCH3(NH3)]+(green trace scaled by 0.6).
The photofragment yield for the H2O and NH3complexes was higher
than for the unmodified methoxide complex, and scaling was performed
to facilitate comparison.
Infrared multiple photon dissociation spectra of the
Discrete Uranyl Anion Complexes
J. Phys. Chem. A, Vol. 112, No. 3, 2008 515
observed is consistent with what would be expected for addition
of a second weak donor ligand (H2O), and then substitution of
a slightly more basic ligand NH3 for H2O. Similarly, the
frequencies measured for the symmetric stretch were very
similar for all three complexes: the ν1value for the unmodified
[UO2OCH3]+was measured at 887 cm-1, and the peak position
is only very modestly shifted to 880 cm-1when H2O is attached,
and to 879 cm-1for NH3. These values are about 20 cm-1higher
than ν1values measured for solvated [UO2OAc]+using Raman
spectroscopy.12As in the case of the hydroxide complexes, the
uranyl stretching frequencies of the methoxide complexes were
not significantly red-shifted by addition of a neutral donor
ligand. This suggests that in the unmodified methoxide complex
[UO2OCH3]+, the uranyl frequency may be shifted to a lower
value as a result of anharmonicity, in a fashion similar to that
suspected to be occurring in the hydroxide complexes. As noted,
red-shifting would be facilitated by high energetic requirements
for the elimination reaction in the unmodified [UO2OCH3]+or
by participation of an excited state.
Although addition of a second donor ligand does not cause
large changes in the uranyl stretching frequencies,43,98it appears
to strengthen the C-O bond in the methoxide ligand. This would
be expected if the methoxide were modestly repelled by
attachment of H2O or NH3to the uranium center. In the spectra
for both [UO2OCH3(H2O)]+and [UO2OCH3(NH3)]+, the C-O
stretch was observed at ca. 1038 and 1040 cm-1, shifted to
higher frequency by ∼50 cm-1compared to unmodified [UO2-
OCH3]+. This trend is directly analogous to what was observed
in the IR spectra of discrete uranyl acetone dication com-
plexes: when an additional donor ligand was added, the binding
of all equatorial ligands was weakened, and the CdO stretching
frequency increased, approaching that of free acetone;60in the
present case, it is the C-O stretch of methoxide that is increased.
Calculations also suggested loosening of the U-OCH3 bond
upon ligation with a neutral donor.
Comparisons of Calculated Frequencies. The changes in
vibrational frequencies can be understood in part by comparison
with frequencies, bond lengths, and angles calculated using
density functional theory. Because calculations of complexes
containing actinide elements are challenging, different combina-
tions of functionals and basis sets were used. These results
provided multiple opportunities for comparison with measure-
ments, in particular using the antisymmetric uranyl stretch,
which was the salient figure of merit in this study. A comparison
of the uranyl frequencies calculated using B3LYP with different
basis sets versus the IRMPD measurements (Figure 6, Table 4)
showed that overall the smallest errors were obtained using the
SDD/D95V approach (Stuttgart RSC for U, D95V for all other
elements). This agreement might be somewhat fortuitous due
to cancellation of errors, considering the D95V basis sets do
not have polarization functions. The LDA/DSPP/DNP results
are also in reasonable agreement, though some of the calculated
trends are incorrect. Calculated values using Stuttgart RSC on
U and 3-21g* and 6-31+g(d) for C, H, N, and O were
systematically 20-30 cm-1higher than measurements for the
solvent complexes, depending on the donor. Overall, differences
between the two sets of calculations were small, but the values
calculated using 6-31+g(d) were slightly better than those
generated using 3-21g* where some trends were incorrect. The
data in Figure 6 may be grouped into three categories: anion
complexes with no donor, with an O-donor (H2O or acetone),
or with an N-donor (NH3or acetonitrile). Calculations for the
O-donor anion complexes were in best agreement with experi-
ments, being within a few cm-1when the SDD/D95V approach
was used, and on the order of 20 cm-1high when the Stuttgart
RSC/3-21g* and Stuttgart RSC/6-31+g(d) approaches were
Figure 6. Uranyl ν3frequencies calculated using B3LYP plotted versus
IRMPD measurements. The line represents the experimental data. Filled
square data points were generated using the SDD pseudopotential
approach (Stuttgart RSC ECP on U, D95V on C, H, N, and O) elements.
Values represented by open squares were generated using Stuttgart RSC
ECP for uranium and 3-21g* for C, H, N, and O, and values represented
by open triangles utilized 6-31+g(d) for C, H, N, and O. Black points
represent [UO2(anion)]+complexes with no neutral donor, red represents
those with O-donating neutrals, and blue represents those with
TABLE 4: Uranyl Antisymmetric Stretching Frequencies (ν3) for Complexes [UO2AS0,1,2]+, Generated Using IRMPD, and
Calculated Using the Same B3LYP, LDA, and PW91 Functionals, Comparing Various Basis Sets
516 J. Phys. Chem. A, Vol. 112, No. 3, 2008
Groenewold et al.
used. Calculations for the complexes containing a neutral that
coordinates via an N atom were slightly less accurate, with
differences ranging from 5-10 cm-1for the Stuttgart RSC/
D95V approach to ∼30 cm-1using the other basis sets for the
first row elements. This suggests that N-donation is slightly more
aggressive in the gas-phase experiment than predicted by theory.
Calculations for the [UO2(anion)]+complexes containing no
neutral donor displayed the poorest agreement with experiment,
being 20 to nearly 60 cm-1higher than the measurements,
depending on the basis set used. The poorer agreement likely
reflects the high energetic requirements for photofragmentation
pathways for these complexes, which is reasonable because they
involve elimination of an oxy radical with concomitant reduction
of the uranium center. The need to vibrationally excite the
uranyl-anion complexes to a higher level increases the op-
portunity for the measured ν3value to be shifted lower as a
result of vibrational anharmonicity, or perhaps by participation
of an excited state, as suggested above.
In contrast, calculations using LDA and PW91 functionals
produced uranyl ν3frequencies that were slightly lower than
measurements (Figure 7). The LDA/DSPP/DNP values (Figure
7, open squares) were generally in good agreement with
measurement, with the salient exception of [UO2OH]+. Com-
pared to experiments, the LDA values for anion complexes with
O-donating ligands were systematically lower than values for
complexes with N-donors, by about 10 cm-1. The value
calculated for the [UO2OH(ACO)2]+was ca. 20 cm-1lower
than the measured value.
The frequencies generated with PW91/ZORA/TZ2P were in
general lower than those generated using LDA/DSPP/DNP or
B3LYP, consistent with our earlier work on the neutral donor
ligands.60The ligand induced shifts are very similar to the results
obtained with the other approaches, however, and systematic
differences between O- and N-donors were not calculated.
Additional insight into the potential interactions from anion
binding can be gained by examining the changes in the
calculated bond lengths and angles, which would also check
the internal consistency of the predicted stretching frequencies.
We selected the calculations performed using B3LYP/Stuttgart
RSC/6-31+g(d) for discussing relationships between calculated
bond lengths and frequencies, which are listed in Table 5, and
trends in bond lengths with varying ligation are depicted
graphically in Figure 8 (detailed structural parameters generated
using B3LYP with three different basis sets are contained in
Supplementary Tables S1- S12, and visual representations are
provided in Figures 8-10). As ligands are added, calculations
show that all distances within the uranyl coordination sphere
increase. The magnitude of the increase depends not only on
the nucleophilic strength of the different ligands but also on
their volumes, and the calculations provide a means to develop
a more quantitative assessment of the effect of ligand addition
to uranyl. The OdUdO bond length is represented by the lower
three traces in Figure 10, and the effect of the anion A, and
Figure 7. Uranyl ν3 frequencies calculated using LDA/DSPP/DNP
and PW91/ZORA/TZ2P, plotted versus IRMPD measurements. The
line represents the experimental data. The open squares were
generated using LDA/DSPP/DNP, and the triangles were generated
using PW91/ZORA/TZ2P. The black points represent [UO2A]+com-
plexes with no neutral donor, red corresponds to complexes with
O-coordinating neutrals, and blue corresponds to those with N-
Figure 8. Lowest-energy conformations of calculated for [UO2OH]+and its solvated complexes. Calculations were performed using the hybrid
B3LYP functional with the Stuttgart RSC/3-21g* basis set approach.
Discrete Uranyl Anion Complexes
J. Phys. Chem. A, Vol. 112, No. 3, 2008 517
subsequent addition of a neutral solvent S is very similar for
the acetate, the hydroxy, and the methoxide complexes. The
uranyl bond elongates by 0.042, 0.044, and 0.048 Å for OAc-,
OH-, and OCH3-, respectively. This is also the order of
increasing anion basicity, resulting in donation of more
electron density to the uranium atom, and attendant repulsion
of the axial oxygen atoms. The amount of donation was
quantified by performing charge analysis calculations with
PW91/ZORA/TZ2P. The Hirshfeld method shows donation to
uranyl of 0.56e (OAc), 0.53e (OH), and 0.65e (OCH3), and the
VDD method gives very similar values of 0.53, 052, and 0.63e,
The addition of a neutral donor to the uranyl anion complexes
induces a further lengthening of the uranyl bonds (Figure 10),
by 0.011-0.016 Å, with the largest shifts occurring for the more
basic neutrals: in this study, addition of ACO caused the largest
OdUdO elongation, consistent with the low uranyl frequency
measured in the IRMPD spectrum. The magnitude of the
elongation on addition of a neutral is much less than that
calculated for the initial attachment of the anion. Hirshfeld
analysis showed that the strongest donating neutral species,
acetone, donates 0.17 e (for one acetone) or 0.25 e (for
two acetones) to a uranyl hydroxide unit in the [UO2OH-
(ACO)1,2]+complexes (which have the greatest OdUdO elon-
gation); these values are significantly lower than those calculated
for the anions.
The calculations also predict that the uranium-anion distance
will be lengthened by addition of a neutral ligand, which is
shown by the middle three traces in Figure 10. Increases ranging
from 0.017 to 0.024 Å occurred in the U-OH length for the
hydroxide complexes, with the magnitude depending on the
basicity of the neutral, and the largest elongation being for
addition of ACO. When a second ACO is added, the U-OH
distance elongates by another 0.036 Å. The U-OCH3 bond
distance experiences very nearly identical increases upon
addition of H2O and NH3to [UO2OCH3]+. The U-anion distance
calculated for the acetate complex is substantially longer than
that for either the hydroxide or the methoxide, and the B3LYP
calculations indicated a bidentate-bound acetate, although
calculation using the LDA and the DNP basis set suggested a
monodentate structure. Using either approach, the U-acetate
length is nearly 0.27 Å longer than for the hydroxide or
methoxide. Addition of H2O or NH3 to the complex causes
elongation of the U-acetate bonds by ∼0.04 and 0.02 Å,
respectively. The fact that H2O produced a longer U-anion
elongation than NH3is contrary to what would be expected on
the basis of calculated coordination energies (NH3 ∼9 kcal/
mol greater than H2O)40but may be consistent with the fact
that uranyl behaves as a hard acid,6,40interacting more strongly
with the oxygen donors.
The U-S bond length in [UO2AS]+increased when NH3was
substituted for H2O, for all three anions studied. Further
U-neutral bond length comparisons involved only the hydroxide
and decreased in the order NH3> ACN > ACO, minimizing
at 2.27 Å. The trend correlates inversely with increasing ligand
nucleophilicity. Addition of a second ACO as the third equatorial
ligand in [UO2OH(ACO)2]+loosens the overall complex: the
U-neutral bond is lengthened by nearly 0.06 Å, and this is
accompanied by a lengthening of the U-OH bond by 0.036 Å,
and the OdUdO bond by nearly 0.01 Å. As the coordination
sphere is completed, distortions of the OdUdO angle from
linearity are lessened, and the value approaches 180°.
TABLE 5: Bond Lengths and OdUdO Bond Angles for
[UO2AS]+Complexes at the B3LYP Level of Theory and
Using the Stuttgart RSC/3-21g* Basis Set
n/a 1.700 n/an/a179.6
Figure 9. Lowest-energy conformations of [UO2OAc]+, [UO2OMe]+, and their solvent complexes with water and ammonia. Calculations were
performed using hybrid B3LYP functional with the Stuttgart RSC/3-21g* basis set approach.
518 J. Phys. Chem. A, Vol. 112, No. 3, 2008
Groenewold et al.
The structure of uranium complexes has been a persistent
topic of research in the chemical community because the
participation of 5f, 6d, and 7s orbitals offers a broad array of
possible structures and reaction pathways. The desire to
understand and then manipulate uranium chemistry has moti-
vated determined investigations of structure and bonding using
spectroscopy and computational chemistry. In principle, these
approaches should be highly complementary, but in practice,
results from each cannot be correlated with each other because
spectroscopy measurements on condensed-phase systems almost
always measure an ensemble of species, whereas calculations
produce data for single discrete species, and do not always
include specific and/or long-range interactions with solvent.
Consequently, it is difficult to use condensed-phase spectro-
scopic measurements to evaluate computational accuracy, which
is badly needed for molecules containing f elements. Infrared
spectra of gas-phase complexes generated using IRMPD provide
data for discrete species that are of great value for evaluating
ligand binding trends and computational chemistry results.
Much emphasis has been placed on the study of uranyl
dication complexes, and our prior IRMPD studies of ACO
complexes enabled comparison of antisymmetric OdUdO and
CdO frequencies with solution-phase measurements and com-
putational results.60However, at mid-pH ranges, uranyl-anion
pairs are more prevalent, and hence in the present study, IRMPD
of ion pairs involving hydroxide, acetate, and methoxide were
examined. The uranyl antisymmetric frequency values were red-
shifted equal to or greater than UO22+coordinated with four or
five neutral donor ligands. Although vibrational anharmonicity
no doubt contributes to these low-frequency values, the intrinsic
frequencies of the ion pair complexes are lower than expected
when compared with solution-phase measurements and with past
and present DFT results. The addition of a solvent neutral S to
the ion pairs did not result in systematic decreases in the ν3
values. But when frequencies for the [UO2AS]+species were
compared for differing neutrals, the ν3 value decreased with
increasing S nucleophilicity, consistent with theory, intuition,
and previous IRMPD measurements.60
The [UO2OCH3]+molecule underwent wavelength-specific
fragmentation reactions, eliminating the OCH3 radical at a
frequency 20 cm-1lower than fragmentations involving rear-
rangement and or loss of H atom(s). DFT modeling suggested
that the OCH3radical loss was activated by absorption at the
uranyl stretching frequency, and the H atom loss/rearrangement
eliminations were activated by absorption at the C-O stretching
frequency. Identifying the cause of this phenomenon remains
an outstanding task: IRMPD anharmonicity together with
absorption nonlinearities (as observed in the spectra of the para-
aminobenzoic acid cation58and [CeOH(ACO)n)3,4]2+cation95)
may contribute to the phenomenon; however, the very large
differences in the different photofragmentation channels suggests
that there may be another factor at work, such as promotion to
an excited-state electronic configuration, which would be
Acknowledgment. Work by G. S. G. and A. K. G. was
supported by the U.S. Department of Energy, Assistant Secretary
for Environmental Management, and the INL Laboratory
Directed Research & Development Program under DOE Idaho
Operations Office Contract DE-AC07-05ID14517. M.J.V.S was
supported in part through a grant from the U.S. National Science
Foundation (CAREER-0239800). Gaussian 03 calculations were
performed at the Wichita State University High-performance
Computing Center (HIPECC), a facility supported by the U.S.
National Science Foundation under Grant No. EIA-0216178 and
Grant No. EPS-0236913, with matching support from the State
of Kansas and HIPECC. W.A.d.J.’s research was performed,
in part, using the Molecular Science Computing Facility in the
William R. Wiley Environmental Molecular Sciences Labora-
tory, a national scientific user facility sponsored by the U.S.
Department of Energy’s Office of Biological and Environmental
Research located at the Pacific Northwest National Laboratory,
which is operated for the Department of Energy by Battelle.
The FOM authors, and authors from V.U.A., were supported
by the Nederlandse Organisatie voor Wetenschappelijk Onder-
zoek (Dutch National Science Foundation). The skillful as-
sistance by the FELIX staff, in particular Dr. B. Redlich, is
gratefully acknowledged. Construction and shipping of the
FTMS instrument was made possible through funding from the
National High Field FT-ICR Facility (grant CHE-9909502) at
the National High Magnetic Field Laboratory, Tallahassee, FL,
as was travel support for one of the INL authors; John Eyler’s
assistance was particularly invaluable.
Supporting Information Available: Complete listings of
structural data (frequencies, bond lengths and angles) calculated
using B3LYP with the Stuttgart RSC/D95V, and Stuttgart RSC/
3-21g* basis sets are provided in tables S1-S12. Geometries
of molecular structures in xyz format. This material is available
free of charge via the Internet at http://pubs.acs.org.
References and Notes
(1) Choppin, G. R.; Rizkalla, E. N. Solution Chemistry of Actinides
and Lanthanides. In Handbook on the Physics and Chemistry of Rare Earths;
Gschneider, J., K. A., Eyring, L., Choppin, G. R., Lander, G. H., Eds.;
Vol. 18, Lanthanides/Actinides: Chemistry; North-Holland: Amsterdam,
1994; pp 559.
(2) Denning, R. G. Struct. Bonding 1992, 79, 215.
(3) Pepper, M.; Bursten, B. E. Chem. ReV. 1991, 91, 719.
(4) Matsika, S.; Zhang, Z.; Brozell, S. R.; Blaudeau, J.-P.; Wang, Q.;
Pitzer, R. M. J. Phys. Chem. A 2001, 105, 3825.
(5) Silva, R. J.; Nitsche, H. Radiochim. Acta 1995, 70/71, 377
(6) Morse, J. W.; Choppin, G. R. ReV. Aquatic Sci. 1991, 4, 1.
(7) Toth, L. M.; Begun, G. M. J. Phys. Chem. 1981, 85, 547.
(8) Burgess, J. Metal Ions in Solution; Ellis Horwood Limited:
Chichester, U.K., 1978.
(9) Choppin, G. Radiochim. Acta 2004, 92, 519.
complexes. Values for unligated [UO2]2+and [UO2A]+complexes are
also included. Values were calculated using B3LYP functional and the
Stuttgart RSC/3-21g* basis set approach. (Note break in y axis at 2.08
Bond lengths plotted versus neutral for [UO2AS]+
Discrete Uranyl Anion Complexes
J. Phys. Chem. A, Vol. 112, No. 3, 2008 519
(10) Brookins, D. G. Geochemical Aspects of RadioactiVe Waste
Disposal; Springer-Verlag: New York, 1984.
(11) Rizkalla, E. N.; Choppin, G. R. Lanthanides and Actinides
Hydration and Hydrolysis. In Handbook on the Physics and Chemistry of
Rare Earths. Volume 18: Lanthanides/Actinides: Chemistry; Gschneidner,
J. K. A., Eyring, L., Choppin, G. R., Lander, G. H., Eds.; North-Holland:
New York, 1994; Vol. 18, pp 529.
(12) Nguyen-Trung, C.; Palmer, D. A.; Begun, G. M.; Peiffert, C.;
Mesmer, R. E. J. Solution Chem. 2000, 29, 101.
(13) Brooker, M. H.; Huang, C.-H.; Sylwestrowicz, J. J. Inorg. Nucl.
Chem. 1980, 42, 1431.
(14) Glebov, V. A. Koord. Khim. 1982, 8, 970.
(15) Nguyen-Trung, C.; Begun, G. M.; Palmer, D. A. Inorg. Chem. 1992,
(16) Jones, L. H. Spectrochim. Acta 1958, 10, 395.
(17) Quiles, F.; Burneau, A. Vibr. Spectrosc. 1998, 18, 61.
(18) Gal, M.; Goggin, P. L.; Mink, J. J. Mol. Struct. 1984, 114, 459.
(19) McGlynn, S. P.; Smith, J. K.; Neely, W. C. J. Chem. Phys. 1961,
(20) Jones, L. H.; Penneman, R. A. J. Chem. Phys. 1953, 21, 542.
(21) Best, S. P.; Clark, R. J. H.; Cooney, R. P. Inorg. Chim. Acta 1988,
(22) Clark, D. L.; Conradson, S. D.; Donohoe, R. J.; Keogh, D. W.;
Morris, D. E.; Palmer, P. D.; Rogers, R. D.; Tait, C. D. Inorg. Chem. 1999,
(23) Kakihana, M.; Nagumo, T.; Okamoto, M.; Kakihana, H. J. Phys.
Chem. 1987, 91, 6128.
(24) Mizuoka, K.; Ikeda, Y. Radiochim. Acta 2004, 92, 631.
(25) Burns, C. J.; Smith, C. C.; Sattelberger, A. P.; Gray, H. B. Inorg.
Chem. 1992, 31, 3724.
(26) Morris, D. E.; Chisholm-Brause, C. J.; Barr, M. E.; Conradson, S.
D.; Gary Eller, P. Geochim. Cosmochim. Acta 1994, 58, 3613.
(27) Duff, M. C.; Coughlin, J. U.; Hunter, D. B. Geochim. Cosmochim.
Acta 2002, 66, 3533.
(28) Tellez Soto, C. A.; Arissawa, M.; Gomez, L., J.; Mondragon, M.
A. Polyhedron 2000, 19, 2353.
(29) Madic, C.; Hobart, D. E.; Begun, G. M. Inorg. Chem. 1983, 22,
(30) Glebov, V. A. Koord. Khim. 1981, 7, 388.
(31) Mizuoka, K.; Ikeda, Y. Inorg. Chem. 2003, 42, 3396.
(32) Basile, L. J.; Sullivan, J. C.; Ferraro, J. R.; LaBonville, P. Appl.
Spectrosc. 1974, 28, 142.
(33) Schreckenbach, G.; Hay, P. J.; Martin, R. L. J. Comput. Chem.
1999, 20, 70.
(34) Vetere, V.; Maldivi, P.; Adamo, C. J. Comput. Chem. 2003, 24,
(35) Van Wullen, C. J. Comput. Chem. 1999, 20, 51.
(36) Kaltsoyannis, N. Chem. Soc. ReV. 2003, 32, 9.
(37) de Jong, W. A.; Harrison, R. J.; Nichols, J. A.; Dixon, D. A. Theor.
Chem. Acc. 2001, 107, 22.
(38) Clavaguera-Sarrio, C.; Ismail, N.; Marsden, C. J.; Begue, D.;
Pouchan, C. Chem. Phys. 2004, 302, 1.
(39) Hay, P. J.; Martin, R. L.; Schreckenbach, G. J. Phys. Chem. A 2000,
(40) Clavaguera-Sarrio, C.; Hoyau, S.; Ismail, N.; Marsden, C. J. J. Phys.
Chem. A 2003, 107, 4515.
(41) Farkas, I.; Banyai, I.; Szabo, Z.; Wahlgren, U.; Grenthe, I. Inorg.
Chem. 2000, 39, 799.
(42) Gutowski, K. E.; Dixon, D. A. J. Phys. Chem. A 2006, 110, 8840.
(43) Gutowski, K. E.; Cocalia, V. A.; Griffin, S. T.; Bridges, N. J.;
Dixon, D. A.; Rogers, R. D. J. Am. Chem. Soc. 2007, 129, 526.
(44) Duncan, M. A. Int. ReV. Phys. Chem. 2003, 22, 407.
(45) Lemaire, J.; Boissel, P.; Heninger, M.; Mauclaire, G.; Bellec, G.;
Mestdagh, H.; Simon, A.; Le Caer, S.; Ortega, J. M.; Glotin, F.; Maitre, P.
Phys. ReV. Lett. 2002, 89.
(46) Moore, D. T.; Oomens, J.; Eyler, J. R.; Meijer, G.; von Helden,
G.; Ridge, D. P. J. Am. Chem. Soc. 2004, 126, 14726.
(47) Oepts, D.; van der Meer, A. F. G.; van Amersfoort, P. W. Infrared
Phys. Technol. 1995, 36, 297.
(48) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C.
M. Science 1989, 246, 64.
(49) Schroder, D.; Roithova, J.; Schwarz, H. Int. J. Mass Spectrom. 2006,
(50) Stewart, I. I. Spectrochim. Acta, Part B 1999, 54, 1649.
(51) Gatlin, C. L.; Turecek, F. Electrospray Ionization of Inorganic and
Organometallic Complexes. In Electrospray Ionization Mass Spectrometry;
John Wiley & Sons: New York, 1997; pp 527.
(52) Chien, W.; Anbalagan, V.; Zandler, M.; Hanna, D.; Van, Stipdonk,
M.; Gresham, G.; Groenewold, G. J. Am. Soc. Mass Spectrom. 2004, 15,
(53) Van Stipdonk, M.; Anbalagan, V.; Chien, W.; Gresham, G.;
Groenewold, G.; Hanna, D. J. Am. Soc. Mass Spectrom. 2003, 14, 1205.
(54) Van Stipdonk, M. J.; Chien, W.; Angalaban, V.; Bulleigh, K.;
Hanna, D.; Groenewold, G. S. J. Phys. Chem. A 2004, 108, 10448.
(55) Van Stipdonk, M. J.; Chien, W.; Bulleigh, K.; Wu, Q.; Groenewold,
G. S. J. Phys. Chem. A 2006, 110, 959.
(56) Oomens, J.; Sartakov, B. G.; Meijer, G.; von Helden, G. Int. J.
Mass Spectrom. 2006, 254, 1.
(57) Moore, D. T.; Oomens, J.; van der Meer, L.; von Helden, G.; Meijer,
G.; Valle, J.; Marshall, A. G.; Eyler, J. R. Chem. Phys. Chem. 2004, 5,
(58) Oomens, J.; Moore, D. T.; Meijer, G.; von Helden, G. Phys. Chem.
Chem. Phys. 2004, 6, 710.
(59) Oomens, J.; Moore, D. T.; von Helden, G.; Meijer, G.; Dunbar, R.
C. J. Am. Chem. Soc. 2004, 126, 724.
(60) Groenewold, G. S.; Gianotto, A. K.; Cossel, K. C.; Van, Stipdonk,
M. J.; Moore, D. T.; Polfer, N.; Oomens, J.; de Jong, W. A.; Visscher, L.
J. Am. Chem. Soc. 2006, 107, 4802.
(61) Sachs, S.; Brendler, V.; Geipel, G. Radiochim. Acta 2007, 95, 103.
(62) Valle, J. J.; Eyler, J. R.; Oomens, J.; Moore, D. T.; van der Meer,
A. F. G.; von Helden, G.; Meijer, G.; Hendrickson, C. L.; Marshall, A. G.;
Blakney, G. T. ReV. Sci. Instrum. 2005, 76, 023103.
(63) Marshall, A. G.; Wang, T.-C. L.; Ricca, T. L. J. Am. Chem. Soc.
1985, 107, 7893.
(64) Bagratashivili, V. N.; Letokov, V. S.; Makarov, A. A.; Ryabov, E.
A. Mutiple Photon Infrared Laser Photophysics and Photochemistry;
Harwood: Chur, Switzerland, 1985.
(65) Oomens, J.; Tielens, A. G. G. M.; Sartakov, B. G.; Von, Helden,
G.; Meijer, G. Astrophys. J. 2003, 591, 968.
(66) Marshall, A. G. Acc. Chem. Res. 1985, 18, 316.
(67) Marshall, A. G.; Hendrickson, C. L.; Jackson, G. S. Mass Spectrom.
ReV. 1998, 17, 1.
(68) Delley, B. J. Chem. Phys. 1990, 92, 508.
(69) Delley, B. J. Chem. Phys. 2000, 113, 7756.
(70) Apra `, E.; et.al. NWChem, A Computational Chemistry Package
for Parallel Computers, Version 4.7 ed.; Pacific Northwest National
Laboratory: Richland, WA, 99352-0999, U.S.A., 2005.
(71) Kendall, R. A.; Apra, E.; Bernholdt, D. E.; Bylaska, E. J.; Dupuis,
M.; Fann, G. I.; Harrison, R. J.; Ju, J.; Nichols, J. A.; Nieplocha, J.;
Straatsma, T. P.; Windus, T. L.; Wong, A. T. Comput. Phys. Commun.
2000, 128, 260.
(72) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;
Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M.
W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.;
Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussion 98, revision A.4;
Gaussian, Inc.: Pittsburgh, PA, 1998.
(73) te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Guerra, C. F.;
Van, Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. J. Comput. Chem.
2001, 22, 931.
(74) Slater, J. C. Phys. ReV. Lett. 1951, 81, 385.
(75) Vosko, S. J.; Wilk, W.; Nusair, M. Can. J. Phys. 1980, 58, 1200.
(76) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(77) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(78) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.;
Pederson, M. R.; Singh, D. J.; Fiolhais, C. 1992, 46, 6671.
(79) Delley, B. Int. J. Quantum Chem. 1998, 69, 423.
(80) Ortiz, J. V.; Hay, P. J.; Martin, R. L. J. Am. Chem. Soc. 1992, 114,
(81) Bergner, A.; Dolg, M.; Kuchle, W.; Stoll, H.; Preuss, H. Mol. Phys.
1993, 80, 1431.
(82) Dolg, M. Modern Methods and Algorithms of Quantum Chemistry;
John von Neumann Institute for Computing: Julich, Germany, 2000; Vol.
(83) Dunning, T. H., Jr.; Hay, P. H. Modern Theoretical Chemistry;
Plenum: New York, 1976; Vol. 3.
(84) Kuchle, W.; Dolg, M.; Stoll, H.; Preuss, H. Mol. Phys. 1991, 74,
(85) Dunning, T. H. J. Chem. Phys. 1989, 90, 1007.
(86) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1993,
(87) Chaban, G. M.; Jung, J. O.; Gerber, R. B. J. Chem. Phys. 1999,
(88) Matsunaga, N.; Chaban, G. M.; Gerber, R. B. J. Chem. Phys. 2002,
(89) Yagi, K.; Hirao, K.; Taketsugu, T.; Schmidt, M. W.; Gordon, M.
S. J. Chem. Phys. 2004, 121, 1383.
(90) Hirshfeld, F. L. Theor. Chim. Acta 1977, 44, 129.
520 J. Phys. Chem. A, Vol. 112, No. 3, 2008
Groenewold et al.
(91) Fonseca Guerra, C.; Handgraaf, J.-W.; Baerends, E. J.; Bickelhaupt,
F. M. J. Comput. Chem. 2004, 25, 189.
(92) Soderholm, L.; Skanthakumar, S.; Neuefeind, J. Anal. Bioanal.
Chem. 2005, 383, 48.
(93) Neuefeind, J.; Soderholm, L.; Skanthakumar, S. J. Phys. Chem. A
2004, 108, 2733.
(94) Quiles, F.; Burneau, A. Vibr. Spectrosc. 2000, 23, 231.
(95) Groenewold, G. S.; Gianotto, A. K.; Cossel, K. C.; Van Stipdonk,
M. J.; Oomens, J.; Polfer, N.; Moore, D. T.; de Jong, W. A.; McIlwain, M.
E. Phys. Chem. Chem. Phys. 2006, 9, 596.
(96) Kasha, M. J. Chem. Phys. 1949, 17, 349.
(97) Lias, S. G. NIST Chemistry WebBook; United States Department
of Commerce, National Institute of Standards and Technology, 2003; Vol.
(98) de Jong, W. A.; Apra, E.; Windus, T. L.; Nichols, J. A.; Harrison,
R. J.; Gutowski, K. E.; Dixon, D. A. J. Phys. Chem. A 2005, 109,
(99) Rodgers, M. T.; Armentrout, P. B. Mass Spectrom. ReV. 2000, 19,
Discrete Uranyl Anion Complexes
J. Phys. Chem. A, Vol. 112, No. 3, 2008 521