Addition of H2O and O2 to acetone and dimethylsulfoxide ligated uranyl(V) dioxocations

Article abstract of DOI:10.1021/jp807651c

Gas-phase complexes of the formula [UO₂(lig)]+ (lig = acetone (aco) or dimethylsulfoxide (dmso)) were generated by electrospray ionization (ESI) and studied by tandem ion-trap mass spectrometry to determine the general effect of ligand charge d

Full text of DOI:10.1021/jp807651c

2350
J. Phys. Chem. A 2009, 113, 2350–2358
Addition of H2O and O2 to Acetone and Dimethylsulfoxide Ligated Uranyl(V) Dioxocations
Christopher M. Leavitt,† Vyacheslav S. Bryantsev,‡ Wibe A. de Jong,§ Mamadou S. Diallo,‡
William A. Goddard III,‡ Gary S. Groenewold,*,| and Michael J. Van Stipdonk*,†
Department of Chemistry, Wichita State UniVersity, Wichita, KS, Materials and Process Simulation Centre,
Beckman Institute 139-74, California Institute of Technology, Pasadena, CA, William R. Wiley EnVironmental
Molecular Sciences Laboratory, Pacific Northwest National Laboratory, Richland, WA, and Interfacial
Chemistry Group, Idaho National Laboratory, Idaho Falls, ID
ReceiVed: August 27, 2008; ReVised Manuscript ReceiVed: December 19, 2008
Gas-phase complexes of the formula [UO2(lig)]+ (lig ) acetone (aco) or dimethylsulfoxide (dmso)) were
generated by electrospray ionization (ESI) and studied by tandem ion-trap mass spectrometry to determine
+
the general effect of ligand charge donation on the reactivity of UO2 with respect to water and dioxygen.
The original hypothesis that addition of O2 is enhanced by strong σ-donor ligands bound to UO2+ is supported
by results from competitive collision-induced dissociation (CID) experiments, which show near exclusive
loss of H2O from [UO2(dmso)(H2O)(O2)]+, whereas both H2O and O2 are eliminated from the corresponding
[UO2(aco)(H2O)(O2)]+ species. Ligand-addition reaction rates were investigated by monitoring precursor and
product ion intensities as a function of ion storage time in the ion-trap mass spectrometer: these experiments
suggest that the association of dioxygen to the UO2+ complex is enhanced when the more basic dmso ligand
was coordinated to the metal complex. Conversely, addition of H2O is favored for the analogous complex ion
that contains an aco ligand. Experimental rate measurements are supported by density function theory
+
calculations of relative energies, which show stronger bonds between UO2 and O2 when dmso is the
coordinating ligand, whereas bonds to H2O are stronger for the aco complex.
Introduction
cation [UO2]+ will bind O2 in the gas phase.23 The tendency to
add O2 to UO2 complexes was dependent on the number and
+
Uranium-oxygen species and their intrinsic reactivity are of
interest because of their importance to the nuclear energy
industry and because their chemistry is fundamental to our
understanding of the migration and fate of uranium in the
environment and to the development of efficient and effective
methods for separation from fission products and other
actinides.1-5 Speciation of uranium is diverse, and includes
multiple oxidation states, oxide forms, and multiuranium
clusters: each species can exhibit variable reactivity, stability,
and solubility.6,7
identity of associated ligands. For example, no binding of O2
was observed for the bare cation but efficient addition occurred
when one or more aco or H2O ligands occupied equatorial
coordination sites. The tendency to add O2 appeared to be
influenced by the amount of charge donated from ligand(s) to
the uranium metal center. Serial dissociation reactions of the
[UO2(aco)3(O2)]+ complex also showed an unusual pattern in
which initial elimination of neutral acetone was followed by
loss of O2, even though there were still σ-donor aco ligands
available for elimination. At the time, it was hypothesized that
binding of oxygen caused oxidization of the U(V) center to
create a U(VI)-superoxide complex, the formation and stability
of which was in some way influenced by the number of donor
ligands. A subsequent computational study illustrated that
binding likely involves formation of a novel two-electron, three-
centered bond between [UO2]+ and O2-, with significant
interaction between the uranyl 5fꢀ and the O2 π*xy orbitals.24
In the present study, we investigated the association reactions
of H2O and O2 to formally U(V) complexes [UO2(aco)]+ and
[UO2(dmso)]+. The aco and dmso species were chosen because
the two ligands contain the same number of atoms and hence
have the same number of vibrational degrees of freedom (DOF)
but differ significantly in their nucleophilicity and gas-phase
basicity. Controlling for the number of vibrational DOF was
important in this study, as this can influence the rates of gas-
phase ligand addition and dissociation reactions by absorbing
and accommodating reaction energy. With respect to the
σ-donating ligands, the U(V) species tends to behave as a Lewis
acid but toward O2 the reactive U(V) complex donates electron
density from the metal center to the ligand, thus acting more
like a Lewis base. By examining the intrinsic reactivity of
Fundamental reactions between uranium and (molecular) O2
have been investigated in the gas and condensed phases. For
example, it was shown that gas-phase U+ reacts with O2 to yield
UO+; this cation reacts with a second O2 molecule to form the
+
reduced uranyl cation UO2 , formally a U(V) species.8 A similar
reaction of U2+ with O2 and other oxidants yielded the (doubly
2+
charged) UO2+ and the UO2 cations.9,10 In the condensed
phase, O2 addition to uranyl centers generally occurs to furnish
η2-peroxo species, and there are examples of materials in which
uranyl molecules (UO22+) are bridged by peroxide in a µ-dioxo
mode.11-17
The gas-phase experiments involving the uranyl cation have
proven particularly useful for determining specific reactivity with
respect to oxidation state, coordination number, and degree of
donation from ligands.17-22 In a previous report by our group,
it was shown that complexes containing the reduced uranyl
* To whom correspondence should be addressed. E-mail:
mike.vanstipdonk@wichita.edu (M.J.V.S.), gary.groenewold@inl.gov (G.S.G.).
† Wichita State University.
‡ California Institute of Technology.
§ Pacific Northwest National Laboratory.
| Idaho National Laboratory.
10.1021/jp807651c CCC: $40.75
2009 American Chemical Society
Published on Web 02/13/2009

Uranyl(V) Dioxocations
J. Phys. Chem. A, Vol. 113, No. 11, 2009 2351
Figure 1. Mass spectra of isolated [UO2(aco)]+ reacting in a mixed
atmosphere containing H2O and O2 for (a) 10 ms, (b) 100 ms, and (c)
1000 ms.
Figure 2. Mass spectra of isolated [UO2(dmso)]+ reacting in a mixed
atmosphere containing H2O and O2 for (a) 10 ms, (b) 100 ms, and (c)
1000 ms.
caused precipitation of BaSO4, and the resulting aqueous UO2I2
solution was separated by decantation. Spray solutions for the
ESI experiments were generated by spiking 1 mL of UO2I2
solution with 5.00 or 3.00 µL of aco or dmso, respectively.
Solutions were infused into the ESI-MS instrument via a syringe
pump maintained at a flow rate of 3-5 µL/min.
The atmospheric pressure ionization stack settings for the
LCQ instrument (lens voltages, quadrupole, octapole voltage
offsets, etc.) were optimized for maximum ion transmission to
the ion trap mass analyzer using the autotune routine within
the LCQ program. The spray needle voltage was maintained at
+5 kV and the N2 sheath gas flow at 25 units (arbitrary to the
LCQ instrument, corresponding to approximately 0.375 L/min).
As in previous studies, the heated capillary was maintained at
a relatively high temperature of 200 °C to facilitate elimination
of neutral iodine to furnish the [UO2(aco)]+ and [UO2(dmso)]+
complexes.23 The ion trap analyzer was operated at a pressure
of ∼1.5 × 10-5 Torr. A helium bath/buffer gas was used to
improve trapping efficiency and functioned as a collision gas
for CID experiments.
[UO2(aco)]+ and [UO2(dmso)]+, we were able to investigate in
a more systematic way the role of σ-donor basicity on the
tendency to add either H2O or O2 and test the hypothesis that
addition of the latter is enhanced by inclusion of strong σ-donors
+
bound to the UO2 center.
Experimental measurements of ligand addition rates, ligand
elimination, and ligand exchange were performed using tandem
(quadrupole) ion trap mass spectrometry. Precursor aco and
dmso-coordinated U(V) dioxocations were generated by ESI.
Comprehensive density functional theory calculations were used
to predict conformations of relevant species complexes, to
determine ligand binding energies, and to predict vibrational
features.
Experimental Methods
Electrospray Ionization Mass Spectrometry (ESI-MS). ESI
mass spectra were generated using a Finnigan LCQ-Deca ion
trap mass spectrometer (ThermoFinnigan, now Thermo Scien-
tific Corp., San Jose CA). In a previous ESI study of O2 addition
+
to UO2 complexes, we relied upon harsh ESI conditions to
Ligand addition reactions of H2O and/or O2 were monitored
+
cause reduction of uranyl ion. In the present experiments, we
by the isolation and storage of aco or dmso ligated UO2 ion
+
enhanced formation of the UO2 complexes by using spray
in the ion trap. The H2O is present in the ion trap because of its
use as a spray solvent. The O2 is present in significant partial
pressures within the ion trap because ESI is an atmospheric
ionization method. Previous investigations have shown that both
H2O and O2 are present at partial pressures sufficient to produce
pseudo-first-order reaction conditions.23 Addition of O2 has been
confirmed in the past using both He gas seeded with O2 and
18O labeled gas. Because of the relatively low concentrations
of aco and dmso used to make the ESI spray solutions,
association reactions to produce aco or dmso adducts were not
observed.
solutions composed in part of UO2I2 in deionized H2O. Because
of the disparate ionization energies and electron affinities of
UO2 and I-, collisions in the ion transfer region of the ESI
+
source tend to cause reductive elimination of I from
[UO2I(aco)n]+ or [UO2I(dmso)n]+ to furnish [UO2(aco)n]+ or
[UO2(dmso)n]+, respectively.
The solution of UO2I2 was created by dissolving 0.05 g of
UO3 in approximately 1-2 µL of H2SO4 followed by gentle
heating for approximately 24 h. The UO2SO4 solution thus
produced was diluted with 10 mL of deionized H2O. The
UO2SO4 solution was added to 10 mL of a 5 mM solution of
BaI2 in deionized H2O. A combination of the two solutions
The multiple-stage CID and gas-phase association reaction
experiments were performed using previously established

2352 J. Phys. Chem. A, Vol. 113, No. 11, 2009
Leavitt et al.
Figure 5. Kinetic profile of association reactions from [UO2(dmso)]+.
of time. Relative abundance values for reactant and product ions
at each reaction time were calculated by dividing the intensity
of each individual ion by the total ion intensity. Experimentally
determined relative ion intensities are reproducible to better than
20% relative standard deviation.
Figure 3. Mass spectra acquired after CID of [UO2(lig)(H2O)(O2)]+
produced from condensation of [UO2(lig)]+ with O2 and H2O, where
lig is (a) acetone and (b) dimethylsulfoxide.
Kinetic Modeling. Reaction rates were modeled, and rate
constants calculated, using the Berkeley Madonna software
package that utilizes a numerical Runga-Kutta integration
algorithm to solve the differential equations based on the
reactions of the system.25 The program uses the downhill
simplex method as documented in Numerical Methods in C for
curve fitting.26 The model used for this fit was based on a
pseudo-first-order approximation of the reactions with respect
to the oxygen and water concentrations. The root-mean-square
(rms) error was used to measure the degree of deviation of the
model from each experimental data point, which was expressed
as a percentage of the average ion abundance over the range of
time in the experiment. Previous kinetics experiments by our
group have shown that rms errors ranging from 1 to 30%
provides acceptable agreement between the experiment and
model.23
Density Functional Theory Calculations. We used the
relativistic effective core potential (RECP) density functional
theory (DFT) to calculate the geometries, binding energies, and
+
frequencies for complexes containing the U(V) UO2 cation.
Our recent work24 has suggested that the hybrid B3LYP
functional27,28 may be underestimating the binding energy of
dioxygen to uranyl(V) systems relative to the binding energy
of acetone ligands. In the quest for more accurate computational
methods for calculating U(V)-ligand bond dissociation energies,
we have tested the newly developed M06-class functionals.29,30
For the smallest U(VI)-superoxo adduct, UO2+-O2(η2), we also
performed coupled-cluster calculations with singles, doubles,
and perturbative triples excitations (CCSD(T)).31-33 Binding
energy values calculated using the M06-L local functional
(without spin-orbit corrections) provided agreement with the
experimental trends (below), and hence the M06-L functional
Figure 4. Kinetic profile of association reactions from [UO2(aco)]+.
The inset shows an expanded plot of the kinetic profile for reaction
times up to 0.3 s.
procedures.18-22 Values for the salient instrumental parameters
were: a) an isolation width between 1.0 and 3.0, b) an activation
Q of 0.300 (used to adjust the qz value for the resonant excitation
of the precursor ion), and c) an activation amplitude ranging of
0-20% (of 5V). The reactant cations (either [UO2(aco)]+ or
[UO2(dmso)]+ were isolated in the ion trap, where they
underwent ligand addition reactions with adventitious H2O and
O2 that were also present in the ion trap. Ion reaction times
were varied from 1 to 5000 ms to observe the temporal variation
in the distribution of the product ions. Kinetic plots were
generated by measuring ion intensities at discrete time intervals
and plotting relative abundance values for each ion as a function
+
was used for comparative UO2 (lig) binding energy calculations
in the present study. For comparison purposes, the results
obtained with the B3LYP functional are included as Supporting
Information in Tables 1S-3S.
The geometries of all complexes were optimized with the
standard Stuttgart small-core (SSC) RECP34 and the correspond-

Uranyl(V) Dioxocations
J. Phys. Chem. A, Vol. 113, No. 11, 2009 2353
SCHEME 1: Main Reaction Pathway for the Addition of H2O and/or O2 to [UO2(aco)]+
TABLE 1: Forward Rate Constant Values for the Addition
the heterogeneous binary species [UO2(aco)(H2O)(O2)]+ at a m/z
of H2O and/or O2 to [UO2(aco)]+
of 378. Ions at m/z 346 and 360, formed by addition of a single
H2O or O2 molecule respectively, were present at lower relative
abundance as was a small peak m/z of 396 that corresponds to
[UO2(aco)(H2O)2(O2)]+. [UO2(aco)(H2O)(O2)]+ became the base
peak at 100 ms reaction time (part b of Figure 1), and the
intensity of the [UO2(aco)(H2O)2(O2)]+ ion increased to ca. 25%.
At the longest reaction time (1000 ms, part c of Figure 1), the
base peak was [UO2(aco)(H2O)2(O2)]+ at m/z of 396. No further
addition of either H2O or O2 was observed for reaction times
extended out to 5000 ms, suggesting the m/z 396 ion represents
a UO2+ complex for which all equatorial coordination sites are
occupied. Note also that addition of a second O2 to the [UO2]+
complexes was not observed, which is consistent with the notion
that O2 addition is a redox process that cannot occur once
uranium has been converted to the VI oxidation state.23 The
reactivity exhibited by [UO2(aco)]+ in the present study is fully
consistent with earlier investigations of related UO2+ complexes
by our group.23
Figure 2 shows the corresponding product ion spectra
generated by isolation and reaction of [UO2(dmso)]+ for 10 ms
(2a), 100 ms (2b), and 1000 ms (2c). The experiments with
[UO2(dmso)]+ were run directly after those with [UO2(aco)]+
to ensure that both species experienced the same (gas phase)
reaction conditions. After 10 ms reaction time (part a of Figure
2), the base ion in the mass spectrum corresponds to formation
of [UO2(dmso)(O2)]+ at m/z of 380. Direct addition of O2
appears to be favored for [UO2(dmso)]+ over [UO2(aco)]+ and
is especially apparent when the relative intensities of [UO2-
(lig)(O2)]+ and [UO2(lig)(H2O)]+ are compared for the aco and
dmso precursor ions. The nearly exclusive addition of O2 to
[UO2(dmso)]+ is consistent with the hypothesis that the basicity
of the σ-donor ligand significantly influences the tendency to
add O2. Other ions observed after the imposed 10 ms reaction
time include those formed by addition of H2O, (O2 and H2O),
and (O2 and 2 H2O) at m/z values of 366, 398, and 416,
respectively. After a reaction time of 100 ms (part b of Figure
2), the [UO2(dmso)]+ precursor ion was depleted to less than
5% relative intensity, and the base peak in the spectrum was
the heterogeneous binary adduct, [UO2(dmso)(H2O)(O2)]+, at
m/z 398. The dominant peak observed after 1000 msec reaction
time (part c of Figure 2) was [UO2(dmso)(H2O)2(O2)]+ at m/z
of 416, and no further addition of either H2O or O2 ligands was
observed for extended reaction times. This was directly analo-
gous to the final product formed from [UO2(aco)]+ by associa-
tion reactions.
rate
kn values
constant, (cm3 s-1
% efficiency
forward reaction
kn
molecule-1) (relative to kADO
)
[UO2(aco)]+ + H2O
k1
k2
k3
k4
k5
6.36 × 10-10
3.74 × 10-10
3.87 × 10-10
1.39 × 10-10
2.13 × 10-09
30
69
18
26
100
[UO2(aco)(H2O)]+ + O2
[UO2(aco)(H2O)(O2)] + H2O
[UO2(aco)]+ + O2
[UO2(aco)(O2)]+ + H2O
ing basis set on U and the standard 6-311G++(d,p) basis set
with diffuse functions on the light atoms. In addition, single-
point energy calculations were performed using more extended
basis sets, namely the SSC basis set augmented with the two
g-functions of the Stuttgart large-core basis set on U (SSC(2g))
and the 6-311++G(3df,3pd) basis set on other elements. Control
+
calculations for complexes of UO2 with H2O and O2 show
(Table 1S of the Supporting Information) that this procedure
leads to small differences in binding energies (<0.35 kcal/mol)
when compared to results obtained after full optimization at
the B3LYP/SSC(3g,1h)/aug-cc-pVTZ level. Full geometry
optimization of UO2+-O2(η2) was also performed at the
CCSD(T)/SSC(2g)/6-311++G(3df,3dp) level, including all
electrons in the correlation treatment.
Electronic structure calculations were performed with the
NWChem program package35 and were based on the unrestricted
formalism. Vibrational frequencies were computed numerically
at the B3LYP/SSC/6-311++G** level. The standard Gibbs free
energy of each species in the gas phase was calculated using
the standard rigid rotor-harmonic oscillator approximation
without scaling.
Results and Discussion
Gas-Phase Addition Reactions of H2O and O2 to [UO2(aco)]+
and [UO2(aco)]+. Competitive addition reactions of H2O and
O2 to [UO2(aco)]+ were conducted by selectively isolating the
relevant precursor ion using a notched rf waveform. It is
important to note that in these experiments only the ion selected,
in this case [UO2(aco)]+, is stored in the ion trap and all other
potential reacting species are resonantly ejected from the ion
trap prior to the imposed reaction period. Figure 1 shows the
temporal evolution of the product ions in spectra generated by
the isolation and storage, without imposed collisional activation,
of [UO2(aco)]+ for 10 ms (1a), 100 ms (1b), and 1000 ms (1c).
The appearance of product ions with m/z values higher than
that of the precursor ion signals formation of adducts by
ion-molecule association reactions that occur during the
imposed isolation/storage time.
Comparison of ion dissociation patterns also clearly showed
stronger O2 binding by the dmso-containing complex. Figure 3
shows CID spectra derived from [UO2(lig)(H2O)(O2)]+, which
was initially formed by condensation of [UO2(lig)]+ with O2
and H2O for lig ) aco (3a) and dmso (3b), respectively. For
After 10 ms reaction time (part a of Figure 1), the base peak
in the product ion spectrum was unreacted [UO2(aco)]+ precur-
sor at m/z 328. The most abundant adduct ion at this time was

2354 J. Phys. Chem. A, Vol. 113, No. 11, 2009
Leavitt et al.
SCHEME 2: Main Reaction Pathway for the Addition of H2O and/or O2 to [UO2(dmso)]+
TABLE 2: Forward Rate Constant Values for the Addition
O2 binding is enhanced by the presence of the more basic
σ-donor ligand.
of H2O and/or O2 to [UO2(dmso)]+
rate
kn values
(cm3 s-1
The multiple-stage isolation and reaction capabilities of the
ion trap also allowed for an examination of displacement
reactions for adducts of the [UO2(aco)]+ and [UO2(dmso)]+
cations. For these experiments, the [UO2(aco)(H2O)]+, [UO2-
(dmso)(H2O)]+, [UO2(aco)(O2)]+, and [UO2(dmso)(O2)]+ ions
were selectively isolated and allowed to react with H2O and O2
in the ion trap for a period of 30 ms. The resulting product ion
spectra are provided in Figure S1 of the Supporting Information.
General reaction pathways observed included:
constant,
% efficiency
(relative to kADO)
forward reaction
kn
molecule-1
)
[UO2(dmso)]+ + O2
k1
k2
k3
4.12 × 10-10
1.30 × 10-09
2.63 × 10-10
76
61
12
[UO2(dmso)(O2)]+ + H2O
[UO2(dmso)(H2O)(O2)]+ + H2O
TABLE 3: Electronic Binding Energies (∆Ee) and Bond
Distances for the Uranyl(V) Superoxo Complex UO2+O2(η2)
Calculated with CCSD(T) and Several Density Functional
Theory (DFT) Methodsa
[UO2(lig)(H2O)]++ O2 f [UO2(lig)(H2O)(O2)]+ (1)
[UO (lig)(O )]++ H O f [UO (lig)(H O) O ]+ (2)
[UO2(lig)(H2O)]++ O2 f [UO2(lig)(O2)]+ + H2O (3)
[UO2(lig)(O2)]++ H2O f [UO2(lig)(H2O)]+ + O2 (4)
method
∆Ee, kJ/mol
UdO, Å
U · · ·O, Å
O-O, Å
M06
-63.4
-81.3
-101.7
-143.2
-145.5
-154.2
-222.3
1.756
1.738
1.735
1.756
1.756
1.751
1.741
2.302
2.293
2.334
2.304
2.302
2.224
2.256
1.294
1.290
1.275
1.294
1.294
1.314
1.276
B3LYPb
M06-L
PBEb
(
)
2
2
2
2
2
2
PW91b
CCSD(T)c
LDAb
For all four precursor ions, the dominant product ion had the
formula [UO2(lig)(H2O)(O2)]+ (reactions 1 and 2) with the
species [UO2(lig)(H2O)2(O2)]+ also observed at a relative
abundance of ∼10%. [UO2(lig)(H2O)]+ also reacted via dis-
placement of H2O by O2 (reaction 3) in both the acetone and
dmso complexes, with the latter complex undergoing this
a Optimized with the SSC/6-311++G** basis set followed by
single point energy calculations with the SSC(2g)/6-311++G-
(3df,3dp) basis set. b Ref 24. c Optimized with the SSC(2g)/
6-311++G(3df,3dp) basis set.
these experiments, the respective binary adduct ions were first
synthesized from [UO2(lig)]+ using ion-molecule association
reactions, then isolated, and finally dissociated using collisional
activation. For both the aco and dmso-containing complexes,
the precursor ion remained the base peak in the CID spectrum,
a result we attribute to rapid association reactions that the
fragmentation products participate in on the time scale of this
experiment. However, despite the influence on ion intensity
distributions by the association reactions, it is clear from the
spectrum in part a of Figure 3 that CID of [UO2(aco)(H2O)(O2)]+
generated [UO2(aco)(O2)]+, [UO2(aco)(H2O)]+, and [UO2(aco)]+
through competitive elimination of H2O, O2, or both ligands.
The [UO2(aco)(H2O)2]+ and [UO2(aco)(H2O)(O2)]+ species were
generated by H2O addition reactions to the [UO2(aco)(O2)]+ and
[UO2(aco)(H2O)]+ fragment ions. A comparison of relative peak
intensities shows that loss of H2O is competitive with loss of
O2, with the former favored by a factor of ∼1.7:1.
For the dmso complex, CID of [UO2(dmso)(H2O)(O2)]+ (part
b of Figure 3), loss of O2 is not competitive with the loss of
H2O. [UO2(dmso)(O2)]+ was produced by exclusive elimination
of the H2O ligand, and the fragment ion accepted one and two
H2O ligands through association reactions to regenerate
[UO2(dmso)(H2O)(O2)]+ and form [UO2(dmso)(H2O)2(O2)]+.
Thus, the comparison of the dissociation spectra shown in Figure
3 reinforces the idea that the precursor ion that contains dmso
more effectively binds the O2 ligand than does the analogous
complex with aco, a result consistent with the hypothesis that
Figure 6. Structures of complexes [UO2(ACO)(H2O)0-2]
+ and [UO2-
(ACO)(H2O)0-2(O2)]+ optimized at the M06-L/SSC/6-311+G** level
of theory.

Uranyl(V) Dioxocations
J. Phys. Chem. A, Vol. 113, No. 11, 2009 2355
fitting the experimental curves, were compared with those
obtained using average dipole orientation theory36-39 to calculate
reaction efficiencies. Table 1 shows the values obtained for the
forward rate constants designated in Scheme 1 for reduced
[UO2(aco)]+. Addition of a single ligand, either H2O or O2, to
[UO2(aco)]+ occurred at a rate efficiency of 30% and 26%
respectively, as kinetically modeled. The subsequent reaction
rates that led to the formation of [UO2(aco)(H2O)(O2)]+ were
much more efficient with [UO2(aco)(H2O)]+ adding O2 and
[UO2(aco)(O2)]+ adding H2O at 69% and 100% of the theoretical
rate constants, respectively. The reaction rate corresponding to
hydration of [UO2(aco)(H2O)(O2)]+ was the least efficient of
all of the reactions, consistent with prior studies that showed
that, as the uranyl coordination sphere filled, condensation
efficiencies decreased.40
Figure 5 is the kinetic plot for the isolation of [UO2(dmso)]+,
and pathways outlined in Scheme 2 were used as a model for
the system. The reactions observed for [UO2(dmso)]+ were less
complex compared to the analogous acetone system due to the
lack of a water adduct to the parent species. This system was
modeled as three subsequent reactions; the first was the addition
of O2 followed by two water addition reactions, and Table 2
shows the experimentally determined forward rate constants.
The reaction rate efficiency for addition of dioxygen to
[UO2(dmso)]+ was 76%, which was significantly higher than
the value modeled for addition of O2 to the aco-containing
complex (26%). The stronger σ-donating dmso ligand increases
the efficiency of O2 addition to the reduced uranyl cation.
Following the initial addition of dioxygen to [UO2(dmso)]+, the
subsequent water additions are not as efficient compared to the
analogous acetone complexes; efficiencies for the addition of
the first and second H2O molecules to [UO2(dmso)(O2)]+ were
61% and 12%, respectively. It is evident that the more basic
ligand promotes the addition of O2 while decreasing the
propensity of H2O addition.
+
Figure 7. Structures of complexes [UO2(DMSO)(H2O)0-2
]
and
[UO2(DMSO)(H2O)0-2(O2)]+ optimized at the M06-L/SSC/6-311+G**
level of theory.
displacement more readily. Multiple water adduct products were
also observed in minor abundance: for acetone [UO2(aco)-
(H2O)2]+ at m/z 364, and for dmso [UO2(dmso)(H2O)2]+ and
[UO2(dmso)(H2O)3]+ at m/z of 388 and 402, respectively. Most
significantly, the displacement of O2 by H2O (reaction 4) was
observed following isolation and storage of [UO2(aco)(O2)]+
in the ion trap (part c of Figure S1 Supporting Information),
whereas the same displacement reaction was not observed when
the analogous [UO2(dmso)(O2)]+ ion was selectively isolated
under the same reaction conditions (part d of Figure S1
Supporting Information). The results from study of displacement
reactions are therefore consistent with the competitive CID
experiments, which suggest that binding and retention of O2 is
enhanced when the more basic σ-donor (dmso) ligand is also
Density Functional Theory Calculations. The binding
energies of the UO2+-O2(η2) complexes calculated with six
density functionals are shown in Table 3. The DFT results are
compared with the benchmark estimates at the CCSD(T)/SSC-
(2g)/6-311++G(3df,3dp) level. PBE41 and PW9142 functionals
yield binding energies that are in good agreement with the
CCSD(T) calculations. In contrast, the M06-L,29 M06,30 and
B3LYP27,28 methods predict binding energies that are, respec-
tively, 52, 73, and 91 kJ/mol lower than the CCSD(T) value,
whereas the LDA method gives a binding energy that is 68 kJ/
mol stronger.
+
bound to the UO2 center.
Experimental Measurements of Gas-Phase Reaction Rates.
In the last series of experiments, the rates for ligand addition
reactions were measured by systematically altering the isolation/
reaction time in the ion trap. Specifically, kinetic plots were
generated by isolating [UO2(lig)]+ in the ion trap from times
ranging from 1 to 2000 ms. The fractional abundances (relative
to total ion abundance) of the precursor and product ions were
measured and plotted as a function of isolation time. Figures 4
and 5 show the kinetic plots generated for the [UO2(aco)]+ and
[UO2(dmso)]+, respectively. For the [UO2(aco)]+ system, the
[UO2(aco)]+ reactant species, along with [UO2(aco)(H2O)]+ and
[UO2(aco)(O2)]+ adducts were all observed at an initial isolation/
reaction time of 1 ms, and the fractional abundance of each
species decreased with increasing reaction time. The relative
abundance of [UO2(aco)(H2O)(O2)]+ increased to a maximum
at approximately 0.15 s, at which point conversion to [UO2-
(aco)(H2O)2(O2)]+ by addition of a second H2O ligand occurred
at longer reaction times. Generation of [UO2(aco)(H2O)2(O2)]+
was essentially complete by 1 s reaction time. On the basis of
the experimental kinetic profiles of the reactant and product ions,
Scheme 1 was generated to function as a kinetic model for the
system. Experimentally determined rate constants, obtained by
The prior UO2 -dioxygen results24 utilized B3LYP and were
+
satisfactory for explaining observed binding preferences in
UO2(aco)n(O2)+ complexes; however, the energetic difference
between aco and O2 elimination was not as great as might have
been expected based on the branching ratio measured in the
collision-induced dissociation spectrum of the n ) 3 complex.
On the other hand, calculations using the PBE and PW91
functionals generated dioxygen-U(V)O2+ binding energies that
were too high. The M06-L functional, which yields binding
energies that are intermediate between values generated using
B3LYP and those generated using PBE, provides results
qualitatively consistent with the relative binding order indicated
by the experiment in the previous24 and present work.
It should be noted that neither DFT nor CCSD(T) calculations
include spin-orbit interactions that are substantial for reaction
energies of actinide compounds in variable oxidation states. The
spin-orbit contributions lower the energies of open-shell U(V)
compounds and thus affect the reaction energies involving

2356 J. Phys. Chem. A, Vol. 113, No. 11, 2009
Leavitt et al.
TABLE 4: Changes in Electronic Energies (with and without Zero-Point Energy (ZPE) Corrections) and Standard Gibbs Free
Energies (kJ/mol) for Binding of Acetone, Dimethylsulfoxide, Water, and Dioxygen to U(V) Compoundsa
reaction
lig ) aco
lig ) dmso
∆Ee + ZPE
∆Ee
∆Ee + ZPE
∆G°
∆Ee
∆G°
+
UO2 + lig
-189.9
-116.5
-105.3
-129.0
-137.2
-137.0
-124.8
-105.1
-185.0
-108.6
-95.4
-122.2
-131.0
-131.7
-117.4
-96.1
-141.8
-81.3
-57.1
-79.0
-85.6
-83.6
-87.9
-55.1
-226.3
-112.4
-101.6
-135.6
-144.4
-144.6
-121.2
-101.8
-222.4
-103.0
-91.8
-129.7
-138.5
-139.7
-111.9
-92.9
-187.9
-65.9
-54.2
-85.9
-96.5
-93.6
-76.4
-51.3
[UO2(lig)]+ + H2O
[UO2(lig)(H2O)]+ + H2O
[UO2(lig)]+ + O2
[UO2(lig)(H2O)]+ + O2
[UO2(lig)(H2O)2]+ + O2
[UO2(lig)(O2)]+ + H2O
[UO2(lig)(H2O)(O2)]+ + H2O
a M06-L/SSC(2g)/6-311++G(3df,3dp) single point energies on M06-L/SSC/6-311++G** optimized geometries. Vibrational frequencies
were computed at the B3LYP/SSC/6-311++G** level.
TABLE 5: M06-L Optimized Distancesa (Angstroms) in
1-12a
c
system
UdO
U· · ·aco/dmso U · · ·H2O
U· · ·O2
O-O
1
2
3
4
5
6
7
8
1.769
1.780
1.782
1.749
1.757
1.765
1.773, 1.780
1.784, 1.793
1.788, 1.800
1.751, 1.757
1.759, 1.765
1.768, 1.768
2.375
2.410
2.420
2.330
2.364
2.410
2.318
2.350
2.370
2.260
2.291
2.331
2.519
2.524
2.342, 2.353 1.285
2.370, 2.363 1.286
2.516
2.536, 2.529b 2.378, 2.381 1.289
2.526
9
2.535, 2.540
2.348, 2.364 1.288
10
11
12
2.522
2.376, 2.378 1.289
2.539, 2.528b 2.387, 2.395 1.292
a Obtained with M06-L/SSC/6-311++G**. b The H2O ligand
trans to the O2 ligand. c The bond adjacent to the aco/dmso ligand is
listed first.
Figure 9. Comparison of asymmetric OdUdO stretching frequencies
for acetone and dimethylsulfoxide ligated UO2+ and UO2+(O2) cations.
binding energies (exclusive and inclusive zero-point energy
(ZPE) corrections) and standard binding free energies of aco,
dmso, H2O, and O2 ligands. Table 5 summarizes the optimized
distances in all studied complexes 1-12.
The results confirm that dmso is a stronger oxygen lone-pair
donor than aco. The binding of the former donor results in a
stronger bond (226.3 kJ/mol for 7 vs 189.9 kJ/mol for 1) and a
shorter intermolecular distance to UO2+ (2.318 Å for 7 vs 2.375
Å for 1). We were also able to find a stable [UO2(dmso)]+
complex coordinated by sulfur but this was bound by only 96.7
kJ/mol. This is consistent with the hard Lewis acid character
of UO2+ favoring complexation with hard bases such as oxygen
donors.47
Figure 8. HOMOs (ꢁ-spin) for 6 and 12, showing the strong overlap
between the UO2+ 5fꢀ and O2 π*xy orbitals. MOs are rendered at 0.027
(e/Å3)1/2 isovalues.
U(VI)/U(V) species. The reported spin-orbit contribution for
several such reactions from the three independent studies43-46
ranges from 28.3 to 29.9 kJ/mol. Thus, our best estimate of the
UO2+-O2(η2) binding energy at the CCSD(T) level after
spin-orbit effects are accounted for is 124.3-125.9 kJ/mol. In
what follows, we use the M06-L functional without spin-orbit
corrections for modeling this type of uranyl coordination
complex, but bear in mind that the better energetic accuracy of
M06-L compared to other DFT methods is fortuitous and the
inclusion of spin-orbit interactions will likely lead to under-
estimation of the binding strength of dioxygen to U(V) systems.
Aside from systematic differences in the energy of the U(V)-O2
bond, binding energies and structural parameters calculated with
several tested functionals follow similar trends with substitution
of aco for dmso and addition of H2O.
Water is a weaker electron donor and forms complexes with
+
UO2 that are significantly less stable than those formed with
aco and dmso ligands. The stability of [UO2(lig)(H2O)n]+
complexes against H2O elimination decreases with increasing
cluster size and is lower for the more strongly coordinated dmso
ligand than for the aco ligand at a fixed cluster size, consistent
with expected metal ion complexation behavior. The UO2+ · · ·
H2O distances closely track the variation in binding energy.
In our earlier communication,24 we suggested that side-on
binding of O2 to form a superoxo complex is due to favorable
overlap between adjacent lobes of the uranyl(V) 5fꢀ and O2
π*xy orbitals. Figure 8 illustrates a two-electron, three-centered
bonding scheme for coordination of O2 in 6 and 12.
The optimized structures of [UO2(aco)]+ and [UO2(dmso)]+
complexes ligated with H2O and O2 are shown in Figures 6
+
and 7. The orientation of the ligands with respect to UO2 is
In agreement with the previous B3LYP calculations,24 the
addition of electron-donating groups appreciably strengthens the
O2 binding energy despite increasing elongation of the U-O2
mainly determined by intermolecular repulsions between the
uranyl oxygen and ligand lone pairs. Table 4 lists the electronic

Uranyl(V) Dioxocations
J. Phys. Chem. A, Vol. 113, No. 11, 2009 2357
bond. Coordination of a single aco and a single dmso enhances
the binding strength by 27.2 and 33.9 kJ/mol, respectively.
Addition of the first water to [UO2(lig)]+ further stabilizes the
interaction with O2 by 8.2-8.8 kJ/mol. Placing the second water
in the remaining position trans to O2 has almost no effect on
binding affinity. This suggests that only ligands adjacent to O2
can directly augment charge transfer in forming the superoxo
bond (Figure 8). Indeed, as shown in Figure 8, there is no
contribution to the HOMO from the water trans to O2.
Interestingly, this water has a shorter U · · ·OH2 distance than
the water adjacent to O2.
were investigated to determine the general effect of ligand
charge donation on the reactivity of UO2+ with respect to water
and dioxygen. The original hypothesis that addition of O2 is
enhanced by strong σ-donor ligands bound to UO2+ is supported
by results from competitive collision-induced dissociation (CID)
experiments, which show near exclusive loss of H2O from
[UO2(dmso)(H2O)(O2)]+, whereas both H2O and O2 are elimi-
nated from the corresponding [UO2(aco)(H2O)(O2)]+ species.
Ligand-addition reaction rates were investigated by monitoring
precursor and product ion intensities as a function of ion storage
time in the ion-trap mass spectrometer: these experiments
+
The UdO bond lengths and stretching frequencies can be
important indicators of the overall basicity of the uranium center.
The calculated UdO bond elongates as the number of equatorial
donor ligands increases. Conversely, this bond length markedly
decreases (by ∼0.02 Å) upon addition of O2, which is compat-
suggest that the association of dioxygen to the UO2 complex
is enhanced when the more basic dmso ligand was coordinated
to the metal complex. Conversely, addition of H2O is favored
for the analogous complex ion that contains an aco ligand.
Experimental rate measurements are supported by density
function theory calculations of relative energies, which show
+
ible with an oxidation of the UO2 metal center occurring via
+
formation of the superoxo complex. Similar variations in UdO
bond distances were found at the B3LYP level but with absolute
values predicted to be ∼0.007 Å longer (Table 3S in the
Supporting Information).
stronger bonds between UO2 and O2 when dmso is the
coordinating ligand, whereas bonds to H2O are stronger for the
aco complex.
Acknowledgment. Work by M. J. Van Stipdonk and C. M.
Leavitt was supported through a grant from the U.S. National
Science Foundation (NSF grant CAREER-0239800). Work by
G. S. Groenewold was supported by the U.S. Department of
Energy, INL Laboratory Directed Research & Development
Program under DOE Idaho Operations Office Contract DE
AC07 05ID14517. Funding for this work was provided by the
National Science Foundation (NIRT CTS Award # 0506951)
and by the US Environmental Protection Agency (STAR Grant
RD-83252501). Work by V. S. Bryantsev, M. S. Diallo, and
W. A. Goddard, III, is performed in part using the MSCF in
EMSL, a national scientific user facility sponsored by the U.S.
DOE, OBER and located at PNNL. W. A. de Jong’s research
was supported by the BES Heavy Element Chemistry program
of the U.S. Department of Energy, Office of Science, and was
performed in part using the Molecular Science Computing
Facility in the William R. Wiley Environmental Molecular
Sciences Laboratory, 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 changes in UdO bond lengths are mirrored by changes
in OdUdO stretching frequencies. Figure 9 illustrates the
expected decrease in the asymmetric uranyl stretching frequen-
cies with increasing the number of donor ligands.48 The red shift
of these vibrational frequencies (∆νa ) 9-17 cm-1) for the
dmso complexes, 7-12, compared to the corresponding aco
complexes, 1-6, is in accordance with the stronger donor
character of dmso.
Calculated changes in standard Gibbs free energies for ligand
addition reactions (∆G°) allow for a direct comparison of the
DFT results with experiment (Table 4). Including zero-point
energy (ZPE) corrections decreases ∆∆Ee values for the
displacement of H2O by O2 in [UO2(lig)(H2O)1-2]+ complexes,
whereas including thermal corrections (T ) 298.15 K) has the
opposite effect on the ∆∆G° of these reactions. It should be
noted that the presence of several low frequency modes (below
100 cm-1) has an adverse effect on the accuracy of the
calculations of entropic contributions within the harmonic
approximation. Thus, the uncertainty in our assessment of any
given ∆G° is not less than 4.2 kJ/mol.
Calculations indicate that [UO2(aco)]+ exhibits a similar
propensity to add H2O (-81.3 kJ/mol) and O2 (-79.0 kJ/mol),
with the former ligand preferred at high temperatures and the
latter ligand preferred at low temperatures. The coordination
of a second ligand to form a binary product [UO2(aco)-
(H2O)(O2)]+ is thermodynamically more favorable for both H2O
(-87.9 kJ/mol) and O2 (-85.6 kcal/mol). The binding free
energy for the subsequent hydration of [UO2(aco)(H2O)-
(O2)]+ is dramatically reduced (-55.1 kJ/mol). In contrast to
[UO2(aco)]+, [UO2(dmso)]+ shows a much stronger ability to
bind O2 (-85.9 kJ/mol) than H2O (-65.9 kJ/mol). Subsequent
water addition reactions to form [UO2(dmso)(O2)(H2O)]+ (-76.4
kJ/mol) and [UO2(dmso)(O2)(H2O)2]+ (-51.3/mol) are thermo-
dynamically less favorable compared to O2 and the analogous
acetone complexes. The predicted binding energy trends are in
excellent agreement with the kinetic data and proposed reaction
pathways for [UO2(aco)]+ and [UO2(dmso)]+ shown in Scheme
1 and Scheme 2.
Supporting Information Available: Tables showing the
effect of the basis set size on binding energies, listing electronic
binding energies, geometric parameters and frequencies for
1-12 calculated at the B3LYP/SSC/6-311++G** level of
theory, and Cartesian coordinates and energies (Hartrees) for
the M06-L/SSC/6-311++G** optimized geometries. This ma-
terial is available free of charge via the Internet at http://
pubs.acs.org.
References and Notes
(1) Greenwood, N. N.; Earshaw, A. Chemistry of the Elements, 2nd
ed.; Butterworth Heinemann: Oxford, Great Britain, 1997; p 1250.
(2) Shulz, W. W.; Navratil, J. D. Science and Technology of Tributyl
Phosphate; CRC Press: Boca Raton, FL, 1984.
(3) Morse, J. W.; Choppin, G. R. ReV. Aquatic Sci. 1991, 4 (1), 1–22.
(4) Silva, R. J.; Nitsche, H. Radiochim. Acta 1995, 70/71, 377–396.
(5) Brookins, D. G. Geochemical Aspects of RadioactiVe Waste
Disposal; Springer-Verlag: New York, 1984.
(6) 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.; North-
Holland: Amsterdam, 1994; Vol. 18, Lanthanides/Actinides: Chemistry, pp
559-590.
Conclusions
Using the combination of ESI and tandem ion-trap mass
spectrometry the gas-phase reactions of [UO2(lig)]+ (lig )
acetone (aco) or dimethylsulfoxide (dmso)) with H2O and O2

2358 J. Phys. Chem. A, Vol. 113, No. 11, 2009
Leavitt et al.
(7) Rizkalla, E. N.; Choppin, G. R., 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.; North-
Holland: Amsterdam, 1994; Vol. 18,Lanthanides/Actinides: Chemistry, pp
529-558.
(31) Purvis, G. D., III.; Bartlett, R. J. J. Chem. Phys. 1982, 76, 1910.
(32) Raghavachari, K.; Trucks, G. W.; Pople, J. A.; Head-Gordon, M.
Chem. Phys. Lett. 1989, 157, 479.
(33) Watts, J. D.; Gauss, J.; Bartlett, R. J. J. Chem. Phys. 1993, 98,
8718–8733.
(34) Ku¨sle, W.; Dolg, M.; Stoll, H.; Preuss, H. J. Chem. Phys. 1994,
100, 7535.
(8) Armentrout, P. B.; Beauchamp, J. L. Chem. Phys. 1980, 50 (1),
27–36.
(9) Cornehl, H. H.; Heinemann, C.; Marcalo, J.; deMatos, A. P.;
Schwarz, H. Angew. Chem., Int. Ed. Engl. 1996, 35, 891–894.
(10) Gibson, J. K.; Haire, R. G.; Santos, M.; Marcu¨alo, J.; Pieres de
Matos, A. J. Phys. Chem. A 2005, 109, 2768–2781.
(11) Kubatko, K.-A. H.; Helean, K. B.; Navrotsky, A.; Burns, P. C.
Science 2003, 302, 1191–1193.
(12) Bhattacharjee, M.; Chaudhuri, M. K.; Purkayastha, R. N. D. Inorg.
Chem. 1986, 25, 2354–2357.
(13) Charpin, P. P.; Folcher, G.; Lance, M.; Vigner, N. E. D. Acta
Crystallogr. 1985, C41, 1302–1305.
(14) Rose, D.; Chang, Y.-D.; Chen, Q.; Zubleta, J. Inorg. Chem. 1994,
33, 5167–5168.
(15) Thuery, P.; Masci, B. Supramol. Chem. 2003, 15, 95–99.
(16) Thuery, P.; Nierlich, M.; Baldwin, B. W.; Komatsuzaki, N.; Hirose,
T. J. Chem. Soc., Dalton Trans. 1999, 1047–1048.
(17) Westland, A. D.; Tarafder, M. T. H. Inorg. Chem. 1981, 20, 3992–
3995.
(18) Van Stipdonk, M.; Anbalagan, V.; Chien, W.; Gresham, G.;
Groenewold, G.; Hanna, D. J. Am. Soc. Mass Spectrom. 2003, 14, 1205.
(19) Chien, W.; Anbalagan, V.; Zandler, M.; Hanna, D.; Van Stipdonk,
M.; Gresham, G.; Groenewold, G. J. Am. Soc. Mass Spectrom. 2004, 15,
777.
(20) Groenewold, G. S.; Van Stipdonk, M. J.; Gresham, G. L.; Chien,
W.; Bulleigh, K.; Howard, A. J. Mass Spectrom. 2004, 39, 752.
(21) Van Stipdonk, M. J.; Chien, W.; Angalaban, V.; Bulleigh, K.;
Hanna, D.; Groenewold, G. S. J. Phys. Chem. A 2004, 108, 10448.
(22) Van Stipdonk, M. J.; Chien, W.; Bulleigh, K.; Wu, Q.; Groenewold,
G. S. J. Phys. Chem. A 2006, 110, 959–970.
(23) Groenewold, G. S.; Cossel, K. C.; Gresham, G. L.; Gianotto, A. K.;
Appelhans, A. D.; Olson, J. E.; Van Stipdonk, M. J.; Chien, W. J. Am.
Chem. Soc. 2006, 128, 3075–3084.
(24) Bryantsev, V. S.; de Jong, W. A.; Cossel, K. C.; Diallo, M. S.;
Goddard III, W. A.; Groenewold, G. S.; Chein, W.; Van Stipdonk, M. J. J.
Phys. Chem. A 2008, 112 (26), 5777–5780.
(25) Zahnley, T.; Macey, R.; Oster, G. Berkeley Madonna, version 8.0.1;
University of California: Berkeley, CA, 2003.
(35) Bylaska, E. J.; de Jong, W. A.; Govind, N.; Kowalski, K.; Straatsma,
T. P.; Valiev, M.; Wang, D.; Apra, E.; Windus, T. L.; Hammond, J.; Nichols,
P.; Hirata, S.; Hackler, M. T.; Zhao, Y.; Fan, P.-D.; Harrison, R. J.; Dupuis,
M.; Smith, D. M. A.; Nieplocha, J.; Tipparaju, V.; Krishnan, M.; Wu, Q.;
Van Voorhis, T.; Auer, A. A.; Nooijen, M.; Brown, E.; Cisneros, G.; Fann,
G. I.; Fruchtl, H.; Garza, J.; Hirao, K.; Kendall, R.; Nichols, J. A.;
Tsemekhman, K.; Wolinski, K.; Anchell, J.; Bernholdt, D.; Borowski, P.;
Clark, T.; Clerc, D.; Dachsel, H.; Deegan, M.; Dyall, K.; Elwood, D.;
Glendening, E.; Gutowski, M.; Hess, A.; Jaffe, J.; Johnson, B.; Ju, J.;
Kobayashi, R.; Kutteh, R.; Lin, Z.; Littlefield, R.; Long, X.; Meng, B.;
Nakajima, T.; Niu, S.; Pollack, L.; Rosing, M.; Sandrone, G.; Stave, M.;
Taylor, H.; Thomas, G.; van Lenthe, J.; Wong A.; Zhang, Z. NWChem, A
Computational Chemistry Package for Parallel Computers, Version 5.1;
Pacific Northwest National Laboratory: Richland, WA, 2007; pp 99352-
0999.
(36) Su, T.; Bowers, M. T. J. Am. Chem. Soc. 1973, 95, 7609–7610.
(37) Su, T.; Bowers, M. T. J. Am. Chem. Soc. 1973, 95, 7611–7613.
(38) Su, T.; Bowers, M. T. Int. J. Mass Spectrom. Ion Phys 1973, 12,
347–356.
(39) Su, T.; Chesnavich, M. J. J. Chem. Phys. 1982, 76, 5183–5185.
(40) Gresham, G. L.; Gianotto, A. K.; Harrington, P. d.B.; Cao, L.; Scott,
J. R.; Olson, J. E.; Appelhans, A. D.; Van Stipdonk, M. J.; Groenewold,
G. S. J. Phys. Chem. A 2003, 107, 8530–8538.
(41) Perdew, J. P.; Burke, K.; Emzerhof, M. Phys. ReV. Lett. 1996, 77,
3865.
(42) Perdew, J. P.; Wang, Y. Phys ReV. B 1992, 46, 12947.
(43) Hay, P. J.; Martin, R. L.; Schreckenbach, G. J. Phys. Chem. A 2000,
104, 6259.
(44) Vallet, V.; Maron, L.; Schimmelpfennig, B.; Leininger, T.; Teichteil,
C.; Gropen, O.; Grenthe, I.; Wahlgren, U. J. Phys. Chem. A 1999, 103,
9285.
(45) Shamov, G. A.; Schreckenbach, G. J. Phys. Chem. A 2005, 109,
10961.
(46) Tsushima, S.; Wahlgren, U.; Grenthe, I. J. Phys. Chem. A 2006,
110, 9175.
(47) Cotton, S. Lanthanide and Actinide Chemistry; John Wiley & Sons:
(26) Flowers, B. H. Numerical Methods in C ++, 2nd ed.; Oxford
University Press: New York, 2000.
Chichester, 2006.
(48) 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, 128, 4802–4813.
(27) Becke, A. D. Phys. ReV A 1988, 38, 3098.
(28) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(29) Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125, 194101.
(30) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215.
JP807651C