Systematic studies of early actinide complexes: Uranium(IV) fluoroketimides

Article abstract of DOI:10.1021/ic700455b

The reaction of (C₅Me₅)₂U(CH ₃)₂ with 2 equiv of N≡C-Ar_F gives the fluorinated uranium(IV) bis(ketimide) complexes (C₅Me ₅)₂U[-N=C(CH₃)(Ar_F)]₂ [where Ar_F = 2-F-C₆H₄ (4), 3-F-C ₆H₄ (5), 4-F-C₆H₄ (6), 2,6-F ₂-C₆H₃ (7), 3,5-F₂-C ₆H₃ (8), 2,4,6-F₃-C₆H₂ (9), 3,4,5-F₃-C₆H₂ (10), and C ₆F₅ (11)]. These have been characterized by single-crystal X-ray diffraction, ¹H and ¹⁹F NMR, cyclic voltammetry, UV-visible-near-IR absorption spectroscopy, and variable-temperature magnetic susceptibility. Density functional theory (DFT) results are reported for complexes 6 and 11 for comparison with experimental data. The most significant structural perturbation imparted by the F substitution in these complexes is a rotation of the fluorinated aryl (Ar_F) group out of the plane defined by the N=C(C_Me)(C_ipso) fragment in complexes 7, 9, and 11 when the Ar_F group possesses two ofluorine atoms. Excellent agreement is obtained between the DFT-calculated and experimental crystal structures for 11, which displays the distortion, as well as for 6, which does not. In 7, 9, and 11, the out-of-plane rotation results in large angles (φ - 53.7-89.4°) between the planes formed by ketimide atoms N=C(C _Me)(C_ipso) and the ketimide aryl groups. Complexes 6 and 10 do not contain o-fluorine atoms and display interplanar angles in the range of φ = 7-26.8°. Complex 4 with a single o-fluorine substituent has intermediate values of φ = 20.4 and 49.5°. The distortions in 7,9, and 11 result from an unfavorable steric interaction between one of the two o-fluorine atoms and the methyl group [-N=C(CH₃)] on the ketimide ligand. All complexes exhibit U^V/U^IV and U ^IV/U^III redox couples, although the distortion in 7, 9, and 11 appears to be a factor in rendering the U^IV/U^III couple irreversible. The potential separation between these couples remains constant at 2.15 ± 0.03 V. The electronic spectra are dominated by unusually intense f-f transitions in the near-IR that retain nearly identical band energies but vary in intensity as a function of the fluorinated ketimide ligand, and visible and near-UV bands assigned to metal (5f)-to-ligand (π*) charge-transfer and interconfiguration (5f² → 5f¹-6d¹) transitions, respectively. Variable-temperature magnetic susceptibility data for these complexes indicate a temperature- independent paramagnetism (TIP) below ～50 K that results from admixing of low-lying crystal-field excited states derived from the symmetry-split ³H₄ 5f² manifold into the ground state. The magnitude of the TIP is smaller for the complexes possessing two o-fluorine atoms (7, 9, and 11), indicating that the energy separation between ground and TIP-admixed excited states is larger as a consequence of the greater basicity of these ligands.

Full text of DOI:10.1021/ic700455b

Inorg. Chem. 2007, 46, 7477−7488
Systematic Studies of Early Actinide Complexes: Uranium(IV)
Fluoroketimides
Eric J. Schelter, Ping Yang, Brian L. Scott, J. D. Thompson, Richard L. Martin, P. Jeffrey Hay,*
David E. Morris,* and Jaqueline L. Kiplinger*
Los Alamos National Laboratory, Los Alamos, New Mexico 87545
Received March 8, 2007
The reaction of (C5Me5)2U(CH3)2 with 2 equiv of N
(C5Me5)2U[ C(CH3)(ArF)]2 [where ArF 2-F-C6H4 (4), 3-F-C6H4 (5), 4-F-C6H4 (6), 2,6-F2-C6H3 (7), 3,5-F2-C6H3
(8), 2,4,6-F3-C6H2 (9), 3,4,5-F3-C6H2 (10), and C6F5 (11)]. These have been characterized by single-crystal X-ray
diffraction, 1H and 19F NMR, cyclic voltammetry, UV
visible near-IR absorption spectroscopy, and variable-temperature
tC−ArF gives the fluorinated uranium(IV) bis(ketimide) complexes
−Nd
)
−
−
magnetic susceptibility. Density functional theory (DFT) results are reported for complexes 6 and 11 for comparison
with experimental data. The most significant structural perturbation imparted by the F substitution in these complexes
is a rotation of the fluorinated aryl (ArF) group out of the plane defined by the NdC(CMe)(Cipso) fragment in complexes
7, 9, and 11 when the ArF group possesses two o-fluorine atoms. Excellent agreement is obtained between the
DFT-calculated and experimental crystal structures for 11, which displays the distortion, as well as for 6, which
does not. In 7, 9, and 11, the out-of-plane rotation results in large angles (φ ) 53.7
formed by ketimide atoms N C(CMe)(Cipso) and the ketimide aryl groups. Complexes 6 and 10 do not contain
o-fluorine atoms and display interplanar angles in the range of φ ) 26.8 . Complex 4 with a single o-fluorine
substituent has intermediate values of φ ) 20.4 and 49.5 . The distortions in 7, 9, and 11 result from an unfavorable
−89.4°) between the planes
d
7−
°
°
steric interaction between one of the two o-fluorine atoms and the methyl group [−NdC(CH3)] on the ketimide
ligand. All complexes exhibit UV/UIV and UIV/UIII redox couples, although the distortion in 7, 9, and 11 appears to
be a factor in rendering the UIV/UIII couple irreversible. The potential separation between these couples remains
constant at 2.15
that retain nearly identical band energies but vary in intensity as a function of the fluorinated ketimide ligand, and
visible and near-UV bands assigned to metal (5f)-to-ligand (
*) charge-transfer and interconfiguration (5f2 5f1-
6d1) transitions, respectively. Variable-temperature magnetic susceptibility data for these complexes indicate a
temperature-independent paramagnetism (TIP) below 50 K that results from admixing of low-lying crystal-field
± 0.03 V. The electronic spectra are dominated by unusually intense f−f transitions in the near-IR
π
f
∼
excited states derived from the symmetry-split 3H4 5f2 manifold into the ground state. The magnitude of the TIP is
smaller for the complexes possessing two o-fluorine atoms (7, 9, and 11), indicating that the energy separation
between ground and TIP-admixed excited states is larger as a consequence of the greater basicity of these ligands.
Introduction
electronic structure of early actinides.1-4 Specifically, studies
on the uranium(IV) bis(ketimide) complexes (C5Me5)2U[-
NdC(Ph)(R)]2 [R ) Ph (1), CH2Ph (2), CH3 (3)] have shown
Complexes of the light actinide elements (Th-Pu) present
an important challenge in understanding electronic structure
and bonding because they frequently defy the conventions
established for both the transition metals and the lanthanides.
Our group has been using organometallic Th and U com-
plexes as tunable platforms for systematically studying the
(1) Morris, D. E.; Da Re, R. E.; Jantunen, K. C.; Castro-Rodriguez, I.;
Kiplinger, J. L. Organometallics 2004, 23, 5142.
(2) Schelter, E. J.; Yang, P.; Hay, P. J.; Martin, R. L.; Scott, B. L.; Da
Re, R. E.; Jantunen, K. C.; Morris, D. E.; Kiplinger, J. L. J. Am. Chem.
Soc. 2007, 129, 5139.
(3) Kiplinger, J. L.; Pool, J. A.; Schelter, E. J.; Thompson, J. D.; Scott,
B. L.; Morris, D. E. Angew. Chem., Int. Ed. 2006, 45, 2036.
(4) Schelter, E. J.; Veauthier, J. M.; Thompson, J. D.; Scott, B. L.; John,
K. D.; Morris, D. E.; Kiplinger, J. L. J. Am. Chem. Soc. 2006, 128,
2198.
* To whom correspondence should be addressed. E-mail:
pjhay@lanl.gov (P.J.H.), demorris@lanl.gov (D.E.M.), kiplinger@lanl.gov
(J.L.K.).
10.1021/ic700455b CCC: $37.00
Published on Web 08/11/2007
© 2007 American Chemical Society
Inorganic Chemistry, Vol. 46, No. 18, 2007 7477

Schelter et al.
spectral resolution was typically 2 nm in the visible region and
4-6 nm in the near-IR region.
numerous spectroscopic signatures typically associated with
transition-metal complexes including charge-transfer (CT)
transitions, enhancements in the intensity of ligand-field
transitions presumably from coupling to these CT states, and
enhancement in the Raman vibrational spectrum when
excited in resonance with the CT and ligand-localized
electronic excited states.1,5 Notably, the metal-based orbitals
Cyclic and square-wave voltammetric data were obtained in the
Vacuum Atmospheres drybox system described above. All data
were collected using a Perkin-Elmer Princeton Applied Research
Corp. (PARC) model 263 potentiostat under computer control with
PARC model 270 software. All sample solutions were ∼5 mM in
complex with a 0.1 M [Bu4N][B(C6F5)4] supporting electrolyte in
THF solvent. The advantageous properties of this quarternary
ammonium salt as an electrolyte for voltammetric studies in low-
dielectric-constant solvents have been noted in several recent
reports,6,7 and the origin of this effect has recently been demon-
strated to be directly related to the greater dissociation of the cation/
anion pair in low-dielectric-constant media such as THF as a result
of the more highly delocalized charge in the [B(C6F5)4]- anion.8
Furthermore, this anion appears to be more inert toward fluoride
abstraction chemistry. All data were collected with the positive-
feedback IR compensation feature of the software/potentiostat
activated to ensure minimal contribution to the voltammetric waves
from uncompensated solution resistance (typically ∼1 kΩ under
the conditions employed). Solutions were contained in PARC model
K0264 microcells consisting of a ∼3 mm diameter Pt disk working
electrode, a Pt wire counter electrode, and a Ag wire quasi-reference
electrode. Scan rates from 20 to 5000 mV/s were employed in the
cyclic voltammetry scans to assess the chemical and electrochemical
reversibility of the observed redox transformations. Half-wave
potentials were determined from the peak values in the square-
wave voltammograms or from the average of the cathodic and
anodic peak potentials in the reversible cyclic voltammograms.
Potential calibrations were performed at the end of each data
collection cycle using the ferrocenium/ferrocene couple as an
internal standard. Electronic absorption and cyclic voltammetric
data were analyzed using Wavemetrics IGOR Pro (version 4.0)
software on a Macintosh platform.
Synthesis. The syntheses of compounds (C5Me5)2U[-NdC(CH3)-
(ArF)]2 [where ArF ) 2-F-C6H4 (4), 3-F-C6H4 (5), 4-F-C6H4 (6),
2,6-F2-C6H3 (7), 3,5-F2-C6H3 (8), 2,4,6-F3-C6H2 (9), 3,4,5-F3-C6H2
(10), and C6F5 (11)] are similar to those previously reported for
their thorium analogues and complexes 2 and 3.2,9 A generic
procedure is reported here for complexes 4-11. In an N2 atmo-
sphere drybox, a 125 mL flask equipped with a magnetic stirbar
was charged with (C5Me5)2U(CH3)2 and 60 mL of toluene. Toluene
solutions of fluorinated nitrile were added to the red-orange (C5-
Me5)2U(CH3)2 solutions, resulting in an immediate color change
to dark red-brown, and the reaction mixture was stirred at room
temperature for 12-15 h. The volatiles were then removed under
reduced pressure to give solids. The solids were dissolved in pentane
(50 mL) and the solutions filtered through a Celite-padded coarse
frit. The filtrate was collected, the solvent removed using dynamic
vacuum, and the resulting solid product collected. The final
purifications varied slightly for each complex. See the Supporting
Information for additional details.
involved in all of these electronic transitions appear to be
principally 5f in character, and the implication is that these
spectral signatures arise because of significant covalent
interactions between the 5f orbitals and those on the ketimide
ligands.
These intriguing spectroscopic results prompted us to
explore in greater detail archetypical transition-metal elec-
tronic structure concepts in these actinide systems through
the preparation and characterization of a more extensive
series of structurally homologous but electronically diverse
uranium(IV) complexes. Recently, we described the synthe-
sis, structures, electrochemistry, and electronic spectroscopy
(including novel luminescent behavior) for a series of
fluorinated 5f06d0 thorium(IV) bis(ketimide) complexes.2
This study was also supported by theoretical calculations of
optimized molecular and electronic structures. In the present
report, we investigate the analogous series of fluorinated
uranium bis(ketimide) complexes (C5Me5)2U[-NdC(CH3)-
(ArF)]2 [where ArF ) 2-F-C6H4 (4), 3-F-C6H4 (5), 4-F-C6H4
(6), 2,6-F2-C6H3 (7), 3,5-F2-C6H3 (8), 2,4,6-F3-C6H2 (9),
3,4,5-F3-C6H2 (10), and C6F5 (11)] with particular attention
to unraveling the added complexities associated with the 5f2
valence electronic configuration in these UIV species. These
new results demonstrate clear ligand-based substituent effects
that have allowed for tuning of the covalent signatures and
the degree of metal-ligand multiple bonding in a series of
uranium(IV) ketimide complexes.
Experimental Section
Instrumentation and Sample Protocols. Electronic absorption
spectral data were obtained for toluene and tetrahydrofuran (THF)
solutions of complexes over the wavelength range 300-1600 nm
on a Perkin-Elmer model Lambda 950 UV-visible-near-IR
spectrophotometer. All data were collected in 1 cm path length
cuvettes loaded in a recirculating Vacuum Atmospheres model HE-
553-2 inert atmosphere (N2) drybox with a MO-40-2 Dri-Train and
run versus a toluene or THF solvent reference. Samples were
typically run at two dilutions, ∼0.1 and ∼0.5 mM, to optimize the
absorbance in the UV-visible and near-IR, respectively. The
Computational Details. The B3LYP hybrid density functional
theory (DFT)10 was employed to optimize the equilibrium molecular
structures of 6, 11, and their anions. The pentamethylcyclopenta-
dienyl (C5Me5) groups in the complexes were used in calculations
(6) Camire, N.; Mueller-Westerhoff, U. T.; Geiger, W. E. J. Organomet.
Chem. 2001, 637, 823.
(7) LeSuer, R. J.; Geiger, W. E. Angew. Chem., Int. Ed. 2000, 39, 248.
(8) LeSuer, R. J.; Buttolph, C.; Geiger, W. E. Anal. Chem. 2004, 76, 6395.
(9) Jantunen, K. C.; Burns, C. J.; Castro-Rodriguez, I.; Da Re, R. E.;
Golden, J. T.; Morris, D. E.; Scott, B. L.; Taw, F. L.; Kiplinger, J. L.
Organometallics 2004, 23, 4682.
(5) Da Re, R. E.; Jantunen, K. C.; Golden, J. T.; Kiplinger, J. L.; Morris,
D. E. J. Am. Chem. Soc. 2005, 127, 682.
(10) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
7478 Inorganic Chemistry, Vol. 46, No. 18, 2007

Uranium(IV) Fluoroketimides
without simplification. The Stuttgart RSC 1997 ECP was employed
for U, which incorporates scalar relativistic effects and replaces
60 core electrons.11 The valence electrons were represented as [8s/
7p/6d/4f], and 6-31G* basis sets were used for C, H, N, and F. All
calculations were carried out with the Gaussian03 suite of codes.12
Natural orbital population analysis was carried out on the total
electron density from the unrestricted DFT calculations.12 The
natural orbitals (NOs) with occupancies of 1.0 give the most
compact representation of the unpaired spin density.
Results and Discussion
Structural Chemistry. Equation 1 presents the synthetic
methods used and the yields obtained in the preparation of
the fluorinated uranium(IV) ketimide complexes. The reac-
Figure 1. Thermal ellipsoid plots of 9 (top) and 10 (bottom) with ellipsoids
projected at the 50% probability level. The U(1) atom resides on a
crystallographically imposed inversion center. Coplanarity is observed in
10 for planes formed by the ketimide aryl groups and atoms N(1)-C(11)-
C(12)-C(13), while in 9, these planes are nearly orthogonal.
∼0.1 Å compared to structurally related uranium(IV) amide
complexes, which possess canonical U-N single bonds
[2.237(3)-2.267(6) Å],14,15 while the structurally relevant
uranium(IV) imido complex (C5Me5)2U(dN-2,4,6-tBu3C6H2)
possesses a UdN bond length of 1.952(12) Å.16 No obvious
trends in the U-N distances for the (C5Me5)2U[-Nd
C(CH3)(ArF)]2 complexes can be deduced on the basis of
simple extent of fluorination or electron-withdrawing ability
of the ketimide ligand substituents.
A primary structural effect was identified for the isos-
tructural (C5Me5)2Th[-NdC(CH3)(ArF)]2 systems, which
apparently masks any anticipated trends among structure,
bonding, and associated spectroscopic and electrochemical
properties. Specifically, F substitution at both ortho positions
of the ketimide aryl group [i.e., (C5Me5)2Th[-NdC(CH3)-
(ArF)]2, where ArF ) 2,6-F2-C6H3, 2,4,6-F3-C6H2, or C6F5]
was found to produce an unfavorable steric interaction that
forces the aryl group to be rotated approximately orthogonal
to the plane formed by the -NdC(CMe)(Cipso) atoms of the
ketimide ligand and results in disruption of conjugation
through the ketimide ligand. As illustrated for complexes 9
and 10 in Figure 1, this distortion is conveniently quantified
by analyzing the torsion angles that describe the smallest
interplanar angle formed by atoms N(1)-C(11)-C(13)-
C(14/18). Looking down the C(11)-C(13) axis, out-of-plane
distortions of the aryl ring result from either clockwise or
counterclockwise rotation of the group along the C(11)-
C(13) axis. The interplanar angles (φ) range from 0° [aryl
groups planar to ketimide atoms NdC(CMe)(Cipso)] to 90°
tion of a red-orange toluene solution of (C5Me5)2U(CH3)2
with 2 equiv of nitrile at room temperature produced the
corresponding uranium(IV) bis(ketimide) complexes 4-11.
Following workup, all complexes were reproducibly isolated
in pure form as brown solids despite the modest yields for
some of them, which reflect the high solubility imparted by
the fluorinated ligands. Single-crystal X-ray crystallography
of compounds 4, 6, 7, and 9-11 confirms the expected
1
structures for these complexes, consistent with the H and
19F NMR data in all cases (Figure 1 and Supporting
Information). Isolation of these species further demonstrates
the compatibility of U and Th ions with fluorinated organic
ligands.2 Representative structures of 9 and 10 are shown in
Figure 1, and the crystallographic parameters for 4, 6, 7,
and 9-11 are listed in Table 1. Table 2 presents salient
geometric parameters for complexes 1, 2, 4, 6, 7, and 9-11.
Single crystals were not obtained for 5 or 8, but their
1
formulations are consistent with their H and 19F NMR
spectroscopic data (see the Supporting Information).
The U-N distances observed for 4, 6, 7, and 9-11 span
from 2.169(9) to 2.220(3) Å and fall within the range
observed for previously reported uranium(IV) ketimide
complexes, 2.04(2)-2.225(5) Å.9,13 Evidence for U-N bond
orders slightly greater than 1 in complexes 4, 6, 7, and 9-11
is deduced from a general shortening of these distances by
(11) Kuchle, W.; Dolg, M.; Stoll, H.; Preuss, H. J. Chem. Phys. 1994, 100,
7535.
(12) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B. Gaussian03, revision
C.02; Gaussian, Inc.: Wallingford, CT, 2004.
(13) Kiplinger, J. L.; Morris, D. E.; Scott, B. L.; Burns, C. J. Organome-
tallics 2002, 21, 3073.
(14) Straub, T.; Frank, W.; Reiss, G. J.; Eisen, M. S. J. Chem. Soc., Dalton
Trans. 1996, 2541.
(15) Peters, R. G.; Scott, B. L.; Burns, C. J. Acta Crystallogr., Sect. C:
Cryst. Struct. Commun. 1999, 55, 1482.
(16) Arney, D. S. J.; Burns, C. J. J. Am. Chem. Soc. 1995, 117, 9448.
Inorganic Chemistry, Vol. 46, No. 18, 2007 7479

Schelter et al.
Table 1. Crystallographic Experimental Parameters for 4, 6, 7, and 9-11
4
6
7
9
10
11
formula
a (Å)
b (Å)
C36H44F2N2U
10.0173(5)
17.6957(9)
18.322(1)
90
90
90
3247.8(3)
4
C36H44F2N2U
17.9790(19)
18.0889(19)
9.9044(11)
90
90
90
3221.1(6)
4
C36H42F4N2U
10.6877(6)
17.5713(9)
19.3511(10)
108.793(1)
90.476(1)
102.700(1)
3343.8(3)
4
C36H40F6N2U
13.5252(5)
14.5744(5)
17.1466(6)
90
96.434(1)
90
3358.7(2)
4
C36H40F6N2U
12.9529(6)
13.8105(6)
18.5471(8)
90
99.7810(10)
90
3269.6(3)
4
C36H36F10N2U
15.9821(4)
16.3139(4)
26.9285(6)
90
90
90
7021.1(3)
8
c (Å)
R (deg)
â (deg)
γ (deg)
V (Å3)
Z
fw
780.76
P212121
141(1)
0.710 73
1.597
5.035
0.0316
0.0694
780.76
Pnma
141(1)
0.710 73
1.610
5.077
0.0735
0.1960
816.75
P1h
141(1)
0.710 73
1.622
4.903
0.0286
0.0689
852.73
C2/c
141(1)
0.710 73
1.686
4.893
0.0197
0.0501
852.73
C2/c
141(1)
0.710 73
1.732
5.027
924.70
Pccn
141(1)
0.710 73
1.750
4.705
0.0318
0.1002
space group
T (K)
λ (Å)
D
calcd (g cm-3
)
µ (mm-1
)
R1 [I > 2σ(I)]a
0.0182
0.0498
wR2 (all data)b
Table 2. Geometric Parameters for 1, 2, 4, 6, 7, and 9-11
1
2
4
6
7
9
10
11
U-N (Å)
2.184(5)
2.203(6)
2.185(5)
2.179(6)
1.280(8)
1.272(9)
1.281(8)
1.293(8)
178.9(6)
176.5(5)
2.184(3)
2.196(4)
2.200(4)
2.169(9)
2.185(3)
2.220(3)
2.190(3)
2.185(3)
1.271(5)
1.251(5)
1.258(5)
1.265(5)
172.8(3)
176.8(3)
169.5(3)
171.7(3)
107.51(12)
104.02(12)
55.3
2.193(2)
1.257(3)
176.78(19)
2.1974(18)
2.204(5)
2.199(5)
NdC (Å)
1.274(4)
162.4(3)
102.40(15)
1.265(6)
1.245(6)
1.289(13)
179.0(8)
1.264(3)
1.243(7)
1.248(7)
U-NdC (deg)
174.1(4)
176.6(4)
172.48(17)
178.4(4)
176.3(4)
N-U-N (deg)
φ (deg)
109.4(2)
107.2(2)
109.92(15)
112.0(5)
26.8a
105.49(11)
60.1
111.83(10)
7.0
104.8(2)
106.1(2)
68.4
49.5
20.4
89.4
73.0
58.7
53.7
a The [-NdC(CH3)(4-F-C6H4)]- ligands of this complex are disordered, resulting in only an estimated twist angle for this complex. The value was
obtained by averaging the values of the various disordered components. The twist angle for the analogous Th complex (C5Me5)2Th[-NdC(CH3)(4-F-
C6H4)]2 (12) is 8.4°.2
(orthogonal aryl groups). Table 2 lists the interplanar angles
(φ) for complexes 4, 6, 7, and 9-11 and reveals two distinct
structure types: complexes 6 and 10 do not contain o-fluorine
atoms and display interplanar angles in the range of φ )
7-26.8° (small aryl rotation), while 7, 9, and 11 each possess
two o-fluorine atoms and exhibit significantly larger values
of φ ) 53.7-89.4° (large aryl rotation). Complex 4 with a
single o-fluorine substituent has intermediate values of φ )
20.4 and 49.5°.
the Supporting Information). For 4-6, 8, and 10, the ketimide
aryl proton resonances were significantly broadened at room
temperature (298 K) and definitive assignments of these
chemical shifts could only be made using variable-temper-
ature (363-388 K) NMR experiments. In contrast, these
ketimide aryl protons were easily observed at room temper-
ature for complexes 7 and 9, which both possess large aryl
rotations and interplanar angles (φ ) 53.7-89.4°) at room
temperature.
In a comparison of the 1H NMR aryl resonances between
the sets of complexes 4-6, 8, and 10 versus 7 and 9, there
should be no significant change in the paramagnetic shift
and relaxation by the dipolar interaction, based on the solid-
state structural data. It can be surmised that the same primary
steric interaction that disrupts the conjugation of the ketimide
ligand also causes disruption of the orbital pathway for Fermi
contact between the UIV ion and aryl protons in 7 and 9.
This is noteworthy because the NMR data demonstrate that
the electronic structure effects associated with the rotation
of the ketimide aryl groups are operative in solution.
DFT Calculations. The structures of the ground states of
6 and 11 were calculated using DFT approaches. Overall,
there is good agreement between the calculated structures
For the 6d05f0 (C5Me5)2Th[-NdC(CH3)(ArF)]2 complexes
possessing two o-fluorine atoms and consequently large
interplanar (φ) angles (ArF ) 2,6-F2-C6H3, 2,4,6-F3-C6H2,
or C6F5), significant perturbations to electronic structure were
observed. Similar electronic structure perturbations might be
expected to occur in the isostructural 6d05f2 (C5Me5)2U[-
NdC(CH3)(ArF)]2 complexes 7, 9, and 11, which also
possess large φ angles. The ground 6d05f2 configuration of
the uranium bis(ketimide) complexes provides several more
spectroscopic handles to survey the impact of this ketimide
ligand rotation on UIV electronic structure. Interestingly, the
1
effect can be qualitatively observed by H NMR spectros-
1
copy. All of the complexes 1-11 exhibit shifted H NMR
resonances due to the paramagnetism of the UIV ion (see
7480 Inorganic Chemistry, Vol. 46, No. 18, 2007

Uranium(IV) Fluoroketimides
Table 3. Comparison of Calculated and Experimental Geometric Parameters for Complexes 6 and 11
(C5Me5)2U[-NdC(CH3)(4-F-C6H4)]2 (6)
(C5Me5)2U[-NdC(CH3)(C6F5)]2 (11)
exptl
calcd
exptl
calcd
R(U-N) (Å)
2.169(9)
1.289(13)
2.446, 2.512
112.0(5)
26.8a
2.185
1.275
2.544
107.1
7.5
2.199(5), 2.204(5)
1.243(7), 1.248(7)
2.499, 2.513
104.8(2)
2.199
1.263
2.518
109.5
63.7
R(N-C) (Å)
R(U-C5Me5(cent)) (Å)
(N-U-N) (deg)
φ (interplanar angle, deg)
65.8
a The [-NdC(CH3)(4-F-C6H4)]- ligands of this complex are disordered, resulting in only an estimated twist angle for this complex. The value was
obtained by averaging the values of the disordered components. The twist angle for the analogous Th complex (C5Me5)2Th[-NdC(CH3)(4-F-C6H4)]2 (12)
is 8.4°.2
Table 4. Calculated Electronic Properties for Complexes 6 and 11 and
the Corresponding Anions [6]- and [11]- a
relative energy (eV)
net spin
N
vertical
adiabatic
U
C
6
0.0
-0.87
0.0
0.0
-0.95
0.0
2.11
2.71
2.12
2.75
-0.11
-0.09
-0.09
-0.08
+0.09
+0.19
+0.09
+0.19
[6]-
11
[11]-
-1.10
-1.24
a Vertical energies (eV) and spin properties of the neutral and negative
ion species are given. Unpaired spin densities are shown for the U metal
center and the ketimide CdN atoms calculated at the equilibrium geometries
of the negative ions.
and the X-ray diffraction results (Table 3). The calculated
U-N bond lengths for 6 and 11 are 2.185 Å [expt ) 2.169-
(9) Å] and 2.199 Å [expt ) 2.199(5) and 2.204(5) Å], and
the calculated U-C5Me5(cent) bond lengths are 2.544 Å (expt
) 2.446 and 2.512 Å) and 2.499 Å (expt ) 2.499 and 2.513
Å), respectively. The major structural difference between the
complexes involving the rotation of the fluorinated aryl ring
out of the NdC(CMe)(Cipso) ketimide plane is well reproduced
by the DFT results. For complex 6, the aryl ring is nearly
coplanar (7.5° calcd; ∼28.6° expt), while in complex 11,
the unfavorable steric interactions between the two o-fluorine
atoms and the methyl group [-NdC(CH3)] on the ketimide
ligand force the aryl ring to be significantly rotated (65.8°
calcd; 63.7° expt).
The electronic properties of the neutral complexes 6 and
11 and their corresponding negative ions are summarized in
Table 4. Anion calculations were performed to assess the
relative agreement between theory and experiment on the
UIV/UIII redox energies. The ground state of each molecule
corresponds to a high-spin 5f2 configuration on the UIV
center. The DFT results indicate 2.1 unpaired electrons on
each metal with minor spin density residing on the CdN
atoms of the ketimide ligands. This can be seen clearly by
examining the ground-state NOs that correspond to the singly
occupied orbitals in Table 5. These orbitals are localized on
the U metal center and comprised entirely of 5f character.
By contrast, a more complicated interpretation arises from
examining the conventional R-spin molecular orbitals (MOs)
of the DFT calculations. In this case, the 5f character, also
totaling two electrons, was spread out over several of the
highest occupied MOs. See the Supporting Information for
Mulliken analysis tables for 6, 11, and their anions.
Figure 2. Cyclic voltammograms for ∼5 mM uranium(IV) bis(ketimide)
complexes in ∼0.1 M [Bu4N][B(C6F5)4]/THF at room temperature. All scans
were collected at 200 mV/s.
withdrawing properties of the pentafluorophenyl group in
11 relative to the 4-fluorophenyl group in 6 correlates with
the fact that complex 11 is more easily reduced to UIII in
the electrochemical studies at -2.34 V (11) vs -2.64 V (6).
While the negative ions formally have a 5f3 ground-state
configuration, some of the unpaired electron density is
delocalized onto the organic ligands (Table 4). Roughly 2.7
unpaired electrons reside on the U metal center according
to the Mulliken analysis. This can also be seen in the three
NOs corresponding to the unpaired electrons for the negative
ions, where significant ligand character is evident (Table 5).
Electrochemistry. Room-temperature voltammetric data
(cyclic and square wave) were obtained for complexes 1-11
in ∼0.1 M [Bu4N][B(C6F5)4]/THF. Typical cyclic voltam-
metric data are illustrated in Figure 2, and the metrical data
As shown in Table 4, the negative ions [6]- and [11]-
are calculated to be stable relative to the neutral species by
0.95 and 1.24 eV, respectively, as gas-phase species without
solvent corrections. The greater electron affinity and electron-
Inorganic Chemistry, Vol. 46, No. 18, 2007 7481

Schelter et al.
Table 6. Summary of Redox Data for (C5Me5)2U[-NdC(R1)(R2)]2 Complexes in ∼0.1 M [Bu4N][B(C6F5)4]/THFa
R1
R2
E1/2(U5+/U4+
)
E1/2(U4+/U3+
)
E1/2(L)
∆E1/2(1)b
∆E1/2(2)c
1
2
3
4
5
6
7
8
C6H5
CH2Ph
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
C6H5
C6H5
C6H5
2-F-C6H4
3-F-C6H4
-0.48 (R)d
-0.46 (R)d
-0.54 (R)d
-0.46 (R)d
-0.45 (R)d
-0.48 (R)d
-0.49 (R)d
-0.42 (R)d
-0.43 (R)d
-0.34 (R)d
-0.26 (R)d
-2.50(R)d
-2.57 (R)d
-2.68 (R)d
-2.62 (R)d
-2.54 (R)d
-2.64 (R)d
-2.82 (C)f
-2.46 (R)d
-2.80 (I)f
-2.48 (R)d
-2.34 (C)f
-2.79 (R)d
2.02
2.11
2.14
2.16
2.09
2.16
2.33
2.04
2.37
2.14
2.08
2.31
2.4
2.42
2.56
2.48
2.60
∼-2.9 (R)d
-2.96 (R)d
-3.02 (I)f
-2.93 (I)f
-3.08 (I)f
g
4-F-C6H4
2,6-F2-C6H3
3,5-F2-C6H3
2,4,6-F3-C6H2
3,4,5-F3-C6H2
C6F5
-2.88 (C), f -3.12 (I)f
-3.18 (I)f
g
2.46
2.75
9
10
11
-2.66 (I)f
2.40
a Legend: (R) ) chemically reversible; (I) ) chemically irreversible; (C) ) wave becomes reversible at fast scan rates. All values are in volts. E1/2 values
d
are versus [(C5H5)2Fe]+/0
.
b ∆E1/2(1) ) E1/2(U5+/U4+) - E1/2(U4+/U3+). c ∆E1/2(2) ) E1/2(U5+/U4+) - E1/2(L1). E1/2 was determined from (Epc + Epa)/2 at
200 mV/s. E1/2 was estimated from (Epc + Epc/2)/2 at 200 mV/s. E1/2 was determined from the peak potential in a square-wave voltammogram. g Distinct
e
f
wave not observed within the available potential window.
from the electrochemical experiments are summarized in
Table 6. These voltammetric data are generally consistent
with those reported previously for uranium(IV) bis(ketimide)
complexes.1,9,13 In particular, the stabilizing influence of the
σ- and π-donating ketimide ligands leads to a ready acces-
sibility of the UV/UIV redox couple in addition to the more
common UIV/UIII couple. The ketimide ligands are also
sources of reductive redox processes (although some of the
fluorinated ketimide ligands have somewhat ambiguous
reduction chemistry as described below), as verified by the
appearance of essentially identical ligand-based voltammetric
behavior in the analogous thorium(IV) metallocene com-
plexes.1,2,9 These reduction steps are invariably coupled to
chemical transformations that render the electron-transfer
step(s) irreversible and lead to additional new oxidation
features in the voltammograms (dashed-line traces in Figure
2). There is no apparent ketimide ligand-based oxidative
electrochemistry. The only other redox processes that can
be observed for these complexes in this solvent/electrolyte
system are attributed to irreversible pentamethylcyclopen-
tadienyl (C5Me5 ) ligand-based oxidation common to all of
-
the tetravalent actinide metallocene complexes at approxi-
mately +1 V versus [(C5H5)2Fe]+/0 1
.
For the majority of the uranium(IV) bis(ketimide) com-
plexes, both of the metal-centered redox transformations (UV/
UIV and UIV/UIII) are chemically reversible, as evidenced by
the appearance of cathodic and anodic waves for both couples
in the cyclic voltammograms (Figure 2). Thus, it is note-
worthy that the putative UIV/UIII couple for the fluorinated
ketimide complexes 7, 9, and 11 shows signs of irreversible
behavior (Figure 2). Note that these are the three complexes
that possess two o-fluorine atoms on the ketimide aryl group
and exhibit the structural distortion discussed above, which
results in disruption of conjugation through the ketimide
ligand. The ultimate origin of this seemingly irreversible
behavior is at present unknown, but preliminary DFT
calculations do indicate a potential structural change in the
ketimide ligand upon reduction for complex 11 that is not
Table 5. NOs Corresponding to Singly Occupied Orbitals in Complexes 6, 11, and Their Corresponding Anions
7482 Inorganic Chemistry, Vol. 46, No. 18, 2007

Uranium(IV) Fluoroketimides
Figure 4. Examples of spectral deconvolutions of the UV-visible (top
panel) and near-IR (bottom panel) regions of the electronic spectra for 6 in
a toluene solution at room temperature. Labels on component fit bands
correspond to those listed in Table 7.
the UV-visible region (top panel, Figure 3) and the near-
IR region (bottom panel, Figure 3) to better compare the
spectral properties of these complexes and facilitate further
discussion. Note in particular that there is good qualitative
correlation among the near-IR spectral data for all 10
complexes but some apparent greater variability in behavior
in the UV-visible spectral region. All spectral data were
successfully fit in both the UV-visible and near-IR regions
using a sum of Gaussian bands, as illustrated for 6 in
Figure 4. A Gaussian band was also required to account for
the low-energy tail of the unresolved high-energy feature in
the UV-visible region, as indicated by the dashed line
(Figure 4). The metrical parameters derived from these fits
for the principal bands in each region are compiled in Table
7. Some summary statistical data have been included for
these metrical parameters. These statistics have no intrinsic
meaning but are simply provided to illustrate the strong
correlation in the electronic transitions among these 10
complexes.
Figure 3. UV-visible (top panel) and near-IR (bottom panel) spectra for
uranium(IV) bis(ketimide) complexes in a toluene solution at room
temperature. Spectra are offset for clarity; ꢀ = 0 at 11 000 cm-1 for all
spectra. The asterisk denotes a small contribution from a solvent vibrational
band that was incompletely subtracted.
seen for complex 6. Additional theoretical and experimental
studies of this redox behavior are underway and will be
reported elsewhere.
A final interesting observation from these electrochemical
data is the consistency in the separation between the two
metal-based redox processes across this range of ligands
[∆E1/2(1) in Table 6]. This same observation was made
previously for a range of (C5Me5)2UIVLL′ complexes, where
L and L′ are ligands capable of stabilizing the UV oxidation
state through both σ- and π-bonding interactions (e.g.,
ketimides, hydrazonates, and imides).1 In this prior study,
the average ∆E1/2(1) value was 2.1 V. Here, the average value
for all 11 bis(ketimide) complexes is 2.15 ( 0.03 V. Thus,
these new fluorinated ketimide complexes preserve this
interesting trend in the redox energetics of these principally
metal-based redox transformations.
The bands observed in the near-IR region can be assigned
unambiguously to intraconfiguration f-f ligand-field transi-
tions on the basis of their relatiVely low oscillator strengths
and their energy range, which are consistent with a 5f2 UIV
ion. The rather broader, less distinct features in the UV-
visible region provide a greater challenge in assignment and
Electronic Absorption Spectroscopy. The electronic
absorption spectral data for complexes 1, 2, and 4-11 are
illustrated in Figure 3. These data have been broken out into
Inorganic Chemistry, Vol. 46, No. 18, 2007 7483

Schelter et al.
Table 7. Summary of Electronic Spectral Data for (C5Me5)2U[-NdC(R1)(R2)]2 Complexes in a Toluene Solution at Room Temperature
electronic transition energy (E) and oscillator strength (f)
MLCT2 f-fA f-fB
complex
nπ*
MLCT1
f-fC
f-fD
E
f
E
f
E
f
E
f
E
f
E
f
E
f
R1
R2
(cm-1
)
(×103) (cm-1
)
(×103) (cm-1
)
(×103) (cm-1
)
(×106) (cm-1
)
(×106) (cm-1
)
(×106) (cm-1
)
(×106)
1
2
4
5
6
7
8
9
C6H5
CH2Ph C6H5
C6H5
20658
21318
21739
20955
21470
20822
20899
13.7
10.9
14.1
8.5
7.5
2.4
7.9
1.9
5.7
6.8
7.9
1.3
17862
2.8
1.5
1.5
2.7
2.0
3.0
1.9
3.5
1.6
2.7
2.3
0.2
16249
2.4
1.7
0.98
1.7
2.1
0.15
2.4
0.14
2.6
0.15
1.4
0.3
9044
9050
9077
9053
9069
9056
9028
9056
9017
9044
9049
6
190
228
410
229
342
206
263
166
292
203
253
24
8278
8326
8560
8492
8519
8445
8479
8427
8486
8463
8448
27
1040
932
505
505
556
832
690
672
795
945
748
61
7980
7999
8101
8012
8069
8110
7962
8120
7964
8082
8040
20
283
307
671
539
731
202
649
152
729
199
446
76
7283
7331
7399
7351
7386
7344
7314
7343
7309
7301
7336
12
71
137
191
180
217
100
233
79
216
79
150
20
18122
18138
18046
18482
18723
17859
19056
18037
18374
18270
123
16708
16834
16605
17015
16900
16438
16923
16604
16847
16712
76
CH3
CH3
CH3
CH3
CH3
CH3
2-F-C6H4
3-F-C6H4
4-F-C6H4
2,6-F2-C6H3
3,5-F2-C6H3
2,4,6-F3-C6H2 20897
3,4,5-F3-C6H2 20846
10 CH3
11 CH3
av
C6F5
20562
21017
118
std dev
require consideration of multiple lines of evidence as
discussed below.
well-defined shoulders in this same energy range and with
approximately the same intensity (Figures 3 and 5) that are
assigned to these same electronic transitions. (Data for only
the lowest-energy nπ* band are provided in Table 7.)
As illustrated in Figure 5, the two lower-energy bands
identified in the uranium bis(ketimide) spectrum have no
analogues in the spectra of either the thorium bis(ketimide)
complex 12 or the uranium bis(alkyl) complex (C5Me5)2U-
(CH3)2. This is the spectral region that engendered the
greatest resonance enhancement in the Raman vibrational
bands localized on the ketimide ligand of 1.5 On the basis
of this observation, these electronic bands were assigned to
metal (U) f ligand (ketimide π*) CT (MLCT) transitions
because such a transition should have a large Franck-
Condon distortion in the ketimide modes between the ground
and MLCT excited states.5 This same set of electronic bands
is observed in all of the uranium bis(ketimide) complexes
reported here (Figure 3), and the MLCT assignment is still
preferred based on the Raman result and the following
additional considerations. First, the electrochemical data
clearly show that the ketimide ligands possess accessible
acceptor orbitals, while the metal center can be readily
oxidized. These results establish the thermodynamic feasibil-
ity for an MLCT optical process.18 Second, the energy of
these optical bands is intermediate between those of the
ligand-field bands (bottom panel, Figure 3) and the ketimide-
localized nπ* bands. DFT calculations indicate that the
energy of the 5f orbital manifold in the uranium bis(ketimide)
complexes is intermediate between those of the ketimide
nonbonding N p orbitals and the ketimide CdN π* orbitals
(see, for example, Figure 2 in ref 19). Although time-
dependent DFT calculations are unable to predict the energies
of transitions associated with the open-shell 5f2 orbital
manifold, the orbital energies serve as a rough basis for
predicting the relative transition energies. In particular, the
energy of the 5fU f ketimide π*CdN MLCT transition should
correspond approximately to the energy separation between
the occupied 5f orbital and the unoccupied ketimide π*CdN
A useful starting point for the assignment of the UV-
visible electronic transitions in these uranium bis(ketimide)
complexes is a comparison of this spectral region for the
analogous thorium bis(ketimide) complex (C5Me5)2Th[-Nd
C(CH3)(4-F-C6H4)]2 (12) and the uranium bis(alkyl) starting
material (C5Me5)2U(CH3)2. Such a comparison is provided
in Figure 5 for 6. The features of particular interest in these
spectral data are indicated by arrows in Figure 5, and these
are among the ones that distinguish complex 6 from its
thorium analogue (12) and (C5Me5)2U(CH3)2.
Inasmuch as there are no readily accessible valence orbitals
on the metal in the thorium bis(ketimide) complex, the
observed UV-visible transitions in complex 12 must be
ligand-based (ketimide and/or C5Me5).17 Furthermore, the
isostructural uranium bis(ketimide) complexes should possess
the same transitions at similar energies and with comparable
intensities because the ligand sets are identical. The distinct
band in the Th complex 12 at ∼22 000 cm-1 (Figure 5) has
been assigned to a set of two nearly degenerate ketimide-
based pN f π*CdN transitions (nπ*) that are relatively weak
because although spin-allowed they are orbitally forbidden.2
The uranium bis(ketimide) complexes all exhibit one or two
(17) Manuta, D. M.; Lees, A. J. Inorg. Chem. 1986, 25, 3212.
(18) Lever, A. B. P., Dodsworth, E. S., Solomon, E. I., Lever, A. B. P.,
Eds.; John Wiley & Sons, Inc.: New York, 1999; Vol. 2.
(19) Clark, A. E.; Martin, R. L.; Hay, P. J.; Green, J. C.; Jantunen, K. C.;
Kiplinger, J. L. J. Phys. Chem. A 2005, 109, 5481.
Figure 5. Comparison of the UV-visible spectra of 6, 12, and (C5Me5)2U-
(CH3)2 in a toluene solution at room temperature.
7484 Inorganic Chemistry, Vol. 46, No. 18, 2007

Uranium(IV) Fluoroketimides
orbital so long as the coulomb (J) and exchange (K) integral
energies do not make too large a contribution to the transition
energy. The magnitudes of J and K scale with the extent of
orbital overlap and should therefore be relatively small for
overlap of 5f and ligand π* orbitals.18 Third, while the
intensity and line widths of these bands may seem anoma-
lously small for fully dipole-allowed MLCT bands (compared
to d-π* transitions in transition-metal complexes), it should
be noted that CT intensities also scale with the extent of
orbital overlap between participating metal and ligand
orbitals. Here, too, the overlap between the 5f and ligand
π*CdN orbitals, while believed to be unusually large for
actinide systems, is small in comparison to d orbital overlap
with ligand orbitals in transition-metal complexes. In fact,
the observed extinction coefficients in these bands for the
uranium bis(ketimide) complexes (ꢀ ) ∼250-300 M-1 cm-1
for 6; Figures 3 and 5) are significantly larger than those
found for the paradigmatic poxygen f 5fU ligand-to-metal CT
energy;21-25 however, the stronger σ- and π-interactions of
the ketimide ligands with the metal center may stabilize some
subset of metal 6d orbitals as well, thereby lowering the
energy of the interconfiguration transition relative to that in,
for example, the (C5Me5)2U(CH3)2 complex. The ultimate
uncertainty in assigning this transition is exacerbated by the
much greater potential for mixing of excited states in these
ketimide complexes as a result of more favorable metal-
ligand orbital energy matching and greater metal-ligand
bonding interactions.
As noted above, there is a remarkable degree of consis-
tency in the f-f spectral region for all of the uranium bis-
(ketimide) complexes considered in this study (bottom panel,
Figure 3). As seen in the summary statistical data in Table
7, the standard deviations in the energies of the four principal
f-f bands among this sample population is only ∼5-25
cm-1 (i.e., comparable to the spectral resolution in this
wavelength range). However, there is some significant
variability in the oscillator strengths of these bands across
this series of complexes suggestive of an origin in some
component of the electronic structure of these complexes.
Variations in both intensities and energies resulting from
changes in ligand- or crystal-field perturbations are typically
correlated according to the nephelauxetic effect.26 Changes
in the magnitude of the ligand field across the series of
ketimide ligands studied here appear to be manifest in the
magnetic susceptibility data, but the observed intensity
changes in the optical spectral data for these f-f bands are
believed to have a different origin. In our previous study of
uranium(IV) metallocene complexes containing a variety of
ligand types, we suggested that the f-f intensities in the
ketimide complexes in particular exhibited substantial en-
hancement relative to the values found in complexes of
simple σ-donor ligands such as halides and alkyls.1 To
provide some quantitative basis for this intensity enhance-
ment, note that the oscillator strengths in the more intense
f-f bands of (C5Me5)2U(CH3)2 and the bis(hydrazonato)
complex (C5Me5)2U[η2-(N,N)-CH2Ph-N-NdCPh2]2 (13)
are 33 × 10-6 (band at 10 930 cm-1) and 47 × 10-6 (band
at 10 100 cm-1), respectively.1 In contrast, the oscillator
strengths in the more intense f-f bands of the uranium bis-
(ketimide) complexes average ∼400 × 10-6 for the four
bands identified in Table 7 over the entire series of
complexes.
2+
transition in typical UO2 complexes possessing strong
metal-ligand covalent bonding (ꢀ ) ∼50-100 M-1 cm-1;
see, for example, Figure 4 in ref 20). Finally, it is possible
that the two bands (MLCT1 and MLCT2) are simply resolved
vibronic components of a transition to a single MLCT excited
state. The spacing between these two bands (Table 7) is in
the range of ∼1300-1700 cm-1 for the entire series (i.e.,
corresponds to a typical mid-IR vibrational mode), which
would be consistent with a ketimide ligand-localized vibra-
tional mode that undergoes a Franck-Condon distortion
between the ground and MLCT excited states, as was already
demonstrated in the resonance Raman data.
A final consideration in the UV-visible spectral region
for the uranium bis(ketimide) complexes is the substantial
intensity in the unresolved high-energy band relative to that
of (C5Me5)2U(CH3)2 (Figure 5). As has been noted previ-
ously, the methyl ligands in (C5Me5)2U(CH3)2 only interact
with the metal center in a σ-donor fashion.1 Because there
are no low-lying methyl ligand-based orbitals available to
participate in ligand-based and/or CT transitions, the UV-
visible spectral region of this complex should be dominated
by metal-based interconfiguration and perhaps C5Me5-metal-
derived electronic transitions. To a first approximation, these
same metal-based (5f2 f 5f16d1) and/or C5Me5 T metal CT
transitions should be present in the uranium bis(ketimide)
complexes with approximately the same energy and intensity
as in the bis(alkyl) species (C5Me5)2U(CH3)2. Pentamethyl-
cyclopentadienyl ligand-localized π-π* transitions are also
possible, but these should be roughly the same in the Th
and U complexes and, based on the data for the Th complex,
occur at higher energy than is shown in Figure 5. A 5f2 f
5f16d1 assignment for this unresolved band in the uranium
bis(ketimide) complexes seems potentially attractive because
such transitions should have large oscillator strengths and
fairly broad bandwidths. Previous studies of classical UIV
coordination complexes place this transition at higher
Studies of the intensity enhancement for parity-forbidden
ligand-field (d-d) optical transitions in transition-metal
complexes were expected to provide a framework for
discerning the mechanism for the observed intensity en-
hancement in the f-f ligand-field bands for the uranium bis-
(ketimide) complexes. The most common enhancement
mechanism for d-d transitions is by a Herzberg-Teller
(21) Krupa, J. C. J. Alloys Compd. 1995, 225, 1.
(22) Krupa, J. C. J. Solid State Chem. 2005, 178, 483.
(23) Godbole, S. V.; Page, A. G.; Sangeeta Sabharwal, S. C.; Gesland, J.
Y.; Sastry, M. D. J. Lumin. 2001, 93, 213.
(24) Poturaj-Gutniak, S.; Taube, M. J. Inorg. Nucl. Chem. 1968, 30, 1005.
(25) Barandiaran, Z.; Seijo, L. J. Chem. Phys. 2003, 118, 7439.
(26) Lever, A. B. P. Inorganic Electronic Spectroscopy, 2nd ed.; Elsevier:
Amsterdam, The Netherlands, 1984.
(20) 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,
38, 1456.
Inorganic Chemistry, Vol. 46, No. 18, 2007 7485

Schelter et al.
process that mixes energetically close CT states into the
ligand-field states.26 Given the energetic proximity of the
MLCT states in the uranium bis(ketimide) complexes, it was
expected that these states could provide the agency for the
observed ∼10-fold intensity enhancements relative to those
seen in the bis(alkyl) complexes noted above by mixing some
CT character into the f-f state. However, if this were the
operative mechanism, there should be an approximately
linear correlation between the oscillator strength of the f-f
band and that of the CT band involved in the mixing because
the matrix element for the transition moment between the
perturbed f orbital states contains linear terms in the CT
transition moment matrix element. No such correlation has
been observed for any of the f-f bands in relation to either
of the MLCT bands or the ligand-based pN f π*CdN bands
in the data in Table 7. Here, however, any trends in intensity
stealing from the higher lying MLCT and ligand-based states
could be masked by competing influences such as the
unfavorable steric factors noted for 7, 9, and 11 that
might artificially suppress (or enhance) the intensities of the
CT and/or ligand-based transitions. Thus, the apparent
large f-f intensity enhancement remains a challenge to
explain.
Figure 6. Magnetic susceptibility of 11 from 2 to 350 K.
be more sensitive to ligand variation than their lanthanide
counterparts. With this electronic structure framework in
mind, the following expectations for magnetic susceptibility
measurements of 4-11 can be applied: (1) J is an effective
3
quantum number and the H4 ground-state term for UIV 5f2
3
will be used; (2) the H4 term in a low-symmetry crystal
Magnetic Susceptibility. The magnetic response of com-
pounds 4-11 was measured between 350 and 2 K to evaluate
changes in the electronic structure based on different ketimide
groups. Before comparison of the series of compounds, it is
useful to discuss the general magnetic behavior of the UIV
ion to establish expectations for comparison between com-
plexes. Actinide ions resist simple, precise interpretation of
their magnetic behaviors for the usual reasons that have been
discussed extensively elsewhere.27,28 The primary consider-
ation is that the magnitude of the spin-orbit coupling is equal
to or larger than the interelectron repulsion resulting in terms
from excited-state configurations that contribute to the
ground-state term. This results in an admixture, making the
quantum number J an ineffective reckoning for the account-
ing of states. The reality of the situation for many UIV
complexes is that the admixture turns out to be small enough
(<10% 6d15f1 contribution to the ground-state 6d05f2 con-
figuration) that Russell-Saunders coupling can be used
successfully,29 especially in the context of a series of
complexes such as those under investigation here.
field (C2V) completely lifts its degeneracy, yielding 2J + 1
) 9 singlet states; (3) appreciable magnetic moments are
only obtained with a population of more than one crystal-
field state due to an orbital (L) component; (4) perturbations
on the electronic structure observed in the magnetic response
will be manifest in the ligand-field manifold.
The response of a representative uranium(IV) bis(ketimide)
complex, 11, is shown in Figure 6. This compound shows
typical magnetic behavior for a UIV 5f2 ion with a nominal
3H4 ground-state term and attains a øT product of 0.97 emu
K mol-1 (2.78 µB) at 350 K, consistent with reported values
for UIV complexes27,29-33 and only coincidentally close to
the expected value of 1.0 emu K mol-1 for a simple S ) 1,
g ) 2 Curie paramagnet. Compared to a J ) 4 Curie
paramagnet, complex 11 exhibits a small but measurable
temperature dependence. As seen in Figure 6, ø versus T
increases slightly with decreasing temperature until it reaches
∼30 K, where it becomes temperature-independent. This
temperature dependence is reflected in øT versus T by the
monotonic decrease with temperature and approach to zero
at 2 K. The broad trend of ø versus T indicates that the first
crystal-field excited state is being thermally depopulated over
the temperature range examined. Previous reports have
shown similar UIV magnetic behavior that was assumed to
result from an excited-state triplet contribution;34 however,
in the low symmetry of 11, the first excited state is also a
singlet. Therefore, the temperature-dependent paramagnetism
must arise from an orbital contribution that is only manifest
with the thermal population of multiple states. Furthermore,
Using the Russell-Saunders coupling scheme, treatment
3
of the UIV ion results in a ground-state H4 term analogous
to the PrIII ion. However, because the ground-state 3H4 term
is split by crystal-field energies larger than the thermal
energies kT achieved in these experiments, the susceptibility
data will never reach the high-temperature Curie value (or
even (10%) for a J ) 4 term of C ) 1.60 emu K mol-1
(3.58 µB). Also, because of larger 5f and ligand orbital
overlap, these crystal-field splittings in U complexes should
(27) Boudreaux, E. A.; Mulay, L. N. Theory and Applications of Molecular
Paramagnetism; Wiley: New York, 1976.
(28) Marks, T. J., Fragala, I. L., Eds. NATO AdVanced Science Institutes
Series C: Mathematical and Physical Sciences; NATO: Dordrecht,
The Netherlands, 1985; Vol. 155.
(30) Gans, P.; Marriage, J. J. Chem. Soc., Dalton Trans. 1972, 46.
(31) Karraker, D. G.; Stone, J. A. Inorg. Chem. 1972, 11, 1742.
(32) Salmon, L.; Thuery, P.; Riviere, E.; Miyamoto, S.; Yamato, T.;
Ephritikhine, M. New J. Chem. 2006, 30, 1220.
(33) Castro-Rodriguez, I.; Meyer, K. Chem. Commun. 2006, 1353.
(34) Hirose, M.; Miyake, C.; Du Preez, J. G. H.; Zeelie, B. Inorg. Chim.
Acta 1988, 150, 293.
(29) Grunzweig-Genossar, J.; Kuznietz, M.; Friedman, F. Phys. ReV. 1968,
173, 562.
7486 Inorganic Chemistry, Vol. 46, No. 18, 2007

Uranium(IV) Fluoroketimides
jugation at the NdC unit, the segregation of these groups in
the magnetic behavior is a manifestation of the weaker
ketimide ligand basicity of group II versus group I. In concert
with the suite of electrochemical and spectroscopy data
collected on these systems as well as the theoretical treat-
ment, we attribute the differences in low-temperature mag-
netic response to ligand-field perturbation of the UIV
paramagnetic states. Importantly, these studies provide
complementary data because the energetic differences be-
tween complexes manifest in the magnetic data are on the
order of 101-102 cm-1, readily discernible in their magnetic
susceptibilities but too small to be manifest in the ambient
temperature optical data discussed above. The differences
between complexes arise from the combination of the
primary effect of fluorination at the ortho positions and the
secondary effect of the extent and position of the fluorination.
The ultimate observation that can be made from the sum
of the magnetic data is that complexes with rotated aryl
groups result in a larger separation between the ground and
excited states responsible for the TIP contribution in their
low-temperature magnetic response. This larger state separa-
tion results in a smaller low-temperature ø versus T response
(and magnetic moment) for complexes with rotated aryl
groups.
Figure 7. Temperature-dependent magnetic susceptibility data for 4, 6,
7, and 9-11 from 2 to 350 K.
the temperature-independent value attained at low temper-
ature is a result of temperature-independent paramagnetism
(TIP) derived from admixing of low-lying crystal-field
excited states into the ground state.
In the collection of the magnetic data for 4-11, we
encountered difficulties in obtaining reproducible data for ø
versus T, especially in the low-temperature regime <30 K.
We attribute this to paramagnetic impurities that contribute
a Curie tail to the data, especially in the context of the small
moments that are being measured in this temperature regime.
The result of the Curie tail is a masking of the low-
temperature TIP behavior. We were able to overcome this
issue by using pulverized, dry samples of single crystals for
measurements. Such samples could not be obtained for
compounds 5 and 8, which are included as Supporting
Information with their acknowledged Curie tails. The ø
versus T data for compounds 4, 6, 7, and 9-11 are shown
in Figure 7. Several of these data sets, especially those for
6, exhibit small Curie tails, which do not affect the outcome
of the comparison within this data. Interestingly, a compari-
son of these data sets reveals marked differences between
their low-temperature responses.
Figure 7 shows the low-temperature magnetic responses
for compounds 4, 6, 7, and 9-11, which are different
depending on the identity of the fluorinated aryl ketimide
substituent. This indicates that, by changing the ligands, the
electronic fine structure of the crystal-field manifold of these
complexes is perturbed. Furthermore, the complexes can be
roughly segregated into two groups based on the magnitude
of their low-temperature magnetic moments. Group I consists
of compounds 7, 9, and 11, while group II consists of 4, 6,
and 10. The distinction between the groups is consistent with
the structural change noted previously upon ortho fluorination
of the aryl rings on the ketimide ligands. Group I, possessing
two o-fluorine atoms on the ketimide aryl group, exhibits
smaller TIP values at low temperature, indicating a larger
energy separation between the ground and TIP-admixed
excited states, while group II has these excited states lower
in energy. On the basis of the theoretical treatment that shows
that ortho fluorination disrupts the -NdC(CH3)(ArF) con-
Conclusions
The series of fluorinated uranium ketimide complexes on
which we report here have provided valuable additional
insight into the hallmarks of covalent bonding interactions
in early actinide ketimide chemistry. Structural chemistry,
experimental data, and DFT have come together to provide
a uniform description of these systems that is highlighted
by (1) stabilization of multiple metal oxidation states (UIII,
UIV, and UV), (2) the appearance of new CT and intercon-
figuration electronic transitions that are classically the domain
of transition-metal chemistry, (3) enhanced intensities in the
intraconfiguration (f-f) transitions, and (4) variable degrees
of TIP at low temperature. All of these elements are a
reflection of the strong metal-ligand bonding interactions
sustained by the ketimide ligands and varied in response to
the different degrees of fluorination of the ketimide aryl
group. Furthermore, the impact of the unfavorable steric
interactions due to the o-fluorine atoms in 7, 9, and 11 that
lead to a structural distortion and disruption of the orbital
delocalization in the ketimide ligand is evidenced in nearly
every line of experimental data. The effects are most
pronounced in the loss of reversibility of the UIV/UIII redox
transformation and the extent of excited-state admixture
3
within the symmetry-split H4 manifold that leads to the
observed TIP at low temperature. Coupled with our recent
studies of the analogous thorium fluoroketimide complexes,
we have established an exciting new link between the
molecular electronic structural properties of the early ac-
tinides and those of the d-block metals and identified some
as-yet unresolved issues. These studies will certainly
provide the motivation for future exploration of covalency
and metal-ligand bonding in this fascinating class of
complexes.
Inorganic Chemistry, Vol. 46, No. 18, 2007 7487

Schelter et al.
Acknowledgment. For financial support of this work, we
acknowledge the LANL Glenn T. Seaborg Institute for
Transactinium Science (Postdoctoral fellowships to E.J.S. and
P.Y.), LANL (Director’s and Frederick Reines Postdoctoral
Fellowships to E.J.S.), the LANL Laboratory Directed
Research & Development Program, and the Division of
Chemical Sciences, Office of Basic Energy Sciences, Heavy
Element Chemistry program. We thank Kimberly C. Jantunen
for executing preliminary reactions. This work was carried
out under the auspices of the National Nuclear Security
Administration of the U.S. Department of Energy at Los
Alamos National Laboratory under Contract DE-AC52-
06NA25396.
Supporting Information Available: Details of materials,
synthetic and crystallographic experimental methods, tables of
Mulliken analysis results for 6, 11, and their anions, magnetic
susceptibility data for 5 and 8, tables with bond lengths and angles
for the structures of 4, 6, 7, 9, 10, and 11, and thermal ellipsoid
plots of 4, 6, 7, and 11. This material is available free of charge
via the Internet at http://pubs.acs.org.
IC700455B
7488 Inorganic Chemistry, Vol. 46, No. 18, 2007