Structural and spectroscopic characterization of a charge-separated uranium benzophenone ketyl radical complex

Article abstract of DOI:10.1021/ja801007q

The reaction of [((^t-BuArO)₃tacn)U^III] (1) with 4,4′-di-tert-butylbenzophenone affords a unique isolable U(IV) ketyl radical species [((^t-BuArO)₃tacn)U^IV (OC?^t-BuPh₂)] (2) supported by XRD data, magnetization measurements, and DFT calculations. Isolation and full characterization of the corresponding diphenyl methoxide complex [(( ^t-BuArO)₃tacn)U^IV(OCT^t-BuPh ₂)] (3) is also presented. The one-electron reduction of benzophenone by [((^AdArO)₃tacn)U^III] (4) leads to a purple U(IV) ketyl radical intermediate [((^AdArO)₃tacn)U ^IV(OC?Ph₂)] (5). This species is highly reactive, and attempts at isolation were unsuccessful and resulted in methoxide complex [((^AdArO)₃tacn)U^IV(OCHPh₂)] (6) from H abstraction and dinuclear paracoupled complex [((^AdArO) ₃tacn)U^IV(OCPhPh-CPh₂O)U^IV(( ^AdArO)₃tacn)] (7).

Full text of DOI:10.1021/ja801007q

Published on Web 04/26/2008
Structural and Spectroscopic Characterization of a
Charge-Separated Uranium Benzophenone Ketyl Radical
Complex
Oanh P. Lam,†,‡ Christian Anthon,‡ Frank W. Heinemann,‡ Joseph M. O’Connor,†
and Karsten Meyer*,‡
Department of Chemistry, UniVersity of California, San Diego, 9500 Gilman DriVe, La Jolla,
California 92093, and Department of Chemistry and Pharmacy, Inorganic Chemistry, UniVersity
of Erlangen-Nu¨rnberg, Egerlandstr. 1, D-91058 Erlangen, Germany
Received February 8, 2008; E-mail: kmeyer@chemie.unierlangen.de
Abstract: The reaction of [((t-BuArO)3tacn)UIII] (1) with 4,4′-di-tert-butylbenzophenone affords a unique
isolable U(IV) ketyl radical species [((t-BuArO)3tacn)UIV(OC•t-BuPh2)] (2) supported by XRD data, magnetiza-
tion measurements, and DFT calculations. Isolation and full characterization of the corresponding diphenyl
methoxide complex [((t-BuArO)3tacn)UIV(OCHt-BuPh2)] (3) is also presented. The one-electron reduction of
benzophenone by [((AdArO)3tacn)UIII] (4) leads to a purple U(IV) ketyl radical intermediate [((AdArO)3-
tacn)UIV(OC•Ph2)] (5). This species is highly reactive, and attempts at isolation were unsuccessful and
resulted in methoxide complex [((AdArO)3tacn)UIV(OCHPh2)] (6) from H abstraction and dinuclear para-
coupled complex [((AdArO)3tacn)UIV(OCPhPhsCPh2O)UIV((AdArO)3tacn)] (7).
complex.14 Subsequent studies demonstrated that stable metal
fluorenone ketyl complexes could be isolated in systems where
Introduction
The syntheses of ketyl radicals through single-electron
reduction of ketones with alkali metals were first discovered
over 100 years ago.1–3 Ever since, this class of compounds has
attracted continuous attention in many areas, such as aiding in
drying and deoxygenating solvents, serving as important
intermediates in organic reactions,4–6 and providing interesting
sources for spectroscopic studies.7–13 However, due to the highly
reactive nature of ketyl radical species, isolation and structural
characterization of benzophenone ketyl complexes have eluded
scientists for many decades. The planar, fused-ring system in
fluorenone ketyls permits more effective electron delocalization
relative to that in benzophenone ketyls, and in 1995, Hou
cleverly exploited this stabilization affect in the successful solid-
state structural characterization of a samarium fluorenone ketyl
the benzophenone analogues proved too unstable for solid-state
characterization.15–18
Despite successes with the structural isolation of fluorenone
ketyls, those of benzophenone ketyls are still sought after
because they have shown to be much more reactive.7,9,12,13
Attempts to isolate metal benzophenone ketyl radicals have often
resulted in hydrogen abstraction to form metal alkoxides,17,19
insertions into M-L bonds,20–28 or reductive coupling to form
(15) Hou, Z.; Jia, X.; Fujita, A.; Tezuka, H.; Yamazaki, H.; Wakatsuki, Y.
Chem.—Eur. J. 2000, 6, 2994–3005.
(16) Hou, Z.; Fujita, A.; Koizumi, T.; Yamazaki, H.; Wakatsuki, Y.
Organometallics 1999, 18, 1979–1985.
(17) Hou, Z.; Fujita, A.; Zhang, Y.; Miyano, T.; Yamazaki, H.; Wakatsuki,
Y. J. Am. Chem. Soc. 1998, 120, 754–766.
(18) Hou, Z.; Fujita, A.; Yamazaki, H.; Wakatsuki, Y. J. Am. Chem. Soc.
1996, 118, 2503–2504.
† University of California, San Diego.
‡ University of Erlangen-Nu¨rnberg.
(19) Adam, R.; Villiers, C.; Ephritikhine, M.; Lance, M.; Nierlich, M.;
Vigner, J. New J. Chem. 1993, 17, 455.
(1) Bechman, F.; Paul, T. Justus Liebigs Ann. Chem. 1891, 266, 1.
(2) Schlenk, W.; Weickel, T. Ber. Dtsch. Chem. Ges. 1911, 44, 1182–
1189.
(20) Ogoshi, S.; Tonomori, K.; Oka, M.; Kurosawa, H. J. Am. Chem. Soc.
2006, 128, 7077–7086.
(3) Schlenk, W.; Thal, A. Ber. Dtsch. Chem. Ges. 1913, 46, 2840–2854.
(4) Skrydstrup, T. Angew. Chem., Int. Ed. Engl. 1997, 36, 345–347.
(5) Molander, G. A. Chem. ReV. 1992, 92, 29–68.
(6) Molander, G. A.; Harris, C. R. Chem. ReV. 1996, 96, 307–338.
(7) Sakamoto, M.; Cai, X.; Fujitsuka, M.; Majima, T. Chem.—Eur. J. 2006,
12, 1610–1617.
(21) Vela, J.; Vaddadi, S.; Cundari, T. R.; Smith, J. M.; Gregory, E. A.;
Lachicotte, R. J.; Flaschenriem, C. J.; Holland, P. L. Organometallics
2004, 23, 5226–5239.
(22) Basuli, F.; Tomaszewski, J.; Huffman, J. C.; Mindiola, D. J. Orga-
nometallics 2003, 22, 4705–4714.
(23) Emslie, D. J. H.; Piers, W. E.; Parvez, M.; McDonald, R. Organo-
metallics 2002, 21, 4226–4240.
(8) Anandhi, R.; Umapathy, S. J. Raman Spectrosc. 2000, 31, 331–338.
(9) Covert, K. J.; Wolczanski, P. T.; Hill, S. A.; Krusic, P. J. Inorg. Chem.
1992, 31, 66–78.
(24) Mindiola, D. J.; Cummins, C. C. Organometallics 2001, 20, 3626–
3628.
(10) Dams, R.; Malinowski, M.; Westdorp, I.; Geise, H. Y. J. Org. Chem.
1982, 47, 248–259.
(25) Castillo, I.; Tilley, T. D. Organometallics 2000, 19, 4733–4739.
(26) Liang, F.; Jacobsen, H.; Schmalle, H. W.; Fox, T.; Berke, H.
Organometallics 2000, 19, 1950–1962.
(11) Mao, S. W.; Nakamura, K.; Hirota, N. J. Am. Chem. Soc. 1974, 96,
5341–5349.
(27) Breen, T. L.; Stephan, D. W. Organometallics 1996, 15, 4509–4514.
(28) Scholz, J.; Go¨rls, H. Inorg. Chem. 1996, 35, 4378–4382.
(29) Fedushkin, I. L.; Skatova, A. A.; Cherkasov, V. K.; Chudakova, V. A.;
Dechert, S.; Hummert, M.; Schumann, H. Chem.—Eur. J. 2003, 9,
5778–5783.
(12) Hirota, N. J. Am. Chem. Soc. 1967, 89, 32–41.
(13) Hirota, N.; Weissman, S. I. J. Am. Chem. Soc. 1964, 86, 2538–2545.
(14) Hou, Z.; Miyano, T.; Yamazaki, H.; Wakatsuki, Y. J. Am. Chem. Soc.
1995, 117, 4421–4422.
9
10.1021/ja801007q CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 6567–6576 6567

A R T I C L E S
Lam et al.
Scheme 1. Formation of U(IV) Benzophenone Ketyl Radical Complex 2 through a One-Electron Reduction of 4,4′-Di-tert-butylbenzophenone
by U(III) Precursor Complex [((t-BuArO)3tacn)UIII] (1) Followed by H Abstraction To Form a U(IV) Diphenyl Methoxide Complex 3
pinacolate complexes.17,29–33 To date, the isolation and structural
characterization of benzophenone ketyl radical complexes are
limited to complexes of alkali/alkaline earth and samarium
metals.15,34–36 Here we demonstrate the use of a triazacy-
clononane-anchored trisaryloxide ligand chelated to uranium to
stabilize a charge-separated di-tert-butyl benzophenone ketyl
radical species. In addition to X-ray diffraction, this complex
has been characterized by electronic absorption spectroscopy,
combustion analysis, magnetism, and DFT studies. Preliminary
reactivity studies on the uranium ketyl complexes have led to
the formation of a U(IV) diphenyl methoxide complex via
H-atom abstraction chemistry and the first crystallographic
evidence of dimer formation resulting from a head-to-tail
reductive para coupling of two benzophenone ketyl ligands.
(2). Purple XRD-quality crystals can be obtained from diffusion
of acetonitrile into a tetrahydrofuran solution of 2.
In the presence of neat 1,4-cyclohexadiene, complex 2
undergoes a color change from purple to green within 12 h,
indicative of H atom abstraction to form the corresponding
U(IV) diphenyl methoxide complex [((t-BuArO)3tacn)UIV(OCHt-
BuPh2)] (3). Recrystallization of 3 from pentane results in XRD-
quality green single crystals.
The molecular structure of 2 reveals the 4,4′-di-tert-butylben-
zophenone ketyl fragment nestled inside the reactive pocket at the
seventh axial position of the trisaryloxide uranium compound
(Figure 1, left). The average U-O(ArO) and U-N(tacn) ligand
distances of 2.207 and 2.683 Å, respectively, are comparable
to those of precursor complex 1 (Table 1). Note that four
resonance structures constitute the overall solid-state molecular
structure of complex 2, with three structures possessing U(IV)
oxidation states and one with U(III) oxidation state (Figure 2,
2a-2d). Interestingly, this one U(III) resonance structure deter-
mines, to a significant extent, how complex 2 structurally and
spectroscopically deviates from that of typical U(IV) 5f2
complexes. For instance, the U1-O4 bond length of 2.178(4)
Å is appreciably longer than that of previously reported U(IV)
5f2 complexes of the (RArO)3tacn3- ligand system37 and of the
Results and Discussion
The reaction of the electron-rich trisaryloxide complex [((t-BuA-
rO)3tacn)UIII] (1) ((t-BuArO)3tacn3- ) trianion of 1,4,7-tris(3,5-
di-tert-butyl-2-hydroxybenzyl)-1,4,7-triazacyclononane) with
benzophenone yielded intractable products that were not isolable.
However, on the basis of literature precedence, we surmised
that the reaction may have formed diphenyl methoxide and
reductive coupling products. Accordingly, the disubstituted 4,4′-
di-tert-butylbenzophenone was chosen as an alternative that
would provide electronic stabilization of the radical and deter
dimerization due to tert-butyl groups blocking access to the para
and pinacol coupling pathways.
Synthesis and Molecular Structure of 2 and 3. Treating a red-
brown hexane solution of [((t-BuArO)3tacn)UIII] (1) with 1 equiv
of disubstituted 4,4′-di-tert-butylbenzophenone affords a purple
precipitate (Scheme 1). The deep purple solids are isolated in
good yield (73%) by vacuum filtration and characterized as a
U(IV) ketyl radical complex [((t-BuArO)3tacn)UIV(OC•t-BuPh2)]
reported U(IV) aryloxide complex [UIV(OAr)4] (d(U-O)av
)
2.136 Å).38 The bond distance of 1.334(6) Å for the C70-O4
bond is shorter than that of a C-O single bond, indicating that
complex 2 may possess an oxidation that is intermediate between
U(III) and U(IV). The U1-O4 bond length of 2.178(4) Å and
the C70-O4 bond length of 1.334(6) Å is comparable to the
bond distances reported for the Sm(III) benzophenone ketyl
complex by Domingos et al. where the Sm-O bond is 2.199 Å
and the C-O bond is 1.322 Å.35 In addition, no significant
disruptions in aromaticity of the phenyl groups are observed in
the molecular structure, most likely due to complex 2 having
two resonance structures where the aromaticity remains intact
(2a and 2d). A notable feature of the structure of 2 is the sum
of the angles around the ketyl carbon, which total up to 359.96°,
confirming that ketyl carbon C70 is indeed sp2-hybridized and
that H abstraction has not occurred.
(30) Cui, C.; Ko¨pke, S.; Herbst-Irmer, R.; Roesky, H. W.; Noltemeyer,
M.; Schmidt, H.-G.; Wrackmeyer, B. J. Am. Chem. Soc. 2001, 123,
9091–9098.
(31) Deacon, G. B.; Forsyth, C. M.; Wilkinson, D. L. Chem.—Eur. J. 2001,
7, 1784–1795.
(32) Hou, Z.; Koizumi, T.; Nishiura, M.; Wakatsuki, Y. Organometallics
2001, 20, 3323–3328.
The molecular structure of complex 3 reveals coordination
of the diphenyl methoxide ligand at the axial position perpen-
dicular to the trisaryloxide plane (Figure 1, right). The average
(33) Villiers, C.; Adam, R.; Lance, M.; Nierlich, M.; Vigner, J.; Ephri-
tikhine, M. J. Chem. Soc., Chem. Commun. 1991, 1144–1145.
(34) Hou, Z.; Jia, X.; Hoshino, M.; Wakatsuki, Y. Angew. Chem., Int. Ed.
Engl. 1997, 36, 1292–1294.
ˆ
(35) Domingos, A.; Lopes, I.; Waerenborgh, J. C.; Marques, N.; Lin, G. Y.;
Zhang, X. W.; Takats, J.; McDonald, R.; Hillier, A. C.; Sella, A.;
Elsegood, M. R. J.; Day, V. W. Inorg. Chem. 2007, 46, 9415–9424.
(36) Clegg, W.; Eaborn, C.; Izod, K.; O’Shaughnessy, P.; Smith, J. D.
Angew. Chem., Int. Ed. Engl. 1997, 36, 2815–2817.
(37) Castro-Rodriguez, I.; Meyer, K. Chem. Commun. 2006, 1353–1368.
(38) Berg, J. M.; Clark, D. L.; Huffman, J. C.; Morris, D. E.; Sattelberger,
A. P.; Streib, W. E.; Sluys, W. G. V. D.; Watkin, J. G. J. Am. Chem.
Soc. 1992, 114, 10811–10821.
9
6568 J. AM. CHEM. SOC. VOL. 130, NO. 20, 2008

XRD Studies on a Benzophenone Ketyl Complex of Uranium
A R T I C L E S
Figure 1. Molecular structures of [((t-BuArO)3tacn)UIV(OC•t-BuPh2)] (2) (left) in crystals of 2·3C4H8O and of [((t-BuArO)3tacn)UIV(OCHt-BuPh2)] (3) (right)
in crystals of 3·C6H6/CH3CN. Thermal ellipsoids are at the 50% probability level. Hydrogens (except H1) and co-crystallized solvent molecules are omitted
for clarity.
Table 1. Selected Structural Parameters for Complexes 2, 3, 6,
and 7 (Bond Distances in Angstroms, Angles in Degrees)
toward the plane than in 2, indicating that 3 has better U-L
orbital involvement and is consistent with 2 having the U(III)
structural parameters
2
3
6
7
resonance structure (2d). Also, compound 2 has a much longer
U1-O4 bond and shorter C70-O4 bond (Figure 3) than 3, again
in compliance with 2 having the U(III) resonance structure (2d).
It should be noted that such a feature can also be attributed to
the delocalization of the negative charge of the ketyl radical
species. In view of the two resonance structures, where the single
electron resides on the ortho- and para-carbons (2b and 2c),
the C70-CAr bonds are expected to be shorter in 2 than in 3,
and indeed that is observed in the structural parameters of 2.
The structural differences between 2 and 3 clearly indicate that
2 deviates from typical U(IV) 5f2 complexes.
Electronic Structure of 2 and 3. In order to elucidate the
electronic structure of complexes 2 and 3, Kohn-Sham DFT
calculations were performed on complexes using the ADF
package (2007.01).39,40 Key geometry parameters are sum-
marized in Table 2, and DFT isosurface plots are shown (Figures
4 and 6).41,42 The experimental and computed distances agree
remarkably well and show only very small deviations of about
0.02 Å.
U1-N1tacn
U1-N2tacn
U1-N3tacn
U1-Nav
2.698(4)
2.700(4)
2.652(4)
2.683
2.208(3)
2.217(3)
2.197(3)
2.207
2.673(4)
2.706(4)
2.678(4)
2.686
2.218(3)
2.228(3)
2.192(3)
2.213
2.646(3)
2.704(3)
2.688(3)
2.678
2.216(3)
2.233(2)
2.200(3)
2.215
2.717(6)
2.643(6)
2.706(5)
2.689
2.196(5)
2.221(4)
2.213(4)
2.210
2.168(4)
2.109(4)
1.366(7)
1.423(7)
U1-O1ArO
U1-O2ArO
U1-O3ArO
U1-Oav
U1-O4
2.178(4)
2.077(3)
2.092(3)
U2-O8
C70-O4
1.334(6)
1.406(5)
1.386(5)
C71-O8
U1-O4-C70
U2-O8-C71
159.6(3)
178.4(3)
171.9(3)
160.2(4)
167.6(4)
Σ
Cketyl
359.96
-0.232
Uout-of-plane shift
-0.216
-0.163
-0.163/4
ligand U-O(ArO) and U-N(tacn) distances in the molecular
structure of 3 remain unperturbed at 2.213 and 2.686 Å,
respectively. The U1-O4 bond distance of 2.077(3) Å and the
C70-O4 distance of 1.406(5) Å in complex 3 are appreciably
different than those of 2. The shorter U1-O4 bond of 2.077(3)
Å is indicative of stronger U-L electrostatic interaction in
complex 3 than in 2 and is in agreement with the U-O distance
reported for the only other known U(IV) diphenylmethoxide
complex (2.071 Å).19 In addition, the U1-O4-C70 angle of
178.4(3)° in 3 is close to linear, whereas the U1-O4-C70 angle
of 2 is bent at an angle of 159.6(3)°.
Defined as the displacement of the uranium atom below the
triangular plane of the three phenolic oxygens, the uranium out-
of-plane shift (Uoop) is an important structural parameter in
complexes of the [((RArO)3tacn)U] system because it provides
an indication of metal-ligand interaction strength.37 This
parameter, which is a function of U-L orbital involvement and
electrostatic interaction, decreases as the formal oxidation state
of the uranium ion increases. For instance, the uranium out-of-
plane shift (Uoop) for the six-coordinate precursor complex 1 is
-0.75 Å. With strongly bound axial ligands, such as in
complexes 2 and 3, Uoop values shift to -0.232 and -0.216 Å,
respectively. In comparison, the uranium ion in 3 is pulled more
The core of complex 3 has an idealized C3 symmetry with a
strong methoxy oxygen ligator coordinating along the z-axis.
From a ligand field point of view, it is expected that methoxide
complex 3 behaves like an f2 system, in which the set of seven
f orbitals are antibonding with respect to the filled orbitals of
the ligands. A scaled ZORA computation on 3 revealed that
SOMO and SOMO′ are mainly of f character and can be
2
2
assigned to orbitals fxyz and fz(x -y ) (Figure 4). In accordance
2
2
with ligand field theory, the fxyz and fz(x -y ) orbitals have δ
symmetry with respect to the U-O axis and contain the two f
electrons since they have only weak π interactions with the
aryloxide oxygens and no interaction with the methoxide group
(39) te Velde, G.; Bickelhaupt, F. M.; Gisbergen, S. J. A. v.; Guerra, C. F.;
Baerends, E. J.; Snijders, J. G.; Ziegler, T. J. Comput. Chem. 2001,
22, 931–967.
(40) Guerra, C. F.; Snijders, J. G.; te Velde, G.; Baerends, E. J. Theor.
Chem. Acc. 1998, 99, 391.
(41) Laaksonen, L. J. Mol. Graphics 1992, 10, 33.
(42) Bergman, D. L.; Laaksonen, L.; Laaksonen, A. J. Mol. Graphics
Modelling 1997, 15, 301.
9
J. AM. CHEM. SOC. VOL. 130, NO. 20, 2008 6569

A R T I C L E S
Lam et al.
Figure 2. Resonance structures of complex 2, 2a-2d.
Figure 3. Comparison of selected bond distances of complexes 2 and 3.
Table 2. Comparison of Computed (DFT) and Actual Bond
Lengths (XRD) for Complexes 2 and 3 (Bond Distances in
Angstroms, Angles in Degrees)
structural parameters
2
computed 2
3
computed 3
U1-Nav
2.683
2.207
2.178
1.334
2.733
2.217
2.190
1.328
159.9
2.700
2.215
2.080
1.385
2.763
2.228
2.105
1.414
175.7
Figure 5. Ligand-field f orbital splitting diagram of 3, an f2 system with
an idealized C3 symmetry and a strong axial ligand.
U1-Oav
U1-O4
C70-O4
U1-O4-C70
159.6
175.5
ligand is observed. Such a spin polarization on the ligand is
often observed for high valent transition metal complexes.43
In C3 symmetry, each of the seven f orbitals are either totally
symmetric (a) or part of a degenerate pair (e). The two π orbitals
transform together and form one e set. The two δ orbitals also
transform together and form another e set. The z3 is unchanged
under a 120° turn and thus is totally symmetric in C3. The ꢀ
orbitals [y(y2-3x2) and x(x2-3y2)] are also unchanged under a
120° turn (refer to figure) and, thus, also totally symmetric in
C3. Since the ꢀ orbitals do not transform together, they may
have individual energies. Their chemical difference is also
apparent since one ꢀ orbital has perfect σ overlap with the
ligands in the plane, while the other has perfect π overlap with
them.
2
2
2
2
(Figure 5). The fy(y -3x ) and fx(x -3y ) orbitals are 3-fold ꢀ
symmetry with respect to the U-O axis and are well matched
for σ and π antibonding overlaps with ligands in the aryloxide
plane. Hence, these orbitals lie higher in energy and are empty.
2
2
The next two higher energy orbitals, fxz and fyz (π antibonding),
3
and the highest orbital, fz (σ antibonding), are also empty.
The total spin density plot of U(IV) f2 complex 3 confirms
2
2
that the fxyz and fz(x -y ) orbitals indeed carry the main part of
the spin; however, a small spin polarization on the coordinated
The ketyl radical complex 2 was modeled to have three
unpaired electrons by computing it as a U(III) complex.
However, the resulting orbitals and spin density plot suggest a
more complex representation (Figure 6). SOMO-2 and SOMO-
2
2
1, of δ-type fxyz and fz(x -y ), are similar to the SOMOs of the
U(IV) complex 3 and are purely uranium f orbital in character.
Conversely, SOMO is not simply a pure uranium f orbital but
rather only 25.8% uranium based (24.0% f and 1.8% d); the
remaining Mulliken AO occupation of the orbital is ligand based.
3
The uranium contribution of the orbital is an fz directionally
optimized for π overlap with the ligand carbonyl oxygen π
orbitals. The ligand contribution shows density exactly where
Figure 4. DFT isosurface contour plots featuring the two degenerate
SOMOs (right), the LUMO (top left), and the resulting spin density (bottom
left) of [((t-BuArO)3tacn)UIV(OCHt-BuPh2)] (3).
(43) Astashkin, A. V.; Neese, F.; Raitsimring, A. M.; Cooney, J. J. A.;
Bultman, E.; Enemark, J. H. J. Am. Chem. Soc. 2005, 127, 16713–
16722.
9
6570 J. AM. CHEM. SOC. VOL. 130, NO. 20, 2008

XRD Studies on a Benzophenone Ketyl Complex of Uranium
A R T I C L E S
Figure 6. DFT isosurface contour plots featuring SOMO-1 and SOMO-2
of metal character (top), the lowest energy SOMO of mainly ligand
character (bottom right), and the resulting spin density (bottom left) of
[((t-BuArO)3tacn)UIV(OC•t-BuPh2)] (2).
one would expect for a reduced diphenyl ketone, that is, mainly
on the ketyl carbon and on the ortho- and para-carbon atoms
in the phenyl rings. Spin and charge densities confirm the
intermediate oxidation state of the uranium center. The Mulliken
charge on the uranium of 1.46 is between those found for the
precursor complex [((t-BuArO)3tacn)UIII] (1) (1.04) and the U(IV)
methoxide complex 3 (1.61). Likewise, the spin densities of
complex 2 (2.37) is between that of complex 1 (3.11) and 3
(2.17). The intermediate oxidation state of the uranium ion may
explain the shift to the UV region of uranium based 5f3 to 5f26d1
transitions commonly observed in the visible region of the
electronic absorption spectra of the [((RArO)3tacn)UIII] system,
which give rise to their red-brown color.44,45 This agreement
between structural, spectroscopic, and the computational data
of these complexes verifies that the scaled ZORA computation
correctly reflects the electronic structure of complex 2.
The DFT calculations have further affirmed that the charge-
separated species is stabilized by a high degree of resonance
and agrees well with structural parameters found in the
molecular structure of complex 2. The spin density map of
complex 2 displays the delocalization of the single unpaired
electron throughout the ligand, including ortho and para
positions of the axial ligand’s phenyl rings. In addition, one-
third of the electron has been found to reside on the uranium
metal center, confirming the resonance structures proposed for
2 in Figure 2. The U1-O4-C70 angle of ketyl complex 2 is
significantly more bent (159.6°) than that of the methoxide
complex 3 (178.4°) and can be attributed to a significant
contribution of the U(III) resonance structure (2d) to the overall
molecular structure of 2 (Figure 2).
Figure 7. Temperature-dependent SQUID magnetization data (1 T) for
complex 2 (top) plotted as magnetic moment (µeff) versus temperature (T)
(blue and magenta) along with uranium CO2 radical anion complex
[((AdArO)3tacn)UIV(CO2•-)] for comparison (green). The data for complex
3 is also shown (bottom). Data were corrected for diamagnetism, and
reproducibility was checked on multiple independently synthesized samples.
in magnetic moment at low temperatures is consistent with a
poorly isolated singlet ground state arising from ligand field
effects.46 The higher magnetic moment at low temperatures of
2 (5 K, µeff ) 1.61 µB) is unusual and is likely due to magnetic
contributions from the single unpaired electron residing on the
disubstituted benzophenone ligand fragment as well as from the
U(III) resonance structure (2d) as previously discussed (Figure
7, top). The room temperature magnetic moment is higher (300
K, µeff ) 3.48 µB) than that found in most U(IV) 5f2 complexes.
This is consistent with the DFT calculations, in which one-
third of the single radical electron is located on the uranium,
hence, raising the magnetic moment. The low temperature
magnetic behavior of 2 is similar to that of the η1-bound uranium
CO2 complex [((AdArO)3tacn)UIV(CO2•-)] previously reported,47
which has also been characterized as a charge-separated U(IV)
complex containing a radical anionic ligand. In conclusion,
magnetization studies reveal that complex 2 is no ordinary U(IV)
5f2 complex and should be viewed in its own class spectro-
scopically as a charge-separated U(III)-L T U(IV)-L•- 5f3
T 5f2 complex.
Magnetism of 2 and 3. Variable temperature magnetization
data were measured on two independently synthesized samples.
The data of complex 2 show a steady drop in µeff as the
temperature is lowered, decreasing from 3.48 µB at 300 K to
1.61 µB at 5 K. Complexes of U(IV) (5f2) commonly show a
variable temperature magnetization trace that exhibits a µeff
range from ∼0.5 µB at 4 K to ∼3.0 µB at 300 K.37 The reduction
In contrast, the magnetization data of complex 3 reveal a
temperature dependency that is much different than that of 2,
keeping in accordance with other U(IV) 5f2 of the (RArO)3tacn
ligand system bearing closed-shell axial ligands.37 The low
temperature magnetic moment for complex 3 (µeff ) 1.14-2.81
(44) Karbowiak, M. J. Phys. Chem. A 2005, 109, 3569–3577.
(45) Karbowiak, M.; Drozdzyn´ski, J. J. Phys. Chem. A 2004, 108, 6397–
6406.
(46) Stewart, J. L.; Andersen, R. A. New J. Chem. 1995, 19, 587–595.
(47) Castro-Rodriguez, I.; Nakai, H.; Zakharov, L. N.; Rheingold, A. L.;
Meyer, K. Science 2004, 305, 1757–1759.
9
J. AM. CHEM. SOC. VOL. 130, NO. 20, 2008 6571

A R T I C L E S
Lam et al.
The pale green color of complex 3 signifies the loss of
extended conjugation of the π system present in ketyl complex
2 due to H abstraction. Consequently, the electronic absorption
spectrum of 3, recorded in toluene, shows the disappearance of
the π-π* transition from the visible region of the electromag-
netic spectrum. Instead, only weak f-f transitions commonly
observed for lanthanide and actinide complexes remain, scattered
throughout the NIR region (Figure 8, bottom).49 The UV/vis/
NIR spectrum of 3 compares well with reported U(IV)
complexes38,50 and demonstrates that electronic absorption
spectroscopy is diagnostic of oxidation state and 5fn electron
configuration.
Synthesis and Molecular Structure of 6 and 7. In light of the
difficulties of isolating the benzophenone ketyl radical complex
with the (t-BuArOH)3tacn ligand system, we attempted to
stabilize the unsubstituted benzophenone ketyl radical by using
amorestericallyencumberingadamantylderivatizedligand(AdArO)3-
tacn3- (trianion of 1,4,7-tris(3-adamantyl-5-tert-butyl-2-hy-
droxybenzyl)-1,4,7-triazacyclononane). Treating a red-brown
frozen benzene solution of trivalent [((AdArO)3tacn)UIII] (4) with
1 equiv of benzophenone results in an intense purple color that
persists for ∼15 s after the solution is thawed before turning
greenish-brown. As judged from its intense colorization, this
transient species is presumably a uranium benzophenone ketyl
radical complex [((AdArO)3tacn)UIV(OC•Ph2)] (5) formed from
an intramolecular single-electron reduction of benzophenone by
U(III) (Scheme 2) analogous to that seen in the formation of
complex 2. Similarly, the intense purple color is attributed to
the π-π* transition of the highly conjugated benzophenone
ketyl radical ligand fragment (λmax ) 550 nm) and is comparable
to that of the purple sodium benzophenone ketyl radical in
toluene (λmax ) 624 nm) (Figure 9). Allowing the reaction to
proceed for 12 h yields a green solution. From this solution,
green solids are isolated by solvent evaporation, and single
crystals suitable for X-ray diffraction were obtained from
diffusion of acetonitrile into a benzene solution. The molecular
structure analysis reveals a U(IV) diphenylmethoxide complex
[((AdArO)3tacn)UIV(OCHPh2)] (6) (Figure 1) similar to 3. One
possible source of hydrogen atom stems from complex decom-
position, which also might explain the low isolated yield of 57%.
Complex 6 can be reproduced in significantly higher yield when
the reaction is done in the presence of hydrogen sources, such
as THF, 1,4-cyclohexadiene, or R3Sn-H (isolated 75% yield).
X-ray diffraction study of 6 reveals a seven-coordinate
uranium complex. The coordination sphere displays the uranium
center encased almost entirely by the macrocyclic hexadentate
ligand, leaving only the seventh axial coordination site available,
where the ligand binds roughly perpendicular to the trisaryloxide
plane. The average U-O(ArO) and U-N(tacn) distances of
2.216 and 2.679 Å (Table 1) are consistent with other U(IV)
complexes of the [((RArO)3tacn)UIV(L)]-type.37 The U1-O4
and C70-O4 bond lengths of 2.092(3) and 1.386(5) Å in 6,
respectively, are in good agreement with compound 3. For
reference, the uranium out-of-plane shift for the six-coordinate
starting complex 4 is Uoop ) -0.88 Å. The uranium out-of-
plane shift for 6 of -0.163 Å is a significant shift from the
six-coordinate U(III) complex, which is consistent with the
strong binding of the methoxide ligand to a highly oxophilic
uranium center. The U1-O4-C70 angle in 6 of 171.9(3)° is
Figure 8. Electronic absorption spectrum of 2 in toluene (top, magenta)
plotted as extinction coefficient (ꢀ) versus wavelength (λ) along with the
spectrum of the sodium benzophenone ketyl radical (blue). The spectrum
of 3 in toluene is also shown (bottom).
µB, 2-300 K) is lower than that of complex 2 (Figure 7, bottom)
but still higher than most [((RArO)3tacn)UIV(L)] complexes,
which usually exhibit a magnetic moment of 0.5 µB at 5 K.37
This could be due to compound 3 possessing magnetic states
lying closer together in energy than others and, thus, get
populated faster at lower temperatures. The general temperature
dependence and room temperature moment of 2.81 µB is typical
for most U(IV) 5f2 complexes but strays from theoretical values
3
calculated for a H4 ground state µeff ) gj(J(J + 1))1/2 ) 3.58
µB. This can be attributed to the quenching of spin-orbit
coupling of uranium complexes leading to a reduced magnetic
moment. Accordingly, U(IV) complexes typically have low
observed magnetic moment compared to theoretical values.
Electronic Absorption of 2 and 3. The majority of the U(IV)
complexes of the trisaryloxide ligand system, reported in the
past, have appeared as pale green solids, almost colorless in
appearance. Hence, the isolation of a deep purple powder
immediately indicates the formation of the uranium ketyl radical
complex similar to the Na-ketyl radical produced for the solvent
drying and deoxygenation process. The electronic absorption
spectrum of 2 in toluene features an intense absorption band
(λmax ) 562 nm, ꢀ ) 5680 M-1 cm-1) in the visible region
(Figure 8, top). This absorption gives rise to the ketyl complex’s
distinctive and intense purple color and is assigned to the π-π*
transition of a highly conjugated ketyl system. The transition
energy of 562 nm is blue-shifted from that of the sodium
benzophenone ketyl radical, which exhibits the π-π* transition
at 618 nm. The shift of λmax to a higher energy wavelength
upon coordination to the uranium metal has also been observed
with previously reported metal ketyl radical complexes.16,17 The
electronic absorption spectrum of the sodium benzophenone
ketyl radical is similar to those reported in the literature.13,48
(49) Marks, T. J. Progress in Inorganic Chemistry; John Wiley & Sons:
New York, 1979; Vol. 25.
(48) Kawai, A.; Hirakawa, M.; Abe, T.; Obi, K.; Shibuya, K. J. Phys. Chem.
A 2001, 105, 9628–9636.
(50) Berg, J. M.; Sattelberger, A. P.; Morris, D. E. Inorg. Chem. 1993, 32,
647–653.
9
6572 J. AM. CHEM. SOC. VOL. 130, NO. 20, 2008

XRD Studies on a Benzophenone Ketyl Complex of Uranium
A R T I C L E S
Scheme 2. Formation of Diphenyl Methoxide Complex 6 and Para-Coupling Dimeric Complex 7 through Intermediate Ketyl Complex 5 from
a One-Electron Reduction of Benzophenone by a U(III) Precursor Complex [((AdArO)3tacn)UIII] (4)
similar to that seen in 3 of 178.4(3)°. Overall, the metrics of 6
are reminiscent of those of 3; this is expected since both
complexes have similar electronic structures.
separated uranium complexes.47,52 These charge-separated
complexes containing radical anionic ligands are appreciated
because they are rare and spectroscopically interesting and,
due to their highly reactive nature, have the potential to
undergo further chemistry.52 Isolating and probing the
electronic structures of these charge-separated complexes are
therefore important endeavors if one wants to understand and
control reactivity. Structural characterization of the ketyl radical
complex [((t-BuArO)3tacn)UIV(OC•t-BuPh2)] (2) along with DFT
calculations have provided us more insight into the reactivity
In addition to the formation of complex 6 from the reaction
of 4 with benzophenone, dinuclear complex [((AdArO)3tacn)UIV-
(OCPhPh-CPh2O)UIV((AdArO)3tacn)] (7) also forms as a
crystalline byproduct in low yield (5%) (Figure 10). Single
crystals suitable for XRD studies of 7 were obtained from slow
diffusion of acetonitrile into a toluene solution of crude 6.
Evidently, the formation of 7 results from head-to-tail para
coupling of ketyl radical complex 5. In 7, where U1 bears the
sp2-hybridized C70, the U1-O4 bond of 2.168(4) Å and the
O4-C70 bond length of 1.366(7) Å are reminiscent of complex
2. The U2 in 7 bears the sp3-hybridized C71 and correspondingly
the U2-O8 bond of 2.109(4) Å and O8-C71 bond length of
1.423(7) Å are reminiscent of complex 3. Breaking of aroma-
ticity can be observed in the nonplanar bridged ring, where there
are four long single bonds and two short double bonds. Complex
7 represents the first structurally characterized product resulting
from para coupling of two benzophenone ketyl radical species.
Pinacol coupling is typically preferred for ketyls; however, there
are a few cases where para coupling takes preference due to
steric demands,9,51 such is the case here with the bulky
adamantyl groups in the [((AdArO)3tacn)UIII] system.
Conclusion
Stabilization and structural characterization of a uranium
benzophenone ketyl radical complex were accomplished by
employing the sterically demanding (RArO)3tacn ligand
system. It has been demonstrated in previous works that the
[((RArO)3tacn)UIII) system can stabilize highly reactive charge-
Figure 10. Molecular structure of complex 7 in crystals of 7·6 CH3CN/5
C7H8 (top) and the core with selected bond lengths (bottom). Thermal
ellipsoids are at the 35% probability level. Hydrogens and co-crystallized
solvent molecules are omitted for clarity.
Figure 9. Electronic absorption spectrum of ketyl intermediate complex 5
(magenta) and sodium benzophenone ketyl (blue).
9
J. AM. CHEM. SOC. VOL. 130, NO. 20, 2008 6573

A R T I C L E S
Lam et al.
Figure 11. Molecular structure of [((AdArO)3tacn)UIV(OCHPh2)] (6) in crystals of 6·CH3CN/0.5 C6H6. Thermal ellipsoids are at the 35% probability level.
Hydrogens (except H1) and co-crystallized solvent molecules are omitted for clarity.
of the [((RArO)3tacn)UIII] system. In addition to examining the
Kohn-Sham DFT calculations were performed on complexes 2
and 3 using the ADF package (2007.01). Analyses and geometry
optimizations of the complexes were performed employing the
BP86 gradient corrected functional53–55 in the scalar zeroth-order
regular approximation (ZORA56–58). The basis set applied is of
TZP quality with frozen cores for the second row elements (1s)
and for uranium (including 5d). In the ZORA computation, the R
and ꢂ spin densities are the sum of all densities of the R and ꢂ
spin-orbitals, respectively. The spin density plots are obtained by
taking the difference between the R and the ꢂ spin densities. ADF
input files are provided in the electronic Supporting Information.
Starting Materials. [(THF)4UI3] and [U(N(SiMe3)2)3] were
prepared as described by Clark et al.59–61 Uranium turnings were
purchased from Oak Ridge National Laboratory (ORNL) and
activated according to literature procedures. Benzophenone (98%)
was purchased from Aldrich and used as received after appropriate
degassing procedures. 4,4′-Di-tert-butylbenzophenone was synthe-
sized according to literature procedures.62
reactivity of complex 2, we also plan to isolate and study other
charge-separated complexes containing radical anions. We are
currently investigating a wide variety of ligands with different
sterics in hopes of achieving the delicate balance between
stabilizing these highly reactive complexes and promoting
controlled reactivity.
Experimental Section
General Methods. All experiments were performed under dry
nitrogen atmosphere using standard Schlenk techniques or an
MBraun inert gas glovebox. Solvents were purified using a two-
column solid-state purification system (Glasscontour System, Irvine,
CA) and transferred to the glovebox without exposure to air. NMR
solvents were obtained from Cambridge Isotope Laboratories,
degassed, and stored over activated molecular sieves prior to use.
Methods. Magnetism data of crystalline powdered samples
(20-30 mg) were recorded with a SQUID magnetometer (Quantum
Design) at 10 kOe (5-300 K for 2 and 6 and 2-300 K for 3).
Values of the magnetic susceptibility were corrected for the
underlying diamagnetic increment (ꢁdia ) -986.01 × 10-6 cm3
mol-1 (2), -988.94 × 10-6 cm3 mol-1 (3), -885.84 × 10-6 cm3
mol-1 (6)) by using tabulated Pascal constants and the effect of
the blank sample holders (gelatin capsule/straw). Samples used for
magnetization measurement were recrystallized multiple times and
checked for chemical composition and purity by elemental analysis
(C, H, and N) and 1H NMR spectroscopy. Data reproducibility was
also carefully checked on independently synthesized samples.
1H NMR spectra were recorded on JEOL 270 and 400 MHz
instruments operating at respective frequencies of 269.714 and
400.178 MHz with a probe temperature of 23 °C in C6D6. Chemical
shifts were referenced to protio solvent impurities (δ 7.15 (C6D6))
and are reported in parts per million.
Complex Synthesis. [((t-BuArO)3tacn)UIV(OC•t-BuPh2)] (2). To
a red-brown hexane solution (6 mL) of [((t-BuArO)3tacn)UIII] (1)
(200 mg, 0.196 mmol) was added 4,4′-di-tert-butylbenzophenone
(57.7 mg, 0.196 mmol) with stirring. Immediately, purple precipitate
formed in a purple solution. The reaction was stirred at room
temperature for 1 h. The purple solids were isolated by vacuum
filtration and washed with small amounts of hexane and dried under
vacuum. Recrystallization by diffusion of acetonitrile into a
tetrahydrofuran solution afforded purple single crystals as 2·3C4H8O
(yield 188 mg, 0.143 mmol, 73%): 1H NMR (290 MHz, benzene-
(53) Becke, A. D. Phys. ReV. A 1988, 38, 3098.
(54) Perdew, J. P. Phys. ReV. B 1986, 33, 8822.
(55) Perdew, J. P. Phys. ReV. B 1986, 34, 7406.
(56) Lenthe, E. V.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1993,
99, 4597.
(57) Lenthe, E. V.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1994,
101, 9783.
(58) Lenthe, E. V.; Ehlers, A. E.; Baerends, E. J. J. Chem. Phys. 1999,
110, 8943.
(59) Avens, L. R.; Bott, S. G.; Clark, D. L.; Sattelberger, A. P.; Watkin,
J. G.; Zwick, B. D. Inorg. Chem. 1994, 33, 2248–2256.
(60) Clark, D. L.; Sattelberger, A. P.; Andersen, R. A. Inorg. Synth. 1997,
31, 307–315.
(61) Clark, D. L.; Sattelberger, A. P.; Bott, S. G.; Vrtis, R. N. Inorg. Chem.
1989, 28, 1771–1773.
(62) Gibson, H. W.; Lee, S.-H.; Engen, P. T.; Lecavalier, P.; Sze, J.; Shen,
Y. X.; Bheda, M. J. Org. Chem. 1993, 58, 3748–3756.
Electronic absorption spectra were recorded from 200 to 2000
nm (Shimadzu (UV-3101PC)) in the indicated solvent.
Results from elemental analysis were obtained from the Analyti-
cal Laboratories at the Friedrich-Alexander-University Erlangen-
Nu¨rnberg (Erlangen, Germany) on Euro EA 3000.
(51) Maury, O.; Villiers, C.; Ephritikhine, M. Tetrahedron Lett. 1997, 38,
6591–6594.
(52) Lam, O. P.; Feng, P. L.; Heinemann, F. W.; O’Connor, J. M.; Meyer,
K. J. Am. Chem. Soc. 2008, 130, 2806–2816.
9
6574 J. AM. CHEM. SOC. VOL. 130, NO. 20, 2008

XRD Studies on a Benzophenone Ketyl Complex of Uranium
A R T I C L E S
d6, 20 °C) δ ) 69.06 (d, 2H, ∆ν1/2 ) 50.4 Hz), 32.17 (s, 6H, ∆ν1/2
) 65.7 Hz), 8.45 (s, 18H (t-Bu H’s on BP), ∆ν1/2 ) 33.1 Hz),
0.96 (d, 9H, ∆ν1/2 ) 30.5 Hz), -1.29 (s, 3H, ∆ν1/2 ) 27.1 Hz),
-2.38 (s, 3H, ∆ν1/2 ) 30.4 Hz), -2.57 (s, 27H (2-t-Bu H’s), ∆ν1/2
) 29.7 Hz), -3.13 (s, 3H, ∆ν1/2 ) 28.6 Hz), -4.61 (s, 2H, ∆ν1/2
) 25.9 Hz), -6.43 (s, 27H (4-t-Bu H’s), ∆ν1/2 ) 44.3 Hz), -7.21
(s, 1H, ∆ν1/2 ) 33.2 Hz), -10.11 (s, 2H, ∆ν1/2 ) 40.5 Hz), -19.42
(s, 1H, ∆ν1/2 ) 33.8 Hz). Anal. Calcd for 2: C, 65.83; H, 7.98; N,
3.20. Found: C, 65.81; H, 7.99; N, 3.37.
molecules is also disordered. Two preferred orientations have been
refined, resulting in occupancies of 82.5(8)% for O7-C84 and
17.5(8)% for O7A-C84A. SAME restraints were applied in the
refinement of the THF molecules. Additional SIMU restraints were
applied in the refinement of the disordered THF. The residual peak
and hole electron densities were 1.966 and -2.954 e ·Å-3. The
absorption coefficient was 2.123 mm-1. The least-squares refine-
ment converged normally with residuals of R1 ) 0.0726, wR2 )
0.1273, and GOF ) 1.098 (all data). C84H128N3O7U, monoclinic,
space group P21/c, a ) 14.794(2), b ) 18.245(2), c ) 29.226(4)
Å, ꢂ ) 96.198(7)°, V ) 7842(2) Å3, Z ) 4, Fcalcd ) 1.296 mg/m3,
F(000) ) 3204, R1(F) ) 0.0480, wR2(F2) ) 0.1138 [I > 2σ(I)].
CCDC reference number: 676962.
Crystallographic details for 3: Green rectangular shaped crystals
grown from slow diffusion of acetonitrile into benzene at room
temperature were coated with Paratone N oil on a microscope slide.
A crystal of approximate dimensions 0.29 × 0.28 × 0.27 mm3
was selected and mounted on a nylon loop. A total of 101 493
reflections (-13 e h e 12, -33 e k e 33, -62 e l e 62) were
collected at T ) 150(2) K in the θ range from 3.24 to 26.37°, of
which 14 880 were unique (Rint ) 0.0465) and 10 662 were
observed [I > 2σ(I)] on a Bruker-Nonius KappaCCD diffractometer
using Mo KR radiation (λ ) 0.71073 Å). The structure was solved
by direct methods (SHELXTL NT 6.12, Bruker AXS, Inc., 2002).
All non-hydrogen atoms were refined anisotropically. Hydrogen
atoms were placed in calculated idealized positions. Two of the
t-butyl groups are disordered. Two alternative positions have been
refined in each case, resulting in site occupancies of 56(2) and
44(2)% for C19-C21 and C19A-C21A and of 82(2) and 18(2)%
for C60-C62 and C60A-C62A, respectively. SIMU, ISOR, and
SADI restraints were applied in the refinement of one of these
t-butyl groups (C59, C60-C62, C60A-C62A). The residual peak
and hole electron densities were 1.256 and -1.669 e ·Å-3. The
absorption coefficient was 2.266 mm-1. The least-squares refine-
ment converged normally with residuals of R1 ) 0.0748, wR2 )
0.0809, and GOF ) 1.173 (all data). C80H114N4O4U, orthorhombic,
space group Pbca, a ) 10.772(2), b ) 27.180(3), c ) 50.052(4)
Å, V ) 14654(3) Å3, Z ) 8, Fcalcd ) 1.300 mg/m3, F(000) ) 5968,
R1(F) ) 0.0484, wR2(F2) ) 0.0750 [I > 2σ(I)]. CCDC reference
number: 676963.
Crystallographic details for 6: Green rectangular shaped crystals
grown from slow diffusion of acetonitrile into benzene at room
temperature were coated with Paratone N oil on a microscope slide.
A crystal of approximate dimensions 0.30 × 0.27 × 0.25 mm3
was selected and mounted on a nylon loop. A total of 61 689
reflections (-18 e h e 18, -29 e k e 28, -29 e l e 29) were
collected at T ) 150(2) K in the θ range from 2.92 to 27.10°, of
which 16 197 were unique (Rint ) 0.0325) and 12 987 were
observed [I > 2σ(I)] on a Bruker SMART diffractometer using
Mo KR radiation (λ ) 0.71073 Å). The structure was solved by
direct methods (SHELXTL NT 6.12, Bruker AXS, Inc., 2002). All
non-hydrogen atoms were refined anisotropically. Hydrogen atoms
were placed in calculated idealized positions. One of the t-butyl
groups is disordered. Two alternative positions have been refined
resulting in site occupancies of 73.8(8) and 26.2(8)% for C31-C33
and C31A-C33A, respectively. SIMU restraints were applied in
the refinement of two further t-butyl groups where no disorder could
be resolved. The residual peak and hole electron densities were
2.762 and -1.497 e ·Å-3. The absorption coefficient was 2.252
mm-1. The least-squares refinement converged normally with
residuals of R1 ) 0.0493, wR2 ) 0.1078, and GOF ) 1.063 (all
data). C87H113N4O4U, monoclinic, space group P21/n, a ) 14.5176(8),
b ) 22.682(2), c ) 23.334(2) Å, ꢂ ) 106.046(1)°, V ) 7384.3(7)
Å3, Z ) 4, Fcalcd ) 1.364 mg/m3, F(000) ) 3148, R1(F) ) 0.0384,
wR2(F2) ) 0.1046 [I > 2σ(I)]. CCDC reference number: 676964.
[((t-BuArO)3tacn)UIV(OCHt-BuPh2)] (3). Stirring a purple solu-
tion of [((t-BuArO)3tacn)UIV(OC•t-BuPh2)] (2) (200 mg, 0.152 mmol)
in neat 1,4-cyclohexadiene (1.5 mL) for 12 h resulted in a cloudy
green solution with a small amount of fine light-gray precipitate.
The light-gray precipitate was filtered off leaving a light-green
solution. The filtrate was evaporated to dryness under reduced
pressure and dried in vacuo. Recrystallization from a concentrated
pentane solution produced green block crystals suitable for X-ray
diffraction as 3·C5H12 (yield 140 mg, 0.106 mmol, 70%): 1H NMR
(290 MHz, benzene-d6, 20 °C) δ ) 51.88 (s, 3H, ∆ν1/2 ) 38.1
Hz), 50.47 (s, 3H, ∆ν1/2 ) 6.6 Hz), 15.94 (d, 3H, ∆ν1/2 ) 7.2 Hz),
5.57 (s, 6H, ∆ν1/2 ) 4.5 Hz), 3.97 (s, 9H (t-Bu H’s on BP), ∆ν1/2
) 3.8 Hz), 3.69 (s, 9H (t-Bu H’s on BP), ∆ν1/2 ) 4.0 Hz), 2.53 (s,
1H, ∆ν1/2 ) 3.3 Hz), 2.47 (s, 6H, ∆ν1/2 ) 5.1 Hz), 1.13 (s, 3H,
∆ν1/2 ) 2.5 Hz), -4.64 (s, 27H (2-t-Bu H’s), ∆ν1/2 ) 3.3 Hz),
-6.95 (s, 3H, ∆ν1/2 ) 5.5 Hz), -7.41 (s, 3H, ∆ν1/2 ) 4.5 Hz),
-8.11 (s, 2H, ∆ν1/2 ) 26.9 Hz), -16.09 (s, 27H (4-t-Bu H’s), ∆ν1/2
) 13.4 Hz). Anal. Calcd for 3: C, 65.78; H, 8.05; N, 3.20. Found:
C, 65.67; H, 8.08; N, 3.34.
[((AdArO)3tacn)UIV(OCHPh2)] (6). A benzene solution (0.5
mL) of benzophenone (29 mg, 0.16 mmol) was added to a thawing
benzene solution (10 mL) of [((AdArO)3tacn)UIII] (4) (200 mg, 0.16
mmol) while stirring. The reaction was allowed to stir at room
temperature for 4 h. Green solids were isolated by solvent
evaporation. Recrystallization of the green powder by slow diffusion
of acetonitrile into a benzene solution produces single rectangular
shaped crystals of 6·CH3CN/0.5 C6H6. The crystals can be isolated
by filtration and dried in vacuo to obtain the bulk solid (yield 120
mg, 0.084 mmol, 52%). Alternatively, 6 can also be synthesized
in higher yields by doing the reaction of [((AdArO)3tacn)UIII] (4)
(100 mg, 0.08 mmol) with benzophenone (15 mg, 0.08 mmol) in
neat 1,4-cyclohexadiene (6 mL). After 12 h, the green reaction
solution is filtered and evaporated to dryness. The crude solids are
washed with hexane and dried (yield 85 mg, 0.06 mmol, 75%): 1H
NMR (400 MHz, benzene-d6, 20 °C) δ ) 45.95 (d, 4H, ∆ν1/2
)
24.7 Hz), 5.83 (d, 6H, ∆ν1/2 ) 25.1 Hz), 3.23 (d, 6H, ∆ν1/2 ) 21.4
Hz), 2.26 (s, 9H, ∆ν1/2 ) 8.5 Hz), 1.21 (d, 4H, ∆ν1/2 ) 8.55 Hz),
1.18 (m, 12H, ∆ν1/2 ) 6.05 Hz), 0.91 (d, 3H, ∆ν1/2 ) 1.11 Hz),
0.84 (t, 12H, ∆ν1/2 ) 1.42 Hz), 0.05 (s, 3H, ∆ν1/2 ) 1.05 Hz),
-3.49 (s, 3H, ∆ν1/2 ) 2.55 Hz), -4.17 (s, 27H (t-Bu H’s), ∆ν1/2
) 2.51 Hz), -4.48 (s, 3H, ∆ν1/2 ) 2.52 Hz), -5.16 (s, 6H, ∆ν1/2
) 26.8 Hz), -5.29 (s, 3H, ∆ν1/2 ) 10.2 Hz), -26.78 (s, 6H, ∆ν1/2
) 25.4 Hz).
Crystallographic Details. Crystallographic details for 2: Purple
prismatic crystals grown from slow diffusion of acetonitrile into a
tetrahydrofuran solution at room temperature were coated with
Paratone N oil on a microscope slide. A crystal of approximate
dimensions 0.36 × 0.25 × 0.20 mm3 was selected and mounted
on a nylon loop. A total of 67 965 reflections (-18 e h e 18,
-22 e k e 22, -36 el e 36) were collected at T ) 100(2) K in
the θ range from 3.42 to 26.37°, of which 15 959 were unique (Rint
) 0.0844) and 11 992 were observed [I > 2σ(I)] on a Bruker-
Nonius KappaCCD diffractometer using Mo KR radiation (λ )
0.71073 Å). The structure was solved by direct methods (SHELXTL
NT 6.12, Bruker AXS, Inc., 2002). All non-hydrogen atoms were
refined anisotropically. Hydrogen atoms were placed in calculated
idealized positions. One of the t-butyl groups (C59-C62) of the
ligand is subjected to rotational disorder around C59. Two preferred
orientations have been refined resulting in occupancies of 74(2)%
for C60-C62 and 26(2)% for C60A-C62A. One of the THF
Crystallographic details for 7: Orange rectangular shaped crystals
grown from slow diffusion of acetonitrile into toluene at room
temperature were coated with Paratone N oil on a microscope slide.
A crystal of approximate dimensions 0.30 × 0.25 × 0.10 mm3
9
J. AM. CHEM. SOC. VOL. 130, NO. 20, 2008 6575

A R T I C L E S
Lam et al.
was selected and mounted on a nylon loop. A total of 69 300
reflections (-22 e h e 22, -23 e k e 23, -35 e l e 35) were
collected at T ) 100(2) K in the θ range from 2.94 to 27.10°, of
which 35 321 were unique (Rint ) 0.0521) and 24 345 were
observed [I > 2σ(I)] on a Bruker SMART diffractometer using
Mo KR radiation (λ ) 0.71073 Å). The structure was solved by
direct methods (SHELXTL NT 6.12, Bruker AXS, Inc., 2002). All
non-hydrogen atoms were refined anisotropically. Hydrogen atoms
were placed in calculated idealized positions. Two of the t-butyl
groups of the ligand are disordered. Two alternative positions each
have been refined to give occupancies of 53(1)% for C25-C27
and 47(1)% for C25′-C27′ and of 82(1)% for C107-C109 and
18(1)% for C167-C169. The crystal structure contains a total of 4
molecules of NCMe and 2.625 molecules of toluene, a number of
which is disordered. The disordered parts include sharing of
alternative positions for toluene or NCMe molecules only but also
sharing of a site between a toluene and an NCMe. SIMU and ISOR
restraints have been applied in modeling the disordered parts of
the structure. Detailed occupancies and further information can be
taken from the deposited data. The residual peak and hole electron
densities were 1.898 and -1.430 e ·Å-3. The absorption coefficient
was 2.017 mm-1. The least-squares refinement converged normally
with residuals of R1 ) 0.0973, wR2 ) 0.1504, and GOF ) 1.065
j
(all data). C190.38H245N10O8U2, triclinic, space group P1, a )
17.321(4), b ) 18.075(4), c ) 28.033(7) Å, R ) 102.643(3), ꢂ )
90.799(3), γ ) 104.456(3)°, V ) 8271(3) Å3
, Z ) 2, Fcalcd ) 1.316
mg/m3, F(000) ) 3411, R1(F) ) 0.0606, wR2(F2) ) 0.1360 [I >
2σ(I)]. CCDC reference number: 676965.
Acknowledgment. Special thanks to Arnie Rheingold, Peter
Gantzel, and Antonio DiPasquale for help with the crystal structures.
This work was funded by DOE Grant DE-FG02-O4ER 15537,
SFB583, and DFG. We thank the GAAN Fellowship Program for
support of O.P.L. This research was also supported in part by NSF
Grant CHE05-18707 (J.M.O.). K.M. congratulates Professor Bernt
Krebs (Uni Mu¨nster) and dedicates this paper on the occasion of
his 70th birthday.
Supporting Information Available: ADF input files. This
material is available free of charge via the Internet at http://
pubs.acs.org.
JA801007Q
9
6576 J. AM. CHEM. SOC. VOL. 130, NO. 20, 2008