Cerium(III) and cerium(IV) bis(η8-pentalene) sandwich complexes: Synthetic, structural, spectroscopic, and theoretical studies

Article abstract of DOI:10.1021/om7002738

The Ce(III) anionic bis(pentalene) sandwich complex K[Ce{C ₈H₄(SiⁱPr₃-1,4)₂} ₂] (1) has been prepared by treatment of CeCl₃ with K ₂[C₈H₄(SiⁱPr₃-1,4) ₂] and crystallographically characterized as its 18-crown-6 complex. Oxidation of 1 with Ag[BPh₄] affords the neutral, formally Ce(IV) sandwich complex [Ce{C₈H₄(SiⁱPr ₃-1,4)₂)₂] (2), whose molecular structure has also been determined. The electronic structure of 2 has been investigated in detail by a combination of magnetic studies, K-edge XANES measurements, gas-phase photoelectron spectroscopy, and density functional calculations.

Full text of DOI:10.1021/om7002738

Organometallics 2007, 26, 3111-3119
3111
Cerium(III) and Cerium(IV) Bis(η8-pentalene) Sandwich
Complexes: Synthetic, Structural, Spectroscopic, and
Theoretical Studies
Gabor Balazs,‡ F. Geoffrey N. Cloke,*,† Jennifer C. Green,*,‡ Robert M. Harker,†
Andrew Harrison,§, Peter B. Hitchcock,† Christian N. Jardine,‡ and Richard Walton‡
The Chemistry Department, School of Life Sciences, UniVersity of Sussex, Brighton, BN1 9QJ, U.K.,
Inorganic Chemistry Laboratory, UniVersity of Oxford, South Parks Road, Oxford, OX1 3QR, U.K., School
of Chemistry, UniVersity of Edinburgh, Joseph Black Building, West Mains Road, Edinburgh, Scotland
EH9 3JJ, and Institut Laue-LangeVin, 6 Rue Jules Horowitz, BP 156-38042, Grenoble Cedex 9, France
ReceiVed March 22, 2007
The Ce(III) anionic bis(pentalene) sandwich complex K[Ce{C8H4(SiiPr3-1,4)2}2] (1) has been prepared
by treatment of CeCl3 with K2[C8H4(SiiPr3-1,4)2] and crystallographically characterized as its 18-crown-6
complex. Oxidation of 1 with Ag[BPh4] affords the neutral, formally Ce(IV) sandwich complex
[Ce{C8H4(SiiPr3-1,4)2}2] (2), whose molecular structure has also been determined. The electronic structure
of 2 has been investigated in detail by a combination of magnetic studies, K-edge XANES measurements,
gas-phase photoelectron spectroscopy, and density functional calculations.
tion. The dominant contribution (∼80%) comes from the π-
(e2u)3fδ1 configuration, in which the two unpaired electrons are
antiferromagnetically coupled in two molecular orbitals, and the
direct product of their spatial symmetries is A1g. Thus, on this
basis, CeCOT2 is best described as having a single metal-
localized valence 4f electron, i.e., as a Ce(III) compound
containing two formally COT1.5- rings. Subsequent studies using
synchrotron radiation-based X-ray absorption spectroscopy
(XANES) also showed that the absorption K-edge of the cerium
center in both CeCOT2 and [CeCOT2]- lies in the trivalent rather
than the tetravalent range.10 Amberger has very recently shown
that optical spectroscopic properties of substituted cerocenes
may be more consistent with a bonding model based on
Ce(III) than Ce(IV).11 Andersen has also discussed the bonding
in CeCOT2 in terms of a molecular Kondo effect, based on
detailed magnetic susceptibility and LIII-edge X-ray absorption
spectroscopy data.12
Cerium complexes derived from the pentalene dianion would
offer an interesting comparison with the related cyclooctatet-
raene complexes discussed above, especially with regard to their
electronic structure. We have developed a synthetic routes to
1,4-trialkylsilyl-substituted pentalene ligands13 and reported the
synthesis and molecular structure of [Th(η-C8H4(SiiPr3-1,4)2],
a pentalene analogue of [ThCOT2].14 In this paper we describe
the synthesis of the analogous anionic cerium(III) and neutral
cerium(IV) pentalene sandwich complexes and a detailed study
of the electronic structure and bonding in the latter.
Introduction
There is currently considerable interest in cerium(IV) orga-
nometallic chemistry, specifically the controversy surrounding
the optimal assignment of oxidation state of cerium in such
complexes and the apparent incompatibility between a highly
oxidizing, formally Ce(IV) metal center and easily oxidized,
“soft” hydrocarbon ligands. Pre-eminent in this debate is the
molecule cerocene (CeCOT2, COT)η-C8H8), first characterized
by Streitweiser et al., who reported photoelectron spectroscopic
and computational studies consistent with a tetravalent formula-
tion, i.e., a Ce4+ center sandwiched between two aromatic
COT2- dianions.1 Raymond also argued strongly, primarily on
structural grounds, for such an ionic formulation of COT-based
complexes of the lanthanides and actinides.2,3 Several recent
structures of cerocenes and their anions with a variety of
substituents are largely consistent with this view.4-6 However,
detailed computational studies by Dolg and co-workers on
CeCOT2 and its actinide analogue ThCOT2 concluded that,
1
while the ground state of ThCOT2 is A1g with a dominant
configuration to this state of π(e2u)4f0d0, CeCOT2, while also
1
possessing a A1g ground state, does not have a dominant π-
(e2u)4 configuration.7-9 In fact, this configuration was found to
contribute only ∼20% to the ground-state wave func-
* Corresponding
authors.
E-mail:
f.g.cloke@sussex.ac.uk;
jennifer.green@chem.ox.ac.uk.
† University of Sussex.
‡ University of Oxford.
§ University of Edinburgh.
Institut Laue-Langevin.
(1) Streitwieser, A.; Kinsley, S. A.; Rigsbee, J. T.; Fragala, I. L.; Ciliberto,
E.; Ro¨sch, N. J. Am. Chem. Soc. 1985, 107, 7786.
(2) Hodgson, K. O.; Raymond, K. N. Inorg. Chem. 1972, 11, 3030.
(3) Raymond, K. N.; Eigenbrot, C. W., Jr. Acc. Chem. Res. 1980, 13,
276.
(8) Dolg, M.; Fulde, P.; Stoll, H.; Preuss, H.; Chang, A.; Pitzer, R. M.
J. Chem. Phys. 1995, 195, 71.
(9) Liu, W. J.; Dolg, M.; Fulde, P. J. Chem. Phys. 1997, 107, 3584.
(10) Edelstein, N. M.; Allen, P. G.; Bucher, J. J.; Shuh, D. K.; Sofield,
C. D.; Kaltsoyannis, N.; Maunder, G. H.; Russo, M. R.; Sella, A. J. Am.
Chem. Soc. 1996, 118, 13115.
(4) Boussie, T. R.; Eisenberg, D. C.; Rigsbee, J.; Streitwieser, A.; Zalkin,
A. Organometallics 1991, 10, 1922.
(5) Kilimann, U.; Herbst-Immer, R.; Stalke, D.; Edelmann, F. T. Angew.
Chem., Int. Ed. Engl. 1994, 33, 1618.
(6) Reissmann, U.; Lameyer, L.; Stalke, D.; Poremba, P.; Edelmann, F.
T. Chem. Commun. 1999, 1865.
(11) Amberger, H. D.; Reddmann, H.; Edelmann, F. T. J. Organomet.
Chem. 2005, 690, 2238.
(12) Booth, C. H.; Walter, M. D.; Daniel, M.; Lukens, W. W.; Andersen,
R. A. Phys. ReV. Lett. 2005, 95, 267202.
(13) Cloke, F. G. N.; Kuchta, M. C.; Harker, R. M.; Hitchcock, P. B.;
Parry, J. S. Organometallics 2000, 19, 5795.
(7) Dolg, M.; Fulde, P.; Kuchle, W.; Neumann, C. S.; Stoll, H. J. Chem.
Phys. 1991, 94, 3011.
(14) Cloke, F. G. N.; Hitchcock, P. B. J. Am. Chem. Soc. 1997, 119,
7899.
10.1021/om7002738 CCC: $37.00 © 2007 American Chemical Society
Publication on Web 05/23/2007

3112 Organometallics, Vol. 26, No. 13, 2007
Balazs et al.
were carried out using the Gaussian 03 package17 with the hybrid
B3LYP18,19 functional using the 6-31++G* basis sets20-26 for C
and H and the Dunning/Huzinaga27 type basis set with the Stuttgart/
Dresden pseudorelativistic core potential for Ce.28 For fragment
calculations the Amsterdam Density Functional program suite ADF
2005.01 was used.29,30 Scalar relativistic corrections were included
via the ZORA method.31-35 The generalized gradient approximation
was employed, using the local density approximation of Vosko,
Wilk, and Nusair36,37 together with the nonlocal exchange correction
by Becke38,39 and nonlocal correlation corrections by Perdew.40 TZP
basis sets were used with triple-accuracy sets of Slater-type orbitals
and two polarization functions added to the main group atoms. A
post-SCF gradient correction was applied.
The geometries of [Ce(C8H6)2] were optimized in the D2, D2d,
and D2h point groups. The optimized geometry within the restraints
of the D2 point group has D2d symmetry. Vertical ionization energies
were estimated, using the optimized structure, from the difference
between the total energy for the molecule and that for the molecular
ion in the appropriate state. For the geometry optimizations in
Gaussian 6-31++G* basis sets for C and H and SDD for Ce
together with the pseudorelativistic core potential of first 28
electrons of Ce were used.
Experimental Section
General Methods. Unless otherwise stated, all experimental
procedures were carried out using standard high-vacuum and
Schlenk techniques, under an atmosphere of dry argon (manipula-
tion of 1 requires Ar of 99.999+% purity) or under dinitrogen in
an MBraun or a Miller-Howe glovebox. Glassware was flame dried
under vacuum prior to use, and Celite filter aid was predried in an
oven at 200 °C. n-Pentane and diethyl ether were distilled from
sodium/potassium alloy, tetrahydrofuran and toluene were distilled
from potassium metal, and pyridine was distilled from CaH2 under
dinitrogen prior to use; toluene-d8 was dried over molten potassium,
then vacuum transferred to, and stored in, an ampule under
dinitrogen prior to use. NMR spectra were recorded at 295 K on a
Bruker DPX 300 MHz spectrometer, with chemical shifts (δ)
reported in ppm, relative to the chemical shifts of the internal
deuterated solvent (1H and 13C) set relative to external TMS.
Coupling constants are quoted in Hz. Electron impact mass spectra
were recorded on a VG Autospec mass spectrometer. Magnetic
measurements on 1 and 2 were carried out on a Quantum Design
SQUID magnetometer in O-ring-sealed KelF capsules at fields of
0.1 and 1 T over the temperature range 5-340 K.
Elemental analyses were carried out by Mikroanalytisches Labor
Pascher, Remagen, Germany, and the University of North London
Elemental Analysis Service, London, UK.
(17) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A.; Vreven, T.; Kudin, K. N.;
Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,
S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A.
D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A.
G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian
03, ReVision C.02; Gaussian, Inc.: Wallingford, CT, 2004.
(18) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37 (2), 785.
(19) Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Chem. Phys. Lett. 1989,
157, 200.
Anhydrous CeCl3 was prepared from CeCl3‚6H2O using tri-
methylsilyl chloride according to the method of Boudjouk.15
AgBPh4 was prepared by precipitation from aqueous solutions of
silver nitrate and sodium tetraphenylborate, followed by drying
under vacuum. 18-Crown-6 was purified by precipitation from
acetonitrile, drying under vacuum, and recrystallization from
heptane. K2[C8H4(SiiPr3-1,4)2]13 and [Ce(η-C8H5(SiMe3-1,3,5)3)2]
(Ce(COT′′′)2)3 were prepared as described elsewhere.
XANES Experiments. XANES data were measured at the
cerium K-edge (∼40.443 keV) on Station 9.2 of the Daresbury SRS.
The synchrotron operates with an average stored energy of 2 GeV
and a typical electron current of 200 mA. The incident X-ray energy
on the sample was selected using a double-crystal Si(220) mono-
chromator, and the second crystal of the monochromator was
detuned to 90% of the maximum intensity to reduce contributions
from higher order harmonics of the selected wavelength. Data were
collected in transmission mode from the cerium-containing materi-
als, with incident and transmitted X-ray intensities measured using
ionization chambers filled with appropriate quantities of noble gas.
The beam was defined as a 12 mm horizontal slit. Data were
measured from a CeB6 standard simultaneously with each sample
to provide a calibration; this was placed between the transmitted
ionization chamber and third ionization chamber. Solutions of Ce-
(COT′′′)2 (0.3 M in toluene), 1 (0.06 M in THF), and 2 (0.03 M in
toluene) were contained in sealed NMR tubes, and solid samples
of CeO2 and CeB6 were pressed into self-supporting discs with
∼50% by mass polyethylene powder. The Ce K-edge region was
scanned from 40.25 to 40.65 keV, and four scans from each sample
and from CeO2 were taken. Although the intrinsic resolution of
experiment, which broadens the features of the XANES region, is
on the order of 10 eV, the precision of the edge energy positions
can be determined with much greater accuracy, as previously
pointed out.10 The data were normalized and summed using the
program EXCALIB, and analysis was performed using the program
EXBACK.16 The edge position was defined as the point of inflection
of the near edge region and determined by measuring the position
of the maximum of the first derivative of X-ray absorption data.
(20) Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971, 54
(2), 724.
(21) Francl, M. M.; Pietro, W. J.; Hehre, W. J.; Binkley, J. S.; Gordon,
M. S.; DeFrees, D. J.; Pople, J. A. J. Chem. Phys. 1982, 77 (7), 3654.
(22) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28 (3), 213.
(23) Hariharan, P. C.; Pople, J. A. Mol. Phys. 1974, 27, 209.
(24) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56
(5), 2257.
(25) Rassolov, V. A.; Pople, J. A.; Ratner, M. A.; Windus, T. L. J. Chem.
Phys. 1998, 109 (4), 1223.
(26) Rassolov, V. A.; Ratner, M. A.; Pople, J. A.; Redfern, P. C.; Curtiss,
L. A. J. Comput. Chem. 2001, 22 (9), 976.
(27) Dunning, T. H.; Hay, P. J. In Modern Theoretical Chemistry;
Plenum: New York, 1976; Vol. 3, pp 1-28.
(28) Cao, X.; Dolg, M. THEOCHEM 2002, 581, 139.
(29) Te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca Guerra,
C.; Van Gisbergen, S. J. A.; Snijders, J. G.; Ziegler, T. J. Comput. Chem.
2001, 22, 931.
(30) Fonseca Guerra, C.; Snijder, J. G.; Te Velde, G.; Baerends, E. J.
Theor. Chem. Acc. 1998, 99, 391.
(31) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1993,
99, 4597.
(32) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1994,
101, 9783.
(33) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1996,
105, 6505.
(34) van Lenthe, E.; Ehlers, A.; Baerends, E. J. J. Chem. Phys. 1999,
110, 8943.
(35) van Lenthe, E.; van Leeuwen, R.; Baerends, E. J.; Snijders, J. G.
Int. J. Quantum. Chem. 1996, 57, 281.
(36) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.
(37) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.
(38) Becke, A. D. Phys. ReV. A 1988, 38, 3098.
(39) Becke, A. D. J. Chem. Phys. 1988, 88, 1053.
(40) Perdew, J. P. Phys. ReV. B 1986, 33, 8800.
DFT Computational Studies. Density functional calculations
(15) So, J.-H.; Boudjouk, P. Inorg. Chem. 1990, 29, 1592.
(16) Binsted, N.; Campbell, J. W.; Gurman, S. J.; Stephensen, P. C.
EXBACK, Daresbury, 1999.

Ce(III) and Ce(IV) Bis(η8-pentalene) Sandwich Complexes
Organometallics, Vol. 26, No. 13, 2007 3113
MHz, C7D8, 295 K, assignments in Figure 1): δ 169.11 (A1),
155.26 (A3), 152.46 (B2), 145.71 (B1), 136.68 (B3), 135.62 (A2),
123.23 (B4), 107.05 (A4), 20.06 (iPrCH3), 19.74 (iPrCH3), 19.63
(iPrCH3), 19.35 (iPrCH3), 17.59 (iPrCH), 15.49 (iPrCH).
X-ray Data Collection, Structure Solution, and Refinement.
Data collection was performed at 173(2) K on an Enraf-Nonius
Kappa CCD diffractometer, with graphite-monochromated Mo KR
radiation (λ ) 0.71073 Å). The molecular structure was solved by
direct methods and refined on F2 by full-matrix least-squares
techniques.
Figure 1. Ring-numbering scheme for NMR assignments.
Photoelectron Spectroscopy. Photoelectron spectra were mea-
sured using a PES laboratories 0078 spectrometer interfaced with
an Atari microprocessor. Spectra were calibrated with He, Xe, and
N2. In order to obtain a suitable vapor pressure, the sample was
held at a temperature of 236 °C.
Results and Discussion
Synthetic Studies. The reaction of CeCl3 with 2 equiv of
K2[C8H4(SiiPr3-1,4)2] in THF afforded tan-colored
K[Ce{C8H4(SiiPr3-1,4)2}2] in ca. 70% yield after workup.
K[Ce{C8H4(SiiPr3-1,4)2}2] is exceptionally sensitive to aerial
oxidation, even minute traces of air resulting in the formation
of a blue color attributed to 2 (Vide infra). Treatment of crude
K[Ce{C8H4(SiiPr3-1,4)2}2] with 18-crown-6 in toluene
yielded analytically pure, crystalline [Ce{C8H4(SiiPr3-1,4)2}2]-
[K(18-crown-6)] (1) after recrystallization from hot toluene (see
Scheme 1).
Preparation of [Ce{C8H4(SiiPr3-1,4)2}2]-[K(18-crown-6)]+, 1.
THF (50 cm3) was added to a stirred mixture of solid CeCl3 (0.26
g, 1.05 mmol) and K2[C8H4(SiiPr3-1,4)2] (1.03 g, 2.08 mmol). The
gold-colored mixture was stirred at room temperature for 18 h to
give a darker brown mixture. The reaction mixture was filtered
through Celite on a frit and washed through with a further 20 mL
of THF. Initial contact of the reaction mixture with the frit caused
the first few drops of solution to turn an intense blue color (due to
trace oxidation to 2); however the remainder of the mixture filtered
to a clear golden-brown solution. Solvent was removed in a vacuum
to give a greenish-brown solid, which was washed on a frit until
pentane washings, which were initially blue, became colorless. The
crude K[Ce{C8H4(SiiPr3-1,4)2}2] was again dried in a vacuum to
give a sand-colored powder, yield 0.71 g (68%). Toluene (7 mL)
was added to a mixture of crude K[Ce{C8H4(SiiPr3-1,4)2}2] (0.189
g, 0.19 mmol) and 18-crown-6 (0.050 g, 0.19 mmol) in an ampule.
Agitation of the mixture resulted in a yellow solution containing a
brown oil that crystallized after storing at 4 °C for several weeks.
The solution was decanted from the crystals, and the product was
washed with pentane, dried in a vacuum, and recrystallized from
hot toluene to afford pure 1. Anal. Calcd for [C64H116CeKO6Si4]:
C, 60.38; H, 9.18. Found: C, 59.95; H, 9.03.
Oxidation of in situ-generated K[Ce{C8H4(SiiPr3-1,4)2}2] with
1 equiv of Ag[BPh4] in THF resulted in formation of an intense
blue solution, from which blue-black crystals of the Ce(IV)
complex [Ce{C8H4(SiiPr3-1,4)2}2] (2) could be isolated (see
Scheme 1) after recrystallization from pentane; 2 may also be
purified by sublimation at 250 °C/10-5 mbar. The UV spectrum
of 2 in pentane solution displays a strong band (ꢀ ) 5 × 103
M-1 cm-1) at 590 nm, assignable to a LMCT transition.
2 exhibits 1H and 13C NMR solution spectra with sharp bands
as expected for a diamagnetic complex; HOESY experiments
allowed assignment of the ring protons and carbons for the
mixture of isomers, but could not distinguish between the
isomers. The ring-numbering scheme for chemical shift assign-
ments is shown in Figure 1 (the top and bottom rings in isomer
A and in isomer B are related by symmetry in solution).
The chemical shifts of the “wingtip” protons A4 and B4 (for
the two isomers A and B) are in a region of the spectrum similar
to those of [Th(η-C8H4(SiiPr3-1,4)2]14 and are essentially tem-
perature invariant. However, interestingly, the other ring protons
A3 and B3 occur at significantly higher field than those in the
thorium analogue, and their chemical shifts move to higher field
with increasing temperature (by ca. 0.5 ppm over the temper-
ature range 185-385 K). The respective 13C resonances do not
change. Variable-temperature (185-385 K) 13C NMR studies
revealed no significant changes in JCH values indicative of any
low-temperature agostic interactions. We deduce that there is a
significant paramagnetic contribution to the shielding arising
from the Ce center. This deduction is supported by the
temperature-independent paramagnetic (TIP) contribution to the
susceptibility and the small HOMO-LUMO gap (Vide infra).
Such shielding, analogous to that of transition metal hydrides,
is known to be potentially highly directional.41 The small
Crystals suitable for single-crystal X-ray diffraction were ob-
tained by vapor diffusion of pentane into a saturated solution of 1
in pyridine.
Preparation of [Ce{C8H4(SiiPr3-1,4)2}2] 2. A prereacted mixture
of CeCl3 (0.25 g, 1.0 mmol) and K2[C8H4(SiiPr3-1,4)2] (0.98 g, 2.0
mmol) in THF (50 cm3) was added to a slurry of AgBPh4 (0.43 g,
1.0 mmol) in THF (20 cm3), which resulted in an immediate color
change from red to deep blue. After stirring for 24 h the solvent
was removed in a vacuum and the product extracted with pentane.
Filtration of the dark blue solution, concentration, and cooling to
-45 °C for 3 days resulted in very dark blue crystals of 2, which
were washed in cold pentane and dried in a vacuum. Yield: 0.81
g (80%). Anal. Calcd for [C52H92CeSi4]: C, 64.40; H, 9.56.
Found: C, 64.38; H, 9.86. MS (EI, 70 eV): m/e 968 (M+, 20%).
1H NMR (300 MHz, C7D8, 295 K, assignments in Figure 1): δ
7.21 (1.9H, d, J ) 3.0 Hz, A4), 5.92 (1.6H, d, J ) 3.0 Hz, B4),
3.04 (1.5H, d, J ) 3.0 Hz, B3), 1.43 (11.3H, overlapping septets,
A+B SiCH{CH3}2), 1.23 (66.6H, overlapping doublets, A+B
SiCH{CH3}2), 0.24 (2.0H, d, J ) 3.0 Hz, A3). 13C{1H} NMR (75.5
Scheme 1

3114 Organometallics, Vol. 26, No. 13, 2007
Balazs et al.
Figure 2. Molecular structure of 1‚(pyridine)2 (thermal ellipsoids at 20%). The isopropyl groups have been omitted for clarity.
Figure 3. Molecular structure of 2 (thermal ellipsoids at 20%). The isopropyl groups have been omitted for clarity.
temperature variations are likely to be due to thermal motion
between conformers (see Structural Studies).
Structural Studies. Crystals of 1 suitable for X-ray diffrac-
tion were obtained by vapor diffusion of pentane into a saturated
solution of 1 in pyridine, and those of 2 were grown by slow
cooling of a saturated solution of 2 in pentane; the structures
are shown in Figures 2 and 3, respectively, with selected bond
distances and angles in Table 2.
The structure of 1 shows a (disorded, Vide infra) bis(η8-
pentalene)cerium anion with a remote K(18-crown-6)(pyridine)2
cation (Figure 2c). In the latter, the potassium atom lies in the
(41) Buckingham, A. D.; Stephens, P. J. J. Chem. Soc. 1964, 2747, 4583.

Ce(III) and Ce(IV) Bis(η8-pentalene) Sandwich Complexes
Organometallics, Vol. 26, No. 13, 2007 3115
Table 1. X-ray Data Collection Parameters for 1 and 2
four diastereomeric forms (meso and racemic isomers of the
staggered and of the eclipsed sandwich structures). However,
two possible diastereoisomers (the chiral form of the staggered
isomer and the meso form of the eclipsed isomer) are expected
to be precluded on steric grounds, due to unfavorable interac-
tions between the silyl ring substituents. This has been previ-
ously observed in the structure of [Th(η-C8H4(SiiPr3-1,4)2],
which exists as distinct, noninterconvertible staggered (meso
form) and semieclipsed (chiral form) isomers.14 These isomers
coexist in a single crystal of [Th(η-C8H4(SiiPr3-1,4)2], manifest
in a disorder of the pentalene framework, an effect ascribed to
domination of the crystal-packing forces by the bulky tri-
isopropylsilyl groups, whose disposition is virtually identical
in both isomers. The disorder was modeled by a 52%:48%
occupancy of the two pentalene ring orientations, i.e., staggered
(meso form) and semieclipsed (chiral form) isomers. Since the
bis(η8-pentalene)cerium sandwich structures in both 1 and 2
exhibit the same disorder and resultant gross features as
[Th(η-C8H4(SiiPr3-1,4)2], they will now be discussed together.
1‚(pyridine)2
2‚(pentane)
formula
fw
space group
a, Å
b, Å
c, Å
C
74H126CeKN2O6Si4
C52H92CeSi4
1041.88
C2/c (No. 15)
21.308(4)
13.376(5)
21.109(5)
90
95.93(2)
90
5984(3)
4
1431.35
I41/acd (No. 142)
24.6965(2)
24.6965(2)
53.5110(5)
90
R, deg
â, deg
γ, deg
V, Å3
Z
90
90
32637.3(5)
16
1.16
F
calc, g/cm3
1.16
T, K
173
0.71
4.28-25.01
173
0.87
2.09-22.99
µ(Mo KR), mm-1
θ range for data
collection, deg
total no. of data
no. of unique data
no. of reflns with I > 2σ(I)
no. of data/restraints/
params
110 448
7140
5432
4269
4150
3831
4150/6/279
7140/28/472
R1 [I > 2σ(I)]
0.050
0.122
0.67, -0.41
0.037
0.096
1.21, -1.05
wR2 [I > 2σ(I)]
max., min. peaks, e/Å3
In 1 the twist angles (as defined by the angles between the
two bridgehead C-C vectors) of the two pentalene rings are
44° for the semistaggered isomer (66% occupancy, Figure 2a)
and 10° for the virtually eclipsed isomer (34% occupancy, Figure
2b). Corresponding values for 2 are 85° for the virtually
staggered isomer (72% occupancy, Figure 3a) and 33° for the
semieclipsed isomer (28% occupancy, Figure 3b), twist angles
center of the crown ether ligand with K-O distances (2.787(5)-
2.811(4) Å) in the expected region. Axial pyridine molecules
complete the coordination sphere of the potassium with the K-N
distances at 2.868(6) Å and a N-K-N angle of 174.3(3)°.
The 1,4-bis(silylated) pentalene dianion has C2-symmetry and
is facially enantiotopic; thus a sandwich compound incorporating
two such ligands would be expected to exist, in principle, in
Table 2. Bond Distances and Angles in 1 and 2
1‚(pyridine)2
2
1‚(pyridine)2
2
Distances (Å) within Rings
Angles (deg) within Rings
C(1A)-C(2A)
C(1A)-C(5A)
C(2A)-C(3A)
C(3A)-C(4A)
C(4A)-C(5A)
C(4A)-C(8A)
C(5A)-C(6A)
C(6A)-C(7A)
C(7A)-C(8A)
C(1B)-C(2B)
C(1B)-C(5B)
C(2B)-C(3B)
C(3B)-C(4B)
C(4B)-C(5B)
C(4B)-C(8B)
C(5B)-C(6B)
C(6B)-C(7B)
C(7B)-C(8B)
1.423(10)
1.448(8)
1.417(11)
1.460(17)
1.429(10)
1.452(9)
1.422(14)
1.410(11)
1.434(9)
1.415(14)
1.446(12)
1.439(15)
1.46(2)
C(1)-C(2)
1.409(7)
1.444(7)
1.396(9)
1.473(7)
1.366(7)
1.473(7)
1.496(8)
1.519(8)
1.509(9)
1.534(8)
1.520(8)
1.526(7)
1.528(7)
1.530(7)
1.516(9)
1.525(9)
1.520(7)
1.522(7)
1.418(7)
1.442(7)
1.425(8)
1.435(11)
1.420(8)
1.425(8)
1.415(12)
1.425(14)
1.46(2)
C(2A)-C(1A)-C(5A)
C(3A)-C(2A)-C(1A)
C(2A)-C(3A)-C(4A)
C(5A)-C(4A)-C(8A)
C(5A)-C(4A)-C(3A)
C(8A)-C(4A)-C(3A)
C(6A)-C(5A)-C(4A)
C(6A)-C(5A)-C(1A)
C(4A)-C(5A)-C(1A)
C(7A)-C(6A)-C(5A)
C(6A)-C(7A)-C(8A)
C(7A)-C(8A)-C(4A)
C(2B)-C(1B)-C(5B)
C(1B)-C(2B)-C(3B)
C(2B)-C(3B)-C(4B)
C(5B)-C(4B)-C(8B)
C(5B)-C(4B)-C(3B)
C(8B)-C(4B)-C(3B)
C(6B)-C(5B)-C(4B)
C(6B)-C(5B)-C(1B)
C(4B)-C(5B)-C(1B)
C(5B)-C(6B)-C(7B)
C(6B)-C(7B)-C(8B)
C(7B)-C(8B)-C(4B)
104.3(6)
112.7(10)
105.4(12)
111.5(7)
107.5(8)
135.0(10)
106.5(7)
137.3(8)
110.1(7)
107.2(10)
112.7(8)
102.1(6)
104.3(10)
112.9(13)
104.9(16)
110.7(11)
107.5(12)
136.2(14)
106.6(11)
138.2(14)
110.4(11)
107.7(15)
111.4(12)
103.6(10)
C(2)-C(1)-C(4)
102.2(4)
114.1(6)
104.1(5)
135.1(4)
107.3(4)
82.5(3)
112.3(4)
104.4(5)
108.1(6)
135.9(5)
109.3(6)
105.8(5)
112.2(5)
104.6(9)
110.9(13)
136.2(12)
105.9(12)
105.3(10)
113.1(8)
C(1)-C(4)
C(2)′-C(2)-C(1)
C(2)-C(2)′
C(2)′-C(2)-C(3)′
C(2)-C(3)′
C(1)-C(2)-C(3)′
C(3)-C(4)
C(4)-C(3)-C(2)′
C(3)-C(2)′
C(4)-C(3)-Ce
C(5)-C(6)
C(3)-C(4)-C(1)
C(5)-C(7)
C(17)-C(14)-C(15)
C(16)′-C(15)-C(15)′
C(16)′-C(15)-C(14)
C(15)′-C(15)-C(14)
C(17)-C(16)-C(15)′
C(14)-C(17)-C(16)
C(171)-C(141)-C(151)
C(141)-C(151)-C(151)′
C(141)-C(151)-C(161)′
C(151)′-C(151)-C(161)′
C(171)-C(161)-C(151)′
C(141)-C(171)-C(161)
C(8)-C(9)
C(8)-C(10)
C(11)-C(12)
C(11)-C(13)
C(18)-C(19)
C(18)-C(20)
C(21)-C(22)
C(21)-C(23)
C(24)-C(26)
C(24)-C(25)
C(14)-C(17)
C(14)-C(15)
C(15)-C(16)′
C(15)-C(15)′
C(16)-C(17)
C(16)-C(15)′
C(141)-C(171)
C(141)-C(151)
C(151)-C(151)′
C(151)-C(161)′
C(161)-C(171)
C(161)-C(151)′
1.435(14)
1.446(14)
1.41(2)
1.420(16)
1.423(14)
Ce-Ring Distances (Å)
Ce-C(1A)
Ce-C(2A)
Ce-C(3A)
Ce-C(4A)
Ce-C(5A)
Ce-C(6A)
Ce-C(7A)
Ce-C(8A)
Ce-C(1B)
Ce-C(2B)
Ce-C(3B)
Ce-C(4B)
Ce-C(5B)
Ce-C(6B)
Ce-C(7B)
Ce-C(8B)
2.844(7)
2.960(7)
2.824(19)
2.592(11)
2.586(6)
2.836(15)
2.991(7)
2.895(6)
2.887(12)
2.921(11)
2.70(3)
Ce-C(2)
2.495(4)
2.496(6)
2.518(13)
2.730(4)
2.735(5)
2.760(11)
2.783(4)
2.784(5)
2.796(13)
2.886(5)
2.889(4)
2.925(8)
Ce-C(15)
Ce-C(151)
Ce-C(3)
1.464(17)
1.433(14)
1.464(17)
Ce-C(16)
Ce-C(161)
Ce-C(1)
Ce-C(14)
Ce-C(141)
Ce-C(17)
Ce-C(4)
2.47(2)
2.566(12)
2.88(3)
Ce-C(171)
2.956(12)
2.754(11)
Fold Angles (deg)
23
Twist Angles (deg)
85 (a), 33 (b)
22 (a), 21 (b)
44 (a), 10 (b)

3116 Organometallics, Vol. 26, No. 13, 2007
Balazs et al.
comparable to those found in [Th(η-C8H4(SiiPr3-1,4)2] (83° and
38°, respectively).
Pentalene ring carbon-carbon bond lengths (1.410(11)-
1.460(17) Å) in 1 and 2 fall within the range observed in
[Th{C8H4(SiiPr3-1,4)2}2] (1.36(2) to 1.49(2) Å),14 with carbon-
carbon bridgehead distances (1.429(10)-1.435(14) Å for the
two isomers of 1; 1.3969(9)-1.46(2) Å for the two isomers of
2) also comparable, within esds, to that of 1.39(2) Å in
[Th{C8H4(SiiPr3-1,4)2}2].14 Cerium-carbon bond distances in
1 and 2 range from 2.47(2) Å at the bridgehead to 2.991(7) Å
at the wingtips of the pentalene ligand, almost identical, within
esds, to the corresponding distances in [Th{C8H4(SiiPr3-1,4)2}2]
(2.543(10) and 2.908(11) Å, respectively).14 Remaining Ce-C
distances in 1 and 2 lie in the range 2.70(3)-2.895(6) Å, and
these compare well with those in [Th{C8H4(SiiPr3-1,4)2}2]
(2.748(10)-2.797(11) Å) and with those found in organosilyl-
substituted cyclopentadienyl cerium(III) compounds, such as
[Ce(η-C5H3{SiMe3}2-1,3)3] (average Ce-C distance 2.83 Å).42
An important metrical parameter with respect to η8-bound
pentalene ligands is the fold angle of the pentalene framework
about the bridgehead C-C bond, which can be inversely
correlated with the size of the bound metal center.43 The fold
angles of 22° and 21° in the two isomers of 1 are slightly less
than that observed in [Th{C8H4(SiiPr3-1,4)2}2] (24°).14 However,
the temptation to attribute this to the relative sizes of Ce3+ (1.07
Å) versus Th4+ (0.99 Å) should be tempered with the fact that
the pentalene ligands in both isomers of 2 are folded by 23°
(Ce4+ radius 0.94 Å). Statistical uncertainties in the fold angles
make such comparisons difficult with central metals of similar
size, but the fold angles in 1 and 2 are clearly smaller than that
in [Ta(C8H4{SiMe3-1,4}2)Cl3] (33°, Ta5+ radius 0.64 Å).
Magnetic Studies. The solid-state magnetic susceptibility of
1 was determined over the temperature range 5-340 K, at a
field of 0.1T. 1 displays Curie behavior over the range 6-150
K with µeff ) 2.19 µB at 300 K, a value approaching that for a
free f1 ion (2.54 µB) and comparable with that found in
[Ce(η-tBuC5H4)3] (2.28 µB at 300 K).42
Figure 4. (a) Detail of the magnetic susceptibility of compound 2
(black circles), the thorium analogue (gray circles), and the sample
holder (open circles), over the temperature range 5-200 K. The
susceptibility of the sample holder has been scaled up by the same
factor used to convert the measured to the molar susceptibility for
2. (b) Residual molar susceptibility of 2 after correction for the
diamagnetism of the sample and the sample holder.
The magnetic susceptibility of 2 and the diamagnetic thorium
analogue was also determined at fields of 0.1 and 1.0 T over
the temperature range 5-340 K. Changing the field over this
range had no significant influence on the susceptibility over
this range, and Figure 4a displays the molar values of
susceptibility in a field of 1 T over the range for both materials.
The response of the Th compound suggests an unexpected
paramagnetic component that most likely arises from the sample
holder. The same figure also displays the response of the sample
holder scaled in the same manner as the sample 2 (i.e., the
susceptibility was multiplied by the factor (molecular weight/
sample mass) for sample 2). Compound 2 has a higher
susceptibility than the Th analogue. The contribution of the Ce
in 2 to the susceptibility was therefore calculated by subtracting
the response of the Th compound, and then the difference in
the scaled response of the sample holder, according to the mass
of the sample holder used. Figure 4b shows the resultant
susceptibility after correction. This led to a residual paramagnetic
component that could be fitted to a Curie-Weiss expression
with negligible Weiss constant and a Curie constant of 0.003-
(3) emu mol-1 (Ce). It is likely that the paramagnetic tail that
is evident in Figure 3b is due to an error in correction for the
sample holder or, alternatively, a trace paramagnetic impurity,12
Figure 5. Normalized XANES spectra.
in which case the Ce component shows TIP with a molar
susceptibility of (4.5 ( 0.3) × 10-4 emu mol-1, significantly
larger than that estimated for cerocene ((1.4 ( 0.2) × 10-4 emu
mol-1).12 This may be accounted for by the smaller HOMO-
LUMO gap in the pentalene complex.
XANES Studies. Figure 5 shows the normalized XANES
spectra of the four compounds studied. It is immediately
apparent from these data that the absorption edge of CeO2 is
shifted to higher energies than any of the organometallic
compounds studied. To quantify this edge shift, we report the
(42) Stultz, S. D.; Andersen, R. A.; Zalkin, A. Organometallics 1990,
9, 115.
(43) Summerscales, O. T.; Cloke, F. G. N. Coord. Chem. ReV. 2006,
250, 1122.

Ce(III) and Ce(IV) Bis(η8-pentalene) Sandwich Complexes
Organometallics, Vol. 26, No. 13, 2007 3117
Table 3. Summary of XANES Data
compound
edge position/ eV
shift from CeB6/eV
Ce(COT′′′)2
2
1
40454.73
40450.49
40450.11
40461.41
4.41
1.91
0
CeO2
11.84
Table 4. IE of 2 and Band Assignments
band
assignment
vertical IE
calculated IE
A
B
C
D
E
4a2
4b1
8b2
8a1
6.32
6.72
7.19
7.53
9.1
6.70
7.14
7.61
7.91
Si-C
Table 5. Relative Energiesa of [Ce(C8H6)2] Optimized with
Different Symmetriesb
∆ kJ/mol (B3LYP)
∆ kJ/mol (BP86)
D2
D2d
D2h
0.000
0.075
4.928
0.000
1.981
15.114
a B3LYP with Gaussian and BP86 with ADF. b The optimized structure
of [Ce(C8H6)2] with the restraints of the D2 point group has effective D2d
symmetry.
edge position in terms of a shift from the edge position of the
CeB6 standard, following the method of Edelstein et al.10
These values are reported in Table 3. The shifts from CeB6
for CeO2 and Ce(COT′′′)2 agree very well with those previously
reported,10 and the values for both 1 and 2 are within the range
expected for Ce(III) compounds.
Photoelectron Spectroscopy. He I and He II photoelectron
(PE) spectra of 2 are shown in Figure 6, and ionizations energies
(IE) of key features are given in Table 4. They resemble closely
the PE spectra of the analogous thorium compound reported
previously and may be assigned in a similar manner.44
The low-energy bands A-D are ill-defined. The presence of
two isomers in the gas phase may well be responsible for the
broadness of the features. They show an intensity increase
relative to band E in the He II spectrum consistent with their
assignment to metal ligand bonding orbitals and band E to Si-C
ionizations. With respect to each other there is no marked
differential intensity change. In general the bands are at lower
IE than the corresponding bands in the Th and U analogues,
suggesting a weaker interaction of the pentalene ligands with
the metal in the Ce compound.44
Figure 6. He I and He II PE spectra of [Ce(C8H6)2].
Computational Studies. Density functional theory (DFT)
calculations were performed on the model compounds
[Ce(C8H6)2]x, where x ) -1, 0, and +1, with the structure
constrained to D2d symmetry. The geometry of [Ce(C8H6)2] was
also optimized in the D2, D2d and D2h point groups. The
optimized geometry within the restraints of the D2 point group
has effective D2d symmetry; thus the energies of the D2 and
D2d geometries are very close. That of the D2h structure is
somewhat higher (Table 5). Thus we may conclude that in the
absence of substituents the barrier to relative rotation of the
rings is very low and that the preferred geometries are controlled
by substituent effects. The calculated geometrical parameters
are given in Table 6.
found experimentally. The Ce-C distances increase slightly in
the anion. The twist angle is primarily influenced by the
substituents, which are neglected in the calculations.
The electronic structure of [Ce(C8H6)2] is similar to that of
the Th analogue. Of the eight π orbitals of the pentalene dianion,
it is the top two occupied ones, π4 and π5, that interact with the
metal orbitals, forming bonds of δ symmetry. For a molecular
symmetry of D2d they give rise to MOs of symmetry a2, b1, b2,
and a1. The orbital energies and principal AO components of
these key orbitals are given in Table 7 for [Ce(C8H6)]2. Iso-
surfaces for LUMO of [Ce(C8H6)]2 and the four highest
occupied orbitals are given in Figure 7.
Given the absence of SiiPr3 groups on the rings of the model
compounds, the agreement between the calculated and experi-
mental structures is excellent. The calculated structural metrics
vary little between the neutral molecule and the anion, as is
The largest single metal contribution is from the Ce 4f:z(x2-
y2) orbital to the 4a2 HOMO, though other metal contributions
are also significant.
Calculated IEs are given in Table 4 and agree well with the
experimental data given the absence of the SiiPr3 substituents
in the model. The spacing of the predicted bands corresponds
(44) Cloke, F. G. N.; Green, J. C.; Jardine, C. N.; Kuchta, M. C.
Organometallics 1999, 18 (6), 1087.

3118 Organometallics, Vol. 26, No. 13, 2007
Table 6. Experimental and Calculated Distances (Å) and Angles (deg) for [Ce(C8H6)2] and [Ce(C8H6)2]-a
Balazs et al.
experimental
calculated
parameter
[Ce(C8H6)2]
[Ce(C8H6)2]-
[Ce(C8H6)2]
[Ce(C8H6)2]-
Ce-C(br)
Ce-C(R)
2.495(4)-2.518(13)
2.730(4)-2.760(11)
2.886(5)-2.925(8)
1.3969(9) - 1.46(2)
1.425(8)-1.473(7)
1.366(7)-1.433(17)
157
2.47(2)-2.592(11)
2.70(3)-2.88(3)
2.921(11)-2.991(11)
1.429(10)-1.435(14)
1.41(2)-1.46(2)
1.410(11)-1.439(15)
158, 159
2.51
2.74
2.88
1.46
1.44
1.42
155.6
92.28
2.59
2.85
2.99
1.46
1.44
1.42
159.4
91.95
Ce-C(â)
C(br)-C(br)
C(br)-C(R)
C(R)-C(â)
ligand fold
twist angle
85, 33
44, 10
a The geometric data are from Gaussian (B3LYP) calculations.
Table 7. Fragment Orbital Contributions to Selected MOs
of [Ce(C8H6)2]a
It is noteworthy that the 1 and 3 positions that undergo the
anomalous chemical shifts are the ones that have significant
contributions to the 4a2 HOMO and 4b1 HOMO-1. It is these
orbitals that are most likely to contribute in second-order
perturbation theory to the paramagnetic shielding.
Geometry optimization of the anion with the BP86 functional
(ADF) leads to a non-aufbau model, as the extra electron is
placed in one of the close lying sets of f orbitals. The fragment
orbital contributions to the MOs are given in Table 8.
In this unrestricted calculation the corresponding R and â
spin orbitals have very similar compositions. On reduction, the
f orbital contribution to the 4a2 and 7b2 orbitals decreases
significantly, whereas the d contribution to the 4b1, 7b2, and
7a1 orbitals varies little. Overall the metal ligand interaction
becomes more ionic. The addition of an f electron to the Ce
reduces the polarization of the ligand orbitals and their mixing
with metal orbitals.
MO
E (eV) % (Pn)2 % Ce
AO-Ce
FO-2×Pn
10e1 (LUMO)
4a2 (HOMO)
-3.68
-4.64
-5.15
-5.67
-5.96
0
66
78
81
79
97
32
21
18
11
f:x
f:z(x2-y2)
d:(x2-y2)
f:z3 + d:xy
d:z2
π5
π5
π4
π4
4b1
7b2
7a1
a The MOs are labeled according to the symmetry labeling of the D2d
point group. Only contributions higher than 5% were considered. The orbital
contributions are taken from a fragment calculation in which the molecule
was broken in two fragments: a Ce and two pentalenes respectively.
The consequent charge distribution in the two species may
be estimated in a variety of ways. In Table 9 we give results
from Mulliken partioning,46-49 the conventional orbital-based
method, and two methods (Hirshfeld50 and Voronoi51) based
on density partitioning.
Mulliken population analysis gives a constant charge of +1
for the Ce(III) and Ce(IV) species. The other two methods show
a lower overall positive charge on Ce and a slight reduction on
addition of the extra electron. The spin density is securely
assigned to the Ce in [Ce(C8H6)2]-. The minor calculated charge
variation between the neutral molecule and the anion is
consistent with the very small shift for the XANES edge
between 1 and 2.
Figure 7. Frontier MOs of [Ce(C8H6)2. The δ character of the
ligand π frontier orbitals, with respect to the metal-ligand axis,
In summary the density functional calculations give a
description of the electronic structure of [Ce(C8H6)2] and its
anion that agrees very well with experimental probes. The partial
covalent character of the four highest occupied MO results in
reduction of the charge on Ce below +4 and below that which
might reasonably be anticipated for more ionic interactions such
as that found in CeO2. The resultant picture is fully consistent
with an oxidation state of IV, which is that arrived at by removal
of the ligands with a full shell, in the same way that the oxidation
state of Mn in the permanganate ion is VII, although the charge
on Mn in no way approaches Mn +7.
closely resembles the e2 orbitals of (C8H8)2- 45
.
very well with the experimental spectra. Both Ce 5d and 4f
orbitals are expected to show increases in their ionization cross
sections with photon energy. The absence of strongly differenti-
ated intensity variations in the PE bands A-D is consistent with
the presence of f or d contributions in all four orbitals.
The small HOMO-LUMO gap of [Ce(C8H6)2] (∼1 eV)
suggests that the charge-transfer band of 2 at 590 nm may be
attributed to excitation from the a2 HOMO into the f manifold.
Table 8. Fragment Orbital Contributions to Selected MOs of the Anion [Ce(C8H6)2]-a
R
â
Irepb
E (eV)
%Ce
%Pn
9.9
Irepb
E (eV)
%Ce
%Pn
8b2 (R HOMO)
1.798
1.547
1.134
-0.349
-0.947
-1.485
-1.768
84.8
unoc
unoc
10.7
20
8a1
10e1
4a2
4b1
7b2
7a1
10e1 (â LUMO)
4a2 (â HOMO)
1.457
-0.306
-0.907
-1.445
-1.720
unoc
8.8
18.3
10.2
10.3
80.6
72.2
77.0
74.2
82.3
72.9
78.0
68.5
4b1
7b2
7a1
11.5
10.7
a The contributions are taken from a ADF calculation. The electronic structure is non-aufbau! Unpaired electron of R spin in orbital 8b2. b Irreducible
representation.

Ce(III) and Ce(IV) Bis(η8-pentalene) Sandwich Complexes
Organometallics, Vol. 26, No. 13, 2007 3119
Table 9. Charges and Spin Densities (ADF; BP86)
[Ce(C8H6)2]
Hirshfeld
[Ce(C8H6)2]-
Mulliken
Voronoi
Mulliken
Hirshfeld
Voronoi
spin dens
Ce
0.98
0.08
0.10
0.00
-0.17
-0.17
0.53
-0.09
-0.07
-0.06
0.06
0.23
-0.10
-0.06
-0.07
0.09
1.00
0.04
0.07
-0.02
-0.21
-0.22
0.43
-0.13
-0.10
-0.08
0.03
0.17
-0.13
-0.08
-0.08
0.05
0.97
0.00
0.00
0.00
0.00
0.00
C (br)
C (R)
C (â)
H (R)
H (â)
0.06
0.08
0.03
0.04
However, there is clearly more than one possible computa-
tional approach. DFT is computationally cheap and has proven
to be successful in predicting structure, interpreting electronic
and PE spectra, and modeling reactivity in relatively large
molecules.52 However it is restricted to systems that can be
described by a single configuration.53 Such single configuration
MO based computational methods are poor at describing
relatively weak interactions, the classic example being two H
atoms at a long distance where they give equal weight to ionic
and neutral forms.54 Valence bond theory is superior in this
respect, as it gives a consistent description of the wave function
as distance is varied.54 The core-like nature of the radial
distribution function for 4f electrons, and the presence of a
centrifugal barrier that confines them within the core,55 leads
to poor overlap with ligand orbitals. This is accentuated in
molecules of high symmetry where hybridization with other
metal orbitals is symmetry restricted. The 4a2 interaction
described above by DFT represents such a case. The alternative
way to represent the electronic structure within an MO context
is to describe the ground state in terms of configuration
interaction between pure ionic configurations, for example, Lf
0d0, L-1f1d0, and L-1f0d1 where L represents the ligand shell
and L-1 a hole in that shell. The configurations must be of the
same symmetry, 1A1, and this requires that the hole in the ligand
set of orbitals is of the same symmetry as the Ce localized
electron. This is often described as the metal electron being
antiferromagnetically coupled to the hole in the ligand shell.7,8,56
In the case of [Ce(C8H6)2] the DFT description indicates that
two configurations would be insufficient, as partial occupation
of both f and d orbitals of a2, b1, b2, and a1 symmetry is required.
However the 4a2 interaction is unique in that it represents one
where symmetry restricts it to an f electron and a ligand set.
The dividing line between description of two electrons being
antiferromagnetically coupled and forming a covalent bond is
difficult to locate; as indicated above, the valence bond
description of a chemical bond moves seamlessly between the
two. It is a matter of degree. A weaker antiferromagnetic
description suggests that a paramagnetic state is likely to be
thermally accessible. Covalent bond formation is accompanied
by a significant HOMO-LUMO gap. The language used is to
a certain extent a matter of training and preference but not a
matter for dispute.
However, in a traditional Ce(III) compound with a localized
f electron, the characteristics of the f1 configuration lead to
paramagnetism with a moment close to that expected for a free
2
ion in a F5/2 state.57 In the case of these formally Ce(IV)
compounds such as CeCOT2 and 2, the ionic configuration
endowing Ce3+ character to the wave function has the f or d
electron spin paired and confined to an orbital of appropriate
symmetry. Thus there exists a real distinction between Ce(III)
and Ce(IV) compounds whatever the real charge on the metal
or the most accurate method of describing the electron distribu-
tion. The primary use of an oxidation state classification is to
inform as to the potential redox and spectroscopic properties
of a metal center. It thus seems appropriate to retain the formal
description of these multiconfigurational ground-state species
as Ce(IV).
Conclusions
The Ce(III) anionic bis(pentalene) sandwich complex
K[Ce{C8H4(SiiPr3-1,4)2}2] (1) has been prepared by treatment
of CeCl3 with K2[C8H4(SiiPr3-1,4)2] and crystallographically
characterized as its 18-crown-6 complex. Oxidation of 1 with
Ag[BPh4] affords the neutral, formally Ce(IV) sandwich com-
plex [Ce{C8H4(SiiPr3-1,4)2}2] (2), whose molecular structure
has also been determined. NMR and magnetic studies indicate
paramagnetism associated with the Ce. XANES measurements
of the K-edge indicate a charge on Ce close to characteristic
values for Ce(III) compounds, the excitation energy for 1 being
only slightly lower in energy than that found for 2. PE spectra
have been measured and assigned. Density functional calcula-
tions give a good account of both the structures and the apparent
charge on Ce in 1 and 2 and the photoelectron spectrum of 2.
They also indicate that in a multiconfiguration treatment of the
ground state of 2 there is likely to be a contribution from a
configuration in which a hole in the ligand shell is strongly
antiferromagnetically coupled to an f electron, both the hole
and the electron having a2 symmetry. A classification of a formal
oxidation state of IV is recommended for 2.
Acknowledgment. We gratefully acknowledge EPSRC for
partial financial support of this project. We thank the Humboldt
Foundation for a Fellowship (G.B.) and the Oxford Supercom-
puting Centre for provision of computing time.
(45) Cloke, F. G. N.; Green, J. C.; Jardine, C. N. Organometallics 1999,
18, 1087.
Note Added in Proof. The permethylated pentalene analog
of 2 has recently been reported: Ashley, A.; Balazs, G.; Cowley,
A.; Green, J. C.; Booth, C. H.; O’Hare, D. Chem. Commun.
2007, 1515.
(46) Mulliken, R. S. J. Chem. Phys. 1955, 23, 1833.
(47) Mulliken, R. S. J. Chem. Phys. 1955, 23, 1841.
(48) Mulliken, R. S. J. Chem. Phys. 1955, 23, 2338.
(49) Mulliken, R. S. J. Chem. Phys. 1955, 23, 2343.
(50) Hirshfeld, F. L. Theor. Chim. Acta 1977, 44 (2), 129.
(51) Fonseca Guerra, C.; Handgraaf, J.-W.; Baerends, E. J.; Bickelhaupt,
F. M. J. Comput. Chem. 2004, 25, 189.
Supporting Information Available: Complete crystallographic
details of the X-ray structures of 1 and 2. This material is available
free of charge via the Internet at http://pubs.acs.org.
(52) Ziegler, T. Chem. ReV. 1991, 91, 7401.
(53) Kohn, W.; Sham, L. J. Phys. ReV. A 1965, 140, 1133.
(54) Coulson, C. A. Valence, 2nd ed.; OUP: Oxford, 1961.
(55) Cowan, R. D. The Theory of Atomic Structure and Spectra;
University of California Press: Berkeley, 1981.
(56) Dolg, M.; Wedig, U.; Stoll, H.; Preuss, H. J. Chem. Phys. 1987,
86, 866.
OM7002738
(57) Figgis, B. N. Introduction to Ligand Fields; Wiley: New York,
1966.