One-electron oxidations of the methyldiphenylphosphonium cyclopentadienylide complexes M(η5-C5H 4PMePh2)(CO)3 (M = Cr, Mo, W): Formation and dimerization of the 17-electron, metal-centered radicals [M(η5- C5H4PMePh2)(CO)3]+

Article abstract of DOI:10.1021/om700688p

Electrochemical oxidations of the group 6 metal complexes M(η⁵-C₅H₄PMePh₂)(CO) ₃ (M = Cr (1), Mo (2), W (3)) occur via one-electron processes at potentials that are highly positive of the values previously observed for the analogous 18-electron anions [M(η⁵-C₅H ₅)(CO)₃]^-. A value of +0.44 V was determined for the electronic effect of the R = [PMePh₂]⁺ group compared to R = H in a C₅H₄R ligand, making it one of the most strongly electron-withdrawing substituents known. On the basis of the experimentally determined oxidation potentials and observations that the oxidized species were long-lived, bulk oxidations of 1, 2, and 3 by [FeCp ₂][B(C₆F₅)₄] were carried out to give the crystalline, dimeric products [M(η⁵-C₅H ₄PMePh₂)(CO)₃]₂[B(C ₆F₅)₄]₂ (M = Cr (1₂ ²⁺), Mo (2₂²⁺), and W (3₂ ²⁺)). These were isolated analytically pure and were characterized by IR (solid state and solution) and NMR (¹H, ¹³C, ¹⁹F, ³¹P) spectroscopy, high-resolution mass spectrometry, and crystallographically. The dimers all contain metal-metal bonds that are comparable in length with or longer than the metal-metal bonds in the isoelectronic, neutral η⁵-C₅H₅ and η⁵-C₅Me₅ analogues, and the metal-metal bond in 1₂²⁺ is the longest nonbridged Cr-Cr bond known. As a result of the apparent weakness of its Cr-Cr bond, 1₂ ²⁺ dissociates significantly in solution to the paramagnetic radical cation 1⁺; the ¹H NMR spectrum of this complex in THF at 298 K exhibits characteristic C₅H₄ resonances at δ26.08 and 13.62.

Full text of DOI:10.1021/om700688p

5890
Organometallics 2007, 26, 5890-5901
One-Electron Oxidations of the Methyldiphenylphosphonium
Cyclopentadienylide Complexes M(η5-C5H4PMePh2)(CO)3 (M ) Cr,
Mo, W): Formation and Dimerization of the 17-Electron,
Metal-Centered Radicals [M(η5-C5H4PMePh2)(CO)3]+
John H. Brownie and Michael C. Baird*
Department of Chemistry, Queen’s UniVersity, Kingston, Ontario K7L 3N6, Canada
Derek R. Laws and William E. Geiger
Department of Chemistry, UniVersity of Vermont, Burlington, Vermont 05405
ReceiVed July 10, 2007
Electrochemical oxidations of the group 6 metal complexes M(η5-C5H4PMePh2)(CO)3 (M ) Cr (1),
Mo (2), W (3)) occur via one-electron processes at potentials that are highly positive of the values
previously observed for the analogous 18-electron anions [M(η5-C5H5)(CO)3]-. A value of +0.44 V was
determined for the electronic effect of the R ) [PMePh2]+ group compared to R ) H in a C5H4R ligand,
making it one of the most strongly electron-withdrawing substituents known. On the basis of the
experimentally determined oxidation potentials and observations that the oxidized species were long-
lived, bulk oxidations of 1, 2, and 3 by [FeCp2][B(C6F5)4] were carried out to give the crystalline, dimeric
products [M(η5-C5H4PMePh2)(CO)3]2[B(C6F5)4]2 (M ) Cr (122+), Mo (222+), and W (322+)). These were
isolated analytically pure and were characterized by IR (solid state and solution) and NMR (1H, 13C, 19F,
31P) spectroscopy, high-resolution mass spectrometry, and crystallographically. The dimers all contain
metal-metal bonds that are comparable in length with or longer than the metal-metal bonds in the
isoelectronic, neutral η5-C5H5 and η5-C5Me5 analogues, and the metal-metal bond in 122+ is the longest
nonbridged Cr-Cr bond known. As a result of the apparent weakness of its Cr-Cr bond, 122+ dissociates
significantly in solution to the paramagnetic radical cation 1+; the 1H NMR spectrum of this complex in
THF at 298 K exhibits characteristic C5H4 resonances at δ 26.08 and 13.62.
D,2c in the crystal structure, which showed significant shortening
of the P-C5H4 bond,2g and in the 13C NMR spectrum, which
showed an unusually high-field chemical shift for the ylide
carbon and a P-C(ylide) coupling constant typical of an
aliphatic carbon-P bond.2h Our more recent crystallographic
and spectroscopic data and DFT calculations on the C5H4-
PMePh2 system strongly confirmed the importance of the
cyclopentadienyl-like structure such as b for free C5H4PMePh2
and, even more so, when C5H4PMePh2 is coordinated in η5-
fashion as in 1, 2, and 3.1
Introduction
We have recently reported the synthesis and reactivity of a
series of group 6 complexes of methyldiphenylphosphonium
cyclopentadienylide (C5H4PMePh2), M(η5-C5H4PMePh2)(CO)3
(M ) Cr (1), Mo (2), W (3)).1 This ligand belongs to an as yet
uncommon class of phosphonium cyclopentadienylides, first
reported in 1956 by Ramirez and Levy,2a who described the
synthesisoftriphenylphosphoniumcyclopentadienylide,C5H4PPh3.2b-f
They found inter alia that this ylide is unusually inert, for
instance being unreactive with ketones, and they attributed its
unusual stability to the charge delocalization implied by
resonance structure b.
The spectroscopic data for complexes 1-3 also demonstrated
that while the cyclopentadienylide ligand is intermediate in
electron donor properties between those of neutral η6-arenes
and the anionic η5-C5H5 ligand, its properties lie rather closer
to that of the latter.1 Therefore, in view of the equilibria that
exist between the metal-metal bonded dimers [Cr(η5-C5H5)-
(CO)3]2 and [Cr(η5-C5Me5)(CO)3]2 and their corresponding 17-
electron, metal-centered persistent radicals,3a,b,f,h-k,o,t-w we
wondered if compounds 1-3 might undergo one-electron
oxidations to give the isoelectronic, persistent cationic metal-
centered radicals [M(η5-C5H4PMePh2)(CO)3]+ (M ) Cr (1+),
Mo (2+), W (3+)) and, if so, to what extent would they dimerize
to the dicationic, metal-metal bonded species [M(η5-C5H4-
Further evidence for extensive delocalization of the π electron
density was found in the relatively high dipole moment of 7.0
(1) Brownie, J. H.; Baird, M. C.; Schmider, H. L. Organometallics 2007,
26, 1433.
(2) (a) Ramirez, F.; Levy, S. J. Org. Chem. 1956, 21, 488. (b) Ramirez,
F.; Levy, S. J. Org. Chem. 1956, 21, 1333. (c) Ramirez, F.; Levy, S. J.
Am. Chem. Soc. 1957, 79, 67. (d) Ramirez, F.; Dershowitz, S. J. Org. Chem.
1957, 22, 41. (e) Ramirez, F.; Levy, S. J. Am. Chem. Soc. 1957, 79, 6167.
(f) Ramirez, F.; Levy, S. J. Org. Chem. 1958, 23, 2035. (g) Ammon, H.
L.; Wheeler, G. L.; Watts, P. H. J. Am. Chem. Soc. 1973, 95, 6158. (h)
Gray, G. A. J. Am. Chem. Soc. 1973, 95, 7736. (i) For a recent computational
study, see: Laavanya, P.; Krishnamoorthy, B. S.; Panchanatheswaran, K.;
Manoharan, M. J. Mol. Struct. (THEOCHEM) 2005, 716, 149.
2+
PMePh2)(CO)3]2 (M ) Cr (122+), Mo (222+), W (322+)).
There seems to exist only a single prior publication dealing
with this type of redox chemistry. Cashman and Lalor reported
in 1971 that Mo(η5-C5H4PPh3)(CO)3 is oxidized by tris(p-
bromophenyl)amminium hexachloroantimonate to give the
10.1021/om700688p CCC: $37.00 © 2007 American Chemical Society
Publication on Web 10/19/2007

M(η5-C5H4PMePh2)(CO)3 (M ) Cr, Mo, W)
Organometallics, Vol. 26, No. 24, 2007 5891
metal-metal bonded, dicationic complex [Mo(η5-C5H4PPh3)-
(CO)3]22+, identified on the basis of elemental analysis and its
IR spectrum.4 However, group 6 metal compounds containing
other ligands provide a rich array of cationic 17-electron
complexes.5 For instance oxidation of M(CO)6 (M ) Cr, Mo,
W) gives the unstable cations [M(CO)6]+, which have lifetimes
on the order of seconds (Cr) or are too unstable to isolate (Mo,
W).6 Substituted derivatives M(CO)6-nLn, M(CO)4L-L, and
M(CO)2(L-L)2 (n ) 1-3; L, L-L ) mono- and bidentate
tertiary phosphines), on the other hand, yield more thermally
stable oxidized complexes, and these have been better character-
ized.7 While oxidation of compounds of the type Cr(η6-arene)-
(CO)3 yields products that are labile8 unless they are generated
in the presence of weakly coordinating anions,8e several
phosphine-substituted complexes [M(η6-arene)(CO)2L]+ (M )
Cr) were found to be stable enough to isolate and characterize.8
However, none of these cationic, metal-centered radicals appear
to form dicationic dimers.
metal-centered radicals M(η5-C5H5)(CO)3 and M(η5-C5Me5)-
(CO)3, which dimerize to give the metal-metal bonded
complexes [M(η5-C5H5)(CO)3]2 and [M(η5-C5Me5)(CO)3]2.3
Although the molybdenum and tungsten dimers do not dissociate
in solution,3 the chromium dimers dissociate as mentioned above
to give the corresponding persistent 17-electron radical mono-
mers.3 We have therefore embarked on a complementary
examination of the redox chemistry of compounds 1-3, finding
that they do indeed undergo oxidation to the corresponding 17-
electron, metal-centered radicals [M(η5-C5H4PMePh2)(CO)3]+
where M ) Cr (1+), Mo (2+), and W (3+), and that these in
turn dimerize to the corresponding metal-bonded complexes
[M(η5-C5H4PMePh2)(CO)3]2 (122+, 222+, and 322+, respec-
2+
tively). As with the above-mentioned compounds [M(η5-C5H5)-
(CO)3]2 and [M(η5-C5Me5)(CO)3]2, we find that the chromium
2+
dimer 12 dissociates extensively in solution to the persistent
radical monomer 1+, but that the heavier metal analogues 22
and 32 dissociate very little, if at all.
2+
2+
Of greater relevance here, however, the anionic group 6
tricarbonyl complexes [M(η5-C5H5)(CO)3]- and [M(η5-C5Me5)-
(CO)3]- (M ) Cr, Mo, W) are readily oxidized to the neutral
Experimental Section
All syntheses were carried out under a dry, deoxygenated argon
atmosphere using standard Schlenk techniques or under nitrogen
in an MBraun Labmaster glovebox. Argon was deoxygenated by
passage through a heated column of BASF copper catalyst and then
dried by passing through a column of 4 Å molecular sieves. NMR
spectra were recorded using Bruker AV 300, AV 500, and AV 600
spectrometers; all 1H and 13C{1H} NMR spectra are referenced to
carbons or residual protons present in the deuterated solvents with
respect to TMS at δ 0, while 31P NMR spectra are referenced to
external 85% H3PO4. Elemental analyses were carried out by
Canadian Microanalytical Service Ltd., Delta, B.C. IR spectra were
acquired on a Perkin-Elmer Spectrum One FT-IR spectrometer at
a spectral resolution of 4 cm-1 (Queen’s) or on a ATI-Mattson
Infinity Series FT-IR spectrometer using Winfirst software and
operating at a resolution of 4 cm-1 (Vermont). A Remspec ZnSe
fiber-optic cable terminating in a gold reflective disk was used for
in situ attenuated total reflectance measurements;9 a spectral window
from 1500 to 2200 cm-1 was available under the experimental
conditions. Electrospray mass spectrometry (ES-MS) data were
collected in the Mass Spectrometry Facility at Queen’s on an
Applied Biosystems/MDS Sciex QSTAR XL QqTOF mass spec-
trometer.
The CH2Cl2, THF, and hexanes used for syntheses were
purchased from Aldrich in 18 L reservoirs packaged under nitrogen
and were dried by passage through columns of activated alumina
(Innovative Technology Solvent Purification System). THF, Et2O,
and CH2Cl2 were then stored over 4 Å molecular sieves to result
in residual water concentrations that were lower than 20 ppm.
Deuterated NMR solvents (Cambridge Isotope Laboratories, Inc.
or CDN Isotopes) were degassed under vacuum and dried by
passage through a small column of activated alumina before being
stored over 4 Å molecular sieves. Most chemicals were obtained
from Aldrich or Strem and were used as received or were purified
by established procedures. The compounds M(η5-C5H4PMePh2)-
(CO)3 (M ) Cr, Mo, W)1 and [FeCp2][B(C6F5)4]10 were synthesized
according to the literature.
(3) (a) Adams, R. D.; Collins, D. E.; Cotton, F. A. J. Am. Chem. Soc.
1974, 96, 749. (b) Landrum, J. T.; Hoff, C. D. J. Organomet. Chem. 1985,
282, 215. (c) Hackett, P.; O’Neill, P. S.; Manning, A. R. J. Chem. Soc.,
Dalton Trans. 1974, 1625. (d) Madach, T.; Vahrenkamp, H. Z. Naturforsch.
1979, 34b, 573. (e) Goh, L.-Y.; D’Aniello, M. J.; Slater, S.; Muetterties, E.
L.; Tavanaiepour, I.; Chang, M.I.; Fredrich, M. F.; Day, V. W. Inorg. Chem.
1979, 18, 192. (f) Cooley, N. A.; Watson, K. A.; Fortier, S.; Baird, M. C.
Organometallics 1986, 5, 2563. (g) Keller, H. J. Z. Naturforsch. 1968, 23b,
133. (h) Madach, T.; Vahrenkamp, H. Z. Naturforsch. 1978, 33b, 1301. (i)
McLain, S. J. J. Am. Chem. Soc. 1988, 110, 643. (j) Cooley, N. A.;
MacConnachie, P. T. F.; Baird, M. C. Polyhedron 1988, 7, 1965. (k) Jaeger,
T. J.; Baird, M. C. Organometallics 1988, 7, 2074. (l) Wrighton, M. S.;
Ginley, D. S. J. Am. Chem. Soc. 1975, 97, 4246. (m) Hughey, J. L.; Bock,
C. R.; Meyer, T. J. J. Am. Chem. Soc. 1975, 97, 4440. (n) Laine, R. M.;
Ford, P. C. Inorg. Chem. 1977, 16, 388. (o) Watkins, W. C.; Jaeger, T.;
Kidd, C. E.; Fortier, S.; Baird, M. C.; Kiss, G.; Roper, G. C.; Hoff, C. D.
J. Am. Chem. Soc. 1992, 114, 907. (p) Goh, L. Y.; Habley, T. W.;
Darensbourg, D. J.; Reibenspies, J. J. Organomet. Chem. 1990, 381, 349.
(q) Guanyang, L.; Wing-Tak, W. J. Organomet. Chem. 1996, 522, 271. (r)
Clegg, W.; Compton, N. A.; Errington, R. J.; Norman, N. C. Acta
Crystallogr. 1988, C44, 568. (s) Rheingold, A. L.; Harper, J. R. Acta
Crystallogr. 1991, C47, 184. (t) Richards, T. C.; Geiger, W. E.; Baird, M.
C. Organometallics 1994, 13, 4494. (u) Woska, D. C.; Ni, Y.; Wayland,
B. B. Inorg. Chem. 1999, 38, 4135. (v) O’Callaghan, K. A. E.; Brown, S.
J.; Page, J. A.; Baird, M. C.; Richards, T. C.; Geiger, W. E. Organometallics
1991, 10, 3119. (w) Yao, Q.; Bakac, A.; Espenson, J. H. Organometallics
1993, 12, 2010. (x) Kadish, K. M.; Lacombe, D. A.; Anderson, J. E. Inorg.
Chem. 1986, 25, 2246. The potentials in this paper have been converted to
the ferrocene reference potential by addition of 0.46 V. (y) Morton, J. R.;
Preston, K. F.; Cooley, N. A.; Baird, M. C.; Krusic, P. J.; McLain, S. J. J.
Chem. Soc., Faraday Trans. 1 1987, 83, 3535.
(4) Cashman, D.; Lalor, F. J. J. Organomet. Chem. 1971, 32, 351.
(5) For reviews see: (a) Baird, M. C. Organomet. Radical Processes
1990, 49. (b) Baird, M. C. Chem. ReV. 1988, 88, 1217.
(6) (a) Pickett, C. J.; Pletcher, D. J. Chem. Soc., Dalton Trans. 1975,
879. (b) Chum, H. L.; Koran, D.; Osteryoung, R. A. J. Organomet. Chem.
1977, 140, 349. (c) Bagchi, R. N.; Bond, A. M.; Colton, R.; Luscombe, D.
L.; Moir, J. E. J. Am. Chem. Soc. 1986, 108, 3352.
(7) (a) Bond, A. M.; Colton, R.; Kevekordes, J. E.; Panagiotidou, P. Inorg.
Chem. 1987, 26, 1430. (b) Bond, A. M.; Colton, R.; Kevekordes, J. E. Inorg.
Chem. 1987, 25, 749. (c) Bagchi, R. N.; Bond, A. M.; Brian, G.; Colton,
R.; Henderson, T. L. E.; Kevekordes, J. E Organometallics 1984, 3, 4. (d)
Bond, A. M.; Carr, S. W.; Colton, R. Inorg. Chem. 1984, 23, 2343. (e)
Bond, A. M.; Carr, S. W.; Colton, R. Organometallics 1984, 3, 541. (f)
Lappert, M. F.; McCabe, R. W.; MacQuitty, J. J.; Pye, P. L.; Riley, P. I. J.
Chem. Soc., Dalton Trans. 1980, 90. (g) Ashford, P. K.; Baker, P. K.;
Connelly, N. G.; Kelly, R. L.; Woodley, V. A. J. Chem. Soc., Dalton Trans.
1982, 477.
(8) (a) Connelly, N. G.; Demidowicz, Z.; Kelly, R. L. J. Chem Soc.,
Dalton Trans. 1975, 2335. (b) Van Order, N.; Geiger, W. E.; Bitterwolf,
T. E.; Rheingold, A. L. J. Am. Chem. Soc. 1987, 109, 5680. (c) Doxsee, K.
M.; Grubbs, R. H.; Anson, F. C. J. Am. Chem. Soc. 1984, 106, 7819. (d)
Zoski, C. G.; Sweigart, D. A.; Stone, N. J.; Rieger, P. H.; Mocellin, E.;
Mann, T. F.; Mann, D. R.; Gosser, D. K.; Doeff, M. M.; Bond, A. M. J.
Am. Chem. Soc. 1988, 110, 2109. (e) Camire, N.; Nafady, A.; Geiger, W.
E. J. Am. Chem. Soc. 2002, 124, 7260.
All electrochemical experiments were conducted (Vermont) using
Princeton Applied Research models PAR-173 and PAR-273 po-
tentiostats interfaced to personal computers. For these experiments,
CH2Cl2 was dried over CaH2, distilled under nitrogen, and purified
further by vacuum transfer from fresh drying agent prior to use,
while [NBu4][B(C6F5)4] was prepared by metathesis of [NBu4]Br
(9) Shaw, M. J.; Geiger, W. E. Organometallics 1996, 15, 13.
(10) Nafady, A.; Chin, T. T.; Geiger, W. E. Organometallics 2006, 25,
1654, and references therein.

5892 Organometallics, Vol. 26, No. 24, 2007
Brownie et al.
Table 1. Selected Bond Lengths and Angles of
[12][B(C6F5)4]2, [22][B(C6F5)4]2, and [32][B(C6F5)4]2
[12][B(C6F5)4]2 [22][B(C6F5)4]2 [32][B(C6F5)4]2
Bond Lengths (Å)
with K[B(C6F5)4] (Boulder Scientific Co.) in methanol,11 recrystal-
lized three times from CH2Cl2/ether, and dried at 353 K under
vacuum for 20 h.
Spectroelectrochemical IR experiments were performed using
Schlenk conditions under an atmosphere of argon after adding
solvent and electrolyte to the electrochemical cell in a Vacuum
Atmospheres drybox under nitrogen. All other experiments were
conducted directly in the drybox under nitrogen. Working electrodes
were glassy carbon electrodes (GCEs, 1-3 mm diameter) supplied
by Bioanalytical Systems. The effective areas of the electrodes were
determined through chronoamperometric measurements of ferrocene
in acetonitrile/0.1 M [NBu4][PF6]. The electrodes were pretreated
by polishing with successively finer diamond paste (Buehler, 3 to
0.25 µm) interspersed by washings with Nanopure water and,
finally, dried in the antechamber of the drybox. Controlled potential
coulometry was carried out using a basket-shaped Pt mesh electrode.
Although the functional reference electrode was Ag/AgCl, separated
from the working compartment by a fine glass frit, all potentials
P(1)-C(1)
1.784(2)
1.792(2)
1.788
1.777(8)
1.782(9)
1.776
1.777(5)
1.789(5)
1.794
P-Me
P-Ph av
M-C5 centroid
C(1)-C(2)
C(1)-C(5)
C(2)-C(3)
C(4)-C(5)
C(3)-C(4)
C(1)-metal
C(2)-metal
C(3)-metal
C(4)-metal
C(5)-metal
1.847
2.010
2.001
1.428(3)
1.425(3)
1.410(3)
1.411(3)
1.406(4)
2.164(2)
2.197(2)
2.253(2)
2.230(2)
2.178(2)
1.447(12)
1.440(12)
1.399(13)
1.435(12)
1.390(15)
2.305(8)
2.319(9)
2.355(9)
2.411(8)
2.337(8)
1.451(7)
1.445(7)
1.419(7)
1.397(7)
1.418(8)
2.298(5)
2.342(5)
2.388(5)
2.367(5)
2.304(5)
Bond Angles (deg)
C(1)-P-Me
C(1)-P-Ph av
Me-P-Ph av
Ph-P-Ph
P-C(1)-C5
centroid
108.92(11)
108.9
110.1
109.66(11)
173.2
107.7(5)
109.8
109.2
110.8(4)
172.0
108.3(2)
109.5
110
109.6(2)
172.9
0/+
reported in this paper are versus the FeCp2 redox couple, as
recommended elsewhere.12 The ferrocene potential was obtained
by adding it to the solution at an appropriate point in the experiment.
The supporting electrolyte in all experiments was [NBu4]-
[B(C6F5)4],12,13 at concentrations of 0.05 or 0.1 M, as noted in the
text. Diagnostics applied to the shapes and positions of cyclic
voltammetric waves were as described earlier.14 Digital simulations
of background-subtracted cyclic voltammetry experiments were
performed using Digisim 3.0 (Bioanalytical Systems). More details
on the electrochemical methodologies can be found in a recent
paper.15
C(1)-C(2)-C(3) 107.7(2)
C(2)-C(3)-C(4) 108.8(2)
C(3)-C(4)-C(5) 108.0(2)
C(4)-C(5)-C(1) 108.3(2)
C(5)-C(1)-C(2) 107.2(2)
106.6(9)
112.1(9)
106.1(8)
108.7(9)
106.6(7)
107.2(4)
109.3(5)
108.0(5)
109.1(5)
106.4(4)
In the case of [12][B(C6F5)4]2, the hydrogen atoms on one of the
dichloromethane molecules were located from the difference
Fourier map, but all of the other hydrogen atoms were calculated
and their contributions were included in the structure factor
calculations. For [22], the positions of all non-hydrogen atoms were
refined anisotropically. The positions for all hydrogen atoms were
calculated and their contributions were included in the structure
factor calculations. However, the structure can only be refined to
R1 ≈ 13% with a THF molecule disordered. ROTAX16c was thus
used to find the twin law. The twin law [1 0 0 0 -1 0 -0.252 0
-1] of merit (fom) 2.48 for 180° rotation was applied to the
refinement and the R1 dropped to ∼10%. The crystal was thus
considered to be merohedrally twinned. SQEEZE16d was then
applied to squeeze out the 1.5 THF molecule, which led to R1 ≈
8.8%. For [32][B(C6F5)4]2, the disordered THF solvent molecule
was refined with DFIX, EADP, and PART. Selected bond lengths
and angles for all three dimers are provided in Table 1, general
crystallographic data in Table 2, and comparisons of metal-metal
bond distances with those of the η5-C5H5 and η5-C5Me5 analogues
in Table 3.
Synthesis of [Cr(η5-C5H4PMePh2)(CO)3]2[B(C6F5)4]2, [12]-
[B(C6F5)4]2. A solution of 0.110 g of 1 (1.3 × 10-4 mol) and 0.051
g of [FeCp2][B(C6F5)4] (1.3 × 10-4 mol) in 5 mL of CH2Cl2 was
stirred for 30 min, during which time it changed from deep
green to yellow-brown. The solution was then layered with
30 mL of hexanes and cooled to -30 °C to give green crystals,
which were filtered and washed with 3 × 20 mL of hexanes.
Yield: 0.104 g (76%). Anal. Found: C 46.90, H 1.57. Calc for
C90H34B2F40P2O6Cr2‚3CH2Cl2: 47.45, H 1.64. IR data are given
in Table 4. 1H NMR (CD2Cl2, 298 K, 600 MHz): δ 28.80 (2H, br
s, C5H4), 13.65 (2H, br s, C5H4), 7.85-7.60 (10H, br, Ph), 1.18
(3H, br, Me). 1H NMR (THF-d8, 298 K, 600 MHz): δ 26.08 (2H,
br, C5H4), 13.62 (2H, br, C5H4), 8.07-7.66 (10H, Ph), 2.34 (3 H,
br s, Me). 1H NMR (THF-d8, 213 K, 600 MHz): δ 7.95-7.75 (10H,
m, Ph), 6.12 (1H, br, C5H4), 5.96 (1H, br, C5H4), 5.85 (1H, br,
C5H4), 5.65 (1H, br, C5H4), 2.89 (3H, br, Me). HR ES-MS (THF):
found 400.0320 m/z; calc 400.0348 (diff 6.8926 ppm). MS-MS
experiments showed that the peak at 400 m/z (target ion) was
responsible for the fragment at 316 m/z (loss of 84 amu ) loss of
3CO).
X-ray crystal structure determinations were performed in the
X-ray Crystallography Laboratory at Queen’s University. Crystals
were mounted on glass fibers with epoxy glue, and data collections
were performed on a Bruker Smart CCD 1000 X-ray diffractometer
with graphite-monochromated Mo KR radiation (λ ) 0.71073 Å)
controlled with a Crysostream Controller 700. No significant decay
was observed during data collections. Data were processed on a
PC using the Bruker AXS Crystal Structure Analysis Package. Data
collection: APEX2 (Bruker, 2006); cell refinement: SAINT
(Bruker, 2005); data reduction: SAINT (Bruker, 2005); structure
solution: XPREP (Bruker, 2005) and SHELXTL (Bruker, 2000);
structure refinement: SHELXTL; molecular graphics: SHELXTL;
publication materials: SHELXTL.16a Neutral atom scattering factors
were taken from Cromer and Waber.16b The raw intensity data were
converted (including corrections for scan speed, background and
Lorentz and polarization effects) to structure amplitudes and their
esd’s using the program SAINT, which corrects for Lp and decay.
Absorption corrections were applied using the program SADABS.
All non-hydrogen atoms were refined anisotropically. The positions
for all hydrogen atoms were calculated (unless otherwise stated),
and their contributions were included in the structure factors and
calculations.
(11) LeSuer, R. J.; Buttolph, C.; Geiger, W. E. Anal. Chem. 2004, 76,
6395.
(12) (a) Gritzner, G.; Kuta, J. Pure Appl. Chem. 1984, 56, 461. (b)
Connelly, N. G.; Geiger, W. E. Chem. ReV. 1996, 96, 877.
(13) LeSuer, R. J.; Geiger, W. E. Angew. Chem., Int. Ed. 2000, 39, 248.
(14) Geiger, W. E. In Laboratory Methods in Electroanalytical Chemistry,
2nd ed.; Kissinger, P. T., Heineman, W. E., Eds.; Marcel Dekker, Inc.;
New York, 1996; Chapter 23.
(15) Nafady, A.; Costa, P. J.; Calhorda, M. J.; Geiger, W. E. J. Am.
Chem. Soc. 2006, 128, 16587.
(16) (a) Bruker AXS, Crystal Structure Analysis Package; Bruker, 2000.
SHELXTL, Version 6.14; Bruker AXS Inc.: Madison, WI, 2005. XPREP,
Version 2005/2; Bruker AXS Inc.: Madison, WI, 2005. SAINT, Version
7.23A; Bruker AXS Inc.: Madison, WI, 2006. APEX2, Version 2.0-2;
Bruker AXS Inc.: Madison, WI. (b) Cromer, D. T.; Waber, J. T.
International Tables for X-ray Crystallography; Kynoch Press: Birmingham,
UK, 1974; Vol. 4, Table 2.2 A. (c) Cooper, R. I., Gould, R. O., Parsons,
S.; Watkin, D. J. J. Appl. Crystallogr. 2002, 35, 168. (d) Sluis, P. v. d.;
Spek, A. L. Acta Crystallogr. 1990, A46, 194.

M(η5-C5H4PMePh2)(CO)3 (M ) Cr, Mo, W)
Organometallics, Vol. 26, No. 24, 2007 5893
Table 2. Crystallographic Data
[12][B(C6F5)4]2 [22][B(C6F5)4]2
C93H40B2Cl6Cr2F40O6P2, C102H58B2F40Mo2O9P2,
[32][B(C6F5)4]2
empirical formula, fw
C
102H58B2F40O9P2W2,
2413.51
180(2)
2462.92
298(2)
2638.74
180(2)
temperature (K)
wavelength (Å)
cryst syst
space group
a (Å)
0.71073
triclinic
P1h
0.71073
monoclinic
P2/n
13.0003(18)
10.0517(14)
38.132(5)
90
92.466(2)
90
4978.3(12)
2
0.71073
monoclinic
P2/n
12.9575(7)
9.9157(6)
37.847(2)
90
9.8319(4)
12.8275(6)
19.6330(9)
70.9350(10)
83.1200(10)
88.6700(10)
2323.09(18)
1
b (Å)
c (Å)
R (deg)
â (deg)
91.9600(10)
90
γ (deg)
V (Å3)
4859.8(5) Å3
2
Z
F (calcd) (Mg/m3)
1.725
0.577
1196
1.643
0.418
2452
1.803
2.535
2580
absorp coeff (mm-1
F(000)
)
cryst size (mm3)
θ range for data
collection (deg)
index ranges
0.15 × 0.15 × 0.10
0.25 × 0.20 × 0.06
0.15 × 0.12 × 0.10
2.09 to 25.00
2.03 to 25.19
1.68 to 25.00
-11ehe11, -15eke15,
-23ele23
16 270
8129 [R(int) ) 0.0191]
99.30%
-15ehe15, -12eke12,
-45ele45
44 366
8919 [R(int) ) 0.0840]
99.50%
-15ehe15, -11eke11,
-44ele44
44 056
8569 [R(int) ) 0.0418]
99.90%
no. of reflns collected
no. of indep reflns
completeness to θ ) 25.00°
absorp corr
multiscan
numerical
multiscan
max. and min. transmn
refinement method
no. of data/restraints/
params
0.9445 and 0.9184
full-matrix least-squares on F2
8129/2/738
0.9754 and 0.9028
full-matrix least-squares on F2
8919/0/641
0.7856 and 0.7023
full-matrix least-squares on F2
8569/4/729
goodness-of-fit on F2
final R indices [I > 2σ(I)]
R indices (all data)
largest diff peak and
1.00
1.054
1.00
R1 ) 0.0351, wR2 ) 0.0940
R1 ) 0.0467, wR2 ) 0.1023
0.352 and -0.449
R1 ) 0.0881, wR2 ) 0.2110
R1 ) 0.1209, wR2 ) 0.2248
1.192 and -0.928
R1 ) 0.0376, wR2 ) 0.1283
R1 ) 0.0453, wR2 ) 0.1369
0.852 and -1.104
hole (e Å-3
)
Table 3. Comparison of Metal-Metal Bond Distances of [M(η5-C5H5)(CO)3]2, [M(η5-C5Me5)(CO)3]2, and
[M(η5-C5H4PMePh2)(CO)3]2[B(C6F5)4]2 (M ) Cr, Mo, W)
2+
metal
[M(η5-C5H5)(CO)3]2
[M(η5-C5Me5)(CO)3]2
[M(η5-C5H4PMePh2)(CO)3]2
Cr
Mo
W
3.281(1)3a
3.235(1)3a
3.222(1)3a
3.311(1),3o 3.310(1)3p
3.284(1),3q 3.281(1)3r
3.288(1)3s
3.3509(7)
3.2764(16)
3.2507(4)
Table 4. Solution (THF) and Solid-State (Fluorolube mulls) IR Data (ν(CO)) of [M(η5-C5H4PMePh2)(CO)3]2[B(C6F5)4]2 (M )
Cr, Mo, W)
metal
THF (cm-1
)
CH2Cl2 (cm-1
)
Fluorolube (cm-1
)
Cr
Mo
W
2037 (s), 1955 (s br), 1914 (s br)
2035, 1955, 1910
2029, 1981, 1943
2036 (vw), 1970, 1949, 1932, 1897 (w sh)
2037 (vw), 1975, 1939, 1932
1972, 1925
2025 (w), 1976 (s br), 1939 (s br), ∼1922 (sh)
2023 (w), 1973 (s br), 1933 (s br), ∼1922 (sh)
Synthesis of [Mo(η5-C5H4PMePh2)(CO)3]2[B(C6F5)4]2, [22]-
[B(C6F5)4]2. In a procedure similar to that described above, a
solution of 0.096 g of 2 (2.2 × 10-4 mol) and 0.186 g of [FeCp2]-
[B(C6F5)4] (2.2 × 10-4 mol) in 10 mL of THF was stirred for 30
min, turning from reddish-green to red. The resulting solution was
layered with 30 mL of hexanes and cooled to -30 °C to yield a
red oil, which slowly solidified. The oil was washed with 3 × 15
mL of hexanes and dried in Vacuo to give a red solid in a yield of
0.207 g (86%). X-ray quality crystals were grown from a solution
of the red material in a 1:4 mixture of CH2Cl2/THF layered with
hexanes and kept at -30 °C. Anal, Found: C 47.30, H 1.67. Calc
for C90H34B2F40P2O6Mo2: 48.12, H 1.53. IR data are given in Table
86.1 (d, JP-C 94.4 Hz, PCCHCH), 8.4 (d, JP-C 60.4 Hz, Me). HR
ES-MS (THF): found 445.9969 m/z; calc 445.9988 (diff 4.1748
ppm).
Synthesis of [W(η5-C5H4PMePh2)(CO)3]2[B(C6F5)4]2, [32]-
[B(C6F5)4]2. A solution of 0.088 g of 3 (1.7 × 10-4 mol) and 0.143
g of [FeCp2][B(C6F5)4] (1.7 × 10-4 mol) in 10 mL of THF was
stirred for 30 min. The resulting solution was layered with 30 mL
of hexanes and cooled to -30 °C to yield a red oil, which slowly
solidified. The oil was washed with 3 × 15 mL of hexanes and
dried in Vacuo to give a red solid in a yield of 0.155 g (77%).
X-ray quality crystals were grown from a dilute THF solution (∼20
mg in 1.5 mL of THF) layered with hexanes and kept at room
1
4. H NMR (THF-d8, 298 K, 600 MHz): δ 7.87-7.72 (10 H, m,
temperature. Anal. Found:
C 44.67, H 1.94. Calc for
Ph), 6.10-6.04 (2H, br m, C5H4) and 5.91 (2H, br m, C5H4), 2.88
(3H, d, Me, 2JP-H 11.0 Hz). 31P NMR (THF-d8, 121 MHz): δ 19.88.
13C NMR (THF-d8, 150 MHz): δ 231.5 (CO), 224.8 (CO), 149.4
(m, o-C6F5), 139.4 (m, p-C6F5), 138.3 (m, m-C6F5), 136.7 (br s,
p-Ph), 133.7 (d, JP-C 11.0 Hz, Ph), 131.5 (d, JP-C 13.2 Hz, Ph),
125.6 (br m, ipso-C6F5), 121.7 (d, JP-C 90.0 Hz, ipso-Ph), 103.0
(d, JP-C 8.8 Hz, PCCHCH), 97.2 (d, JP-C 11.0 Hz, PCCHCH),
C90H34B2F40P2O6W2: 44.62, H 1.41. IR data are given in Table 4.
1H NMR (THF-d8, 298 K, 600 MHz): δ 7.80-7.71 (10H m, Ph),
2
6.05 (4H, br m, C5H4), 2.86 (3H, br d, JP-H 12.8 Hz, Me). 13C
NMR (THF-d8, 150 MHz): δ 218.5 (s, CO), 211.7 (s, CO), 149.4
(m, o-C6F5), 139.4 (m, p-C6F5), 137.3 (m, m-C6F5), 136.8 (d, JP-C
2.8 Hz, p-Ph), 133.8 (d, JP-C 12.5 Hz, Ph), 131.5 (d, JP-C 13.9 Hz,
Ph), 125.6 (br m, ipso-C6F5), 121.4 (d, JP-C 90.2 Hz, ipso-Ph), 102.8

5894 Organometallics, Vol. 26, No. 24, 2007
Brownie et al.
process as given in eq 1. This conclusion was confirmed by
bulk coulometry (see below).
Cr(η5-C5H4PMePh2)(CO)3 (1) - e- h
[Cr(η5-C5H4PMePh2)(CO)3]+ (1+) (1)
A second, chemically irreversible anodic wave of ap-
proximately one-half of the height of the first wave was observed
at Epa ) ca. 0.95 V (Figure SM1, Supporting Information), but
the products of this second oxidation were not studied. The free
cyclopentadienylide ligand itself oxidizes (irreversibly) at Epa
) 0.1 V. Owing to the fact that, as will be shown, the
molybdenum and tungsten analogues 2 and 3 undergo fast
dimerization reactions when oxidized, the oxidation of 1 was
also investigated under conditions, specifically at lower tem-
peratures and higher concentrations, that would be expected to
favor dimerization of 1+. CV scans of a 3.6 mM solution of 1
in CH2Cl2 at temperatures from ambient to 238 K failed to show
any changes that would indicate involvement of dimerization
of 1+ in detectable amounts in solution. Simulations of CV scans
of 1 (Figure 1) provide values of the charge-transfer coefficient,
â ) 0.57 (where â ) 1 - R), and the standard heterogeneous
electron-transfer rate constant, ks ) 0.04 cm s-1, for the
reversible one-electron oxidation. The diffusion coefficient of
1 was measured as Do ) 5 × 10-6 cm2 s-1
.
Bulk electrolysis of 1 at Eappl ) 0.3 V resulted in clean
conversion to the radical cation 1+ as the color of the solution
remained yellow. The coulometry was consistent with a one-
electron process, with values of 0.90 and 0.95 F/equiv being
obtained in two separate experiments. The starting material 1
was efficiently regenerated to greater than 90% of its original
concentration by back-electrolysis at Eappl ) -0.6 V. When
monitored by in situ IR spectroscopy (Figure 2), the anodic
electrolysis revealed the presence of three carbonyl bands for
Figure 1. Experimental (circles) and simulated CVs for 0.6 mM
1 in CH2Cl2/0.1 M [NBu4][B(C6F5)4]. Top, V ) 2.0 V s-1; bottom,
V ) 0.3 V s-1. Relevant simulation parameters: ks ) 0.04 cm s-1
â ) 0.57.
,
the 17-electron species 1+, at 2035, 1955, and 1910 cm-1
,
comparable with the values of νCO of [122+][B(C6F5)4]2 prepared
from the oxidation of 1 by [FeCp2][B(C6F5)4] (see below) and
at considerably higher frequencies than νCO of the neutral 1
(1912 and 1809 cm-1).1 Assuming that 1809 cm-1 may be taken
as the average of the asymmetric νCO of 1, an average shift of
+124 cm-1 in νCO is calculated for the one-electron oxidation,
in concert with expectations.17 Taken together, the voltammetry,
electrolysis, and IR spectroscopy indicate that the first oxidation
of 1 is a quasi-Nernstian, chemically reversible, one-electron
process involving a largely metal-based orbital; note that we
have previously shown that the HOMO of 1 is indeed a metal-
based orbital.1 There is no evidence for appreciable amounts
of the chromium dimer dication, 122+, under these conditions
of temperature and concentration.
(d, JP-C 8.3 Hz, PCCHCH), 94.4 (d, JP-C 11.1 Hz, PCCHCH), 84.3
(d, JP-C 97.1 Hz, PCCHCH), 8.4 (d, JP-C 59.6 Hz, Me). 31P NMR
(THF-d8, 121 MHz): δ 20.76. HR ES-MS(THF): found 532.0424
m/z; calc 532.0406 (diff 3.5448 ppm).
Oxidation of 1 with [Ph3C][B(C6F5)4]. A solution of 0.400 g
of 1 (9.99 × 10-4 mol) and 0.968 g of [Ph3C][B(C6F5)4] (1.05 ×
10-3 mol) in 15 mL of THF was stirred for 30 min, the green
solution slowly turning yellow. The solution was then layered with
∼100 mL of hexanes and cooled to -30 °C, and a green solid
precipitated and was separated from the supernatant. The solid was
washed with 4 × 10 mL of hexanes before being dried in Vacuo to
yield 0.917 g (85%) of [12][B(C6F5)4]2. This material had an IR
spectrum identical to that of the material prepared using [FeCp2]-
[B(C6F5)4] as oxidant.
Oxidation of 1 with [FeCp2][PF6]. A solution of 0.130 g of 1
(3.25 × 10-4 mol) and 0.108 g of [FeCp2][PF6] (3.25 × 10-4 mol)
in 10 mL of THF was stirred for 1.5 h, turning to a deep brownish-
yellow-green color. As the reaction proceeded, it was monitored
by IR spectroscopy, which showed the growth of peaks at 2030,
1941, and 1906 cm-1, identical to those of the [12][B(C6F5)4]2
obtained using [FeCp2][B(C6F5)4], and a new set of peaks (already
Cyclic voltammograms of 2 in CH2Cl2/0.05 M [NBu4]-
[B(C6F5)4] revealed two chemically irreversible features when
the potential was scanned from -1.1 to +0.1 V (Figure 3). The
anodic wave at Epa ≈ -0.19 V is due to the apparent
one-electron oxidation of 2 to 2+, and the cathodic feature at
more negative potentials arises from a follow-up reaction of
2+. Whereas the anodic wave has the shape of an electrochemi-
present after 5 min) at 1917, 1821, and 1808 cm-1
.
(17) Shifts of >100 cm-1 are widely observed for νCO when the positive
charge on a metal carbonyl complex is increased by one (Nakamoto, K.
Infrared and Raman Spectra of Inorganic and Coordination Compounds,
4th ed.; Wiley, New York, 1986; pp 292-293). For recent experimental
and theoretical work showing that the origin of this effect is predominantly
electrostatic, see: (a) Goldman, A. S.; Krogh-Jespersen, K. J. Am. Chem.
Soc. 1996, 118, 12159. (b) Willner, H.; Aubke, F. Angew. Chem., Int. Ed.
1997, 36, 2402. (c) Ehlers, A. W., Ruiz-Morales, Y.; Baerends, E. J.; Ziegler,
T. Inorg. Chem. 1997, 36, 5031.
Results and Discussion
Electrochemical Oxidation of 1, 2, and 3. Cyclic voltam-
mograms (CVs) of 1 in CH2Cl2 with 0.05 M [NBu4][B(C6F5)4]
exhibited an oxidation wave at E1/2 ) -0.38 V (Figure 1) having
the general characteristics of a chemically reversible one-electron

M(η5-C5H4PMePh2)(CO)3 (M ) Cr, Mo, W)
Organometallics, Vol. 26, No. 24, 2007 5895
Figure 2. Infrared data for spectroelectrochemistry of 3.0 mM 1 in CH2Cl2/0.05 M [NBu4][B(C6F5)4] at 298 K. Arrows indicate the
increase or decrease in intensity of infrared absorptions as oxidation proceeds.
Although there are a number of mechanisms that could
account for these observations, the voltammetric behavior is
most strongly reminiscent of an anodic EC mechanism, involv-
ing fast dimerization of 2+ (eqs 2 and 3), and a cathodic two-
electron EEC process resulting in regeneration of the original
18-electron complex 2 (eqs 4 and 5).
Mo(η5-C5H4PMePh2)(CO)3 - e- h
[Mo(η5-C5H4PMePh2)(CO)3]+ (E) (2)
2[Mo(η5-C5H4PMePh2)(CO)3]+ h
[Mo(η5-C5H4PMePh2)(CO)3]22+ (C) (3)
[Mo(η5-C5H4PMePh2)(CO)3]22+ + 2e- f
2[Mo(η5-C5H4PMePh2)(CO)3]2 (EE) (4)
[Mo(η5-C5H4PMePh2)(CO)3]2 f
2Mo(η5-C5H4PMePh2)(CO)3 (C) (5)
In the cathodic direction the “C” step (eq 5) involves
homolytic cleavage of the Mo-Mo bond of the dimer. Since
the data are insufficient to differentiate between the proposed
EEC mechanism and an ECE cathodic mechanism in which
fragmentation of the metal-metal bond occurs after the ac-
ceptance of the first electron, simulations were performed using
only the EEC process for the cathodic reaction. There is
significant literature precedent for these types of overall
Figure 3. Experimental (circles) and simulated CVs for 0.7 mM
mechanisms in describing the redox-based formation and
2 in CH2Cl2/0.1 M [NBu4][B(C6F5)4]. Top, V ) 0.5 V s-1; bottom,
cleavage of metal-metal bonded organometallic dimers.18 In
V ) 2.0 V s-1. Relevant simulation parameters: ks (oxidation) )
the present case, the best fit to CV scans over the scan rate
0.1 cm s-1, ks (reduction) ) 0.0025 cm s-1, Kdim ) 1.0 × 106 M-1
,
range of 0.5 to 2 V s-1 employed a dimerization equilibrium
kf ) 1.0 × 106 M-1 s-1
.
constant, Kdim, of 1 × 106 M -1 and a dimerization rate constant,
k
dim, of 106 M-1 s-1. It is clear from the cathodic wave shape
that the charge-transfer coefficient, R, is less than 0.5 for the
reduction process, and a good fit was obtained with R ) 0.35.
Although the simulations agreed with experiment when the
cally reversible process (i.e., one having relatively fast hetero-
geneous charge-transfer kinetics), the cathodic wave has the
characteristics of an electrochemically irreversible process, being
quite broad and exhibiting Epc values that are highly scan rate
dependent (e.g., -0.77 V at V ) 0.1 V s-1 and -0.82 V at V )
1 V s-1).
effective heterogeneous electron transfer was 0.0025 cm s-1
,
(18) Chong, D. S.; Nafady, A.; Costa, P. J.; Calhorda, J. M.; Geiger, W.
E. J. Am. Chem. Soc. 2005, 127, 15676, and references therein.

5896 Organometallics, Vol. 26, No. 24, 2007
Brownie et al.
Figure 4. Spectroelectrochemical data for 3.0 mM 2 in CH2Cl2/0.05 M [NBu4][B(C6F5)4] at 298 K. Arrows indicate the increase or decrease
in signal intensity as oxidation proceeds. CVs taken at V ) 0.1 V s-1
.
this value should be taken only as representative of a family of
rate constants, R values, and E1/2 values that would be consistent
with the cathodic process.
The cyclic voltammetric characteristics of 3 in CH2Cl2/0.05
M [NBu4][B(C6F5)4] were very similar to those of the molyb-
denum analogue, with only slight shifts in potential being
noted.19 At a scan rate of 0.1 V s-1, the chemically irreversible
oxidation peak appeared at Epa ) -0.20 V and the irreversible
cathodic product peak was found at Epc ) -0.81 V. More
detailed work on this system was not pursued, owing to its
apparent mechanistic similarity to 2.
When bulk anodic electrolysis of 2 was conducted at Eappl
)
0.15 V, the pale yellow solution turned orange-pink as the red
dimer dication, 222+, was formed. Voltammetry indicated
essentially quantitative formation of 222+ as the single product,
having an irreversible cathodic wave at Epc of approximately
-0.77 V, and back-reduction of this solution at Eappl ) -1.3 V
regenerated 2 in greater than 95% yield. One nuance of these
experiments is that the coulometry count was consistently
somewhat lower than that expected for a one-electron process,
being 0.7 to 0.8 F/equiv in three separate electrolyses. In situ
IR spectroscopy was therefore used to monitor the carbonyl
region of the reactant and product. As shown in Figure 4,
oxidation resulted in disappearance of the symmetric and
asymmetric νCO of 2 at 1915 and 1810 cm-1,1 respectively, and
the appearance of at least three new terminal carbonyl bands,
at 2029, 1981, and 1943 cm-1 and very close to the ν(CO) of
[Mo(η5-C5H4PMePh2)(CO)3]2[B(C6F5)4]2 in THF (Table 4). The
fact that there is an isosbestic point at 1920 cm-1 demonstrates
As shown previously,1 methyldiphenylphosphonium cyclo-
pentadienylide is less electron donating than is the formally
anionic η5-cyclopentadienyl ligand. This accounts for the fact
that compounds 1-3 oxidize at potentials that are considerably
more positive than do their 18-electron analogues, [M(η5-C5H5)-
(CO)3]-, for which E1/2 values of -0.82,3t -0.75,3x and -0.85
V3x were observed for the chromium, molybdenum, and tungsten
complexes, respectively. The [PMePh2]+ group therefore sta-
bilizes this family of 18-electron complexes against one-electron
oxidation by 440-650 mV. The ligand electronic effect must
be taken from the shift of 440 mV observed for the reversible
processes of the Cr complexes. In the context of other acceptor
(19) For both compounds 2 and 3, a second irreversible anodic wave is
observed at approximately Epa ) 1.40 V, apparently due to the oxidation
of the respective dimer dications formed near the electrode surface.
that the new absorption bands arise from a single product, the
2+
dimer dication 22
.

M(η5-C5H4PMePh2)(CO)3 (M ) Cr, Mo, W)
Organometallics, Vol. 26, No. 24, 2007 5897
Scheme 1. Synthetic Route to the Complexes [M(η5-C5H4PMePh2)(CO)3]2[B(C6F5)4]2 (M ) Cr, Mo, W)
substituents, this group has a greater electronic effect than a
bromine atom, an aldehyde group, or a trifluoromethyl group
(shifts of 170, 280, and 320 mV, respectively) and about the
same effect as the strongly electron-withdrawing perfluoro-
phosphazine group, N3P3F5.20,2l Put another way, the [PMePh2]+
group is at least as electron withdrawing as an amino group is
electron donating, evidenced by the fact that the NH2 substituent
effect on a C5H4R ring is only -370 mV.21,22 The electrostatic
effect of the phosphonium moiety is undoubtedly the major
reason for this result.
ion. Note that ν(CO) of [(η5-C5H4PMePh2)Mo(CO)3I]I are
observed at 2055 and 1979 cm-1,1 quite different from the ν-
(CO) observed here for the product of oxidation by [FeCp2]-
[PF6] and thus excluding an analogous fluoro complex.
X-ray Crystal Structures of [M(η5-C5H4PMePh2)(CO)3]2-
[B(C6F5)4]2 (M ) Cr, Mo, and W). X-ray quality crystals of
[12][B(C6F5)4]2 were grown from a solution of CH2Cl2 layered
with hexanes and kept at -30 °C; there are 1.5 CH2Cl2
molecules of solvation per mole of Cr in the unit cell, consistent
with the elemental analyses. Initially crystals of the molybdenum
and tungsten analogues were grown from 1:4 CH2Cl2/THF
solutions layered with hexanes and kept at -30 °C, but twinned
crystals were obtained in both cases. While the crystals of [22]-
[B(C6F5)4]2 were of sufficient quality for the structure to be
solved with acceptable errors on bond lengths and angles, the
twinned crystals of [32][B(C6F5)4]2 did not yield acceptable
crystallographic data. It was however possible to obtain un-
twinned X-ray quality crystals of the latter from a very dilute
THF solution layered with hexanes and kept at room temperature
over the course of three weeks. Crystals of the molybdenum
and tungsten dimeric complexes each contain 1.5 THF molecules
of solvation per mole of metal in the unit cell, but in these cases
the solvent molecules were removed under reduced pressure
prior to obtaining elemental analyses. The structures of [12]-
[B(C6F5)4]2 and [32][B(C6F5)4]2 were solved by direct methods,
but structural refinement of [22][B(C6F5)4]2 required the use of
ROTAX16c to find the twin law and then application of
SQUEEZE16d to squeeze out the THF molecules of solvation.
The crystal structures of all three complexes show that the
cations are present as dimers in the solid state. Selected bond
lengths and angles are presented in Table 1, while crystal-
lographic data are presented in Table 2, and the complete
structures of [12][B(C6F5)4]2 and [22][B(C6F5)4]2 are presented
in Figures 5 and 7, respectively; that of [32]2+, which is
isomorphous with [22]2+, is not shown. For purposes of clarity,
the structures of the cations, [12]2+ and [22]2+, with the anions
deleted, are also shown in Figures 6 and 8. The Cr-Cr bond
distance of the chromium complex is 3.3509(7) Å, apparently
the longest Cr-Cr bond distance known for a compound not
containing some type of bridging ligands. The longest Cr-Cr
bond distance reported heretofore is 3.471(1) Å, observed in
the bridged fulvalene (Fv) compound [Cr2(µ-Fv)(CO)6], in
which the geometry of the bridging ligand requires a long Cr-
Cr bond.24 The metal-metal bond of [12][B(C6F5)4]2 is signifi-
cantly longer than the Cr-Cr bonds in the analogous compounds
[Cr(η5-C5H5)(CO)3]2 (3.281(1) Å) and [Cr(η5-C5Me5)(CO)3]2
(3.3107 (7) and 3.310(1) Å for two polymorphs); comparisons
are shown in Table 3.3a,o,p
Bulk Syntheses of [M(η5-C5H4PMePh2)(CO)3]2[B(C6F5)4]2
(M ) Cr, Mo and W). The syntheses of [M(η5-C5H4PMePh2)-
(CO)3]2[B(C6F5)4]2 (M ) Cr, Mo and W) were carried out in
CH2Cl2 (Cr) or THF (Mo, W) by combining 1, 2, or 3 with
equimolar amounts of ferrocenium tetrakisperfluorophenylbo-
rate, [FeCp2][B(C6F5)4] (Scheme 1). Separation from ferrocene
was readily achieved by recrystallizing the products from
solutions of CH2Cl2 (M ) Cr) or THF (M ) Mo, W) by layering
with hexanes. Although 1+ is clearly the dominant species in
the chromium system under these conditions, the electrolysis
solution was ESR silent at room temperature, apparently a
consequence of the fast relaxation phenomena often encountered
with this family of metal-centered radicals.3y It is also important
to note that the three complexes do not take part in chlorine
atom abstraction reactions with CH2Cl2, which is a common
concern when handling metal-centered radicals.5
Oxidation of 1 could also be effected using [Ph3C][B(C6F5)4]
since the trityl cation (Ph3C+) has a calculated reduction
potential of E1/2 ≈ -0.12 V versus FeCp2
.
0/+ 23 The IR spectrum
of the material formed was identical to that obtained using
[FeCp2][B(C6F5)4] and the yield was similar, and we note that
the trityl ion has been used previously as an oxidant with the
compound Mo(η5-C5H5)(CO)(dppe)H (dppe ) Ph2PCH2CH2-
PPh2).23 In contrast, while use of [FeCp2][PF6] as an oxidant
with 1 did result in the formation of the oxidized product 1+
(IR), the formation of other products was evident even within
the first 5 min of reaction. In addition to the ν(CO) observed
for 1+, peaks at 1917, 1821, and 1808 cm-1 were also observed
and were the dominant features of the spectrum after 1.5 h.
While the peaks at 1917 and 1808 cm-1 are identical to those
of 1 in THF, we cannot propose a process that would result in
the re-formation of 1, and so these peaks are presumably to be
attributed to an unknown product that has incorporated fluoride
(20) Saraceno, R. A.; Riding, G. H.; Allcock, H. R.; Ewing, A. G. J.
Am. Chem. Soc. 1988, 110, 980.
(21) For a compilation of substituent effects on the redox potentials of
cyclopentadienyl complexes see: Lu, S.; Strelets, V. V.; Ryan, M. F.; Pietro,
W. J.; Lever, A. B. P. Inorg. Chem. 1996, 35, 1013. Using the notation of
this paper, the [PMePh2]+ group has an EL(L) value of 0.77 V [the EL(L)
value of the cyclopentadienyl ligand is 0.33 V]. For reference to the C5H4-
CF3 ligand see: Gassman, P. G.; Winter, C. H. J. Am. Chem. Soc. 1986,
108, 4228.
The metal-metal bond distances in [22][B(C6F5)4]2 and [32]-
[B(C6F5)4]2 are 3.2764(16) and 3.2507(4) Å, respectively, shorter
than the Cr-Cr bond of [12][B(C6F5)4]2, as in the trend observed
for the isoelectronic η5-C5H5 and η5-C5Me5 tricarbonyl deriva-
(22) Britton, W. E.; Kashyap, R.; El-Hashash, M.; El-Kady, M.;
Herberhold, M. Organometallics 1986, 5, 1029.
(23) Cheng, T.-Y.; Szalda, D. J.; Zhang, J.; Bullock, R. M. Inorg. Chem.
2006, 45, 4712.
(24) McGovern, P. A.; Vollhardt, K. P. C. Chem. Commun. 1996, 1593.

5898 Organometallics, Vol. 26, No. 24, 2007
Brownie et al.
Figure 5. Complete molecular structure of [12][B(C6F5)4]2.
for these cationic complexes is not observed for the neutral
complexes 1, 2, and 3, in which P(1)-C(1)-C5(centroid) bond
angles are nearly linear,1 but a similar feature is found in the
crystal structure of [Mo(η5-C5H4PMePh2)(CO)3I][I], in which
the P atom is also bent out of the C5 ring plane.1 The P(1)-
C(1) bonds, used as an indication of the P-C bond order, are
elongated in the oxidized species relative to the neutral
tricarbonyl starting materials,1 and again a similar lengthening
of the P(1)-C(1) bond is observed in [Mo(η5-C5H4PMePh2)-
(CO)3I][I].1 The chromium complex has the longest P(1)-C(1)
bond distance of 1.784(2) Å, but in all three complexes this
bond length approaches those of the P-Ph single bonds.
The C5 ring is in all cases bound in an η5 fashion with small
variations in the M-C bond distances; carbons C(1), C(2), and
C(5) are slightly closer to the metal than are C(3) and C(4).
The M-C5 bond distances are also shorter on average than those
observed in the neutral complexes, consistent with the above
suggestion of smaller metal covalent radii in the oxidized
complexes. The C5 ring bond lengths and angles do not differ
significantly from those of the neutral precursors, and the
structures of the borate anions are consistent with other
structures reported for this anion.25
Nature of the Complexes in Solution: IR, NMR, and MS
Data. Having established dimeric structures for all three
oxidized species in the solid state, we carried out a series of
spectroscopic studies (IR and NMR spectroscopy and ES-MS
on solutions, IR spectroscopy on the crystalline solids) to assess
the possibility of dissociation of the cationic dimers to mono-
meric species in solution. The related compounds [Cr(η5-C5H5)-
(CO)3]2 and [Cr(η5-C5Me5)(CO)3]2 do dissociate significantly,
while the analogous molybdenum and tungsten dimers do not
dissociate thermally at all.3
Figure 6. Structure of [12][B(C6F5)4]2 with the anions omitted for
clarity.
tives of chromium, molybdenum, and tungsten.3 The complex
[22][B(C6F5)4]2 has a Mo-Mo bond length similar to those of
[Mo(η5-C5H5)(CO)3]2 (3.235(1) Å) and [Mo(η5-C5Me5)(CO)3]2
(3.278(14) and 3.281(1) Å, P21/c and P21/n polymorphs),3p,r
while [32][B(C6F5)4]2 has a shorter W-W bond length than does
[W(η5-C5Me5)(CO)3]2 (3.288(1) Å).3s
The fact that the Cr-Cr bond of [12][B(C6F5)4]2 is longer
than that of the sterically very hindered [Cr(η5-C5Me5)(CO)3]2
may be due in part to electrostatic repulsions generated by
bringing two positively charged units together although, of
course, the covalent radii of the chromium atoms in the cationic
complex would also be smaller. Because of steric crowding
arising from the metal-metal bonds, the CO ligands are splayed
away from the adjacent metal center; consistent with this
hypothesis, the M-C-O bonds are not linear but are bent away
from the other monomer unit, and in all three complexes the
PMePh2 groups of the cyclopentadienylide ligands are orientated
away from each other, thus minimizing steric interactions.
In all three cases the P(1)-C(1)-C5(centroid) angles are
<180° such that the phosphorus atoms of the ligands are bent
away from the metal centers. The out-of-plane bending observed
IR spectra were obtained for all three compounds [M(η5-C5H4-
PMePh2)(CO)3]2[B(C6F5)4]2 (M ) Cr, Mo, W) in THF solutions
and in the solid state as Fluorolube mulls; all are soluble and
stable in THF at room temperature. Interestingly, dilute THF
solutions of [12][B(C6F5)4]2 at room temperature are yellow
(25) For several examples see: (a) Yang, X.; Stern, C. L.; Marks, T. J.
Organometallics 1991, 10, 840. (b) O’Connor, A. R.; Nataro, C.; Golen, J.
A.; Rheingold, A. L. J. Organomet. Chem. 2004, 689, 2411.

M(η5-C5H4PMePh2)(CO)3 (M ) Cr, Mo, W)
Organometallics, Vol. 26, No. 24, 2007 5899
Figure 7. Molecular structure of [22][B(C6F5)4]2.
Figure 8. Structure of [22][B(C6F5)4]2 with the anions omitted for
clarity.
Figure 10. IR spectra of [22][B(C6F5)4]2 in solution and as a mull.
The peaks at 1514 and 1643 cm-1 are attributed to the [B(C6F5)4]-
anion.
Figure 9. IR spectra of [12][B(C6F5)4]2 in solution and as a mull.
The peaks at 1514 and 1643 cm-1 are attributed to the [B(C6F5)4]-
anion.
Figure 11. IR spectra of [32][B(C6F5)4]2 in solution and as a mull.
The peaks at 1514 and 1643 cm-1 are attributed to the [B(C6F5)4]-
anion.
rather than the green of the crystalline dimer, while those of
[22][B(C6F5)4]2 and [32][B(C6F5)4]2 are red, the color of the
crystalline dimers. Data for the carbonyl stretching modes are
listed in Table 4, and the IR spectra are shown in Figures 9-11.
As can be seen, while the average frequency of the mull
spectrum of the chromium complex is similar to the averages
of the solid-state spectra of the molybdenum and tungsten
complexes, the peak pattern of the chromium complex differs
significantly from those of the other spectra. The difference is
probably a result of the chromium complex assuming a different
space group (Table 2), since the dimeric dication is the
dominant, if not exclusive, form of all three metal complexes
in the solid state.
The mull and solution IR spectra of the chromium dimer are
markedly different, however, in contrast with the corresponding
pairs of solid- and solution-phase spectra of the molybdenum

5900 Organometallics, Vol. 26, No. 24, 2007
Brownie et al.
and tungsten dimers, which are very similar. Indeed, the pattern
in the carbonyl stretching region of [12][B(C6F5)4]2 in either
THF or CH2Cl2 (see below) is very similar to that of the
monomeric Cr(η5-C5Me5)(CO)3, for which ν(CO) are observed
at 1994 (s) and 1876 (s, br) cm-1 in THF and 1994 (s) and
1890 (s, br) in toluene.3o Thus the IR data in Table 4 and Figures
2 and 8 are consistent with the presence of the 1+ ion in solution.
The carbonyl frequencies are, of course, higher for 1+ than for
Cr(η5-C5Me5)(CO)3 as a result of the positive charge in the
former. Assuming a double degeneracy of the 1890 cm-1 band
(νasym) in Cr(η5-C5Me5)(CO)3, a weighted average shift of +43
cm-1 in νCO is calculated for substitution of the C5Me5 ring by
C5H4PMePh2.
no spin-spin couplings could be observed in any of the
resonances. The combination of significant broadening of the
1H resonances and the very unusual chemical shifts of the C5H4
ring hydrogens is consistent with the species in solution being
paramagnetic,26 but an attempt to carry out a variable-temper-
ature experiment failed because of precipitation of [12]-
[B(C6F5)4]2 at lower temperatures.
As shown during our IR studies, however, this complex is
very soluble in THF, and 1H NMR spectra were therefore readily
obtained in THF-d8 over the temperature range 323 to 183 K.
No precipitation of either the green dimer or any other solid
was noted when the sample was cooled to 183 K, but the
solution, which is golden yellow at room temperature, turned
green reversibly on cooling. These color changes arise, of
Our observations that the mull IR spectra of [22][B(C6F5)4]2
and [32][B(C6F5)4]2 are nearly identical not only to each other
but to the corresponding solution spectra confirm our conclu-
sions based on the electrochemical and spectroelectrochemical
2+
course, from reversible interconversion of the green 12 and
the yellow 1+.
1
The H NMR spectrum of [12][B(C6F5)4]2 in THF-d8 at 298
2+
2+
K was very similar to that observed in CD2Cl2, exhibiting two
broad C5H4 resonances at δ 26.08 (∆ν1/2 ) 565 Hz) and 13.62
(∆ν1/2 ) 480 Hz), three broad phenyl resonances in the region
evidence that the dimeric ions 22 and 32 are the dominant
species in solution as well as in the solid state. Indeed, the IR
spectra, which exhibit three strong peaks (one a shoulder) in
the region 1900-1980 cm-1 in the solution spectra, are very
similar in pattern to the spectra of [Cr(η5-C5Me5)(CO)3]2 in THF
(1919, 1902, 1876 cm-1) and toluene (1920, 1900, 1875 cm-1),3o
although again the frequencies of the cationic complexes are
higher. Of particular interest is the rather weak feature at about
2025 cm-1 in all the spectra. Being close to the value of the
highest energy band of 1+ (2037 cm-1), it must be considered
as possibly arising from a small amount of the monomer 2+.
We attempted to test this possibility by recording the spectra
δ 8.07-7.66, and a broad P-Me resonance at δ 2.34 (∆ν1/2
)
95 Hz). However, lowering the temperature of the sample from
323 to 183 K resulted in major changes, the C5H4 resonances
broadening somewhat, shifting from δ 23.72 (∆ν1/2 ) 470 Hz)
and 12.85 (∆ν1/2 ) 385 Hz) at 323 K to δ 28.70 (∆ν1/2 ) 820
Hz) and 14.10 (∆ν1/2 ) 795 Hz) at 253 K, and weakening
relative to the resonance of CH2Cl2 added and used as an internal
reference. Concomitantly, weak new resonances began to
appeared in the “normal” C5H4 region, at δ ∼5.5-6.2.
By 233 K, the integrated intensities of the two C5H4
resonances, now at δ 29.97 and 12.70, had become extremely
low and had been replaced by four somewhat broadened
resonances of equal intensity, at δ 6.12, 5.96, 5.85, and 5.65.
The chemical shifts of the latter four resonances are similar to
the C5H4 chemical shifts of 1,1 2,1 3,1 222+, and 322+, supporting
the assignments. That there were four separate resonances at
this very low temperature rather than the two observed for the
other compounds at room temperature probably arises because
the cationic dimer is “frozen” in a conformation in which the
four H atoms are not equivalent. Indeed, the observed lack of
symmetry is consistent with the solid-state structure in which
the four ring protons are nonequivalent because of the fixed
position of the PMePh2 group; at these low temperatures, slowed
rotation of the PMePh2 would be expected and would give rise
to magnetic nonequivalence of the C5H4 protons.
2+
of 22 at different solution temperatures, but essentially no
difference in the relative intensities of the weakest and strongest
carbonyl bands was observed between 273 and 310 K in CH2-
Cl2. This result argues for assignment of the 2037 cm-1 band
to the dimer 222+, rather than the monomer 2+, but we cannot
rule out the presence, suggested in NMR data (see below), of
minor amounts of the monomer radical cation under these
conditions.
High-resolution electrospray mass spectra (THF solutions)
were also obtained as part of the characterization of the cationic
products of the reactions of 1, 2, and 3 with [FeCp2][B(C6F5)4],
and for all three systems, the mass spectra exhibited no peaks
attributable to dimeric species. The only molecular ions observed
were those of the monomeric, 17-electron radicals, 1+, 2+, and
3+, for which the experimental m/z values differed by only
6.8926, 4.1748, and 3.5448 ppm, respectively, from the
calculated m/z values. Also observed in all cases were excellent
isotopic peak distribution matches with the calculated distribu-
tions expected for each molecular ion, and an MS-MS
experiment of the molecular ion of 1+ revealed a major fragment
ion with m/z of 316, corresponding to the loss of three carbonyls.
The conclusions to be drawn from this information in isolation
would be that the species in solution are monomeric. However,
in view of the overwhelming mass of electrochemical and IR
data discussed above, the molybdenum and tungsten dimers must
dissociate under the conditions of the mass spectroscopy
experiment.
Simultaneously with the above changes, the P-Me resonance
(δ 2.34 at 298 K) decreased in intensity as a new resonance at
δ 2.79 appeared and gained in intensity. The original P-Me
resonance had disappeared completely at 213 K, leaving only
a broadened P-Me resonance at δ 2.89. The latter was not a
singlet, but whether the splitting observed was a result of
coupling to 31P or of “freezing” of more than one conformation
or a combination of both could not be determined. Changes in
the closely spaced phenyl resonances as the temperature was
lowered were relatively subtle and shed no new light on the
situation.
1
To further investigate the equilibration implied by the H
NMR experiments were also carried out to characterize the
NMR data, the Evans method27a-c was used to determine the
average magnetic moment of the chromium atoms of [Cr(η5-
C5H4PMePh2)(CO)3]2[B(C6F5)4]2 in CD2Cl2 at 298 K and in
THF over the temperature range 323 to 233 K. The magnetic
1
solution behavior of the oxidized species; these included H,
1
13C, and 31P NMR experiments and variable-temperature H
experiments. A 600 MHz 1H NMR spectrum of [12][B(C6F5)4]2
in CD2Cl2 at 298 K exhibited two very broad C5H4 resonances
at δ 28.90 (∆ν1/2 ) 535 Hz) and 13.65 (∆ν1/2 ) 470 Hz), three
somewhat broadened phenyl resonances at δ 7.85, 7.65, and
7.60, and a broad methyl resonance at δ 1.18 (∆ν1/2 ) 75 Hz);
(26) Bertini, I.; Luchinat, C.; Parigi, G. Solution NMR of Paramagnetic
Molecules: Applications to Metallobiomolecules and Models; Elsevier
Science B.V.: Amsterdam, 2001.

M(η5-C5H4PMePh2)(CO)3 (M ) Cr, Mo, W)
Organometallics, Vol. 26, No. 24, 2007 5901
case could H-1H coupling be resolved, but variable-temper-
ature experiments conducted in the temperature range 323-
213 K showed that the phenyl resonances decoalesced somewhat
at 323 K and exhibited three peaks. The C5H4 resonances began
to overlap as the temperature was lowered. The 31P NMR
spectrum exhibited only one resonance, at δ 19.88 and slightly
shifted from that of 2 (δ 18.25 in CD2Cl2). The 13C NMR
spectrum was very similar to that previously reported for [Mo-
(η5-C5H4PMePh2)(CO)3I][I],1 with nearly identical chemical
shifts.
1
Table 5. Magnetic Susceptibilities of [12][B(C6F5)4]2 at
Different Temperatures
temperature (K)
solvent
µ (µB) ((0.05)
323
308
298
273
253
233
298
THF-d8
THF-d8
THF-d8
THF-d8
THF-d8
THF-d8
CD2Cl2
1.83
1.78
1.80
1.93
1.78
1.46
1.71
The 1H NMR spectrum of [32][B(C6F5)4]2 at 298 K exhibited
two broad phenyl resonances at δ 7.80-7.71, the P-Me
resonance as a broad doublet at δ 2.86, and the C5H4 resonances
susceptibilities were measured using cyclohexane and the
residual proton signal of CD2Cl2 as references, and diamagnetic
corrections and magnetic susceptibilities of solvents were taken
from literatures sources.27 The resulting magnetic moments
determined at each temperature are shown in Table 5. As can
be seen, the experiments carried out at higher temperatures
yielded for both solvents magnetic moments of ∼1.8 µB, very
close to the spin-only value for one unpaired electron.27
However, the average magnetic moment decreased to 1.46 µB
1
as a broad band centered at δ 6.05. Variable-temperature H
NMR spectra were obtained in the temperature range 198-323
K, and a third phenyl resonance was observed and the C5H4
multiplet narrowed above 298 K.
Summary
on cooling to 233 K, a result consistent with significant
The group 6 metal complexes M(η5-C5H4PMePh2)(CO)3 (M
) Cr (1), Mo (2), W (3)) are oxidized via one-electron processes,
both electrochemically and in bulk by [FeCp2][B(C6F5)4], to give
the corresponding 17-electron, metal-centered radicals [M(η5-
C5H4PMePh2)(CO)3]+ (M ) Cr (1+), Mo (2+), and W (3+)) as
the [B(C6F5)4]- salts. These in turn dimerize to give the metal-
metal bonded complexes [M(η5-C5H4PMePh2)(CO)3]2[B(C6F5)4]2
(M ) Cr (122+), Mo (222+), and W (322+)), which have been
isolated analytically pure and were characterized by IR (solid
state and solution) and NMR (1H, 13C, 19F, 31P) spectroscopy,
high-resolution mass spectrometry, and crystallographically. The
dimers all contain metal-metal bonds that are comparable in
length with or longer than the metal-metal bonds in the
isoelectronic, neutral η5-C5H5 and η5-C5Me5 compounds, and
2+
dimerization of 1+ to the diamagnetic 12
.
These observations extend our conclusions, discussed above
and based on visual, electrochemical, and IR spectroscopic
evidence, that 122+ undergoes reversible dissociation in solution
to 1+. However, although our electrochemical investigation of
this equilibrium in CH2Cl2 at temperatures from ambient to 238
K failed to show any changes that would indicate a degree of
dimerization of 1+, the magnetic moment data suggest that
significant dimerization occurs in THF at temperatures as high
as 233 K. In addition, the low-temperature 1H NMR data suggest
that little or no monomer is present at or below 213 K in THF.
Similar observations have been made previously for [Cr(η5-
C5H5)(CO)3]2 and [Cr(η5-C5Me5)(CO)3]2,3 of course, but in one
sense the [12][B(C6F5)4]2 equilibrium stands in stark contrast
to the two neutral monomer-dimer systems. Exchange pro-
cesses between monomer and dimer for both neutral systems
2+
the metal-metal bond in 12 is the longest nonbridged Cr-
Cr bond known. As a result of the apparent weakness of its
2+
Cr-Cr bond, 12 dissociates significantly in solution to the
1
are sufficiently rapid on the NMR time scale that averaged H
paramagnetic, radical cation 1+. The redox behavior of the
neutral compounds 1, 2, and 3 is qualitatively similar to that of
the anionic, isoelectronic analogues [M(η5-C5H5)(CO)3]- and
[M(η5-C5Me5)(CO)3]-, with the [PMePh2]+ substituent exhibit-
ing an electron-withdrawing effect greater than that of the
trifluoromethyl group and thus imparting dramatic changes in
the redox potentials.
resonances are observed.3 However, for [12][B(C6F5)4]2, separate
C5H4 and P-Me resonances are observed over a wide temper-
ature range, and thus monomer-dimer exchange for this system
must be slow on the NMR time scale. We are unaware of any
other such system where this type of behavior has been
observed.
1
The H, 31P, and 13C NMR spectra of [22][B(C6F5)4]2 and
[32][B(C6F5)4]2 were obtained in THF-d8 since these compounds
exhibit low solubilities in CD2Cl2. For both complexes, the
proton and phosphorus resonances were somewhat broadened,
possibly evidence for the presence of a small amount of
monomeric radicals in the solutions.
Acknowledgment. We thank the Government of Ontario
(Ontario Graduate Scholarship to J.H.B.), Queen’s University
(Queen’s Graduate Award to J.H.B.), the Natural Sciences and
Engineering Research Council (Discovery Grant to M.C.B.),
and the National Science Foundation (CHE-0411703 to W.E.G.)
for funding this research.
1
Thus the H NMR spectrum of [22][B(C6F5)4]2 exhibited a
poorly resolved P-Me doublet at δ 2.88, C5H4 resonances as
broad, featureless bands in the region δ 6.10-6.04 and at δ
5.91, and phenyl resonances in the region δ 7.87-7.72. In no
Supporting Information Available: One cyclic voltammogram
of 1 showing irreversible second anodic wave. Crystallographic
details, including figures of [12][B(C6F5)4]2, [22][B(C6F5)4]2, and
[32][B(C6F5)4]2, showing complete numbering schemes and thermal
ellipsoid figures, and tables of positional and thermal parameters
and bond lengths and angles. This material is available free of
charge via the Internet at http://pubs.acs.org.
(27) (a) Evans, D. F. J. Chem. Soc. 1959, 2003. (b) Crawford, T. H.;
Swanson, J. J. Chem. Educ. 1971, 48, 382. (c) Sur, S. K. J. Magn. Reson.
1989, 82, 169. (d) Earnshaw, A. Introduction to Magnetochemistry;
Academic Press: New York, 1968. (e) Abeles, T. P.; Bos, W. G. J. Chem.
Educ. 1967, 44, 438. (f) Weast, R. C., Lide, D. R., Eds. CRC Handbook of
Chemistry and Physics, 70th ed.; CRC Press: Boca Raton, FL, 1990; pp
E-134-139.
OM700688P