Stable mononuclear, 17-electron molybdenum(III) carbonyl complexes. Synthesis, structure, thermal decomposition, and Cl- addition reactions

Article abstract of DOI:10.1021/ja9540738

The first examples of stable, mononuclear 17-electron carbonyl complexes of Mo(III) have been synthesized, isolated, and characterized by IR and EPR spectroscopy. Oxidation of Cp*MoCl(CO)(PMe₃)₂ (1; E( 1/2 ) = -0.48 V vs Fc/Fc⁺), Cp*MoCl(CO)(dppe) (2; E( 1/2 ) = -0.44 V), and CpMoCl(CO)(dppe) (3; E( 1/2 ) = -0.25 V) with Fc⁺PF₆^- yields [1]PF₆, [2]PF₆, and [3]PF₆, respectively. The IR stretching vibration of the 17-electron oxidation products are 136-153 cm^-1 blue-shifted with respect to the corresponding stretching vibrations of the parent Mo(II) compounds. The room temperature EPR spectra show observable coupling to the Mo and P nuclei and indicate a trans geometry for 1⁺ and a cis geometry for 2⁺ and 3⁺. The four-legged piano stool geometry of 2⁺ with the phosphines atoms in relative cis positions has been confirmed by a single-crystal X-ray analysis. The X-ray data for [2]PF₆·THF are the following: monoclinic, P2₁/n, a = 13.7394(14) A?, 6 = 20.421(2) A?, c = 14.857(2) A?, β = 99.119(8)°, V = 4115.8(8) A?³, Z = 4 D(x) = 1.469 g·cm^-3, λ(Mo Kα) =0.71073 A?, μ (Mo Kα) = 0.562 mm^-1, R(F) = 4.59%, R(wF²) = 10.77% for 4299 data with E(o) > 4σ(F(o)) The thermal stability of the cations decreases with increasing carbonyl stretching frequencies in the order 1⁺ > 2⁺ >> 3⁺. The decarbonylated product of thermolysis of 3⁺, [CpMoClF(dppe)(MeCN)]⁺PF₆^-, [4]PF₆, which is believed to arise via a 15-electron [CpMoCl(dppe)]⁺ intermediate, has been isolated and characterized by ¹H, ³¹P, and ¹⁹F NMR spectroscopies as well as a single-crystal X-ray analysis. The structure of 4⁺ shows a distorted pseudooctahedral geometry with the Cp ring and the F atom occupying the two axial coordination sites and the two phosphine atoms of the dppe occupying two cis equatorial sites. The data for [4]PF₆·MeCN are as follows: monoclinic, P2₁/n, a = 13.160(2) A?, b = 10.942(2) A?, c = 25.010(3) A?, β = 95.225(10)°, V = 3586.4(9) A?³, Z = 4, D(x) = 1.557 g·cm^-3, μ(Mo Kα) = 0.71073 A?, R(F) = 4.11% R(wF²) = 11.81% for 5964 data with F(o) > 4σ(F(o)). The nucleophilic addition of Cl^- to 1⁺, 2⁺, or 3⁺ is followed by redox processes to afford mixtures of the parent Mo(II) carbonyl complexes 1-3 and Mo(IV) or Mo(V) products, depending on the nature of the ligands. The mechanism of these reactions has been elucidated through parallel chemical and electrochemical studies.

Full text of DOI:10.1021/ja9540738

J. Am. Chem. Soc. 1996, 118, 3617-3625
3617
Stable Mononuclear, 17-Electron Molybdenum(III) Carbonyl
Complexes. Synthesis, Structure, Thermal Decomposition, and
Cl- Addition Reactions
James C. Fettinger, D. Webster Keogh,* and Rinaldo Poli*
Contribution from the Department of Chemistry and Biochemistry, UniVersity of Maryland,
College Park, Maryland 20742
ReceiVed December 4, 1995X
Abstract: The first examples of stable, mononuclear 17-electron carbonyl complexes of Mo(III) have been synthesized,
isolated, and characterized by IR and EPR spectroscopy. Oxidation of Cp*MoCl(CO)(PMe3)2 (1; E1/2 ) -0.48 V
-
vs Fc/Fc+), Cp*MoCl(CO)(dppe) (2; E1/2 ) -0.44 V), and CpMoCl(CO)(dppe) (3; E1/2 ) -0.25 V) with Fc+PF6
yields [1]PF6, [2]PF6, and [3]PF6, respectively. The IR stretching vibration of the 17-electron oxidation products
are 136-153 cm-1 blue-shifted with respect to the corresponding stretching vibrations of the parent Mo(II) compounds.
The room temperature EPR spectra show observable coupling to the Mo and P nuclei and indicate a trans geometry
for 1+ and a cis geometry for 2+ and 3+. The four-legged piano stool geometry of 2+ with the phosphines atoms
in relative cis positions has been confirmed by a single-crystal X-ray analysis. The X-ray data for [2]PF6‚THF are
the following: monoclinic, P21/n, a ) 13.7394(14) Å, b ) 20.421(2) Å, c ) 14.857(2) Å, â ) 99.119(8)°, V )
4115.8(8) Å3, Z ) 4, Dx ) 1.469 g‚cm-3, λ(Mo KR) ) 0.71073 Å, µ (Mo KR) ) 0.562 mm-1, R(F) ) 4.59%,
R(wF2) ) 10.77% for 4299 data with Fo > 4σ(Fo). The thermal stability of the cations decreases with increasing
carbonyl stretching frequencies in the order 1+ > 2+ . 3+. The decarbonylated product of thermolysis of 3+,
[CpMoClF(dppe)(MeCN)]+PF6-, [4]PF6, which is believed to arise via a 15-electron [CpMoCl(dppe)]+ intermediate,
has been isolated and characterized by 1H, 31P, and 19F NMR spectroscopies as well as a single-crystal X-ray analysis.
The structure of 4+ shows a distorted pseudooctahedral geometry with the Cp ring and the F atom occupying the
two axial coordination sites and the two phosphine atoms of the dppe occupying two cis equatorial sites. The data
for [4]PF6‚MeCN are as follows: monoclinic, P21/n, a ) 13.160(2) Å, b ) 10.942(2) Å, c ) 25.010(3) Å, â )
95.225(10)°, V ) 3586.4(9) Å3, Z ) 4, Dx ) 1.557 g‚cm-3, µ(Mo KR) ) 0.71073 Å, R(F) ) 4.11%, R(wF2) )
11.81% for 5964 data with Fo > 4σ(Fo). The nucleophilic addition of Cl- to 1+, 2+, or 3+ is followed by redox
processes to afford mixtures of the parent Mo(II) carbonyl complexes 1-3 and Mo(IV) or Mo(V) products, depending
on the nature of the ligands. The mechanism of these reactions has been elucidated through parallel chemical and
electrochemical studies.
Introduction
(CO)2 complexes8,9 fully decarbonylate at the mild CH2Cl2
reflux.10 In addition, a greater kinetic lability is expected for
the metal-CO bond in odd-electron systems.1,2 We report here
the synthesis, isolation, and characterization of the first examples
of stable mononuclear, 17-electron Mo(III) carbonyl derivatives.
Stable CO derivatives of Mo(III) were previously limited to
dinuclear, metal-metal bonded, diamagnetic compounds, e.g.
[Cp2Mo2(µ-SR)2(CO)3L]2+ (R ) Me, Ph; L ) CO, MeCN,
t-BuCN),11-13 [Cp2Mo2(µ-SMe)3(CO)2]+,14 and [Cp2Mo2(µ-
SMe)(µ-S2CH2)(CO)2]+.15
Organometallic radicals, especially those characterized by a
17-electron configuration, are receiving a great deal of attention
in view of their involvement as intermediates in a variety of
fundamental processes.1,2 Stable 17-electron compounds con-
taining carbon monoxide as a ligand are generally confined to
the low oxidation states. For complexes of molybdenum, very
few such species have been isolated to date and these all contain
the metal in the oxidation state +1, e.g. [Mo(CO)2(L-L)2]+
(L-L ) dmpe, dppe),3,4 [Mo(CO)2(CNxylyl)2(PEt2Ph)2]+,5
[Mo(CO)2L2(t-Bu-DAB)]+ (L ) MeCN, MeNC, t-BuNC; DAB
) 1,4-diaza-1,3-butadiene),6 Tp*Mo(CO)3 [Tp* ) hydridotris-
(3,5-dimethylpyrazolyl)borate].7 Upon increasing the oxidation
state, it is expected that the metal-CO bond weakens. For
instance, although the CpMoX(CO)3 (X ) halogen) complexes
are thermally stable toward CO loss, the corresponding Cp-
MoX3-
Experimental Section
General Data. All operations were carried out under an atmosphere
of argon with standard Schlenk-line techniques. Solvents were purified
by conventional methods and distilled under argon prior to use. FT-
(8) Haines, R. J.; Nyholm, R. S.; Stiddard, M. H. B. J. Chem. Soc. (A)
1966, 1966, 1606-1607.
(9) Green, M. L. H.; Lindsell, W. E. J. Chem. Soc. (A) 1967, 686-687.
(10) Gordon, J. C.; Lee, V. T.; Poli, R. Inorg. Chem. 1993, 32, 4460-
4463.
(11) Courtot-Coupez, J.; Gue´guen, M.; Guerchais, J. E.; Pe´tillon, F. Y.;
Talarmin, J.; Mercier, R. J. Organomet. Chem. 1986, 312, 81-95.
(12) El Khalifa, M.; Gue´guen, M.; Mercier, R.; Pe´tillon, F. Y.; Saillard,
J.-Y.; Talarmin, J. Organometallics 1989, 8, 140-148.
(13) Gue´guen, M.; Pe´tillon, F. Y.; Talarmin, J. Organometallics 1989,
8, 148-154.
X Abstract published in AdVance ACS Abstracts, April 1, 1996.
(1) Baird, M. C. Chem. ReV. 1988, 88, 1217-1227.
(2) Trogler, W. C., Ed. Organometallic radical processes; Trogler, W.
C., Ed.; Elsevier: Amsterdam, 1990; Vol. 22.
(3) Reimann, R. H.; Singleton, E. J. Organomet. Chem. 1971, 32, C44-
C46.
(4) Connor, J. A.; Riley, P. I. J. Chem. Soc., Dalton Trans. 1979, 1231-
1237.
(5) Conner, K. A.; Walton, R. A. Inorg. Chem. 1986, 25, 4422-4430.
(6) Bell, A.; Walton, R. A. Polyhedron 1986, 5, 845-858.
(7) Shiu, K.-B.; Lee, L.-Y. J. Organomet. Chem. 1988, 348, 357-360.
(14) Gomes de Lima, M. B.; Guerchais, J. E.; Mercier, R.; Pe´tillon, F.
Y. Organometallics 1986, 5, 1952-1964.
(15) Tucker, D. S.; Dietz, S.; Parker, K. G.; Carperos, V.; Gabay, J.;
Noll, B.; Rakowski Dubois, M. Organometallics 1995, 14, 4325-4333.
0002-7863/96/1518-3617$12.00/0 © 1996 American Chemical Society

3618 J. Am. Chem. Soc., Vol. 118, No. 15, 1996
Fettinger et al.
IR spectra were recorded on a Perkin-Elmer 1800 spectrophotometer
with a liquid IR cell equipped with KBr plates. NMR spectra were
recorded on Bruker WP200, AF200, and AM400 spectrometers; the
peak positions are reported with positive shifts downfield of TMS as
calculated from the residual solvent peaks (1H), downfield of external
85% H3PO4 (31P), or downfield of external TFA (19F). For each 31P
and 19F-NMR spectrum a sealed capillary containing H3PO4 or TFA
was immersed in the same NMR solvent used for the measurement
and this was used as the reference. Samples for EPR spectra were
placed in 3-mm glass tubes and were recorded on a Bruker ER200
spectrometer upgraded to ESP 300 equipped with an X-band microwave
generator, using DPPH (g ) 2.004) as an external calibrant. Cyclic
voltammograms were recorded with an EG&G 362 potentiostat
connected to a Macintosh computer through MacLab hardware/software.
The electrochemical cell used consisted of a modified Schlenk tube
with a Pt counter electrode sealed through uranium glass/Pyrex glass
seals. The cell was fitted with a Ag/AgCl reference electrode and a Pt
disk working electrode. All half-wave potentials were measured and
are reported with respect to the ferrocene/ferricenium couple. The
ferrocene was added to the solution at the end of each measurement as
an internal standard. No IR correction was applied during the
measurements. The elemental analyses were carried out by M-H-W
Synthesis of Cp*MoCl(dppe)(CO) (2). Cp*MoCl4 (0.260 g, 0.778
mmol) was added to a THF solution (25 mL) of dppe (0.310, 0.778
mmol), containing Na/Hg (0.060 g, 2.60 mmol in 7 g of Hg). Stirring
for approximately 5 min gave a light brown colored solution with a
brown precipitate. The argon atmosphere was replaced with CO and
the solution was allowed to stir overnight, during which time the
solution became clear and dark orange in color. The THF was removed
under reduced pressure, the residue was extracted with toluene (10 mL),
and the resulting solution was then filtered through Celite. The Celite
was rinsed with toluene (2 × 2 mL) and the rinsings added to the
original toluene extract. The toluene solution was concentrated to ca.
2 mL and heptane (20 mL) was added, precipitating an orange solid.
The mixture was placed at -80 °C for 1 h, after which time the solid
was filtered, washed with heptane (2 × 5 mL), and dried under vacuum.
Yield: 0.318 g, 60%. Anal. Calcd (found) for C37H39ClMoOP2: C,
64.12 (63.91); H, 5.67 (5.59). IR (THF): 1835 cm-1
.
1H-NMR
(C6D6): δ 7.77 (m, 4 H, Ph2PCH2CH2PPh2), 7.61 (m, 20 H, Ph2PCH2-
CH2PPh2), 1.55 (br, 4H, Ph2PCH2CH2PPh2), 1.41 (s, 15H, C5Me5). 31P-
{1H}-NMR (C6D6): δ 90.2 (d, JPP ) 38 Hz, 1 P), 64.5 (d, JPP ) 38
Hz, 1 P).
Synthesis of [Cp*MoCl(dppe)(CO)]PF6 ([2]PF6). By a procedure
analogous to that described above for 1+, the interaction of 0.383 g
(0.508 mmol) of 2 with 0.180 g (0.515 mmol) of Fc+PF6- yielded 320
g (75% yield) of [2]PF6. Anal. Calcd (found) for C37H39ClF6MoOP3:
C, 53.03 (52.6); H, 4.69 (4.90). IR (THF): 1971 cm-1. EPR (THF):
Laboratories, Phoenix, AZ, Galbraith Laboratories Inc., Knoxville, TN,
16,17
and Desert Analytics Laboratory, Tucson, AZ. Cp*MoCl4
and
{CpMoCl2}n18 were prepared as previously described. PMe3 (Aldrich),
dppe (Strem), and CO (Air Products) were used without any further
purification.
g ) 2.003 (dd with Mo satellites; aP1 ) 28.1 G; aP2 ) 17.7 G; aMo
29.8 G).
)
Synthesis of Cp*MoCl(PMe3)2(CO) (1). Cp*MoCl4 (0.460 g, 1.23
mmol) was added to a THF solution (20 mL) of PMe3 (0.255 mL,
2.46 mmol) containing Na/Hg (0.090 g, 3.91 mmol in 10 g of Hg)
giving a purple colored solution, which was stirred at room temperature.
Over a period of 12 h the color changed from purple to red-brown to
yellow-brown. An aliquot was taken for an NMR in C6D6 to confirm
the identity of Cp*MoCl(PMe3)219 and to verify the complete consump-
tion of the intermediate Cp*MoCl2(PMe3)2 (see Results). The THF
solution was filtered through Celite and then evaporated to dryness
under reduced pressure. The residue was dissolved in heptane (10 mL)
and the argon atmosphere was then replaced with CO, resulting in an
immediate color change to red-orange. The heptane solution was
Synthesis of CpMoCl(dppe)(CO) (3). CpMoCl2(dppe) was pre-
pared in situ from {CpMoCl2}n (0.544 g, 2.34 mmol) and dppe (0.932
g, 2.34 mmol) in CH2Cl2 (5 mL) as previously described,20 and
recovered as a solid by reducing the volume to ca. 1 mL and
precipitation with heptane (10 mL). The solid was washed with heptane
(2 × 5 mL) and then suspended in THF (40 mL). The THF suspension
was transferred via cannula onto Na/Hg (0.058 g, 2.52 mmol in 6 g of
Hg) and the atmosphere of argon was replaced with CO. After stirring
overnight the solution was dark orange in color. The THF solution
was concentrated to ca. 10 mL and an orange solid was precipitated
by adding heptane (30 mL). The solid was washed with heptane (3 ×
5 mL) and dried under vacuum. Yield: 0.867 g, 60%. The
spectroscopic properties of this material are identical to those previously
1
concentrated to ca. /2 of the original volume and then placed at -80
reported.21 IR (THF): 1853 cm-1
.
1H-NMR (C6D6): δ 7.88 (m, 4 H,
°C overnight. Orange crystals were isolated and dried under vacuum.
Yield: 0.160 g, 29%. MS (CI, negative ions): 448 ([M]-, 100%),
Ph2PCH2CH2PPh2), 7.74 (m, 8H, Ph2PCH2CH2PPh2), 7.63 (m, 8H, Ph2-
PCH2CH2PPh2), 4.50 (s, 5 H, Cp), 2.69 (br, Ph2PCH2CH2PPh2), 1.77
(br, Ph2PCH2CH2PPh2). 31P{1H}-NMR (CDCl3): δ 92.9 (d, JPP ) 38
Hz, 1 P), 68.0 (d, JPP ) 37 Hz, 1 P).
372 ([M]- - PMe3, 93%). IR (THF): 1782 cm-1 1H-NMR (C6D6):
.
δ 1.66 (s, 15 H, C5Me5), 1.30 (vt, JPH ) 4 Hz, 18 H, PMe3). 31P{1H}-
NMR (C6D6): δ 18.5.
-
Synthesis of [Cp*MoCl(PMe3)2(CO)]PF6 ([1]PF6). Cp2Fe+PF6
Synthesis of [CpMoCl(dppe)(CO)]PF6 ([3]PF6). By a procedure
analogous to that described above for [1]PF6, the interaction of
(0.179 g, 0.512 mmol) was added to a stirring THF solution (10 mL)
of 1 (0.229 g, 0.512 mmol). An immediate reaction occurred quenching
the blue color of the Fc+. The THF was concentrated under reduced
pressure to approximately 3 mL and the product 1+PF6- was precipitated
with heptane (15 mL), filtered, washed with heptane (3 × 5 mL), and
dried under vacuum. Yield: 0.207 g, 68%. Anal. Calcd (found) for
C17H33ClF6MoOP3: C, 34.51 (33.27); H, 5.62 (5.68). Low C analyses
(erratically in the range of 30.33-33.27% from different commercial
laboratories, with differences greater than 0.5% even for duplicate
analyses on the same preparation) for this compound have been
repeatedly obtained on several different preparations, in spite of the
homogeneity of the crystalline samples upon optical inspection and
the spectroscopic purity. We tentatively attribute this phenomenon to
incomplete combustion during the analytical procedure. No EPR-active
impurity is observed in the EPR spectrum (see Results section) and no
diamagnetic impurity is observed in the NMR spectrum (in d8-THF),
-
0.240 g (0.385 mmol) of 3 with 0.143 g (0.409 mmol) of Fc+PF6
yielded 207 g (70% yield) of [3]PF6. Anal. Calcd (found) for
C31H29ClF6MoOP3‚(C7H16)0.25: C, 51.12 (51.21); H, 4.19 (4.26). The
presence of heptane was confirmed by 1H-NMR spectroscopy. IR
(THF): 2002 cm-1. EPR (THF): g ) 2.002 (dd with Mo satellites;
aP1 ) 28.6 G; aP2 ) 16.6 G; aMo ) 30.2 G).
General Procedure For Studying the Thermal Stability of the
Carbonyl Cations. Approximately 0.050 g of [1]PF6, [2]PF6, and [3]-
PF6 was dissolved in THF (5 mL). The solutions were heated in an
oil bath thermostated at 55 °C. The concentrations of the carbonyl
cations were then analyzed by monitoring the intensity of the νCO bands
in the IR spectra of the solutions.
Synthesis of [CpMoClF(dppe)(MeCN)]PF6 ([4]PF6). A THF
solution (5 mL) of [3]PF6 (0.164 g, 0.214 mmol) was heated for 1 h at
55 °C. The red precipitate that had formed was filtered off and dried
under vacuum. Yield: 0.036 g, 44% based on [CpMoClF(dppe)]PF6.
Anal. Calcd (found) for C31H29ClF7MoP3: C, 49.06 (48.83); H, 3.85
(4.02). The red solid was dissolved in MeCN and crystals of [4]PF6
(0.030 g) were obtained by slow diffusion of a Et2O layer. 1H-NMR
(d3-MeCN): δ 8.29 (m, 2H, Ph2PCH2CH2PPh2), 7.78 (m, 4H, Ph2-
the latter showing only the solvent resonances. IR (THF): 1920 cm-1
.
EPR (THF): g ) 2.005 (t with Mo satellites; aP ) 23.0 G; aMo
26.3 G).
)
(16) Murray, R. C.; Blum, L.; Liu, A. H.; Schrock, R. R. Organometallics
1985, 4, 953-954.
PCH2CH2PPh2), 7.48 (m, 14H, Ph2PCH2CH2PPh2), 5.17 (t, 5H, JPH
)
(17) Abugideiri, F.; Keogh, D. W.; Kraatz, H.-B.; Poli, R.; Pearson, W.
J. Organomet. Chem. 1995, 488, 29-38.
7.2 Hz, C5H5), 3.25 (m, 4H, Ph2PCH2CH2PPh2). 31P{1H}-NMR
(18) Linck, R. G.; Owens, B. E.; Poli, R.; Rheingold, A. L. Gazz. Chim.
Ital. 1991, 121, 163-168.
(20) Krueger, S. T.; Owens, B. E.; Poli, R. Inorg. Chem. 1990, 29, 2001-
2006.
(21) Sta¨rker, K.; Curtis, M. D. Inorg. Chem. 1985, 24, 3006-3010.
(19) Abugideiri, F.; Keogh, D. W.; Poli, R. J. Chem. Soc., Chem.
Commun. 1994, 2317-2318.

Stable Mononuclear, 17-e Mo(III) CO Complexes
J. Am. Chem. Soc., Vol. 118, No. 15, 1996 3619
Table 1. Crystal Data for [2]PF6‚THF and [4]PF6‚MeCN
(d3-MeCN): δ 55.5 (br, d, 1P, JPF ) 73 Hz), 45.6 (br, d, 1P, JPF ) 53
Hz). 19F-NMR (d3-MeCN): δ 13.17 (dd, MoF, JFP1 ) 75 Hz, JFP2
61 Hz), 9.47 (d, PF6, JFP ) 706 Hz).
)
[2]PF6‚THF
[4]PF6‚MeCN
empirical formula
formula wt
temp, K
wavelength, Å
crystal system
space group
unit cell dimensions
a, Å
C41H47ClF6MoO2P3
910.09
153(2)
0.71073
monoclinic
P2(1)/n
C35H35ClF7MoN2P3
840.95
153(2)
0.71073
monoclinic
P2(1)/n
Synthesis of Cp*MoCl2(dppe) [6]. Cp*MoCl4 (0.382 g, 1.02 mmol)
was added to a THF solution (20 mL) of dppe (0.406 g, 1.02 mmol)
containing Na(sand) (0.053 g, 2.31 mmol) and napthalene (0.020 g,
0.16 mmol) giving a brown colored solution, which was stirred at room
temperature. Over a period of 12 h the color changed from brown to
red-brown. The THF was removed under reduced pressure and the
residue extracted into CH2Cl2 (15 mL). The CH2Cl2 solution was
13.7394(14)
20.421(2)
14.857(2)
99.119(8)
4115.8(8)
4
13.160(2)
10.942(2)
25.010(3)
95.225(10)
3586.4(9)
4
1
filtered and concentrated to /3 of its original volume. A brown solid
b, Å
c, Å
was precipitated by adding heptane (15 mL). The solid was dried
and isolated. Yield: 0.457g, 67%. The compound was recrystallized
by layering CH2Cl2 solution with heptane. Anal. Calcd (found) for
C36H39Cl2MoP2: C, 61.73 (61.69); H, 5.61 (5.55). EPR (THF): g )
1.987 (t; aMo ) 28.9 G, aP ) 27.2 G).
General Procedure for the Reaction of Carbonyl Cations with
PPN+Cl-. Approximately 0.040 g of [1]PF6, [2]PF6, and [3]PF6 was
dissolved in THF (5 mL). An excess (2 to 3 equiv) of PPN+Cl- was
added to each of the solutions. An immediate color change was
observed in all three solutions. The reactions were monitored by IR,
EPR, and NMR spectroscopies. For the NMR measurements, an aliquot
of the solution was transferred into a separate Schlenk tube and
evaporated to dryness, and the residue was redissolved in the suitable
deuterated solvent (see Results).
Β, deg
vol, Å3
Z
density (calcd),
1.469
1.557
mg/m3
absn coeff, mm-1
F(000)
0.562
1868
none
1704
independent reflcns 5367 [Rint ) 0.0572]
6310 [Rint ) 0.0368]
refinement method
full-matrix
full-matrix
least-squares on F2
least-squares on F2
6310/0/445
data/restraints/
parameters
goodness-of-fit
on F2
5366/0/495
1.057
1.098
final R indices
[I > 2σ(I)]
R
X-ray Crystallography. (a) [2]PF6‚THF. A reddish-orange crystal
with dimensions 0.58 × 0.38 × 0.05 mm was placed and optically
centered on the Enraf-Nonius CAD-4 diffractometer. The crystal cell
parameters and orientation matrix were determined from 25 reflections
in the range 33.5 < 2θ < 39.2°, and confirmed with axial photographs.
Three nearly orthogonal standard reflections were monitored at 1-h
intervals of X-ray exposure and showed insignificant variation in
intensity. Seven ψ-scan reflections were collected over the range 7.1
< θ < 13.1°; the absorption correction was applied with transmission
factors ranging from 0.2938 to 0.5577.22 The data were corrected for
Lorentz and polarization factors. The space group was uniquely
determined by systematic absences from the data. Direct methods
allowed the location of the heavy atoms, i.e., the Mo, Cl, P, and F
atoms, and fragments of the phenyl rings and carbonyl ligand. The
remaining non-hydrogen atoms were found from two subsequent
difference-Fourier maps. Although the molecule of interest and the
counterion (PF6) were located and refined relatively well, the area of
the unit cell involving the THF molecule was found to be a highly
randomized “ball” of electron density. Many difference maps and trial
orientations were tested and optimized with the final result consisting
in three different orientations, the sum of the occupancies being 1.
Hydrogen atoms attached to carbon atoms were placed in calculated
positions. All of the non-hydrogen atoms were refined anisotropically,
except for the disordered THF molecule. The selected crystal data are
reported in Table 1 and selected bond distances and angles are collected
in Table 2.
0.0459
0.1077 [4299 data]
1.114 and -1.112
0.0411
0.1181 [5964 data]
0.872 and -0.911
Rw
largest diff peak
and hole, e‚A-3
Table 2. Selected Bond Distances (Å) and Angles (deg) for
.
[2]PF6 THF
Mo(1)-C(1)
Mo(1)-Cl(1)
Mo(1)-P(1)
2.053(6)
2.473(2)
2.5555(13)
Mo(1)-P(2)
2.4902(13)
2.006(5)
Mo(1)-CNTa
C(1)-Mo(1)-Cl(1) 78.89(14) Cl(1)-Mo(1)-P(2)
127.67(5)
C(1)-Mo(1)-P(1) 133.13(14) CNTa-Mo(1)-C(1) 109.9(2)
C(1)-Mo(1)-P(2)
Cl(1)-Mo(1)-P(1)
P(1)-Mo(1)-P(2)
82.74(14) CNTa-Mo(1)-Cl(1) 113.1(2)
84.66(5)
73.38(4)
CNTa-Mo(1)-P(1) 116.9(2)
CNTa-Mo(1)-P(2) 119.2(2)
a CNT ) center of gravity of atoms C(10)-C(14).
Table 3. Selected Bond Distances (Å) and Angles (deg) for
.
[4]PF6 MeCN
Mo(1)-F(1)
Mo(1)-N(6)
Mo(1)-Cl(1)
1.951(2)
2.182(3)
2.4984(10)
Mo(1)-P(1)
Mo(1)-P(2)
Mo(1)-CNTa
2.5313(10)
2.5479(9)
1.990(4)
F(1)-Mo(1)-N(6)
F(1)-Mo(1)-Cl(1)
F(1)-Mo(1)-P(1)
F(1)-Mo(1)-P(2)
77.30(10) Cl(1)-Mo(1)-P(2)
80.41(6) P(1)-Mo(1)-P(2)
87.51(3)
81.33(3)
(b) [4]PF6‚MeCN. A reddish-orange crystal with dimensions 0.50
× 0.50 × 0.45 mm was placed and optically centered on the Enraf-
Nonius CAD-4 diffractometer. The crystal cell parameters and
orientation matrix were determined from 25 reflections in the range
23.8 < 2θ < 36.0° and confirmed with axial photographs. Six standard
reflections were monitored at 1-h intervals of X-ray exposure and
showed an insignificant variation of intensity. Four ψ-scan reflections
were collected over the range 7.8 < θ < 10.1°; the absorption correction
was not applied as the various reflections were not internally or
comparably consistent. The data were corrected for Lorentz and
polarization factors. The space group was uniquely determined by
systematic absences from the data. Direct methods allowed the location
of the Mo, Cl, P, and F atoms, and fragments of the phenyl rings. The
remaining non-hydrogen atoms were found from three subsequent
difference-Fourier maps. A molecule of solvation, MeCN, was also
found in the asymmetric unit and refined well. Hydrogen atoms
attached to carbon atoms were placed in calculated positions. All of
the non-hydrogen atoms were refined anisotropically. The selected
68.74(6) CNTa-Mo(1)-F(1) 174.1(1)
75.65(6) CNTa-Mo(1)-N(6) 102.3(1)
N(6)-Mo(1)-Cl(1) 82.87(8) CNTa-Mo(1)-Cl(1) 105.4(1)
N(6)-Mo(1)-P(1)
93.77(8) CNTa-Mo(1)-P(1) 105.5(1)
N(6)-Mo(1)-P(2) 152.43(8) CNTa-Mo(1)-P(2) 105.1(1)
Cl(1)-Mo(1)-P(1) 148.90(3)
a CNT ) center of gravity of atoms C(1)-C(5).
crystal data are reported in Table 1 and selected bond distances and
angles are collected in Table 3.
Results
Synthesis and Cyclic Voltammetric Studies of 1, 2, and
3. Compound 1 has been synthesized by addition of CO to the
recently reported stable 16-electron Cp*MoCl(PMe3)219 (see eq
1). The formation of 1 is fast and quantitative; work-up of the
product must be carried out immediately because if 1 is left
under a CO atmosphere, a much slower substitution of one of
the PMe3 ligands by CO occurs, as indicated by the growth of
(22) North, A. C. T.; Phillips, D. C.; Mathews, F. S. Acta Crystallogr.,
Sect. A 1968, A24, 351-359.

3620 J. Am. Chem. Soc., Vol. 118, No. 15, 1996
Fettinger et al.
CO stretching vibrations in the IR spectrum at 1946 and 1851
cm-1, these being tentatively assigned to compound Cp*MoCl-
(CO)2(PMe3). Another problem of this synthesis is that the
reduction of Cp*MoCl4/2PMe3 with Na proceeds through the
previously described23 Cp*MoCl2(PMe3)2 intermediate and the
last reduction step from Mo(III) to Mo(II) is rather slow, leaving
variable amounts of the less soluble Mo(III) intermediate in the
final carbonylated solution, which then co-crystallizes with the
desired compound 1. The amount of Mo(III) impurity can be
eliminated by utilizing a slight excess of Na and longer reaction
times; this procedure inevitably causes overreduction, lowering
the yield and introducing another impurity in the final solution,
namely Cp*MoH(PMe3)3, but the latter species has a higher
solubility and does not hamper the isolation of pure 1. The
reported somewhat low yield of 1 in the Experimental Section
(29%) is due to this overreduction procedure and to the
subsequent short contact time with CO, required to obtain the
ultra-pure material required for the studies presented here. The
complex has been isolated also several times in higher yields,
i.e. greater than 75%, but a 10-15% impurity of the Mo(III)
complex Cp*MoCl2(PMe3)2 is always present in those cases.
Compound 2 may also be synthesized by a procedure identical
to that described for 1, proceeding through the CO addition to
the 16-electron, Cp*MoCl(dppe), which is synthesized from
Cp*MoCl4, Na/Hg, and dppe in the appropriate stoichiometric
amounts.24 It was found, however, that the straight reduction
from Cp*MoCl4 under an atmosphere of CO improved the yield
and is a more convienent one-step synthesis (eq 2). Compound
3 has previously been synthesized by interaction of CpMoCl-
(CO)3 with dppe in refluxing toluene.21 We report here an
alternative preparation of 3 by reduction of CpMoCl2(dppe),
which is quantitatively generated in situ from {CpMoCl2}n and
dppe in CH2Cl2,20 with 1 equiv of Na/Hg under an atmosphere
of CO (eq 3). The IR, 1H, and 31P-NMR spectroscopic
properties are fully consistent with 18-electron, diamagnetic,
Mo(II) complexes for compounds 1, 2, and 3.
Figure 1. EPR spectra (X band) of [1]PF6, [2]PF6, and [3]PF6 (solvent
) THF, room temperature).
oxidation studies on 3 have not been described. It was, however,
stated that 3+ is less stable than the iodide analogue.25
Synthesis and Characterization of 1+, 2+, and 3+ by
Chemical Oxidation. The interaction of compounds 1, 2, and
3 with 1 equiv of Fc+PF6- in THF at room temperature affords
stable solutions of the corresponding cations, 1+, 2+, and 3+
-
(see eq 4). The cations can be isolated as PF6 salts without
decomposition by precipitation from THF with heptane, and
were characterized by chemical analysis and by IR and EPR
spectroscopies. The CO ligands offer a convenient marker for
the electron density on the metal center. The IR data for both
the neutral Mo(II) complexes and the carbonyl cations show
standard textbook trends, related to the different donating
abilities of PMe3 versus dppe and Cp* versus Cp. Upon
oxidation, blue shifts of 138 (1), 136 (2), and 150 cm-1 (3) for
νCO are observed as expected, in agreement with the removal
of electron density from the Mo center. The EPR data give
useful structural information (Figure 1). Complex 1+ shows a
binomial triplet indicating equivalent phosphines, whereas both
2+ and 3+ show a doublet of doublets, indicating the inequiva-
lence of the two phosphorus nuclei. Since all structurally
characterized CpMoIII complexes with four addtional mono-
dentate donors exhibit the ubiquitous 4-legged piano stool
structure,27 for instance [(ring)MoCl(PMe3)3)]+ (ring ) Cp,
Cp*),28 it is logical to propose that 1+, 2+, and 3+ also adopt
such a geometry. However, the two phosphorus donors are
located in relative trans positions in 1+, whereas they are
necessarily cis for the two complexes with the chelating dppe
ligand. The proposed structure for 2+ has been confirmed by
a single-crystal X-ray analysis (vide infra). The EPR parameters
for 1+ show a larger P (23.0 G) and a smaller Mo (26.3 G)
hyperfine coupling constant with respect to Cp*MoCl2(PMe3)2
(aP ) 15 G, aMo ) 36 G).23 Complexes 2+ and 3+ exhibit
spectral parameters similar to those of 1+: aMo ) 29.8 and 30.2;
average aP ) 22.9 and 22.6, respectively.
Cp*MoCl(PMe3)2 + CO f Cp*MoCl(PMe3)2(CO) (1)
Cp*MoCl4 + dppe + 3Na + CO f
Cp*MoCl(dppe)(CO) (2)
CpMoCl2(dppe) + Na + CO f CpMoCl(dppe)(CO)
(3)
All three compounds 1, 2, and 3 exhibit reversible oxidation
waves in the cyclic voltammogram at relatively low potentials,
E1/2 ) -0.48 V for 1 and -0.44 V for 2 vs the ferrocene
standard in THF. The reversible oxidation of 3 in CH2Cl2 had
already been reported by Gipson et al. at E1/2 ) -0.25 V on
the same scale.25,26 While it is clear that the carbonyl cations
are stable at least on the CV time scale, the question is whether
they are sufficiently stable to be isolated. Lau and Gipson
attempted bulk electrolysis of the iodide anologue of 3, CpMoI-
(dppe)(CO), in CH2Cl2 at low temperatures and generated the
corresponding cation, but found a reversion back to the neutral
Mo(II) parent complex upon warming the mixture to room
temperature. The oxidation of 3 in an IR spectrochemical cell
has been reported to give 3+, characterized by a stretching
vibration at 2011 cm-1, but bulk electrolysis or chemical
(ring)MoClL2(CO) + Cp2Fe+PF6- f
[(ring)MoClL2(CO)]+PF6- + Cp2Fe (4)
(ring ) Cp*, L2 ) (PMe3)2, dppe; ring ) Cp, L2 ) dppe)
Thermal Stability of 1+, 2+, and 3+. In order to probe the
stability of 1+, 2+, and 3+ , thermolysis in THF was performed
and monitored by IR spectroscopy. Complex 1+ shows the
greatest stability with less than 5% decomposition after 30 min
at 55 °C, and complex 2+ is only slightly less stable with less
than 10% decomposition after 30 min under the same condi-
tions. The least stable by far was 3+, which showed 75%
decomposition within 30 min at 55 °C. The decomposition of
3+ produced the neutral parent molecule, 3, which accumu-
(23) Baker, R. T.; Calabrese, J. C.; Harlow, R. L.; Williams, I. D.
Organometallics 1993, 12, 830-841.
(24) Full details of the preparation of this compound, which has also
been crystallographically characterized, will be reported in a subsequent
contribution.
(25) Lau, Y. Y.; Gipson, S. L. Inorg. Chim. Acta 1989, 157, 147-149.
(26) Lau, Y. Y.; Huckabee, W. W.; Gipson, S. L. Inorg. Chim. Acta
1990, 172, 41-44.
(27) Poli, R. J. Coord. Chem. B 1993, 29, 121-173.
(28) Fettinger, J. C.; Kraatz, H.-B.; Poli, R.; Rheingold, A. L. Acta
Crystallogr., Sect. C 1995, C51, 364-367.

Stable Mononuclear, 17-e Mo(III) CO Complexes
J. Am. Chem. Soc., Vol. 118, No. 15, 1996 3621
lated in solution, and a red precipitate, which was isolated,
crystalized from a MeCN/Et2O layering, and identified as
[CpMoClF(dppe)(MeCN)]+PF6-, [4]PF6. The NMR properties
of 4+ are fully consistent with an 18-electron Mo(IV) complex.
The 19F NMR showed two resonances in a 6:1 ratio, a doublet
at 9.47 ppm (JPF ) 706 Hz) for the PF6 anion, and a doublet of
doublets at 13.17 ppm (JPF ) 75 Hz, JPF ) 61 Hz), assigned to
the Mo-bound fluoride, split by the two inequivalent phosphorus
complex by this electron transfer process, this may reasonably
be expected to add Cl- to afford an 18-electron adduct
Cp*MoCl3(PMe3)2, and this product could further eliminate a
PMe3 ligand, generating the previously described30 neutral Mo-
(IV) complex, Cp*MoCl3(PMe3). This complex is paramagnetic
(S ) 1) and is characterized by a broad 1H-NMR resonance at
δ 2.3 ppm (w1/2) 184 Hz). The parallel study of the interaction
between [Cp*MoCl2(PMe3)2]+ and Cl- in stoichiometric amounts
indeed produced the expected resonances of free PMe3 (1H: δ
1.31; 31P: δ -61) for Cp*MoCl3(PMe3).
1
nuclei. The H NMR showed a triplet for the Cp resonance
corresponding to coupling to two similar phosphine ligands
which, in addition to the 19F data, would seem to indicate that
the two phosphorus donors occupy inequivalent equatorial sites
while the F occupies the axial site in a pseudo-octahedral
structure, the Cp ligand occupying the opposite axial site. This
assumption has been confirmed by a single-crystal X-ray
analysis (vide infra). In addition to being thermally unstable,
compound 3+ is also light sensitive, decomposing to 4+ after
an overnight exposure of a THF solution to fluorescent light.
Irradiation with a UV lamp caused a faster (<10 min)
decomposition to the same red precipitate obtained by the
thermal treatment. A possible route to compound [4]PF6 will
be proposed in the Discussion section.
With the final product being a complex of Mo(V), the
electrochemistry of Cp*MoCl3(PMe3) was investigated. This
compound shows an irreversible oxidation wave with an
anodic peak potential of Epa ) +0.22 V, i.e., too high to engage
in an electron transfer process with 1+. We then considered
that the addition of excess Cl- to afford an 18-electron
[Cp*MoCl4(PMe3)]- species, or an equilibrium with a small
amount of the bis-PMe3 adduct, may render the metal more
susceptible to oxidation, through the increase of electron density
at the metal. However, an equilibrium with a bis-phosphine
adduct is discounted because the addition of free PMe3 to
Cp*MoCl3(PMe3) causes no change either to the resonance of
the 16-electron Mo(IV) species or to the resonance of free PMe3,
and does not produce new resonances attributable to a new
diamagnetic species. The interaction between Cp*MoCl3(PMe3)
and excess Cl- was therefore investigated next. The addition
of excess PPN+Cl- to the CV cell containing Cp*MoCl3(PMe3)
did not change the cyclic voltammogram, except for a slight
decrease in the peak height for the oxidation of Cp*MoCl3-
(PMe3). This experiment indicates that, if an equilibrium with
an 18-electron [Cp*MoCl4(PMe3)]- is involved (eq 6), its
position must be shifted toward the left. The existence of such
an equilibrium, however, is demonstrated by the NMR and
electrochemical experiments descibed below. The interaction
between Cp*MoCl3(PMe3) with PPN+Cl- showed that the
characteristic resonance of the Mo complex slightly shifted to
ca. δ 1.9 (w1/2 ) 224 Hz). New diamagnetic resonances were
Reaction of 1+, 2+, and 3+ with PPN+Cl-. One of the
predicted characteristics of the 17-electron Mo(III) carbonyl
cationic complexes was the reactivity toward nucleophiles. The
expected outcome of the reaction with Cl- was the product of
CO substitution, to yield the known class of stable paramagnetic
(ring)MoCl2L2 compounds.27 These reactions, however, were
complicated by electron transfer processes, which have been
fully rationalized as shown below with the aid of IR, NMR,
and EPR monitoring and the parallel study of related chemical
and electrochemical processes.
Treatment of 1+ with excess Cl- caused an immediate
consumption of the molybdenum complex as shown by IR
spectroscopy. However, the IR spectrum also showed the
growth of an absorption corresponding to the νCO of the neutral
parent Mo(II) material, 1. 1H-NMR confirmed the identity of
1, but did not show any other NMR active product. The EPR
spectrum of the reaction mixture, however, revealed the forma-
tion of a paramagnetic material which exhibits a doublet with
Mo satellites (g ) 1.984, aMo ) 45 G, aP ) 28 G). This material
was shown to be the 17-electron Mo(V) complex, Cp*MoCl4-
(PMe3) (5), by comparison with the EPR spectrum of a genuine
sample prepared from Cp*MoCl4 and PMe3, and with the
spectrum reported in the literature.16 The nature of the observed
products demands the stoichiometry of eq 5.
1
not observable in the H- and 31P-NMR spectra. This result
suggests that any 18-electron [Cp*MoCl4(PMe3)]- species must
be in rapid exchange with the 16-electron neutral complex on
the NMR time scale. The same resonance at δ 1.9 is also
obtained from the interaction of the previously described31
[Cp*MoCl4]- with PMe3. In this latter experiment, no residual
peak due to [Cp*MoCl4]- (reported at -13.3 ppm),31 is
observed. Therefore, of the two available equilibria relating
the 18-electron [Cp*MoCl4(PMe3)]- complex with its possible
dissociation products (eqs 6 and 7), the first one allows both
3[Cp*MoCl(PMe3)2(CO)]+ + 3 Cl- f
2Cp*MoCl(PMe3)2(CO) + Cp*MoCl4(PMe3) +
CO + PMe3 (5)
Cp*MoCl3(PMe3) + Cl- h [Cp*MoCl4(PMe3)]- (6)
[Cp*MoCl4]- + PMe3 f [Cp*MoCl4(PMe3)]- (7)
The result of this reactivity originally appeared confusing, a
Mo(III) complex leading to a Mo(V) and a Mo(II) complex by
simply adding Cl-. In order to elucidate the reaction path, a
number of control experiments were carried out. Since the
reaction obviously involved electron transfer steps, the electro-
chemical properties of the possible intermediates was considered,
beginning with the expected Cp*MoCl2(PMe3)2 product of CO
substitution by Cl-. It was, as a matter of fact, already known
from previous investigations carried out in this laboratory that
Cp*MoCl2(PMe3)2 is a fairly good reducing agent (E1/2 ) -0.84
V)29 and is therefore capable of reducing 1+ to 1 (E1/2 ) -0.48
V). Following the formation of the [Cp*MoCl2(PMe3)2]+
species to be present at equilibrium, whereas the second one is
totally shifted to the right. Finally, an independent cyclic
voltammetric study of 5 showed an irreversible reduction with
a cathodic peak potential Epc ) -1.0 V, showing that the
primary reduction product, [Cp*MoCl4(PMe3)]-, immediately
decomposes. This decomposition leads to the formation of
Cp*MoCl3(PMe3) by Cl- dissociation, as clearly demonstrated
by the observation of an irreversible oxidation wave with Epa
) +0.22 V during the reverse scan, corresponding exactly to
the observed irreversible oxidation wave for Cp*MoCl3(PMe3),
(30) Abugideiri, F.; Gordon, J. C.; Poli, R.; Owens-Waltermire, B. E.;
Rheingold, A. L. Organometallics 1993, 12, 1575-1582.
(31) Abugideiri, F.; Brewer, G. A.; Desai, J. U.; Gordon, J. C.; Poli, R.
Inorg. Chem. 1994, 33, 3745-3751.
(29) Poli, R.; Owens, B. E.; Linck, R. G. Inorg. Chem. 1992, 31, 662-
667.

3622 J. Am. Chem. Soc., Vol. 118, No. 15, 1996
Fettinger et al.
broadened again upon further cooling to 193 K. Given the result
of the Cl- addition to 1+ (Vide supra), a likely possibility for
the EPR active product seems to be [CpMoCl3(dppe)]+, 7+. For
the 3+/Cl- system, the initial product of CO substitution by
Cl-, namely CpMoCl2(dppe) (E1/2 ) -0.33 V),27 is able to
reduce 3+ to 3 (E1/2 ) -0.25 V), being in turn oxidized to the
Mo(IV) complex [CpMoCl2(dppe)]+. Unlike the corresponding
Cp*/dppe system, [CpMoCl2(dppe)]+ can add another Cl- to
give the known neutral 18-electron, 7, as already reported in
the literature.21 We have further confirmed this by generating
[CpMoCl2(dppe)]+ from CpMoCl2(dppe) and Cp2FePF6, fol-
lowed by addition of PPNCl. Identical 1H and 31P-NMR
resonances with those reported21 for 7 were immediately
produced. Next, a cyclic voltammetric investigation of com-
pound 7 was carried out, which showed two reversible oxida-
tions at E1/2 ) -0.39 and -0.02 V. Since two different isomers,
eq-eq and ax-eq, are present in slow equilibrium for this
compound,21 it seems reasonable to assign each wave to the
oxidation of a different isomer. The one isomer showing a
reversible wave at -0.39 V is a sufficiently strong reductant to
take 3+ to 3, becoming oxdized to 7+ in the process, while the
other isomer remains a neutral Mo(IV) species. As stated above,
the 31P NMR of the final products of the Cl- addition showed
the singlet resonance of the eq-eq isomer of 7. Thus, the
oxidation wave at E1/2 ) -0.02 V can be assigned to the eq-
eq isomer, which remains unoxidized during the Cl- addition
process, while the wave at E1/2 ) -0.39 V is assigned to the
ax-eq isomer. This also implies that, given the reversibility
of the oxidation process, complex 7+ should also adopt a
pseudo-octahedral ax-eq configuration.
Figure 2. Cyclic voltammogram of Cp*MoCl4(PMe3).
see Figure 2. Thus, this cyclic voltammetric experiment
confirms that equilibrium 6 must be shifted largely to the left
and, in addition, it demonstrates the feasibility of the oxidation
of [Cp*MoCl4(PMe3)]- by 1+. In conclusion, electrochemical
and NMR studies give consistent results and both point to the
oxidation of [Cp*MoCl4(PMe3)]- by 1+ to generate the observed
Mo(V) product.
The reaction of 2+ with excess Cl- also proceeds immediately
and quantitatively as shown by IR and EPR spectroscopies. The
IR spectrum of the reaction mixture revealed once again the
formation of the parent Mo(II) material, 2, whose identity was
also confirmed by 1H- and 31P-NMR spectroscopies in CDCl3.
In contrast to the reaction of 1+ with Cl- described above, no
EPR active material was produced. Instead, a paramagnetically
shifted and broadened peak at -8 ppm was observed in the
1H-NMR spectrum, indicating the formation of another para-
magnetic, EPR-inactive material. Proceeding along similar lines
of thought as illustrated above for the corresponding reaction
of 1+, the electrochemistry of the expected product of CO
substitution, e.g. 6, was investigated, showing that the compound
is reversibly oxidized at E1/2 ) -0.58 V. Similar to the Cp*/
PMe3 system, 6 has sufficient reducting power to take the
starting material, 2+, to the parent Mo(II) complex, 2, producing
in turn the Mo(IV) complex [Cp*MoCl2(dppe)]+, 6+. A genuine
sample of 6+ was prepared by interacting 6 with 1 equiv of
Fc+PF6-, which showed a 1H-NMR spectrum in CDCl3 identical
In order to confirm the identity of the EPR active material
as 7+, genuine samples were generated by two different
pathways. The first one involved the interaction of CpMoCl4
with dppe in THF. There was an immediate reaction with the
precipitation of a light brown colored solid. This solid is
insoluble in THF but dissolves in CH2Cl2, suggesting the ionic
formulation [CpMoCl3(dppe)]+Cl- rather than the neutral adduct
CpMoCl4(dppe), and showed the same EPR signal as the product
of the Cl- addition to 3+. Compound 7+ was also prepared
without formulation ambiguities by chemical oxidation of 7 with
-
Cp2Fe+PF6 in THF . This product, unlike the corresponding
Cl- salt, was soluble in THF, and again showed the same EPR
signal as the product of the 3+/Cl- reaction, finally confirming
the stoichiometry of eq 9. All of the results of the Cl- addition
to 1+, 2+, and 3+ are summarized in Scheme 1.
to that of the Cl- addition reaction product.
A
1H-NMR
experiment also showed that 6+ does not interact with Cl-,
explaining the difference in final outcome between this system
and the 1+/Cl- system. Therefore, the reaction between 2+ and
Cl- is described by eq 8.
(3 + 2n)[CpMoCl(dppe)(CO)]+ + (3 + 2n)Cl- f
(2 + n)CpMoCl(dppe)(CO) + n eq,eq-CpMoCl3(dppe) +
ax,eq-[CpMoCl3(dppe)]+ + (1 + n)CO (9)
2[Cp*MoCl(dppe)(CO)]+ + Cl- f
Cp*MoCl(dppe)(CO) + [Cp*MoCl2(dppe)]+ + CO (8)
X-ray Analyses. An ORTEP diagram of 2+, the first
mononuclear Mo(III) carbonyl complex to be crystallographi-
cally characterized, is shown in Figure 3. The molecule exhibits
a typical four-legged piano-stool geometry with an η5-bound
Cp* ring and the dppe ligand in the expected cis configuration,
the other two positions being occupied by a chloride and a
carbonyl ligand. The molecule is structurally similar to other
known isoelectronic CpMoIII complexes not containing a
carbonyl ligand.27 The Mo-CO distance (2.053(6) Å) is
comparable with known diamagnetic Mo(III) dimers containing
CO ligands, i.e. [CpMo(CO)2(µ-S-t-Bu)]2(BF4) (average 2.016(16)
Å),11 [Cp2Mo2(CO)3(MeCN)(µ-SPh)2](BF4)2 (average 2.01(4)
Å),12 and [(CpMoCO)2(µ-SMe)(S2CH2)]OTf (average 2.068-
(12) Å).15 The Mo-Cl distance (2.473(2) Å) compares well
with those in other half-sandwich Mo(III) complexes, e.g. 2.489-
The reaction of 3+ with excess Cl- proceeds similarly to that
of 1+ in that the final products are a Mo(V) and a Mo(II)
material. Once again, the reaction with Cl- consumes all of
the 3+ and the IR confirms the production of the neutral parent
1
complex, 3. The H- and 31P-NMR spectra of the reaction
products in CDCl3 indicated two NMR active materials, one
being 3 and the other being eq,eq-CpMoCl3(dppe), 7, by
comparison to literature characterization.21 However, the EPR
spectrum of the reaction mixture in THF or CH2Cl2 at room
temperature shows a strong, slightly broadened singlet with Mo
satellites (g ) 1.948, aMo ) 52 G). This spectrum slightly
sharpened upon cooling to 213 K but not sufficiently to allow
the phosphorus hyperfine splitting to be observed, and then it

Stable Mononuclear, 17-e Mo(III) CO Complexes
J. Am. Chem. Soc., Vol. 118, No. 15, 1996 3623
Scheme 1
Figure 4. ORTEP view of the 4+ ion. Ellipsoids are drawn at the
40% probability level.
An ORTEP view of 4+ is shown in Figure 4. The molecule
exhibits a pseudo-octahedral geometry when the η5-Cp ring is
considered to occupy a single coordination position. The F atom
occupies the trans positions with respect to the ring, whereas
the Cl atom, the MeCN molecule, and the dppe ligand occupy
equatorial sites. All of the equatorial atoms are bent signifi-
cantly away from the Cp ring, with values of the CNT-Mo-
equatorial atom angles between 102.5° and 105.5°. This
geometrical characteristic is also found for other 18-electron
CpMoIV complexes, e.g. CpMoCl3(dppe),21 CpMoCl3[P(OCH2)3-
CEt]2,34 and CpMoCl3(PMe2Ph)2.30 The average Mo-P bond
length (2.515(1) Å) in the title complex corresponds well with
the Mo-Peq distance (2.521(2) Å) found in CpMoCl3(dppe)21
as well as the average distance (2.554(1) Å) found in
CpMoCl3(PMe2Ph)2.30 The Mo-Cl distance (2.4984(10) Å) is
similar to that found for the average Mo-Cleq (2.527(1) Å)
distance in CpMoCl3(PMe2Ph)2 and the average Mo-Cl distance
(2.476(3) Å) found for CpMoCl3(dppe). Very few MoIV-F
complexes have been crystallographically characterized. The
Mo-F distance in the title complex (1.951(2) Å) is almost
identical with that found in [MoOF(Ph2PCHdCHPPh2)2]+ (1.95-
(1) Å),35 which has the F trans to the oxo ligand. The
comparison seems reasonable due to the psuedo-octahedral
geometry of the two molecules as well as the presence of strong
trans-influencing ligands, Cp and O, in each of the compounds.
A strong trans influence for the Cp ligand has also been observed
in other compounds, e.g. Cp*WBr3(CO)2.36
(1) Å (average) in Cp*MoCl2(PMe3)2,23 2.471(3) Å (average)
in CpMoCl2(PMe3)2,32 and 2.427(9) Å in [CpMoCl(PMe3)3]+.33
The two Mo-P distances in [2]PF6 differ by over 0.05 Å, with
the P trans to the CO ligand having the longer distance, this
being a possible manifestation of the stronger trans influence
of the CO ligand compared to the Cl- ligand. The average
Mo-P distance (2.523 Å) is similar to the average Mo-P
distances (2.51 Å) found in Cp*MoCl2(PMe3)2,23 and slightly
longer than the Mo-P bonds in CpMoBr2(dppe), which average
2.460(4) Å.20 Another typical feature of the four-legged piano-
stool Mo(III) molecules is the distinct “slip” of the ring with
two of the C atoms distinctly closer than the other three, the
difference between the longest and the shortest Mo-C bond
being as much as 0.18 Å.27 The structure of [2]PF6 shows a
similar phonemenon, the largest ∆(Mo-C) being 0.13 Å.
Discussion
While the electrochemical oxidation of compound Cp*MoCl-
(CO)3 is irreversible and occurs at the high potential of +0.55
V vs Fc/Fc+, the isoelectronic complex Cp*MoCl(PMe3)3 is
oxidized reversibly at -1.32 V27 (a stabilization of almost 2
V!) and the oxidation product, the 17-electron [Cp*MoCl-
(PMe3)3]+, has been isolated and crystallographically character-
ized.28 We therefore reasoned that the mixed-ligand system
Cp*MoCl(CO)(PMe3)2, 1, might undergo a reversible oxidation
and generate a sufficiently stable Mo-CO compound. The
results shown in this contribution show that our expectations
have been fulfilled. The first stable 17-electron paramagnetic
(32) Krueger, S. T.; Poli, R.; Rheingold, A. L.; Staley, D. L. Inorg. Chem.
1989, 28, 4599-4607.
(33) Abugideiri, F.; Kelland, M. A.; Poli, R.; Rheingold, A. L. Orga-
nometallics 1992, 11, 1303-1311.
(34) Poli, R.; Kelland, M. A. J. Organomet. Chem. 1991, 419, 127-
136.
(35) Cotton, F. A.; Eglin, J. L.; Wiesinger, K. J. Inorg. Chim. Acta 1992,
195, 11-23.
(36) Blackburn, H.; Kraatz, H.-B.; Poli, R.; Torralba, R. C. Polyhedron
1995, 14, 2225-2230.
Figure 3. ORTEP view of the 2+ ion. Ellipsoids are drawn at the
40% probability level.

3624 J. Am. Chem. Soc., Vol. 118, No. 15, 1996
Fettinger et al.
Mo(III) carbonyl complexes have been synthesized, spectro-
scopically and structurally characterized, and their thermal
stability and reactivity toward Cl- have been studied. Com-
Scheme 2
plexes 1+, 2+, and 3+ are easily prepared by chemical oxidation
-
of the neutral parent molecules with 1 equiv of Fc+PF6
.
Thermal decomposition studies indicate that the stability of the
carbonyl cation is directly related to the strength of the Mo-
CO bond (i.e. the lower the νCO, the more stable the cation).
Since weak Mo-CO bonds are expected for relatively high
oxidation state metals, it seems reasonable to suggest that the
electronic nature of the ancillary ligands helps stabilize these
cationic complexes. The stability increase on going from 3+
to 2+ could be attributed to both the greater donor power and
greater steric protection of the Cp* vs the Cp ligand.
One interesting aspect of the carbonyl cations is their
reactivity toward nucleophiles. All three carbonyl cations react
instantaneously with Cl-, but the nature of the end products
depends on the possibilities for the addition of more than one
Cl- as well as potential electron transfer processes. Scheme 1
shows that the pathways for the production of three different
types of complexes (5, 6+, and 7+) are related to each other.
As is shown, the first step always involves the attack of the
Cl- to the Mo center, replacing the CO ligand, and generating
a neutral Mo(III) complex. A rapid electron transfer reaction
occurs, involving the starting carbonyl cation, producing a Mo-
(IV) cation (a) and a neutral Mo(II) carbonyl complex. At this
point the 2+ system stops due to the inability to add another
Cl- by either forming a very crowded 18-electron trichloride
bis-phosphine complex or overcoming the chelate effect of the
dppe ligand. The 1+ system does not have the barrier of the
chelate effect to overcome and can easily displace a PMe3 ligand
in exchange for an additional Cl-, producing the 16-electron
complex, Cp*MoCl3(PMe3) (b). The followup chemistry has
been fully rationalized with the independent studies presented
in the Results section, and involves the ultimate formation of
compound 5. In the case of 3+, the smaller size of the Cp ligand
with respect to Cp* allows a Cl- ion to bind (a) and form the
18-electron CpMoCl3(dppe) (c). Cyclic voltammetric studies
on CpMoCl3(dppe) showed that 7+ can be formed in the
presence of a fair oxidant, i.e. 3+. For the same reason invoked
to explain the stability of 6+ in the 2+ system, product 7+
remains a salt and does not add Cl- due to the chelate effect of
the dppe ligand and steric constraints. Thus, for all three
systems, the products of Cl- addition are the neutral parent
carbonyl complex and one of the higher oxidation state
complexes, 5, 6+, or 7+.
The first step of Scheme 1 deserves further discussion. This
is a ligand exchange on a 17-electron organometallic complex,
which is known to proceed in general by an associative
pathway.2 However, for Cp- and Cp*-containing Mo(III)
organometallics, we have recently demonstrated the possible
involvement of dissociative (e.g. 15-electron) intermediates.37
This occurs, however, in the phosphine exchange process on
the neutral CpMoCl2L2 complexes, i.e. a system in which the
incoming ligand is relatively bulky (a condition disfavoring an
associative path), the outgoing ligand is not strongly bound,
and the unsaturated intermediate can be stablized by π-donor
ancillary ligands and by the spin reorganization energy (both
the latter two conditions favoring a dissociative path). A spin
quartet ground state for the 15-electron CpMoCl2(PH3) model
compound has been recently calculated.38 However, the halide
exchange on the same system was shown to be associative.29
For the process under consideration here, the coulombic effect
of the opposite charges of complex and incoming nucleophile
provides a strong bias in favor of the associative mechanism. It
is, however, intriguing to consider the labilization of the CO
ligand to give a cationic 15-electron complex, [(ring)MoClL2]+.
This labilization could play an important role in the mechanism
of thermal decomposition, since no strong nucleophilic reagent
is present in that case. The formation of the same product by
either thermal or photochemical decomposition of 3+ would
seem to point toward a CO dissociation mechanism, since
irradiation is known to promote CO dissociation processes.39
At the same time, the much faster rate of the Cl- addition
reaction with respect to the thermal decomposition process
indicates that the former must follow an associative pathway,
because if this were a dissociative one, its rate would be
regulated by the rate of CO dissociation and, consequently,
should not be faster than the thermal decomposition process.
To date, no stable CpMoIII 15-electron complex has been
isolated. Given the highly unsaturated nature of these hypo-
thetical 15-electron complexes, their reactivity and tendency to
reestablish a coordinately saturated system is expected to be
quite high. This is manifested, during the decarbonylation of
-
3+, by the F- abstraction from PF6
. Scheme 2 shows a
possible pathway to the production of 4+. It is important to
remember that the neutral complex 3 is also observed among
the reaction products. The first step involves the formation of
the highly reactive 15-electron intermediate, which quickly
reacts with a molecule of PF6-, producing a neutral Mo(III)
complex, CpMoClF(dppe). At this point, the system may enter
the familiar electron transfer process seen in Scheme 1,
producing [CpMoClF(dppe)]+ as well as the neutral carbonyl
complex, 3. This process is expected because the effect of the
fluoride ligand in the coordination sphere should be that of
lowering the potential necessary for the oxidation of the 17-
electron CpMoXY(dppe) (X,Y ) halide). Although fluoride
complexes are currently not known and it is therefore not
possible to directly substantiate this hypothesis, it should be
observed that the oxidation of CpMoX2(dppe) becomes more
facile in the order I < Br < Cl, i.e. the lighter halide ligand
increases the tendency of the metal to be oxidized.20 This trend
has been attributed to the better ability of the lighter halide to
(37) Cole, A. A.; Fettinger, J. C.; Keogh, D. W.; Poli, R. Inorg. Chim.
Acta 1995, 240, 355-366.
(38) Cacelli, I.; Poli, R.; Rizzo, A. to be submitted for publication.
(39) Collman, J. P.; Hegedus, L. S. Principles and Applications of
Organotransition Metal Chemistry; University Science Books: Mill Valley,
CA, 1980.

Stable Mononuclear, 17-e Mo(III) CO Complexes
J. Am. Chem. Soc., Vol. 118, No. 15, 1996 3625
stabilize the higher oxidization state via π-donation. After
formation of [CpMoClF(dppe)]+ via electron transfer, the
reaction ceases presumably because of the insolubility of the
complex in THF as well as the unavailability of additional X-.
The insoluble red precipitate obtained from the thermal decom-
position reaction indeed analyzes correctly for [CpMoClF(dppe)]-
[PF6] (see Experimental Section). The addition of a molecule
of MeCN occurs during the crystalization to finally produce
4+.
that further engage in redox processes with the unreacted Mo-
(III)-carbonyl compounds to ultimately lead to a variety of
different products, whose nature and mechanism of formation
has been fully elucidated.
Acknowledgment. This work was supported by the donors
of the Petroleum Research Fund (Grant 25184-AC3-C). Ad-
ditional support from the Alfred P. Sloan Foundation and the
National Science Foundation through awards to R.P. is also
grately acknowledged. We would also like to thank Dr. Yui
Fai Lam for his help with the 19F-NMR experiments. The
upgrade of the EPR instrument was made possible in part by
an NSF shared-equipment grant (CHE-9225064).
Conclusion
We have reported in this contribution that mononuclear,
paramagnetic Mo(III)-carbonyl compounds can be stabilized
by strongly donating and sterically protecting phosphine and
Cp ligands. Their structure and EPR properties are similar to
those of an extensive class of half-sandwich Mo(III) complexes.
The strength of the Mo-CO has been probed by thermal
decompositions, indicating that the bond cleavage occurs more
easily for the complex with the weakest CO bond. The results
of the thermal and photochemical decomposition of 3+ suggest
the formation of an extremely reactive 15-electron intermediate
that is able to abstract a fluoride ion from PF6-. Attack of the
17-electron carbonyl complexes by Cl- probably occurs, on the
other hand, associatively to afford products of CO substitution
Supporting Information Available: X-ray structural de-
termination of [2]PF6‚THF and [4]PF6‚MeCN: complete crys-
tallographic data, atomic coordinates, and full tables of distances
and angle and anisotropic thermal parameters (30 pages). This
material is contained in many libraries on microfiche, im-
mediately follows this article in the microfilm version of the
journal, can be ordered from the ACS, and can be downloaded
from the Internet; see any current masthead page for ordering
information and Internet access instructions.
JA9540738