Accessibility of 17-electron structures for cyclopentadienylchromium(III) compounds. 1. Experimental studies on the dichloride and dimethyl compounds

Article abstract of DOI:10.1021/om960330l

Adducts of the CpCrX₂ (X = Cl, CH₃) fragments with the bidentate ligands Me₂PCH₂PMe₂ (dmpm), Me₂PCH₂CH₂PMe₂ (dmpe), and Ph₂PCH₂CH₂PPh₂ (dppe) are described. Depending on the amount of ligand used, dinuclear 2:1 adducts or mononuclear 1:1 adducts are obtained. The dinuclear adducts have a CpX₂Cr(μ-η¹:η¹-L-L)CrX₂Cp structure featuring 15-electron Cr centers, as shown by X-ray structural determinations on [CpCrCl₂]₂(dppe), [CpCrCl₂]₂(dmpe), and [CpCr(CH₃)₂]₂(dmpe). The reaction between [CpCrCl₂]₂(L-L) and L-L affords the mononuclear CpCrCl₂(L-L) derivatives. A ³¹P-NMR monitoring of this reaction for L-L = dmpm reveals a quantitative reaction, followed by release of free dmpm. This phenomenon is interpreted as a dmpm-induced isomerization of [CpCrCl₂]₂(dmpm) to [CpCrCl(dmpm)]⁺[CpCrCl₃]^-. The mononuclear adducts have a 15-electron configuration CpCrX₂(η¹-L-L) with a dangling phosphine ligand, both in the solid state, as shown by an X-ray crystal structure of CpCrCl₂(dmpm), and in solution, as shown by ³¹P-NMR, UV-visible, and EPR spectroscopies. The hypothetical 17-electron configuration where the bidentate L-L ligand adopts a chelating configuration, CpCrX₂(η²-L-L), is not experimentally observed for any of the systems investigated.

Full text of DOI:10.1021/om960330l

+
+
Organometallics 1996, 15, 4211-4222
4211
Accessibility of 17-Electr on Str u ctu r es for
Cyclop en ta d ien ylch r om iu m (III) Com p ou n d s. 1.
Exp er im en ta l Stu d ies on th e Dich lor id e a n d
Dim eth yl Com p ou n d s
J ames C. Fettinger,1a Sundeep P. Mattamana,1a Rinaldo Poli,*,1a and
Robin D. Rogers1b
Department of Chemistry and Biochemistry, University of Maryland,
College Park, Maryland 20742, and Department of Chemistry,
Northern Illinois University, DeKalb, Illinois 60115
Received May 3, 1996X
Adducts of the CpCrX2 (X ) Cl, CH3) fragments with the bidentate ligands Me2PCH2-
PMe2 (dmpm), Me2PCH2CH2PMe2 (dmpe), and Ph2PCH2CH2PPh2 (dppe) are described.
Depending on the amount of ligand used, dinuclear 2:1 adducts or mononuclear 1:1 adducts
are obtained. The dinuclear adducts have a CpX2Cr(µ-η1:η1-L-L)CrX2Cp structure featuring
15-electron Cr centers, as shown by X-ray structural determinations on [CpCrCl2]2(dppe),
[CpCrCl2]2(dmpe), and [CpCr(CH3)2]2(dmpe). The reaction between [CpCrCl2]2(L-L) and L-L
affords the mononuclear CpCrCl2(L-L) derivatives. A 31P-NMR monitoring of this reaction
for L-L ) dmpm reveals a quantitative reaction, followed by release of free dmpm. This
phenomenon is interpreted as a dmpm-induced isomerization of [CpCrCl2]2(dmpm) to
[CpCrCl(dmpm)]+[CpCrCl3]-. The mononuclear adducts have a 15-electron configuration
CpCrX2(η1-L-L) with a dangling phosphine ligand, both in the solid state, as shown by an
X-ray crystal structure of CpCrCl2(dmpm), and in solution, as shown by 31P-NMR, UV-
visible, and EPR spectroscopies. The hypothetical 17-electron configuration where the
bidentate L-L ligand adopts a chelating configuration, CpCrX2(η2-L-L), is not experimentally
observed for any of the systems investigated.
In tr od u ction
electron, four-legged piano stool structure (I; Chart 1).4
Three-legged piano stool Mo(III) compounds (II), isolobal
with the large number of octahedral Werner-type com-
plexes, are unknown. On the other hand, cyclopenta-
dienylchromium(III) compounds are almost invariably
15-electron, three-legged piano stool systems with a spin
quartet ground-state configuration,5-8 just like octahe-
dral Werner-type complexes. It is fair to state that Cr-
(III) has a greater tendency to behave as a Werner-type
system, even with soft organic ligands, whereas the
properties of the Mo(III) center are modified by the soft
ligand environment toward the more typical behavior
of low-valent organometallics (i.e. tendency to reach an
electronically saturated configuration). This phenom-
enon is best exemplified by the comparison of the
structures of [CpqMX2]2 (M ) Cr, Mo; Cpq ) variety of
η5-C5R4R′ ligands; X ) Cl, Br, I) systems. For M ) Mo,
the structure features four bridging halides and a
metal-metal bond to achieve a saturated 18-electron
configuration (III),9-12 whereas the structure of Cr
systems has only two bridging halides and no metal-
metal bond giving rise to two antiferromagnetically
In classical (Werner-type) coordination chemistry, the
behavior of Cr(III) and Mo(III) is fairly similar. In each
case, the overwhelming tendency is to form octahedral
complexes.2 In the valence-electron formalism, these
are 15-electron systems, 12 electrons being donated by
the ligand set and 3 additional electrons residing in the
metal-based t2g set of orbitals in a parallel fashion, to
give rise to a spin quartet ground state. In spite of its
larger size with respect to Cr3+, the Mo3+ ion does not
have any tendency to increase its coordination number
in order to reach a more saturated electronic configu-
ration when being surrounded by the classical ligands
that characterize Werner-type coordination chemistry
(e.g. halides, N-donor ligands, ethers, tertiary phos-
phines, etc.).2 The only example of higher coordination
for Mo(III) appears to be [Mo(CN)7]4- 3
.
When softer carbon-based ligands are present, on the
other hand, the coordination chemistry of Mo(III) is
profoundly different from the corresponding chemistry
of Cr(III). Mo(III) gives rise to an extensive series of
organometallic complexes based on the half-sandwich
motif, the metal being bonded to one cyclopentadienyl
ring and four additional ligands, resulting in a 17-
(4) Poli, R. J . Coord. Chem. B 1993, 29, 121-173.
(5) Fischer, E. O.; Ulm, K.; Kuzel, P. Z. Anorg. Allg. Chem. 1963,
319, 253-265.
(6) Grohmann, A.; Ko¨hler, F. H.; Mu¨ller, G.; Zeh, H. Chem. Ber.
1989, 122, 897-899.
(7) Theopold, K. H. Acc. Chem. Res. 1990, 23, 263-270.
(8) Thomas, B. J .; Noh, S. K.; Schulte, G. K.; Sendlinger, S. C.;
Theopold, K. H. J . Am. Chem. Soc. 1991, 113, 893-902.
(9) Grebenik, P. D.; Green, M. L. H.; Izquierdo, A.; Mtetwa, V. S.
B.; Prout, K. J . Chem. Soc., Dalton Trans. 1987, 9-19.
(10) Fromm, K.; Hey-Hawkins, E. Z. Anorg. Allg. Chem. 1993, 619,
261-270.
X Abstract published in Advance ACS Abstracts, September 1, 1996.
(1) (a) University of Maryland. (b) Northern Illinois University.
(2) Wilkinson, G.; Gillard, R. D.; McCleverty, J . A. Comprehensive
Coordination Chemistry; Pergamon Press: Oxford, U. K., 1988; Vol.
3.
(3) Rossman, G. R.; Tsay, F.-D.; Gray, H. B. Inorg. Chem. 1973, 12,
824-829.
(11) Shin, J . H.; Parkin, G. Polyhedron 1994, 13, 1489-1493.
S0276-7333(96)00330-5 CCC: $12 00 © 1996 American Chemical Society

+
+
4212 Organometallics, Vol. 15, No. 20, 1996
Fettinger et al.
Bruker ER200 spectrometer upgraded to ESP300 and equipped
with an X-band microwave generator. The elemental analyses
were carried out by M-H-W Laboratories, Phoenix, AZ. CrCl3-
(THF)3 was prepared as previously described.20 dppe, dmpm,
and dmpe (Strem) were used without further purification.
CpNa was prepared from freshly cracked CpH and sodium
sand in THF, isolated as a white CpNa‚xTHF solid by
precipitation with heptane from a concentrated THF solution,
and titrated with 0.1 M HCl to determine the effective
molecular weight. A 1.4 M solution of MeLi in heptane
(Aldrich) was used as received. CpCrCl2(PMe3) and CpCr-
(CH3)2(PMe3) were prepared according to the procedure de-
scribed in the literature.6
Ch a r t 1
P r ep a r a t ion of [Cp Cr Cl2]2(d p p e). To a solution of
CpCrCl2(THF) in 25 mL of THF, prepared in situ from CrCl3-
(THF)3 (375 mg, 1.00 mmol) and CpNa(THF)0.38 (124 mg, 1.07
mmol), was added dppe (180 mg, 0.45 mmol) dissolved in 5
mL of THF at room temperature. The solution turned slightly
darker blue. After ca. 2 h of stirring the solution was
evaporated to dryness and the residue taken up in 25 mL of
CH2Cl2. The solution was then filtered and concentrated to
ca. 5 mL. The product was precipitated by addition of 20 mL
of heptane, filtered out, washed with heptane, and dried under
vacuum. Yield: 253 mg (69%). Anal. Calcd for C36H34Cr2-
Cl4P2: C, 55.84; H, 4.43. Found: C, 55.73; H, 5.09. 1H-NMR
(CDCl3, δ): 255 (br, w1/2 ) 5500 Hz, Cp), 9.2 and 7.2 (br, w1/2
) 104 and 52 Hz, Ph), -22 (br, CH2). UV/vis (CH2Cl2, nm (ꢀ,
M-1 cm-1)): λmax 660 (390), 482 (140). Crystals of this
compound were grown by layering a CH2Cl2 solution of
CpCrCl2(dppe) (see below) with heptane.
coupled S ) 3/2, 15-electron metal centers (IV).13-16 The
only 17-electron Cr(III) system that appears to be stable
(albeit only below room temperature) is CpCr(η3-
C3H5)2,17 where the metal is completely surrounded by
soft organic ligands. It could be argued that differences
in sterics (the smaller size of Cr3+ with respect to Mo3+
and the smaller size of the Cp ligand with respect to
the isoelectronic XL2 set18 of ligands) and metal-ligand
bond strength (the M-L bond dissociation energy
generally increases upon descending a group of transi-
tion metals)19 are responsible for this variation of
structure and stability, but other indirect observations,
to be presented and discussed later in this paper, show
that these effects alone do not explain all facts.
In this paper, we present experimental studies carried
out on cyclopentadienyl chromium(III) systems with
bidentate phosphine ligands, CpCrX2(L-L) (X ) Cl,
CH3). These studies were spurred by the idea that the
added chelate effect may force the system to adopt a
17-electron configuration (I) akin that of Mo(III) ana-
logues. Our results show that the 15-electron configu-
ration is strongly preferred for Cr(III) even under these
circumstances.
P r ep a r a tion of [Cp Cr Cl2]2(d m p m ). To a solution of
CpCrCl2(THF) in 25 mL of THF, prepared in situ from CrCl3-
(THF)3 (260 mg, 0.69 mmol) and CpNa(THF)0.38 (90 mg, 0.78
mmol), was added dmpm (0.05 mL, 0.31 mmol) at room
temperature. The solution turned slightly darker blue. After
ca. 2 h of stirring the solution was evaporated to dryness and
the residue taken up in 25 mL of CH2Cl2. The solution was
then filtered and concentrated to ca. 5 mL. The product was
precipitated by addition of 20 mL of heptane, filtered out,
washed with heptane, and dried under vacuum. Yield: 113
mg (71.5%). Anal. Calcd for C15H24Cr2Cl4P2: C, 35.18; H, 4.72.
Found: C, 34.25; H, 4.33. 1H-NMR (CDCl3, δ): 253 (br, w1/2
) 3200 Hz, Cp), -27.8 and -33.4 (br, total w1/2 ) 1900 Hz,
PCH3 and PCH2). UV/vis (CH2Cl2, nm (ꢀ, M-1 cm-1)): λmax
639 (1020), 477 (275).
)
Exp er im en ta l Section
P r ep a r a tion of [Cp Cr Cl2]2(d m p e). CpCrCl2(THF) was
prepared in situ from CrCl3(THF)3 (450 mg, 1.20 mmol) and
CpNa(THF)0.38 (148 mg, 1.28 mmol) in 30 mL of THF. The
resulting blue solution was evaporated to dryness and ex-
tracted in CH2Cl2 (25 mL). To this solution was added dmpe
(90 µL, 0.54 mmol) at room temperature. The solution turned
slightly darker blue with formation of a blue precipitate within
5 min. After ca. 2 h of stirring, the solution was left standing
overnight without stirring, producing additional product as
blue needle-shaped crystals. The product was filtered out,
washed with heptane, and dried under vacuum. Yield: 121
mg (43.7%). Anal. Calcd for C16H26Cr2Cl4P2: C, 36.53; H, 4.98.
Found: C, 37.02; H, 4.67. 1H-NMR (CDCl3, δ): 262 (br, w1/2
) 2900 Hz, Cp), -32.1 (br, w1/2 ) 1100 Hz, PCH3 and PCH2).
UV/vis (CH2Cl2, nm (ꢀ, M-1 cm-1)): λmax ) 599 (1130), 472
(200). The needles obtained as described above were used for
the X-ray analysis.
P r ep a r a tion of Cp Cr Cl2(d p p e). To a solution of CpCrCl2-
(THF) in 25 mL of THF, prepared in situ from CrCl3(THF)3
(400 mg, 1.06 mmol) and CpNa(THF)0.38 (133 mg, 1.15 mmol),
was added dppe (450 mg, 1.12 mmol) dissolved in 10 mL of
THF at room temperature. The solution turned slightly darker
blue. After ca. 2 h of stirring the solution was evaporated to
dryness and the residue taken up in 25 mL of CH2Cl2. The
solution was then filtered and concentrated to ca. 5 mL. The
Gen er a l Meth od s. All operations were carried out under
an atmosphere of dinitrogen or argon with standard Schlenk-
line techniques. Solvents were purified and dried by conven-
tional methods and distilled under argon prior to use. FT-IR
spectra were recorded on a Perkin-Elmer 1800 spectropho-
tometer with KBr disks (Nujol mulls). NMR spectra were
obtained with Bruker WP200 and AF200 spectrometers; the
peak positions are reported with positive shifts downfield of
TMS as calculated from the residual solvent peaks (1H) or
downfield of external 85% H3PO4 (31P). For each 31P-NMR
spectrum, a sealed capillary containing H3PO4 was immersed
in the same NMR solvent used for the measurement and this
was used as the reference. EPR spectra were recorded on a
(12) Desai, J . U.; Gordon, J . C.; Kraatz, H.-B.; Lee, V. T.; Owens-
Waltermire, B. E.; Poli, R.; Rheingold, A. L.; White, C. B. Inorg. Chem.
1994, 33, 3752-3769.
(13) Ko¨hler, F. H.; de Cao, R.; Ackermann, K.; Sedlmair, J . Z.
Naturforsch. 1983, 38B, 1406.
(14) Ko¨hler, F. H.; Lachmann, J .; Mu¨ller, G.; Zeh, H.; Brunner, H.;
Pfauntsch, J .; Wachter, J . J . Organomet. Chem. 1989, 365, C15.
(15) Morse, D. B.; Rauchfuss, T. B.; Wilson, S. R. J . Am. Chem. Soc.
1988, 110, 8234-8235.
(16) Morse, D. B.; Rauchfuss, T. B.; Wilson, S. R. J . Am. Chem. Soc.
1990, 112, 1860.
(17) Angermund, K.; Do¨hring, A.; J olly, P. W.; Kru¨ger, C.; Roma˜o,
C. C. Organometallics 1986, 5, 1268-1269.
(18) Green, M. L. H. J . Organomet. Chem. 1995, 500, 127-148.
(19) Connor, J . A. Top. Curr. Chem. 1971, 71, 71-110.
(20) Herwig, W.; Zeiss, H. H. J . Org. Chem. 1958, 23, 1404.

+
+
Cyclopentadienylchromium(III) Compounds
Organometallics, Vol. 15, No. 20, 1996 4213
blue product was precipitated by addition of 20 mL of heptane,
filtered out, washed with heptane, and dried under vacuum.
Yield: 444 mg (71.5%). Anal. Calcd for C31H29CrCl2P2: C,
63.5; H, 4.9. Found: C, 61.1; H, 5.0. The low analysis is
attributed to inevitable contamination by the dimeric species,
[CpCrCl2]2(dppe) (see Results section). 1H-NMR (CDCl3, δ):
255 (br, w1/2 ) 4500 Hz, Cp), 9.2 and 7.2 (br, w1/2 ) 130 and
64 Hz, Ph), -22 (br, w1/2 ) 1800 Hz, CrPCH2). 31P-NMR
(CDCl3, δ): 83.8 (br, w1/2 ) 330 Hz, dangling P), -12.8 (free
dppe). UV/vis (CH2Cl2, nm): λmax ) 657, 486.
P r epar ation of CpCr Cl2(dm pm ). To a solution of CpCrCl2-
(THF) in 30 mL of THF, prepared in situ from CrCl3(THF)3
(550 mg, 1.46 mmol) and CpNa(THF)0.38 (190 mg, 1.65 mmol),
was added dmpm (0.250 mL, 1.58 mmol) at room temperature.
The solution turned slightly darker blue. After ca. 1 h of
stirring the solution was evaporated to dryness and the residue
taken up in 45 mL of toluene. The solution was then filtered,
concentrated to ca. 25 mL, and placed at -20 °C, yielding a
heterogeneous blue precipitate containing needle-shaped crys-
tals, one of which was used for the X-ray analysis. Yield: 214
mg (45.3%). Anal. Calcd for C10H19CrCl2P2: C, 37.0; H, 5.9.
Found: C, 35.7; H, 5.7. The low analysis is attributed to
inevitable contamination by the dimeric species, [CpCrCl2]2-
(dmpm) (see Results section), which also accounts for the
heterogeneity. 1H-NMR (CDCl3, δ): 253 (br, w1/2 ) 3200 Hz,
(br, w1/2 ) 1154 Hz, CH2), -38.0 (br, w1/2 ) 1730 Hz, CH2).
31P-NMR (CD3CN, δ): -142.8 (septet, J PF ) 704 Hz).
Rea ction of [Cp Cr Cl(d p p e)]+P F 6- w ith P P N+Cl-. Com-
pound [CpCrCl(dppe)]+PF6- (18 mg, 0.028 mmol) and PPN+Cl-
(16 mg, 0.028 mmol) were dissolved together in 1 mL of CD3-
CN. The resulting 31P-NMR spectrum showed the resonances
of CpCrCl2(dppe) and free dppe, in addition to those of the
-
PPN+ and PF6 ions (see Results section).
F or m a t ion of P P N+[Cp Cr Cl3]- a n d R ea ct ion w it h
d m p m . CpCrCl2(THF) was prepared in situ from CrCl3(THF)3
(150 mg, 0.40 mmol) and CpNa(THF)0.38 (46 mg, 0.40 mmol)
in 15 mL of THF. The blue solution was evaporated to
dryness, and the residue was extracted in CH2Cl2 (20 mL).
The resulting solution was filtered, and PPN+Cl- (230 mg, 0.40
mmol) was added. An aliquot of the solution was used for the
1H-NMR characterization: δ 216 (br, w1/2 ) 1900 Hz) in
acetone-d6 and 232 (br, w1/2 ) 3000) in CDCl3. To the
remaining solution was added dmpm (65 µL, 0.41 mmol),
which changed the color to darker blue, and a second aliquot
1
of the solution was inspected by H- and 31P-NMR in acetone-
d6, showing two resonances in the region which is character-
istic of the CpCrIII protons, i.e. at δ 249 and 216 in an
approximate 1:2 ratio.
1H-NMR Stu d y of th e Rea ction betw een Cp Cr Cl2-
(d m p m ) a n d P P N+Cl-. CpCrCl2(dmpm) (17 mg, 0.052 mmol)
and PPN+Cl- (25 mg, 0.043 mmol) were dissolved together in
1 mL of CD2Cl2. The 1H-NMR spectrum of the solution showed
two resonances in the region characteristic of CpCrIII protons,
at δ 245 and 225 in an approximate 2:1 ratio. The same
experiment was also run in acetone-d6, to yield resonances at
δ 249 and 216 in an approximate 1:2 ratio.
Cp), 1.5 (br, w1/2 ) 95 Hz, CrPCH2P(CH3)2), -34.7 (br, w1/2
)
800 Hz, CrPCH3 and CrPCH2). 31P-NMR (CDCl3, δ): -28.1
(br, w1/2 ) 260 Hz, dangling P), -54.8 (free dmpm). 31P-NMR
(C6D6, δ): -8.1 (br, w1/2 ) 220 Hz), -54.8 (free dmpm). UV/
vis (CH2Cl2, nm): λmax ) 630, 480. µeff ) 3.80 µB (by the NMR
method21 in CD3CN).
P r ep a r a tion of [Cp Cr Me2]2(d m p e). CpCrCl2(THF) was
prepared in situ from CrCl3(THF)3 (280 mg, 0.74 mmol) and
CpNa(THF)0.38 (91 mg, 0.79 mmol) in 25 mL of THF. To the
resulting blue solution was added dmpe (60 µL, 0.36 mmol),
yielding a blue precipitate. The mixture was then cooled to
-78 °C and treated with MeLi (1.0 mL of a 1.4 M heptane
solution, 1.4 mmol), following by slow warming to room
temperature. The mixture turned purple. After being stirred
at room temperature for 1 h, the mixture was evaporated to
dryness. The residue was extracted with toluene (20 mL), the
mixture was filtered, and the resulting solution was concen-
trated to ca. 5 mL and cooled to -20 °C, yielding 54 mg (35%)
P r ep a r a t ion of Cp Cr Cl2(d m p e). CpCrCl2(THF) was
prepared in situ from CrCl3(THF)3 (230 mg, 0.61 mmol) and
CpNa(THF)0.38 (75 mg, 0.65 mmol) in 25 mL of THF. The
resulting blue solution was evaporated to dryness, extracted
in CH2Cl2 (25 mL), and filtered. To this solution was added
dmpe (0.110 mL, 0.66 mmol) at room temperature. The
solution turned slightly darker blue. After ca. 2 h of stirring
the solution was then filtered and concentrated to ca. 3 mL.
The product was precipitated by addition of 15 mL of heptane,
filtered out, washed with heptane, and dried under vacuum.
Yield: 118 mg (57.3%). 1H-NMR (CDCl3, δ): 250 (br, w1/2
)
3200 Hz, Cp), 1.0 (br, w1/2 ) 40 Hz, free dmpe + CrPCH2CH2P-
(CH3)2) -33.2 (br, w1/2 ) 1100 Hz, CrPCH3 and CrPCH2). 31P-
NMR (CDCl3, δ): -25.1 (br, w1/2 ) 96 Hz, dangling P), -46.6
(free dmpe). UV/vis (CH2Cl2, nm): λmax ) 612, 474.
31P -NMR St u d y of t h e [Cp Cr Cl2]2(d m p m ) + d m p m
Rea ction . [CpCrCl2]2(dmpm) (14 mg, 0.027 mmol) was dis-
solved in 1 mL of CDCl3. This solution was then introduced
in a NMR tube. To the NMR tube was then added dmpm (4.3
µL, 0.027 mmol) and the reaction monitored by 31P-NMR (see
Results section).
P r ep a r a tion of [Cp Cr Cl(d p p e)]+P F 6-. To a solution of
CpCrCl2(dppe) (137 mg, 0.23 mmol) in 20 mL of THF was
added TlPF6 (80 mg, 0.23 mmol). A white precipitate (TlCl)
formed immediately, while the color of the solution became
blue-purple. After being stirred at room temperature for 5 h,
the solution was filtered, concentrated to ca. 10 mL, and cooled
to -20 °C. This resulted in the formation of purple colored
crystals, which were filtered off and dried under vacuum.
Yield: 24 mg (16.5%). Diffusion of heptane layer into the
mother solution gave additional blue solid to bring the total
yield to 36 mg (29%). The elemental analysis (C, H) for this
material gave unacceptably low results, which are attributed
to contamination by TlCl. In agreement with this hypothesis,
dissolution of the crystalline product in CH3CN leads to the
subsequent formation of small amounts of a white precipitate.
1H-NMR (CD3CN, δ): 314 (br, w1/2 ) 4615 Hz, Cp), 12.8, 10.2
(br, w1/2 ) 54 Hz, p-Ph), 9.9 (br, w1/2 ) 49 Hz, m-Ph), -15
of dark purple crystals. 1H-NMR (C6D6, δ): 188 (br, w1/2
)
2600 Hz, Cp), -17.1 (br, w1/2 ) 640 Hz, PCH3), -33.6 (br, w1/2
) ca. 1000, PCH2). No resonances were observed in the 31P-
NMR spectrum. Anal. Calcd for C20H38Cr2P2: C, 54.12; H,
8.56. Found: C, 53.78; H, 8.52. UV/vis (THF, nm): λmax
)
530, 390. A single crystal was used for the X-ray analysis.
F or m a tion of [Cp Cr Me2]2(d m p m ). By a procedure iden-
tical to that described above for the preparation of [CpCrCl2]2-
(dmpe), a purple solution of [CpCrMe2]2(dmpm) was prepared
from CrCl3(THF)3 (260 mg, 0.69 mmol), CpNa(THF)0.38 (89 mg,
0.77 mmol), dmpm (50 µL, 0.32 mmol), and MeLi (1.0 mL, 1.4
mmol). The residue after evaporation of the final THF solution
showed solubility in toluene and ether, but a solid material
could not be obtained from either solution. 1H-NMR (C6D6):
187 (br, w1/2 ) 2900 Hz, Cp), -15.7 (br, w1/2 ) 640 Hz, PCH3),
-31.7 (br, w1/2 ) 1000 Hz, PCH2). No resonances were
observed in the 31P-NMR spectrum. UV/vis (THF, nm): λmax
) 573, 387.
Attem p ts To Gen er a te a Solu tion of [Cp Cr Me2]2(d p p e).
By a procedure identical to that described above for the
preparation of [CpCrCl2]2(dmpe), the preparation of a solution
of [CpCrMe2]2(dppe) was attempted starting from CrCl3(THF)3
(150 mg, 0.40 mmol), CpNa(THF)0.38 (52 mg, 0.45 mmol), dppe
(75 mg, 0.19 mmol), and MeLi (0.60 mL, 0.84 mmol). The
resulting toluene solution, however, had a green rather than
purple color. The investigation of this material was aban-
doned.
F or m a t ion of Cp Cr Me2(L-L) (L-L ) d m p e, d m p m ,
d p p e). Solutions of these three complexes were obtained by
(21) Lo¨liger, J .; Scheffold, R. J . Chem. Educ. 1972, 49, 646-647.

+
+
4214 Organometallics, Vol. 15, No. 20, 1996
Fettinger et al.
Ta ble 1. Cr ysta llogr a p h ic Da ta for All Com p ou n d s
compound
[CpCrCl2]2(dmpe)
[CpCrCl2]2(dppe)
CpCrCl2(dmpm)
CpCrMe2(dmpe)
formula
fw
space group
a, Å
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
Z
C36H34Cl4Cr2P2
774.38
P21/n
9.7894(14)
14.974(2)
12.041(2)
90
105.937(14)
90
1697.2(4)
2
C16H26Cl4Cr2P2
526.11
P21/n
6.6354(7)
13.7755(14)
12.2663(12)
90
97.881(2)
90
C10H19Cl2CrP2
324.09
P1h
C20H38Cr2P2
444.44
P21/n
6.6168(6)
13.954(2)
12.4150(11)
90
90.046(8)
90
1146.3(2)
2
6.441(3)
9.995(3)
12.479(3)
72.03(2)
83.41(3)
79.82(3)
750.5(4)
2
1110.6(2)
2
T, K
λ, Å
153(2)
0.71073
1.515
173(2)
0.71073
1.573
153(2)
0.710 73
1.434
1.301
1.53
153(2)
0.710 73
1.288
1.088
1.23
F
calcd, g/cm3
µ, mm-1
1.076
1.601
T
max/Tmin
R(Fo)a (I > 2σ(I))
0.1294
0.2766
0.0564
0.1002
0.0505
0.1151
0.1498
0.3718
Rw(Fo2)b (I > 2σ(I))
a
b
2
2
2
R ) ∑||Fo| - |Fc||/∑|Fo|. Rw ) [∑w(|Fo| - |Fc| )2/∑w(|Fo| )2]1/2; w ) 1/σ2(|Fo|).
Ta ble 2. Selected Bon d Len gth s (Å) a n d An gles
(d eg) for [Cp Cr Cl2]2(d m p e)a
procedures identical with that described above for the prepa-
ration of [CpCrCl2]2(dmpe), except that twice the molar amount
of the diphosphine ligand was used. Crystalline products,
however, could not be isolated from these solutions. Cooling
a heptane solution of CpCrMe2(dmpm) to -80 °C did produce
a crystalline sample, but after the mother liquor was decanted
away at low temperature and the sample dried under vacuum,
the solid melted to a viscous oil upon warming to room
temperature. The solution of CpCrCl2(dppe) had a greenish
color, while the solutions of the other two complexes were
purple. All the three compounds, however, gave NMR spectra
consistent with the presence of the desired Cr(III) products
with a η1-coordination for the diphosphine ligand.
Cp Cr Me2(d m p e). 1H-NMR (C6D6, δ): 185 (br, w1/2 ) 2900
Hz, Cp), 1.2-0.8 (m, free dmpe + CrPCH2CH2P(CH3)2), -13.0
(br, w1/2 ) 650 Hz, CrPCH3), -34.0 (br, w1/2 ) 970 Hz,
CrPCH2). 31P-NMR (C6D6, δ): -6.5 (br, w1/2 ) 170 Hz,
dangling P). A resonance at δ -46.6 due to free dmpe is also
observed in the 31P-NMR. UV/vis (THF, nm): λmax ) 561, 390.
Cp Cr Me2(d m p m ). 1H-NMR (C6D6, δ): 185 (br, w1/2 ) 2500
Hz, Cp), 2.8 (br, w1/2 ) 160 Hz, CrPCH2P(CH3)2), -15.7 (br,
w1/2 ) 625 Hz, CrPCH3), -31.5 (br, w1/2 ) 950 Hz, CrPCH2).
Cr(1)-Cl(1)
Cr(1)-Cl(2)
2.298(2)
2.284(2)
Cr(1)-P(1)
Cr(1)-CNT
2.414(2)
1.884(3)
Cl(1)-Cr(1)-Cl(2)
Cl(1)-Cr(1)-P(1)
99.02(6) Cl(2)-Cr(1)-P(1)
93.01(6)
90.88(6) Cl(2)-Cr(1)-CNT 124.3
Cl(1)-Cr(1)-CNT 123.0
P(1)-Cr(1)-CNT
118.4
a
CNT is the Cp ring centroid.
determined the space group as the monoclinic P21/n (No. 14).
Direct methods (SHELXS-86) allowed the successful location
of the four heavy atoms in the asymmetric unit (Cr, two Cl
and P). The remaining non-hydrogen atoms were found from
two subsequent difference-Fourier maps. The structure was
refined with SHELXL-93. Hydrogen atoms were placed in
calculated positions and used for structure factor calculations
but not refined. The low diffracting power of the crystal and
the resulting low number of observed reflections allowed only
the Cr, Cl, and P atoms to be refined anisotropically. Crystal
data are assembled in Table 1. Other data are available as
Supporting Information.
[Cp Cr Cl2]2(d m p e). Crystals of this compound were too
small for measurement on a regular diffractometer. Therefore,
the structure of this compound was determined with a Siemens
SMART diffractometer equipped with a CCD area detector at
Northern Illinois University. A blue needle with dimensions
0.30 × 0.04 × 0.04 mm was mounted on the diffractometer,
and the intensity data were gathered at -100 °C. The space
group was determined to be the centric P21/n (No. 14) from
the systematic absences. A summary of data collection
parameters is given in Table 1. The geometrically constrained
hydrogen atoms were placed in calculated positions and
allowed to ride on the bonded atom with B ) 1.2Ueq(C). The
methyl hydrogen atoms were included as a rigid group with
31P-NMR (C6D6, δ): 109 (br, w1/2 ) 640 Hz, dangling P).
A
resonance at δ -54.8 due to free dmpm is also observed in the
31P-NMR. UV/vis (THF, nm): λmax ) 600, 387.
Cp Cr Me2(d p p e). 1H-NMR (C6D6, δ): 186 (br, w1/2 ) 3000
Hz, Cp), -15.0 (br, w1/2 ) 900 Hz, CrPCH2). 31P-NMR (C6D6,
δ): 103 (br, w1/2 ) 500 Hz, dangling P). A resonance at δ -12.8
due to free dppe is also observed in the 31P-NMR.
Th er m a l Tr ea tm en t of Cp Cr Me2(d m p e) in THF a n d
Tolu en e. A THF solution of CpCrMe2(dmpe), prepared in situ
as described above, was heated to the reflux temperature for
10 h. No color change was noticed during this period, and a
1H- and 31P-NMR investigation of the final solution exhibited
only the resonances of the unreacted starting compound. A
thermal treatment in refluxing toluene slowly deposited a
brown insoluble precipitate, which did not dissolve in CH2-
Cl2, and whose nature was not further investigated.
rotational freedom at the bonded carbon atom (B ) 1.2Ueq
-
(C)). Refinement of non-hydrogen atoms was carried out with
anisotropic temperature factors. Selected bond distances and
angles are reported in Table 2.
X-r a y Cr ysta llogr a p h y. [Cp Cr Cl2]2(d p p e). A blue crys-
tal with dimensions 0.40 × 0.15 × 0.10 mm was placed and
optically centered on the Enraf-Nonius CAD-4 diffractometer.
The cell parameters and crystal orientation matrix were
determined from 25 reflections in the range 9.2 < 2θ < 40.6°
and were confirmed with axial photographs. Data collection
and reduction and the structure solution and refinement were
routine. No correction for decays was necessary, while the
Cp Cr Cl2(η1-d m p m ). A blue needle with dimensions 0.15
× 0.15 × 0.45 mm was placed on the Enraf-Nonius CAD-4
diffractometer. The cell parameters and crystal orientation
matrix were determined from 25 reflections in the range 15.6
< 2θ < 21.8° and confirmed with axial photographs. Data
collection and reduction and the structure solution and refine-
ment were routine. No correction for decays was necessary,
while the data were corrected for absorption by the ψ-scan
method (Tmax/Tmin ) 1.53). The averaging of equivalent
reflections gave Rint ) 0.0644. Intensity statistics clearly
favored the centrosymmetric space group P1h (No. 2). Direct
data were corrected for absorption by the ψ-scan method (Tmax
/
Tmin ) 1.42). The averaging of equivalent reflections gave Rint
) 0.0676. The systematic absences from the data uniquely

+
+
Cyclopentadienylchromium(III) Compounds
Organometallics, Vol. 15, No. 20, 1996 4215
Ta ble 3. Selected Bon d Len gth s (Å) a n d An gles
(d eg) for Cp Cr Cl2(d m p m )a
carbons produce a substantial amount of the dimer (eq
2) as verified by H-NMR.
1
The room-temperature 1H-NMR spectrum (in CDCl3)
shows a broad and paramagnetically shifted resonance
centered at δ 268 (w1/2 ) 3400 Hz), which is assigned
to the Cp protons in the mononuclear CpCrCl2(THF)
product. Although the NMR of this compound has not
previously been reported to the best of our knowledge,
similar resonances have been assigned to the Cp protons
in analogous compounds (e.g. δ 231 for CpCrCl2(PMe3)
and CpCrCl2(PEt3),6 233.6 for [CpCrCl3]-, and 247.0 for
CpCrCl2(py)13). In addition, resonances due to the
Cr(1)-Cl(1)
Cr(1)-Cl(2)
2.281(2)
2.295(2)
Cr(1)-P(1)
Cr(1)-CNT
2.410(2)
1.882(7)
Cl(1)-Cr(1)-Cl(2) 100.08(8) Cl(2)-Cr(1)-P(1)
88.59(7)
Cl(1)-Cr(1)-P(1)
92.80(7) Cl(2)-Cr(1)-CNT 124.8(2)
Cl(1)-Cr(1)-CNT 122.5(2)
P(1)-Cr(1)-CNT
119.3(2)
a
CNT is the Cp ring centroid.
methods allowed the location of the Cr, two Cl, and two P
atoms. Subsequent difference-Fourier maps revealed the
location of all of the remaining non-hydrogen atoms. Hydrogen
atoms were placed in calculated positions and used for
structure factor calculations but not refined. Crystal data are
assembled in Table 1, and selected bond distances and angles
are reported in Table 3.
coordinated THF ligand are evident at δ 37.5 (w1/2
)
1000 Hz) for the R protons and 11.8 (w1/2 ) 280 Hz) for
the â protons (see Table 4). The latter assignment is
based on the expected larger contact shift and greater
broadening for the protons that are closer to the
paramagnetic metal center. Similar shifts for the THF
R and â protons have been reported for other d3, 15-
electron complexes.22,23 The same spectrum shows also
a less shifted Cp resonance at δ 158 (w1/2 ) 2400 Hz),
assigned to the antiferromagnetic dimer [CpCrCl2]2
[Cp Cr (CH3)2]2(d m p e). A black crystal with dimensions
0.40 × 0.15 × 0.05 mm was placed on the Enraf-Nonius CAD-4
diffractometer. The cell parameters and crystal orientation
matrix were determined from 25 reflections in the range 16.1
< 2θ < 20.1° and confirmed with axial photographs. Data
collection and reduction and the structure solution and refine-
ment were routine. No correction for decays was necessary,
while the data were corrected for absorption on the basis of
ψ-scan reflections (Tmax/Tmin ) 1.23). The averaging of equiva-
lent reflections gave Rint ) 0.0785. The systematic absences
from the data clearly determined the nonstandard centrosym-
metric monoclinic space group P21/n (No. 14). Direct methods
allowed the location of the Cr and P atoms. The remaining
non-hydrogen atoms were found from two subsequent differ-
ence-Fourier maps. Hydrogen atoms were placed in calculated
positions and used for structure factor calculations but not
refined. A final difference-Fourier map included many peaks
with the largest near the two heavy atoms and as large as
2.13 e‚Å-3; it is clearly apparent that one or more other
moieties are partially occupying the same space as the
molecule of interest, but their elucidation proved futile with
many minor peaks and no clear solution as to what the moiety-
(ies) should have been. Possibilities for the nature of these
other low-occupancy components are one or more different
orientations of the molecule of interest and/or solvent mol-
ecules. A figure showing the spacial relationship between the
residual peaks and the main molecule is provided in the
Supporting Information. Crystal data for the [CpCr(CH3)2]2-
(dmpe) molecule are assembled in Table 1. All other data are
available as Supporting Information.
1
(reported at ca. δ 150 in the literature13). The H-NMR
of the product after several cycles of vacuum drying and
redissolution in CDCl3 showed the increase of the dimer
resonance at δ 158 at the expenses of the resonances
assigned to the mononuclear THF adduct, whereas the
addition of THF to the NMR solution resulted in the
opposite change. We observed only one Cp resonance
for the dimeric [CpCrCl2]2 compound, while two differ-
ent resonances, assigned to cis and trans isomers, have
previously been reported.13
A2. Din u clea r [Cp Cr Cl2]2(L-L) Ad d u cts. The ad-
dition of the bidentate ligands Ph2PCH2CH2PPh2 (dppe),
Me2PCH2CH2PMe2 (dmpe), and Me2PCH2PPMe2 (dmpm)
to the THF solution of CpCrCl2(THF) in a 1:2 ratio leads
to the formation of compounds that analyze as Cp2Cr2-
Cl4(L-L) (eq 3).
2CpCrCl2(THF) + L-L f Cp2Cr2Cl4(L-L) + 2THF
(3)
L-L ) dppe, dmpe, dmpm
The products are fairly soluble in CH2Cl2 for L-L )
dmpm and dppe and are recovered as powders by
addition of heptane. The dmpe complex is only spar-
ingly soluble in CH2Cl2 and precipitates directly from
solution. The 1H-NMR spectra for all derivatives (Table
4) show broad, shifted resonances for the Cp protons
around δ 250, similar to those found for other CpCrIII
complexes. This indicates that these phosphine deriva-
tives have a three-legged piano stool coordination
geometry and a spin quartet ground state as has been
established for other 15-electron CpCrIII derivatives, for
instance [CpCrX3]Li (X ) Cl, Me, Ph).24 The aliphatic
phosphine proton resonances (all dmpm and dmpe
protons and the methylene protons of dppe) are upfield
shifted to the δ -20 to -35 region [cf. δ -33.3 for
CpCrCl2(PMe3)].6 No 31P-NMR resonance is observed
for any of these compounds, presumably because of the
proximity of the phosphorus nucleus to the paramag-
netic metal center. The optical spectra are in accord
Resu lts
A. Ch lor id e System s. A1. Th e [Cp Cr Cl2]2 P r e-
cu r sor . Compounds [CpCrCl2]2 and CpCrCl2(THF)
have previously been obtained from [Cp2Cr][CpCrCl3]
by standing in chloroform or by dissolution in THF,
respectively,5,13 the ionic precursor being in turn ob-
tained from chromocene. THF can be easily removed
from the mononuclear complex to generate the dinuclear
one.5 We have generated these compound more conve-
niently by direct transmetalation of NaCp and CrCl3-
(THF)3 in THF (eq 1), by analogy to the more recent
CrCl3(THF)3 + NaCp f
CpCrCl2(THF) + NaCl + 2THF (1)
2CpCrCl2(THF) a [CpCrCl2]2 + 2THF
(2)
synthesis of the Cp* analogue.14 Although the direct
product of this reaction is the THF adduct, removal of
the THF solvent and dissolution in chlorinated hydro-
(22) Poli, R.; Mui, H. D. J . Am. Chem. Soc. 1990, 112, 2446-2448.
(23) Poli, R.; Gordon, J . C. Inorg. Chem. 1991, 30, 4550-4554.
(24) Arndt, P.; Kurras, E.; Otto, J . Z. Chem. 1983, 23, 443-445.

+
+
4216 Organometallics, Vol. 15, No. 20, 1996
Ta ble 4. NMR P r op er ties of Cp Cr III Com p ou n d sa
1H-NMR
other protons
Fettinger et al.
compd
Cp
31P-NMR
solvent
CpCrCl2(THF)
[CpCrCl2]2
268
158
255
262
253
255
250
253
188
187
186
185
185
THF, 37.5 (R), 11.8 (â)
CDCl3
CDCl3
CDCl3
CDCl3
CDCl3
CDCl3
CDCl3
CDCl3
C6D6
C6D6
C6D6
C6D6
C6D6
[CpCrCl2]2(dppe)
[CpCrCl2]2(dmpe)
[CpCrCl2]2(dmpm)
CpCrCl2(dppe)
CpCrCl2(dmpe)
CpCrCl2(dmpm)
[CpCrMe2]2(dmpe)
[CpCrMe2]2(dmpm)
CpCrMe2(dppe)
CpCrMe2(dmpe)
CpCrMe2(dmpm)
Ph, 9.2-7.2; PCH2, -22
PCH3 and PCH2, -32.1
PCH3, -27.8; PCH2, -33.4
Ph, 9.2-7.2; PCH2, -22
+83.8
-25.1
-28.1
PCH3 and PCH2, -33.2; PCH3,b ca. 1.0
PCH3 and PCH2, -34.7; PCH3,b 1.5
PCH3, -17.1; PCH2, -33.6
PCH3, -15.7; PCH2, -31.7
PCH2, -15.0
c
+103
-6.5
+109
PCH3, -13.0; PCH2, -34.0; PCH3,b ca. 1.0
PCH3, -15.7; PCH2, -31.5, PCH3,b 2.8
Solvent ) CDCl3; T ) 294 ( 1 K. Dangling group. c 31P-NMR: -8.2 in C6D6.
a
b
F igu r e 2. ORTEP view of compound [CpCrCl2]2(dmpe)
showing the labeling scheme used. The ellipsoids are drawn
at the 50% probability level.
Cp*CrMe2(PMe3) [2.426(2) Å]6 and in other trialkyl-
phosphine-containing, non-cyclopentadienyl-substituted
Cr(III) coordination compounds, e.g. [CrCl2(dmpe)2]BPh4
[2.445(5) Å].26 The structure of the dmpm complex as
obtained from the synthetic methodology described
above is assumed to be identical with those of the dppe
and dmpe analogues.
A3. Mon on u clea r Cp Cr Cl2(L-L) w ith Da n glin g
P h osp h in e. When the interaction between CpCrCl2-
(THF) and L-L was carried out in a 1:1 molar ratio,
solutions containing the mononuclear compounds
CpCrCl2(L-L) were obtained (eq 4). The same solutions
could also be obtained by addition of 1 equiv of L-L to
the isolated dinuclear [CpCrCl2]2(L-L) compounds.
F igu r e 1. ORTEP view of compound [CpCrCl2]2(dppe)
showing the labeling scheme used. The ellipsoids are drawn
at the 40% probability level, and hydrogen atoms are
omitted for clarity.
with the proposed geometry and spin state, the maxi-
mum absorptions being essentially superimposable with
those of CpCrCl2(PMe3) (vide infra). These spectra have
the same general shape as those reported for [CpCrCl3]-
and [CpCr(H2O)3]2+ 25
with the position of the absorp-
,
tion maxima being very close to those reported for the
trichloro complex (624 and 492 nm).
The molecular geometry is confirmed by a single-
crystal X-ray investigation on compounds [CpCrCl2]2-
(dppe) and [CpCrCl2]2(dmpe), whose ORTEP drawings
are shown in Figures 1 and 2. In each compound, each
chromium atom is bonded to a Cp ring, two chlorine
atoms, and one phosphorus atom of the bidentate ligand,
which serves as bridge between the two metal centers.
The crystals of the dppe compound had a very low
diffracting power, resulting in a rather unprecise struc-
tural determination. For this reason, no emphasis is
placed on the metric parameters for this compound, and
full results are reported and discussed only for the dmpe
compound (see Table 2). The average distances to the
chlorine atoms (2.291(7) Å) and Cp ring (center of
gravity, 1.884(3) Å) compare quite well with analogous
distances in compounds [CpCrCl2]2 [Cr-Cl(terminal),
2.274(4) Å; Cr-Cp, 1.867 Å]13 and [Cp*CrCl2]2 [Cr-Cl-
(terminal), 2.290(4) Å; Cr-Cp*, 1.88 Å],14 while the
Cr-P distance of 2.414(2) Å is similar to those found in
CpCrCl2(THF) + L-L f CpCrCl2(L-L) + THF
L-L ) dppe, dmpe, dmpm
(4)
The physical and spectroscopic properties of these
solutions are very similar to those deriving from the 2:1
1
interaction of eq 3. In particular, the H-NMR spectra
(Table 4) show the downfield contact shifted Cp reso-
nance and the upfield contact-shifted phosphine (ali-
phatic R proton) resonance that are characteristic of the
quartet state for the 15-electron CpCrX2L structure. The
spin state has been confirmed by a magnetic suscepti-
bility measurement (Evans’ method) on CpCrCl2(dmpm)
(µeff ) 3.80 µB). However, contrary to the above di-
nuclear complexes, the 31P-NMR spectrum of all these
mononuclear complexes is characterized by a broad,
(26) Salt, J . E.; Wilkinson, G.; Motevalli, M.; Hursthouse, M. B. J .
Chem. Soc., Dalton Trans. 1986, 1141-1154.
(25) Spreer, L. O.; Shah, I. Inorg. Chem. 1981, 20, 4025-4027.

+
+
Cyclopentadienylchromium(III) Compounds
Organometallics, Vol. 15, No. 20, 1996 4217
collected in Table 3. All distances and angles for this
compound are quite close to those discussed above for
the dinuclear [CpCrCl2]2(dmpe) structure (cf. Table 3
with Table 2).
The reaction between the isolated dinuclear com-
pound [CpCrCl2]2(dmpm) and 1 equiv of free dmpm in
CDCl3 (eq 6) was monitored by 31P-NMR in order to gain
[CpCrCl2]2(dmpm) + dmpm a 2CpCrCl2(dmpm)
(6)
further insight into the apparent equilibrium between
mononuclear and dinuclear phosphine adducts. The
results of this experiment, however, are unexpected and
rather puzzling. A representative run is illustrated in
Figure 4. The immediate recording of the spectrum
(within 1 min from the introduction of the phosphine
in the NMR tube) shows the expected formation of the
CpCrCl2(η1-dmpm) monomer as indicated by the broad
paramagnetically shifted 31P-NMR resonance for the
dangling phosporus atom and the complete disappear-
ance of the free phosphine resonance. Subsequently,
however, the free phosphine resonance reappears in the
spectrum and grows until it reaches a constant inten-
sity. The results of this experiment were reproduced
several times. This observation indicates that eq 6 is
quantitative but that the monomer product subse-
quently engages in a slower equilibrium which involves
release of free phosphine in solution.
The addition of the phosphine and all other operations
were conducted at constant temperature; thus, a change
of the equilibrium position for the hypothetical equilib-
rium between the dinuclear CpCl2Cr(µ-L-L)CrCl2Cp and
the mononuclear CpCrCl2(η1-L-L) as the experiment
progresses is not a possible explanation of this phenom-
enon. From the kinetic point of view, the occurrence of
a quantitative reaction followed by its coming back to
an observable equilibrium position is not acceptable
for a transformation as mechanistically simple as
the present one. The explanation of this phenomenon,
therefore, must involve the quantitative formation of
CpCrCl2(η1-dmpm) followed by a slower release of dmpm
to afford a new compound, i.e. different than the starting
dinuclear [CpCrCl2]2(dmpm). Incidentally, this behav-
ior is consistent with the observation of the instability
of CpCrCl2(L-L) toward loss of diphosphine (vide supra).
Since no new phosphorus resonance other than those
of free dmpm and CpCrCl2(η1-dmpm) appears in the
spectrum, any new compound must have either no
coordinated phosphine or a chelating dmpm ligand (i.e.
both P donors bonded to the paramagnetic metal center).
A reasonable hypothesis for this equilibrium is indicated
in eq 7. The relative intensity of the free phosphine and
F igu r e 3. ORTEP view of compound CpCrCl2(η1-dmpm)
showing the labeling scheme used. The ellipsoids are drawn
at the 40% probability level.
contact-shifted resonance (see Table 4), which is as-
signed to the dangling phosphorus nucleus. These
spectra also invariably exhibit the sharp resonance of
free L-L, even when the ligand L-L had been added in
substoichiometric amount. Solid samples of crude
CpCrCl2(L-L), obtained by vacuum drying the solutions
and washed with heptane in order to eliminate any
unreacted free L-L, gave again, upon dissolution in
CDCl3, a 31P-NMR spectrum exhibiting the broad reso-
nance of the dangling ligand and the sharper resonance
due to free L-L. This phenomenon can only indicate the
presence of a rapid equilibrium (on the time scale of the
solid dissolution and the recording of the NMR spec-
trum) between the mononuclear compound on one side
and the dinuclear compound and free ligand on the
other side; see eq 5. The observation of separate
2CpCrCl2(L-L) a Cp2Cr2Cl4(L-L) + L-L
(5)
resonances for the dangling phosphorus and free ligand
shows, on the other hand, that the ligand exchange is
slow on the NMR time scale. Equilibrium (5) prevented
us from obtaining analytically pure solid samples of the
mononuclear compound, and indeed, the analytical
results on the isolated mononuclear complexes always
show low C and H contents, consistent with the presence
of substantial amounts of dinuclear material. The
methyl protons of the dangling dmpm and dmpe ligands
show a broad peak at δ 1.0-1.5, which is the same
region as the protons of free phosphine, indicating that
the 5- and 6-bond distances of these protons from the
paramagnetic Cr center allow the transmission of a
negligible amount of spin density onto these nuclei. The
resonances due to the aliphatic â protons of the dangling
dmpe and dppe ligands could not be identified unam-
biguously.
We were fortunate to obtain single crystals of complex
CpCrCl2(dmpm) by low-temperature crystallization, and
these were investigated by X-ray crystallography. The
structure (Figure 3) shows the expected three-legged
piano stool and the dangling phosphine ligand. The
conformation chosen by the dangling phosphine is such
as to place the long arm (CH2PMe2) of the coordinated
phosphorus atom anti with respect to the Cp ring, e.g.
the conformation around the Cr-P bond can be de-
scribed as a staggered pseudo-ethane with the two
largest groups in relative anti positions. The same
conformation across the Cr-P bond is observed for the
dinuclear dppe and dmpe complexes (see Figures 1 and
2). Selected distances and angles for this compound are
2CpCrCl2(η1-dmpm) a
[CpCrCl(η2-dmpm)]+[CpCrCl3]- + dmpm (7)
dangling phosphine resonances at equilibrium (Figure
4) shows that equilibrium 7 is largely shifted to the left.
A concurrent 1H-NMR investigation did not help sub-
stantiate the hypothesis of equilibrium 7, because the
alleged ionic product is formed only in small amounts
1
and its H-NMR resonances would be broad and largely
overlapping with those of the CpCrCl2(η1-dmpm) com-
pound.

+
+
4218 Organometallics, Vol. 15, No. 20, 1996
Fettinger et al.
F igu r e 4. 31P-NMR study of the reaction between [CpCrCl2]2(dmpm) and dmpm. Solvent ) CDCl3, room temperature.
prepared and reacted with Cl- to regenerate the neutral
Sch em e 1
CpCrCl2(L-L) species. Attempts to prepare such com-
pound with L-L ) dmpm by eq 9 were not successful,
because of the interference of the formation of complexes
between the Tl+ cation and the phosphine ligand. The
use of AlCl3 as Lewis acid led to similar problems.
These results will be reported separately later.27 How-
ever, we were successful in the synthesis of the analo-
gous dppe complex. Reaction of CpCrCl2(dppe) with the
equimolar amount of TlPF6 in THF leads to precipita-
-
tion of TlCl and formation of [CpCrCl(dppe)]+PF6 (eq
9). A related ionic complex with a chelating diphosphine
ligand, [(η5-C5Me5)Cr(CH3)(dmpe)]+PF6-, has been pre-
An important implication of this hypothesis is that
viously prepared by the same strategy.8
among the two isomeric forms [CpCrCl2]2(µ-η1:η1-dmpm)
and [CpCrCl(η2-dmpm)]+[CpCrCl3]-, only the former
would be produced under kinetically controlled condi-
tions during the synthesis, while the latter could only
form through the catalytic intervention of free dmpm.
Indeed, the preparation of the neutral dmpm dimer was
carried out by using a defect amount of free phosphine.
A possible mechanistic pathway that leads from the
dinuclear [CpCrCl2]2(µ-L-L) structure to the isomeric
ionic form is depicted in Scheme 1. The rate-limiting
step during the isomerization reaction could be either
the intramolecular substitution of chloride by the
dangling phosphorus donor (step b) or the replacement
of L-L by free Cl- (step c). The position of the equilib-
rium for step b would also be expected to affect the rate
of isomerization if step c is rate limiting. When the
same experiment was carried out with a substoichio-
metric amount (0.75 equiv) of dmpm, the same results
were obtained, except that the free dmpm resonance
reached this time a smaller relative intensity.
The NMR spectroscopic properties of [CpCrCl-
-
(dppe)]+PF6 are as expected (see Experimental Sec-
tion). In particular, no 31P-NMR resonance other than
the typical septet feature of the PF6- anion is observed.
1
The Cp resonance in the H-NMR spectrum is further
shifted dowfield with respect to those of neutral CpCrCl2-
(PR3) complexes, i.e. to δ 314. Addition of the stoichio-
metric amount of [Ph3PNPPh3]+Cl- (PPN+Cl-) produces
the spectrum shown in Figure 5, featuring the reso-
nances of the dangling phosphorus of CpCrCl2(η1-dppe)
(obtained through equilibrium 8) and of free dppe
(supposedly obtained through an equilibrium process
identical to that shown for the dmpm system in eq 7),
in addition to the resonances of the PF6- anion and the
PPN+ cation. Therefore, it is established that equilib-
rium 8 can be shifted to the right (by Tl+) or to the left
(by Cl-).
The equilibrium of step c in Scheme 1 can be simpli-
fied as illustrated in eq 10. The occurrence of equilib-
The individual steps in Scheme 1 have been tested
stoichiometrically. The equilibrium of step b can be
simplified as illustrated in eq 8. A salt of [CpCrCl-
CpCrCl2(dmpm) + Cl- a [CpCrCl3]- + dmpm (10)
CpCrCl2(THF) + PPNCl f [PPN][CpCrCl3] + THF
CpCrCl2(L-L) a [CpCrCl(L-L)]+Cl-
(8)
(11)
rium (10) has been demonstrated by investigating the
reaction in both directions. The [PPN]+[CpCrCl3]-
compound was prepared by addition of PPN+Cl- to
CpCrCl2(L-L) + TlPF6 f
[CpCrCl(L-L)]+PF6- + TlCl (9)
(L-L)]+ with a noncoordinating anion (PF6-) has been
(27) Mattamana, S. P.; Poli, R. Unpublished results.

+
+
Cyclopentadienylchromium(III) Compounds
Organometallics, Vol. 15, No. 20, 1996 4219
F igu r e 6. ORTEP view of compound [CpCrMe2]2(dmpe)
showing the labeling scheme used. The ellipsoids are drawn
at the 40% probability level.
eluded our characterization attempts, and the dmpm
compound could not be obtained in the crystalline state.
A crystal structure of compound [CpCrMe2]2(dmpe) (see
Figure 6) shows the expected molecular dinuclear
geometry, identical to that already established for the
[CpCrCl2]2(L-L) (L-L ) dppe, dmpe) complexes (vide
supra). A severe disorder problem (see Experimental
Section) did not allow a satisfactory refinement of the
structure; therefore, the X-ray result is used only as a
verification of the molecular geometry.
These dinuclear compounds exhibit similar optical
and NMR properties, which are consistent with their
formulation as 15-electron compounds. Neither com-
pound shows a 31P-NMR signal, consistent with binding
of both phosphorus atoms to a paramagnetic Cr center.
The 1H-NMR resonance of the Cp hydrogen atoms
(Table 4) is found at δ 187-188, showing a substantial
decrease of spin density at the ring hydrogen nuclei on
going from the dichloride to the dimethyl derivatives.
The optical spectra show a similar pattern as the
corresponding dichloride compounds, with a blue shift
of the absorption bands by ca. 100 nm, which corre-
sponds to a change of color from blue for the dichloride
compounds to purple for the dimethyl analogues. This
is as expected, given the stronger ligand field of alkyl
ligands with respect to Cl.
F igu r e 5. 31P-NMR of the solution obtained by adding
[Ph3PNPPh3]Cl to [CpCrCl(dppe)]PF6. Solvent ) CD3CN,
room temperature.
CpCrCl2(THF) (which is in equilibrium with [CpCrCl2]2
as discussed above; see eq 2) in CH2Cl2; see eq 11. The
1
compound shows a broad H-NMR resonance at δ 216
in acetone-d6, 225 in CD2Cl2, and 232 in CDCl3, the
latter one comparing fairly well with the value of 233.6
reported for the corresponding chromocenium salt in
CDCl3.13
The reaction of either CpCrCl2(dmpm) with PPN+Cl-,
or [CpCrCl3]- with dmpm, in stoichiometric amounts
afford the same equilibrium mixture of CpCrCl2(dmpm)
and [CpCrCl3]- in approximately 1:2 ratio in acetone-
d6 or 2:1 ratio in CD2Cl2.
The result of the addition of dmpm to [CpCrCl2]2-
(dmpm) (Figure 4), initially carried out in CDCl3, is
therefore satisfactorily explained by the isomerization
process outlined in Scheme 1. Since the product of this
isomerization is ionic, we tested the same process in the
more polar solvents acetone (ꢀ ) 20.7) and acetonitrile
(ꢀ ) 37.5; vs ꢀ ) 4.806 for chloroform). The results were
essentially identical, featuring the initial disappearance
of the free phosphine resonance, followed by its recovery.
The final relative intensity of CpCrCl2(η1-dmpm) and
free dmpm resonances was essentially the same in the
three solvents. Analogous experiments on the dmpe
system were hampered by the low solubility of the
dinuclear starting material. Addition of 1 equiv of dmpe
to [CpCrCl2]2(µ-dmpe) caused the rapid dissolution of
the precipitate, but the phenomenon of disappearance
and subsequent recovery of the free phosphine reso-
nance was not observed. Equally, such a phenomenon
was not observed for the dppe system. However, in both
cases the final solution exhibited the resonance of
residual free phosphine, even when this was added in
a substoichiometric amount.
B2. Mon on u clea r [Cp Cr Me2]2(L-L) Com p ou n d s.
Solutions containing the mononuclear CpCrMe2(L-L)
complexes have been obtained by the same synthetic
procedure that affords the dinuclear derivatives de-
scribed above, except for the use of twice the amount of
the diphosphine ligand (eq 13). These products are
CpCrCl2(THF) + L-L + 2MeLi f
CpCrMe2(L-L) + 2LiCl + THF (13)
extremely soluble in hydrocarbons and could not be
crystallized as pure materials. Crystals of CpCrMe2-
(dmpm) were obtained at low temperature, but these
melted upon warming to room temperature. As for the
dinuclear systems described above, the dppe complex
appeared less stable, but its presence in the final
solution is nevertheless demonstrated by NMR spec-
troscopy (vide infra). In all these solutions, free phos-
phine is always observed. A possible equilibrium be-
tween the mononuclear and the dinuclear compound
(or an ionic isomer of the latter as for the chloride
system described above) seems thus indicated. In-
deed, the cationic and anionic alkyl complexes [CpCrMe-
B. Meth yl System s. B1. Din u clea r [Cp Cr Me2]2-
(L-L) Com p ou n d s. Treatment of the chloride com-
plexes Cp2Cr2Cl4(L-L) with 2 equiv of methyllithium
provided access, in each case, to the dinuclear methy-
lated derivatives [CpCrMe2]2(L-L), eq 12.
Cp2Cr2Cl4(L-L) + 4MeLi f
[CpCrMe2]2(L-L) + 4LiCl (12)
-
(dmpe)]+PF6 and Li[CpCrMe3] are known.8,24 How-
L-L ) dmpe, dmpm, dppe
ever, given the extreme air sensitivity of these com-
pounds, we did not pursue a detailed study of such equi-
librium process.
All these compounds are quite soluble in apolar
solvents. The dppe system is exceedingly sensitive and

+
+
4220 Organometallics, Vol. 15, No. 20, 1996
Fettinger et al.
The presence in solution of the molecular 15-electron
species CpCrMe2(η1-L-L) with a dangling phosphine
ligand is directly indicated by the broad 31P-NMR
resonance attributed to the uncoordinated P nucleus
1
(Table 4). In addition, H-NMR resonances attributed
to the dangling PMe2 groups are also observed for the
dmpm and dmpe compounds, these being sufficiently
distinct from the resonance of the free phosphine which
is also present. It is interesting to compare the 31P-
NMR paramagnetic shift experienced by the various
derivatives on going from free phosphine to the dangling
phosphine in the dichloride and dimethyl complexes.
dppe is shifted downfield from δ -12.8 to +83.8 upon
coordination to the “CpCrCl2” moiety, a shift of 95 ppm,
and a further downfield shift of 19 ppm is observed upon
replacing the two Cl with Me ligands. For the dmpe
derivatives, these shifts are 21.7 and 18.4 ppm, whereas
for the dmpm derivatives they are 26.7 and 137 ppm,
respectively. Thus, the chemical shift qualitatively
moves downfield in the series L-L < CpCrCl2(L-L) <
CpCrMe2(L-L) but, from the quantitative point of view,
the chemical shifting pattern is different for each system
for reasons that are presently unclear. We should also
observe that there is a considerable solvent dependence
of the dangling phosphorus chemical shift, this value
for CpCrCl2(dmpm) being -28.1 ppm in CDCl3 and -8.2
in C6D6. The other dichloride derivatives were not
sufficiently soluble in C6D6 to record a 31P-NMR spec-
trum, whereas the dimethyl derivatives decomposed in
CDCl3 to afford blue solutions of the corresponding
dichlorides. The optical spectra of solutions containing
the CpCrMe2(L-L) materials (vide infra) are essentially
superimposable with those of the corresponding di-
nuclear [CpCrMe2]2(L-L) complexes.
F igu r e 7. Visible spectra of the CpCrX2(L-L) species (X
) Cl, Me; L-L ) dppe, dmpe, dmpm) in comparison with
those of CpCrX2(PMe3).
plexes of Cr(III),29 the electronic relaxations are appar-
ently much faster for spin quartet half-sandwich com-
pounds, leading to EPR silent solutions at room
temperature.30 In fact, we find that compounds CpCrX2-
(PMe3) (X ) Cl, CH3) are EPR silent in isotropic solution
down to the freezing point of the toluene solvent.
Furthermore, addition of an excess PMe3 to these
solution does not generate any isotropic EPR signal. At
the liquid-N2 temperature, a typical EPR spectrum for
3
a S ) /2 d3 center is observed such as that reported30
for CpCrCl(CH3)(PMe3). A hypothetical 17-electron
structure, however, must necessarily have an orbitally
nondegenerate spin doublet ground state, for which
relatively sharp EPR signals are expected even at room
temperature, as have been reported for the correspond-
ing Mo(III) systems, CpMoCl2(PR3)2.31-33 Solutions of
any of the compounds CpCrX2(L-L) (X ) Cl, Me; L-L )
dppe, dmpe, dmpm) do not show an EPR signal down
to the freezing point of the solvent,34 therefore indicating
that the 17-electron structure CpCrX2(η2-L-L) is ener-
Compound CpCrMe2(dmpe) was found to be thermally
stable in refluxing THF (no noticeable decomposition
over 10 h). This result is in contrast with the observed
spontaneous reduction by alkyl radical elimination of
compound Cp*Cr(CH2Ph)2(THF) upon treatment with
the bidentate 2,2′-bipyridyl (bipy) ligand, to afford the
16-electron Cr(II) product Cp*Cr(CH2Ph)(bipy).28
C. Absen ce of 17-Electr on Sp ecies. We have
demonstrated the adoption of 15-electron structures for
getically inaccessible. The spectra at liquid-nitrogen
3
temperature are, again, consistent with S )
/
d3
2
systems (e.g. see the spectrum of CpCr(CH3)2(dmpe) in
Figure 8) and do not show signals in the g ) 2 region
that could be attributed to a S ) 1/2 species. Estimating
an EPR detection limit of 10-5 M and given the
concentrations used for the experiment (up to 5 × 10-2
M), we estimate an upper limit of 2 × 10-4 for the
molar fraction of CpCrX2(η2-L-L) in solution at 300 K.
This, in turn, sets a lower limit of 21.2 kJ (5.1 kcal) for
the free energy of CpCrX2(η2-L-L) relative to CpCrX2-
(η1-L-L).
1
compounds CpCrX2(L-L) (X ) Cl, Me) by H- and 31P-
NMR in solution and by an X-ray structural determi-
nation on CpCrCl2(dmpm) in the solid state. However,
the data discussed so far cannot rule out the presence
of a minor equilibrium amount of a 17-electron species
(e.g. CpCrX2(η2-L-L)). Such a 17-electron species should
display dramatically different optical properties with
respect to those characteristic of the 15-electron, spin
quartet geometry. The optical spectra of solutions of
the mononuclear CpCrX2(L-L) complexes (also contain-
ing all the other species in equilibrium with them for X
) Cl; vide supra), on the other hand, match very well
the spectrum of the corresponding CpCrX2(PMe3) (see
Figure 7). This indicates that if a 17-electron species
is present, its relative amount must be very small.
Direct evidence for a 17-electron species should be
provided by EPR spectroscopy. All the 15-electron
configurations for Cr(III) are characterized by a spin
quartet ground state. Although relatively sharp EPR
spectra are observed for Werner-type octahedral com-
Discu ssion
This work was spurred by the question of what
determines the structural and electronic preference of
(29) Wertz, J . E.; Bolton, J . R. Electron Spin Resonance. Elementary
Theory and Practical Applications; McGraw-Hill: New York, 1972.
(30) Barrera, J . A.; Wilcox, D. E. Inorg. Chem. 1992, 31, 1745-1752.
(31) Krueger, S. T.; Poli, R.; Rheingold, A. L.; Staley, D. L. Inorg.
Chem. 1989, 28, 4599-4607.
(32) Krueger, S. T.; Owens, B. E.; Poli, R. Inorg. Chem. 1990, 29,
2001-2006.
(33) Poli, R.; Owens, B. E.; Krueger, S. T.; Rheingold, A. L.
Polyhedron 1992, 11, 2301-2312.
(34) EPR signals due to 17-electron Cr(III) species have been
observed in our laboratory for other CpCrX2L2 systems (e.g. X ) CN).
These results will be reported in due course.
(28) Bhandari, G.; Kim, Y.; McFarland, J . M.; Rheingold, A. L.;
Theopold, K. H. Organometallics 1995, 14, 738-745.

+
+
Cyclopentadienylchromium(III) Compounds
Organometallics, Vol. 15, No. 20, 1996 4221
F igu r e 8. EPR spectrum of compound CpCr(CH3)2(dmpe)
at 77 K in a frozen toluene glass.
half-sandwich group 6 M(III) compounds: 15-electron,
3
1
S ) /2, for Cr and 17-electron, S ) /2, for Mo. The
paucity of 17-electron Cr(III) systems has also been
pointed out in a recent review article.18 A simple
answer to this question could be based on a combination
of steric and bond energy factors: Cr3+ is much smaller
than Mo3+, thus one could imagine that the interligand
repulsion in the 17-electron Cr system exceeds the M-L
bond strength. A similar situation was discussed for
the Cr-Cr bond formation in the [CpCr(CO)2(L)]2
dimers (L ) CO or P(OMe)3).35 In addition, the M-L
bond strength is expected to be smaller for Cr relative
to Mo.19 However, the situation must be more complex.
To illustrate why, we simply need to compare the
nonexistence of the 17-electron CpCrX2(PR3)2 systems
for either X ) Cl or CH3 with the existence and stability
of isosteric 16-electron CpVX2(PR3)2 (X ) Cl or alkyl).36,37
In fact, the vanadium systems are stable with sterically
more demanding phosphine ligands (e.g. PEt3) than
those with which Cr cannot form more than a three-
legged piano stool structure. For instance, CpCrCl2-
(PMe3) will not form a hypothetical 17-electron CpCrCl2-
(PMe3)2. Fourteen-electron complexes of type CpVX2L
have been reported only with sterically encumbering
ligands, e.g. CpV(CH2CMe3)2(PMe3).38 This comparison
demonstrates that Cr(III) has the steric ability to form
17-electron, four-legged piano stool structures. Another
interesting observations is that the half-sandwich V(II)
system CpVXL2 (X ) halide or alkyl, L ) tertiary
phosphine), isoelectronic with the Cr(III) system, also
exists as a spin quartet 15-electron species and has no
tendency to bind another ligand.
F igu r e 9. Qualitative orbital interaction diagram (top)
and reaction coordinate (bottom) for the reaction between
[M] and L ([M] ) 15-electron CpVXL2 or CpCrX2L (left) or
14-electron CpVX2L (right)).
provides only two electrons. The experimentally deter-
mined magnetic properties indicate a quartet ground
state for the V(II) and Cr(III) complexes and a spin
triplet ground state for the V(III) complexes. The
hypothetical interaction with an additional 2-electron
donor, resulting in the formation of a new metal-ligand
bond, requires the availability of an empty metal orbital.
Thus, the process will take place for the d3 metal centers
only at the expense of pairing two electrons to produce
a spin doublet product, while the ligand addition to the
d2 V(III) center does not require electron pairing (see
Figure 9). Hence, the formation of a new V(III)-L bond
to afford a 16-electron CpVX2L2 complex is only hin-
dered sterically, whereas the formation of a new V(II)-L
or Cr(III)-L bond to afford a 17-electron CpVXL3 or
CpCrX2L2 complex is a trade-off of the energetic cost of
pairing the electrons (plus the increase of steric repul-
sion) and the energetic gain of the new M-L bond. The
results already available in the literature indicate that
this trade-off favors the less saturated and higher spin
geometries for V(II) and Cr(III), whereas it favors the
more saturated, spin-paired geometry for Mo(III). The
well-documented stability of spin doublet, 17-electron
CpMoX2L2 systems4 finds a rationalization in the ex-
pected stronger Mo-L bonds and lower Mo pairing ener-
gy (the cost of pairing the electrons is less in the more
diffuse 4d orbitals of Mo with respect to the 3d orbitals
of Cr or V). These qualitative considerations are sup-
ported by preliminary theoretical calculations, which
will be presented in detail in a separate contribution.41
The results presented in this contribution show that
the thermodynamic help of the chelate effect is not
sufficient to tilt the balance in favor of a more saturated
A rationalization of the different structural preference
for half-sandwich compounds of Cr(III) and V(II) on one
hand, and V(III) on the other, may be found when
considering the energetic effect of electron pairing.39,40
For the three-legged piano stool structure, the three
valence orbitals that remain available after formation
of the metal-ligand σ bonds are occupied by three metal
electrons for the complexes of the d3 Cr(III) and V(II)
metal centers, while the corresponding d2 V(III) center
(35) Watkins, W. C.; J aeger, T.; Kidd, C. E.; Fortier, S.; Baird, M.
C.; Kiss, G.; Roper, G. C.; Hoff, C. D. J . Am. Chem. Soc. 1992, 114,
907-914.
(36) Nieman, J .; Teuben, J . H.; Huffman, J . C.; Caulton, K. G. J .
Organomet. Chem. 1983, 255, 193-204.
(37) Hessen, B.; Teuben, J . H.; Lemmen, T. H.; Huffman, J . C.;
Caulton, K. G. Organometallics 1985, 4, 946-948.
(38) Hessen, B.; Buijink, J .-K. F.; Meetsma, A.; Teuben, J . H.;
Helgesson, G.; Kakansson, M.; J agner, S.; Spek, A. L. Organometallics
1993, 12, 2268-2276.
(39) Poli, R. Comments Inorg. Chem. 1992, 12, 285-314.
(40) Poli, R. Chem. Rev., in press.
(41) Cacelli, I.; Poli, R.; Rizzo, A. To be published.

+
+
4222 Organometallics, Vol. 15, No. 20, 1996
Fettinger et al.
structure for the half-sandwich CpCrX2(L-L) (X ) Cl,
CH3) system. The formation of a chelate ring should
provide an ideal increase of 8.0 eu in translational
entropy with respect to the formation of the correspond-
ing structure with two enthalpically equivalent mono-
dentate ligands,42 which corresponds to a free energy
contribution (-T∆S) of 2.4 kcal/mol at 298 K. The
enthalpic equivalence is not necessarily guaranteed
because a hypothetical CpCrX2L2 (L ) monodentate
ligand) may adopt a trans geometry, while the CpCrX2-
(L-L) structure would presumably adopt a cis geometry
as found for the structurally characterized CpMoBr2-
(dppe).32 In addition, there is likely to be some strain
in the metallacycle, which should be minimal, however,
for the five-membered rings with the dmpe and dppe
ligands.43 These enthalpic effects oppose the entropic
help; thus, an overall free energy gain of 2.4 kcal/mol
is the maximum expected under ideal circumstances.
The sum of this chelate effect gain and the estimated
minimum energy of 5.1 kcal/mol for the 17-electron
CpCrX2(η2-L-L) with respect to the 15-electron CpCrX2-
(η1-L-L) indicates that the hypothetical 17-electron
CpCrX2(L)2 with X ) Cl or L and L ) phosphine ligand
could be 7.5 kcal/mol or more higher in energy than the
CpCrX2L + L system.
instead of adding another ligand and reaching a 17-
electron configuration as is the case of Mo(III), because
the cost of pairing the electron into the required spin
doublet configuration exceeds the energetic gain of
forming the new bond. The chelate effect of ligands such
as dmpe is not sufficient to overcome this unfavorable
energy differential. The stabilization of the less satu-
rated configuration by metal-ligand π bonding is not
sufficiently important for this system, because analo-
gous results were obtained for the dichloride and the
dimethyl systems. It is not unlikely, though, that 17-
1
electron half-sandwich complexes with a spin /2 ground
state may exist with other ligand systems34 or that they
could represent intermediates in organometallic reac-
tions such as ligand exchange. Experiments to test
these hypotheses are currently in progress.
Ack n ow led gm en t. We are grateful to the National
Science Foundation (Grant CHE-9508521) for support
of this work. The EPR spectrometer was upgraded in
part with NSF funds (Grant CHE-9225064).
Su p p or tin g In for m a tion Ava ila ble: For compounds
[CpCrCl2]2(dppe), [CpCrCl2]2(dmpe), CpCrCl2(dmpm), and
[CpCr(CH3)2]2(dmpe), tables of crystal data, fractional atomic
coordinates and U values, bond distances and angles, aniso-
tropic thermal parameters, and hydrogen atom coordinates
and U values (24 pages). Ordering information is given on
any current masthead page.
Con clu sion
We propose that half-sandwich Cr(III) compounds
prefer to adopt a 15-electron, spin quartet configuration
(42) Myers, R. T. Inorg. Chem. 1978, 17, 952-958.
(43) Li, C.; Cucullu, M. E.; McIntyre, R. A.; Stevens, E. D.; Nolan,
S. P. Organometallics 1994, 13, 3621-3627.
OM960330L