Competition between steric and electronic control of structure in Ru(CO)2L2L′ complexes

Article abstract of DOI:10.1021/om9700775

Magnesium reduction of RuCl₂(CO)₂L₂ in the presence of equimolar L in THF gives Ru(CO)₂L₃ (L = PPh₃ (1), PMePh₂ (2), PEt₃ (3), PⁱPr₂Me (4)). The corresponding reduction of RuCl₂(CO)₂(PEt₃)₂ in the presence of equimolar L′ (L′ = P(2-furyl)₃ (5) or AsPh₃ (6)) gives Ru(CO)₂(PEt₃)₂L′ but gives a mixture of Ru(CO)₂(PEt₃)_3-nL_n (n = 0-3) species when L′ = PPh₃. Comparisons show that 3 or 5 reacts slowly with L″ = (H)₂, CO, or PhC≡CPh to form Ru(CO)₂L″(PEt₃)₂ and free PEt₃ or P(2-furyl)₃ but rapidly with 4 or 6 to give the analogous products. The reaction of PhC≡CPh with Ru(CO)₂(PEt₃)₂L′ is faster for L′ = PEt₃ than for P(2-furyl)₃. All of these reactions are proposed to take place by preliminary ligand loss of L′, this being slower for 3 and 5 than for 1, 4, and 6. Reaction of O₂ with complexes containing the readily dissociated L′ species gives simply Ru(η²-O₂)(CO)₂L₂, but for Ru(CO)₂-(PEt₃)₃, this is accompanied by an apparent bimolecular electron transfer involving the intact complex to give Ru(CO)(CO₃)(PEt₃)₃. X-ray structure determinations of Ru(CO)₂(PEt₃)₃ (bis equatorial CO in trigonal bipyramid (TBP)), Ru(CO)₂(PⁱPr₂Me)₃ (two isomers: bis axial CO in TBP and also square pyramidal), and Ru(η²-PhCCPh)(CO)₂(PEt₃)₂ (cis carbonyls and trans phosphines) are reported. It is shown that all of the Ru(CO)₂L₃ species exist in solution as two isomers in rapid equilibrium. Ab initio MP2 calculations on the unhindered Ru(CO)₂-(PH₃)₃ model shows a preference for a trigonal bipyramidal structure with only a weak preference for CO to be at the equatorial site. It is shown that this pattern cannot be generalized to all π-acid ligands since ethylene is calculated to have a strong preference for an equatorial site in a TBP. Integrated quantum chemical and molecular mechanics calculations on Ru(CO)₂(PEt₃)₃ and Ru(CO)₂(PⁱPr₂Me)₃ give structures in excellent agreement with the X-ray results and confirm that the geometry and relative energetic preference for the observed structural isomers is strongly influenced, or even dominated by, the steric effect of the phosphine ligands.

Full text of DOI:10.1021/om9700775

Organometallics 1997, 16, 1979-1993
1979
Com p etition betw een Ster ic a n d Electr on ic Con tr ol of
Str u ctu r e in Ru (CO)2L2L′ Com p lexes
Masamichi Ogasawara,*,† Feliu Maseras,‡ Nuria Gallego-Planas,‡
Kazumori Kawamura,†,§ Kazuhiko Ito,†,§ Koichiro Toyota,†,§ William E. Streib,†
Sanshiro Komiya,†,| Odile Eisenstein,‡ and Kenneth G. Caulton†,
Department of Chemistry and Molecular Structure Center, Indiana University,
Bloomington, Indiana 47405-4001, and LSDSMS (UMR 5636), Case Courier 14, Baˆtiment 15,
Universite´ de Montpellier 2, 34095 Montpellier Cedex 5, France
Received February 3, 1997X
Magnesium reduction of RuCl2(CO)2L2 in the presence of equimolar L in THF gives Ru-
(CO)2L3 (L ) PPh3 (1), PMePh2 (2), PEt3 (3), PiPr2Me (4)). The corresponding reduction of
RuCl2(CO)2(PEt3)2 in the presence of equimolar L′ (L′ ) P(2-furyl)3 (5) or AsPh3 (6)) gives
Ru(CO)2(PEt3)2L′ but gives a mixture of Ru(CO)2(PEt3)3-nLn (n ) 0-3) species when L′ )
PPh3. Comparisons show that 3 or 5 reacts slowly with L′′ ) (H)2, CO, or PhCtCPh to
form Ru(CO)2L′′(PEt3)2 and free PEt3 or P(2-furyl)3 but rapidly with 4 or 6 to give the
analogous products. The reaction of PhCtCPh with Ru(CO)2(PEt3)2L′ is faster for L′ ) PEt3
than for P(2-furyl)3. All of these reactions are proposed to take place by preliminary ligand
loss of L′, this being slower for 3 and 5 than for 1, 4, and 6. Reaction of O2 with complexes
containing the readily dissociated L′ species gives simply Ru(η2-O2)(CO)2L2, but for Ru(CO)2-
(PEt3)3, this is accompanied by an apparent bimolecular electron transfer involving the intact
complex to give Ru(CO)(CO3)(PEt3)3. X-ray structure determinations of Ru(CO)2(PEt3)3 (bis
equatorial CO in trigonal bipyramid (TBP)), Ru(CO)2(PiPr2Me)3 (two isomers: bis axial CO
in TBP and also square pyramidal), and Ru(η2-PhCCPh)(CO)2(PEt3)2 (cis carbonyls and trans
phosphines) are reported. It is shown that all of the Ru(CO)2L3 species exist in solution as
two isomers in rapid equilibrium. Ab initio MP2 calculations on the unhindered Ru(CO)2-
(PH3)3 model shows a preference for a trigonal bipyramidal structure with only a weak
preference for CO to be at the equatorial site. It is shown that this pattern cannot be
generalized to all π-acid ligands since ethylene is calculated to have a strong preference for
an equatorial site in a TBP. Integrated quantum chemical and molecular mechanics
calculations on Ru(CO)2(PEt3)3 and Ru(CO)2(PiPr2Me)3 give structures in excellent agreement
with the X-ray results and confirm that the geometry and relative energetic preference for
the observed structural isomers is strongly influenced, or even dominated by, the steric effect
of the phosphine ligands.
In tr od u ction
RuHCl(CO)(PPh3)3 + AgClO4 + 2MeCN f
[RuH(CO)(MeCN)2(PPh3)2]ClO4 + AgCl + PPh3
Many Ru(CO)5 derivatives of the type Ru(CO)4L and
Ru(CO)3L2 have been reported, and they comprise the
core of zero-valent ruthenium chemistry, together with
clusters derived from Ru3(CO)12.1 In the 1970s, Roper
and co-workers reported the synthesis of Ru(CO)2-
(PPh3)3 by a multistep route (eq 1).2 The most valuable
feature of this complex is its unusual willingness to
react by loss of one PPh3 ligand to make a range of Ru-
(CO)2(PPh3)2Ln complexes containing unusual ligands
L (L ) CF2,3 S2,4 etc.). Roper suggested that this
[RuH(CO)(MeCN)2(PPh3)2]ClO4 + CO f
([RuH(CO)2(MeCN)(PPh3)2]ClO4)
V + PPh3
[RuH(CO)2(PPh3)3]ClO4
[RuH(CO)2(PPh3)3]ClO4 + tBuOK f
Ru(CO)2(PPh3)3 + KClO4 + tBuOH (1)
phosphine dissociation presumably occurred because
three bulky PPh3 ligands in one molecule are sterically
unfavorable. The implication was that this molecule
reacts by a dissociative mechanism, in spite of the rarity
(and thus presumed high energy) of isolable 16-electron
Ru(0) compounds. We reported recently the reduction
of cis,cis,trans-RuCl2(CO)2(PtBu2Me)2 with Mg in THF
to give unsaturated Ru(CO)2(PtBu2Me)2.5 This molecule
may be a model of the probable “high-energy intermedi-
† Indiana University.
‡ Universite´ de Montpellier.
§ Temporary visiting researchers at Indiana University (1995) on
leave from the Department of Chemistry, Hiroshima University, J apan.
| On sabbatical leave from the Department of Applied Chemistry,
Tokyo University of Agriculture and Technology, J apan.
E-mail: caulton@indiana.edu; eisenst@lsd.univ-montp2.fr.
X Abstract published in Advance ACS Abstracts, April 1, 1997.
(1) Bennett, M. A.; Bruce, M. I.; Matheson, T. W. Comprehensive
Organometallic Chemistry; Pergamon Press: New York, 1982; Vol. 4,
p 691.
(2) Cavit, B. E.; Grundy, K. R.; Roper, W. R. J . Chem. Soc., Chem.
Commun. 1972, 60.
(3) Clark, G. R.; Hoskins, S. V.; J ones, T. C.; Roper, W. R. J . Chem.
Soc., Chem. Commun. 1983, 719.
(4) Clark, G. R.; Russell, D. R.; Roper, W. R.; Walker, A. J .
Organomet. Chem. 1977, 136, C1.
S0276-7333(97)00077-0 CCC: $14.00 © 1997 American Chemical Society

1980 Organometallics, Vol. 16, No. 9, 1997
Ogasawara et al.
ate” from Roper’s complex. However, Ru(CO)2(PtBu2-
Me)2 shows unusual stability (i.e., persistence).5,6 The
complex Ru(CO)2(PtBu2Me)2 also represents the only
X-ray structurally characterized d8 four-coordinate com-
plex whose structure is not planar. Ab initio calcula-
tions clarified that the bent OC-Ru-CO unit in the
complex is stabilizing the molecule and that its struc-
ture is not due to the bulkiness of the PtBu2Me ligands
because calculations on the model compound Ru(CO)2-
(PH3)2 reproduced quantitatively the experimental angles
at the metal. Successful isolation of Ru(CO)2(PtBu2Me)2
as a persistent complex suggests the possibility that
dissociation of a phosphine ligand L from Ru(CO)2L3,
generating Ru(CO)2L2, is not so thermodynamically
unfavorable, even though the phosphine ligands L are
not sterically demanding.
In spite of its attractive reactivity, Ru(CO)2(PPh3)3
remains as the only example of an isolated Ru(CO)2L3
derivative.7 Presumably, some of the original reaction
steps are not applicable to other analogous complexes
with different phosphines due to its complicated syn-
thetic route. In this paper, we describe a very conve-
nient new synthesis of Roper’s complex Ru(CO)2(PPh3)3,
which has also enabled us to produce a range of Ru-
(CO)2L2L′ species involving mixed phosphine8 and phos-
phine/arsine complexes (L * L′) with a range of steric
and electronic properties. This achievement also gives
us an opportunity to clarify the prediction of accessible
Ru(CO)2L2 species mentioned above. Furthermore, in
the course of the investigation, we found an interesting
relationship between the structures of the complexes
and the steric properties of the phosphines. We report
here on our observations. Part of this work has been
reported in a preliminary communication.9
was a generous loan from J ohnson Matthey and used as
received. RuCl2(CO)2(PEt3)2,10 RuCl2(CO)2(PMePh2)2,11 and
12
RuCl2(CO)2(PPh3)2 were synthesized as reported. 1H (300
MHz) NMR spectra were recorded on a Varian XL300 spec-
trometer. 31P NMR spectra were obtained on a Nicolet NT-
360 spectrometer at 146 MHz or on a Varian XL300 spectrom-
eter at 122 MHz. 1H NMR chemical shifts are reported in ppm
downfield of tetramethylsilane, using residual solvent reso-
nances as internal standards. 31P NMR chemical shifts are
relative to an external standard, 85% H3PO4. Infrared spectra
were recorded on a Nicolet 510P FT-IR spectrometer. Elemen-
tal analyses were performed on a Perkin-Elmer 2400 CHNS/O
Elemental Analyzer at the Chemistry Department, Indiana
University.
cis,cis,tr a n s-Ru Cl2(CO)2(P iP r 2Me)2. Carbon monoxide
was passed through
a solution of ruthenium trichloride
hydrate (1.50 g, 6.5 mmol) in 2-methoxyethanol (35 mL) at
130 °C until the solution color changed to pale yellow. After
a small amount of insoluble material was filtered away, Pi-
Pr2Me (1.85 g, 14 mmol) was added and the solution was
refluxed for 10 min. The solution was concentrated to ca. 8
mL under reduced pressure and cooled to -20 °C, yielding two
crops of white crystals; yield 2.59 g (5.3 mmol, 81%). 1H NMR
(C6D6, 23 °C): δ 0.92 (dvt, J HP ) J HH ) 7.2 Hz, 12H, PCCH3),
1.20 (dvt, J HP ) J HH ) 7.2 Hz, 12H, PCCH3), 1.47 (vt, J HP
)
3.9 Hz, 6H, PCH3), 2.19 (m, 4H, PCHMe2). 31P{1H} NMR
(C6D6, 23 °C): δ 28.7 (s). IR (C6D6): νCO 2033, 1966 cm-1. Anal.
Calcd for RuC16H34Cl2O2P2: C, 39.03; H, 6.96. Found: C,
39.26; H, 7.07.
Activa tion of Ma gn esiu m Tu r n in gs. The typical proce-
dure to generate the activated magnesium turnings is followed.
Magnesium turnings (25 mg, 1.03 mmol) and THF (1 mL) were
placed in a Schlenk flask, and 1,2-dibromoethane (20 µL, 0.23
mmol) was added via syringe. The mixture was gently stirred
until the evolution of ethylene ceased, to give the activated
magnesium (0.80 mmol). This magnesium was used im-
mediately in another reaction without further treatment.
Ru (CO)2(P P h 3)3 (1). Magnesium turnings (55 mg, 2.27
mmol) were activated with 1,2-dibromoethane (41 µL, 0.47
mmol) in THF (1.5 mL) in a Schlenk flask. To the flask, a
mixture of cis,cis,trans-RuCl2(CO)2(PPh3)2 (1.36 g, 1.80 mmol)
and triphenylphosphine (0.49 g, 1.87 mmol) in THF (40 mL)
was added by means of cannula transfer. The mixture was
stirred at 60 °C, until all of the magnesium was consumed
(ca. 10 h). During this period, the color of the solution changed
from colorless to pale yellow. The volatiles were removed, and
the pale yellow residue was extracted with benzene (40 mL
4). After the insoluble material was filtered away, the solution
was concentrated to ca. 20 mL. Addition of methanol (ca. 100
mL) to this solution gave the yellow-orange precipitate.
Washing with MeOH and pentane gave the title compound in
pure form; yield 1.55 g (1.62 mmol, 92%). All of the spectro-
scopic data are consistent with those reported previously.2,13
Ru (CO)2(P MeP h 2)3 (2). A mixture of cis,cis,trans-RuCl2-
(CO)2(PMePh2)2 (300 mg, 0.48 mmol) and PMePh2 (105 mg,
0.52 mmol) in THF (10 mL) was added to a THF suspension
of activated magnesium (12 mg, 0.49 mmol) by means of
cannula transfer. The mixture was stirred for 12 h at room
temperature. During this period, the color of the solution
changed from colorless to pale yellow. The volatiles were
removed, and the yellow residue was extracted with hot
Exp er im en ta l Section
Gen er a l. All manipulations were carried out using stan-
dard Schlenk and glovebox techniques under prepurified
argon. Benzene, heptane, pentane, THF, and toluene were
dried over sodium benzophenone ketyl, distilled, and stored
in gastight solvent bulbs. Methanol and 2-methoxyethanol
were degassed under vacuum and used without further
purification. Benzene-d6 and toluene-d8 were dried over
sodium metal and vacuum-distilled prior to use. Phosphines
(PEt3, PMePh2, and PPh3), triphenylarsine, and 1,2-dibromo-
ethane were purchased from Aldrich Chemical Co. and used
without purification. PiPr2Me was synthesized from PiPr2Cl
(Aldrich) and MeLi, distilled, and stored under argon. Tris-
(2-furyl)phosphine was a generous gift from Professor Masa-
hiko Saburi of Saitama Institute of Technology, Japan. Diphen-
ylacetylene was purchased from Aldrich Chemical Co. and
used after purifying by sublimation under reduced pressure.
Gaseous reagents (H2, O2, and CO) were purchased from Air
Products and used as received. Ruthenium trichloride hydrate
(5) Ogasawara, M.; Macgregor, S. A.; Streib, W. E.; Folting, K.;
Eisenstein, O.; Caulton, K. G. J . Am. Chem. Soc. 1995, 117, 8869.
(6) Ogasawara, M.; Macgregor, S. A.; Streib, W. E.; Folting, K.;
Eisenstein, O.; Caulton, K. G. J . Am. Chem. Soc. 1996, 118, 10189.
(7) There are very few reports of other Ru(CO)2(phosphine)3 com-
plexes with some special phosphines. With tridentate chelate phos-
phines, see: (a) Siegl, W. O.; Lapporte, S. J .; Collman, J . P. Inorg.
Chem. 1973, 12, 674. (b) Letts, J . B.; Mazanec, T. J .; Meek, D. W.
Organometallics 1983, 2, 695. With PF3, see: (c) Udovich, C. A.; Clark,
R. J . J . Organomet. Chem. 1976, 36, 355.
(8) Two Ru(CO)2L2L′ complexes are reported as the ligand exchange
products from Ru(CO)2(PPh3)3, see: Bohle, D. S.; Roper, W. R.
Organometallics 1986, 5, 1607.
(9) Ogasawara, M.; Maseras, F.; Gallego-Planas, N.; Streib, W. E.;
Eisenstein, O.; Caulton, K. G. Inorg. Chem. 1996, 35, 7468.
heptane (15 mL
4). The hot filtrate was cooled to room
temperature to give two crops of yellow microcrystals; yield
299 mg (0.40 mmol, 83%). 1H NMR (C6D6, 23 °C): δ 1.68 (m,
9H, PCH3), 6.98 (m, 18H, o- and p-H), 7.53 (m, 12H, m-H).
31P{1H} NMR (C6D6, 23 °C): δ 28.3 (s). IR (C6D6): νCO 1896,
(10) Lupin, M. S.; Shaw, B. L. J . Chem. Soc. A 1968, 741.
(11) Gill, D. G.; Mann, B. E.; Shaw, B. L. J . Chem. Soc., Dalton
Trans. 1973, 311.
(12) Stephenson, T. A.; Wilkinson, G. J . Inorg. Nucl. Chem. 1966,
28, 945.
(13) Gaffney, T. R.; Ibers, J . A. Inorg. Chem. 1982, 21, 2851.

Ru(CO)2L2L′ Complexes
Organometallics, Vol. 16, No. 9, 1997 1981
1844 cm-1. Anal. Calcd for RuC41H39O2P3: C, 64.99; H, 5.19.
Found: C, 64.86; H, 5.17.
Ru(CO)2(PEt3)2(PPh3) (major product) and Ru(CO)2(PEt3)-
(PPh3)2 (minor product), with a small amount of 1 and 3.
Ru(CO)2(PEt3)2(PPh3): 31P{1H} NMR (C6D6, 23 °C) δ 31.8 (d,
J PP ) 14 Hz, 2P, PEt3), 45.9 (t, J PP ) 14 Hz, 1P, PPh3).
Ru(CO)2(PEt3)(PPh3)2: 31P{1H} NMR (C6D6, 23 °C) δ 27.7 (t,
J PP ) 68 Hz, 1P, PEt3), 50.4 (d, J PP ) 68 Hz, 2P, PPh3).
Ru (H)2(CO)2(P Et3)2. A benzene (5 mL) solution of Ru-
(CO)2(PEt3)3 (120 mg, 0.24 mmol) was placed in a Schlenk
flask, and H2 gas was passed through the solution overnight
at room temperature with stirring. The solution, now color-
less, was evaporated to dryness, and the residue was extracted
with pentane (2 mL 3). The pentane solution was concen-
trated to ca. 1 mL, then cooled to -78 °C, yielding colorless
crystals. The crystalline complex melted into a colorless oil
at room temperature; yield 61 mg (0.16 mmol, 66%). 1H NMR
(C6D6, 23 °C): δ -7.95 (t, J HP ) 24.0 Hz, 2H, Ru-H), 1.01
(tvt, J HP ) J HH ) 7.8 Hz, 18H, PCCH3), 1.45 (qvt, J HH ) 7.6
Hz, J HP ) 3.2 Hz, 12H, PCH2Me). 31P{1H} NMR (C6D6, 23
°C): δ 41.4 (s). IR (C6D6): νCO 1995, 1952 cm-1. Anal. Calcd
for RuC14H32O2P2: C, 42.52; H, 8.16. Found: C, 42.49; H, 7.82.
This complex also can be synthesized from 5 or 6 with
H2 gas.
Ru (CO)2(P Et3)3 (3). A colorless solution of cis,cis,trans-
RuCl2(CO)2(PEt3)2 (650 mg, 1.40 mmol) and PEt3 (176 mg, 1.49
mmol) in THF (15 mL) was added to a THF suspension of
activated magnesium (36 mg, 1.50 mmol) obtained as described
above. The mixture continued to stir for 12 h at room
temperature to give a yellow solution. The solvent was
evaporated to dryness under reduced pressure. The residue
was extracted with pentane (7 mL 3), concentrated to ca. 3
mL, and cooled to -40 °C, yielding two crops of yellow crystals
of the title compound (665 mg, 93%). 1H NMR (C6D6, 23 °C):
δ 1.06 (m, 27H, PCCH3), 1.55 (m, 18H, PCH2Me). 31P{1H}
NMR (C6D6, 23 °C): δ 31.2 (s). IR (C6D6): νCO 1883, 1827 cm-1
.
Anal. Calcd for RuC20H45O2P3: C, 46.96; H, 8.87. Found: C,
46.78; H, 8.60.
Ru (CO)2(P iP r 2Me)3 (4). A THF (15 mL) solution of cis,cis,-
trans-RuCl2(CO)2(PiPr2Me)2 (800 mg, 1.62 mmol) and PiPr2-
Me (225 mg, 1.70 mmol) was added to a THF suspension of
activated magnesium (44 mg, 1.80 mmol) in a Schlenk flask.
The mixture was stirred for 12 h at room temperature. During
this period, the color of the solution changed from colorless to
light orange. The volatiles were removed, and the orange
residue was extracted with pentane (5 mL 3). The pentane
solution was concentrated to ca. 2 mL and cooled to -40 °C,
yielding orange crystals; yield 752 mg (1.36 mmol, 84%). 1H
NMR (C6D6, 20 °C): δ 1.09 (m, 18H, PCCH3), 1.16 (m, 18H,
PCCH3), 1.22 (m, 9H, PCH3), 1.89 (m, J HP ) 3.2 Hz, 6H,
PCHMe2). 31P{1H} NMR (C6D6, 20 °C): δ 40.0 (s). IR
Ru (CO)3(P Et3)2. Ru(CO)2(PEt3)3 (100 mg, 0.20 mmol) was
dissolved in benzene (5 mL), and carbon monoxide was passed
through the solution at room temperature for 6 h with stirring.
During this period, the solution color changed from pale yellow
to colorless. The volatiles were removed under reduced
pressure, and the white residue was extracted with pentane
(2 mL
3). After a small amount of insoluble material was
(Nujol): νCO 1865 cm-1
.
Anal. Calcd for RuC23H51O2P3: C,
filtered away, the solution was concentrated to ca. 1 mL and
cooled to -40 °C, yielding colorless needles of the title
compound; yield 67 mg (0.16 mmol, 81%). 1H NMR (C6D6, 23
°C): δ 1.02 (dt, J HP ) 16.6 Hz, J HH ) 7.6 Hz, 18H, PCCH3),
1.49 (qvt, J HH ) 7.6 Hz, J HP ) 3.7 Hz, 12H, PCH2Me). 31P-
49.90; H, 9.28. Found: C, 49.68; H, 8.95.
Ru (CO)2(P Et3)2[P (2-fu r yl)3] (5). A THF (15 mL) solution
of cis,cis,trans-RuCl2(CO)2(PEt3)2 (501 mg, 1.08 mmol) and tris-
(2-furyl)phosphine (254 mg, 1.09 mmol) was added to a THF
suspension of activated magnesium (27 mg, 1.11 mmol). The
mixture was stirred for 12 h at room temperature to give a
yellow solution with a small amount of Mg remaining. The
solvent was removed under reduced pressure. The resulting
solid was extracted with pentane (10 mL 3), and the pentane
extract was concentrated to ca. 5 mL. Cooling to -40 °C
yielded yellow crystals; yield 560 mg (0.90 mmol, 83%). 1H
NMR (C6D6, 23 °C): δ 1.08 (m, 18H, PCCH3), 1.52 (m, 12H,
PCH2Me), 6.08 (m, 3H, 5-furyl), 6.78 (m, 3H, 4-furyl), 7.19 (m,
{1H} NMR (C6D6, 23 °C): δ 42.7 (s). IR (C6D6): νCO 1879 cm-1
.
Anal. Calcd for RuC15H30O3P2: C, 42.75; H, 7.18. Found: C,
42.62; H, 6.85. This complex also can be synthesized from 5
or 6 with CO gas.
Ru (η2-P h CCP h )(CO)2(P Et3)2. (a ) By Ma gn esiu m Re-
d u ction of cis,cis,tr a n s-Ru Cl2(CO)2(P Et3)2. A THF (10 mL)
solution of cis,cis,trans-RuCl2(CO)2(PEt3)2 (302 mg, 0.65 mmol)
and diphenylacetylene (119 mg, 0.67 mmol) was added to a
THF suspension of activated magnesium (16 mg, 0.65 mmol).
The mixture was stirred for 12 h, and the solvent was
evaporated to dryness. The resulting solid was extracted with
pentane (10 mL 3), concentrated to ca. 3 mL, and cooled to
0 °C, yielding two crops of yellow crystals of the title compound
(300 mg, 80%). 1H NMR (C6D6, 23 °C): δ 0.80 (m, 18H,
PCCH3), 1.24 (m, 12H, PCH2Me), 7.06 (m, 2H, p-H), 7.29 (m,
4H, m-H), 8.09 (m, 4H, o-H). 31P{1H} NMR (C6D6, 23 °C): δ
30.1 (s). IR (C6D6): νCO 1954, 1892 cm-1; νCC 1754 cm-1. Anal.
Calcd for RuC28H40O2P2: C, 58.83; H, 7.05. Found: C, 58.76;
H, 6.92.
(b) F r om Ru (CO)2(P Et3)3 a n d P h CCP h . To a solution
of Ru(CO)2(PEt3)3 (11 mg, 0.021 mmol) in C6D6 (0.5 mL) was
added diphenylacetylene (6.0 mg, 0.034 mmol). The solution
was kept stirring at room temperature. After 24 h of stirring,
the 31P NMR spectrum showed >95% conversion of 3 into Ru-
(η2-PhCCPh)(CO)2(PEt3)2 and free PEt3. This complex also can
be synthesized from 5 or 6 with PhCCPh.
3H, 3-furyl). 31P{1H} NMR (C6D6, 23 °C): δ -15.5 (t, J PP
)
24 Hz, 1P, P(2-furyl)3), 35.6 (d, J PP ) 24 Hz, 2P, PEt3). IR
(C6D6): νCO 1900, 1844 cm-1. Anal. Calcd for RuC26H39O5P3:
C, 49.92; H, 6.28. Found: C, 49.84; H, 6.15.
Ru (CO)2(P Et3)2(AsP h 3) (6). A THF (15 mL) solution of
cis,cis,trans-RuCl2(CO)2(PEt3)2 (500 mg, 1.08 mmol) and tri-
phenylarsine (340 mg, 1.11 mmol) was added to a THF
suspension of activated magnesium (28 mg, 1.16 mmol). The
mixture was stirred for 24 h at room temperature to give a
dark brown solution with a small amount of Mg remaining.
The solvent was removed under reduced pressure. The
resulting solid was extracted with pentane (10 mL 3), and
the pentane extract was concentrated to ca. 3 mL. Cooling to
-40 °C yielded tan colored crystals; yield 513 mg (0.73 mmol,
68%). 1H NMR (C6D6, 20 °C): δ 0.97 (tvt, J PH ) J HH ) 7.5
Hz, 18H, PCCH3), 1.32 (tvt, J HH ) 7.5 Hz, J PH ) 3.0 Hz, 12H,
PCH2Me), 7.03-7.09 (m, 9H, Ph), 7.82 (m, 6H, Ph). 31P{1H}
NMR (C6D6, 20 °C): δ 33.8 (s). IR (C6D6): νCO 1889, 1833 cm-1
.
Rea ction of 3 w ith O2. A C6D6 solution of Ru(CO)2(PEt3)3
(15 mg, 0.029 mmol) was placed in an NMR tube with a Teflon
stopcock. The headspace was degassed by a freeze-pump-
thaw cycle, and a calibrated amount of O2 gas (0.03 mmol)
was introduced into the tube. Shaking the tube caused a
solution color change from pale yellow to pale orange. The
31P{1H} NMR spectrum of this soluition, after 20 min, showed
conversion to Ru(CO3)(CO)(PEt3)3 with some other minor
products as well as remaining starting complex. See text for
detail. Ru(CO3)(CO)(PEt3)3: 31P{1H} NMR (C6D6, 23 °C): δ
22.6 (d, J PP ) 22 Hz, 2P), 31.9 (t, J PP ) 22 Hz, 1P). IR (C6D6):
Anal. Calcd for RuC32H45AsO2P2: C, 54.94; H, 6.48. Found:
C, 55.00; H, 6.24.
Magn esiu m Redu ction of Ru Cl2(CO)2(P Et3)2 with P P h 3.
cis,cis,trans-RuCl2(CO)2(PEt3)2 (303 mg, 0.65 mmol) was re-
duced with activated magnesium (16 mg, 0.66 mmol) in the
presence of PPh3 (172 mg, 0.66 mmol) in THF (12 mL). After
the mixture was stirred for 12 h at room temperature, the
solvent was evaporated to dryness. The residue was extracted
with pentane (10 mL 3) and filtered. The pentane extract
was evaporated to dryness under reduced pressure to give oily
products. The 31P{1H} NMR spectrum showed formation of
νCO 1917 cm-1, νCO 1669 cm-1
.
3

1982 Organometallics, Vol. 16, No. 9, 1997
Ogasawara et al.
R u (η2-O2)(CO)2(P E t 3)2. A solution of Ru(CO)2(PEt3)2-
(AsPh3) (20 mg, 0.029 mmol) in C6D6 (0.5 mL) was placed in
an NMR tube fitted with a Teflon stopcock. The solution was
frozen in liquid N2, the headspace was evacuated, and excess
O2 (1 atm) was introduced into the tube. On warming to room
temperature and shaking the tube, the solution color changed
immediately from pale yellow to pale orange. The remaining
O2 was then removed from the tube by a freeze-pump-thaw
cycle. Although 1H and 31P{1H} NMR and IR spectra showed
clean conversion to Ru(η2-O2)(CO)2(PEt3)2, the pure complex
could not be isolated because of contamination with AsPh3.
1H NMR (C6D6, 20 °C): δ 0.93 (dt, J HP ) J HH ) 7.6 Hz, 18H,
PCCH3), 1.54 (qvt, J HH ) 7.6 Hz, J HP ) 3.6 Hz, 12H, PCH2-
Me). 31P{1H} NMR (C6D6, 20 °C): δ 33.8 (s). IR (C6D6): νCO
Ta ble 1. Cr ysta llogr a p h ic Da ta for Ru (CO)2(P Et3)3
formula: C20H45O2P3Ru
a ) 14.229(2) Å
b ) 11.216(1) Å
c ) 16.572(2) Å
â ) 104.17(1)°
V ) 2564.45 Å3
Z ) 4
fw ) 511.57
space group: P21/n
T ) -171 °C
λ ) 0.710 69 Åa
Fcalcd ) 1.325 g cm-3
µ ) 8.1 cm-1
R(Fo)b ) 0.0235
Rw(Fo)c ) 0.0244
a
b
Graphite monochromator. R ) ∑ Fo
-
Fc /∑ Fo
. )
c Rw
[∑w( Fo - Fc )2/∑w Fo
]
1/2, where w ) 1/σ2( Fo ).
2
1995, 1925 cm-1; νOO 839 cm-1
.
Ru (H)2(CO)2(P iP r 2Me)2. A benzene (5 mL) solution of
Ru(CO)2(PiPr2Me)3 (122 mg, 0.22 mmol) was placed in a
Schlenk flask, and H2 gas was passed through the solution
for 10 min at room temperature with stirring. The solution,
now colorless, was evaporated to dryness, and the residue was
extracted with pentane (2 mL 3). The pentane solution was
concentrated to ca. 1 mL, then cooled to -78 °C, yielding fine
white needles; yield 66 mg (0.16 mmol, 71%). 1H NMR (C6D6,
23 °C): δ -8.25 (t, J HP ) 23.6 Hz, 2H, Ru-H), 0.93 (dvt, J HP
) J HH ) 7.0 Hz, 12H, PCCH3), 1.13 (br, 6H, PCH3), 1.15 (dvt,
J HP ) J HH ) 7.0 Hz, 12H, PCCH3), 1.66 (m, 4H, PCHMe2).
31P{1H} NMR (C6D6, 23 °C): δ 54.6 (s). IR (C6D6): νCO 1995,
1952 cm-1. Anal. Calcd for RuC16H36O2P2: C, 45.38; H, 8.57.
Found: C, 45.40; H, 8.17.
Ru (CO)3(P iP r 2Me)2. To a benzene (5 mL) solution of Ru-
(CO)2(PiPr2Me)3 (100 mg, 0.18 mmol) carbon monoxide was
bubbled for 10 min at room temperature. The solution color
changed from orange to colorless during this period. The
volatiles were removed under reduced pressure, and the
residue was extracted with pentane (3 mL 3). The filtrate
was concentrated to ca. 2 mL and cooled to -78 °C to give the
title compound as colorless needles; yield 78 mg (0.17 mmol,
96%). 1H NMR (C6D6, 23 °C): δ 0.98 (dvt, J HP ) J HH ) 7.2
Hz, 12H, PCCH3), 1.16 (dvt, J HP ) J HH ) 7.2 Hz, 12H, PCCH3),
1.19 (vt, J HP ) 6.1 Hz, 6H, PCH3), 1.83 (m, 4H, PCHMe2). 31P-
F igu r e 1. ORTEP drawing of Ru(CO)2(PEt3)3, 3, with
selected atom labels.
cm-1
. Anal. Calcd for RuC16H34O4P2: C, 42.38; H, 7.56.
Found: C, 41.95; H, 7.26.
X-r a y Str u ctu r e Deter m in a tion . (a ) Ru (CO)2(P Et3)3.
A single crystal was obtained by cleaving a large piece of the
sample in a nitrogen atmosphere glovebag. The crystal was
mounted using silicone grease, and it was then transferred to
a goniostat where it was cooled to -171 °C for characterization
and data collection. The compound is thermochromic, the
bright yellow crystal becoming colorless at low temperature.
A preliminary search for peaks and then analysis using the
programs DIRAX and TRACER revealed a primitive mono-
clinic cell (Table 1). Following intensity data collection (6° <
2θ < 55°), the additional conditions h + 1 ) 2n for h01 and k
) 2n for 0k0 uniquely determined the space group P21/n. Four
standards measured every 300 data showed no signfiicant
trends. The data were corrected for absorption (max and min
factors: 0.818 and 0.924).
The structure was solved using a combination of direct
methods (MULTAN78) and Fourier techniques. The position
of the ruthenium atom was obtained from an initial E-map.
The positions of the remaining non-hydrogen atoms were
obtained from subsequent iterations of a least-squares re-
finement, followed by a difference Fourier calculation. All of
ethyl groups of one of the three PEt3 ligands were disor-
dered, having a major (85% occupancy) and a minor (15%
occupancy) orientation of each CH2 moiety. Hydrogens were
included in fixed calculated positions, with thermal param-
eters fixed at one plus the isotropic thermal parameter of the
carbon to which it was bonded. Although only three of the
non-hydrogen atoms have different positions in the two
orientations of the disordered ligand, all 15 of the hydrogens
have different positions. All were included with the appropri-
ate occupancies and labels. In the final cycles of refinement,
the non-hydrogen atoms were varied with anisotropic thermal
parameters. The final difference map was featureless, the
largest peak being 0.47 and the deepest hole -0.28 e/Å3. The
results are shown in Figure 1, Table 2, and the Supporting
Information.
{1H} NMR (C6D6, 23 °C): δ 51.1 (s). IR (C6D6): νCO 1869 cm-1
.
Anal. Calcd for RuC17H34O3P2: C, 45.43; H, 7.62. Found: C,
45.31; H, 7.22.
Ru (η2-P h CCP h )(CO)2(P iP r 2Me)2. Ru(CO)2(PiPr2Me)3 (126
mg, 0.23 mmol) and diphenylacetylene (45 mg, 0.25 mmol)
were dissolved into 5 mL of benzene, and the solution was
stirred for 10 min at room temperature. The solvent was
removed under reduced pressure, and the residue was ex-
tracted with pentane (3 mL
3). The pentane solution was
concentrated to ca. 4 mL and cooled to -40 °C to give yellow
crystals; yield 129 mg (0.22 mmol, 95%). 1H NMR (C6D6, 23
°C): δ 0.59 (vt, J HP ) 2.7 Hz, 6H, PCH3), 0.89 (dvt, J HP ) J HH
) 6.8 Hz, 12H, PCCH3), 1.02 (dvt, J HP ) J HH ) 7.2 Hz, 12H,
PCCH3), 1.72 (m, 4H, PCHMe2), 7.04 (m, 2H, p-H), 7.29 (m,
4H, m-H), 8.02 (m, 4H, o-H). 31P{1H} NMR (C6D6, 23 °C): δ
40.0 (s). IR (C6D6): νCO 1956, 1894 cm-1; νCC 1752 cm-1. Anal.
Calcd for RuC30H44O2P2: C, 60.09; H, 7.40. Found: C, 59.96;
H, 7.28.
Ru (η2-O2)(CO)2(P iP r 2Me)2. Ru(CO)2(PiPr2Me)3 (100 mg,
0.18 mmol) was placed in a Schlenk flask and dissolved in
benzene (5 mL). The head space of the flask was evacuated
by freeze-pump-thaw cycles, and O2 gas (1 atm) was intro-
duced to the flask. The solution was stirred at room temper-
ature for 10 min, then evaporated to dryness. The residue
was redissolved into pentane (5 mL). After a small amount
of insoluble material was filtered away, the solution was
concentrated to ca. 1 mL and cooled to -78 °C, yielding orange
crystals of the product (68 mg, 0.15 mmol, 83%). 1H NMR
(C6D6, 23 °C): δ 0.94 (dvt, J HP ) J HH ) 7.2 Hz, 12H, PCCH3),
1.03 (vt, J HP ) 3.2 Hz, 6H, PCH3), 1.19 (dvt, J HP ) J HH ) 7.2
Hz, 12H, PCCH3), 1.95 (m, 4H, PCHMe2). 31P{1H} NMR (C6D6,
23 °C): δ 38.6 (s). IR (C6D6): νCO 1993, 1923 cm-1; νOO 884

Ru(CO)2L2L′ Complexes
Organometallics, Vol. 16, No. 9, 1997 1983
Ta ble 2. Selected Bon d Dista n ces (Å) a n d An gles
(d eg) for Ru (CO)2(P Et3)3
space group to Cc or C2/c. An initial choice of C2/c was later
proven correct by the successful solution of the structure. Four
standard reflections measured every 300 data showed no
significant trends. The data were corrected for absorption;
transmission factors ranged from 0.76 to 0.86.
The structure was solved using a combination of direct
methods (MULTAN78) and Fourier techniques. The position
of the ruthenium atom was obtained from an initial E-map.
The positions of the remaining atoms, including nearly all of
the hydrogens, were obtained from subsequent iterations of a
least-squares refinement, followed by a difference Fourier
calculation. All hydrogens were initially placed in idealized
positions. In the final cycles of refinement, the non-hydrogen
atoms were varied with anisotropic thermal parameters. The
final difference map was featureless, the largest peak being
0.47 and the deepest hole -0.42 e/Å3. Results are shown in
Table 6 and the Supporting Information.
Com p u ta tion a l Deta ils. Ab initio calculations on the
model system Ru(CO)2(PH3)3 were carried out at the MP2
computational level. Effective core potentials were used to
replace the 28 innermost electrons of the Ru atom,14 as well
as the 10 core electrons of each P atom.15 The basis set is of
valence double-ς quality for all atoms,14-16 supplemented with
polarization functions on C, O, and P.17 Geometry optimiza-
tions were performed within Cs symmetry, and Ru-P-H
angles were restricted to a single value that was subsequently
optimized.
In the IMOMM calculations, the quantum mechanics de-
scription was applied to the same model system used in the
pure ab initio calculations described above: Ru(CO)2(PH3)3.
The ab initio computational level was exactly the same. For
the molecular mechanics part, the MM3 force field was
applied.18 The torsional constants for the dihedral angles
terminating at the Ru atom (i.e., Ru-P-C-C, Ru-P-C-H)
were set to zero. The bond distances for the atoms linking
the quantum mechanics and the molecular mechanics parts
were frozen to 1.42 Å (P-H, ab initio description) and 1.843
Å (P-C, molecular mechanics description). Apart from this,
IMOMM geometry optimizations were full, with no symmetry
restrictions.
Ru(1)-P(6)
Ru(1)-P(13)
Ru(1)-P(20)
Ru(1)-C(2)
2.3761(5)
2.3422(5)
2.3524(5)
1.8883(19)
Ru(1)-C(4)
O(3)-C(2)
O(5)-C(4)
1.8964(19)
1.1679(24)
1.1613(24)
P(6)-Ru(1)-P(13)
95.785(19) P(13)-Ru(1)-C(4) 88.60(6)
P(6)-Ru(1)-P(20) 102.347(19) P(20)-Ru(1)-C(2) 84.20(6)
P(6)-Ru(1)-C(2)
P(6)-Ru(1)-C(4)
113.32(6)
113.09(6)
P(20)-Ru(1)-C(4) 87.32(6)
C(2)-Ru(1)-C(4) 133.58(8)
P(13)-Ru(1)-P(20) 161.571(18) Ru(1)-C(2)-O(3) 176.63(17)
P(13)-Ru(1)-C(2)
85.64(6)
Ru(1)-C(4)-O(5) 175.14(17)
Ta ble 3. Cr ysta llogr a p h ic Da ta for
Ru (CO)2(P iP r 2Me)3
formula: C23H51O2P3Ru
a ) 16.930(2) Å
b ) 17.932(2) Å
c ) 9.540(1) Å
R ) 93.27(1)°
â ) 91.90(1)°
γ ) 102.04(1)
V ) 2824.88 Å3
Z ) 4
fw ) 553.65
space group: P1h
T ) -173 °C
λ ) 0.710 69 Åa
Fcalcd ) 1.302 g cm-3
µ ) 7.4 cm-1
R(Fo)b ) 0.0345
Rw(Fo)c ) 0.0360
a
b
Graphite monochromator. R ) ∑ Fo
-
Fc /∑ Fo
. )
c Rw
[∑w( Fo - Fc )2/∑w Fo
]
1/2, where w ) 1/σ2( Fo ).
2
(b) Ru (CO)2(P iP r 2Me)3. The complex is not only air-
sensitive but also has a melting point slightly below room
temperature. Fortunately, it is not sensitive to moisture and
CO2. The sample was handled on a dry-ice-cooled glass plate
in a nitrogen atmosphere glovebag. A single crystal was
obtained by cleaving a large piece of the sample, and it was
then mounted on a glass fiber using silicone grease. The
crystal was then transferred to a goniostat using a hand-held
nitrogen cold stream, and it was then cooled to -173 °C for
characterization and data collection (6° < 2θ < 45°).
A
preliminary search for peaks revealed a triclinic cell (Table
3). An initial choice of space group P1h was later proven correct
by the successful solution of the structure. Four standards
measured every 300 data showed no significant trends. The
data were corrected for absorption (max and min factors: 0.862
and 0.924).
Resu lts
The structure was solved using a combination of direct
methods (MULTAN78) and Fourier techniques. The positions
of the ruthenium atoms were obtained from an initial E-map.
The positions of the remaining non-hydrogen atoms were
obtained from subsequent iterations of least-squares refine-
ment, followed by a difference Fourier calculation. There were
two molecules in the asymmetric unit. An isopropyl group and
the methyl group on one of the phosphine ligands on Ru2 are
disordered. The occupancies for the two orientations in the
disorder were refined, and one orientation was dominant. After
a small adjustment to normalize the total, the occupancies
were 60.8% and 39.2%; these were fixed in the remaining
refinement. Hydrogens were included in fixed calculated
positions with thermal parameters fixed at one plus the
isotropic thermal parameter of the carbon atom to which it
was bonded.
P r ep a r a tion of th e Com p lexes. Magnesium re-
duction of cis,cis,trans-RuCl2(CO)2(PPh3)2 in THF in the
presence of 1 equiv of PPh3 gives clean conversion to
Ru(CO)2(PPh3)3, 1, according to eq 2. Our spectroscopic
cct-RuCl2(CO)2L2 + L + Mg f Ru(CO)2L3 + MgCl2
(2)
L ) PPh3 (1), PMePh2 (2), PEt3 (3), PiPr2Me (4)
data for this molecule agree with those reported previ-
ously.2,13 This new synthetic method is much simpler
than the original synthetic route reported by Roper.2
Dichloride complexes, RuCl2(CO)2L2, are readily avail-
able from RuCl3‚nH2O in high yield for a wide range of
phosphines L.10-12 Following this method makes it
possible to synthesize analogous complexes with other
phosphines, such as PMePh2 (2), PEt3 (3), and PiPr2Me
(4). In all cases, reactions proceed cleanly and quanti-
In the final cycles of refinement, the non-hydrogen atoms
were varied with anisotropic thermal parameters. The largest
peak in the final difference map was 1.1 and the deepest hole
was -0.5 e/Å3. Results are shown in Figure 2, Table 4, and
the Supporting Information
(c) Ru (η2-P h CtCP h )(CO)2(P Et3)2. A single crystal was
obtained by cleaving a large piece of the sample in a nitrogen
atmosphere glovebag. The crystal was mounted using silicone
grease, and it was then transferred to a goniostat where it
was cooled to -172 °C for characterization and data collection
(6° < 2θ < 55°). A preliminary search for peaks and then
analysis using the programs DIRAX and TRACER revealed a
C-centered monoclinic cell (Table 5). Following intensity data
collection, the additional condition 1 ) 2n for h01 limited the
(14) Hay, P. J .; Wadt, W. R. J . Chem. Phys. 1985, 82, 299.
(15) Wadt, W. R.; Hay, P. J . J . Chem. Phys. 1985, 82, 284.
(16) Hehre, W. J .; Ditchfield, R.; Pople, J . A. J . Chem. Phys. 1972,
56, 2257.
(17) (a) Krishnan, R.; Binkley, J . S.; Seeger, R.; Pople, J . A. J . Chem.
Phys. 1980, 72, 650. (b) Francl, M. M.; Pietro, W. J .; Hehre, W. J .;
Binkley, J . A.; Gordon, M. S.; DeFrees, D. J .; Pople, J . A. J . Chem.
Phys. 1982, 77, 3654.
(18) Allinger, N. L. MM3(92); QCPE: Bloomington, IN, 1992.

1984 Organometallics, Vol. 16, No. 9, 1997
Ogasawara et al.
F igu r e 2. ORTEP drawings of the two isomers of Ru(CO)2(PiPr2Me)3, 4, with selected atom labels.
Ta ble 4. Selected Bon d Dista n ces (Å) a n d An gles
(d eg) for Ru (CO)2(P iP r 2Me)3
Ta ble 6. Selected Bon d Dista n ces (Å) a n d An gles
(d eg) for Ru (P h C2P h )(CO)2(P Et3)2
Ru(1)-P(7)
Ru(1)-P(15)
Ru(1)-P(23)
Ru(1)-C(3)
Ru(1)-C(5)
Ru(2)-P(35)
Ru(2)-P(43)
2.3423(12)
2.3400(11)
2.3547(11)
1.907(4)
1.910(4)
2.3594(11)
2.3876(11)
Ru(2)-P(51)
Ru(2)-C(31)
Ru(2)-C(33)
O(4)-C(3)
2.3589(11)
1.900(4)
1.902(4)
1.156(5)
1.147(5)
1.160(5)
1.154(5)
Ru(1)-P(20)
Ru(1)-P(27)
Ru(1)-C(2)
Ru(1)-C(3)
Ru(1)-C(16)
Ru(1)-C(18)
2.3640(6)
2.3632(7)
2.1538(23)
2.1538(22)
1.9014(24)
1.9017(25)
O(17)-C(16)
O(19)-C(18)
C(2)-C(3)
1.152(3)
1.151(3)
1.286(3)
1.461(3)
1.461(3)
C(2)-C(10)
C(3)-C(4)
O(6)-C(5)
O(32)-C(31)
O(34)-C(33)
P(20)-Ru(1)-P(27) 175.158(20) C(3)-Ru(1)-C(16) 143.70(10)
P(7)-Ru(1)-P(15) 116.81(4) P(35)-Ru(2)-C(31) 88.44(12)
P(7)-Ru(1)-P(23) 115.33(4) P(35)-Ru(2)-C(33) 85.41(13)
P(20)-Ru(1)-C(2)
P(20)-Ru(1)-C(3)
87.39(6)
88.54(6)
C(3)-Ru(1)-C(18) 113.15(10)
C(16)-Ru(1)-C(18) 103.14(10)
P(7)-Ru(1)-C(3)
P(7)-Ru(1)-C(5)
95.82(12) P(43)-Ru(2)-P(51) 105.31(4)
90.04(13) P(43)-Ru(2)-C(31) 106.72(13)
P(20)-Ru(1)-C(16) 90.81(7)
P(20)-Ru(1)-C(18) 90.71(8)
Ru(1)-C(2)-C(3)
72.63(14)
Ru(1)-C(2)-C(10) 138.03(17)
P(15)-Ru(1)-P(23) 127.81(4) P(43)-Ru(2)-C(33) 106.59(14)
P(27)-Ru(1)-C(2)
P(27)-Ru(1)-C(3)
87.85(6)
87.05(6)
C(3)-C(2)-C(10)
Ru(1)-C(3)-C(2)
Ru(1)-C(3)-C(4)
C(2)-C(3)-C(4)
149.28(23)
72.63(14)
139.40(17)
147.94(23)
P(15)-Ru(1)-C(3)
P(15)-Ru(1)-C(5)
P(23)-Ru(1)-C(3)
P(23)-Ru(1)-C(5)
C(3)-Ru(1)-C(5)
86.61(12) P(51)-Ru(2)-C(31) 84.24(12)
88.46(12) P(51)-Ru(2)-C(33) 87.34(13)
86.20(12) C(31)-Ru(2)-C(33) 146.68(18)
93.63(12) Ru(1)-C(3)-O(4)
173.61(17) Ru(1)-C(5)-O(6)
P(27)-Ru(1)-C(16) 91.57(7)
P(27)-Ru(1)-C(18) 92.85(8)
176.9(3)
178.0(3)
C(2)-Ru(1)-C(3)
34.74(9)
Ru(1)-C(16)-O(17) 177.97(23)
C(2)-Ru(1)-C(16) 108.98(10) Ru(1)-C(18)-O(19) 177.45(23)
C(2)-Ru(1)-C(18) 147.84(10)
P(35)-Ru(2)-P(43) 100.30(4) Ru(2)-C(31)-O(32) 177.6(3)
P(35)-Ru(2)-P(51) 154.39(4) Ru(2)-C(33)-O(34) 177.3(4)
Ru(CO)2(PEt3)3 (at 31.2 ppm). When the reduction is
carried out with a Ru:PPh3 ratio of 1:1.7, the formerly
major AM2 31P{1H} NMR pattern observed from the 1:1
molar ratio and assigned to Ru(CO)2(PEt3)2(PPh3) is now
smaller and the formerly minor AM2 pattern assigned
to Ru(CO)2(PEt3)(PPh3)2 is now larger. The fact that
reduction of cis,cis,trans-RuCl2(CO)2(PEt3)2 in the pres-
ence of PPh3 can lead to a product containing more than
one PPh3 and less than two PEt3 suggests possible
ligand redistribution can be a complicating factor; that
is, trapping of the transient reduction product Ru(CO)2-
(PEt3)2 is not the only process taking place. To test the
possibility of ligand redistribution subsequent to form-
ing Ru(0), Ru(CO)2(PPh3)3 was reacted with equimolar
PEt3 for 1 h at 23 °C in THF. This yields all possible
Ru(CO)2(PEt3)n(PPh3)3-n species, together with free
PPh3. This confirms the hypothesis that pure Ru(CO)2-
(PEt3)2(PPh3) could undergo ligand redistribution even
in the absence of added free PR3. The complex Ru(CO)2-
(PPh3)3 is particularly susceptible to reaction with PEt3,
since it has been suggested to participate in the equi-
librium (eq 4). Because of their complexity, we did not
Ta ble 5. Cr ysta llogr a p h ic Da ta for
Ru (P h C2P h )(CO)2(P Et3)2
formula: C28H45O2P2Ru
a ) 19.353(2) Å
b ) 11.368(1) Å
c ) 26.558(3) Å
â ) 103.45(1)°
V ) 5682.69 Å3
Z ) 8
fw ) 571.64
space group: C2/c
T ) -172 °C
λ ) 0.710 69 Åa
Fcalcd ) 1.336 g cm-3
µ ) 6.86 cm-1
R(Fo)b ) 0.0313
Rw(Fo)c ) 0.0320
a
b
Graphite monochromator. R ) ∑ Fo
-
Fc /∑ Fo
. )
c Rw
[∑w( Fo - Fc )2/∑w Fo
]
1/2, where w ) 1/σ2( Fo ).
2
tatively, with the solution color change from colorless
to yellow. The isolated yield of the complexes is in the
range of 83-93%, depending on the solubility and
crystallinity of the complexes.
Reduction of cis,cis,trans-RuCl2(CO)2(PEt3)2 in the
presence of equimolar PPh3 according to eq 3 for 12 h
gives Ru(CO)2(PEt3)2(PPh3) as the main product. This
cct-RuCl2(CO)2(PEt3)2 + L′ + Mg f
Ru(CO)2(PEt3)2L′ + MgCl2 (3)
Ru(CO)2L3 yk1z Ru(CO)2L2 + L
(4)
L′ ) PPh3, P(2-furyl)3 (5), AsPh3 (6)
k-1
assignment comes from the observed AM2 pattern(s) in
the 31P NMR spectrum and also knowing the 31P NMR
chemical shifts of Ru(CO)2(PPh3)3 (at 50.4 ppm) and of
investigate in detail the kinetics or thermodynamics of
these redistributions.

Ru(CO)2L2L′ Complexes
Organometallics, Vol. 16, No. 9, 1997 1985
Tris(2-furyl)phosphine (PFr3) is a phosphine with a
steric profile close to that of PPh3,19 but with much less
σ-basicity.20 It is thus a promising candidate as a
leaving group from an Ru(CO)2L2L′ species. In addition,
the electronic property of this phosphine as a very weak
σ-donor may avoid the sort of ligand redistribution in a
Ru(CO)2L2L′ species, which was observed in the PEt3/
PPh3 system as described above. In order to have
selectivity in ligand loss from Ru(CO)2L2L′, we chose
PEt3 as the tightly binding ligand L. As expected,
reduction according to eq 3 proceeds cleanly within 12
h. Recrystallization of the pentane extract gives a
mixed phosphine complex Ru(CO)2(PEt3)2(PFr3), 5, in
pure form in 83% yield. Similarly, the phosphine/ar-
sine complex Ru(CO)2(PEt3)2(AsPh3), 6, was prepared
from cis,cis,trans-RuCl2(CO)2(PEt3)2 and AsPh3 in 68%
isolated yield. The lower yield of 6 can be attributed to
the low crystallinity of 6, since an NMR-scale experi-
ment showed almost quantitative conversion of cis,cis,-
trans-RuCl2(CO)2(PEt3)2 into 6.
plex 3 also shows similar NMR and IR characteristics
to those of 2.
The structural conclusions from the spectroscopic data
described above were confirmed by single-crystal X-
ray diffraction of 3. As shown in Figure 1, in the solid
state, two carbonyls of complex 3 occupy equatorial
positions. These equatorial ligands have a C2-Ru1-
C4 angle of 133.58(8)°, which is somewhat above the
ideal 120° angle. The P6-Ru-C2/4 angles are then
identical at 113.32(6)° and 113.09(6)°. The two axial
phosphines bend away from the equatorial phosphine
(angles 95.785(19)° and 102.347(19)°), so that the (trans)
P13-Ru1-P20 angle is only 161.571(18)°. The equato-
rial P6-Ru1 distance, 2.3761(5) Å, is statistically
significantly longer than the axial P-Ru distances (2.35
Å average). The equatorial and one of the axial phos-
phines have only one CH3 group anti to the Ru-P
bond, but the second axial phosphine has no CH3 anti
to the Ru-P bond. Both axial PEt3 groups are nearly
staggered with respect to the equatorial Ru(CO)2P
subunit.
The structure of Roper’s complex 1 has been assigned
as A, based on its IR spectrum,2 which is different from
those of 2 and 3. This structural difference might
originate in the steric properties of the phosphine
ligands. Indeed, the IR spectrum of 4 in pentane shows
a strong band at 1867 cm-1, consistent with the struc-
ture A expected for a phosphine, PiPr2Me, of cone angle
146°, which is nearly identical with that of PPh3 (145°).23
However, this band in the PiPr2Me complex has a
significant shoulder at ca. 1850 cm-1, which is too high
a frequency and excessively intense to be a 13CO
isotopomer of the trans carbonyl species. If this band
is due to an isomer, it should be possible to alter the
mole fractions of each by selective solvation. Any second
isomer B or C has a dipole moment. Therefore, it
should be favored in a polar solvent. In fact, the IR
spectra of 4 in THF or ethanol show increasing amounts
of the lower frequency band and also greater separations
of this band from that of the trans CO isomer.9 If the
assignment of the lower energy band mentioned above
is correct, there must also be a symmetric stretch at
about 60 cm-1 higher frequency. Assuming a C-Ru-C
angle of about 147° (based on an X-ray structure
determination of 4, vide infra), this band should have
an intensity only 9% of that of the asymmetric stretch.
Although we note a weak band in the expected location,
its anticipated weakness makes assignment uncertain.
Stronger back-donation is expected for the second
isomer than that in the trans CO isomer. The observa-
tion of νCO of the second isomer at lower frequency than
that of the trans CO isomer is consistent with this
structural assignment.
Sp ectr oscop ic a n d Cr ysta llogr a p h ic Ch a r a cter -
iza tion of th e Com p lexes. All of the complexes Ru-
(CO)2L3 (1-4, L ) PPh3, PMePh2, PEt3, PiPr2Me) show
only one 31P NMR resonance at room temperature,
which could be taken to indicate structure A. However,
rapid fluxionality of the molecules in solution is expected
for these five-coordinate complexes, so infrared spec-
troscopy (a “faster” technique) is a more reliable method
for structural assignment of these complexes. Only one
νCO band is predicted for the trans structure A, two CO
absorptions with unequal intensity for the cis structure
B, and two νCO bands with approximately equal inten-
sity for C. The IR spectrum of 2 in C6D6, which shows
two strong νCO bands, is inconsistent with structure A.
The IR intensities of complex 2 can be used to calculate
the angle between the two CO diatomic oscillators of
117°,21 which is surely too large for an axial and
equatorial location of two carbonyls ( 90°, structure C)
but is consistent with structure B. The 31P{1H} NMR
spectrum of 2 in toluene at -93 °C shows considerable
broadening of the signal (δ 28.3, w1/2
250 Hz); this
chemical shift is almost the same as that at 23 °C.
Unfortunately, then, it is not possible to halt this
fluxionality, which time-averages the axial and equato-
rial phosphine sites of B. The observed broadening can
be attributed to slow rotation about the Ru-P and P-C
bonds in the ligands in 2 at this temperature.22 Com-
Against this background of solution behavior, the
X-ray structure of crystalline 4 is especially interesting.
The unit cell contains equal amounts of two crystallo-
graphically-independent molecules. One is quite ac-
curately a trans CO TBP (structure A) with idealized
C3v symmetry (Figure 2a). The second is not well-
described as a trigonal bipyramid. Instead, it is better
described as a square pyramid (Figure 2b). This is a
somewhat subtle point since both B (TBP) and D
(square pyramid) have C2v symmetry. Perhaps the
(19) There is no report on the steric profile of tris(2-furyl)phosphine.
However, the cone angle of tri(N-pyrrolyl)phosphine, which also has
five-membered heteroaryl substituents, has been reported as identical
with that of PPh3. Moloy, K. G.; Petersen, J . L. J . Am. Chem. Soc.
1995, 117, 7696.
(20) (a) Allen, D. W.; Taylor, B. F. J . Chem. Soc., Dalton Trans. 1982,
51. (b) Allen, D. W.; Bell, N. A.; Fong, S. T.; March, L. A.; Nowell, I.
W. Inorg. Chim. Acta 1985, 99, 157.
(21) Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chemistry, 5th
ed.; Wiley: New York, 1988; p 1935.
(22) Notheis, J . U.; Heyn, R. H.; Caulton, K. G. Inorg. Chim. Acta
1995, 229, 187.
(23) (a) Tolman, C. A. Chem. Rev. 1977, 77, 313. (b) White, D.;
Coville, N. J . Adv. Organomet. Chem. 1994, 36, 95.

1986 Organometallics, Vol. 16, No. 9, 1997
Ogasawara et al.
of PFr3 decreases the electron density at the Ru center,
and thus diminishes back-donation to the carbonyl
ligands. The π-accepting properties of PFr3 may also
contribute to the coordination of PFr3 in an equatorial
plane. The molecule is persistent to ligand redistribu-
tion, which was observed on the PEt3/PPh3 system
described above; heating a toluene solution of Ru(CO)2-
(PEt3)2(PFr3) up to 110 °C does not show any ligand
redistribution products.
Spectroscopic characteristics of 6 (L′ ) AsPh3) are
very similar to those of 5. In the 31P{1H} NMR
spectrum, 6 shows only one singlet and even at -90 °C
the signal is observed as a sharp resonance at the same
chemical shift. These observations are consistent with
the two phosphines in 6 being equivalent. In the IR
spectrum of 6, two CO stretches are shown with unequal
intensity. All of these data suggest the structure of 6
as B, with triphenylarsine at the third equatorial
position.
Th eor etica l Ca lcu la tion s of Isom er P r efer en ce.
It thus seems that small-to-medium-sized phosphines
yield an isomer where the carbonyls adopt equatorial
sites, while large phosphines (e.g., PPh3) adopt a struc-
ture which minimizes phosphine/phosphine repulsion:
the trans dicarbonyl isomer. This indicates that the
reported electronic preference of CO for an equatorial
site24 is small enough to be ultimately dominated by
steric effects, but the solid state structure of Ru(CO)2-
(PEt3)3 does agree with an earlier molecular orbital
analysis which found back-donation most efficient in the
equatorial sites of a TBP.
Full optimization by ab initio calculations at the MP2
level were carried out for Ru(CO)2(PH3)3. Three ener-
getically-close minima were located with geometries
similar to A, B, and C. The two most stable isomers,
essentially (i.e., within 0.2 kcal‚mol-1) isoenergetic, were
found to have two equatorial CO ligands, isomer B, and
one equatorial and one axial CO ligand, isomer C. The
third isomer, A, with two axial CO ligands was calcu-
lated to be only 3 kcal‚mol-1 above the two other
isomers. This work will focus on isomers A and B.
Isomer C is not considered further because of the fac
arrangement of three bulky ligands.
decisive factor favoring the square pyramidal assign-
ment is thus the increase of the C-Ru-C angle from
the expected 120-130° value in a TBP to 147°. Clearly,
the structure adopted is controlled by an attempt to
minimize L/L repulsion, and apparently, increasing only
the L-Ru-L* angle leads to higher energy than also
distorting the C-Ru-C angle to a square pyramidal
form. This concerted displacement follows the conven-
tional path of Berry pseudorotation. Since both isomers
exist in the solid state, we have attempted to investigate
the solid state IR spectrum. However, the complex 4 is
a solid with a low melting point (mp ) ca. 15 °C) and
our attempt was unsuccessful.
Finding the two isomers of 4, both in the solid state
and in solution, prompted us to reinvestigate Roper’s
complex 1, whose geometry was described as structure
A.2 Indeed, the solution IR spectrum of 1 in C6D6 shows
a small signal at 1856 cm-1 in addition to an originally
reported CO absorption at 1908 cm-1. The expected
second νCO band of the minor isomer is estimated at ca.
1910 cm-1, where it overlaps with the strong signal of
the main isomer. Again, νCO of the minor isomer (with
bent OC-Ru-CO) is observed at a lower frequency than
that of the major isomer (with trans CO ligands), as
shown for complex 4.
Although attempted synthesis of Ru(CO)2(PEt3)2-
(PPh3) leads to a ligand redistribution to form a mix-
ture of Ru(CO)2(PEt3)2(PPh3) and Ru(CO)2(PEt3)(PPh3)2,
their 31P NMR spectra give quite interesting struc-
tural information of these molecules. In the 31P NMR
spectra, each of the complexes is observed as an AM2
pattern. However, there is a huge difference between
2
their J PP coupling constants: 14 Hz for Ru(CO)2(PEt3)2-
(PPh3), which is thus attributed to structure B, and
68 Hz for Ru(CO)2(PEt3)(PPh3)2, which is attributed
to structure A. This structural change is caused by
the averaged cone angle of the three phosphines in
Ru(CO)2(PEt3)2(PPh3) being 136° while that in Ru-
(CO)2(PEt3)(PPh3)2 is 141°.23 These observations strong-
ly support the influence of steric bulk of the phos-
phines in Ru(CO)2(phosphine)3 on the geometry of the
complex.
The most relevant structural parameters of structures
A and B are shown in Scheme 1 (trans and cis). The
structure of the trans CO isomer, optimized in Cs
symmetry, has essentially a TBP geometry. The struc-
ture of the cis CO isomer deviates only modestly from
TBP geometry. The main deviation is the opening of
the angle between the two CO ligands (128.5°) and a
slight bending of the phosphine toward this enlarged
angle (P-Ru-P ) 174.1°). The bond lengths have some
interesting patterns. The longest Ru-P bond is ob-
tained for the equatorial phosphine in the cis isomer.
The Ru-C bond is shorter and accordingly the C-O bond
is longer for the cis isomer. This bond length pattern
is in accord with a well-established fact:24 σ-donor
ligands (phosphine) make the weakest bond at the
equatorial site, while, in contrast, π-acceptors (CO)
make the strongest bond.
Similarly, the 31P{1H} NMR spectrum of 5 (L′ ) PFr3)
shows an AM2 pattern. The chemical shift of PFr3 in
this compound is shifted 61 ppm downfield from that
of pure PFr3 (δ -76.5 in C6D6 at 23 °C), which proves
that it is bound to Ru by a phosphorus atom and not by
2
the furyl ring. The J PP coupling constant, 24 Hz, is
consistent with a 90° P-Ru-P′ angle of structure B
with the PFr3 at an equatorial position and inconsistent
with structure A ( 80 Hz expected). Two CO stretches
with unequal intensity are seen in the IR spectrum at
1900 and 1842 cm-1 in C6D6, confirming structure B.
Although these values are low enough to be consistent
with zero-valent Ru, they are shifted to somewhat
higher frequency compared to those of Ru(CO)2(PEt3)3
(1883 and 1825 cm-1). The weaker σ-donor character
While the geometry pattern is in agreement with the
commonly-accepted effect of a π-acceptor ligand at the
equatorial site, the energy pattern is surprising. Such
a small difference in energy (3 kcal‚mol-1) between the
(24) Rossi, A. R.; Hoffmann, R. Inorg. Chem. 1975, 14, 365.

Ru(CO)2L2L′ Complexes
Organometallics, Vol. 16, No. 9, 1997 1987
Sch em e 1
Sch em e 2
cis and trans structures suggest that there is no strong
site preference for the CO ligand. Since this result was
surprising, it was necessary to determine if this pattern
would also hold for a monocarbonyl complex and also if
it is characteristic of other π-acids.
site, it has a C3v symmetry. The orbitals of the C2v
metal fragment, which have the proper symmetry to
interact with the HOMO 5σ of CO, are the occupied x2
- y2 orbital and the empty LUMO made of a mixture of
py and s orbitals.25 In the case of a C3v metal fragment,
2
The geometry of the model species Ru(CO)(PH3)4 was
therefore optimized. Two minima were located, corre-
sponding to the two different isomers with an equatorial
and axial CO ligand. As in the dicarbonyl system, a
small difference in energy (2 kcal‚mol-1) in favor of the
equatorial CO isomer is obtained. Therefore, a mono-
carbonyl also shows the same effect as the dicarbonyl
complex. However, this absence of electronic site pref-
erence for the equatorial site cannot be generalized to
all π-accepting ligands. For example, calculations were
carried out on Ru(PH3)4(C2H4). In this system, the only
minimum is found for ethylene at the equatorial site
and lying in the equatorial plane. All of the other
structures, including that with axial ethylene, are about
20 kcal‚mol-1 higher in energy.
there is only one empty metal orbital (made of dz , pz,
and s) to interact with the HOMO of CO (Scheme 2).
The occupied x2 - y2 orbital interacts in a four-electron
destabilizing way with the ligand attached at the
equatorial site. Such destabilization is absent when the
ligand is attached at the axial site. The back-donation
from the metal into the two π*CO of the ligand is larger
when associated with the HOMO of the C2v metal
fragment. Thus, while back-donation clearly favors the
equatorial site, a σ interaction disfavors it because of
the four-electron repulsion. This four-electron repulsion
is large in the case of CO since the HOMO directly
points toward x2 - y2. It is of lesser importance in the
case of olefin since the π orbital is closer to the nodal
planes of x2 - y2. Thus, the behavior of CO is strongly
influenced by the σ M-CO interactions and, therefore,
shows some similarities in behavior with a phosphine.
This difference between ethylene and CO should be kept
It remains to be understood why CO and C2H4 ligands
behave differently. This surprising result can be un-
derstood by considering the interactions of a ML4
fragment with a ligand to form the TBP species. If the
ligand is positioned at the equatorial site, the ML4
fragment has a C2v symmetry, while if it is at the axial
(25) Albright, T. A.; Burdett, J . K.; Whangbo, M.-H. In Orbital
Interactions in Chemistry; Wiley: New York, 1985.

1988 Organometallics, Vol. 16, No. 9, 1997
Ogasawara et al.
Ta ble 7. Va lu es of Selected Bon d Dista n ces a n d
Bon d An gles Op tim ized for Com p lex 3A a t th e
IMOMM(MP 2:MM3) Com p u ta tion a l Level
Ta ble 9. Va lu es of Selected Bon d Dista n ces a n d
Bon d An gles Op tim ized for Com p lex 4A a t th e
IMOMM(MP 2:MM3) Com p u ta tion a l Level. X-r a y
Va lu es a r e In clu d ed in th e P a r en th eses for
Com p a r ison
Ru-P(2)
Ru-P(3)
Ru-P(4)
2.363
2.370
2.361
Ru-C(5)
Ru-C(6)
1.910
1.903
Ru-P(15)
Ru-P(23)
Ru-P(7)
2.371 (2.340)
2.392 (2.355)
2.395 (2.342)
Ru-C(3)
Ru-C(5)
1.904 (1.907)
1.908 (1.910)
P(2)-Ru-P(3)
P(2)-Ru-P(4)
P(3)-Ru-P(4)
P(2)-Ru-C(5)
P(3)-Ru-C(5)
116.3
P(4)-Ru-C(5)
P(2)-Ru-C(6)
P(3)-Ru-C(6)
P(4)-Ru-C(6)
C(5)-Ru-C(6)
91.5
87.8
91.0
89.5
178.2
122.7
121.0
90.4
P(15)-Ru-P(23) 127.1 (127.8) P(7)-Ru-C(3)
94.5 (95.8)
88.4 (88.5)
92.3 (93.6)
90.8 (90.0)
P(15)-Ru-P(7)
P(23)-Ru-P(7)
P(15)-Ru-C(3)
P(23)-Ru-C(3)
117.4 (116.8) P(15)-Ru-C(5)
115.5 (115.3) P(23)-Ru-C(5)
89.9
87.2 (86.6)
87.4 (86.2)
P(7)-Ru-C(5)
C(3)-Ru-C(5) 174.3 (173.6)
Ta ble 8. Va lu es of Selected Bon d Dista n ces a n d
Bon d An gles Op tim ized for Com p lex 3B a t th e
IMOMM(MP 2:MM3) Com p u ta tion a l Level. X-r a y
Va lu es a r e In clu d ed in P a r en th eses for
Com p a r ison
Ta ble 10. Va lu es of Selected Bon d Dista n ces a n d
Bon d An gles Op tim ized for Com p lex 4B a t th e
IMOMM(MP 2:MM3) Com p u ta tion a l Level. X-r a y
Va lu es a r e In clu d ed in P a r en th eses for
Com p a r ison
Ru-P(6)
Ru-P(13)
Ru-P(20)
2.480 (2.376)
2.347 (2.342)
2.349 (2.352)
Ru-C(4)
Ru-C(2)
1.891 (1.896)
1.889 (1.888)
Ru-P(43)
Ru-P(35)
Ru-P(51)
2.476 (2.388)
2.372 (2.359)
2.364 (2.359)
Ru-C(31)
Ru-C(33)
1.892 (1.900)
1.898 (1.902)
P(6)-Ru-P(13)
97.8 (95.8)
P(6)-Ru-C(2)
110.1 (113.3)
P(6)-Ru-P(20) 102.3 (102.3) P(13)-Ru-P(20) 159.9 (161.6)
P(6)-Ru-C(4) 113.2 (113.1) C(4)-Ru-C(2)
136.7 (133.6)
P(43)-Ru-P(35) 102.2 (105.3) P(43)-Ru-C(33) 105.9 (106.6)
P(43)-Ru-P(51) 100.3 (100.3) P(35)-Ru-P(51) 157.4 (154.4)
P(43)-Ru-C(31) 108.6 (106.7) C(31)-Ru-C(33) 145.6 (146.7)
in mind when the different behavior by these π-accept-
ing ligands is considered.
exception of some Ru-P distances (in particular, Ru-
Peq for isomer B), which have errors as large as 0.1 Å,
agreement between the computed and experimental
bond distances and angles is very good. The improve-
ment of bond angles with respect to pure ab initio
calculations on the model system is clear. The IMOMM
computed value for the Pax-Ru-Pax angle, which was
171.1° for the model system, is 159.9° for complex 3 (X-
ray, 161.6°) and 157.4° for complex 4B (X-ray, 154.4°).
The experimental difference among bond angles be-
tween the three equatorial phosphine ligands in complex
4A (127.8°, 116.8°, and 115.3°) was absent from the ab
initio calculation on the model system but is reproduced
by the IMOMM calculation, with values of 127.1°,
117.4°, and 115.5°.
Even more interesting than the agreement of the
IMOMM predictions with experimental geometries are
the energetics these calculations provide. Complex 3,
isomer B, the only one existing in the crystal, is
computed to be more stable than isomer A by 3.0
kcal‚mol-1. The relationship between the two isomers
is reversed for complex 4, with A being 2.8 kcal‚mol-1
more stable than B. Electronic effects associated with
the quantum mechanics part of IMOMM always favor
the isomeric form B, with the two carbonyl ligands in
equatorial positions, by 3.1 kcal‚mol-1 in the case of 3
and by 1.7 kcal‚mol-1 in the case of 4. Steric effects,
represented by the molecular mechanics part of IM-
OMM, mark the difference between 3 and 4. Steric
effects always favor the isomeric form A, with the two
carbonyl ligands in axial positions, but they do it by
quite different magnitudes: the preference is quantified
as 0.1 kcal‚mol-1 for complex 3 and 4.5 kcal‚mol-1 for
complex 4. Therefore, the picture emerging from these
calculations is quite clear. There is an electronic
preference for the placement of the π-acidic carbonyl
ligands in equatorial positions (isomer B) though,
quantitatively, this preference is always smaller than
5 kcal‚mol-1 and definitely smaller than was a priori
expected.24 There is a steric preference toward the
placement of the bulkier phosphine ligands in the
equatorial positions (isomer A), the weight of this
Th eor etica l Eva lu a tion of Ster ic vs Electr on ic
Effects. The pure ab initio calculations on the model
Ru(CO)2(PH3)3 system presented in the previous section
provide a satisfactory characterization of the electronic
effects, but they can only provide qualitative suggestions
on what might be the steric effects. These steric effects
were therefore quantitatively assessed through theo-
retical calculations with the IMOMM computational
scheme. This is a recently developed method that
allows a geometry optimization of the equilibrium and
transition state geometries of large molecular systems
by integrating molecular orbital (MO) calculations for
a small model system and molecular mechanics (MM)
calculations for the remainder of the system.26 In this
method, the energy of the real system is expressed as a
sum of the MO energy of the small model system and
the MM energy of the real system, excluding the part
already calculated with the MO method. Using this
energy and its gradient with respect to the nuclear
coordinates, one can fully optimize the structures of the
real system. With a proper choice of the model system,
this computational scheme is able to provide ab initio
quality results on large systems at a price only slightly
higher than that of the ab initio calculations on model
systems. The method has been successfully applied to
the study of some transition metal systems, like Pt-
27
(PR3)2H2 and OsO4(NR3).28
IMOMM calculations were carried out on complexes
3 (Ru(CO)2(PEt3)3) and 4 (Ru(CO)2(PiPr2Me)3), those for
which there are X-ray structures available. These
calculations yield two local minima for each of the
systems, corresponding to isomers A and B. Experi-
mental X-ray structures are available for species 3B,
4A, and 4B, while 3A had not been previously identified.
The most significant parameters of the optimized struc-
tures are presented in Tables 7-10, together with the
corresponding experimental values. With only the
(26) Maseras, F.; Morokuma, K. J . Comput. Chem. 1995, 16, 1170.
(27) Matsubara, T.; Maseras, F.; Koga, N.; Morokuma, K. J . Phys.
Chem. 1996, 100, 2573.
(28) Ujaque, G.; Maseras, F.; Lledo´s, A. Theor. Chim. Acta 1996,
94, 67.

Ru(CO)2L2L′ Complexes
Organometallics, Vol. 16, No. 9, 1997 1989
preference depending on the nature of the phosphine
ligand. In small phosphines, like PEt3, this steric effect
is negligible and isomer B is the more stable form. In
larger phosphines, like PiPr2Me, steric effects are strong
enough to invert the small electronic preference for
isomer B and isomer A becomes the more stable form.
A final aspect of the IMOMM calculation deserving
discussion concerns the Ru-Peq distance in the isomeric
form B. This particular bond has some significance,
since it is the likely candidate to be broken in the
phosphine dissociation process that is the necessary
preliminary step for all the reactivity of these com-
plexes.5 On the one hand, it certainly presents the
longest Ru-P distance, as one may expect from the
weaker bond. It does so in all of our tests, both
experimental (X-ray on 3 and 4) and theoretical (pure
ab initio on model system, IMOMM on 3 and 4), with
this lengthening being overestimated in the theoretical
calculations. This simple correlation between bond
strength and bond length is, however, not so straight-
forward as it may appear. This is the only possible
conclusion from the comparison between complexes 3B
and 4B. It is clear both from theory and experiment
that this isomeric form B is destabilized with respect
to form A in the case of complex 4. One should,
therefore, expect a weaker bond in 4B, as it is indeed
proven by its larger reactivity. However, the Ru-Peq
distances are practically the same in 3B and 4B (3
longer than 4 by 0.012 Å in X-ray; 4 longer than 3 by
0.004 Å in IMOMM). However, the increase in the
phosphine cone angle is manifested in an increase of
the Pax-Ru-Peq angle (3, X-ray, 95.8°; IMOMM, 97.8°.
4, X-ray, 100.3°; IMOMM 100.3°).
F igu r e 3. Infrared spectra (νCO region) of Ru(CO)2(PEt3)3,
3, in pentane and THF.
trum of 5 in the presence of equimolar PFr3 at 80 °C.
There is no broadening of each signal (or coalescence).
Thus, any such dissociation of PFr3 from 5 is too slow
to observe by this NMR technique. However, direct
confirmation of such a dissociation will be discussed
later. It is also noteworthy that this solution shows no
detectable production of Ru(CO)2(PEt3)(PFr3)2 under
these conditions. However, we cannot simply conclude
from these results that tris(2-furyl)phosphine is more
weakly binding to the Ru than PEt3 (vide infra).
The solid state structure of Ru(η2-PhCtCPh)(CO)2-
(PEt3)2 is shown in Figure 4.29 The molecule adopts a
structure in which the two phosphines are trans, and
the two carbonyls, Ru, and the entire PhCCPh ligand
are coplanar. The structure might be described as tri-
gonal bipyramidal with axial phosphines, but the angle
between the two carbonyls is only 103.14(10)°, which
more closely resembles an octahedral angle of Ru(II).
The molecule has very close to C2v symmetry. The Ru-
C16 and Ru-C18 distances are equal, as are the Ru-
C2 and Ru-C3 distances. Consistent with the alkyne
acting as only a two-electron donor, the Ru-C2/3 dis-
tances are long (2.15 Å) and the C2-C3 distance is
lengthened only modestly (to 1.286(3) Å) from that in
the free alkyne (1.204(12) Å).30 The alkyne phenyl rings
are bent back to C2-C3-C4 ) 147.9(2)° and C3-C2-
C10 ) 149.3(2)°. The three ethyl groups on each phos-
phorus adopt the common conformation with one methyl
These theoretical results prompted us to reevaluate
our spectral data in search of evidence for a second (or
a third) isomer in the experimental system. Since NMR
spectra can be obscured by rapid fluxional averaging of
signals from more than one isomer, we reevaluated our
IR spectra. The two strong bands in the spectrum of 3
in pentane (assigned to isomer with structure B) are
accompanied by an additional weak absorption at 1860
cm-1 (Figure 3), which we assign to the isomer of
structure A. Consistent with the lack of dipole moment
of the trans dicarbonyl 3A, there is less of it at
equilibrium in the more polar solvent THF.
Rea ctivity of th e Com p lexes. The reactivity of the
new complexes was studied for selected complexes, 3,
4, 5, and 6, with H2, CO, PhCCPh, and O2. Complexes
3, 5, and 6 all react with H2, CO, and PhCtCPh to give
Ru(H)2(CO)2(PEt3)2, Ru(CO)3(PEt3)2, and Ru(η2-PhCt
CPh)(CO)2(PEt3)2, respectively, together with formation
of free PEt3 for 3, PFr3 for 5, and AsPh3 for 6. In all
cases, the reactions are clean and proceed quantita-
tively, judging from their 31P{1H} NMR spectra. The
reactions are very slow for 3 and 5, and their half-lives
are on the order of hours. Complex 6 is much more
reactive than 3 and 5; all of the reactions mentioned
above are complete within 10 min. All of these reactions
can be explained as proceeding with an initial dissocia-
tion of a phosphine or arsine ligand from the complexes
to give the common transient Ru(CO)2(PEt3)2. In cases
of reactions with 5 or 6, the dissociation of PFr3 from 5
and AsPh3 from 6 is highly selective; no production of
free PEt3 is observed. To test for such a dissociative
equilibrium, we have recorded the 31P{1H} NMR spec-
(29) Gagne´, M. R.; Takats, J . Organometallics 1988, 7, 561. Bird-
whistell, K. R.; Tonker, T. L.; Templeton, J . L. J . Am. Chem. Soc. 1987,
109, 1401. Marinelli, G.; Streib, W. E.; Huffman, J . C.; Caulton, K. G.;
Gagne´, M. R.; Takats, J .; Dartiguenave, M.; Chardon, C.; J ackson, S.
A.; Eisenstein, O. Polyhedron 1990, 19, 1867.
(30) Indiana University Molecular Structure Center Report No.
7774, available from the Chemistry Library, Indiana University.

1990 Organometallics, Vol. 16, No. 9, 1997
Ogasawara et al.
Complex 4, which has different structures (square
pyramid and trans-TBP) than 3 (cis- and trans-TBP),
shows much higher reactivity than 3. However, its
reaction patterns toward H2, CO, and PhCtCPh are
identical to those of 3 to give Ru(H)2(CO)2(PiPr2Me)2,
Ru(CO)3(PiPr2Me)2, and Ru(η2-PhCtCPh)(CO)2(PiPr2-
Me)2, respectively, together with dissociated equimolar
PiPr2Me. It also reacts with O2 to give Ru(η2-O2)(CO)2(Pi-
Pr2Me)2 and PiPr2Me. All of these reactions are very
clean and quite fast (they are finished within an
observation time of 10 min).
Tr is(2-fu r yl)p h osp h in e: A Str on g π-Accep tor .
The unusual phosphine tris(2-furyl)phosphine was ini-
tially employed in 5 because of its extremely weak
σ-basicity.20 Thus, it was expected that PFr3 in 5 would
be a leaving group from 5 and that introduction of PFr3
into the complex would enhance its reactivity. J udging
from the reaction patterns of 5 toward H2, CO, and
PhCCPh (vide supra), this first objective has been
realized. However, is complex 5 more reactive than 3?
During the preliminary reaction study described above,
we could not detect any dramatic reactivity enhance-
ment in 5 vs 3. The reaction progress of 3 and 5 toward
F igu r e 4. ORTEP drawing of Ru(η2-PhCCPh)(CO)2(PEt3)2
with selected atom labels.
anti to the Ru-P bond and the other two methyls nearly
in the plane of the three CH2 groups. The conformation
about each Ru-P bond makes the ethyl groups stag-
gered with respect to all of the Ru-C bonds.
The reactions of complexes 3, 5, and 6 with O2 are
quite different from each other. A benzene solution of
6 reacts immediately with dioxygen gas (immediate
color change) to give Ru(η2-O2)(CO)2(PEt3)2. Although
the spectroscopic data (1H and 31P{1H} NMR and IR)
showed the quantitative conversion to the dioxygen
adduct, isolation of the product in pure form could not
be achieved due to contamination by released AsPh3.
The reaction of 3 and O2 is more complex; the 31P{1H}
NMR spectrum of the initial reaction mixture shows two
sets of AM2 patterns (III and I) and a signal of Et3PO
in addition to resonances of some other minor products
(one of which is Ru(η2-O2)(CO)2(PEt3)2). The AM2 pat-
terns indicate formation of tris(triethylphosphine) spe-
cies, one of which, III, isomerizes to the other (I) in 2
days at room temperature (see Scheme 3).31 The IR
spectrum of this solution shows a strong absorption in
the Ru-CO region at 1917 cm-1 and a weaker band at
1669 cm-1, which is consistent with structure I. An-
other stereoisomer, II, is possible for this species; how-
ever, a facial arrangement of the bulkier ligands, trieth-
ylphosphines, is not probable since this orientation of
the three phosphines was distinctly less favorable (could
not be detected) in Ru(CO)2(PEt3)3 (structure C). Struc-
ture I also has a push/pull interaction between carbon-
ate oxygen and CO, which is less effective in II. As for
the other initial product, we propose structure III. A
mixture of Ru(η2-O2)(CO)2(PEt3)2 and PEt3 does not form
I, which indicates that I does not form via phosphine
dissociation from 3. The different reaction patterns be-
tween 3 and 6 toward O2 suggest that tighter binding
of the three phosphine ligands is important for forma-
tion of the carbonate complex. Similar carbonate com-
plexes are reported for reactions between Ru(CO)2(L3)
and O2, where L3 are tridentate phosphines. In Ru(CO)2-
(L3), the tridentate phosphines resist dissociation due
to the chelate effect.7ab,32 A reaction between 5 and O2
gives many products and is impossible to characterize.
PhCCPh was therefore monitored by 31P{1H} NMR, and
the results are shown in Figure 5. Quite unpredictably,
the PFr3/PEt3 complex, 5, is somewhat less reactive than
3. Since the reaction between the isolable four-
coordinate complex Ru(CO)2(PtBu2Me)2 and PhCCPh
proceeds in the time of mixing,6 a probable common
intermediate from 3 or 5, Ru(CO)2(PEt3)2, should react
with PhCCPh immediately. Thus, the rate-determining
step of the reaction can be assumed to be phosphine
dissociation from 3 or 5. The lower reactivity of 5
indicates that PFr3 in 5 binds tighter than PEt3 in 3.
Combined with the weak σ-base character of PFr3, these
observations can be used to propose a stronger π-acidity
of PFr3. However, this point, the tighter coordination
of PFr3, is a little bit confusing, since the dissociation
of PFr3 from 5 is highly selective. The X-ray crystal
structure of 3 and the ab initio calculations on Ru(CO)2-
(PH3)3 show that the Ru-Peq bond is longer than the
b
Ru-Pax distance, i.e., RuPEt3 is the weakest Ru-P
bond in 3. Thus, the order of the bond strength in 3
and 5 can be estimated as Ru-PEt3b < Ru-PFr3 < Ru-
c
PEt3 . The second inequality makes PFr3 the leaving
group from 5; the first inequality makes 5 less reactive
than 3.
Dissocia tion of a P h osp h in e fr om Ru (CO)2L3.
Although it has been suggested that 1 undergoes
phosphine dissociation in solution, as shown in eq 4, and
its reactivity has supported this idea, there is no direct
evidence of this equilibrium.2 We have tried to detect
evidence of eq 4 by monitoring the 31P{1H} NMR spectra
of Ru(CO)2L3 in the presence of free L for L ) PEt3 and
PiPr2Me. In the case of L ) PEt3, the signals for both
(31) Roundhill, D. M. Comprehensive Coordination Chemistry; Per-
gamon Press: New York, 1987; Vol. 5, p 463. Collman, J . P. Acc. Chem.
Res. 1968, 1, 136. Curtis, M. D.; Han, K. R. Inorg. Chem. 1985, 24,
378. Roper, W. R. J . Organomet. Chem. 1986, 300, 167.
(32) Christian, D. F.; Roper, W. R. J . Chem. Soc., Chem. Commun.
1971, 1271.

Ru(CO)2L2L′ Complexes
Organometallics, Vol. 16, No. 9, 1997 1991
Sch em e 3
F igu r e 5. Progress of the reactions between PhCCPh (5.9
10-2 mmol) and 3 or 5 (5.9
10-2 mmol) in C6D6 (0.5
mL) at 20 °C.
3 and free PEt3 are sharp even at 110 °C in toluene and
do not show any obvious line broadening in their NMR
spectrum. This is consistent with a low reactivity of 3
toward H2, CO, and PhCCPh to give bis(triethylphos-
phine) complexes. However, 31P{1H} NMR signals of a
mixture of 4, which is the most reactive complex in our
hands, and free PiPr2Me show increased broadening at
higher temperatures, as shown in Figure 6. For the
mixture of 4 and PiPr2Me, the line width of free PiPr2-
Me is first-order in [4]. In addition, the line width of
the 31P{1H} NMR signal of 4 at 85 °C is independent of
the concentration of 4 and of added free PiPr2Me.
Furthermore, spin-saturation transfer was observed
between the signals of 4 and free PiPr2Me by 31P NMR
at 75 °C. Saturation of the free phosphine resonance
led to a considerable decrease in the intensity of the
resonance of coordinated phosphine. All of these ob-
servations confirm the dissociative mechanism (eq 4) for
line broadening. The rate constants k1 were determined
over 10 °C increments between 55 and 105 °C from the
line widths of the free PiPr2Me.33 An Eyring plot of
these data is linear, and the activation parameters for
the dissociation of the phosphine from 4 are ∆Hq ) 11.3
( 0.4 kcal‚mol-1, ∆Sq ) -16.7 ( 1.0 cal‚mol-1‚K-1, and
∆Gq (at 85 °C) ) 17.3 ( 0.8 kcal‚mol-1. The negative
∆Sq is unexpected for a dissociative process, and it
contributes significantly to raising the modest ∆Hq to a
larger ∆Gq. We suggest that this negative ∆Sq comes
in part from the need to first rearrange the less crowded
F igu r e 6. Observed variable-temperature 122 MHz 31P
NMR spectra of a mixture of Ru(CO)2(PiPr2Me)3, 4 (δ 40.4,
0.14 mmol), and PiPr2Me (δ -9, 0.21 mmol) in toluene-
d8 (0.7 mL). The rate constants for phosphine loss from 4
are shown.
trans dicarbonyl A to structure D, Scheme 4. The
greater crowding in D will contribute to a negative ∆Sq
as will additional bending of L toward L* as D proceeds
toward the transition state (we know the structure of
Ru(CO)2L2).5 For example, it has been shown that ∆S°
for a similar rearrangement (of Os(CO)2(C2F4)(PPh3)2)
is -14.3 cal‚mol-1‚K-1 34
Apparently, after D is formed,
.
there is only a minimum release of entropy until point
q is reached. The ∆Hq and ∆Gq obtained here lie in the
range of those reported for the H2 dissociation from
dihydrogen complexes.35 Dihydrogen represents one of
(34) Burrell, A. K.; Clark, G. R.; Rickard, C. E. F.; Roper, W. R.;
Ware, D. C. J . Organomet. Chem. 1990, 398, 133.
(35) (a) Zhang, K.; Gonzalez, A. A.; Hoff, C. D. J . Am. Chem. Soc.
1989, 111, 3627. (b) Millar, J . M.; Kastrup, R. V.; Melchior, M. T.;
Howath, I. T.; Hoff, C. D.; Crabtree, R. H. J . Am. Chem. Soc. 1990,
112, 9643. (c) Halper, J .; Cai, L. S.; Desrosiers, P. J .; Lin, Z. R. J . Chem.
Soc., Dalton Trans. 1991, 717. (d) Gusev, D.; Kuhlman, R. L.; Renkema,
K. B.; Eisenstein, O.; Caulton, K. G. Inorg. Chem. 1996, 35, 6775.
(33) The rate constants, k1, were calculated from the following
equation: k1 ) πR[∆-1/(πT2)], where R is PiPr2Me(free)/4 molar ratio,
∆ is the free phosphine line width, and T2 is the spin-spin relaxation
time of the free phosphine resonance.

1992 Organometallics, Vol. 16, No. 9, 1997
Ogasawara et al.
Sch em e 4
the weakest ligands in transition metal chemistry. The
kinetic parameters obtained here (being an upper limit
on the dissociation energy) indicate how easy the
phosphine dissociation is in 4. In contrast, ∆Hq for loss
coordinate species Ru(CO)2L2 (which leads immediately
to Ru(O2)(CO)2L2) and also the ease of (one-electron)
oxidation of intact Ru(CO)2L2L′. We propose that the
relatively electron-rich Ru(CO)2(PEt3)3 (which only slowly
of CO from Ru(CO)5 is 27.6 kcal‚mol-1 36
.
loses PEt3) reacts in part by electron transfer with O2
-•
and that the geminate pair [Ru(CO)2(PEt3)3+•; O2
]
collapses to form the peroxycarbonate III. The presence
of the peroxy linkage makes III metastable, and it
ultimately rearranges to carbonate.
Discu ssion
The complexes Ru(CO)2L2L′ participate in either an
X or Y equilibrium, in solution, and this is very
dependent on the steric and electronic properties of the
phosphines. The order of the steric properties (cone
angle) of the phosphines23 which we employed in this
study is PiPr2Me (146°) > PPh3 (145°) > (PPh3)2(PEt3)
(141°, average) > (PPh3)(PEt3)2 (136°, average) PMe-
Ph2 (136°) > PEt3 (132°). This order clearly explains
the relationship between the steric characteristics of the
phosphines and the structure of the complexes; the first
three give equilibrium X, the latter three are Y. In
This work has shown the continuum of behavior, both
structural and reactivity, which can come from system-
atic variation of the phosphine identity in Ru(CO)2L3.
Our synthetic method also allows detection of further
subtle features which arise in mixed phosphine or
phosphine/arsine species Ru(CO)2L2L′. While it was
always clear that different structures were of similar
energy for a five-coordinate complex, there were only
very rarely instances of two structures coexisting at
detectable levels and those were only in the solid state,
where intramolecular preferences could be subject to
solid state packing effects. This is the first study where
systematic modification of the groups L and L′ permit
mapping of the changeover of the preferred isomer from
electronically to sterically dictated. For example, we
show how a cis isomer can be sterically destabilized and
a trans isomer results. What could not have been
anticipated, however, is that destabilizing the cis-TBP
form gives rise to not one but two alternatives, the trans
and the distorted cis. Moreover, the cone angle similar-
ity between PiPr2Me and PPh3 led to reinvestigation and
the discovery that 1 is not simply a trans isomer. The
phenomenon of multiple isomers in solution is general.
The detected structure of the distorted cis isomer relied
wholly on the coexistence of this form with the trans
isomer in the crystal studied by X-ray diffraction. This
has the additional benefit of showing how a bulky
phosphine can increase the thermal dissociative reactiv-
ity of Ru(0). Moreover, these results, with phosphines
of cone angle less than 146°, provide a context for better
understanding why PtBu2Me, with cone angle 161°,
destabilizes a Ru(CO)2L3 species to the point where
Ru(CO)2(PtBu2Me)2 can be isolated and will not interact
with PtBu2Me.6 The resulting four-coordinate, zero-
valent Ru species then shows a remarkably high but
also sterically-selective reactivity.
contrast, the electronic properties (pKa value shown
below) of the phosphine do not fit with the structural
behavior: PiPr2Me37 > PEt3 (8.69) > (PPh3)(PEt3)2 (6.70,
average) > (PPh3)2(PEt3) (4.72, average) > PMePh2
(4.57) > PPh3 (2.73). It is known that the equatorial
sites in a five-coordinate d8 complex permit the greatest
back-donation.24 Thus, the strongest π-acceptor will
occupy equatorial sites (Y) until phosphine/phosphine
steric repulsion becomes intolerable (e.g., PPh3 ) L )
L′), and then structure X is preferred. However, the
electronic site preference is small in the case of CO, and
thus, it is relatively easy to manipulate the structures
by steric control only. A comprehensive study of M(CO)4-
(ER3) species (M ) Fe, Ru, O5; E ) P, As, Sb; R ) Me,
Ph) has shown an equilibrium in solution between axial-
and equatorial-ER3 isomers and related these to both
steric and electronic factors.38
The varied reactivity of Ru(CO)2L2L′ with O2 is
proposed to depend on the relative ability of the four-
Thus, even for a very “ordinary” phosphine like PEt3,
our studies show the coexistence of two isomers of Ru-
(CO)2L3. As the phosphine becomes larger, two isomers
are still present, but their coordination geometry is
considerably different from trigonal bipyramidal. Since
these results apply even to the frequently employed
(36) Hastings, W. R.; Roussel, M. R.; Baird, M. C. J . Chem. Soc.,
Dalton Trans. 1990, 203.
(37) Although the pKa value of PiPr2Me has not been reported, there
is a report postulating that PiPr2Me is more basic than PEt3. See:
Vasteg, S.; Heil, B.; Marko´, L. J . Mol. Catal. 1979, 5, 189.
(38) Martin, L. R.; Einstein, F. W. B.; Pomeroy, R. K. Inorg. Chem.
1985, 24, 2777.

Ru(CO)2L2L′ Complexes
Organometallics, Vol. 16, No. 9, 1997 1993
phosphine PPh3, the generality and significance of the
present results seems established. Several other ex-
amples of coexistence of several TBP isomers have been
reported previously. Thus, Os(CO)2(C2H4)3 exists as
diaxial dicarbonyl and axial/equatorial dicarbonyl iso-
mers,39 and Os(C2F4)(CO)2(PPh3)2 coexists34 as diequa-
torial and diaxial dicarbonyl isomers which do not
coalesce by 31P NMR at 25 °C. The work reported here
shows the dramatic impact of the phosphine identity
on the chemical reactivity of these molecules.
We began this project with the objective of finding
some compound Ru(CO)2L2L′ which would serve as a
“stable” (long shelf-life as a solid) precursor on dissolving
(by eq 4) to four-coordinate, reactive Ru(CO)2L2 for
ligands L where the isolation of this 16-electron species
eluded us, due to unfavorable thermodynamics.6 This
was, in fact, the special utility of Roper’s complex Ru-
(CO)2(PPh3)3. This goal has been achieved in the form
of Ru(CO)2(PEt3)2(AsPh3) and Ru(CO)2(PiPr2Me)3, while
Ru(CO)2(PEt3)3 stands as a useful comparison compound
in terms of its greatly reduced reactivity (in the absence
of outer sphere electron transfer, with a reagent like
O2). The study of Ru(CO)2(PEt3)2[P(2-furyl)3] reveals
that the P(2-furyl)3 ligand binds more tightly than might
have been predicted.
Ack n ow led gm en t. F.M. thanks the CNRS for a
research associate position. N.G.-P.’s stay in Montpel-
lier was financed by the European Community via the
Human Capital and Mobility Program (CHRX-CT93-
0152). The U.S./France Collaboration (K.G.C./O.E.) was
financed by a bilateral NSF/CNRS (PICS) program. We
thank Professor Masahiko Saburi of Saitama Institute
of Technology and J ohnson Mathey/Aesar for material
support.
Su p p or tin g In for m a tion Ava ila ble: Tables of crystal-
lographic data, fractional coordinates, and thermal parameters
and crystal structures for [Ru(CO)2(PEt3)3], [Ru(CO)2(PiPr2-
Me)], and [Ru(PhC2Ph)(CO)2(PEt3)2] (15 pages). Ordering
information is given on any current masthead page.
(39) Kiel, G.-Y.; Takats, J .; Grevels, F.-W. J . Am. Chem. Soc. 1987,
109, 2227.
OM9700775