Syntheses, structure, and Berry pseudorotation of ruthenium-, iron-, and cobalt-phosphorane complexes

Article abstract of DOI:10.1016/S0022-328X(01)01263-3

Transition metal complexes bearing a phosphorane fragment with two 3-methylcatecholate substituents, Cp′(CO)₂M{P(OC₇H₆O)₂} (Cp′ = η⁵-C₅H₅, η-C₅Me₅; M = Ru,

Full text of DOI:10.1016/S0022-328X(01)01263-3

Journal of Organometallic Chemistry 646 (2002) 204–211
www.elsevier.com/locate/jorganchem
Syntheses, structure, and Berry pseudorotation of ruthenium-,
iron-, and cobalt-phosphorane complexes
Hiroshi Nakazawa *, Kazumori Kawamura, Tsuyoshi Ogawa, Katsuhiko Miyoshi *
Department of Chemistry, Graduate School of Science, Hiroshima Uni6ersity, Higashi-Hiroshima 739-8526, Japan
Received 20 July 2001; accepted 18 September 2001
Abstract
Transition metal complexes bearing
Cp%(CO)2M{P(OC7H6O)2} (Cp%=h5-C5H5, h5-C5Me5; M=Ru, Fe) and L(CO)3Co{P(OC7H6O)2} (L=PPh3, PPh2Me), were
prepared. The X-ray structure of (h5-C5Me5)(CO)2Ru{P(OC7H6O)2} was determined. From the variable-temperature 31P-NMR
studies, the activation parameters for Berry pseudorotation around the pentacoordinate phosphorus were determined and
discussed. © 2002 Elsevier Science B.V. All rights reserved.
a
phosphorane fragment with two 3-methylcatecholate substituents,
¸¹¹¹¹º
¸¹¹¹¹º
¸¹¹¹¹º
Keywords: Metallaphosphorane; Hypervalent bonds; Berry pseudorotation; Activation parameters
tation for iron, ruthenium, and cobalt phosphoranes
with 3-methylcatecholate substituents.
1. Introduction
Pentacoodinate phosphorus compounds, phospho-
ranes, have attracted considerable attention because
they have 3-center-4-electron bonds (hypervalent
2. Results and discussion
bonds) in the apical positions and are an interesting
2.1. Berry pseudorotation process of
family of chemical species violating the octet rule [1].
metallaphosphorane
One behavior of phosphoranes is apical–equatorial re-
arrangement, so called Berry pseudorotation [2]. Al-
though barriers to Berry pseudorotation of several
cholate substituents in addition to a transition metal
A metallaphosphorane having two 3-methylcate-
organophosphoranes have been evaluated [3], those of
phosphorane having a transition metal fragment as a
substituent, metallaphosphorane, have not been re-
ported. In 1999, we found that a metallaphosphorane
with two 3-methylcatecholate substituents on a phos-
phorane phosphorus is suitable to estimate a barrier to
Berry pseudorotation from the variable-temperature
31P-NMR studies, and reported the first activation
parameters of Berry pseudorotation for a ruthenium
phosphorane [4]. In this paper, we report the synthesis,
structure, and activation parameters of Berry pseudoro-
fragment on a phosphorane phosphorus is suitable for
investigation on a pseudorotation process. Scheme 1
shows a graph for possible isomers of the metallaphos-
phorane and the relationship among them through
Berry pseudorotation according to the convention [2a].
The phosphorus is surrounded by a transition metal
and four oxygens classified into two kinds of oxygens
(O1 and O2, which are close to and far from the methyl
substituent, respectively). A symbol in an oval is to
denote a particular trigonal–bipyramidal (tbp) configu-
ration by the apical atoms. An underlined symbol shows
the optical isomer of the parent compound. An atom
drawn between symbols next to each other shows a
pivot atom about which pseudorotation converts one
isomer into another. There are ten possible configura-
tions, which can be linked by 11 Berry pseudorotations.
* Corresponding authors. Tel.: +81-824-247420; fax: +81-824-
240729.
0022-328X/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(01)01263-3

H. Nakazawa et al. / Journal of Organometallic Chemistry 646 (2002) 204–211
205
2.2. Preparation of metallaphosphoranes
We newly prepared five metallaphosphoranes, all of
which have two 3-methylcatecholate substituents on the
phosphorus. These complexes are shown in Eqs. (1) and
(2), together with 1a reported previously [4].
(1)
(2)
Ruthenium and iron phosphorane complexes (1b, 2a,
and 2b) were prepared from [Cp%(CO)2M{P(OPh)3}]+
(Cp%=h5-C5H5 or h5-C5Me5) and 3-methylcatechol in
the presence of a base (Eq. (1)). In this reaction,
nucleophilic attack of catecholate at the coordinating
Scheme 1.
phosphite
phosphorus
takes
place
to
give
tion pathways via two structures which have a transi-
tion-metal fragment in an apical position. Therefore,
the two signals observed in the 31P-NMR are attributed
to O1O1/O2O2/O1O1/O2O2 and O1O2/O1O2.
The conversion feature mentioned above is further
supported by the 13C-NMR data. Each isolated com-
plex (1b, 2a, and 2b) shows three doublets due to CO.
The rapid O1O2/O1O2 conversion results in the loss of
the diastereotopic relationship between the two car-
bonyl ligands. Thus, O1O2/O1O2 would give one dou-
blet assigned to CO. In contrast, during the
O1O1/O2O2 conversion and equally the O1O1/O2O2
conversion, the phosphorus chirality is not lost, giving
rise to two doublets in the 13C-NMR spectra.
Lattman et al. reported the reaction of
[(PPh3)(CO)3Co]− with ClP(OC6H4O)2 to give cobalt
phosphorane (PPh3)(CO)3Co{P(OC6H4O)2} [6]. We ap-
plied their method for the preparation of cobalt phos-
phoranes bearing two 3-methylcatecholates. The
reaction of [Co(CO)4]− with ClP(OC7H6O)2 was first
attempted to obtain (CO)4Co{P(OC7H6O)2}. However,
the desired complex could not be obtained. According
to the 31P-NMR spectra of the reaction mixture, the
added phosphorane disappeared but no symptom of
cobalt phosphorane formation was observed and it was
found that the reaction was not clean. In contrast, the
Cp%(CO)2M{P(OC7H6O)2} [5] (although the complex
¸¹¹¹¹º
should be described as Cp%(CO)2M{P(OC7H6O)2}, the
tie bar is omitted hereafter for simplicity). The product
was isolated as a yellow powder in a moderate yield.
The elemental analysis data showed that the isolated
complexes were chemically pure metallaphosphoranes,
but the spectroscopic data suggested a mixture of some
isomers. In each case, the 31P-NMR spectrum at room
temperature shows two signals with an approximately
equal intensity in the range of metallaphosphoranes.
Due to equatophilicity of a transition-metal fragment,
four isomers bearing a transition-metal fragment in an
apical position (MO1, MO2, MO1, and MO2) can be
ruled out as possible isomers of the isolated metal-
laphosphoranes. O1O1 versus O2O2, O1O2 versus
O1O2, and O2O2 versus O1O1 are interconvertible by
one transformation pathway with the transition-metal
fragment in an equatorial position as a pivotal group.
Therefore, two isomers in each pair are not distin-
guished in NMR spectra at room temperature. The
interconversion between O1O2 and O1O2 corresponds
to enantiomerization. O1O1/O2O2 is enantiomeric to
O1O1/O2O2, but they are not interconvertible to each
other because the inversion requires four transforma-

206
H. Nakazawa et al. / Journal of Organometallic Chemistry 646 (2002) 204–211
Table 1
3a and 3b may take a trans trigonal-bipyramidal struc-
ture.
Summary of crystal data for 1b
Formula
C26R27O6PRu
2.3. Crystal structure of 1b
Formula weight
Crystal system
Space group
Cell constants
567.54
Monoclinic
P21/c
The structure of 1b was determined by X-ray diffrac-
tion (XRD) analysis. The crystal data are listed in
Table 1. The orientation of one catecholate (C13ꢀC19,
O3, and O4) could be fixed, but that of the other was
not. The unit cell consists of two geometrical isomers
(O2O2 and O1O2) and their enantiomers (O2O2 and
O1O2), and the two geometrical isomers are disor-
dered. Therefore, the methyl position in the disordered
catecholate was fixed in the two positions estimated
from difference Fourier maps with a respective weight
of 0.5. Eventually, O1O1 and O1O1 did not exist in the
crystal, and this is presumably due to the intermolecu-
lar packing requirement. The ORTEP drawing is dis-
played in Fig. 1 and the selected bond distances and
angles are listed in Table 2, respectively, for one of the
isomers.
,
a (A)
8.9630(2)
13.2350(5)
21.9530(8)
101.482(2)
2552.1(1)
4
,
b (A)
,
c (A)
i (°)
V (A )
3
,
Z
Dcalcd (g cm−3
)
1.477
7.15
v, (cm−1
)
Crystal size (mm)
Radiation
2qmax (°)
Unique data
0.40×0.25×0.25
Mo–Ka (u=0.71073 A)
55.8
5640
4877
0.052
0.107
,
Unique data with Fo\3|(Fo)
R
Rw
The X-ray structure clearly shows that 1b is a ruthe-
nium phosphorane complex. The ruthenium has a nor-
mal piano-stool configuration. The five-coordinate
phosphorus adopts a tbp geometry with the ruthenium
atom in one equatorial position. The tbp geometry is
slightly distorted: the two apical bonds bend away from
the ruthenium fragment (the OapꢀPꢀOap angle is 167°),
and in the equatorial plane, the angles of RuꢀPꢀOeq are
slightly larger (122.8 and 127.1°) than an ideal angle of
120°, though the sum of the equatorial angles is 359.2°.
We previously reported the X-ray structure of ruthe-
reaction of [Co(CO)3L]− (L=PPh3, PPh2Me) with
ClP(OC7H6O)2 led to the formation of the cobalt phos-
phoranes (Eq. (2)). The yellow oily complexes encoun-
tered considerable difficulty in the purification process.
An impurity showing a resonance at about −25 ppm
in the 31P-NMR spectrum could not be removed, which
interfered with obtaining analytically pure spectro-
scopic data of 3a and 3b, as well as the correct elemen-
tal analysis data. However, the 31P-NMR spectra are
very informative. Complex 3a shows three doublets at
48.85, 48.91, and 56.84 ppm with JPP=ca. 250 Hz, and
3b shows similarly three doublets at 43.27, 51.45, and
51.56 ppm with JPP=ca. 250 Hz. As mentioned in the
case of the iron and ruthenium phosphoranes, the
cobalt complexes are expected to show two resonances
assignable to the phosphorane with very close chemical
shifts, which may be two doublets at 48.85 and 48.91
ppm for 3a and those at 51.45 and 51.56 ppm for 3b. A
signal at 56.84 ppm for 3a is assigned to PPh3 and that
at 43.27 ppm for 3b to PPh2Me. The upfield shift due to
the change from PPh3 to PPh2Me is quite reasonable.
Lattman et al. reported for (PPh3)(CO)3Co-
{P(OC6H4O)2} that the 31P-NMR spectrum showed an
AB pattern with chemical shifts at 47 and 56 ppm with
nium
phosphorane
complex,
Cp(CO)2Ru-
{P(OC10H6O)2} (4) [4], closely related to 1b. The distor-
tion from the tbp geometry around the phosphorane
phosphorus was also observed for 4: the two apical
bonds bend away from the ruthenium fragment like for
1b, but the angles of RuꢀPꢀOeq are smaller than 120°,
unlike for 1b. The larger RuꢀPꢀOeq angles for 1b may
be due to the bulky Cp* ligand. The RuꢀP bond
,
distance is slightly longer for 1b (2.343 A) than for 4
,
(2.311 A), which may be also due to the bulky Cp*
ligand. A phosphorane fragment tends to take an orien-
tation where the s* orbital of the 3-center-4-electron
bond can accept electron density from the HOMO of
the Cp%(CO)2M fragment which is shown in Fig. 2(a).
Therefore, the ideal phosphorane orientation from the
electronic point of view should have the same torsion
angles (C1ꢀRu1ꢀP1ꢀO4 and C2ꢀRu1ꢀP1ꢀO6). How-
ever, the torsion angles for 1b are 18.2° for
C1ꢀRu1ꢀP1ꢀO4 and 72.1° for C2ꢀRu1ꢀP1ꢀO6 (Fig.
2(b)). The deviation from the ideal phosphorane orien-
tation is larger than that for 4, presumably due to the
bulky Cp* ligand.
JPP=258 Hz. But the phosphorus assignment has not
been made. Now, we can reasonably assign the signal at
47 ppm to the phosphorane. The X-ray crystal structure
of the Lattman’s complex revealed that the cobalt
adopts a trigonal-bipyramidal geometry with three car-
bonyls in equatorial positions and with two phosphorus
ligands in axial positions. Since the JPP values are
almost the same between Lattman’s and our complexes,

H. Nakazawa et al. / Journal of Organometallic Chemistry 646 (2002) 204–211
207
Fig. 1. ORTEP drawing of 1b (40% probability ellipsoids), showing the numbering system. All hydrogen atoms are omitted for clarity.
2.4. Acti6ation parameters for Berry pseudorotation
process
The 31P-NMR signals of metallaphosphoranes
Table 2
,
Selected bond distances (A) and angles (°) for 1b
treated in this paper were temperature dependent. In
the previous paper [4], the VT NMR measurements
were performed for a toluene solution of 1a. For the
corresponding iron complex, 2a, since the signals did
not coalesce even at the boiling point of toluene (248
K), the solvent was changed to p-xylene (b.p.=411 K).
The spectral change with temperature is shown in Fig.
3. While the spectrum shows two singlets at room
temperature, the signals broaden at higher temperatures
and coalesce at 383 K. The broad resonance slightly
sharpens as the temperature is further raised to 393 K,
which was the experimental upper limit imposed by the
boiling point of the solvent. The two sharp singlets
were reproduced when the sample was cooled down to
room temperature, indicating that no decomposition of
the complex takes place during the VT NMR
measurements.
This spectral behavior is explained on the basis of
Berry pseudorotation around the phosphorane phos-
phorus. As mentioned above, O1O1/O2O2 and O1O1/
O2O2 show the identical 31P-NMR chemical shift, and
O1O2/O1O2 shows another singlet at room tempera-
ture. Even at 286 K, the 31P-NMR spectrum shows two
singlets, indicating that energy barriers of pseudorota-
tion between O1O1 and O2O2 (equally between O1O1
and O2O2); and O1O2 and O1O2 are low. In contrast,
Bond distances
Ru1ꢀP1
Ru1ꢀC1
Ru1ꢀC2
P1ꢀO3
P1ꢀO4
P1ꢀO5
2.343(1)
1.887(6)
1.896(5)
1.647(4)
1.784(4)
1.701(8)
1.867(8)
1.138(7)
1.137(7)
1.392(5)
1.334(6)
1.38(1)
C13ꢀC18
C14ꢀC15
C15ꢀC16
C16ꢀC17
C17ꢀC18
C18ꢀC19
C20ꢀC21
C20ꢀC25
C21ꢀC22
C22ꢀC23
C23ꢀC24
C24ꢀC25
C25ꢀC26
1.392(7)
1.396(7)
1.367(9)
1.39(1)
1.383(8)
1.480(9)
1.37(1)
1.37(2)
1.41(2)
1.34(2)
1.49(2)
1.36(2)
1.45(2)
P1ꢀO6
O1ꢀC1
O2ꢀC2
O3ꢀC13
O4ꢀC14
O5ꢀC20
O6ꢀC21
C13ꢀC14
1.35(1)
1.363(7)
Bond angles
P1ꢀRu1ꢀC1
P1ꢀRu1ꢀC2
C1ꢀRu1ꢀC2
Ru1ꢀP1ꢀO3
Ru1ꢀP1ꢀO4
Ru1ꢀP1ꢀO5
Ru1ꢀP1ꢀO6
O3ꢀP1ꢀO4
O3ꢀP1ꢀO5
O3ꢀP1ꢀO6
O4ꢀP1ꢀO5
O4ꢀP1ꢀO6
90.4(2)
88.6(2)
93.8(3)
122.8(1)
96.9(2)
127.1(2)
94.6(3)
88.6(2)
109.3(3)
81.5(3)
92.8(3)
167.7(3)
O5ꢀP1ꢀO6
83.8(4)
113.8(3)
112.7(4)
118.1(7)
114.7(7)
175.0(5)
175.4(5)
113.1(4)
111.5(4)
112.3(9)
111.3(9)
P1ꢀO3ꢀC13
P1ꢀO4ꢀC14
P1ꢀO5ꢀC20
P1ꢀO6ꢀC21
Ru1ꢀC1ꢀO1
Ru1ꢀC2ꢀO2
O3ꢀC13ꢀC14
O4ꢀC14ꢀC13
O5ꢀC20ꢀC21
O6ꢀC21ꢀC20

208
H. Nakazawa et al. / Journal of Organometallic Chemistry 646 (2002) 204–211
ment with the experimental VT NMR data and af-
forded activation parameters via an Eyring analysis
(Fig. 4) [7].
Similar spectral changes were observed for 2b, 3a,
and 3b. For the Cp* derivatives, 1b and 2b, since higher
temperature was required to observe the coalescence
temperature, n-butyl benzene (b.p.=456 K) was used
as a solvent. Even in the solvent, the signals for 1b did
not coalesce. For 3a and 3b, in contrast, a temperature
lower than a room temperature was requested, and then
THF was used as a solvent. Although 3a and 3b have
not been isolated in a pure form, several VT NMR
measurements were performed for samples containing
the impurity in a different ratio (10–30% based on the
cobalt phosphorane) and the same results were ob-
tained. Therefore, it can be said that the impurity does
not affect the energy barriers to Berry pseudorotation.
Activation parameters thus obtained are summarized
in Table 3. Although the DH" for 1b could not be
obtained due to a higher coalescence temperature than
the boiling point of the solvent used, the DH" could be
estimated to be greater than 67.9 kJ mol−1 estimated
for 2b under the same conditions.
Fig. 2. (a) HOMO of the Cp*(CO)2M. (b) Newman projection along
the P1ꢀRu1 bond for 1b.
The activation parameters evaluated here correspond
to the isomerization between O1O1/O2O2 and O1O2/
O1O2, and equally to that between O1O1/O2O2 and
O1O2/O1O2 (see Scheme 1). In any case, the isomeriza-
tion takes place via a relatively unstable isomer having
a transition metal fragment in an apical position. The
isomer is depicted in Fig. 5 where (a) shows the equato-
rial plane and (b) is a plane containing a transition
Fig. 3. Variable-temperature experimental (left) and simulated (right)
31P-NMR spectra of 2a.
Table 3
Activation parameters DH", DS" and DG" for 1a,* 1b, 2a, 2b, 3a,
and 3b
Complex
DH"
DS"
DG"
(kJ mol−1
)
(J mol−1 K−1
)
(kJ mol−1
)
1a
1b
2a
2b
3a
3b
42.190.5
\67.9
51.490.8
67.991.4
38.690.3
38.590.2
−91.191.6
–
−85.792.5
−51.893.7
−96.591.1
−97.190.8
73.190.7 (340 K)
–
84.291.3 (383 K)
89.792.1 (420 K)
67.890.5 (303 K)
67.990.3 (303 K)
* See Ref. [4].
Fig. 4. Eyring plot for the rotation of the phosphorane ligand along
the FeꢀP axis in 2a derived from variable-temperature NMR data.
the two singlets coalesce into one broad resonance at
383 K. This means that isomerization between O1O1/
O2O2 and O1O2/O1O2 (equally between O1O1/O2O2
and O1O2/O1O2) takes place faster than the NMR
time scale at temperatures higher than 383 K. Line
shape analysis yielded calculated spectra in good agree-
Fig. 5. One of the isomers in a process of pseudorotation of metal-
laphosphorane. (a) An equatorial plane; (b) a plane containing a
transition metal in the apical position and parallel to (a).

H. Nakazawa et al. / Journal of Organometallic Chemistry 646 (2002) 204–211
209
metal in the apical position and parallel to (a). With
3. Experimental
interconversion via the unstable isomer, two factors
should be considered: (i) a steric factor (steric repulsion
between equatorial substituents on (a) and an apical
transition metal fragment on (b)) and (ii) an electronic
factor (p-back donation from a transition metal to a
phosphorane fragment which has been pointed out to
be important for Cp(CO)2M(phosphorane) (M=Fe
and Ru)) [4,5,8].
The point which should be noted first in Table 3 is
that the activation parameters for 3a and 3b are the
same. The factor (i) seems to be the same for 3a and 3b
because the phosphine on the cobalt may keep the
position trans to the phosphorane during the pseudoro-
tation. Therefore, the same activation parameters
derived from 3a and 3b show that the extent of p-back
donation from Co(CO)3L to the phosphorane is the
same in 3a and 3b, indicating that the p-back donation
is weak if any.
3.1. General remarks
All reactions were carried out under an atmosphere
of dry nitrogen by using standard Schlenk tube tech-
niques. All solvents were purified by distillation: ether,
THF, benzene, toluene, p-xylene, and n-butylbenzene
were distilled from sodium–benzophenone, hexane and
pentane were distilled from sodium metal, and CH2Cl2
was distilled from P2O5. They were stored under an N2
atmosphere. Other reagents were used without further
purification. Cp(CO)2Fe{P(OPh)3}]PF6 [9], Cp*(CO)2-
Fe{P(OPh)3}]PF6 [9], Cp*(CO)2RuCl [10], [Co(CO)3-
(PPh3)]2 [11], [Co(CO)3(PMePh2)]2 [11], and ClP-
(OC7H6O)2 [12,13] were prepared according to the re-
spective published procedures with some modifications.
IR spectra were recorded on a Shimadzu FTIR-
8100A spectrometer. A JEOL LA-300 multinuclear
spectrometer was used to obtain 1H-, 13C-, and 31P-
The next point noted is that DH9 is greater for the
iron and ruthenium complexes than for the cobalt
complexes. This can be explained from both factors (i)
and (ii). Let us consider the factor (ii) first. The p-back
donation from a filled d orbital to empty s* orbitals of
phosphorane apical bonds have been reported to stabi-
lize metallaphosphoranes having a transition metal
fragment in an equatorial position, such as
Cp(CO)2M(phosphorane) (M=Fe, Ru). However, the
p-back donation is not expected for the isomer shown
in Fig. 5. Therefore, the pseudorotation, for example
from O1O1 to MO2, requests more energy for the Fe
and Ru complexes. Next, let us consider the factor (i).
The Cp ligand protrudes below (b) in Fig. 5, which
leads to considerable steric repulsion between the Cp
ligand and the equatorial substituents giving a greater
activation energy. The effect is more prominent for
bulkier Cp* derivatives, consistent with the greater
DH" (1a vs. 1b and 2a vs. 2b; greater than 16.5 kJ
mol−1 compared with the corresponding Cp
derivative).
1
NMR spectra. H- and 13C-NMR data were referenced
to Si(CH3)4 as an internal standard. 31P-NMR data
were referenced to 85% H3PO4 as an external standard.
Elemental analyses were performed on a Perkin–Elmer
2400CHN elemental analyzer.
3.2. Preparation of [Cp*(CO)2Ru{P(OPh)3}]BF4
A solution of Cp*(CO)2RuCl (613 mg, 1.87 mmol)
and P(OPh)3 (560 mg, 1.87 mmol) in CH2Cl2 (50 ml)
was added to a suspension of AgBF4 (364 mg, 1.87
mmol) in CH2Cl2 (20 ml). The mixture was stirred for
10 h at room temperature (r.t.). After filtration to
remove AgCl formed, the solvent was removed from
the filtrate under reduced pressure to give a yellow
powder, which was washed with benzene and then
ether. Crystallization from CH2Cl2–hexane gave color-
less crystals (967 mg, 1.40 mmol, 75%). Anal. Calc. for
C30H30BF4O5PRu: C, 52.27; H, 4.39. Found: C, 52.02;
H, 4.15%. IR (wCO, cm−1, in CH2Cl2): 2066, 2022.
1H-NMR (l, in CDCl3): 1.97 (d, JPH=3.1 Hz, 15H,
C5(CH3)5), 7.13 (d, JHH=7.8 Hz, 6H, o-C6H5), 7.25 (m,
3H, p-C6H5), 7.40 (t, JHH=7.7 Hz, 6H, m-C6H5).
13C-NMR (l, in CDCl3): 9.62 (s, C5(CH3)5), 104.34 (d,
Finally let us compare DH"s for the Ru and Fe
complexes. It has been reported that the Cp(CO)2Ru
fragment has stronger p-back donation ability than the
corresponding iron fragment [4]. Therefore, DH" is
expected to be greater for a ruthenium complex than
for an iron complex. Actually DH" for 1b is greater
than that for 2b. However, DH"s for 1a and 2a show
the opposite tendency. For Cp%(CO)2M(phosphorane)
(M=Ru, Fe), the RuꢀP bond distances (2.311–2.377
JPC=1.9 Hz, C5(CH3)5), 120.22 (d, JPC=5.0 Hz, o-
C6H5), 126.39 (s, p-C6H5), 130.40 (s, m-C6H5), 149.80
(d, JPC=9.9 Hz, m-C6H5), 195.11 (d, JPC=21.8 Hz,
CO). 31P-NMR (l, in CDCl3): 132.36 (s).
,
A) are longer than the FeꢀP bond distances (2.272–
,
2.291 A) [4,5b]. Therefore, steric repulsion is expected
3.3. Preparation of Cp*(CO)2Ru{P(OC7H6O)2} (1b)
to be smaller for 1a than for 2a, leading to smaller
DH" for 1a than for 2a. DH"s for 1a and 1b (the
difference is only 9.3 kJ mol−1) may be the results of a
critical balance of the two antagonistic (steric and
electronic) factors.
A solution of [Cp*(CO)2Ru{P(OPh)3}]BF4 (371 mg,
0.54 mmol) in THF (5 ml) was added to a solution of
3-methylcatechol (134 mg, 1.08 mmol) and Et3N (0.075
ml, 0.54 mmol) in THF (5 ml). The mixture was stirred

210
H. Nakazawa et al. / Journal of Organometallic Chemistry 646 (2002) 204–211
for 3 h at 55 °C, and the volatile compounds were
removed under reduced pressure. The yellow residue
was extracted with ether (20 ml×2), and the ether
extract was evaporated to dryness. Crystallization from
ether/hexane gave pale yellow crystals of 1b (200 mg,
0.35 mmol, 66%). Anal. Calc. for C26H27O6PRu: C,
THF): 2024, 1975. 1H-NMR (l, in CDCl3): 1.80 (d,
PH=1.7 Hz, 15H, C5(CH3)5), 2.25 (s, 6H,
J
OC6H3(CH3)O), 6.57–6.76 (m, 6H, OC6H3(CH3)O).
13C-NMR (l, in CDCl3): 9.84 (s, C5(CH3)5), 15.28 (s,
OC6H3(CH3)O), 97.92 (s, C5(CH3)5), 107.32 (d, JPC
=
6.8 Hz, OC6H3(CH3)O), 107.70 (d, JPC=8.1 Hz,
OC6H3(CH3)O), 118.80 (s, OC6H3(CH3)O), 119.33 (s,
OC6H3(CH3)O), 119.35 (d, JPC=6.8 Hz, OC6H3-
(CH3)O), 119.63 (d, JPC=6.8 Hz, OC6H3(CH3)O),
121.69 (s, OC6H3(CH3)O), 122.27 (s, OC6H3(CH3)O),
144.81 (d, JPC=2.5 Hz, OC6H3(CH3)O), 145.01 (s,
OC6H3(CH3)O), 145.65 (d, JPC=2.5 Hz, OC6H3-
(CH3)O), 145.95 (d, JPC=1.9 Hz, OC6H3(CH3)O),
212.08 (d, JPC=41.6 Hz, CO), 212.33 (d, JPC=41.0
Hz, CO), 212.63 (d, JPC=41.0 Hz, CO). 31P-NMR (l,
in CDCl3): 74.63 (s), 75.16 (s).
55.02; H, 4.80. Found: C, 54.72; H, 4.70%. IR (wCO
,
1
cm−1, in THF): 2034, 1983. H-NMR (l, in CDCl3):
1.92 (d, JPH=3.3 Hz, 8H, C5(CH3)5), 1.93 (d, JPH=3.3
Hz, 7H, C5(CH3)5), 2.27 (s, 6H, OC6H3(CH3)O), 6.60–
6.78 (m, 6H, OC6H3(CH3)O). 13C-NMR (l, in CDCl3):
10.25 (s, C5(CH3)5), 15.35 (s, OC6H3(CH3)O), 101.54
(d, JPC=3.1 Hz, C5(CH3)5), 101.57 (d, JPC=1.9 Hz,
C5(CH3)5), 107.56 (d, JPC=8.7 Hz, OC6H3(CH3)O),
107.82 (d, JPC=9.3 Hz, OC6H3(CH3)O), 119.33 (s,
OC6H3(CH3)O), 119.59 (s, OC6H3(CH3)O), 119.84 (d,
JPC=7.5 Hz, OC6H3(CH3)O), 120.00 (d, JPC=8.7 Hz,
OC6H3(CH3)O), 121.98 (s, OC6H3(CH3)O), 122.32 (s,
OC6H3(CH3)O), 143.98 (s, OC6H3(CH3)O), 144.30 (s,
OC6H3(CH3)O), 144.72 (s, OC6H3(CH3)O), 145.10 (s,
OC6H3(CH3)O), 198.20 (d, JPC=24.2 Hz, CO), 198.39
(d, JPC=23.6 Hz, CO), 198.55 (d, JPC=23.6 Hz, CO).
31P-NMR (l, in CDCl3): 50.57 (s), 51.18 (s).
3.6. Preparation of (PPh3)(CO)3Co{P(OC7H6O)2} (3a)
[Co(PPh3)(CO)3]2 (257 mg, 0.32 mmol) was dissolved
in THF (15 ml) and then NaK2.8 (168 mg, 1.27 mmol)
was added. After 45 min stirring at r.t., the mixture was
filtered to remove unreacted NaK2.8 and the insoluble
materials. The solution containing K[Co(PPh3)(CO)3]
was cooled at −78 °C and ClP(OC7H6O)2 (190 mg,
0.61 mmol) was added. The reaction mixture was
warmed to r.t., then the volatile solvent was removed to
give a brown oil. After washing the oil with hexane (50
ml×3), ether (10 ml) was added to extract 3a. The
solvent was removed from the ether extract under re-
duced pressure to give a yellow oil (3a) with some
impurity. 31P-NMR (l, in CDCl3): 48.85 (d, JPP=249.9
Hz, phosphorane), 48.91 (d, JPP=238.1 Hz, phospho-
rane), 56.84 (d, JPP=252.5 Hz, PPh3).
3.4. Preparation of Cp(CO)2Fe{P(OC7H6O)2} (2a)
A treatment of [Cp(CO)2Fe{P(OPh)3}]PF6 (443 mg,
0.70 mmol) with 3-methylcatechol (174 mg, 1.40 mmol)
and Et3N (0.39 ml, 2.80 mmol) in a similar manner to
that of 1b gave a yellow powder of 2a (142 mg, 0.31
mmol, 45%). Anal. Calc. for C21H17O6PFe: C, 55.78; H,
3.79. Found: C, 55.55; H, 3.76%. IR (wCO, cm−1, in
1
CDCl3): 2046, 1998. H-NMR (l, in CDCl3): 2.27 (s,
3H, OC6H3(CH3)O), 2.28 (s, 3H, OC6H3(CH3)O), 5.02
(d, 5H, JPH=0.8 Hz, C5H5), 6.64–6.82 (m, 6H,
OC6H3(CH3)O). 13C-NMR (l, in CDCl3): 15.25 (s,
OC6H3(CH3)O), 85.17 (s, C5H5), 107.62 (d, JPC=8.7
3.7. Preparation of (PMePh2)(CO)3Co{P(OC7H6O)2}
(3b)
Hz, OC6H3(CH3)O), 107.87 (d,
JPC=9.4 Hz,
OC6H3(CH3)O), 119.80 (s, OC6H3(CH3)O), 120.03 (s,
OC6H3(CH3)O), 120.14 (d, JPC=10.6 Hz, OC6H3-
(CH3)O), 120.28 (d, JPC=11.2 Hz, OC6H3(CH3)O),
122.38 (s, OC6H3(CH3)O), 122.72 (s, OC6H3(CH3)O),
143.84 (s, OC6H3(CH3)O), 144.12 (s, OC6H3(CH3)O),
144.51 (s, OC6H3(CH3)O), 144.93 (s, OC6H3(CH3)O),
210.20 (d, JPC=44.1 Hz, CO), 210.26 (d, JPC=44.7
Hz, CO), 210.27 (d, JPC=44.1 Hz, CO). 31P-NMR (l,
in CDCl3): 70.97 (s), 71.59 (s).
The complex was prepared from [Co(PPh2Me)-
(CO)3]2 in a manner similar to that of 3a. The yellow oil
of 3b has some impurity, which could not be removed
completely. 31P-NMR (l, in CDCl3): 43.27 (d, JPP
=
246.7 Hz, PPh2Me), 51.45 (d, JPP=247.4 Hz, phospho-
rane), 51.56 (d, JPP=246.7 Hz, phosphorane).
3.8. X-ray structure determination for 1b
3.5. Preparation of Cp*(CO)2Fe{P(OC7H6O)2} (2b)
Crystallographic and experimental details of X-ray
crystal structure analysis for 1b are given in Table 1. A
suitable crystal of 1b was mounted on a glass fiber. All
measurements were made on a MacScience DIP2030
imaging plate area detector with graphite monochro-
A treatment of [Cp*(CO)2Fe{P(OPh)3}]PF6 (559 mg,
0.80 mmol) with 3-methylcatechol (198 mg, 1.59 mmol)
and Et3N (0.11 ml, 0.96 mmol) in a similar manner to
that of 1b gave yellow crystals of 2b (263 mg, 0.50
mmol, 63%). Anal. Calc. for C26H27O6PFe: C, 59.79; H,
5.21. Found: C, 59.77; H, 5.24%. IR (wCO, cm−1, in
,
mated Mo–Ka radiation (u=0.71073 A). The crystal-
to-detector distance was 100 mm with the detector at
the zero swing position. Readout was performed in the

H. Nakazawa et al. / Journal of Organometallic Chemistry 646 (2002) 204–211
211
(b) R.R. Holmes, Pentacoordinated Phosphorus, ACS Mono-
graph Series 175 and 176, vols. I and II, American Chemical
Society, Washington, DC, 1980.
0.050 mm pixel mode. The data were collected at a
temperature of 20091 K to a maximum 2q value of
55.8°. A total of 90 oscillation images were collected,
each being exposed for 4.0 min with an oscillation angle
of 2.0°. Cell parameters and intensities for the reflection
were estimated by the program packages of MACDENZO
[14].
[3] (a) J. Moc, K. Morokuma, J. Am. Chem. Soc. 117 (1995) 11790,
and references cited therein;
(b) T. Uchimura, M. Uebayasi, T. Hirose, S. Tsuzuki, A. Ylin-
iemela¨, K. Tanabe, K. Taira, J. Org. Chem. 61 (1996) 1608;
(c) S. Kojima, K. Kajiyama, M. Nakamoto, K.Y. Akiba, J. Am.
Chem. Soc. 118 (1996) 12866.
The structure was solved by direct methods [15] and
expanded using Fourier techniques [16]. Non-hydrogen
atoms except for the disordered OC6H3(CH3)O were
refined anisotropically, while the rest were refined
isotropically. Hydrogen atoms were included but were
not refined. The final cycle of full-matrix least-squares
refinement was calculated on the basis of 4877 observed
reflections (Fo\3|(Fo)) and 299 variable parameters,
and converged (the largest parameter shift was less than
0.19 times its estimated S.D.) with unweighted and
weighted agreement factors R=0.052 and Rw=0.104.
Neutral atom scattering factors were taken from
Cromer and Waber [17]. All calculations were per-
formed on an SGI Indy O2 computer using the program
system TEXSAN crystallographic software package of
Molecular Structure Corporation [18].
[4] H. Nakazawa, K. Kawamura, K. Kubo, K. Miyoshi,
Organometallics 18 (1999) 2961.
[5] (a) H. Nakazawa, K. Kubo, K. Miyoshi, J. Am. Chem. Soc. 115
(1993) 5863;
(b) K. Kubo, H. Nakazawa, T. Mizuta, K. Miyoshi,
Organometallics 17 (1998) 3522.
[6] M. Lattman, S.A. Morse, A.H. Cowley, J.G. Lasch, N.C. Nor-
man, Inorg. Chem. 14 (1985) 1364.
[7] (a) DMNR5: D.S. Stephenson, G. Binch, Program 365, Quantum
Chemistry Program Exchange, Indiana University, Bloomington,
IN 47405;
(b) DNMR5 (IBM-PC version): C.B. Lemaster, L.C. leMaster,
N.S. true, Program QCMP 059, Quantum Chemistry Program
Exchange, Indiana University, Bloomington, IN 47405.
[8] S.K. Chopra, J.C. Martin, Heteroat. Chem. 2 (1991) 71.
[9] D. Catheline, D. Astruc, Organometallics 3 (1984) 1094.
[10] T. Blockmore, J.D. Cotton, M.I. Bruce, F.G.A. Stone, J. Chem.
Soc. A (1968) 2931.
[11] A.R. Manning, J. Chem. Soc. A (1968) 1135.
[12] F. Ramirez, A.J. Bigler, C.P. Smith, Tetrahedron 24 (1968) 5041.
[13] K.B. Dillon, R.N. Reev, T.C. Waddington, J. Chem. Soc.,
Dalton Trans. (1978) 1465.
[14] MACDENZO: D. Gewirth, (with the cooperation of the program
authors Z. Otwinowski, W. Minor), The MACDENZO Manual—
Acknowledgements
This work was supported by a Grant-in-Aid for
Science Research (Number 12640539 and 12640540)
from the Ministry of Education, Science, Sports and
Culture of Japan.
A
Description of the Program DENZO
,
XDISPLAYF, and
SCALEPACK, Yale University, New Haven, CT, 1995.
[15] SIR92: A. Altomare, M.C. Burla, M. Camalli, M. Cascarano, C.
Giacovazzo, A. Guagliardi, G. Polidori, J. Appl. Crystallogr. 27
(1994) 435.
[16] DIRDIF94: P.T. Beurskens, G. Admiraal, G. Beurskens, W.P.
Bosman, R. de Gelder, R. Israel, J.M.M. Smits, The DIRDIF-94
Program System, Technical Report of the Crystallography Labo-
ratory; University of Nijmegen, The Netherlands, 1994.
[17] D.T. Cromer, J.T. Waber, International Tables for X-ray Crys-
tallography, vol. IV, The Kynoch Press, Birmingham, England,
1974; table 2.2 A.
References
[1] K.Y. Akiba (Chapter 2), T. Kawashima (Chapter 6), Y. Ya-
mamoto, K.Y. Akiba (Chapter 9), in: K.Y. Akiba (Ed.), Chem-
istry of Hypervalent Compounds, Wiley VCH, New York, 1999.
[2] (a) J. Emsley, D. Hall, The Chemistry of Phosphorus, Harper &
Row, New York, 1976;
[18] TEXSAN: Single-Crystal Structure Analysis Software, version 1.6,
Molecular Structure Corporation, The Woodlands, TX, 1993.