Sterically promoted zirconium-phosphorus π-bonding: Structural investigations of [Cp2Zr(Cl){P(H)Dmp}] and [Cp2Zr{P(H)Dmp}2] (Dmp = 2,6-Mes2C6H3)

Article abstract of DOI:10.1016/S0020-1693(99)00335-7

The new terminal zirconocene-organophosphanido complexes [Cp₂Zr(Cl){P(H)Dmp}] (1) and [Cp₂Zr{P(H)Dmp}₂] (2) (Dmp = 2,6-Mes₂C₆H₃) bearing the sterically demanding ligand Dmp have been prepared and structurally characterized. A flattened pyramidal geometry for the phosphorus atom of 1 and a shortened Zr-P bond length of 2.638(1) ? provide evidence for moderate Zr-P π-bonding. Compound 2, however, displays both pyramidal and planar phosphorus atoms. The corresponding Zr-P bond lengths of 2.726(2) and 2.519(2) ?, in conjunction with the phosphorus geometries, indicate that one phosphanido ligand is engaged in substantial π-bonding to the zirconium center while minimal, if any, of such interactions are present for the other phosphanido ligand. The solid state structures and ³¹P NMR spectra for 1 and 2 are very different than previously reported [Cp₂Zr(Cl){P(H)Mes*}] and [Cp₂Zr{P(H)Mes*}₂] (Mes* = 2,4,6-(t)Bu₃C₆H₂) which carry the related sterically demanding group Mes*. Comparisons to other structurally characterized zirconocene(IV) and hafnocene(IV) complexes having terminal phosphanido ligands suggest that upon increasing the steric congestion at the phosphorus atoms greater π interactions with the metal centers are afforded if such steric interactions do not prevent achieving proper alignment of the PR₂ group for optimal M-P π-bonding. (C) 2000 Elsevier Science S.A.

Full text of DOI:10.1016/S0020-1693(99)00335-7

www.elsevier.nl/locate/ica
Inorganica Chimica Acta 297 (2000) 181–190
Sterically promoted zirconium–phosphorus p-bonding: structural
investigations of [Cp2Zr(Cl){P(H)Dmp}] and [Cp2Zr{P(H)Dmp}2]
(Dmp=2,6-Mes2C6H3)
Eugenijus Urnezˇius, Stephen J. Klippenstein, John D. Protasiewicz *
;
Department of Chemistry, Case Western Reser6e Uni6ersity, Cle6eland, OH 44106-7078, USA
Received 6 April 1999; accepted 14 June 1999
Abstract
The new terminal zirconocene–organophosphanido complexes [Cp2Zr(Cl){P(H)Dmp}] (1) and [Cp2Zr{P(H)Dmp}2] (2)
(Dmp=2,6-Mes2C6H3) bearing the sterically demanding ligand Dmp have been prepared and structurally characterized. A
,
flattened pyramidal geometry for the phosphorus atom of 1 and a shortened ZrꢀP bond length of 2.638(1) A provide evidence for
moderate ZrꢀP p-bonding. Compound 2, however, displays both pyramidal and planar phosphorus atoms. The corresponding
,
ZrꢀP bond lengths of 2.726(2) and 2.519(2) A, in conjunction with the phosphorus geometries, indicate that one phosphanido
ligand is engaged in substantial p-bonding to the zirconium center while minimal, if any, of such interactions are present for the
other phosphanido ligand. The solid state structures and 31P NMR spectra for 1 and 2 are very different than previously reported
[Cp2Zr(Cl){P(H)Mes*}] and [Cp2Zr{P(H)Mes*}2] (Mes*=2,4,6-tBu3C6H2) which carry the related sterically demanding group
Mes*. Comparisons to other structurally characterized zirconocene(IV) and hafnocene(IV) complexes having terminal phos-
phanido ligands suggest that upon increasing the steric congestion at the phosphorus atoms greater p interactions with the metal
centers are afforded if such steric interactions do not prevent achieving proper alignment of the PR2 group for optimal MꢀP
p-bonding. © 2000 Elsevier Science S.A. All rights reserved.
Keywords: Crystal structures; Zirconium complexes; Metallocene complexes; Phosphanido ligand complexes
zirconocene–organophosphanido complexes, [Cp2Zr-
(Cl){P(H)Dmp}] (1), and [Cp2Zr{P(H)Dmp}2] (2)
(Dmp=2,6-Mes2C6H3). These materials were prepared
during the course of our work to examine the chemistry
of the zirconium phosphinidene complex [Cp2Zr=
PDmp(PMe3)] which also contains the sterically de-
manding Dmp group to render stability to ZrꢀP
multiple bonds [14]. Recent structural work has un-
veiled novel molecular architectures for the alkali metal
salts of DmpPH2 [15,16]. Compounds 1 and 2 represent
the first examples of structurally characterized transi-
tion metal complexes derived from this primary steri-
cally encumbered phosphine. Analysis of structurally
characterized Group IV metallocene complexes con-
taining terminal phosphanido groups suggests that the
geometry and p-bonding of terminal PR2 groups may
be largely dominated by steric factors.
1. Introduction
Group IV metallocenes containing organophos-
phanido ligands (PR2) have a rich and diversified chem-
istry [1]. Such materials have been investigated for
applications in the synthesis of early–late hetero-
bimetallic complexes [2]. In addition, these materials
are active intermediates in catalytic PP-bond forming
reactions [3,4]. More recently, attention has focused on
the pursuit of related compounds having metal–phos-
phorus multiple bonds for utilization in phosphinidene
transfer reactions [1,5–13]. In the present study we
have prepared and structurally characterized two new
* Coresponding author. Tel.: +1-216-368 5060; fax: +1-216-368
3006.
0020-1693/00/$ - see front matter © 2000 Elsevier Science S.A. All rights reserved.
PII: S0020-1693(99)00335-7

182
E. Urne; zˇius et al. / Inorganica Chimica Acta 297 (2000) 181–190
2. Experimental
filter. The volatiles were removed under vacuum to
afford 0.439 g of deep-red impure 2. H NMR analysis
1
All reactions were performed under a nitrogen atmo-
sphere using Schlenk techniques, Vacuum Atmo-
spheres’ or Braun MB-150M dry boxes. DmpPH2 was
prepared as described previously [17]. Solvents were
distilled from sodium/benzophenone solutions. 1H
NMR spectra were recorded on Varian Gemini 200 or
300 spectrophotometers and referenced to residual sol-
(versus ferrocene) of this material revealed a maximal
yield of 82% for 2. Analytically pure samples were
obtained by recrystallization from 6:1 hot hexanes–
benzene solutions at −35°C. X-ray quality crystals
were obtained by slow pentane diffusion into a satu-
rated THF solution of [Cp2Zr{P(H)Dmp}2] at room
1
temperature. H NMR (C6D6): l 7.02 (m, 2H), 6.94 (m,
1
vent H signals. 31P NMR spectra were recorded on the
4H), 6.88 (s, 8H), 5.28 (s, 10H), 3.70 (d, JHP=256 Hz,
2H), 2.26 (s, 12H), 2.17 (s, 24H). 31P{1H} NMR
(C6D6): l 17.4 (br s). 31P NMR (C6D6): l 17.4 (br d,
Varian Gemini 300 instrument operating at 121.4 MHz
and referenced to external 85% H3PO4. Elemental
analyses were performed at Oneida Research Services,
Whitesboro, NY.
JHP=256 Hz). Anal. Calc. for C58H62P2Zr: C, 76.36; H,
6.85. Found: C, 76.16; H, 6.82%.
2.1. Li[P(H)Dmp]·Et2O
2.4. X-ray crystallography
To a slurry of 1.56 g of DmpPH2 (4.51 mmol) in 20
ml Et2O were added 2.82 ml of 1.6 M nBuLi (4.51
mmol) in hexanes. The resulting slurry was stirred for
30 min and the precipitate was filtered off and dried in
vacuo for 1.5 h. The isolated pale-yellow
Li[P(H)Dmp]·Et2O (1.40 g, 74% yield) has one molecule
Diffraction data were collected with a Siemens P4
,
instrument (Mo Ka radiation (u=0.71073 A)). Crystals
were mounted in the dry-box under nitrogen and sealed
in 0.4 mm capillary tubes. Crystals were judged to be
acceptable based on omega scans and rotation photo-
graphy. A random search located reflections to generate
the reduced primitive cell, and cell lengths were corrob-
orated by axial photography. Additional reflections
with 2q values near 25° were appended to the reflection
array and yielded the refined cell constants. Data were
collected as presented in Table 1 and were corrected for
absorption (empirical „ scans). Examination on the
diffractometer yielded monoclinic cells for both 1 and
2.
1
of coordinated diethyl ether as ascertained by H NMR
1
spectroscopy. H NMR (C6D6): l 6.99 (t, JHH=7 Hz,
1H), 6.84 (d, JHH=7 Hz, 2H), 6.83 (s, 4H), 3.26 (q,
J=7 Hz, 4H), 2.21 (s, 6H), 2.12 (s, 12H), 1.06 (t, J=7
Hz, 6H); PꢀH not observed. 31P NMR (C6D6): l −145
(broad).
2.2. [Cp2Zr(Cl){P(H)Dmp}] (1)
Computations were performed using SHELXTL PLUS
,
A solution of 0.401 g of Li[P(H)Dmp]·Et2O (0.941
mmol) in 3 ml DME was added to a stirred solution of
0.380 g of [Cp2ZrCl2] (1.30 mmol) in 7 ml of DME. A
deep-red color developed immediately. After stirring for
2 h the solution was filtered through a fritted glass filter
and the solvent was removed under vacuum. Recrystal-
lization at 0°C from hot toluene yielded 0.347 g of
orange–red crystalline 1 (crude yield 61%, contami-
nated with trace amounts of 2 and [Cp2ZrCl2]). Further
recrystallization does provide purer 1 but with appre-
Version 5 (Siemens Analytical) and data were refined
2 2
against F2 (wR2=S[w(Fo2−Fc ) ]/S[w(Fo2)2]1/2). Sys-
tematic absences for 1 were consistent with the space
group P2(1)/n. Systematic absences for 2 were consis-
tent with the space group Cc. Refinement of the oppo-
site enantiomorph of 2 resulted in higher R values
(R1=0.0471, wR2=0.1061) and a higher absolute
structure factor (0.37(6)). Most non-hydrogen atoms
were located with direct methods, and the remaining
non-hydrogen atoms were located in successive differ-
ence maps. All of the non-hydrogen atoms were refined
anisotropically. For compound 1 all hydrogen atoms
were located in successive difference maps and their
positions refined. For compound 2 hydrogen atoms on
the phosphorus atoms were located and positions
refined, while remaining hydrogen atoms were gener-
ated at idealized positions.
1
ciable losses in yield. H NMR (C6D6): l 7.06–7.00 (m,
3H), 6.95 (s, 4H), 5.58 (s, 5H), 5.57 (s, 5H), 4.60 (d,
J
HP=250 Hz, 1H), 2.30 (s, 12H), 2.22 (s, 6H). 31P
NMR (C6D6): l 25 (d, JHP=250 Hz). Anal. Calc. for
C34H36ClPZr: C, 67.80; H, 6.02. Found: C, 67.27; H,
5.92%.
2.3. [Cp2Zr{P(H)Dmp}2] (2)
2.5. Computation
To a stirred solution of 0.355 g of Li[P(H)Dmp]·Et2O
(0.833 mmol) in 20 ml of toluene was slowly added a
solution of 0.119 g of [Cp2ZrCl2] (0.411 mmol) in 20 ml
toluene. The resulting deep-red solution was stirred for
a further 45 min and then filtered through a fritted glass
Non-local density functional theory (DFT) was em-
ployed in the determination of the optimized structures
and energies of [Cp2Zr(H)PH2] and [Cp2Zr(Cl)PH2].
These determinations employed the B3-LYP (Becke-3

E. Urne
;
zˇius et al. / Inorganica Chimica Acta 297 (2000) 181–190
183
Table 1
Crystal data and structure refinement for 1 and 2
1
2
Empirical formula
Formula weight
Temperature (K)
C
602.27
293(2)
34H36ClPZr
C58H62P2Zr
912.24
293(2)
,
Wavelength (A)
0.71073
monoclinic
P2(1)/n
0.71073
monoclinic
Cc
Crystal system
Space group
Unit cell dimensions
,
a (A)
10.3728(13)
14.8103(14)
19.5584(13)
90
19.742(3)
12.285(2)
22.598(3)
90
,
b (A)
,
c (A)
h (°)
i (°)
100.006(7)
115.114(9)
k (°)
90
2958.9(5)
90
V (A3)
4962.4(16)
Z
4
4
Dcalc (Mg m−3
)
1.352
0.537
1248
1.221
0.322
1920
Absorption coefficient (mm−1
F(000)
)
Crystal size (mm)
Crystal color/shape
q Range for data collection (°)
Limiting indices
0.36×0.36×0.16
brown–orange/plate
2.09–23.00
−1BhB11, −1BkB16, −21BlB21
5356
0.36×0.30×0.12
red–brown/irregular
1.99–24.50
−1BhB22, −1BkB14, −26BlB24
4962
Reflections collected
Independent reflections
Refinement method
Data/restraints/parameters
Goodness-of-fit on F2
Final R indices [I\2|(I)]
R indices (all data)
Absolute structure parameter
Extinction coefficient
Largest difference peak and hole (e A
4119 (Rint=0.0308)
full-matrix least-squares on F2
4119/0/443
4494 (Rint=0.0284)
full-matrix least- squares on F2
4494/2/557
1.082
1.047
R1=0.0401, wR2=0.0756
R1=0.0743, wR2=0.0864
na
0.0008(2)
0.298 and −0.277
R1=0.0465, wR2=0.0921
R1=0.0754, wR2=0.1045
0.01(5)
0.00022(12)
0.273 and −0.276
−3
,
)
Lee–Yang–Parr) hybrid functional [18] and a 6-31G*
basis set for the C and H atoms [19]. The LANL2DZ
basis set, which includes the Hay and Wadt relativistic
effective core potential, was employed for the Zr atoms
[20]. GAUSSIAN94 quantum chemistry software was em-
ployed in these calculations [21].
Likewise, addition of 0.5 equiv. of Li[P(H)Dmp]·Et2O
to [Cp2ZrCl2] results in a ratio of 1:2 of about 6.0:1.0.
Compounds 1 and 2 are difficult to separate from
[Cp2ZrCl2] and from one another by simple recrystal-
lization when formed under these conditions. The
propensity for the addition of two PR2 groups to
[Cp2ZrCl2] was somewhat diminished by performing the
reaction in dimethoxyethane (DME). Addition of 1
equiv. of Li[P(H)Dmp]·Et2O to [Cp2ZrCl2] in DME
results in a ratio of 2.5:1.0 for 1:2. Under the optimal
3. Results and discussion
conditions for formation of
1
(0.72 equiv.
The addition of 1 equiv. of Li[P(H)Dmp]·Et2O [17] to
zirconocene dichloride in toluene rapidly leads to a
deep blood-red solution containing a mixture of
[Cp2Zr(Cl){P(H)Dmp}] (1) and [Cp2Zr{P(H)Dmp}2] (2)
(Scheme 1). The presence of both 1 and 2 was indicated
by a pair of 31P NMR resonances at l 23.1 ppm (d,
Li[P(H)Dmp]·Et2O to [Cp2ZrCl2], DME), crude yields
of 61% (contaminated by traces of 2 and [Cp2ZrCl2])
for 1 could be obtained. The reaction of 2 with
J
HP=250 Hz) and 18.0 ppm (d, JHP=257 Hz), respec-
1
tively. H NMR (C6D6) analysis of the material after
removal of toluene under vacuum revealed an approxi-
mate equilibrium ratio of 1.0:1.6 for 1 to 2 (in addition
to unreacted [Cp2ZrCl2]). The formation of 2 in prefer-
ence to 1 was further exaggerated if [Cp2ZrCl2] was not
fully dissolved prior to addition of Li[P(H)Dmp]·Et2O.
Scheme 1.

184
E. Urne; zˇius et al. / Inorganica Chimica Acta 297 (2000) 181–190
[Cp2ZrCl2] was explored as an alternative synthesis of
1, but the resulting mixture still contained a distribution
of species. Similar phosphanido redistributions and
labile mixtures of mono- and bis-phosphanido zircono-
cenes have been observed for reactions of LiPPh2 with
[Cp2ZrCl2] [22].
Isolation of the bis-phosphanido species 2 was opti-
mized by increasing the amount of Li[P(H)Dmp]·Et2O
relative to [Cp2ZrCl2] to 2.1:1 and performing the reac-
tion in toluene or benzene. Such reactions yielded 2 as
Scheme 2.
1
the predominate species (ca. 82% yield by H NMR).
Invariably, traces of 1 and a third species (3) were also
present and thus work-up required extensive recrystal-
lization to purify 2 to a level sufficient for satisfactory
elemental analysis. Once free of contaminants, com-
pound 2 is extremely stable in solution in the absence of
water and oxygen. In fact, solutions of pure 2 can be
heated to 80°C (sealed NMR tube, C6D6) for hours
with no signs of decomposition. This stability is in stark
contrast to the related bis-phosphanido complex
[Cp2Zr{P(H)Mes*}2], which could only be isolated as
an impure material and in low yields owing to its
tendency to undergo decomposition to Mes*PH2
and the cyclometallation product derived from a
zirconocene phosphinidene complex (eq. (1)) [23].
ture of 3 has tentatively been assigned as the anionic
phosphinidene complex [Cp2ZrꢁPDmp{LiP(H)Dmp}]
(Scheme 2) based on comparisons to the 31P NMR of
known anionic phosphinidene complexes ([Cp2Zrꢁ
PMes*{K(H)(THF)2}]2 l 565 ppm, and [Cp*2ZrꢁPMes-
{LiCl(DME)}] l 537.6) and independent synthesis by
reaction of 2 with 1 equiv. of BuLi. Unfortunately we
have been unable, thus far, to isolate 3 as X-ray quality
single crystals.
The structures of compounds 1 and 2 have been
investigated by a pair of single-crystal X-ray diffraction
experiments. For clarity of discussion, the results for 2
are presented first. Single crystals of 2 were grown by
pentane diffusion into a saturated THF solution of 2,
and the results of the structural determination are
presented in Fig. 1. The most significant and notewor-
thy features of 2 are the details regarding the ZrꢀP
n
The
postulated
intermediate
phosphinidene
[Cp2ZrꢁPMes*] has been trapped as the PMe3 adduct
[Cp2Zr(ꢁPMes*)(PMe3)] [8,24–27]. The disparate stabil-
ities of 2 and [Cp2Z{P(H)Mes*}2] may be a result of the
diverse nature of ZrꢀP bonding in the corresponding
structures of these materials (see below), or an inability
for 2 to attain conformations required for elimination
of ArPH2, or perhaps the presence of a more cy-
clometallation resistant ligand. Addition of PMe3 to 2
in THF or toluene leads to slow formation of
[Cp2Zr(ꢁPDmp)(PMe3)] and DmpPH2, albeit in low
yields [28]1.
,
bonding. The ZrꢀP distances of 2.519(2) and 2.726(2) A
,
are very disparate (0.207 A) and nearly span the whole
range of known ZrꢀPR2 bond lengths. The correspond-
ing two phosphorus atoms display two very different
geometries. The geometry at P1 is essentially planar as
(1)
Attempts to consume 1 completely by further in-
creasing the ratio of Li[P(H)Dmp]·Et2O to [Cp2ZrCl2]
led to increasing amounts of 3 and DmpPH2 that were
also difficult to separate from 2 by recrystallization.
Compound 3 displays 31P NMR signals at l 496
(broad) and −103 ppm (dm, JHP=263 Hz)2. The struc-
1 Identified by NMR, preliminary X-ray structure analysis, and
independent synthesis from [Cp2Zr(Me)Cl] and LiP(H)Dmp.
2 [Cp2ZrꢁPDmp{DmpP(H)Li}] (3). 1H NMR (C6D6): l 7.11–7.00
(m, 3H), 6.93 (m, 1H), 6.89 (m, 4H), 6.82 (s, 2H), 6.77–6.74 (m, 4H),
5.49 (s, 10H), 2.40 (s, 6H), 2.27 (dm, JHP=261 Hz, 1H), 2.19 (m,
24H), 2.12 (s, 6H). 31P{1H} NMR (toluene): l 496 (br s), −103 (m).
31P NMR (toluene): l 496 (br s), −103 (dm, JHP=263 Hz).
Fig. 1. Thermal ellipsoid plot for [Cp2Zr{P(H)Dmp}2] (2). Selected
,
bond distances (A) and angles (°): ZrꢀP1: 2.519(2), ZrꢀP2: 2.726(2),
P1ꢀH1: 1.17(8), P2ꢀH2: 1.36(7), P1ꢀZrꢀP2: 91.30(7), ZrꢀP1ꢀH1:
106(4), ZrꢀP1ꢀC101: 142.8(3), C101ꢀP1ꢀH1: 109(4), ZrꢀP2ꢀH2:
92(3), ZrꢀP2ꢀC201: 121.4(2), C201ꢀP2ꢀH2: 88(3).

E. Urne
;
zˇius et al. / Inorganica Chimica Acta 297 (2000) 181–190
185
Scheme 3.
gauged by the sum of the bonds angles (SBA(P1)=
358(6)°), while the geometry at P2 is pyramidal
(SBA(P2)=301(4)°). These values are mirrored by
ZrꢀPꢀC bond angles of 142.8(3) and 121.4(2)°, respec-
tively. The interesting asymmetric environments exhib-
ited by the two phosphanido ligands in
2 are
reminiscent of the structure of [Cp2Hf(PEt2)2], the first
example of a bis-phosphanido complex to exhibit this
type of structural feature [29]. This particular structure
contains both planar and pyramidal phosphorus atoms
((SBA(P1)=358.8(3)°) and (SBA(P2)=311.9(3)°) and
,
displays correspondingly short (2.488(1) A) and long
,
(2.682(1) A) HfꢀP bond lengths.
Adoption of a planar geometry for one of the phos-
phanido ligands in 2 or [Cp2Hf(PEt2)2] can be at-
tributed to the desire for the complex to achieve an 18
valence electron count configuration. However, not all
such Group IV metallocenes choose to adopt planar
phosphanido groups to achieve an 18 valence electron
count (as discussed below). A planar phosphanido
group should also be aligned perpendicular to the plane
of the three atoms in the wedge of the metallocene for
optimal p-bonding to the a1 p-accepting molecular or-
bital of the metallocene fragment (structure I, Scheme
3). The arrangement of the phosphanido unit in 2 is
offset by 18° from this orientation for maximal p-bond-
ing [30]. For some recent important discussions of
ligand lone pair p-interactions with transition metal
centers see Ref. [3].
Fig. 2. Thermal ellipsoid plot for [Cp2Zr(Cl)(P(H)Dmp)] (1). Selected
,
bond distances (A) and angles (°): ZrꢀP: 2.638(1), ZrꢀCl: 2.449(1),
PꢀC1: 1.841(4), PꢀH: 1.29(5), PꢀZrꢀCl: 97.05(4), HꢀPꢀZr: 104(2),
C1ꢀPꢀZr: 125.7(1), HꢀPꢀC1: 101(2).
the Dmp(H)P group of compound 1 models, at least in
the solid state, the transition state for interconversion
of the two types of phosphanido groups of 2.
Table 2 lists the key structural features of some
related structurally characterized Zr(IV) metallocene
complexes having a single terminal phosphanido ligand.
Analysis of these data reveals some interesting facts.
First, a range of geometries is observed for the phos-
phorus centers, ranging from essentially planar to pyra-
midal geometries. A rough correlation appears to exist
between the steric congestion about each phosphorus
center and the metal–phosphorus bond lengths. For
example, the complex having the most congested phos-
phorus center, [Cp2Zr(Cl){P(SiMe3)Mes*}], contains
the shortest ZrꢀP bond length. The shorter MꢀP dis-
tances also scale with the planarity of the phosphanido
ligand. Increasing rehybridization to a planar geometry
would allow the phosphorus centers to engage in in-
creased p-bonding to the metal center. It is well known,
however, that phosphorus(III) resists structural rear-
rangements from pyramidal to planar [31]. The data
from Table 2 suggest that such a rehybridization at
phosphorus may be facilitated by steric congestion. No
correlation between ZrꢀP and ZrꢀCl bond distances is
readily discerned that would allow a dissection of po-
The structure of compound 1 provides a comparison
of the bonding of a single Dmp(H)P group to a zir-
conocene center. The details of the molecular environ-
ments for 1 are presented in Fig. 2. Again, the most
interesting feature is the nature of the zirconium–phos-
phorus bonding. Careful examination of the structural
parameters of 1 establishes that the Dmp(H)P group of
1 is essentially intermediate to the two limiting types of
Dmp(H)P groups in 2. Firstly, the ZrꢀP bond length of
,
2.638(1) A of 1 is close to the average ZrꢀP distances in
,
2 (2.622 A). Secondly, a flattened pyramidal geometry
for the phosphanido ligand in 1 is indicated by the sum
of the bond angles at the phosphorus atom (SBA(P)=
331(3)°), which is also close to the average of the two
corresponding values in 2 (average SBA(P)=335°).
Lastly, the ZrꢀPꢀC bond angle of 125.7(1)° in 1 is
between the two corresponding values for 2. As such,

186
E. Urne; zˇius et al. / Inorganica Chimica Acta 297 (2000) 181–190
Table 2
Selected structurally characterized zirconocene(IV) and hafnocene(IV) terminal mono-phosphanido complexes
31P{1H} NMR
S
BA(P) (°)
PꢀZrꢀX angle (°)
Ref.
,
,
Phosphanido complexes
dMP (A)
dMX (A)
[Cp2Zr(Cl){P(SiMe3)Mes*}]
[Cp2Zr(Cl){P(H)Mes*}]
[Cp2Zr(Cl){P(SiMe3)2}]
[Cp*2Hf(H){P(H)Ph}] a
[Cp2Zr(CH3){P(SiMe3)2}]
[Cp2Zr(Cl)(P(H)Dmp)] (1)
[Cp2% Zr(Cl){P(H)Trip}] c
[Cp02 Zr(Cl){P(H)Cy}] d
2.541(4)
2.543(3)
2.547(6)
2.549(8)
2.629(3)
2.638(1)
2.6381(8)
2.6539(9)
2.485
2.494(3)
156
83.2
−108.9
30.5
−121.7
25.0
−4.6
71.7
359.5 (0.7)
350.5
100.39
98.61(9)
[55]
[35]
[56]
[57]
[56,61]
this work
[58]
2.432(5)
344.4 (0.8)
100.6(4)
b
b
b
2.35(6)
349.2 (0.3)
330.7 (2.8)
328.6 (2.0)
308 (2)
98(1)
2.449(1)
2.4547(6)
2.4442(9)
97.05(4)
93.55(2)
99.79(4)
[59]
a Cp*=C5Me5.
b Data unavailable owing to structural disorder.
c Cp%=C5H4Me, Trip=2,4,6-iPr3C6H2.
d Cp0=C5EtMe4.
tential competing p-bonding between zirconium from
chloride or phosphorus. Oddly, exchange of chloride
for methyl in [Cp2Zr(Cl){P(SiMe3)2}] lengthens the
complex [Cp2Zr(Cl)PH2] produced a geometry for
[Cp2Zr(Cl)PH2] having pyramidal PH2 group
a
(SBA(P)=286.3°) and ZrꢀP and ZrꢀCl bond distances
,
,
ZrꢀP bond length (0.082 A), while effecting little
of 2.669 and 2.450 A, respectively. Minimization of
change in the geometry of the phosphorus atom. Inter-
estingly, exchange of phosphorus for arsenic in
[Cp2Zr(Cl){E(SiMe3)2}] (E=P, As) results in a signifi-
cant change in ligand geometry from SBA(P)=344.4° to
[Cp2Zr(H)PH2] also favored a complex having a pyra-
midal PH2 unit (SBA(P)=293.9°) with ZrꢀP and ZrꢀH
,
bond distances of 2.656 and 1.856 A, respectively.
Minimization of [Cp2Zr(Cl)PH2] while constraining the
PH2 unit to a planar geometry led to a structure 6.8
kcal mol−1 higher in energy than the structure having a
,
BA(As)=328.8° (dZrCl=2.445(1) A, dZrAs=2.7469(7)
S
,
A) [32]. Such an effect may be the result of the larger
covalent radius of arsenic, and hence, a more relaxed
steric environment at the arsenic center as compared to
the phosphorus center.
,
pyramidal PH2 group and led to dZrP=2.520 A and
,
d
ZrCl=2.498 A.
The crystal structure of [Cp2Zr(Cl){P(H)Mes*}] [35]
The above analysis would also imply that the elec-
tronically favored geometry for zirconocene phos-
phanido groups is pyramidal and that zircon-
ium–phosphorus p-bonding alone is not a suffi-
cient driving force for planarization of phosphanido
ligands. Barrier heights for pyramidal inversion at
phosphorus in the hypothetical species [Cp2Ti(Cl)PR2]
[33]3 have been calculated using the approximate molec-
ular orbital method partial retention of diatomic differ-
ential overlap (PRDDO) [34]. Stabilization of a planar
phosphorus geometry by ligand-to-metal p-bonding
was suggested to facilitate the inversion process. A
barrier height of 6.7 kcal mol−1 was found for
[Cp2Ti(Cl)PH2] (SBA(P)=305.6° for pyramidal geome-
(4, Mes*=2,4,6-tBu3C6H2) allows a direct comparison
of the influence of the aryl groups on the geometry of
the phosphanido groups for two [Cp2Zr(Cl){P(H)Ar}]
,
complexes. The ZrꢀP bond length of 2.543(3) A in 4 is
shorter than the same bond length in 1. The ZrꢀPꢀC
angle of 128.4(2)° in 4 is slightly smaller than the
analogous angle in 1. A somewhat planar geometry for
the Mes*(H)P group in 4 is suggested by the sum of the
bond angles at phosphorus (SBA(P)=350°) and corrob-
orated by a shorter ZrꢀP bond distance. A greater
degree of planarity for 4 compared to 1 is consistent
with previous arguments that a single Mes* group
imposes greater steric distortions than a single Dmp
group [36].
try),
while
for
[Cp2Ti(Cl){P(H)Ph}]
and
The phosphanido group of 1 is approximately an
average of the two types of phosphanido groups of 2.
No such relationship connects the structures of 4 and
[Cp2Zr{P(H)Mes*}2] (5) [23]. For example, compound
5 contains two longer and nearly identical ZrꢀP bond
[Cp2Ti(Cl)PPh2] a planar phosphanido ligand was pre-
dicted to be favored over a pyramidal phosphanido
group. The latter results were ascribed to the phenyl
groups being bulkier than hydrogens and the ability of
the phenyl groups to find resonance stabilization with
the P-atom lone pair. Our preliminary calculations
,
lengths of 2.681(5) and 2.682(5) A. The two ZrꢀPꢀC
bond angles in 5 of 119.0(3) and 119.3(3)° are smaller
than the value for 4. Based on this information, the
geometry at each phosphanido group of 5 would thus
be expected to be identical, but differences in the sum
of the bond angles at each phosphorus center in 5
(SBA(P)=321(2) and 356(2)°) would suggest the pres-
ence of two types of phosphanido groups. Owing to the
using
density
functional
theory
[18,21]
on
[Cp2Zr(X)PH2] also show a preference for pyramidal
phosphanido groups. Energy minimization of a model
3 The first example of a structurally characterized simple Ti(IV)
phosphanido complex has recently been reported (see Ref. [33]).

E. Urne
;
zˇius et al. / Inorganica Chimica Acta 297 (2000) 181–190
187
Table 3
Selected structurally characterized zirconocene(IV) and hafnocene(IV) terminal bis-phosphanido complexes
31P{1H} NMR
SBA(P) (°)
PꢀZrꢀP angle (°)
Ref.
,
Phosphanido complexes
dMP (A)
[Cp2Hf{PEt2}2]
2.488(1), 2.682(1)
2.519(2), 2.726(2)
2.536(3), 2.694(3)
2.553(1), 2.654(1)
2.600(2), 2.634(2)
2.63(2)
–
17.4
144.1
−98.8
−71.2
39.0
358.8 (0.3), 311.9 (0.3)
357.8 (5.7), 301.4 (4.2)
359.9 (0.6), 326.5 (0.6)
360.0 (0.2), 336.0 (0.2)
98.64(3)
91.30(7)
103.75(9)
99.4(1)
96.95(7)
96.1(7)
[29]
this work
[60]
[61]
[62]
[Cp2Zr{P(H)Dmp}2] (2)
[(C5H4SiMe3)2Zr{PPh2}2]
[Cp2Hf{P(SiMe3}2]
[Cp2% Zr{P(SiMe3}2] a
[Cp*2Zr{P(H)Mes}2] b
[Cp2Zr{P(H)Mes*}2]
358.3 (0.1), 351.2 (0.1)
c
[27]
[23]
2.681(5), 2.682(5)
51.6
356.0 (2.8), 321.3 (2.8)
97.86(9)
a Cp%=C5H4Me.
b Cp*=C5Me5, Mes=2,4,6-Me3C6H2.
c Hydrogen atoms not located.
Fig. 3. Variable temperature 31P{1H} NMR (toluene) of [Cp2Zr{P(H)Dmp}2] (2).
Although the steric shielding provided by Dmp and
Mes* groups are close to one another, they differ
somewhat in the three-dimensional form of their screen-
ing. Both types of systems have been successfully em-
ployed to stabilize unusual environments involving
main group elements [37–41]. The particular steric
demands of the Dmp and Mes* ligands at metal centers
have been contrasted, and it has been suggested that
introduction of two bulky aryls (MAr2) produces
greater steric congestion and geometrical distortions for
compounds containing two Dmp ligands as compared
lower quality of the structure of 5 (R=0.098%) the
hydrogen atom positions may not be well defined.
Regardless, the relationship between 4 and 5 is very
different than that between 1 and 2. As pointed out
previously, p-bonding between a phosphanido group
and zirconium will require the group to reorient itself
so that the phosphorus p orbital can donate to the a1
p-accepting molecular orbital of the metallocene frag-
ment. Steric interactions in 5 between the two bulky
Mes*(H)P groups may be responsible for preventing
the alignment of one of the two groups for p-bonding.

188
E. Urne; zˇius et al. / Inorganica Chimica Acta 297 (2000) 181–190
to complexes containing two Mes* groups [36,42,43].
The aryl groups of 2 and 5 are more insulated from one
another and connected by three atoms as opposed to a
single atom, however. A potentially better comparison
to a pair of iron(II) bis-thiolate complexes bearing both
kinds of hindered aryls is clouded by the fact that the
Mes* containing species is dimeric ([{Mes*S}2Fe]2) and
the process which interconverts the two type of phos-
phanido groups to be observed on the NMR time-scale
at −126°C (l 270.2 and −15.3 ppm) (Scheme 3) [29].
The asymmetry of the phosphanido groups of 2 in
solution is also not evinced by room temperature
31P{1H} NMR spectroscopy. Upon cooling the reso-
nance broadens and at about −60°C disappears into
the baseline (Fig. 3). At about −109°C very broad
signals centered at l 149, 25 and −130 ppm appear in
an approximate 0.36:0.30:0.34 ratio. Calculation of the
weighted chemical shift for this distribution of species
gives l 17, in exact agreement with the room tempera-
ture shift for 2. Solubility problems prevented further
measurements at reduced temperatures. The signals of
equal magnitude at l 149 and −130 are assigned to the
planar and pyramidal phosphorus atoms of 2, respec-
tively [1]. The signal at l 25 is tentatively assigned as a
second isomer of 2 (2%, Scheme 4) having two equivalent
Dmp(H)P groups with geometries comparable to those
found in 1 (31P{1H} NMR: l 25). Thus, compound 2,
unlike other structurally characterized metallocene(IV)
bis-phosphanido complexes, appears to have a delicate
balance of steric and electronic forces to afford signifi-
cant concentrations of the two forms of 2. Both isomers
2 and 2% could have stereoisomers, owing to the pres-
ence of chiral pyramidal phosphanido groups [8,24–27],
but the range of chemical shifts is probably too large to
be explained by such stereoisomers. Low-temperature
31P NMR spectroscopy of [Cp2Zr{P(H)Mes*}2] showed
only a single resonance down to −80°C [23]. Variable-
the
Dmp
containing
species
is
monomeric
[{DmpS}2Fe]) [44,45]. Comparisons can also be drawn
to phosphorus atoms double bonded directly to one
another in the diphosphenes DmpPꢁPDmp and
Mes*PꢁPMes*. The PꢁP bond length in the former
material is slightly shorter than in the latter (1.985(2)
,
vs. 2.034(2) A) suggesting that a closer approach of two
DmpP units is easier than for two Mes*P units [17,46].
A
full analysis will require additional structural
determinations.
Further data for some Zr(IV) and Hf(IV) metal-
locenes bearing two terminal phosphanido ligands are
presented in Table 3. A greater range of MꢀP bond
lengths and ligand geometries is observed for these
complexes than in analogous mono-phosphanido spe-
cies. Again, the phosphanido groups displaying shorter
MꢀP bonds are more planar. As observed for the
mono-phosphanido complexes, as the phosphorus cen-
ters become increasingly planar, the metal–phosphorus
bond lengths decrease. Notably, the range of metal–
phosphorus bond lengths spans a greater range of
values for the bis-phosphanido complexes than for the
mono-phosphanido complexes. The less direct correla-
tion of sterics to the type of bonding mode for bis-
phosphanido metallocenes may reflect the difficulty for
a planar phosphanido group to orient itself so that the
phosphorus p-orbital may be in alignment to the a1
p-accepting molecular orbital of the metallocene frag-
ment, as proposed for [Cp*Zr{P(H)Mes)2}] [8,24–27].
Further computational studies will be needed to exam-
ine these issues.
1
temperature H NMR, however, did show evidence for
hindered PꢀC bond rotation and inversion at
phosphorus.
The importance of steric congestion for inducing
planar geometries and p-bonding for P(III) has been
previously noted in the chemistry of phospholes. These
five-membered heterocycles (below) might be expected
to be aromatic if a planar geometry is adopted by the
phosphorus center. The phospholes that exhibit the
greatest degree of aromaticity and the lowest barriers to
pyramidal inversion (interconversion of A and A%),
however, are the ones containing the most sterically
demanding groups (such as R=Mes*) [47,48]. The role
of extreme sterics has been further noted in select
phosphametallacycles where delocalization of the phos-
phorus lone pair is unlikely [25,49,50]. On the other
The fascinating display of two types of phosphanido
groups in 2 in the solid state is shared by several of the
other complexes in Table 3. Detecting this phenomenon
in solution however is not always possible. Low-tem-
perature 31P NMR studies of [Cp2Hf(PEt2)2], for exam-
ple, revealed a single time-averaged signal even upon
cooling. Replacing the diethylphosphanido groups by
more bulky dicyclohexylphosphanido groups allowed
Scheme 4.

E. Urne
;
zˇius et al. / Inorganica Chimica Acta 297 (2000) 181–190
189
References
hand, pyramidal inversion of phosphines can be pro-
moted by main group Lewis acids by a mechanism
involving backside attack of empty p-orbital of the
main group species on the smaller lobe of the phospho-
rus lone pair. However, this inversion leads to coordi-
nation of the phosphorus center (see Ref. [50]).
[1] E. Hey-Hawkins, Chem. Rev. 94 (1994) 1661–1717.
[2] D.W. Stephan, Coord. Chem. Rev. 95 (1989) 41–107.
[3] N. Etkin, M.C. Fermin, D.W. Stephan, J. Am. Chem. Soc. 119
(1997) 2954–2955.
[4] M.C. Fermin, D.W. Stephan, J. Am. Chem. Soc. 117 (1995)
12645–12646.
[5] D.W. Stephan, Angew. Chem., in press.
[6] A.H. Cowley, Acc. Chem. Res. 30 (1997) 445–451.
[7] T.L. Breen, D.W. Stephan, Organometallics 16 (1997) 365–369.
[8] T.L. Breen, D.W. Stephan, J. Am. Chem. Soc. 117 (1995)
11914–11921.
[9] J.B. Bonanno, P.T. Wolczanski, E.B. Lobkovsky, J. Am. Chem.
Soc. 116 (1994) 11159–11160.
[10] C.C. Cummins, R.R. Schrock, W.M. Davis, Angew. Chem., Int.
Ed. Eng. 32 (1993) 756–759.
[11] A.H. Cowley, A.R. Barron, Acc. Chem. Res. 21 (1988) 81–87.
[12] F. Mathey, Angew. Chem., Int. Ed. Eng. 26 (1987) 275–286.
[13] G. Huttner, K. Evertz, Acc. Chem. Res. 19 (1986) 406–413.
The conclusion suggested by the above analyses is
that zirconium(IV) or hafnium(IV) metallocenes con-
taining planar phosphanido groups may require the
assistance of sterically encumbering group(s) to per-
suade the phosphorus center to adopt an undesired
planar geometry. The zirconium or hafnium center may
take advantage of this situation and engage in p-bond-
ing to the now planarized phosphanido group. This
effect may be tempered in complexes of the form
[Cp2Zr{PRR%}2] if the presence of the second phos-
phanido group sterically hinders adoption of the proper
orientation for ZrꢀP p-bonding to occur. Previous de-
scriptions of the bonding of phosphanido ligands to
metal centers have emphasized only the importance of
the metal center to achieve an 18-electron valence
configuration. It must also be remembered that while
phosphorus(III) resists structural rearrangements from
pyramidal to planar, planar geometries for amido lig-
ands are the norm [51–54]. Planar metal amides, how-
ever, do not necessarily mandate NꢀM p-d p-bonding
[54].
[14] E. Urne; zˇius, J.D. Protasiewicz, manuscript in preparation.
[15] G.W. Rabe, S. Kheradmandan, L.M. Liable-Sands, I.A. Guzei,
A.L. Rheingold, Angew. Chem., Int. Ed. Eng. 37 (1998) 1404–
1407.
[16] G.W. Rabe, S. Kheradmandan, G.P.A. Yap, Inorg. Chem. 37
(1998) 6541–6543.
[17] E. Urne
; zˇius, J.D. Protasiewicz, Main Group Chem. 1 (1996)
369–372.
[18] A.D. Becke, J. Chem. Phys. 98 (1993) 5648–5652.
[19] W.J. Hehre, L. Radom, P.P. Schleyer, J.A. Pople, Ab Initio
Molecular Orbital Theory, Wiley, New York, 1986.
[20] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 299–310.
[21] M.J. Frisch, H.B. Schlegel, P.M.W. Gill, B.G. Johnson, M.A.
Robb, J.R. Cheeseman, T. Keith, G.A. Petersson, J.A. Mont-
gomery, K. Raghavachari, M.A. Al-Laham, V.G. Zakrzewski,
J.V. Ortiz, J.B. Foresman, J. Cioslowski, B.B. Stefanov, A.
Nanayakkara, M. Challacombe, C.Y. Peng, P.Y. Ayala, W.
Chen, M.W. Wong, J.L. Andres, E.S. Replogle, R. Gomperts,
R.L. Martin, D.J. Fox, J.S. Binkley, D.J. Defrees, J. Baker,
J.J.P. Stewart, M. Head-Gordon, C. Gonzalez, J.A. Pople,
GAUSSIAN 94, Revision C.3, Gaussian, Pittsburgh, PA, 1995.
[22] S.R. Wade, M.G.H. Wallbridge, G.R. Willey, J. Chem. Soc.,
Dalton Trans. (1983) 2555–2559.
4. Supplementary material
[23] E. Hey-Hawkins, S. Kurz, J. Organomet. Chem. 479 (1994)
125–133.
[24] T.L. Breen, D.W. Stephan, Organometallics 15 (1996) 4223–
4227.
[25] T.L. Breen, D.W. Stephan, J. Am. Chem. Soc. 118 (1996)
4204–4205.
[26] T.L. Breen, D.W. Stephan, Organometallics 15 (1996) 4509–
4514.
Full experimental details for the X-ray crystallo-
graphic analysis of 1 and 2. Crystallographic data for
the structural analysis has been deposited with the
Cambridge Crystallographic Data Centre. Copies of
this information may be obtained free of charge from:
The Director, CCDC, 12 Union Road, Cambridge,
CB21EZ,UK (fax:+44-1223-336-033; email: deposit@
ccdc.cam.ac.uk or www: http://www.ccdc.cam.ac.uk).
[27] Z. Hou, T.L. Breen, D.W. Stephan, Organometallics 12 (1993)
3158–3167.
[28] E. Urne; zˇius, J.D. Protasiewicz, unpublished results.
[29] R.T. Baker, J.F. Whitney, S.S. Wreford, Organometallics 2
(1983) 1049–1051.
Acknowledgements
[30] (a) M.C. Kuchta, G. Parkin, Coord. Chem. Rev. 176 (1998)
323–372. (b) K.G. Caulton, New. J. Chem. 18 (1994) 25–41.
[31] W. Kutzelnigg, Angew. Chem., Int. Ed. Eng. 23 (1984) 272–295.
[32] E. Hey-Hawkins, F. Lindenberg, Organometallics 13 (1994)
4643–4644.
[33] A.V. Firth, D.W. Stephan, Inorg. Chem. 37 (1998) 4732–4734.
[34] J.R. Rogers, T.P.S. Wagner, D.S. Marynick, Inorg. Chem. 22
(1994) 3104–3110.
We thank the National Science Foundation (CHE-
9733412) and the Department of Chemistry at CWRU
for support of this work. J.D.P. also wishes to acknowl-
edge Professor Stephen J. Lippard for his invaluable
crystallographic lessons and for sharing his passion for
relating chemical structure to chemical function and
reactivity.
[35] J. Ho, R. Rousseau, D.W. Stephan, Organometallics 13 (1994)
1918–1926.

190
E. Urne; zˇius et al. / Inorganica Chimica Acta 297 (2000) 181–190
[36] R.S. Simons, L. Pu, M.M. Olmstead, P.P. Power, Organometal-
lics 16 (1997) 1920–1925.
[50] M.D. Fryzuk, G.R. Giesbrecht, S.J. Rettig, Inorg. Chem. 37
(1998) 6928–6934.
[37] A.H. Cowley, N.C. Norman, Prog. Inorg. Chem. 34 (1986)
1–63.
[51] M.A. Dewey, D.A. Knight, A. Arif, J.A. Gladysz, Chem. Ber.
125 (1992) 815–824.
[38] P.J. Brothers, P.P. Power, in: R. West, F.G.A. Stone (Eds.),
Multiply Bonded Main Group Metals and Metalloids, Academic
Press, San Diego, CA, 1996, pp. 1–63.
[39] A.H. Cowley, J. Organomet. Chem. 400 (1990) 71–80.
[40] N.C. Norman, Polyhedron 12 (1993) 2431–2446.
[41] P.P. Power, J. Chem. Soc., Dalton. Trans. (1998) 2939–2951.
[42] X.-W. Li, W.T. Pennington, G.H. Robinson, Organometallics 14
(1995) 2109–2111.
[43] R.J. Wehmschulte, W.J. Grigsby, B. Schiemenz, R.A. Bartlett,
P.P. Power, Inorg. Chem. 35 (1996) 6694–6702.
[44] P.P. Power, S.C. Shoner, Angew. Chem., Int. Ed. Eng. 30 (1991)
330–332.
[45] J.J. Ellison, K. Ruhlandt-Senge, P.P. Power, Angew. Chem., Int.
Ed. Eng. 33 (1994) 1178–1180.
[46] M. Yoshifuji, I. Shima, N. Inamoto, K. Hirotsu, T. Higuchi, J.
Am. Chem. Soc. 103 (1981) 4587–4589.
[47] G. Keglevich, Z. Bo¨cskei, G.M. Keseru¨, K. Ujsza`szy, L.D. Quin,
J. Am. Chem. Soc. 119 (1997) 5095–5099.
[48] L. Nyula´szi, L. Soo´s, G. Keglevich, J. Organomet. Chem. 566
(1998) 29–35.
[52] S.C. Goel, M.Y. Chiang, D.J. Rauscher, W.E. Buhro, J. Am.
Chem. Soc. 115 (1993) 160–169.
[53] M.A. Matchett, M.Y. Chiang, W.E. Buhro, Inorg. Chem. 33
(1994) 1109–1114.
[54] M.F. Lappert, P.P. Power, A.R. Sanger, R.C. Srivastava, Metal
and Metalloid Amides, Wiley, New York, 1980.
[55] A.M. Arif, A.H. Cowley, C.M. Nunn, M. Pakulski, J. Chem.
Soc., Chem. Commun. (1987) 994–995.
[56] E. Hey, M.F. Lappert, J.L. Atwood, S.G. Bott, Polyhedron 7
(1988) 2083–2086.
[57] G.A. Vaughan, G.L. Hillhouse, A.L. Rheingold, Organometal-
lics 8 (1989) 1760–1765.
[58] E. Hey-Hawkins, S. Kurz, G. Baum, Z. Naturforsch., Teil B 50
(1995) 239–242.
[59] U. Segerer, E. Hey-Hawkins, Polyhedron 16 (1997) 2537–2545.
[60] A.-M. Larsonneur, R. Choukroun, J.–C. Daran, T. Cuenca, J.C.
Flores, P. Royo, J. Organomet. Chem. 444 (1993) 83–89.
[61] L. Weber, G. Meine, R. Boese, N. Augart, Organometallics 6
(1987) 2484–2488.
[62] F. Lindenberg, E. Hey-Hawkins, J. Organomet. Chem. 435
(1992) 291–297.
[49] T.L. Breen, D.W. Stephan, Organometallics 15 (1996) 5729–
5737.
.