Synthesis and reactivity of the first η6-rhodium(I) and η6-iridium(I) complexes of 2,6-bis(trimethylsilyl)phosphinines

Article abstract of DOI:10.1002/ejic.200390095

It has been shown that 2 equiv. of 2,3,5,6-tetraphenylphosphinine (2) react with [Rh(COD)₂][BF₄] to yield the bis(η¹-phosphinine)Rh^I complex 6, whose X-ray crystal structure is presented. On the other hand, 2,6-bis(trimethylsilyl)phosphinines 4 and 5 react with Rh⁺ and Ir⁺ precursors to yield the first (η⁶-phosphinine)Rh^I and -Ir^I complexes 7-10, and 11-12, respectively. The X-ray crystal structure of complex 8 is presented. Reaction of these η⁶-phosphinine complexes with water or ethanol yields the first (ν⁵-phosphacyclohexadieny1)Rh^I and -Ir^I complexes 13-17, resulting from the formal 1,1-addition of RO^- and H⁺. DFT calculations comparing isoelectronic (η⁶-phosphinine)- and (η⁶-benzene)Fe⁰ and -Rh^I complexes allows the rationalization of the large difference in reactivity of these complexes. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejic.200390095

FULL PAPER
Synthesis and Reactivity of the First η6-Rhodium(
I
) and η6-Iridium(
)
I
Complexes of 2,6-Bis(trimethylsilyl)phosphinines
Marjolaine Doux,[a] Louis Ricard,[a] Francois Mathey,[a] Pascal Le Floch,*[a] and
¸
[a]
´
Nicolas Mezailles*
Keywords: Phosphinines / P ligands / Pi interactions / Rhodium / Iridium
It has been shown that 2 equiv. of 2,3,5,6-tetraphenylphos-
phinine (2) react with [Rh(COD)2][BF4] to yield the bis(η1-
phosphinine)RhI complex 6, whose X-ray crystal structure is
presented. On the other hand, 2,6-bis(trimethylsilyl)phosphi-
water or ethanol yields the first (η5-phosphacyclohexadien-
yl)RhI and -IrI complexes 13−17, resulting from the formal
1,1-addition of RO− and H+. DFT calculations comparing iso-
electronic (η6-phosphinine)- and (η6-benzene)Fe0 and -RhI
nines 4 and 5 react with Rh+ and Ir+ precursors to yield the complexes allows the rationalization of the large difference
first (η6-phosphinine)RhI and -IrI complexes 7−10, and 11−12,
in reactivity of these complexes.
respectively. The X-ray crystal structure of complex 8 is pre- ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
sented. Reaction of these η6-phosphinine complexes with
Germany, 2003)
Introduction
be a versatile and effective catalyst for a number of car-
bonylation reactions[9] and for the silylformylation of al-
kynes.[10] More recently, analogous [(η6-benzene)Ir]ϩ com-
plexes were obtained by Oro et al. on reduction of an [(η6-
benzene)Ir(H)2]ϩ complex by ethylene.[11]
Therefore, we were interested in developing a method
leading to the first η6-phosphinine complexes with Group-
9 metal centers. We now wish to report on the syntheses
of the [(η6-phosphinine)M]ϩ (M ϭ Rh, Ir) complexes, and
present DFT calculations comparing the isoelectronic (pho-
sphinine)- and (benzene)Fe0 and -RhI derivatives. Studies
on their reactivity, leading to the first (η5-phosphacyclohex-
adienyl)RhI and -IrI complexes are also presented.
The coordination chemistry of phosphinines has been
studied for many years now. These ligands usually bind to
electron-rich metal centers through the lone pair on the
phosphorus atom.[1] Comparatively, fewer η6 complexes
have been synthesized with these ligands. The sandwich bis-
η6 complexes were usually synthesized by co-condensation
techniques between the ligand (in large excess) and the va-
porized metal() (Ti, V, Cr, Ho).[2] The Group-6 mono-η6
complexes were synthesized either by starting with an η1-
coordinated phosphinine, or by a phosphinine bearing two
bulky groups at the ortho positions.[3] The only Group-7 η6
complex was obtained by photolysis of a mixed (car-
bonyl)(η1-phosphinine) complex.[4] For Group 8, the zero-
valent Fe complexes were obtained by Zenneck et al. in the
mid 1990s,[5] and very recently we reported on the syntheses
of [(η6-phosphinine)RuCp*] complexes.[6] Most interest-
ingly, [(η6-phosphinine)Fe(COD)] complexes were found to
be active in the cyclotrimerization of alkynes and in the
synthesis of pyridine derivatives.[5a]
Results and Discussion
The syntheses of (phosphinine)Fe0 complexes relies on
the co-condensation of Fe0 with phosphinine and COD.
Obviously, this method can not be duplicated to yield the
The [(η6-arene)Rh]ϩ complexes have been known for
more than 30 years now and were first synthesized by
Haines et al. in 1969.[7] Since then, a number of [(η6-
arene)Rh(L)(LЈ)]ϩ have been isolated.[8] Among them, the
zwitterionic complex [(η6-PhBPh3)Rh(COD)] appeared to
[a]
´
´ ´
Laboratoire ‘‘Heteroelements et Coordination’’, UMR CNRS
7653, Ecole Polytechnique,
91128 Palaiseau Cedex, France
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
Scheme 1
Eur. J. Inorg. Chem. 2003, 687Ϫ698  2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim 1434Ϫ1948/03/0204Ϫ0687 $ 20.00ϩ.50/0
687

´
M. Doux, L. Ricard, F. Mathey, P. Le Floch, N. Mezailles
FULL PAPER
desired isoelectronic RhI and IrI complexes. Two different
methods were thus devised. The first one, Method A, relies
on the chloride abstraction of [M(diene)Cl]2 (M ϭ Rh, Ir;
diene ϭ COD, NBD) by GaCl3 at low temperatures, fol-
lowed by the coordination of the phosphinine. It is hoped
that the phosphinine adopts a η6-coordination mode.
Method B relies on the classical displacement of one diene
ligand from the [Rh(diene)2][BF4] (diene ϭ COD, NBD)
precursor by phosphinine (Scheme 1).
(2)
Moreover, since the η1 complexes are usually favored, it
was necessary to study the different factors that lead to the
η6 complexes, namely the nature of the metal precursor and
the substitution pattern of the ligand. In order to test the
importance of this second factor, we chose four different
phosphinines, 2Ϫ5. Among these, two had not been previ-
ously reported (2Ϫ3) [Equation (1)].
This complex was fully characterized by usual NMR
techniques as well as by elemental analysis. In the 31P NMR
spectrum, the complex is characterized by a doublet at δ ϭ
175 ppm (coordination chemical shift ∆δ ϭ Ϫ33 ppm) and
a large coupling constant 1JRh,P of 166.5 Hz, clearly indicat-
ing the η1-coordination mode.[1] This coordination mode
was confirmed by the crystal structure analysis of 6. An
ORTEP drawing of 6 is presented in Figure 1, along with
selected bond lengths and angles. As expected, the structure
reveals a square-planar geometry around the rhodium
˚
center. The RhϪP bond lengths are 2.281 and 2.301 A, and
are typical, as are the CϭC bond lengths within the ring
˚
(1.386Ϫ1.410 A). The two CϪPϪC internal angles of the
phosphinine ligands are 106.4 and 107.0°, and fall in the
usual range for coordinated phosphinine. The CϭC bond
˚
lengths of the COD moiety [1.353(7) and 1.388(7) A] com-
pare with those in the isoelectronic Fe0 complex.[5a]
(1)
The phosphinines were synthesized according to the
method developed several years ago in our laboratories.[12]
This method relies on the high reactivity of the 1,3,2-diaza-
phosphinine 1 with alkynes. In a first step, 1 reacts with 1
equiv. of the first alkyne in a [4ϩ2]/retro [4ϩ2] sequence to
produce a 1,2-monoazaphosphinine, which is not isolated,
but whose complete formation is checked by 31P NMR
spectroscopy. This intermediate then reacts with the second
alkyne, at higher temperature, in another [4ϩ2]/retro [4ϩ2]
sequence to yield the desired phosphinine. The phosphin-
ines are then easily purified by column chromatography.
Phosphinines 2 and 3, which are white solids, were fully
characterized by usual NMR spectroscopic techniques and
elemental analysis.
Figure 1. ORTEP view of complex 6; ellipsoids are scaled to en-
close 50% of the electron density; phenyl groups have been omitted
˚
for clarity; selected bond lengths [A] and angles [°]; numbering is
arbitrary and differs from the ones used in NMR spectra: Rh1ϪP1
2.301(1), P1ϪC1 1.738(5), P1ϪC5 1.738(5), C1ϪC2 1.407(6),
C2ϪC3 1.386(7), C3ϪC4 1.391(7), C4ϪC5 1.410(6), Rh1ϪP2
2.281(1), P2ϪC6 1.721(5), P2ϪC10 1.740(5), C6ϪC7 1.397(7),
C7ϪC8 1.388(7), C8ϪC9 1.396(7), C9ϪC10 1.405(6), C63ϪC64
1.353(7), C59ϪC60 1.388(7); C5ϪP1ϪC1 106.4(2), C6ϪP2ϪC10
107.0(2)
A first series of experiments showed that the 2,6-diphenyl
derivatives only led to the formation of the bis(η1) complex,
irrespective of the method or the number of equivalents of
phosphinine used, as indicated by 31P NMR spectroscopy
When a bulkier trimethylsilyl group replaces one phenyl
[Equation (2)]. Thus, 2 equiv. of the 2,3,5,6-tetraphenyl de- group, such as in 3, a different outcome is observed. In fact,
rivative 2 reacts with [Rh(COD)2]ϩ to quantitatively yield the two methods yield different products. Method A leads
complex 6.
to a mixture of two η1 complexes as can be seen by 31P
Eur. J. Inorg. Chem. 2003, 687Ϫ698
688

The First η6-RhI and η6-IrI Complexes of 2,6-Bis(trimethylsilyl)phosphinines
FULL PAPER
NMR spectroscopy, corresponding to different stereoiso-
As can be predicted, the first method is faster and the
mers formed as a result of the bulkiness of the ligand which reaction is completed between Ϫ78 °C and room temper-
probably precludes their free rotation [Equation (3)].
ature, whereas the second method requires a few hours of
heating at 40 °C. The main NMR spectroscopic character-
istics of the ligands and of these six complexes are summar-
ized in Table 1.
As can be seen from the data, all the complexes are char-
acterized by a strong shift of the signals in the 31P NMR
spectra (∆δ ranges from Ϫ150 to about Ϫ200 ppm). Unlike
the η1 coordination of phosphinines to Rh (1JRh,P
ഠ
160 Hz), the signals for complexes 7 and 8 appear as weakly
coupled (small Rh-P coupling constant of 7.6 and 9.1 Hz,
respectively), or as singlets for 9 and 10. The signals for the
iridium complexes 11 and 12 appear as singlets but at a
much higher field than the Rh analogs (∆δ ഠ Ϫ50 ppm).
The strong shift of the signal for 4-H in the 1H NMR spec-
tra is also indicative of the η6-coordination mode. This
ranges from δ ϭ 5.7 ppm in 7 to δ ϭ 6.7 ppm in 9 (com-
pared with δ ϭ 7.1 ppm for 4). In the 13C NMR spectra,
the most significant change on η6-coordination is indicated
by the chemical shift of the C-4 signal. In the Rh complexes,
(3)
However, when 3 was treated with [Rh(COD)2][BF4] at
room temperature for 4 h, the desired η6 complex was
formed [Equation (4)] as the major product (about 60% by
1H NMR spectroscopy). The signal for the complex appears
as a doublet which is strongly shifted to a higher field (δ ϭ
1
84 ppm, ∆δ ϭ Ϫ124 ppm, JRh,P ϭ 6 Hz). The very small
coupling constant between the phosphinine and the rhod-
ium center is also very indicative of this type of coordina-
3
tion. Nevertheless, a mixture of the above-mentioned η1
it appears as a doublet of doublets with JC,P of about
10 Hz (compared with 3JC,P ϭ 20 Hz in the free ligand) and
1
complexes (sets of doublets at δ ഠ 190 ppm, JRh,P
ഠ
3
1JC,Rh of about 2 Hz. The same JC,P value is observed in
160 Hz) was also formed during the reaction (about 40%
1
the Ir complexes 11 and 12. The π coordination of the phos-
phinine was confirmed by two X-ray crystal analyses. Yel-
low crystals of 7 and 8 were obtained by slow diffusion of
hexane into a CH2Cl2 solution of the complex. As the two
structures are very similar, only the ORTEP plot of complex
8 is shown in Figure 2, together with selected bond lengths
and bond angles. The ORTEP plot of complex 7 can be
found in the Supporting Information (see footnote on the
first page of this article).
by H NMR spectroscopy).
The bonds between P and Rh are rather long [2.464(1)
(4)
˚
˚
A] compared with 2.279 and 2.304 A in 6. It should be
˚
noted that the bond between Rh and C3 (2.188 A) is much
˚
shorter, by more than 0.18 A, than the bonds between the
Unfortunately, due to the large amount of by-products,
the isolation of the pure η6 complex was not achieved. In a
third attempt, 2,6-bis(trimethylsilyl) derivatives were used.
The two different methods depicted above led to the desired
η6 complexes [Equation (5)]. It appears then that two very
bulky groups are needed to favor η6-versus η1 coordination.
As can be seen with the syntheses of complexes 11 and 12,
this approach can be successfully extended to the Ir derivat-
ives.
rhodium atom and the other carbon atoms. Note that the
atoms labeled C3 in the different crystal structures are
named C-4 in the NMR discussions. A close inspection of
the CϪC bond lengths in these two complexes shows that
what can be perceived as a minor change (Me or Ph sub-
stituents) can induce significant changes in the complexes.
˚
There are two long [C1ϪC2 (1.414 A) and C4ϪC5 (1.420
˚
˚
A)] bonds and two short [C2ϪC3 (1.409 A) and C3ϪC4
˚
(1.392 A)] bonds in 7 (R ϭ Me). The reverse is seen in 8
(R ϭ Ph). The aromaticity of the phosphinine ring does not
seem to be significantly perturbed by the coordination to
the Rh center, as indicated by the typical bond lengths and
angles. These values compare with those found in non-
strained macrocycles recently synthesized.[13]
A study of the reactivity of the series of six η6-phosphin-
ine complexes was undertaken. First, the stability of the
complexes towards water and ethanol was tested. All of the
complexes were quantitatively converted, within minutes at
room temperature and with 5 equiv. of water or ethanol, to
the previously unknown (η5-phosphacyclohexadienyl)rhod-
(5) ium and -iridium complexes (Scheme 2). As all the com-
Eur. J. Inorg. Chem. 2003, 687Ϫ698
689

´
M. Doux, L. Ricard, F. Mathey, P. Le Floch, N. Mezailles
FULL PAPER
Table 1. NMR characteristics of ligands and complexes, in CDCl3 unless otherwise stated
31P NMR
1JP,Rh
1H NMR
δ4-Η
4JH,P
13C NMR
2JC,P/1JC,Rh
δ
δC-2,6
1JC,P/1JC,Rh
δC-3,5
δC-4
3JC,P/1JC,Rh
4
265.0
269.4
101.4
103.2
107.7
116.6
60.4
Ϫ
Ϫ
7.6
9.1
0
0
Ϫ
Ϫ
7.1
7.3
5.9
5.7
6.7
6.6
6.0
6.0
0
0
0
1.3
0
1.5
0
0
164.0
166.5
134.5
141.1
130.3
132.2
84.1
89.1
76.5/1.7
104.5/0
101.2/3.0
104.6/2.3
150.4
154.9
131.5
137.9
127.9
133.2
11.0
10.8
2.8/1.7
0/0
3.2/0
5.7/2.3
133.9
133.4
97.1
94.1
101.5
100.7
22.8
21.3
8.6/3.2
10.4/3.5
12.9/2.3
12.0/2.3
5[a]
7
8
9
10
11[b]
12[b]
123.7
102.1
127.7
5.4
90.7
10.3
[c]
[c]
[c]
[c]
[c]
[c]
58.1
[a]
[b]
[c]
In C6D6. In CD2Cl2. Hydrolyzed during 13C NMR recording.
spectively, despite the coordination to an Rh center. The
reaction of 8 with ethanol yields complex 16 which displays
a signal at a lower field (δ ϭ Ϫ8.5 ppm). The same trend is
observed with the Ir complexes. For all of these complexes,
when 31P NMR spectroscopy is coupled with 1H NMR
spectroscopy, the singlets split into doublets, with very large
1JH,P coupling constants of 586 Hz (13) and 626 Hz (17).
1
This proves the R2P(OR)H structure. When the H NMR
characteristics of these complexes are compared with the
starting η6-phosphinine complexes, it should be noted that
the signal for 4-H is not significantly perturbed, showing
that C-4 is still attached to the Rh center. This is also ob-
served in the 13C NMR spectra. In fact, the signals for C-
4 and C-3,5 are only slightly shifted: δ ϭ 91.8 ppm and
125.9 ppm, respectively, in 13, compared with δ ϭ 97.1 and
131.5 ppm, respectively, in 7. However, the signal for C-2,6
is very significantly shifted in the new complexes: δC-2,6
ranges from δ ϭ 134.5 ppm in 7 to δ ϭ 69.7 ppm in 13.
These data suggest the proposed new η5-coordination
mode. The crystal structures of 14, 15 and 17 confirm this
coordination mode. Suitable crystals were obtained by slow
diffusion of hexane into THF solutions of the complexes.
ORTEP views of 14 and 17 are presented in Figures 3 and
4, together with selected bond lengths and angles. The OR-
TEP view of complex 15 is very similar to that of 14, and
can be found in the Supporting Information.
Figure 2. ORTEP view of complex 8; ellipsoids are scaled to en-
˚
close 50% of the electron density; selected bond lengths [A] and
angles [°]; numbering is arbitrary and differs from the ones used
in NMR spectra: Rh1ϪP1 2.464(1), Rh1ϪC1 2.388(4), Rh1ϪC2
2.366(4), Rh1ϪC3 2.188(4), Rh1ϪC4 2.325(4), Rh1ϪC5 2.437(4),
P1ϪC1 1.777(4), P1ϪC5 1.772(5), C1ϪC2 1.389(6), C2ϪC3
1.427(6), C3ϪC4 1.431(6), C4ϪC5 1.392(6), Si1ϪC1 1.920(4),
Si1ϪC6 1.858(6), Si1ϪC7 1.863(6), Si1ϪC8 1.861(6), C24ϪC25
1.402(7), C28ϪC29 1.400(7); C5ϪP1ϪC1 105.8(2), P1ϪC1ϪSi1
114.0(2), P1ϪC5ϪSi2 112.3(2)
plexes behave similarly, only three were fully characterized
in their reaction with water, and two in their reaction with
ethanol.
The most significant NMR spectroscopic data of the
complexes 13Ϫ17 are collected in Table 2. In the 31P NMR
spectra of complexes 13 and 14, the signals for both com-
plexes appear as singlets at δ ഠ Ϫ19.8 and Ϫ16.2 ppm, re-
From the three structures it can be seen that the phos-
phorus atom adopts a pyramid-type configuration. The η5
coordination of the carbon part of the ring is also apparent.
The presence of an OϪGaCl3 bond in 14 and 15 can be
rationalized by the reaction of the transient PϪOH with
GaCl4Ϫ, which releases HCl [Equation (6)]. A similar type
Scheme 2
690
Eur. J. Inorg. Chem. 2003, 687Ϫ698

The First η6-RhI and η6-IrI Complexes of 2,6-Bis(trimethylsilyl)phosphinines
Table 2. NMR characteristics of complexes 13Ϫ17, in CDCl3
FULL PAPER
31P
1H
13C
JC,P/JC,Rh
δ
δ4-Η
4JH,P
δPH
1JH,P
δC-2,6
JC,P/JC,Rh
δC-3,5
δC-4
JC,P/JC,Rh
13
14
15
16
17
Ϫ19.8
Ϫ15.3
Ϫ5.8
Ϫ8.5
Ϫ1.8
5.04
5.43
5.9
5.77
6.42
2.7
3.2
3.2
3.0
1.6
6.69
7.13
6.77
7.25
6.90
586
584
590
621
626
69.7
67.5
51.5
63.8
48.0
59.4/3.6
51.1/0
broad
42.7/3.0
45.6
125.9
129.0
112.0
133.1
114.5
3.6/3.6
4.6/4.6
5.7
2.9/2.9
4.8
91.8
92.9
98.6
92.0
98.9
29.0/3.6
25.4/2.7
27.6
22.7/3.5
30.3
of reactivity has been observed between a [CpFe(η6-
phosphinine)][AlCl4Ϫ] complex and water.[14]
(6)
Apart from these features, the CϪC and RhϪC bond
lengths are very similar to those of the η6 complexes. In the
final products, a formal 1,1-addition of water or methanol
takes place at the phosphorus atom. The mechanism most
likely involves the reaction of the nucleophile at the electro-
positive phosphorus center, followed by the reaction of the
electrophile. This type of mechanism has been reported by
Dimroth in 1983 for the syntheses of tricarbonyl[η5-(1λ5-
Figure 3. ORTEP view of complex 14; ellipsoids are scaled to en-
˚
close 50% of the electron density; selected bond lengths [A] and
angles [°]; numbering is arbitrary and differs from the ones used
in NMR spectra: Rh1ϪP1 2.7959(7), Rh1ϪC1 2.312(2), Rh1ϪC2
2.285(2), Rh1ϪC3 2.246(2), Rh1ϪC4 2.357(2), Rh1ϪC5 2.442(2),
P1ϪC1 1.759(2), P1ϪC5 1.756(2), P1ϪO1 1.542(2), C1ϪC2
1.433(3), C2ϪC3 1.417(3), C3ϪC4 1.442(3), C4ϪC5 1.413(3),
C24ϪC25 1.407(3), C28ϪC29 1.411(3); C5ϪP1ϪC1 107.0(1)
phosphinine)]metal(0) complexes (metal
ϭ
Group-6
metal).[15]
Having tested the reactivity of the complexes towards
protic sources, we then investigated the ease with which the
diene moiety can be displaced. This factor is a prerequisite
for the catalytic behavior of these complexes. Unfortunately,
the reaction of the η6 complexes with four-electron donor
ligands, such as bipy or dppe, or even two-electron donor
ligands, such as phosphanes, only led to the displacement
of the phosphinine moiety within a few hours. In fact, this
behavior is not unprecedented, and Valderrama et al. noted
in 1982 that the arene ring in [Rh(C6Me6)(CO)2][PF6] was
very labile.[8g] It is not only displaced by bipy (instantan-
eously) but also by acetone or acetonitrile. In 2000, Werner
et al. synthesized various [(η6-arene)Rh(PiPr3)(η2-COE)]ϩ
complexes with electron-rich and electron-poor arenes.
They showed that both the alkene and the arene were dis-
placed by ligands such as alkynes.[8m] Therefore, we tried
to hydrogenate the diene moiety of these complexes under
various pressures of H2 in the presence of two-electron
donors, hoping that hydrogenation would be faster than
displacement of the phosphinine. Unfortunately, the typical
solvent in which hydrogenation is fast (MeOH), destroys
the η6 coordination of the phosphinine, as shown above. In
other solvents, the hydrogenation was not fast enough and
displacement of the phosphinine always occurred.
Figure 4. ORTEP view of complex 17; ellipsoids are scaled to en-
˚
close 50% of the electron density; selected bond lengths [A] and
angles [°]; numbering is arbitrary and differs from the ones used in
NMR spectra: Ir1ϪP1 2.755(1), Ir1ϪC1 2.242(3), Ir1ϪC2
2.265(3), Ir1ϪC3 2.262(3), Ir1ϪC4 2.250(3), Ir1ϪC5 2.362(3),
P1ϪC1 1.751(3), P1ϪC5 1.735(3), P1ϪO1 1.592(2), C1ϪC2
1.447(4), C2ϪC3 1.410(4), C3ϪC4 1.428(4), C4ϪC5 1.441(4),
C16ϪC17 1.430(4), C20ϪC21 1.408(5); C5ϪP1ϪC1 106.4(2)
The high sensitivity of the above-mentioned complexes
towards nucleophilic attack prompted us to theoretically in-
vestigate, by DFT, the nature of the interaction between the
Eur. J. Inorg. Chem. 2003, 687Ϫ698
691

´
M. Doux, L. Ricard, F. Mathey, P. Le Floch, N. Mezailles
FULL PAPER
phosphinine ligand and the metal fragment. Considering noted in the structures of 7Ϫ9). This distortion had already
the important difference in stability between cationic been observed in [Rh(η6-arene)(olefin)L] complexes (L ϭ
Group-9 complexes and their isoelectronic neutral iron() 2e-donor ligand).
complexes synthesized by Zenneck et al., we extended our
The most significant information is given by the NBO
investigation to the η6-Fe0 species. Furthermore, in order to analysis,[16] which reveals that in rhodium complexes I and
draw a comparison between phosphinines and arenes, the II, the arene ligands are electronically deficient (see
corresponding benzene derivatives were also studied. There- Table 4). Thus, the sum of the NBO charges at the ligand
fore, a series of four complexes was studied: the [Rh(η6- reveals that ϩ0.340 e has been transferred to the metallic
arene)(C2H4)2]ϩ cationic species I (arene ϭ C6H6) and II fragment in I, and ϩ0.423 e in the case of the phosphinine
(arene ϭ C5H5P), and the neutral species [Fe(η6-ar- complex II. Most significant is the fact that in the coordin-
ene)(C2H4)2] III (arene ϭ C6H6) and IV (arene ϭ C5H5P) ated phosphinine ligand, the phosphorus atom now bears a
(Scheme 3).
more important positive charge of ϩ0.825, vs. ϩ0.658 in
the free ligand (calculated at the same level of theory). This
simple comparison accounts for the higher reactivity of
these η6-ligated phosphinines towards nucleophilic attack.
The NBO analysis also reveals that the Rh atom is more
positively charged in the benzene complex I (ϩ0.404) than
in II (ϩ0.274). In line with experimental observations (e.g.
stability towards hydrolysis), the charge distribution pattern
of the phosphinine ring is not significantly perturbed by
coordination to the zero-valent [Fe(C2H4)2] fragment, and
the NBO charges compare with those calculated for the free
ligand. Thus, in complex IV the phosphorus atom bears a
charge of ϩ0.685, and the sum of the charges on all the
atoms of the ring is close to zero (ϩ0.035). Note that in III
and IV a significant amount of electron-density has now
been delocalized on the ethylene ligands, as shown by the
sum of the charges (Ϫ0.204 in III and Ϫ0.160 in IV).
Interestingly, bond dissociation energies indicate that, al-
though the bond strength of the ethylene ligands are quite
similar in both types of complexes, benzene and phosphin-
ine are more firmly bound to Fe in III and IV, than to Rh
in I and II. This correlates very well with the experimental
observations that shows that substitution of the ethylene
ligands can occur in the case of Fe0 complexes, but com-
Scheme 3
All the geometric parameters of the calculated structures
IϪIV are summarized in Table 3.
Table 3. Calculated geometries and binding energies De (non-ZPE-
corrected) of complexes IϪIV
I[a]
II[a]
1.781
III[a]
IV[a]
1.776
P1ϪC2
C1ϪC2
C2ϪC3
C3ϪC4
P1—M
C1—M
C2—M
C3—M
C4—M
C5ϪC6
C7ϪC8
MϪC5
1.421
1.402
1.418
1.405
1.396
1.423
2.617
1.399
1.417
2.481
2.365
2.444
2.135
2.166
2.474
2.431
2.329
1.399
1.400
2.197
2.197
57.21
61.19
58.44
2.217
2.173
2.135 petes with that of the arene group in cationic RhI com-
1.402
2.179
1.406
2.071
1.402
1.403
plexes. The bond dissociation energies are reported in
Table 3. These data suggest that the contribution of the
2.091
benzene and phosphinine ligands is nearly equivalent in
2.086
MϪC7
these two types of complexes.
De (ring)
De (C5ϪC6)
De (C7ϪC8)
55.86
60.60
57.17
48.12
60.55
45.24
46.39
Therefore, to gain a deeper understanding of the
areneϪmetal interactions in these complexes, a CDA ana-
lysis of IϪIV was undertaken.[17] This method developed
by Frenking and co-workers already proved to be a power-
[a]
Binding energies are expressed in kcal/mol.
A first remark concerns the overall geometries of the cal- ful tool in analyzing donorϪacceptor interactions in terms
culated complexes that are analogous to that of the experi- of the common DewarϪChattϪDuncansson (DCD) model
mental structure. In each case, the D2 (phosphinine concepts. The results given by the CDA analysis are sum-
complexes) and C2v (benzene complexes) symmetries corre- marized in Table 5. We did not limit our investigation to
spond to a minimum on the potential energy surface. Some the arene ligands, and the analysis of the ethyleneϪmetal
discrepancies can be noted between the experimental and interaction was also carried out. Care was taken to differen-
theoretical structures in the case of the (phosphinine)Fe tiate between the ethylene ligands in the phosphinine com-
and -Rh complexes; however, as mentioned above in the plexes II and IV, for obvious symmetry reasons.
discussion of experimental structures of complexes 7 and 8,
A first important point concerns the residual term (∆),
the difference in the substitution pattern of the ring plays a which is important in analyzing the donorϪacceptor inter-
very important role. Note that the nature of the olefin most action. In all cases, the values calculated are small and com-
likely plays a role in these distortions. The same can be said pare with those already reported in the literature for various
of the benzene complexes. Apart from this, all the coordin- ligands that were described following the DCD model.
ated ligands show a slightly inverse boat conformation (as These values were found to be comparable for the ethylene
692
Eur. J. Inorg. Chem. 2003, 687Ϫ698

The First η6-RhI and η6-IrI Complexes of 2,6-Bis(trimethylsilyl)phosphinines
Table 4. Calculated NBO charge distribution for complexes IϪIV
FULL PAPER
I[a]
II[a]
0.825
III[a]
IV[a]
qP
qC1
qC2
0.685
Ϫ0.244
Ϫ0.230
Ϫ0.258
Ϫ0.247
Ϫ0.593
Ϫ0.622
Ϫ0.245
qC3
Ϫ0.227
qC4
Σqring
Ϫ0.236
0.423
Ϫ0.256
0.035
0.340
0.080
qM
0.404
Ϫ0.435 (1.445)
0.274
0.124
Ϫ0.495 (1.410)
0.125
qC5 (P)[a]
qC7 (P)[a]
Σqolefin (C5ϪC6)
Σqolefin (C7ϪC8)
Ϫ0.421 (1.436)
Ϫ0.432 (1.451)
0.165
Ϫ0.491 (1.437)
Ϫ0.492 (1.428)
Ϫ0.074
0.256
Ϫ0.204
0.138
Ϫ0.086
[a]
Wiberg bond indices.
Table 5. Results of the charge decomposition analysis of complexes IϪIV
II
I
III
IV
Arene ring:
d[a]
0.649
0.121
5.364
0.759
0.161
4.714
0.912
0.386
2.363
1.062
0.398
2.669
b[b]
d/b
b/(d ϩ b)
15.71%
Ϫ0.592
Ϫ0.009
17.5%
Ϫ0.613
Ϫ0.015
29.73%
Ϫ0.563
Ϫ0.029
27.26%
Ϫ0.540
Ϫ0.058
r[c]
∆Х[d]
C-5ϭC-6, C-7ϭC-8:
d
b
0.430
0.268
1.604
38.39%
Ϫ0.377
Ϫ0.013
0.452, 0.385
0.508
0.342
1.485
40.23%
Ϫ0.380
Ϫ0.014
0.500, 0.495
0.331, 0.333
1.510, 1.480
39.83%, 40.22%
Ϫ0.403, Ϫ0.387
Ϫ0.015, Ϫ0.014
0.257, 0.241
1.759, 1.597
36.25%, 38.50%
Ϫ0.376
d/b
B(d/d ϩ b)
r
∆
Ϫ0.016
[a]
[b]
[c]
[d]
d ϭ donation. b ϭ back donation. r ϭ repulsion. ∆ ϭ residual.
ligands in all the complexes studied; however, they were ranges from 36.25 to 40.23%, and the corresponding Wib-
found to be higher for the two arene ligands in the iron erg bond indices, which vary from 1.41 to 1.45.
complexes. This is consistent with the finding that the π-
In conclusion, we have devised two different methods
accepting capacity of both arenes is enhanced in the zero- that allowed us to synthesize the first (η6-phosphinine)RhI
valent iron complexes III and IV. As can be seen in Table 4, and -IrI complexes. The first studies of the reactivity of
the b/(b ϩ d) ratio increases by about 10% on going from these complexes with ethanol or water quantitatively
the rhodium to the iron complexes. Consequently, one may yielded the first (η5-phosphacyclohexadienyl)RhI and -IrI
propose that an increase in the back donation (b) would complexes, which resulted from the formal 1,1-addition of
increase covalency, and hence a larger absolute value for ROϪ, followed by Hϩ. DFT calculations allowed us to ra-
this residual term results. On examining the d/b and b/(b ϩ tionalize the large differences in reactivity between the (η6-
d) ratios, one can assume that in both types of complexes, phosphinine or benzene)rhodium cationic complexes and
the arenes essentially behave as π-donors, the percentage of the isoelectronic Fe0 complexes. A study on the chemical
π-acceptance [given by the b/(d ϩ b) ratio] does not exceed reactivity of these complexes is underway in our laborat-
29.73%. It also appears that absolute values of donation (d) ories.
are slightly greater in the case of the phosphinine com-
plexes, but care must be taken since it should be emphasized
Experimental Section
that the absolute values of d and b are not important, only
their ratios are relevant. In view of these results, it is obvi-
ous that the contribution of the phosphinine and benzene
groups to the electronic exchange in the complex are rela-
tively similar. On the other hand, the contribution of the
ethylene ligands also appears to be comparable in both
General: All reactions were routinely performed under argon or
nitrogen by using Schlenk and glove-box techniques and dry deoxy-
genated solvents. Dry THF and hexane were obtained by distilla-
tion from Na/benzophenone, and dry CDCl3 from P2O5. CD2Cl2
˚
was dried and stored, like CDCl3, overn 4 A Linde molecular
types of complexes, as shown by the b/(d ϩ b) ratio, which sieves. Nuclear magnetic resonance spectra were recorded with a
Eur. J. Inorg. Chem. 2003, 687Ϫ698 693

´
M. Doux, L. Ricard, F. Mathey, P. Le Floch, N. Mezailles
FULL PAPER
Bruker AC-200 SY spectrometer operating at 300.0 MHz for 1H,
(m, 20 H, CH of C6H5), 7.50 (AXXЈ, vt, ΣJH,P ϭ 7.5, 1 H, 4-H)
75.5 MHz for 13C and 121.5 MHz for 31P. Solvent peaks were used ppm. 13C NMR (75.5 MHz, CDCl3, 298 K): δ ϭ 25.8 (s, CH2 of
as internal references relative to Me4Si for 1H and 13C chemical
shifts (ppm); 31P chemical shifts were relative to a 85% H3PO4 ex-
ternal reference. Coupling constants are given in Hz. The following
COD), 70.0 (s, CH of COD), 127.8 (s, CH of Ph), 128.3 (s, CH of
Ph), 128.5 (s, CH of Ph), 128.6 (s, CH of Ph), 129.5 (s, CH of Ph),
132.2 (s, CH of Ph), 134.9 (AXXЈ, vt, ΣJC,P ϭ 23, C-4), 138.9 (m,
1
abbreviations are used: s ϭ singlet, d ϭ doublet, t ϭ triplet, m ϭ C-α of Ph), 140.5 (t, JC,P ϭ 2.3, C-2,6), 150.5 (m, C of Ph) 161.5
multiplet, quint ϭ quintuplet, v ϭ virtual. IR data were collected
with a PerkinϪElmer 297 spectrometer. Mass spectra were ob-
tained at 70 eV with an HP 5989B spectrometer coupled to an HP
5980 chromatograph by the direct inlet method. Elemental analyses
were performed by the ‘‘Service d’analyse du CNRS’’, at Gif
sur Yvette, France. Diazaphosphinine 1,[12a] phosphinines 4,[18]
5,[12a] [Rh(COD)Cl]2,[19] [Rh(NBD)Cl]2,[20] [Ir(COD)Cl]2,[21]
[Rh(COD)2][BF4][8b,22] were prepared according to reported pro-
cedures.
(m, C-3,5) ppm. 31P NMR (121.5 MHz, CD2Cl2, 298 K): δ ϭ
1
175.15 (d, JP,Rh ϭ 166.5) ppm. C66H54P2RhBF4 (1098.81): calcd.
C 72.14, H 4.95; found C 72.43, H 5.15.
Synthesis of the η6-Phosphinine Complexes. Method A: A solution
of GaCl3 (63.4 mg, 0.36 mmol) in CH2Cl2 was added by cannula
into a mixture of [M(diene)Cl]2 (0.15 mmol) and phosphinine
(0.3 mmol) in CH2Cl2 at Ϫ78 °C. The resulting mixture was then
stirred for 30 min at room temperature. The solid was then precipit-
ated with hexane, filtered, and dried under vacuum yielding the
title complex. Method B: CH2Cl2 (10 mL) was added to a mixture
of [Rh(diene)2][BF4] (0.3 mmol) and phosphinine (0.3 mmol). The
solution was heated at 40 °C until the reaction was complete, as
shown by 31P NMR spectroscopy. After evaporation of the solvent
under vacuum, the solid was washed with hexane and dried to yield
the title complex.
Synthesis of 2,3,5,6-Tetraphenylphosphinine (2): 5 equiv. of di-
phenylacetylene (1.78 g, 10.0 mmol) were added to a solution of 1
(0.42 g, 2.0 mmol) in 8 mL of toluene. The mixture was then heated
at 130 °C for 12 h. The completion of the reaction was checked
by 31P NMR spectroscopy. The solution was then cooled to room
temperature, and the title product purified by column chromato-
graphy on silica gel. Excess diphenylacetylene was eluted first with
hexane, and phosphinine 2 was then eluted with a mixture of hex-
ane/toluene (80:20). Yield 55%, 440 mg. 1H NMR (300 MHz,
CDCl3, 298 K): δ ϭ 7.26Ϫ7.40 (m, 20 H, CH of Ph), 7.70 (d,
4JH,P ϭ 3.6, 1 H, 4-H) ppm. 13C NMR (75.5 MHz, CDCl3, 298 K):
δ ϭ 127.5Ϫ130.9 (m, CH of Ph), 137.4 (d, 3JC,P ϭ 12.5, C-4), 142.4
Synthesis of Complex 7: 3,5-Dimethyl-2,6-bis(trimethylsilyl)phos-
phinine (4) (80 mg, 0.3 mmol) was used with Method
{[Rh(COD)Cl]2 (74 mg, 0.15 mmol)} and Method
A
B
{[Rh(COD)2][BF4] (121 mg, 0.3 mmol)}. Method A: yield 87%,
180 mg; Method B: yield 66%, 112 mg. Slow diffusion of hexane
into a THF solution of the complex yielded X-ray quality crystals.
1H NMR (300 MHz, CDCl3, 298 K): δ ϭ 0.49 (s, 18 H, SiCH3),
2.20Ϫ2.28 (m, 8 H, CH2 of COD), 2.64 (s, 6 H, CH3), 4.70 (s, 4
H, CH of COD), 5.92 (s, 1 H, 4-H) ppm. 13C NMR (75.5 MHz,
2
2
(d, JC,P ϭ 24.5, C-α of Ph), 142.5 (s, C-α of Ph), 145.8 (d, JC,P
ϭ
10.9, C-3,5), 169.6 (d, JC,P ϭ 54.8, C-2,6) ppm. 31P NMR
(121.5 MHz, CDCl3, 298 K): δ ϭ 206.0 (s) ppm. C29H21P (400.45):
calcd. C 86.98, H 5.29; found C 86.97, H 5.44.
1
3
CDCl3, 298 K): δ ϭ 1.2 (d, JC,P ϭ 16.8, SiCH3), 24.4 (s, CH3),
1
Synthesis of 5,6-Diphenyl-2-(trimethylsilyl)phosphinine (3): 1 equiv.
of (trimethylsilyl)acetylene (0.20 g, 2.0 mmol) was added to a solu-
tion of 1 (0.42 g, 2.0 mmol) in 8 mL of toluene. The mixture was
then heated at 60 °C for 12 h. The complete formation of the
monoazaphosphinine was observed by 31P NMR spectroscopy. 3
equiv. of diphenylacetylene (1.07 g, 6.0 mmol) were then added to
the mixture, and the mixture was heated at 130 °C for 16 h. The
solution was then cooled to room temperature and the title product
purified by column chromatography. Excess diphenylacetylene was
eluted first with hexane, and phosphinine 3 was then eluted with a
mixture of hexane/toluene (80:20). 3 was recovered as a white solid
after solvent evaporation. Yield 75%, 480 mg. 1H NMR (300 MHz,
32.0 (s, CH2 of COD), 83.3 (d, JC,Rh ϭ 11.6, CH of COD), 97.1
(dd, 3JP,C ϭ 8.6, 1JC,Rh ϭ 3.2, C-4), 131.5 (dd, 2JC,P ϭ 2.8, 1JC,Rh ϭ
1
1
1.7, C-3,5), 134.5 (dd, JC,P ϭ 76.5, JC,Rh ϭ 1.7, C-2,6) ppm. 31P
1
NMR (121.5 MHz, CDCl3, 298 K): δ ϭ 101.42 (d, JRh,P ϭ 7.6)
ppm. C21H37BF4PRhSi2 (566.37): calcd. C 44.53, H 6.58; found C
44.02, H 6.18.
Synthesis of Complex 8: 3,5-Diphenyl-2,6-bis(trimethylsilyl)phos-
phinine (5) (118 mg, 0.3 mmol) was used with Method
{[Rh(COD)Cl]2 (74 mg, 0.15 mmol)} and Method
A
B
{[Rh(COD)2][BF4] (122 mg, 0.3 mmol)}. Method A: yield 77%,
188 mg; Method B: yield 69%, 143 mg. Slow diffusion of hexane
into a THF solution of the complex yielded yellow crystals suitable
4
CDCl3, 298 K): δ ϭ 0.32 (d, JH,P ϭ 0.8, 9 H, SiMe3), 7.02Ϫ7.16
1
4
2
for X-ray analysis. H NMR (300 MHz, CDCl3, 298 K): δ ϭ 0.26
(m, 10 H, CH of Ph), 7.45 (dd, JH,P ϭ 3.0, JH,H ϭ 8.1, 1 H, 4-
H), 8.03 (dd, 3JH,P ϭ 10.6, 2JH,H ϭ 8.1, 1 H, 3-H) ppm. 13C NMR
(75.5 MHz, CDCl3, 298 K): δ ϭ 0.6 (d, JC,P ϭ 5.7, SiMe3),
127.2Ϫ131.3 (m, CH of Ph), 137.9 (d, JC,P ϭ 12.5, C-4), 142.8 (s,
C-α of Ph), 143.4 (d, JC,P ϭ 23.6, C-α of Ph), 146.4 (d, JC,P
10.3, C-3,5), 170.0 (d, JC,P ϭ 62.3, C-2), 170.7 (d, JC,P ϭ 77.1,
C-6) ppm. 31P NMR (121.5 MHz, CDCl3, 298 K): δ ϭ 232.7 (s)
ppm. C20H21PSi (320.44): calcd. C 74.96, H 6.61; found C 75.27,
H 6.25.
(s, 18 H, SiCH3), 2.34Ϫ2.37 (m, 4 H, CH2 of COD), 2.39Ϫ2.53 (m,
4
3
4 H, CH2 of COD), 4.85 (s, 4 H, CH of COD), 5.72 (d, JH,P
ϭ
1.3, 1 H, 4-H, 7.51Ϫ7.57 (m, 10 H, CH of Ph) ppm. 13C NMR
3
(75.5 MHz, CDCl3, 298 K): δ ϭ 2.2 (dd, 3JC,P ϭ 9.8, 3JC,Rh ϭ 2.3,
2
2
ϭ
1
1
1
SiCH3), 32.1 (s, CH2 of COD), 84.8 (d, JC,Rh ϭ 11.5, CH of
3
1
COD), 94.1 (dd, JC,P ϭ 10.4, JC,Rh ϭ 3.5, C-4), 129.5Ϫ136.8 (s,
Ph), 137.9 (s, C-3,5), 141.1 (d, 1J
ϭ 104.5, C-2,6) ppm. 31P
C,P
1
NMR (121.5 MHz, CDCl3, 298 K): δ ϭ 103.16 (d, JRh,P ϭ 9.1)
ppm. C31H41Cl4GaPRhSi2 (815.24): calcd. C 45.67, H 5.07; found
C 45.23, H 4.76.
Synthesis of Complex 6: Phosphinine 2 (240 mg, 0.6 mmol) and
[Rh(COD)2][BF4] (121 mg, 0.3 mmol) were weighed in air and then
placed under nitrogen in a Schlenk flask. CH2Cl2 (10 mL) was then
added by syringe into the mixture, which was stirred for 2 h at 40
°C. The solvent was then evaporated under vacuum and the COD
extracted with 2 mL of hexane. The solid was then dried under
vacuum to yield pure 6 as a yellow powder. Yield 95%, 313 mg. 1H
Synthesis of Complex 9: 3,5-Dimethyl-2,6-bis(trimethylsilyl)phos-
phinine (4) (80 mg, 0.3 mmol) was used with Method
{[Rh(NBD)Cl]2 (69 mg, 0.15 mmol)} and Method
A
B
{[Rh(NBD)2][BF4] (112 mg, 0.3 mmol)}. Beige solid. Method A:
yield 58%, 117 mg; Method B: yield 75%, 124 mg. 1H NMR
NMR (300 MHz, CDCl3, 298 K): δ ϭ 1.97Ϫ2.01 (m, ΣJ ϭ 13, 8 (300 MHz, CDCl3, 298 K): δ ϭ 0.46 (s, 18 H, SiCH3), 1.34 (s, 2 H,
H, CH2 of COD), 4.00 (m, ΣJ ϭ 13, 4 H, CH of COD), 6.92Ϫ7.26
CH2 of NBD), 2.57 (s, 6 H, CH3), 3.58 (s, 2 H, CH of NBD), 4.12
694
Eur. J. Inorg. Chem. 2003, 687Ϫ698

The First η6-RhI and η6-IrI Complexes of 2,6-Bis(trimethylsilyl)phosphinines
FULL PAPER
(m, ΣJ ϭ 7.3, 4 H, CH of NBD), 6.68 (s, 1 H, 4-H) ppm. 13C NMR ppm. 31P NMR (121.5 MHz, CDCl3, 298 K): δ ϭ Ϫ19.81 (d,
3
(75.5 MHz, CDCl3, 298 K): δ ϭ 1.3 (d, JC,P ϭ 9.7, SiCH3), 24.7
1JP,H ϭ 586) ppm. C21H38Cl3GaOPRhSi2 (672.66): calcd. C 37.50,
2
1
(s, CH3), 47.1 (d, JC,Rh ϭ 2, CH of NBD), 48.3 (d, JC,Rh ϭ 8.8, H 5.69; found C 37.21, H 5.25.
CH of NBD), 60.1 (d, 3JC,Rh ϭ 4, CH2 of NBD), 101.5 (dd, 3JC,P ϭ
Synthesis of Complex 14: Complex 8 (73 mg, 0.09 mmol) was used.
Yield 100%, 71 mg. 1H NMR (300 MHz, CD2Cl2, 298 K): δ ϭ 0.20
(s, 18 H, SiMe3), 2.16Ϫ2.45 (m, 8 H, CH2 of COD), 4.48 (s, 4 H,
1
1
12.9, JC,Rh ϭ 2.3, C-4), 127.9 (d, JC,Rh ϭ 3.2, C-3,5), 130.3 (dd,
1JC,P ϭ 101.2, JC,Rh ϭ 3.0, C-2,6) ppm. 31P NMR (121.5 MHz,
1
CDCl3, 298 K): δ ϭ 107.77 ppm. C20H33BF4PRhSi2 (550.33):
calcd. C 43.65, H 6.04; found C 42.27, H 5.75.
4
1
CH of COD), 5.53 (d, JH,P ϭ 3.4, 1 H, 4-H), 7.13 (d, JH,P
ϭ
581.2, 1 H, PH), 7.18Ϫ7.48 (m, 10 H, CH of Ph) ppm. 13C NMR
(75.5 MHz, CD2Cl2, 298 K): δ ϭ 1.4 (d, JC,P ϭ 3.0, SiMe3), 31.5
(s, CH2 of COD), 69.0 (dd, JC,P ϭ 53.5, JC,Rh ϭ 53.5, C-2,6),
75.9 (d, JC,Rh ϭ 12.2, CH2 of COD), 94.3 (dd, JC,P ϭ 25.7,
3
Synthesis of Complex 10: 3,5-Diphenyl-2,6-bis(trimethylsilyl)phos-
phinine (5) (118 mg, 0.3 mmol) was used with Method
1
1
A
B
1
3
{[Rh(NBD)Cl]2 (69.1 mg, 0.15 mmol)} and Method
{[Rh(NBD)2][BF4] (112 mg, 0.3 mmol)}. Brown solid. Method A: 1JC,Rh ϭ 3.1, C-4), 128.5Ϫ129.7 (s, CH of Ph), 130.3 (t, JC,Rh
ϭ
1
yield 62%, 148 mg; Method B: yield 71%, 144 mg. 1H NMR
2JC,P ϭ 3.0, C-3), 139.7 (d, 3JC,P ϭ 11.3, C-α of Ph) ppm. 31P NMR
(300 MHz, CDCl3, 298 K): δ ϭ 0.18 (d, 4JH,P ϭ 1.7, 18 H, SiCH3), (121.5 MHz, CD2Cl2, 298 K): δ ϭ Ϫ15.31 (d, JP,H ϭ 581.2) ppm.
1
1.49 (t, JH,H ϭ 1.7, 2 H, CH2 of NBD), 3.76 (t, 3JH,H ϭ 1.7, 2 H, C31H42Cl3GaOPRhSi2 (796.80): calcd. C 46.73, H 5.31; found C
3
CH of NBD), 4.44 (m, ΣJ ϭ 10, 4 H, CH of NBD), 6.6 (d, 4JH,P ϭ 46.32, H 4.97.
1.5, 1 H, 4-H), 7.35Ϫ7.52 (m, 10 H, CH of Ph) ppm. 13C NMR
Synthesis of Complex 15: Complex 11 (100 mg, 0.13 mmol) was
(75.5 MHz, CDCl3, 298 K): δ ϭ 2.3 (dd, 3JC,P ϭ 9.3, 3JC,Rh ϭ 3.1,
used. Yield 100%, 99 mg. 1H NMR (300 MHz, CD2Cl2, 298 K):
δ ϭ 0.39 (s, 18 H, SiCH3), 1.96Ϫ2.22 (m, 8 H, CH2 COD), 2.24 (s,
3 H, CH3), 2.93 (d, JH,H ϭ 14.8, 4 H, CH2 of COD), 4.12 (s, 4 H,
2
1
SiCH3), 46.9 (d, JC,Rh ϭ 1.7, CH of NBD), 49.2 (d, JC,Rh ϭ 8.7,
3
CH of NBD), 61.1 (d, JC,Rh ϭ 6.8, CH2 of NBD), 100.7 (dd,
3JC,P ϭ 12, JC,Rh ϭ 2.3, C-4), 129.3Ϫ130.5 (s, Ph), 132.2 (d,
1
4
1
CH of COD), 5.99 (d, JH,P ϭ 3.2, 1 H, 4-H), 6.77 (d, JH,P
ϭ
1JC,P ϭ 104.6, 1JC,Rh ϭ 2.3, C-2,6), 133.2 (dd, 1JC,P ϭ 5.7, 1JC,Rh ϭ
2.3, C-3,5) ppm. 31P NMR (121.5 MHz, CDCl3, 298 K): δ ϭ
116.57 ppm. C30H37Cl4GaPRhSi2 (799.20): calcd. C 45.09, H 4.67;
found C 44.71, H 4.34.
582.4, 1 H, PH) ppm. 13C NMR (75.5 MHz, CDCl3, 298 K): δ ϭ
1.2 (d, 3JC,P ϭ 3.4, SiCH3), 22.7 (d, 3JC,P ϭ 8.6, CH3), 33.0 (s, CH2
of COD), 51.5 (m, C-2,6), 58.6 (s, CH of COD), 98.6 (m, C-4),
112.0 (t, JC,P ϭ 5.7, C-3,5) ppm. 31P NMR (121.5 MHz, CD2Cl2,
3
1
Synthesis of Complex 11: 3,5-Dimethyl-2,6-bis(trimethylsilyl)phos-
298 K): δ ϭ Ϫ5.79 (d, JP,H ϭ 582.4) ppm. C21H38Cl3GaIrOPSi2
phinine (4) (80 mg, 0.3 mmol) was used with Method
A
(761.97): calcd. C 33.10, H 5.03; found C 32.88, H 4.65.
{[Ir(COD)Cl]2 (101 mg, 0.15 mmol)}. Beige solid. Method A: yield
84%, 197 mg. 1H NMR (300 MHz, CDCl3, 298 K): δ ϭ 0.46 (s, 18
H, SiCH3), 2.15Ϫ2.18 (m, 8 H, CH2 of COD), 2.69 (s, 3 H, CH3),
4.55 (s, 4 H, CH of COD), 6.0 (s, 1 H, 4-H) ppm. 13C NMR
Reactivity with Ethanol: The same procedure was used for all the
complexes: 5 equiv. of ethanol were added to a CH2Cl2 solution of
the complex (100 mg). The solution instantaneously turned pale
yellow and the 31P NMR spectra showed the formation of a unique
product. The solution was then concentrated to dryness and the
title complex was obtained as a yellow solid. Yield 100%.
3
(75.5 MHz, CD2Cl2, 298 K): δ ϭ 0.5 (d, JP,C ϭ 9.7, SiCH3), 23.4
(d, 3JP,C ϭ 2.4, CH3), 32.8 (s, CH2 of COD), 67.8 (s, CH of COD),
3
1
90.7 (d, JP,C ϭ 10.3, C-4), 123.7 (d, JP,C ϭ 102.1, C-2,6), 127.7
(d, JP,C ϭ 5.4, C-3,5) ppm. 31P NMR (121.5 MHz, CD2Cl2,
2
Synthesis of Complex 16: Complex 8 (73 mg, 0.09 mmol) was used.
Yield 100%, 77.4 mg. 1H NMR (300 MHz, CDCl3, 298 K): δ ϭ
298 K): δ ϭ 60.45 ppm. C21H37Cl4GaIrPSi2 (780.41): calcd. C
32.32, H 4.78; found C 31.95, H 4.31.
3
0.20 (s, 18 H, SiCH3), 1.39 (t, JH,H ϭ 6.9, 3 H, CH3 of EtO), 2.28
(d, J ϭ 8.8, 4 H, CH2 of COD), 2.31Ϫ2.49 (m, 4 H, CH2 of COD),
Synthesis of Complex 12: 3,5-Diphenyl-2,6-bis(trimethylsilyl)phos-
3
3
3.75 (dq, JH,P ϭ 14.9, JH,H ϭ 6.9, 2 H, CH2 of EtO), 4.63 (s, 4
phinine (5) (118 mg, 0.3 mmol) was used with Method
A
4
4
H, CH of COD), 5.77 (d, JH,P ϭ 3.0, 1 H, 4-H), 7.25 (d, JH,P
ϭ
{[Ir(COD)Cl]2 (101 mg, 0.15 mmol)}. Brown solid. Method A:
yield 79%, 215 mg. 1H NMR (300 MHz, CD2Cl2, 298 K): δ ϭ 0.15
(s, 18 H, TMS), 2.17Ϫ2.21 (m, 8 H, CH2 of COD), 4.70 (s, 4 H,
CH of COD), 6.01(s, 1 H, 4-H), 7.34Ϫ7.48 (s, 10 H, H of Ph) ppm.
31P NMR (121.5 MHz, CD2Cl2, 298 K): δ ϭ 58.09 ppm.
621, 1 H, PH), 7.35Ϫ7.38 (m, 4 H, CH of Ph), 7.49Ϫ7.50 (m, 6 H,
H of Ph) ppm. 13C NMR (75.5 MHz, CDCl3, 298 K): δ ϭ 1.3 (m,
2
SiCH3), 15.0 (s, CH3), 30.3 (s, CH2 of COD), 62.2 (d, JC,P ϭ 6.8,
1
1
1
CH2), 63.8 (dd, JC,P ϭ 42.7, JC,Rh ϭ 3, C-2,6), 79.0 (d, JC,Rh
ϭ
4
1
12.3, CH of COD), 92.0 (d, JC,P ϭ 22.7, JC,Rh ϭ 3.5, C-4),
1
Reactivity with H2O: The same procedure was used for all the com-
plexes: 5 equiv. of water were added to a CH2Cl2 solution of the
complex. The solution instantaneously turned pale yellow and the
31P NMR spectra showed the formation of a unique product. The
solution was then concentrated to dryness and the title complex
was obtained as a yellow solid. Yield 100%.
129.4Ϫ129.6 (s, Ph), 133.1 (t, JC,Rh ϭ
3JC,P ϭ 2.9, C-3,5), 138.6
(d, JC,P ϭ 11.9, Ph) ppm. 31P NMR (121.5 MHz, CDCl3, 298 K):
2
1
δ ϭ Ϫ8.51 (d, JP,H ϭ 621) ppm. C33H47Cl4GaOPRhSi2 (861.31):
calcd. C 46.02, H 5.50; found C 45.71, H 5.19.
Synthesis of Complex 17: Complex 11 (100 mg, 0.13 mmol) was
used. Yield 100%, 107 mg. 1H NMR (300 MHz, CDCl3, 298 K):
Synthesis of Complex 13: Complex 7 (70 mg, 0.1 mmol) was used. δ ϭ 0.41 (s, 18 H, SiCH3), 1.24 (t, 3JH,H ϭ 7, 3 H, CH3), 2.13Ϫ2.16
3
Yield 100%, 67 mg. Slow diffusion of hexane into a CH2Cl2 solu-
(m, 8 H, CH2 of COD), 2.41 (s, 3 H, CH3), 3.49 (dq, JH,P ϭ 15.2,
tion of the complex yielded X-ray quality crystals. 1H NMR 3JH,H ϭ 7.1, 2 H, CH2), 4.08 (s, 4 H, CH of COD), 6.42 (d, 4JH,P ϭ
1
(300 MHz, CDCl3, 298 K): δ ϭ 0.46 (s, 18 H, SiCH3), 2.06 (d, 1.6, 1 H, 4-H), 6.90 (d, JH,P ϭ 626, 1 H, PH) ppm. 13C NMR
JH,H ϭ 8.5, 4 H, CH2 COD), 2.28 (s, 3 H, CH3), 2.38 (d, JH,H
ϭ
(75.5 MHz, CDCl3, 298 K): δ ϭ 1.3 (s, SiCH3), 16.3 (s, CH3), 23.2
3
14.8, 4 H, CH2 of COD), 4.12 (s, 4 H, CH of COD), 5.04 (d,
(d, JC,P ϭ 9.4, CH3), 33.3 (s, CH2 of COD), 48.0 (d, 1JC,P ϭ 45.6,
1
2
4JH,P ϭ 2.7, 1 H, 4-H), 6.69 (d, JH,P ϭ 586, 1 H, PH) ppm. 13C C-2,6), 61.3 (s, CH of COD), 63,5 (d, JP,C ϭ 6.5, CH2), 98.9 (d,
NMR (75.5 MHz, CDCl3, 298 K): δ ϭ 1.3 (m, SiCH3), 24.23 (d, 3JP,C ϭ 30.3, C-4), 114.5 (d, JP,C ϭ 4.8, C-3,5) ppm. 31P NMR
2
1
1
3JC,P ϭ 10.8, CH3), 32.3 (s, CH2 of COD), 69.7 (dd, JC,P ϭ 59.4, (121.5 MHz, CDCl3, 298 K): δ ϭ Ϫ1,82 (d, JP,H ϭ 626) ppm.
1JC,Rh ϭ 3.6, C-2,6), 75.0 (d, 1JC,Rh ϭ 12.5, CH of COD), 91.8 (dd,
3JC,P ϭ 29,1JC,Rh ϭ 3.6, C-4), 125.9 (t, 1JC,Rh ϭ 3JC,P ϭ 3.6, C-3,5)
C23H43Cl4GaIrOPSi2 (826.48): calcd. C 33.42, H 5.24; found C
33.08, H 4.83.
Eur. J. Inorg. Chem. 2003, 687Ϫ698
695

´
M. Doux, L. Ricard, F. Mathey, P. Le Floch, N. Mezailles
FULL PAPER
Theoretical Methods: The geometries of complexes IϪIV were
optimized using the gradient-corrected density functional theory
(DFT), using Becke’s three-parameter hybrid method B3LYP.[16]
The basis set used is identical to that used by Frenking and co-
[metal fragment]; (ii) the mixing of the unoccupied MOs of the
ligand and the occupied MOs of the metal fragment; this value
(noted b) accounts for the back donation [metal fragment] Ǟ li-
gand; (iii) the mixing of the occupied MOs of the ligand and the
workers in many studies.[17] The basis set, which is known as ‘‘Basis occupied MOs of the metal fragment; this term (noted ϽrϾ), which
Set II’’, incorporates the Hay and Wadt small-core relativistic ef-
fective core potential and double-ζ valence basis set (441/2111/N1)
(N ϭ 4, 3, 2 for first, second, and third transition metal rows, re-
spectively) for transition metals, in conjunction with all-electron 6-
describes the repulsive polarization ligand o [metal fragment], is
negative because electronic charge is removed from the overlapping
area of the occupied orbitals; (iv) the residual term (∆), which re-
sults from the mixing of the unoccupied MOs of the two respective
31G(d) basis sets for the main-group elements, H, C, and P.[25] All fragments; usually, this term is very close to zero for closed-shell
optimized structures reported here have only positive eigenvalues
of the Hessian matrix, i.e. they are minima on the potential energy
surface. The charge distribution in the optimized structures was
calculated with the NBO partitioning scheme. Inspection of the
arenes (benzene or phosphinine) and ethyleneϪmetal fragment
donorϪacceptor interactions was performed using the charge-de-
interactions. This value constitutes an important probe in deter-
mining whether the bonding studied can really be classified as a
donorϪacceptor interaction following the DewarϪChattϪ
Duncansson model. Important deviations from ∆ ϭ 0 imply that
the bond studied is more conventionally described as a normal co-
valent bond between two open-shell fragments. A more detailed
composition analysis (CDA).[26] In the CDA method, the molecular presentation of the CDA method and the interpretation of the re-
orbitals (canonical, natural, or KohnϪSham) of the complex are
expressed in terms of MOs of appropriately chosen fragments. In
the cases studied, the KohnϪSham orbitals of the B3LYP/II calcu-
lations are formed in the CDA procedure as a linear combination
of the MOs of the arene and those of the remaining fragment
[M(C2H4)2] (positively charged for M ϭ Rh and neutral for M ϭ
Fe). The same approach was used for the analysis of the ethylene
ligand, [M(arene)(C2H4)] (positively charged for M ϭ Rh and neut-
ral for M ϭ Fe) being considered as the second fragment. In both
cases, the arene and ethylene ligands, as well as the considered frag-
ments were computed in the geometry of the complex. The orbital
sults can be found in the literature. CDA calculations were per-
formed with the program CDA version 2.1. More details about the
method and a discussion of its applications and results can be
found in the literature. These calculations were performed with the
Gaussian-98 program.[27]
X-ray Structure Determinations: Crystals of complexes 6Ϫ8, 14, 15,
17 suitable for X-ray diffraction were obtained by slow diffusion of
hexane into a dichloromethane solution of the complexes. Data
were collected with a Nonius Kappa CCD diffractometer, using an
˚
Mo-Kα (λ ϭ 0.71070 A) X-ray source and a graphite monochrom-
contributions are divided into four parts: (i) the mixing of the occu- ator. Experimental details are described in Tables 6 and 7. The crys-
pied MOs of the ligand and the unoccupied MOs of the metal
fragment; this value (noted d) represents the donation ligand Ǟ
tal structures were solved using SIR 97[28] and SHELXL-97.[29]
ORTEP drawings were made using ORTEP III for Windows.[30]
Table 6. Details of X-ray structure determinations
6
8
Crystal color/shape
Crystal size [mm]
Empirical formula
orange/plate
0.18 ϫ 0.16 ϫ 0.12
C67H56Cl6GaP2Rh
yellow/cube
0.16 ϫ 0.16 ϫ 0.16
C31H41Cl4GaPRhSi2
815.22
Formula mass
1308.39
Temperature [K]
150.0(10)
Mo-Kα
0.71069
150.0(10)
Mo-Kα
0.71069
Radiation (graphite monochromator)
˚
Wavelength [A]
Crystal system
Space group
monoclinic
P21/n
14.348(5)
17.724(5)
47.681(5)
94.120(5)
12094(6)
8, 1.437
monoclinic
Cc
21.309(9)
11.185(5)
30.250(5)
90.446(6)
7210(5)
8, 1.502
1.633
27.45
3312
˚
a [A]
˚
b [A]
˚
c [A]
β [°]
3
˚
V [A ]
Z, calcd. density [g·cmϪ3
]
Absorption coefficient [cmϪ1
2Θmax [°]
]
1.077
25.03
5328
F(000)
Index ranges h,k,l
Refl.collected/independent
Refl. used
Ϫ16/17, Ϫ21/21, Ϫ56/31
38747/19776
13955
0.0378
Ϫ26/27, Ϫ13/14, Ϫ39/38
25079/14757
13036
(Rint
)
0.0569
Absorption correction
Parameters refined
0.8297 min, 0.8816 max
1385
10
0.0561, 0.1657
1.040
1.034(0.093)/Ϫ1.371(0.093)
Ϫ
0.7802 min, 0.7802 max
757
17
0.0368, 0.0906
1.001
1.199(083)/Ϫ0.738(083)
Ϫ0.002(8)
Reflection/parameter ratio
Final R1, wR2 [I Ͼ 2σ(I)]
Goodness-of-fit on F2
Ϫ3
˚
Diff. peak/hole [e·A
]
Flack parameter
696
Eur. J. Inorg. Chem. 2003, 687Ϫ698

The First η6-RhI and η6-IrI Complexes of 2,6-Bis(trimethylsilyl)phosphinines
FULL PAPER
Table 7. Details of X-ray structure determinations
14
17
Crystal color
Crystal shape
yellow
plate
colorless
needle
Crystal size [mm]
Empirical formula
Formula mass
0.22 ϫ 0.22 ϫ 0.09
C32H44Cl5GaOPRhSi2
881.70
0.20 ϫ 0.16 ϫ 0.16
C23H43Cl4GaIrOPSi2
826.44
Temperature [K]
150.0(10)
150.0(10)
Radiation (graphite monochromator)
Wavelength [A]
Mo-Kα
0.71069
Mo-Kα
0.71069
˚
Crystal system
Space group
monoclinic
P21/n
8.371(5)
25.173(5)
17.777(5)
monoclinic
P21/n
13.421(5)
14.379(5)
17.895(5)
˚
a [A]
˚
b [A]
˚
c [A]
β [°]
91.970(5)
3744(3)
4, 1.564
1.649
30.01
1792
110.090(5)
3243.3(19)
4, 1.693
5.398
30.01
3
˚
V [A ]
Z, calcd. density [g·cmϪ3
]
Absorption coefficient [cmϪ1
2Θmax [°]
]
F(000)
1628
Index ranges h,k,l
Refl.collected/independent
Refl. used
Ϫ11/11,Ϫ33/35, Ϫ24/24
19140/10843
9382
Ϫ18/18, Ϫ20/18, Ϫ24/25
16753/9447
7817
(Rint
)
0.0177
0.0228
Absorption correction
Parameters refined
0.7130 min., 0.8657 max.
382
24
0.0332, 0.1238
0.943
0.800(106)/Ϫ1.008(106)
0.4115 min., 0.4788 max.
307
25
0.0293, 0.0765
1.023
1.692(127)/Ϫ1.328(127)
Reflection/parameter ratio
Final R1, wR2 [I Ͼ 2σ(I)]
Goodness-of-fit on F2
Ϫ3
˚
Diff. peak/hole [e·A
]
[2b]
Organomet. Chem. 1997, 528, 77Ϫ81.
V: C. Elschenbroich,
Compound 6 and 14 each contain a highly disordered CH2Cl2 mo-
lecule. They were accounted for by using the Platon SQUEEZE
function. The structure of compound 6 calls for an additional com-
ment. Both GaCl4 anions present in the asymmetric unit appear to
be disordered. When average positions are used, R1 converges to
about 7.9%. The anion labeled GA1 has fewer problems than the
second one, and the disorder could not be refined. For the anion
labeled GA2, a model consisting of a major site, accounting for
70% of the electron-density, and a minor site (30% at GA2A) was
refined. This model implies that the neighboring CH2Cl2 solvate is
similarly disordered; not taking this into account results in appar-
M. Nowotny, B. Metz, W. Massa, J. Graulich, K. Biehler, W.
Sauer, Angew. Chem. Int. Ed. Engl. 1991, 30, 547Ϫ550, Angew.
[2c]
Chem. 1991, 103, 601Ϫ604.
Cr: C. Elschenbroich, F. Bär,
E. Bilger, D. Mahrwald, M. Nowotny, B. Metz, Organometal-
lics 1993, 12, 3373Ϫ3378. [2d] Ho: P. L. Arnold, F. G. N. Cloke,
P. B. Hitchcock, Chem. Commun. 1997, 481Ϫ482.
[3] [3a]
H. Vahrenkamp, H. Nöth, Chem. Ber. 1972, 105,
[3b]
1148Ϫ1157.
J. Deberitz, H. Nöth, Chem. Ber. 1973, 106,
2222Ϫ2226. [3c] K. C. Nainan, C. T. Sears, J. Organomet. Chem.
1978, 148, C31ϪC34.
[4] [4a]
F. Nief, C. Charrier, F. Mathey, M. Simalty, J. Organomet.
[4b]
˚
Chem. 1980, 187, 277Ϫ285.
J. Fisher, A. De Cian, F. Nief,
ent contacts of about 1.4 A between the chlorine atoms of GA2A
Acta Crystallogr., Sect. B 1979, 35, 1686Ϫ1695.
and the H atoms of the solvate. Compound 7 is twinned, twin oper-
ator [1 0 0 0 Ϫ1 0 0 0 Ϫ1], BASF ϭ 0.41. CCDC-188543 (6),
-188544 (7), -188545 (8), -188546 (14), -188547 (15), and -188548
(17) contain the supplementary crystallographic data for this paper.
These data can be obtained free of charge at www.ccdc.cam.ac.uk/
conts/retrieving.html [or from the Cambridge Crystallographic
Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; Fax: (in-
ternat.) ϩ 44-1223/336-033; E-mail: deposit@ccdc.cam.ac.uk].
[5a] F. Knoch, F. Kremer, U. Schmidt, U. Zenneck, P. Le Floch,
F. Mathey, Organometallics 1996, 15, 2713Ϫ2719. [5b] D. Böhm,
H. Geiger, F. Knoch, S. Kummer, P. Le Floch, F. Mathey, U.
[5]
Schmidt, U. Zenneck, Phosphorus, Sulfur Silicon 1996, 109/110,
[5c]
173Ϫ176.
P. Le Floch, F. Knoch, F. Kremer, F. Mathey, J.
Scholz, W. Scholz, K.-H. Thiele, U. Zenneck, Eur. J. Inorg.
Chem. 1998, 119Ϫ126.
N. Mezailles, L. Ricard, P. Le Floch, F. Mathey, Organometal-
[6]
[7]
´
lics 2001, 20, 3304Ϫ3307.
Supporting Information Available: ORTEP drawings of complexes
7 and 15 (see footnote on the first page of this article).
M. J. Nolte, G. Gafner, L. M. Haines, J. Chem. Soc., Chem.
Commun. 1969, 1406Ϫ1407.
[8] [8a]
R. R. Schrock, J. A. Osborn, Inorg. Chem. 1970, 9,
[8b]
Acknowledgments
The authors thank the CNRS and the Ecole Polytechnique for sup-
porting this work.
2339Ϫ2343.
R. R. Schrock, J. A. Osborn, J. Am. Chem.
[8c]
Soc. 1971, 93, 3089Ϫ3091.
M. Green, T.A. Kuc, J. Chem.
Soc., Dalton Trans. 1972, 832Ϫ839. [8d] M. Green, G. J. Parker,
[8e]
J. Chem. Soc., Dalton Trans. 1974, 333Ϫ343.
R. Uson, L.
A. Oro, C. Foces-Foces, F. Cano, A. Vegas, M. Valderrama, J.
Organomet. Chem. 1981, 215, 241Ϫ253. [8f] R. Uson, L. A. Oro,
C. Foces-Foces, F. Cano, H. Garcia-Blanco, M. Valderrama, J.
Organomet. Chem. 1982, 229, 293Ϫ304. [8g] M. Valderrama, L.
[1]
´
For a recent review: N. Mezailles, F. Mathey, P. Le Floch, Prog.
Inorg. Chem. 2001, 49, 455Ϫ550.
[2] [2a]
Ti: P. L. Arnold, F. G. N. Cloke, K. Khan, P. Scott, J.
Eur. J. Inorg. Chem. 2003, 687Ϫ698
697

´
M. Doux, L. Ricard, F. Mathey, P. Le Floch, N. Mezailles
FULL PAPER
[8h]
[21]
A. Oro, Can. J. Chem. 1982, 60, 1044Ϫ1047.
R. R. Burch,
J. L. Herde, J. C. Lambert, C. V. Senoff, Inorg. Synth. 1974,
15, 18Ϫ20.
E. L. Muetterties, V. W. day, Organometallics 1982, 1, 188Ϫ197.
[8i] M. Valderrama, M. Scotti, R. Ganz, L. A. Oro, F. J. Lahoz,
C. Foces-Foces, F. H. Cano, J. Organomet. Chem. 1985, 288,
[22] [22a]
M. Green, T. A. Tuc, S. H. Taylor, J. Chem. Soc., Chem.
[22b]
Commun. 1970, 1553Ϫ1554.
R. R. Schrock, J. A. Osborn,
97Ϫ107. [8j] E. Bittersmann, K. Hildebrand, A. Cervilla, P. La-
J. Am. Chem. Soc. 1971, 93, 3089Ϫ3090.
[23] [23a]
[23b]
[8k]
A. D. Becke, Phys. Rev. A 1988, 38, 3098Ϫ3100.
C.
huerta, J. Organomet. Chem. 1985, 287, 255Ϫ263.
J. R.
[23c]
Lee, W. Yang, R.G. Parr, Phys. Rev. 1988, B41, 785Ϫ789.
Bleeke, A. J. Donaldson, Organometallics. 1988, 7, 1588Ϫ1596.
[23d]
[8l]
A.D. Becke, J. Chem. Phys. 1993, 98, 5648Ϫ5652.
D. J.
M. A. Esteruelas, L. A. Oro, Coord, Chem. Rev. 1999, 195,
[8m]
Stevens, F. J. Devlin, C. F. Chablowski, M. J. Frisch, J. Phys.
Chem. 1994, 80, 11623Ϫ11627.
G. Frenking, I. Antes, M. Böhme, S. Dapprich, A. W. Ehlers,
V. Jonas, A. Neuhaus, M. Otto, R. Stegmann, A. Veldkamp, S.
F. Vyboishchikov, Reviews in Computational Chemistry (Eds.:
K. B. Lipkowitz, D. B. Boyd), VCH, New York, 1996, vol. 8,
pp. 63Ϫ144.
P. J. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 299Ϫ310.
S. Dapprich, G. Frenking, CDA 2.1, Marburg, 1995. The pro-
gram is available on-line: ftp://chemie.uni-marburg.de (/pub/
cda).
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M.
A. Robb, J. R. Cheeseman, V. G. Zakrzewski, Jr. J. A.
Montgomery, R. E. Stratmann, J. C. Burant, S. Dapprich, J. M.
Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J.
Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Po-
melli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P.
Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, A.
G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challac-
ombe, P. M. W. Gill, B. Johnson, W.Chen, M. W. Wong, J. L.
Andres, C. Gonzalez, M. Head-Gordon, E. S. Replogle, J. A.
Pople, Gaussian 98, Revision A-11, Gaussian, Inc., Pittsburgh
PA, 1998.
A. Altomare, M. C. Burla, M. Camalli, G. Cascarano, C. Gia-
covazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori, R.
Spagna, SIR97, an integrated package of computer programs
for the solution and refinement of crystal structures using single
crystal data.
G. M. Sheldrick, SHELXL-97, Universität Göttingen,
Göttingen, Germany, 1997.
557Ϫ618.
H. Werner, G. Canepa, K. Ilg, J. Wolf, Organo-
metallics. 2000, 19, 4756Ϫ4766.
[24]
[9]
[9a] I. Amer, H. Alper, J. Am. Chem. Soc. 1990, 112, 3674Ϫ3676.
[9b]
J. Q. Zhou, H. Alper, J. Chem. Soc., Chem. Commun. 1991,
[9c]
233Ϫ234.
H. Alper, J. Q. Zhou, J. Org. Chem. 1992, 57,
[9d]
3729Ϫ3731.
3328Ϫ3331.
J. Q. Zhou, H. Alper, J. Org. Chem. 1992, 57,
[9e]
J. Q. Zhou, H. Alper, J. Chem. Soc., Chem.
[25]
[26]
[9f]
Commun. 1993, 316Ϫ317.
K. Totland, H. Alper, J. Org.
C. M. Crudden, H. Alper, J.
[9g]
Chem. 1993, 58, 3326Ϫ3329.
Org. Chem. 1994, 59, 3091Ϫ3097.
[10] [10a]
J. Q. Zhou, H. Alper, Organometallics 1994, 13,
[27]
[10b]
1586Ϫ1591.
F. Monteil, I. Matsuda, H. Alper, J. Am.
[10c]
Chem. Soc. 1995, 117, 4419Ϫ4420.
J. Q. Zhou, G. Facey,
B. R. James, H. Alper, Organometallics 1996, 15, 2496Ϫ2503.
[11] [11a]
F. Torres, E. Sola, M. Martin, J. A. Lopez, F. J. Lahoz, L.
[11b]
A. Oro, J. Am. Chem. Soc. 1999, 121, 10632Ϫ10633.
Torres, E. Sola, M. Martin, C. Ochs, G. Picazo, J. A. Lopez,
F.
F. J. Lahoz, L. A. Oro, Organometallics 2001, 20, 2716Ϫ2724.
[12] [12a]
N. Avarvari, P. Le Floch, F. Mathey, J. Am. Chem. Soc.
[12b]
1996, 118, 11978Ϫ11979.
N. Avarvari, P. Le Floch, L.
Ricard, F. Mathey, Organometallics 1997, 16, 4089Ϫ4098.
N. Avarvari, N. Maigrot, L. Ricard, F. Mathey, P. Le Floch,
Chem. Eur. J. 1999, 5, 2109Ϫ2118.
F. Nief, J. Fischer, Organometallics 1986, 5, 877Ϫ883.
K. Dimroth, H. Kaletsch, J. Organomet. Chem. 1983, 247,
271Ϫ285.
A. E. Reed, L. A. Curtiss, F. Weinhold, Chem. Rev. 1988, 88,
899Ϫ926.
See for example: G. Frenking, N. Fröhlich, Chem. Rev. 2000,
100, 717Ϫ774.
[13]
[14]
[15]
[28]
[16]
[17]
[18]
[29]
[30]
´
L. Cataldo, S. Choua, T. Berclaz, M. Goeffroy, N. Mezailles,
L. Ricard, F. Mathey, P. Le Floch, J. Am. Chem. Soc. 2001,
123, 6654Ϫ6661.
Louis J. Farrugia, ORTEP-3, Department of Chemistry, Uni-
versity of Glasgow.
[19]
[20]
G. Giordano, R. H. Crabtree, Inorg. Synth. 1990, 28, 88Ϫ90.
E. W. Abel, M. A. Bennett, G. Wilkinson, J. Chem. Soc.
1959, 3178.
Received June 28, 2002
[I02356]
698
Eur. J. Inorg. Chem. 2003, 687Ϫ698