Proton induced P-H and Mo-H bond activation at the phosphide bridged dimolybdenum complexes [Mo2Cp2(μ-H)(μ-PHR)(CO) 4] (R = Cy, 2,4,6-C6H2R′3; R′ = H, Me, tBu)

Article abstract of DOI:10.1039/b412875c

The new hydride complexes [Mo₂Cp₂(μ-PHR)(CO) ₄] having bulky substituents (R = 2,4,6-C₆H ₂¹Bu₃ = Mes*, R = 2,4,6-C ₆H₂Me₃ = Mes) have been prepared in good yield by addition of Li[PHR] to the triply bonded [Mo₂Cp ₂(CO)₄] and further protonation of the resulting anionic phosphide complex [Mo₂Cp₂(μ-PHR)(CO)₄] ^-. Protonation of the Mes* compound with either [H(OEt ₂)₂][B{3,5-C₆H₃(CF₃) ₂}₄] or HBF₄·OEt₂ gives the cationic phosphinidene complex [Mo₂Cp₂(μ-H) (u-PMes*)(CO)₄]⁺ in high yield. In contrast, protonation of the analogous hydride compounds with Mes or Cy substituents on phosphorus give the corresponding unsaturated tetracarbonyls [Mo₂Cp ₂(μ-PHR)(CO)₄]⁺, which are unstable at room temperature and display a cis geometry. Decomposition of the latter give the electron-precise pentacarbonyls [Mo₂Cp₂(μ-PHR)(μ-CO) (CO)₄]⁺, also displaying a cis arrangement of the metal fragments. In the presence of BF₄^- as external anion, fluoride abstraction competes with carbonylation to yield the neutral fluorophosphide hydrides [Mo₂Cp₂(μ-H)(μ-PFR)(CO) ₄]. Similar results were obtained in the protonation reactions of the hydride compounds having a Ph substituent on phosphorus. In that case, using HCl as protonation reagent gave the chloro-complex [Mo,ClCp₂(μ- PHPh)(CO)₄] in good yield. The structures and dynamic behaviour of the new compounds are analyzed on the basis of solution IR and ¹H, ³¹P, ¹⁹F and ¹³C NMR data as well as the X-ray studies carried out on [Mo₂Cp₂(μ-H)(μ-PHMes)(CO) ₄] (cis isomer), [Mo₂Cp₂(μ-H)(μ-PFMes) (CO)₄] (trans isomer), [Mo₂Cp₂(μ-PHCy)(μ- CO)(CO)₄](BF₄) and [Mo₂ClCp₂(μ- PHPh)(CO)₄].

Full text of DOI:10.1039/b412875c

View Article Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
Proton induced P–H and Mo–H bond activation at the phosphide
bridged dimolybdenum complexes [Mo2Cp2(l-H)(l-PHR)(CO)4]
(R = Cy, 2,4,6-C6H2Rꢀ3; Rꢀ = H, Me, tBu)
Celedonio M. Alvarez,a M. Angeles Alvarez,a Daniel Garc´ıa-Vivo´,a M. Esther Garc´ıa,a
Miguel A. Ruiz,*a David Sa´ez,a Larry R. Falvello,b Tatiana Solerb and Patrick Hersonc
a Departamento de Qu´ımica Orga´nica e Inorga´nica/IUQOEM, Universidad de Oviedo,
E-33071, Oviedo, Spain. E-mail: mara@fq.uniovi.es
b Departamento de Qu´ımica Inorga´nica/ICMA, Universidad de Zaragoza, E-50009, Zaragoza,
Spain
c Laboratoire de Chimie Inorganique et Materiaux, Universite´ P. et M. Curie, 75252, Paris,
Cedex 05, France
Received 20th August 2004, Accepted 29th October 2004
First published as an Advance Article on the web 15th November 2004
t
The new hydride complexes [Mo2Cp2(l-H)(l-PHR)(CO)4] having bulky substituents (R = 2,4,6-C6H2 Bu3 = Mes*,
R = 2,4,6-C6H2Me3 = Mes) have been prepared in good yield by addition of Li[PHR] to the triply bonded
[Mo2Cp2(CO)4] and further protonation of the resulting anionic phosphide complex [Mo2Cp2(l-PHR)(CO)4]−.
Protonation of the Mes* compound with either [H(OEt2)2][B{3,5-C6H3(CF3)2}4] or HBF4·OEt2 gives the cationic
phosphinidene complex [Mo2Cp2(l-H)(l-PMes*)(CO)4]+ in high yield. In contrast, protonation of the analogous
hydride compounds with Mes or Cy substituents on phosphorus give the corresponding unsaturated tetracarbonyls
[Mo2Cp2(l-PHR)(CO)4]+, which are unstable at room temperature and display a cis geometry. Decomposition of the
latter give the electron-precise pentacarbonyls [Mo2Cp2(l-PHR)(l-CO)(CO)4]+, also displaying a cis arrangement of
−
the metal fragments. In the presence of BF4 as external anion, fluoride abstraction competes with carbonylation to
yield the neutral fluorophosphide hydrides [Mo2Cp2(l-H)(l-PFR)(CO)4]. Similar results were obtained in the
protonation reactions of the hydride compounds having a Ph substituent on phosphorus. In that case, using HCl as
protonation reagent gave the chloro-complex [Mo2ClCp2(l-PHPh)(CO)4] in good yield. The structures and dynamic
behaviour of the new compounds are analyzed on the basis of solution IR and 1H, 31P, 19F and 13C NMR data as well
as the X-ray studies carried out on [Mo2Cp2(l-H)(l-PHMes)(CO)4] (cis isomer), [Mo2Cp2(l-H)(l-PFMes)(CO)4]
(trans isomer), [Mo2Cp2(l-PHCy)(l-CO)(CO)4](BF4) and [Mo2ClCp2(l-PHPh)(CO)4].
procedure to generate a planar phosphinidene ligand bridging
Introduction
a dimetal centre, this being an alternative to synthetic methods
Recently we reported a two-step process to convert the phos-
relying on the direct use of dichlorophosphines.2,3 Second, the
phide hydride complex [Mo2Cp2(l-H)(l-PHR)(CO)4] (1a, R =
proton-induced elimination of hydrogen occurring at 1a is
t
2,4,6-C6H2 Bu3) into the phosphinidene derivative [Mo2Cp2(l-
not a common process for bridging hydride ligands, although
PR)(CO)4] via the corresponding phosphinidene–hydride cation
it is a well established reaction for terminal hydride atoms,
[Mo2Cp2(l-H)(l-PR)(CO)4]+.1 The key transformation of the
which is likely to involve dihydrogen-bonded intermediates.4
above synthesis occurs after protonation of the hydride complex,
Finally, the oxidative addition of the P–H bond leading to
which presumably induces dihydrogen elimination to give first
the phosphinidene–hydride complex denotes the involvement of
an unsaturated cation which then would experience the oxidative
very reactive cationic intermediates (formation of this product
addition of its P–H bond to yield the electron-precise hydride
occurs instantaneously at room temperature) and has little
complex finally isolated, which displays a planar trigonal
precedent itself. Although P–H bond cleavage is a common
phosphinidene bridge (Scheme 1).
There are several points of interest in the above transforma-
process for coordinated phosphine ligands (generally requiring
thermal of photochemical activation of the complex, however)5
tions. In the first place, they represent a novel and high yield
we are only aware of a few examples of bridging phosphide
ligands experiencing related reactions. These all involve the
6
formation of triply bridging phosphinidene ligands at Os3
or FeCoRu7 centres (Scheme 2). Interestingly, the reverse
reaction (P–H reductive elimination) seems to be thermo-
dynamically favoured at anionic derivatives containing bent
phosphinidene bridges, as observed for the cluster [Os3(l-
H)(l2-PPh)(CO)10]−,8 and proposed for the dimanganese anion
[Mn2(l-H)(l-PCy)(CO)8]−(Scheme 2).9
Taking into account the above considerations and given our
current interest in the chemistry of both phosphinidene-bridged
complexes,10,11 and unsaturated binuclear cations,12 we decided
to study in more detail the protonation reaction leading to the
above mentioned phosphinidene cation. In order to analyze the
influence of the organic substituent on phosphorus, we have
also studied the protonation reactions of the hydride-phosphide
complexes [Mo2Cp2(l-H)(l-PHR)(CO)4] having mesityl (1b),
Scheme 1 Transformations involved in the protonation reaction of
compound 1a [M = Mo(CO)2Cp].
4 1 6 8
D a l t o n T r a n s . , 2 0 0 4 , 4 1 6 8 – 4 1 7 9
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4
©

View Article Online
structure, as they exhibit separated resonances for each of the Cp
groups or carbonyl ligands. At room temperature, however, just
a broad cyclopentadienyl resonance is observed in the 1H NMR
spectrum, thus suggesting the operation of a fluxional process.
Although we have not studied this dynamic process in detail, it
seems to be similar to the one detected for the related complexes
having PHCy5 or PHPh17 bridges. The proposed process in the
latter cases implies the eventual chemical equivalence of both
cyclopentadienyl groups and all four carbonyls ligands.17
The structure of the mesityl derivative 1b is more complex. Its
IR spectrum is significantly different from that of 1a or related
[Mo2Cp2(l-H)(l-PR2)(CO)4] complexes in that it exhibits an
additional high-frequency C–O stretching band (1969 cm−1)
of medium intensity not present in the spectra of related
tetracarbonyl complexes. This can be attributed to the presence
in solution of a cis isomer in addition to the usual trans isomer
(Chart 2), as confirmed by NMR data to be discussed later on.
Scheme 2 Reported type of processes relating bridging phosphide and
phosphinidene ligands (terminal ligands on metal atoms omitted for
clarity, see text).
cyclohexyl (1c) or phenyl (1d) R groups. Moreover, in or-
der to analyze the expected influence of the external anion
on the behaviour of the unsaturated cations likely to be
12b
generated,
we have systematically used both HBF4·OEt2
and [H(OEt2)2](BArꢀ4), [Arꢀ = 3,5-C6H3(CF3)2]13 as protonation
reagents. The anions present in these acids can be considered as
weakly or extremely weak coordinating ligands, respectively,14
and the tetraarylborate anion has been found by us and
Chart 2
12b,15
others to greatly increase the stability of reactive cations.
Finally, a few experiments have been also preformed using
HCl, and acid providing coordinating chloride anions. As will
be shown, our results suggest that very reactive unsaturated
cations are invariably formed after the protonation of the
hydride complexes [Mo2Cp2(l-H)(l-PHR)(CO)4]. The fate of
these cations, however, is strongly dependent both on the nature
of the R group on phosphorus and on the external anion.
Upon crystallization from toluene–petroleum ether mixtures,
crystals of cis-1b were formed. Relevant bond distances and
angles are collected in Table 1. The structure of this molecule
(Fig. 1) displays two almost eclipsed MoCp(CO)2 moieties
bridged by the phosphide and hydride ligands. Out of the two
possible cis isomers, the crystal contains the one with the bulky
mesityl group pointing away from the cyclopentadienyl ligands.
This isomer should be more stable on steric grounds. On the
other hand, the intermetallic separation for cis-1b [3.2868(2)
Results and discussion
˚
A] is only marginally longer than the values found for related
ꢀ
18
˚
trans-[Mo2Cp2(l-H)(l-PRR )(CO)4] complexes (3.25–3.28 A),
Synthesis and structural characterization of hydride complexes
1a,b
which might be indicative of a slight steric congestion in this
molecule. All other interatomic distances and angles are similar
to the values displayed by the related trans-complexes.
Dimolybdenum complexes of the type [Mo2Cp2(l-H)(l-
PHR)(CO)4] (R = Me, Ph,16,17 Cy)5 have been previously pre-
6a
It should be noted that compound 1b represents the first
example of a tetracarbonyl complex of type [M2Cp2(l-H)(l-
PRRꢀ)(CO)4] exhibiting a cis geometry. The only precedent for
this geometry in the phosphide hydride complexes of the group
pared through the P–H cleavage reaction of the corresponding
primary phosphine PH2R on either [Mo2Cp2(CO)6] or the triply
bonded [Mo2Cp2(CO)4]. Attempts to prepare in this way similar
hydride complexes with bulkier substituents as Mes* (1a) or
Mes (1b) were however unsuccessful. In order to synthesize
compounds 1a,b we have designed a new two-step procedure
involving first the addition of the corresponding phosphide
anion Li[PHR] to the triply bonded [Mo2Cp2(CO)4], and then
protonation of the resulting carbonyl anions Li[Mo2Cp2(l-
PHR)(CO)4] (2a,b) with either H3PO4 or (NH4)PF6.
5
6 metals is found in the complex [Mo2{l-(g -C5H4)2SiMe2}(l-
H)(l-PMe2)(CO)4],19 where the cis geometry is forced by the
presence of the linked cyclopentadienyl ligands. However, related
cis/trans isomerism has been previously found for the thiolate-
t
bridged complexes [Mo2Cp2(l-H)(l-SR)(CO)4] (R = Me, Bu,
Ph).20
In solution, compound 1b displays an equilibrium mixture
of the cis isomer found in the crystal and its trans isomer,
with the cis/trans ratio increasing on lowering the temperature
(see Experimental section). At room temperature broad but
separated 31P NMR resonances are observed for both isomers,
The IR spectrum of 1a exhibits C–O stretching bands with
a pattern comparable to those of previously known complexes
of type [Mo2Cp2(l-H)(l-PRRꢀ)(CO)4]. These have been shown
in all cases to display MoCp(CO)2 fragments in a transoid
relative arrangement with respect to the average Mo2(l-H)(l-
1
while only average H NMR resonances are observed for the
P) plane (Chart 1), as confirmed crystallographically on the
1
PMe2, PtBu2 or PPhEt derivatives. Low-temperature H
◦
18a
18b
18c
˚
Table 1 Selected bond lengths (A) and angles ( ) for compound cis-1b
or 13C NMR spectra of 1a are consistent with the proposed
Mo(1)–Mo(2)
Mo(1)–H(1)
Mo(1)–P(1)
Mo(1)–C(1)
Mo(1)–C(2)
Mo(2)–H(1)
Mo(2)–P(1)
Mo(2)–C(3)
Mo(2)–C(4)
P(1)–H(2)
3.2868(2)
1.77(4)
2.4323(5)
1.962(3)
1.966(2)
1.95(4)
2.4264(6)
1.969(3)
1.966(2)
1.30(3)
C(1)–Mo(1)–C(2)
C(1)–Mo(1)–P(1)
C(1)–Mo(1)–H(1)
C(1)–Mo(1)–Mo(2)
C(4)–Mo(2)–C(3)
C(4)–Mo(2)–P(1)
C(4)–Mo(2)–H(1)
C(4)–Mo(2)–Mo(1)
Mo(2)–P(1)–Mo(1)
C(15)–P(1)–H(2)
C(3)–Mo(2)–P(1)
76.40(9)
114.11(6)
66.9(14)
90.40(6)
80.57(10)
79.04(7)
127.6(12)
117.59(7)
85.14(2)
99.4(12)
111.17(7)
P(1)–C(15)
1.829(2)
Chart 1
D a l t o n T r a n s . , 2 0 0 4 , 4 1 6 8 – 4 1 7 9
4 1 6 9

View Article Online
cis and trans for 2b, although we have not studied this in detail.
We note that retention of the trans geometry has been crystallo-
graphically verified in the related pairs of complexes [Mo2Cp2(l-
H)(l-PR2)(CO)4]/[Mo2Cp2(l-PR2)(CO)4]−, (R = Ph,21 Me).22
Protonation reactions of complex 1a
Addition of a slight excess of [H(OEt2)2](BArꢀ4) or HBF4·OEt2
to dichloromethane solutions of compound 1a causes its rapid
transformation into the corresponding phosphinidene hydrides
[Mo2Cp2(l-H)(l-PMes*)(CO)4]X, [X = BArꢀ4 (3a), BF4 (3aꢀ),
Chart 3]. As stated in the Introduction section, this reaction
is thought to occur via an unsaturated phosphide complex
(Scheme 1), but all attempts to detect any intermediate species
by carrying out the above reactions at lower temperatures were
unsuccessful.
Fig. 1 ORTEP view of the molecular structure of compound cis-1b,
with thermal ellipsoids drawn at 30% probability (carbon-bound
hydrogen atoms and some of the labels omitted for clarity).
cyclopentadienyl, mesityl or hydride ligands. The 400.13 MHz
1H NMR spectrum of the complex recorded at 243 K exhibits
already separated resonances for the P–H, aromatic, cyclopenta-
dienyl and hydride groups, but severe broadening of the methyl
resonances is apparent in the spectrum. At this point, the main
resonances of the cis and trans isomers can be safely assigned on
the basis of their relative intensities and the fact that the trans
isomer should exhibit two distinct Cp resonances while the cis
isomer should exhibit a single one. The origin of the broadening
of the methyl resonances is revealed by the 1H NMR spectrum
at 193 K, which then exhibits three separated methyl resonances
for each isomer. Thus it is concluded that rotation of the mesityl
ring around its P–C bond becomes slow at low temperatures on
the NMR timescale so as to give distinct resonances for the ortho
methyl groups of the cis isomer. This might be another reflection
of the existence of some steric congestion in the cis geometry.
The IR spectra of the anionic complexes 2a,b display C–
O stretching bands with a pattern similar to those of the
neutral hydride derivatives 1a,b but shifted ca. 100 cm−1 to
lower frequencies as expected (Table 2). Thus, it is reasonable to
assume that they exhibit a similar geometry, that is trans for 2a or
Chart 3
The cationic hydride complex in compounds 3a or 3aꢀ is quite
acidic and is easily deprotonated just by water to give quanti-
tatively the phosphinidene complex [Mo2Cp2(l-PMes*)(CO)4]
(4), first prepared in low yield by Arif et al.2
Spectroscopic data for the cations in 3a or 3aꢀ (Table 2) suggest
a structure closely related to that of their neutral derivative
4 (Chart 3). In the first place, these cations exhibit strongly
deshielded 31P NMR resonances at ca. 730 ppm, as usually ob-
served for trigonal (four-electron donor) phosphinidene bridges.
Moreover, the pattern of the C–O stretching bands is similar to
that for 4, but frequencies for the cations are some 80 cm−1
higher as expected. Therefore, we assume that the cation in
compounds 3a and 3aꢀ displays a transoid relative arrangement
of their MoCp(CO)2 fragments with respect to the Mo2P plane.
The hydride resonances in these cations appear at ca. −9 ppm
and exhibit a strong P–H coupling of 51 Hz, to be compared
1
Table 2 Selected IR and 31P{ H} NMR data for new compounds
Type
Ra
Compd.
m(CO)b/cm−1
d(P)c/ppm
[Mo2Cp2(l-H)(l-PHR)(CO)4]
Li[Mo2Cp2(l-PHR)(CO)4]
Mes*
Mes
Mes*
Mes
Mes*
Mes*
Mes*
Mes
Mes
Cy
Mes
Cy
Ph
Cy
Cy
1a
1b
2a
2b
3a
3aꢀ
4
5b
5bꢀ
5c
6b
6c
6d
7c
7cꢀ
7d
7dꢀ
8
1956 (w, sh), 1940 (vs), 1868 (s)
1969 (m), 1935 (vs), 1884 (m, sh), 1869 (s)
1884 (w), 1853 (vs), 1835 (m, sh), 1781 (s, br)e
1877 (m), 1838 (vs), 1789 (s), 1770 (m)e
2026 (w), 2000 (vs), 1967 (s), 1951 (s)
2025 (w), 1996 (vs), 1966 (s), 1950 (s)
1958 (w), 1921 (vs), 1880 (s), 1856 (s)
1989 (vs)f
82.2
77.0 (br), 63.3 (br)d
[Mo2Cp2(l-H)(l-PR)(CO)4](BArꢀ4)
[Mo2Cp2(l-H)(l-PR)(CO)4]BF4
[Mo2Cp2(l-PR)(CO)4]
731.5
724.7
685.6
[Mo2Cp2(l-PHR)(CO)4](BArꢀ4)
[Mo2Cp2(l-PHR)(CO)4]BF4
[Mo2Cp2(l-PHR)(CO)4](BArꢀ4)
[Mo2Cp2(l-H)(l-PFR)(CO)4]
71.9g
1988 (vs)f
73.0g
1999 (w, sh), 1977 (vs), 1955 (m)
1978 (w), 1945 (vs), 1881 (s)
1965 (w, sh), 1942 (vs), 1880 (s)
1947 (vs), 1887 (s)
2056 (vs), 2002 (s), 1778 (w, br)
2047 (vs), 1993 (s), 1784 (w, br)
2060 (vs), 2010 (s), 1775 (w, br)
2057 (vs), 1999 (s), 1771 (w, br)
1993 (s), 1944 (vs), 1869 (m)
123.1
368.3 (br, d)h
417.2 (d)i
375.3 (d)j
83.5
[Mo2Cp2(l-PHR)(l-CO)(CO)4](BArꢀ4)
[Mo2Cp2(l-PHR)(l-CO)(CO)4]BF4
[Mo2Cp2(l-PHR)(l-CO)(CO)4](BArꢀ4)
[Mo2Cp2(l-PHR)(l-CO)(CO)4]BF4
[Mo2ClCp2(l-PHR)(CO)4]
88.0
64.1
Ph
Ph
Ph
61.9
124.9k
a Mes* = 2,4,6-C6H2 Bu3; Mes = 2,4,6-C6H2Me3. b Recorded in dichloromethane solution, unless otherwise stated. c Recorded in CD2Cl2 solutions
at 290 K and 121.50 MHz, unless otherwise stated; d relative to external 85% aqueous H3PO4; J in Hertz. d trans and cis isomers, respectively (see
text) ratio trans : cis = 2. At 193 K, d (trans) 80.4; d (cis) 64.3; ratio trans:cis = 0.5. e Recorded in bis(2-methoxyethyl)ether solution. f Other bands of
the cation could not be identified unambiguously due to partial decomposition of the complex at room temperature. g Recorded at 243 K. h J(PF)
855. i Recorded in CDCl3 solution, J(PF) 880. j J(PF) 911. k If recorded at 161.09 MHz and 203 K then d 123.8 (isomer A) and d 134.8 (isomer B),
ratio A/B = 5 (see text).
t
4 1 7 0
D a l t o n T r a n s . , 2 0 0 4 , 4 1 6 8 – 4 1 7 9

View Article Online
with the value of 37 Hz for the phosphide hydride 1a. This is
considered to be related to the higher strength of the Mo–P
bond in compounds 3a or 3aꢀ, which should have considerable
multiple character.
6c in refluxing dichloromethane, an observation of interest
when considering the reaction pathways operative in all these
reactions, as we will discuss later on.
The protonation reactions of the phenyl derivative 1d reveal
that steric effects are most relevant to determine the stability and
evolution of the cationic intermediates under study. Thus, even
when using [H(OEt2)2](BArꢀ4) we could not detect the presence of
the expected tetracarbonyl cation 5d. Instead, this reaction leads
directly to the pentacarbonyl derivative [Mo2Cp2(l-PHPh)(l-
CO)(CO)4](BArꢀ4) (7d), its yield being increased under a CO
atmosphere as anticipated. Expectedly, no tetracarbonyl cation
could be detected in the reaction of 1d with HBF4·OEt2, but
now the fluoro-derivative [Mo2Cp2(l-H)(l-PFPh)(CO)4] (6d)
was the major species formed, along with some pentacarbonyl
[Mo2Cp2(l-PHPh)(l-CO)(CO)4](BF4) (7dꢀ), the ratio 6d : 7dꢀ
being ca. 6 : 1. As observed for the cyclohexyl substrate, the
formation of the fluoro-derivative 6d can be fully suppressed by
carrying out the protonation reaction under a CO atmosphere,
whereas refluxing dichloromethane solutions of 7dꢀ gives 6d in
moderate yield.
Protonation reactions of compounds 1b–d
The hydride complexes having a substituent on phosphorus
smaller than Mes*, such as Mes, Cy or Ph behave in a way
strongly dependent on that group and also on the nature of
the external anion. However, no phosphinidene species were
detected under any of the conditions examined.
The mesityl derivative 1b reacts readily with either
[H(OEt2)2](BArꢀ4) or HBF4·OEt2 at low temperature to give deep
purple–violet solutions shown (by NMR) to contain a single P-
containing species as major product in each case. The available
spectroscopic data for these species (to be discussed later on)
allow us to identify them as the corresponding phosphide
complexes [Mo2Cp2(l-PHMes)(CO)4]X, (X = BArꢀ4 (5b), BF4
(5bꢀ)] (Chart 4). These cations are thermally unstable, but their
evolution in dichloromethane solution at room temperature is
strongly dependent on the external anion. Thus, the tetraarylbo-
rate salt 5b gives a complex mixture of products which could not
be characterised. In contrast, the tetrafluoroborate salt 5bꢀ gave
the fluorophosphide derivative [Mo2Cp2(l-H)(l-PFMes)(CO)4]
(6b) in moderate yield.
Compound 6d was previously prepared in 11% yield through
the reaction of 1d with [Ph3C][PF6].17 Although the mechanism
responsible for this reaction was not investigated at the time,
we can now envisage that initial hydride abstraction from 1d
would yield the hypothetical tetracarbonyl cation 5d, which
seems electrophilic enough to readily abstract fluoride from the
−
available external ion (BF4 or PF6−).
Solution structure of compounds 5
The available spectroscopic data for compounds 5b, 5bꢀ and
5c indicate that these 32 electron complexes all have the same
structure in solution (Table 2 and Experimental section). In the
first place, the data for 5b and 5bꢀ are very similar to each
other, which suggest the absence of significant cation–anion
interactions. We note, however, that the C–O stretching bands
of cations 5b and 5bꢀ could not be completely identified due
to significant decomposition of the corresponding solutions at
room temperature. Fortunately, a clean IR spectrum can be
obtained for 5c. The pattern of the C–O stretching bands for
this complex is very different from that of the common transoid
structures. This can be explained by assuming a cis relative
arrangement of the MoCp(CO)2 moieties with respect to the
Mo2P plane in these cations, as actually indicated by the 1H and
13C NMR spectra. The latter spectra exhibit single resonances for
the Cp ligands in each case, either at room or low temperatures,
which excludes a fluxional behaviour. These spectroscopic
features are not compatible with transoid geometries, which
would render inequivalent cyclopentadienyl groups. Finally, the
13C NMR spectrum of 5bꢀ is further consistent with the proposed
cis geometry, as it shows just two distinct carbonyl resonances,
these corresponding to the ligands placed either cis or trans with
respect to the P atom.
Chart 4
The cyclohexyl derivative 1c reacts with [H(OEt2)2](BArꢀ4)
to give the corresponding tetracarbonyl [Mo2Cp2(l-
PHCy)(CO)4](BArꢀ4) (5c), analogous to 5b. This salt is
also thermally unstable but evolves differently from 5b, as it
experiences spontaneous carbonylation at room temperature
to give the pentacarbonyl derivative [Mo2Cp2(l-PHCy)(l-
CO)(CO)4](BArꢀ4) (7c) in medium yield. As expected,
compound 7c is formed in good yield when the protonation
reaction is carried out under a CO atmosphere, even at low
temperature.
The tetraarylborate anion has some (although not enough)
stabilizing influence on the tetracarbonyl cation in 5c. This is ev-
idenced by the fact that no tetracarbonyl cation can be detected
in the reaction of 1c and HBF4·OEt2, even at low temperature.
Instead, a mixture containing similar amounts of the corre-
sponding pentacarbonyl [Mo2Cp2(l-PHCy)(l-CO)(CO)4](BF4)
(7cꢀ) and the fluoro-derivative [Mo2Cp2(l-H)(l-PFCy)(CO)4]
(6c) is obtained. The formation of the latter complex can be fully
suppressed by carrying out the protonation reaction under CO,
in which case only the pentacarbonyl 7cꢀ is formed. Interestingly,
the latter can be converted with good yield into the neutral
Structural characterization of fluorophosphide complexes
Spectroscopic data for compounds 6b–d suggest a close struc-
tural similarity between these neutral species and the corre-
sponding hydride precursors 1b–d, so the cyclohexyl and phenyl
compounds exhibit trans geometries while the mesityl product
displays both cis and trans isomers, as will be discussed later
on. The structure of the mesityl species 6b has been determined
through an X-ray study, and it is shown in Fig. 2, while the
relevant bond distances and angles are collected in Table 3. In the
crystal, compound 6b exhibits a transoid relative arrangement
of the MoCp(CO)2 moieties with respect to the averaged Mo2(l-
˚
H)(l-P) plane. The intermetallic length [3.2629(6) A] is slightly
shorter than that found in the cis isomer of 1b, this being
consistent with the slight steric congestion suspected for the
latter. The transoid arrangement of the MoCp(CO)2 fragments
in 6b, however, forces a twist of the mesityl ring by some 40◦ with
D a l t o n T r a n s . , 2 0 0 4 , 4 1 6 8 – 4 1 7 9
4 1 7 1

View Article Online
The other two resonances exhibit a high P–C coupling of ca.
20 Hz, which is consistent with them being cis to phosphorus.
Finally, one of the latter resonances also displays C–F coupling.
By taking into account that three-bond couplings are usually
strongly dependent on the dihedral angle (φ) defined by the
24b,27
bonds involved
(here C–Mo–P–F) we then assign the F-
coupled 13C NMR resonance (d 239.0 ppm) to that carbonyl
group placed opposite to the F atom with respect to the Mo2P
plane (φ ca. 130◦).
We should remark the fact that compound 6c, having a PFCy
bridge, displays fluxional behaviour in solution while its parent
compound 1c, with a similarly sized PHCy bridge, behaves as a
rigid molecule on the NMR time scale. This gives credit to our
previous hypothesis that electronic effects are prevalent in the
dynamic behaviour of [M2Cp2(l-H)(l-PRRꢀ)(CO)4] complexes,
with the electron-withdrawing substituents on the phosphide
ligand increasing the rate of the fluxional process.5
As anticipated above, the mesityl compound 6b displays both
trans and cis isomers in solution, as found for its precursor 1b.
Although we have not studied the corresponding equilibrium
in detail, the presence of a small amount of the cis isomer at
room temperature can be inferred from the IR spectrum, which
displays a weak high frequency band at 1978 cm−1, and from the
fact that the 31P NMR spectrum displays a very broad resonance
for the phosphide ligand at room temperature. From the solution
data for complexes 1a–d and 6b–d we have found that only
those compounds having a mesityl substituent on phosphorus
(l-PHMes or l-PFMes) exhibit in solution significant amounts
of the cis isomer. Currently we can not formulate a satisfactory
explanation for this experimental finding.
Fig. 2 ORTEP view of the molecular structure of compound trans-6b,
with thermal ellipsoids drawn at 30% probability (carbon-bound
hydrogen atoms and some of the labels omitted for clarity).
◦
˚
Table 3 Selected bond lengths (A) and angles ( ) for compound 6b
Mo(1)–Mo(2)
Mo(1)–H(1)
Mo(1)–P(1)
Mo(1)–C(6)
Mo(1)–C(7)
Mo(2)–H(1)
Mo(2)–P(1)
Mo(2)–C(13)
Mo(2)–C(14)
P(1)–F(1)
3.2629(6)
1.86(4)
2.3959(9)
1.972(4)
1.963(4)
1.82(3)
C(6)–Mo(1)–C(7)
C(6)–Mo(1)–P(1)
C(6)–Mo(1)–H(1)
C(6)–Mo(1)–Mo(2)
C(14)–Mo(2)–C(13)
C(14)–Mo(2)–P(1)
78.14(16)
116.43(12)
78.9(11)
83.91(13)
80.12(17)
73.74(12)
134.8(11)
119.28(12)
85.07(3)
2.4303(10) C(14)–Mo(2)–H(1)
1.938(4)
1.968(4)
1.623(2)
1.836(3)
C(14)–Mo(2)–Mo(1)
Mo(2)–P(1)–Mo(1)
C(15)–P(1)–F(1)
Structural characterization of pentacarbonyl compounds
96.01(13)
94.79(11)
The structure of the tetrafluoroborate salt of the cyclohexyl
cation 7cꢀ has been determined through an X-ray study. There
are two independent but similar cations in the unit cell; one
is depicted in Fig. 3, while the relevant bond distances and
angles are collected in Table 4. The cation displays two almost
eclipsed MoCp(CO)2 moieties bridged by cyclohexylphosphide
and carbonyl ligands. These bridges define a quite puckered (ca.
120◦) MoPMoC central skeleton, which allows the terminal CO
ligands to define two almost parallel planes. As found for the
mesityl compound 1b, the substituent ring on phosphorus is
pointing away from the cyclopentadienyl ligands thus minimiz-
ing the steric repulsions.
P(1)–C(15)
C(13)–Mo(2)–P(1)
respect to the plane normal to the intermetallic vector. This is
surely the origin of the slight asymmetry of the Mo–P bonds
˚
[2.396(1) and 2.430(1) A]. All other geometrical parameters are
similar to those found in 1b and are not unusual. Incidentally,
one of the cyclopentadienyl rings appears disordered in two
positions (with 66 and 33% occupancy) related by a libration or
rotatory movement. This phenomenon is relatively common for
cyclopentadienyl complexes.23
In solution, the IR spectra of compounds 6 exhibit C–O
stretching frequencies with a pattern similar to their phosphide
precursors 1, but shifted some 10 cm−1 to higher frequency as
a consequence of the replacement of H by F at the phosphide
bridge. This also causes a strong deshielding of ca. 300 ppm
at the 31P nuclei, which then are found in the NMR spectra
at ca. 400 ppm. These resonances appear as highly coupled
doublets with J(PF) values around 900 Hz, as expected for
directly bonded 31P and 19F atoms.24 Comparison between the 31P
and 19F data for 6d (not given in the original report by Woodward
et al.)17 and those for the mesityl compound 6b reveal significant
differences concerning the fluorine atom, which for 6b exhibits
a lower P–F coupling (855 vs. 911 Hz) and higher chemical shift
(−93.7 vs. −122.0 ppm). This might be another indication of
steric pressure at the mesityl compound.
The intermetallic bond length in either of the two in-
˚
dependent cations (ca. 3.19 A) is significantly shorter than
˚
those found for 1b [3.2868(2) A] or other neutral phos-
phide hydride complexes of type [Mo2Cp2(l-H)(l-PR2)(CO)4]
The phenyl compound 6d was reported to be fluxional in
solution.17 The same is found for the cyclohexyl compound
6c, which at room temperature displays a single 1H NMR
cyclopentadienyl resonance, itself inconsistent with a trans
1
structure. At 243 K, however, the H and 13C NMR spectra
are consistent with the static structure, with two inequivalent Cp
rings and four distinct carbonyl ligands. By recalling that 2J(PC)
couplings in complexes of the type [MCpX(CO)2(PR3)] (M =
Mo, W; X = halogen, alkyl, hydride, etc.) usually follow the order
Fig. 3 ORTEP view of the molecular structure of the cation in
compound 7cꢀ, with thermal ellipsoids drawn at 30% probability
(carbon-bound hydrogen atoms and some of the labels omitted for
clarity).
J
cis > Jtrans,
25,26 we can safely assign the singlet resonances at 234.0
and 231.8 ppm to the carbonyl ligands trans to phosphorus.
4 1 7 2
D a l t o n T r a n s . , 2 0 0 4 , 4 1 6 8 – 4 1 7 9

View Article Online
◦
ꢀ
˚
Table 4 Selected bond lengths (A) and angles ( ) for compound 7c
finding. Currently, we can not offer an explanation for these
contradictory observations.
The H NMR spectra for all compounds 7 exhibit a single
1
Mo(1)–Mo(2)
Mo(1)–C(1)
Mo(1)–P(12)
Mo(1)–C(11)
Mo(1)–C(12)
Mo(2)–C(1)
Mo(2)–P(12)
Mo(2)–C(21)
Mo(2)–C(22)
C(1)–O(1)
3.1984(12)
2.273(12)
2.414(3)
1.981(12)
2.033(11)
2.149(11)
2.443(3)
2.031(12)
2.047(12)
1.152(13)
1.847(9)
C(11)–Mo(1)–C(12)
C(11)–Mo(1)–P(12)
C(11)–Mo(1)–C(1)
C(11)–Mo(1)–Mo(2)
C(22)–Mo(2)–C(21)
C(22)–Mo(2)–P(12)
C(22)–Mo(2)–C(1)
C(22)–Mo(2)–Mo(1)
Mo(2)–P(12)–Mo(1)
Mo(2)–C(1)–Mo(1)
C(1)–Mo(1)–P(12)
82.1(5)
81.9(3)
131.0(4)
90.9(3)
83.1(4)
134.4(3)
82.9(5)
89.1(3)
82.36(8)
92.6(4)
75.6(3)
resonance for the cyclopentadienyl ligands, which suggests that
the cis geometry of the cation in the solid state is retained
in solution. The 13C NMR spectra of 7c indicated, however,
the presence of a fluxional process only involving the carbonyl
ligands. Thus, the spectrum recorded at 203 K is consistent with
the static structure, exhibiting three carbonyl resonances corre-
sponding respectively to the bridging carbonyl [d 270.0 ppm,
J(HP) 22 Hz], two terminal carbonyls cis to phosphorus [d
228.4 ppm, J(HP) 17 Hz] and two terminal carbonyls trans
to phosphorus [d 220.3 ppm, J(HP) ca. 0 Hz]. However, just
a broad averaged carbonyl resonance at ca. 230 ppm is observed
in the spectrum recorded at 290 K. Apart from this, other 13C
P(12)–C(121)
18
˚
(ca. 3.25–3.28 A). This shortening effect is surely due to
the orbital contraction derived from the positive charge in
the molecule. On the other hand, the intermetallic sepa-
rations in 7cꢀ are still higher than that measured for the
isoelectronic diphosphine-bridged complex [W2Cp2(l-Cl)(l-
CO)(CO)2(l-Ph2PCH2PPh2)](PF6), [3.040(3) A]. This differ-
ence can be related to the different number of ligands (two
vs. three) bridging the dimetal centre in these two cations. The
positive charge of the complex is likely to be responsible also for
the significant lengthening of the Mo–carbonyl bonds, which
display values around 2.03 A, well above the usual values of ca.
1.95 A found in neutral complexes as 1b, 6b and related species.
1
resonances remained essentially unchanged, and the H NMR
spectrum was also roughly temperature-independent.
The above observations can be explained by assuming the
existence for compounds 7 of a fluxional process as illustrated in
Scheme 3. The key proposal is that opening of the bridging CO
can occur. This would lead to an intermediate A containing
a MoCp(CO)3P moiety where cis/trans isomerization could
easily occur in a similar way to that found in mononuclear
MoCp(CO)2LR (L = phosphine ligand, R = H, alkyl) species.32
By extending this type of rearrangements, complete scrambling
of the CO ligands is then achieved.
28
˚
˚
˚
This lengthening is justified by the decreased back p-donation
from the metal to the CO ligand and is also consistent with the
relatively easy decarbonylation experienced by complexes 7 (i.e.
the transformation 7cꢀ/6c).
In the crystal, compound 7cꢀ displays two significant an-
ion/cation interactions. The most significant one occurs in
molecule A, where there is a close P(12) · · · F(22) distance of
◦
˚
3.538 A, defining a P(12) · · · F(22)-B(2) angle of 108.5 . By
˚
considering a normalized P–H bond length of ca. 1.42 A,
˚
we can estimate the H · · · F separation to be around 2.15 A,
well below the sum of covalent radii for F and H atoms (ca.
˚
2.67 A). All this points to the presence of a classical linear
P–H · · · FBF3 hydrogen bond of moderate strength.29,30 As we
will see later on, our NMR data suggest that some of this
cation–anion interaction is retained in solution. In contrast
with this, molecule B does not exhibit any significant contacts
involving the P–H hydrogen atom. Instead, some weaker and less
−
defined contacts between fluorine atoms of BF4 and H atoms
of the cyclopentadienyl ligands [F(13) · · · H(4111) = 2.351 A;
˚
˚
F(11) · · · H(3153) = 2.391 A] can be identified.
Scheme 3 Fluxional process proposed for compounds 7 in solution.
Spectroscopic data in solution for all four compounds 7 are
consistent with the solid state structure of 7cꢀ and also reveal
that (a) the cis structure of the cation is retained in solution and
(b) significant anion–cation interactions remain in the solution
of the tetrafluoroborate salts. The latter is clearly denoted by
comparison of the C–O stretching bands in the pairs 7c/7cꢀ and
7d/7dꢀ. All these spectra exhibit three C–O stretching bands, one
in the bridging region and the other two in the region of terminal
carbonyls and with the relative intensities expected for M(CO)2
oscillators having CMC angles below 90◦.31 However, the
tetrafluoroborate salts exhibit C–O frequencies systematically
lower (5–10 cm−1) than the corresponding tetraarylborate salts
Protonation reactions with HCl
In order to gain further insight into the processes leading
to the fluoro-derivatives 6, we carried out some protonation
experiments with HCl, an acid having now a clearly coor-
dinating anion. Reaction of the latter with the cyclohexyl
compound 1c led to a mixture of products which could not
be separated nor properly identified. In contrast, bubbling HCl
through a dichloromethane solution of the phenyl compound
1d gave rapidly the neutral chloro-derivative [Mo2ClCp2(l-
PHPh)(CO)4] (8) in good yield.
−
(Table 2), which suggests significant ion-pairing for the BF4
salts. In line with this, the 31P shielding of the phosphide ligands
experiences significant shifts upon changing the external anion.
The structure of compound 8 is shown in Fig. 4, while the
relevant bond distances and angles are collected in Table 5. The
molecule is made up by two MoCp(CO)2 units bridged by the
PHPh group. The chloride ligand is terminally bonded to one
of the Mo atoms, in a position cis to the phosphide ligand. A
metal–metal bond of order one should be formulated for this
complex according to the EAN rule. This is consistent with the
1
The H NMR data for compounds 7c and 7cꢀ reveal that
the above cation–anion interaction involves the P–H hydrogen
atom, as both its chemical shift and P–H coupling are strongly
modified when replacing the tetraarylborate anion (dH 4.75 ppm,
1J(HP) 354 Hz) by tetrafluoroborate (dH 5.52 ppm, 1J(HP)
381 Hz). The downfield shift observed for 7cꢀ is expected if
a hydrogen bond interaction (P–H · · · F) exists in solution, as
found in the crystal. However, this should also induce a decrease
in the value of the P–H coupling, contrary to the experimental
˚
intermetallic separation of 3.2315(3) A, a value intermediate
ꢀ
˚
between those measured in the cation 7c [3.1981(1) A] and the
˚
neutral hydrides 1b and 6b (ca. 3.27 A). There are significant
differences in the Mo–C lengths of carbonyl ligands which can
D a l t o n T r a n s . , 2 0 0 4 , 4 1 6 8 – 4 1 7 9
4 1 7 3

View Article Online
Chart 5
In fact, we propose that isomer B has the same structure
as A, but with the chloride ligand placed opposite to the P–H
bond with respect to the Mo2P plane (Chart 5). The presence
of two asymmetric isomers, however, does not itself explain
the appearance of just a single cyclopentadienyl resonance at
290 K for compound 8. Although we have not studied this
process in detail, it is quite likely that chloride dissociation
could take place in solution to a small extent. This would
generate small amounts of the tetracarbonyl cation similar to
compounds 5 (Scheme 4). Recombination of Cl− with that
intermediate cation could happen at either of the molybdenum
atoms, thus justifying the averaging of the inequivalent Cp
resonances. Besides, recombination of Cl− could also take
place on either side of the Mo2P plane, thus rationalising
the interconversion between isomers A and B. Although this
hypothesis is simple and attractive, other possibilities such as
intramolecular rearrangements can not be excluded at this time.
Fig. 4 ORTEP view of the molecular structure of compound 8, with
thermal ellipsoids drawn at 30% probability (carbon-bound hydrogen
atoms and some of the labels omitted for clarity).
◦
˚
Table 5 Selected bond lengths (A) and angles ( ) for compound 8
Mo(1)–Mo(2)
Mo(1)–Cl(1)
Mo(1)–P(1)
Mo(1)–C(1)
Mo(1)–C(2)
Mo(2)–P(1)
Mo(2)–C(3)
Mo(2)–C(4)
P(1)–H(2)
3.2315(3)
2.5296(8)
2.4381(8)
2.018(4)
2.006(3)
2.3857(8)
1.969(4)
1.934(4)
1.22(4)
C(1)–Mo(1)–C(2)
C(1)–Mo(1)–P(1)
C(1)–Mo(1)–Cl(1)
C(2)–Mo(1)–Cl(1)
C(4)–Mo(2)–C(3)
C(4)–Mo(2)–P(1)
Cl(1)–Mo(1)–P(1)
C(3)–Mo(2)–P(1)
C(31)–P(1)–H(2)
Mo(2)–P(1)–Mo(1)
83.21(14)
121.36(11)
84.64(11)
152.61(11)
80.84(16)
84.71(11)
82.84(3)
119.13(12)
100.2(19)
84.11(2)
P(1)–C(31)
1.822(3)
be attributed to the different coordination numbers around
Mo(1) and Mo(2). Thus, those carbonyls on Mo(1) exhibit
˚
higher M–C lengths (ca. 2.01 A) than those on Mo(2) (ca.
˚
1.95 A). A similar structural feature has been also found in the
=
isoelectronic complex [WMoClCp2(l-HC CHPh)(CO)4], an
species prepared from the alkyne compound [WMoCp2(l-g :g -
2
2
CHCPh)(CO)4] and HCl.33 In contrast, the equally isoelectronic
=
dimolybdenum complexes [Mo2Cp2X(l-HC CH2)(CO)4] (X =
Cl, Br, O2CMe, O2CCF3) were found to display a semibridging
carbonyl, as judged from spectroscopic data and an X-ray
structure analysis of the trifluoroacetate derivative.34 As for the
˚
Mo–Cl bond, the value of 2.5296(8) A in compound 8 compares
well with the corresponding figure in the above MoW complex
˚
(W–Cl = 2.508 A) and can be considered as normal. Finally,
the differences in the coordination environments around the
metal atoms seem to be partially compensated by the phosphide
ligand, which binds Mo(2) more strongly, as judged from the
Scheme 4 Dissociative process proposed for compound 8 in solution
(R = Ph).
˚
significantly different P(1)–M lengths [2.386(1) vs. 2.438(1) A].
Compound 8 exhibits dynamic behaviour in solution. This
is readily apparent since the H NMR spectrum at room tem-
Reactions pathways in the protonation reactions of the hydride
complexes 1
1
perature exhibits just a single cyclopentadienyl resonance, this
being inconsistent with the solid state structure. Analysis of the
low-temperature NMR data (Table 2 and Experimental section)
reveals the presence of two isomers A and B interconverting
The initial aim of the present work was to gain some insight
into the protonation reactions of hydride phosphide complexes
of the type [Mo2Cp2(l-H)(l-PHR)(CO)4] 1a–d, especially con-
cerning the activation of the P–H bond and the involvement
of intermediate cations having multiple metal–metal bonds.
By combining the results obtained when using acids having
anions of distinct coordination abilities (BArꢀ4−, BF4−, Cl−)
interacting with substrates having different electronic and steric
characteristics at the phosphide ligand, we can get a general
picture of the dominant processes under operation (Scheme 5).
Upon protonation, the first step in all cases would be the
elimination of dihydrogen from the incoming proton and the
bridging hydride. As we have noted above, this is a well known
process for terminal M–H bonds,4 but to our knowledge a
similar reaction involving a bridging hydride ligand has not
1
rapidly on the NMR timescale (Chart 5). Separate 31P, H and
13C NMR resonances could be observed at 203 K, (ratio A : B =
5 : 1 at this temperature), which revealed that both isomers
display inequivalent cyclopentadienyl resonances. The 13C NMR
spectrum of the major isomer A is fully consistent with the
crystal structure of compound 8, and displays four inequivalent
carbonyl resonances in the terminal region at 239.4(s), 237.7(s),
237.2 [d, J(CP) 20 Hz] and 227.4 [d, J(CP) 14 Hz] ppm, their
multiplicity being consistent with two ligands trans and two
other cis with respect to the phosphide ligand. From the available
data for the minor isomer, we conclude that its structure would
be quite similar to that of the major species.
4 1 7 4
D a l t o n T r a n s . , 2 0 0 4 , 4 1 6 8 – 4 1 7 9

View Article Online
to occur in the reaction of the triply bonded [M2Cp2(CO)4]
ꢀ
6a,17
with PHRR ligands.
Although some nucleophilic attacks
on phosphide ligands have been reported, these usually require
the use of strong nucleophiles, such as organolithium reagents.35
In our case, it seems unlikely that the phosphorus atom at the
tetracarbonyl cations 5 would be electrophilic enough to abstract
fluoride from the tetrafluoroborate ion. Instead, we propose
that tetracarbonyls 5 would experience to a small extent the
P–H bond cleavage leading to phosphinidene hydrides of type
3 even for small R groups on phosphorus. In the presence
of BF4−, however, this would allow the binding of fluoride
at the more accessible phosphinidene phosphorus atom, thus
explaining the formation of fluorophosphide complexes 6b-d.
Nucleophilic attack at the P atom is a well established reaction
for terminally bound phosphinidene ligands,11 but we can quote
no related examples involving bridging phosphinidene ligands
−
or BF4 donors.
Conclusion
Proton attack on the hydride-bridged compounds 1 in-
duces dihydrogen elimination yielding the unsaturated cations
[Mo2Cp2(l-PHR)(CO)4]+ (5), which then evolve differently
depending mainly on the size of the R group on phosphorus.
For the very bulky Mes* group, rapid oxidative addition
of the P–H bond occurs to give the hydride phosphinidene
cation [Mo2Cp2(l-H)(l-PMes*)(CO)4]+. For other R groups
the same process is proposed to occur to a small extent, this
Scheme 5 Reaction pathways in the protonation reactions of hydride
complexes 1 [M = Mo(CO)2Cp; R = Mes*, Mes, Cy, Ph].
been reported so far. Indeed, our data indicate that the P–
H bond is not involved at this stage. For example, no deu-
terium was found at the hydride position of the corresponding
products when reacting the deuterated substrates [Mo2Cp2(l-
D)(l-PHMes*)(CO)4] or [Mo2Cp2(l-D)(l-PHCy)(CO)4] with
HBF4·OEt2. This strongly suggests that the hydride ligand in
compounds 3 and 6 corresponds to the original P–H hydrogen
atom.
−
being followed by fluoride abstraction from external BF4
at the phosphinidene position, thus explaining the formation
of the fluorophosphide products [Mo2Cp2(l-H)(l-PFR)(CO)4].
Otherwise, the unsaturated tetracarbonyl cations evolve by
spontaneous carbonylation to give the electron-precise pen-
−
tacarbonyls [Mo2Cp2(l-PHR)(l-CO)(CO)4]+. The use of BArꢀ4
as a counterion expectedly suppresses the fluoride abstraction
process, but adds little stability to the unsaturated tetracarbonyl
cations of type 5.
The H+-induced dehydrogenation of the hydrides 1a-d would
give the corresponding tetracarbonyl cations 5, which are
observable species when the phosphide ligand has mesityl or
cyclohexyl groups. The chemical evolution of these unsaturated
cations would then be critically dependent on the substituent
on phosphorus. By considering that the electronic similarity of
the Mes*, Mes and Ph substituents is not paralleled by the
products obtained, we conclude that the chemical behaviour
of these unsaturated cations is mainly governed by the size of
the substituents on phosphorus. For the biggest Mes* group,
oxidative addition of the P–H bond to the Mo–Mo double bond
is the dominant process. This possibly relieves some of the steric
pressure in the cation, as it leaves the very bulky Mes* group at a
less congested trigonal (rather than tetrahedral) phosphinidene
environment. At the same time, the dimetal centre thus becomes
electron-precise.
Experimental
General comments
All manipulation and reactions were carried out under a
nitrogen atmosphere using standard Schlenk techniques. Sol-
vents were purified according to literature procedures and
distilled prior to use.36 Petroleum ether refers to that fraction
distilling in the range 65–70 ◦C. Compounds [Mo2Cp2(CO)4],37
t
[Li(THF)3][PHR] (R = 2,4,6-C6H2 Bu3 or Mes*,38 2,4,6-
C6H2Me3 or Mes),39 [Mo2Cp2(l-H)(l-PHR)(CO)4], (R = Cy,5
Ph)16 and [H(OEt2)2](BArꢀ4), [Arꢀ = 3,5-C6H3(CF3)2]13 were
prepared as described previously. Compounds [Mo2Cp2(l-D)(l-
PHR)(CO)4] were prepared as the non-deuterated species, but
using 85% D3PO4 in D2O at the protonation step. All other
reagents were obtained from the usual commercial suppliers
and used as received. Chromatographic separations were carried
out using jacketed columns refrigerated by tap water. Silica
gel (230–400 mesh), Florisil (100–200 mesh) or aluminium
oxide (activity I, 150 mesh) were purchased from Aldrich and
degassed under vacuum prior to use. The latter oxide was mixed
afterwards under nitrogen with the appropriate amount of water
to reach the activity desired. Filtrations were performed using
diatomaceous earth. NMR spectra were routinely recorded
For the cationic intermediates having smaller R groups no
P–H cleavage occurs at a significant extent, and spontaneous
carbonylation (which obviously requires partial decomposition
of the cation) is the dominant process that relieves the electronic
−
unsaturation of the dimetal centre. In the presence of the BF4
counterion, however, fluoride abstraction competes efficiently
with carbonylation, especially for the smallest Ph derivative. As
the fluoride complexes 6b-d are either obtained from detectable
tetracarbonyls (5bꢀ or 5cꢀ) or through decarbonylation reactions
of pentacarbonyls (7cꢀ or 7dꢀ), we propose that the tetracarbonyl
cations 5 are at the origin of all the fluoroderivatives 6.
Fluoride abstraction, however, could occur in different ways.
Initial fluoride coordination at the metallic position12,28 is
unlikely, as this would yield a compound similar to the chloro-
complex 8, which would not be expected to rearrange any fur-
ther. Alternatively, the fluoride abstraction might occur directly
at the P atom, thus generating an unsaturated intermediate A
containing a fluorophosphine PHFR ligand which then would
experience the oxidative addition of the P–H bond, as presumed
1
1
at 300.13 (1H), 188.28 (19F{ H}), 121.50 (31P{ H}), 100.62
(13C{ H}) MHz in CD2Cl2 at room temperature unless otherwise
1
indicated. Chemical shifts (d) are given in ppm, relative to
internal TMS (1H, 13C), internal CFCl3 (19F) or external 85%
aqueous H3PO4 solutions (31P). Coupling constants (J) are given
in Hz.
Preparation of [Mo2Cp2(l-H)(l-PHMes*)(CO)4] (1a).
A
diglyme solution (30 ml) containing ca. 1.5 mmol of
D a l t o n T r a n s . , 2 0 0 4 , 4 1 6 8 – 4 1 7 9
4 1 7 5

View Article Online
[Mo2Cp2(CO)4] was prepared in situ from [Mo2Cp2(CO)6] (0.
745 g, 1.5 mmol) according to the literature procedure.37 This
was transferred using a cannula over solid [Li(THF)3][PHMes*],
freshly prepared from H2PMes* (0.435 g, 1.57 mmol) and BuLi
(0.990 ml of an 1.6 M solution in hexanes, 1.58 mmol) in THF
(15 ml), by literature methods.38 The mixture was then stirred
for 10 min to yield a dark green solution of the anionic complex
2a (see text). Aqueous 85% H3PO4 (0.4 ml, excess) was then
added, and the mixture was stirred for 5 min to give a deep-
red solution. Removal of the solvent under vacuum gave a
residue which was extracted with dichloromethane–petroleum
ether (1 : 1) and filtered. Removal of solvents from the filtrate
gave a residue which was then dissolved in the minimum amount
of dichloromethane and chromatographed on alumina (activity
IV, 40 × 4 cm). Elution with dichloromethane–petroleum ether
(1 : 6) gave a red band. Removal of the solvents from the latter
yielded compound 1a as a red microcrystalline solid (0.890 g,
83%) (Found: C, 53.67; H, 5.64. C32H41O4PMo2 requires C,
53.94; H, 5.80%); dH (400.13 MHz, 243 K) 7.94 [d, J(HP) 355,
1H, PH], 7.46, 7.28 (2 × s, 2 × 1H, C6H2), 5.52, 5.18 (2 × s,
Preparation of [Mo2Cp2(l-H)(l-PMes*)(CO)4](BF4) (3aꢀ).
solution of compound 1a (0.900 g, 1.26 mmol) in
A
dichloromethane (20 ml) was stirred with HBF4·OEt2 (500 ll
of a 54% Et2O solution, 2.73 mmol) for 5 min to give a black–
green mixture. Solvent was then removed under vacuum and
the residue was washed with dichloromethane–petroleum etlher
(1 : 4, 3 × 10 ml) to give a black microcrystalline solid (0.925 g,
92%) (Found: C, 48.03; H, 5.09. C32H40BF4Mo2O4P requires C,
48.14; H, 5.05%); dH (200.13 MHz) 7.63 [d, J(HP) 4, 2H, C6H2],
t
t
5.76 (s, 10H, Cp), 1.45 (s, 18H, Bu), 1.37 (s, 9H, Bu), −9.12
[d, J(HP) 51, 1H, l-H]; dC 228.6 [d, J(CP) 30, 2 × CO], 222.9
(s, 2 × CO), 155.9 [s, C4(C6H2)], 151.9 [s, C2(C6H2)], 142.4 [d,
J(CP) 23, C1(C6H2)], 125.2 [d, J(CP) 8, C3(C6H2)], 94.9 (s, Cp),
39.5 (s, 2 × CMe3), 35.6 (s, CMe3), 33.7 (s, 6 × Me) 31.0 (s, 3 ×
Me).
Preparation of [Mo2Cp2(l-PMes*)(CO)4] (4). A solution of
compound 1a (1.00 g, 1.4 mmol) in dichloromethane (20 ml)
was stirred with HBF4·OEt2 (300 ll of a 54% Et2O solution,
2.1 mmol) for 5 min to give a black–green solution of compound
3aꢀ. Distilled water (0.5 ml, excess) was then added, and the
mixture was further stirred for 5 min and then filtered through
alumina (activity IV, 20 × 2 cm). Removal of the solvents from
the filtrate yielded compound 4 as a black microcrystalline solid
(0.950 g, 95%) (Found: C, 53.98; H, 5.39. C32H39O4PMo2 requires
C, 54.09; H, 5.49%). Spectroscopic data for this product were in
agreement with those reported for compound 4 in reference 2.
t
2 × 5H, Cp), 1.64, 1.30, 1.22 (3 × s, 3 × 9H, Bu), −12.98 [d,
J(HP) 36, 1H, l-H]; dH (200.13 MHz, 290 K) 7.92 [d, J(HP) 353,
1H, PH], 7.38 [d, J(HP) 2, C6H2], 5.31 (br, 10H, Cp), 1.44 (br,
18H, tBu), 1.30 (s, 9H, tBu), −12.88 [d, J(HP) 37, 1H, l-H]; dC
(100.62 MHz, 213 K) 245.9 [d, J(CP) 27, CO], 242.3 [d, J(CP)
19, CO], 239.0, 237.8 (2 × s, 2 × CO), 155.4 [d, J(CP) 8, C2 or
C6(C6H2)], 155.2 [s, C4(C6H2)], 149.6 [s, C6 or C2(C6H2)], 138.1
[d, J(CP) 12, C1(C6H2)], 122.9, 122.5 [2 × s, br, C3,5(C6H2)], 92.2,
91.6 (2 × s, Cp), 38.9, 38.7, 35.0 (3 × s, 3 × CMe3), 34.6, 33.7,
31.0 (3 × s, Me).
Preparation of solutions of [Mo2Cp2(l-PHMes)(CO)4](BArꢀ4)
(5b). Compound 1b (0.030 g, 0.051 mmol) and [H(OEt2)2]
(BArꢀ4) (0.055 g, 0.054 mmol) were agitated in CH2Cl2 (5 ml)
or CD2Cl2 (0.6 ml) at 0 ◦C for 1 min to give purple–violet
solutions shown (by NMR) to contain compound 5b as a major
species. These solutions decomposed progressively upon storage
at room temperature or any further manipulation (filtration,
crystallization, etc.). dH (400.13 MHz, 243 K) 9.08 [d, J(HP)
417, 1H, PH], 7.76 (s, 8H, Arꢀ), 7.59 (s, 4H, Arꢀ), 7.01 (s, 2H,
C6H2), 5.83 (s, 10H, Cp), 2.32 (s, 3H, Me), 2.21 (s, 6H, Me).
Preparation of [Mo2Cp2(l-H)(l-PHMes)(CO)4] (1b). A so-
lution of the anionic complex 2b was prepared as de-
scribed above, from ca. 1 mmol of [Mo2Cp2(CO)4] and
[Li(THF)3][PHMes], the latter freshly prepared from H2PMes
(0.160 g, 1.05 mmol). The dark green resulting solution was
then stirred at 0 ◦C with an excess of [NH4][PF6] for 20 min to
give a brown mixture. Removal of the solvent under vacuum
gave a residue which was extracted with toluene and filtered.
Removal of the solvent from filtrate gave a residue which
was then extracted with toluene–petroleum ether (1 : 1) and
chromatographed on Florisil (40 × 4 cm). Elution with the same
mixture gave first a minor brown fraction and then a red–orange
fraction. Removal of the solvents from the latter fraction yielded
compound 1b as an orange microcrystalline solid (0.325 g, 54%).
The crystals used in the X-ray study were grown by slow diffusion
of petroleum ether into a toluene solution of the complex at
Preparation of solutions of [Mo2Cp2(l-PHMes)(CO)4](BF4)
(5bꢀ). The procedure is identical to that described for 5b but
using HBF4·OEt2 (15 ll of a 54% Et2O solution, 0.109 mmol)
instead, and a temperature of 243 K. This gives purple–
violet solutions containing compound 5bꢀ as major species.
Manipulation of these solutions at room temperature causes
the decomposition of the complex to give compound 6b and
other uncharacterized products. dH (400.13 MHz, 243 K) 9.01
[d, J(HP) 417, 1H, PH], 7.05 (s, 2H, C6H2), 5.86 (s, 10H, Cp),
2.35 (s, 6H, Me), 2.23 (s, 3H, Me); dC (100.62 MHz, 223 K) 228.1
(s, br, 2 × CO), 225.5 (s, br, 2 × CO), 141.8 [s, C4(C6H2)], 141.3
[d, J(CP) 8, C2(C6H2)], 130.9 [s, C3(C6H2)], 126.8 [d, J(CP) 43,
C1(C6H2)], 96.0 (s, Cp), 24.1 (s, 2 × Me), 20.8 (s, Me).
Preparation of solutions of [Mo2Cp2(l-PHCy)(CO)4](BArꢀ4)
(5c). The procedure is identical to that described for 5b,
but using compound 1c (0.028 g, 0.050 mmol) instead. This
gives dark–brown solutions containing compound 5b as major
species. These solutions decompose progressively upon storage
or manipulation at room temperature to give compound 7c as
major product. dH (200.13 MHz) 7.93 [d, J(HP) 387, 1H, PH],
7.75 (s, 8H, Arꢀ), 7.58 (s, 4H, Arꢀ), 5.81 (s, 10 H, Cp), 1.60–2.10
(m, 11H, Cy).
◦
−20 C (Found: C, 47.20; H, 4.02. C23H23Mo2O4P requires C,
47.12; H, 3.95%); dH (200.13 MHz, 290 K) 7.35 [br, d, J(HP)
340, 1H, PH], 6.88 [d, J(HP) 3, 2H, C6H2], 5.17 (s, 10H, Cp),
2.25 (s, 3H, Me), 2.18 (br, s, 6H, Me), −12.2 [br, d, J(HP) 42,
1H, l-H]; dH (400.13 MHz, 193 K, isomer cis-1b) 7.75 [d, J(HP)
333, 1H, PH], 7.09, 6.80 (2 × s, 2 × 1H, C6H2), 5.23 (s, 10H, Cp),
2.70, 2.28, 1.45 (3 × s, 3 × 3H, Me), −12.29 [d, J(HP) 40, 1H,
l-H]; dH (400.13 MHz, 193 K, isomer trans-1b): 7.12 [d, J(HP)
360, 1H, PH], 7.05, 6.80 (2 × s, 2 × 1H, C6H2), 5.21, 5.20 (2 × s,
2 × 5H, Cp), 2.79, 2.27, 1.68 (3 × s, 3 × 3H, Me), −12.14 [d,
J(HP) 41, 1H, l-H]. Ratio cis : trans = 0.5 : 1 (293 K), 1.2 : 1
(243 K), 2 : 1 (193 K).
Preparation of [Mo2Cp2(l-H)(l-PMes*)(CO)4][BArꢀ4] (3a).
A solution of compound 1a (0.090 g, 0.126 mmol) in
dichloromethane (10 ml) was stirred with [H(OEt2)2](BArꢀ4)
(0.132 g, 0.13 mmol) for 5 min to give a dark green solution.
Solvent was then removed under vacuum and the residue washed
with petroleum ether (2 × 5 ml) to give compound 3a as a black
powder (0.178 g, 90%). dH (200.13 MHz) 7.73 (s, 8H, Arꢀ), 7.46
[d, J(HP) 5, 2H, C6H2], 7.56 (s, 4H, Arꢀ), 5.66 (s, 10H, Cp), 1.45
(s, 18H, tBu), 1.37 (s, 9H, tBu), −9.27 [d, J(HP) 51, 1H, l-H].
Preparation of [Mo2Cp2(l-H)(l-PFMes)(CO)4] (6b). A solu-
tion of compound 1b (0.090 g, 0.154 mmol) in dichloromethane
(15 ml) was stirred with HBF4·OEt2 (70 ll of a 54% Et2O
solution, 0.508 mmol) for 10 min at room temperature. Removal
of the solvent under vacuum gave a brown residue which was
extracted with toluene and filtered to give an orange solution.
The latter was concentrated under vacuum and then added
a layer of diethyl ether and petroleum ether. After complete
diffusion of these solvents at −20 ◦C, orange crystals of
4 1 7 6
D a l t o n T r a n s . , 2 0 0 4 , 4 1 6 8 – 4 1 7 9

View Article Online
compound 6b were obtained (0.031 g, 33%). These crystals were
used in the X-ray diffraction study of the complex (Found: C,
45.54; H, 3.98. C23H22FMo2O4P requires C, 45.71; H, 3.64%);
dH (200.12 MHz) 6.91 (s, 2H, C6H2), 5.27 (s, br, 10H, Cp), 2.29
(s, 6H, Me), 2.27 (s, 3H, Me), −10.77 [d, J(HP) 45, 1H, l-H]; dF
−93.7 [d, JFP = 855, PF].
from 1d (0.020 g, 0.037 mmol) and [H(OEt2)2](BArꢀ4) (0.040 g,
0.040 mmol) compound 7d was isolated as a red–orange powder
(0.047 g, 89%) (Found: C, 43.92; H, 1.81. C53H28BF24Mo2O5P
requires C, 44.38; H, 1.97%).
Preparation of [Mo2Cp2(l-PHPh)(l-CO)(CO)4](BF4) (7dꢀ).
The procedure is identical to that described for 7cꢀ. By start-
ing from 1d (0.050 g, 0.092 mmol) and HBF4·OEt2 (30 ll
of a 54% Et2O solution, 0.218 mmol) compound 7dꢀ was
isolated as a red powder (0.056 g, 92%) (Found: C, 39.05;
H, 2.52. C21H16BF4Mo2O5P requires C, 38.33; H, 2.45%); dH
(200.13 MHz) 7.60–7.40 (m, 5H, Ph), 6.67 [d, J(HP) 384, 1H,
PH], 5.43 (s, 10H, Cp).
Preparation of [Mo2Cp2(l-H)(l-PFCy)(CO)4] (6c). A solu-
tion of compound 7cꢀ (0.090 g, 0.136 mmol) in dichloromethane
(15 ml) was refluxed for 1 h to give a yellow-orange so-
lution. Solvent was then removed under vacuum and the
residue chromatographed on a silica gel column. Elution with
dichloromethane/petroleum ether (1 : 1) gave a yellow fraction.
Removal of the solvents from the latter fraction yielded com-
pound 6c (0.033 g, 42%) as an orange microcrystalline solid
(Found: C, 41.35; H, 3.84. C20H22FMo2O4P requires C, 42.27;
H, 3.90%); dH (400.13 MHz, CDCl3, 290 K) 5.26 (s, br, 10H,
Cp), 1.20–2.70 (m, 11H, Cy), −10.84 [d, J(HP) 42, 1H, l-H];
dH (400.13 MHz, CDCl3, 243 K) 5.26, 5.21 (2 × s, 2 × 5H,
Cp), 1.20–2.70 (m, 11H, Cy), −10.92 [d, J(HP) 42, 1H, l-H]; dF
(282.36 MHz, CH2Cl2) −126 [d, JFP = 880, PF]; dC (100.62 MHz,
CDCl3, 223 K) 243.1 [d, J(CP) 21, CO], 239.0 [dd, J(CP) 23,
J(CF) 13, CO], 234.0, 231.8 (2 × s, 2 × CO), 91.3, 91.0 (2 × s,
2 × Cp), 54.9 [dd, J(CP), J(CF) = 16.5, 17.5, C1(Cy)], 31.1, 30.0
[2 × s, C2, C6(Cy)], 27.5 [d, 12.3, C3 or C5(Cy)], 27.4 [d, J(CP)
11.5, C5 or C3(Cy)], 26.3 [s, C4(Cy)].
Preparation of [Mo2ClCp2(l-PHPh)(CO)4] (8). Hydrogen
chloride diluted in nitrogen was gently bubbled through a solu-
tion of compound 1d (0.070 g, 0.129 mmol) in dichloromethane
(15 ml) for 15 min to give a dark red solution. The solvent was
then removed under vacuum and the residue crystallized by slow
diffusion of petroleum ether into a concentrated toluene solution
of the product. Red crystals of compound 8 (0.054 g, 72%) were
thus obtained. The crystals used in the X-ray diffraction study
of the complex were grown in this way (Found: C, 41.11; H, 2.64.
C20H16ClMo2O4P requires C, 41.50; H, 2.77%); dH (200.13 MHz,
290 K) 7.5–6.5 (m, 5H, Ph), 6.75 [d, J(HP) 368, 1H, PH], 5.23
(s, 10H, Cp); dH (400.13 MHz, 233 K) 7.66 (m, 2H, Ph), 7.42
(m, 3H, Ph), 6.81 [d, J(HP) 363, 1H, PH], 5.27 (s, 10H, Cp); dH
(400.13 MHz, 203 K), Isomer A: 7.60–7.40 (m, 5H, Ph), 6.78
[d, J(HP) 364, 1H, PH], 5.30, 5.26 (2 × s, 2 × 5H, Cp); isomer
B: 6.11 [d, J(HP) 388, 1H, PH], 5.19, 5.18 (2 × s, 2 × 5H, Cp);
Ratio A/B = 5 at 203 K; dC (100.62 MHz, 213 K), Isomer A:
239.4, 237.7 (2 × s, 2 × CO), 237.2 [d, J(CP) 20, CO], 227.4 [d,
J(CP) 14, CO], 138.1 [d, J(CP) 43, C1(Ph)], 130.8 [d, J(CP) 8,
C2(Ph)], 128.6 [s, C4(Ph)], 128.0 [d, J(CP) 10, C3(Ph)], 94.1, 91.5
(2 × s, 2 × Cp). Isomer B: 91.0, 89.9 (2 × s, 2 × Cp). Other
resonances from this minor isomer could not be identified in the
spectrum.
Preparation of [Mo2Cp2(l-H)(l-PFPh)(CO)4] (6d). A solu-
tion of compound 7dꢀ (0.090 g, 0.137 mmol) in dichloromethane
(15 ml) was refluxed for 80 min to give an orange solution.
Solvent was then removed and the residue chromatographed on
alumina (activity IV). Elution with dichloromethane–petroleum
ether (1 : 2) gave an orange fraction. Removal of solvents from
the latter fraction yielded compound 6d (0.027 g, 35%) as an
orange solid (Found: C, 42.44; H, 2.98. C20H16FPMo2O4 requires
C, 42.72; H, 2.87%). Spectroscopic data for this product were
in agreement with those reported for compound 6d in reference
17. dF (CH2Cl2) −122.0 [d, JFP = 911, PF].
X-Ray structure determination for compounds 1b and 8
Preparation of [Mo2Cp2(l-PHCy)(l-CO)(CO)4](BArꢀ4) (7c).
Carbon monoxide was gently bubbled through a solution of
compound 1c (0.050 g, 0.091 mmol) in dichloromethane (10 ml).
Solid [H(OEt2)2](BArꢀ4) (0.100 g, 0.099 mmol) was then added
and the mixture was stirred for 10 min to give a red solution.
The solvent was then removed under vacuum and the residue
was washed with toluene (2 × 5 ml) and petroleum ether (2 ×
5 ml) to give compound 7c (0.116 g, 89%) as a red powder
(Found: C, 44.01; H, 2.46. C53H34BF24Mo2O5P requires C, 44.19;
H, 2.38%); dH (400.13 MHz) 7.72 (s, 8H, Arꢀ), 7.56 (s, 4H, Arꢀ),
5.41 [d, J(HP) 1, 10H, Cp], 4.75 [dd, J(HP) 354, J(HH) 11, 1H,
PH], 2.40–1.20 (m, 11H, Cy); dC (100.62 MHz, 203 K) 270.0 [d,
J(CP) 22, l-CO], 228.4 [d, J(CP) 17, 2 × CO], 220.3 (s, 2 × CO),
161.1 [q, J(CB) 50, C1(Arꢀ)], 134.2 [s, C2(Arꢀ)], 128.3 [q, J(CF)
32, C3(Arꢀ)], 123.9 [q, J(CF) 272, CF3], 116.9 [s, C4(Arꢀ)], 92.8 (s,
Cp), 39.1 [d, J(CP) 25, C1(Cy)], 35.5 [s, C2(Cy)], 27.9 [d, J(CP)
12, C3(Cy)], 25.2 [s, C4(Cy)].
Data were collected on a Bruker Smart-CCD-1000 diffractome-
ter using graphite-monochromated Cu-Ka radiation at room
temperature (1b) or 100 K (8). The program SMART40 was
used for collecting frames of data, indexing reflections, and
determining lattice parameters. The frames were integrated by
the program SAINT,40 and absorption correction was applied
with SADABS.41 The structure was solved by Patterson inter-
pretation and phase expansion using DIRDIF,42 and refined
with full-matrix least squares on F2 using SHELXL97.43 All
non-hydrogen atoms were refined anisotropically. All hydrogen
atoms were located in the Fourier map and refined isotropically.
Crystallographic data and structure refinement details for both
1b and 8 are collected in Table 6.
X-Ray structure determination for compound 6b
Diffraction data were gathered at room temperature using a
four-circle serial diffractometer.44 Absorption corrections were
based on an ellipsoidal model derived from three-dimensional
w-scans,45 in this case from 18 w-scans of reflections which had
equivalent Eulerian v-angles between −30.2 and +51.0◦. The
structure was solved by direct methods46 and refined to F2 by
full-matrix least-squares analysis using all diffraction data.43
Non-hydrogen atoms were refined anisotropically. The fully
occupied non-methyl hydrogen atom sites were observed in a
difference map, but with one exception these and the partially
occupied H-atom sites of the disordered Cp (see below) were
placed at calculated positions and refined as riding atoms
with isotropic displacement parameters set to 1.2 times the
equivalent isotropic displacement parameters of their respective
parent atoms. The methyl groups were treated as variable-
metric rotators, with initial torsion angles set by the observed
Preparation of [Mo2Cp2(l-PHCy)(l-CO)(CO)4](BF4) (7cꢀ).
The procedure is similar to that described for 7c, but using
HBF4·OEt2 instead. Starting from 1c (0.090 g, 0.164 mmol)
and HBF4·OEt2 (50 ll of a 54% Et2O solution, 0.363 mmol),
and using diethyl ether and petroleum ether during workup,
compound 7cꢀ (0.088 g, 81%) was obtained as a red powder. The
crystals used in the X-ray study were grown by slow diffusion of
petroleum ether into a dichloromethane solution of the complex
at −20 ◦C (Found: C, 36.01; H, 3.17. C21.5H23BClF4Mo2O5P
1
(7cꢀ· CH2Cl2) requires C, 36.55; H, 3.27); dH (200.13 MHz) 5.52
2
[dd, J(HP) 381, J(HH) 11, 1H, PH], 5.51 (s, 10H, Cp), 2.50–1.30
(m, 11H, Cy).
Preparation of [Mo2Cp2(l-PHPh)(l-CO)(CO)4](BArꢀ4) (7d).
The procedure is identical to that described for 7c. By starting
D a l t o n T r a n s . , 2 0 0 4 , 4 1 6 8 – 4 1 7 9
4 1 7 7

View Article Online
Table 6 Crystal data for compounds 1b, 6b, 7cꢀ and 8
Compound
1b
6b
7cꢀ·1/2CH2Cl2
8
Mol. formula
Mr
Crystal system
Space group
Crystal color
Crystal shape
Crystal size/mm
C23H23O4Mo2P
586.26
Monoclinic
C2/c
Orange
C23H22O4FMo2P
604.26
C21.5H23BClO5F4Mo2P
706.53
Monoclinic
P21/n
Red
Block
C20H16ClO4Mo2P
578.63
Triclinic
Triclinic
¯
¯
P1
P1
Red
Red
Prism
Block
Prism
0.30 × 0.22 × 0.09
Cu-Ka (1.54182)
33.8525(4)
8.5459(1)
16.3577(2)
90
110.995(1)
90
4418.11(9)
8
0.27 × 0.22 × 0.16
Mo-Ka (0.71073)
8.5683(9)
10.8886(13)
13.719(3)
94.019(15)
103.809(10)
112.294(8)
1131.5(3)
2
0.10 × 0.10 × 0.10
0.32 × 0.25 × 0.18
Cu-Ka (1.54182)
8.6350(1)
9.8969(1)
13.7036(1)
104.295(1)
102.891(1)
106.691(1)
1031.057(18)
2
˚
Radiation (k/A)
Mo-Ka (0.71073)
˚
a/A
20.6073(19)
9.5398(7)
27.706(3)
90
101.031(8)
90
5346.1(9)
8
˚
b/A
˚
c/A
a/◦
b/◦
c /◦
3
˚
V/A
Z
Dc/g cm−3
l/cm−1
1.763
10.21
1.774
12.14
1.756
11.6
1.864
12.09
Diffractometer
T/K
Smart-CCD
100
x-Scan
2.80–70.3
18766
CAD4
299
x-Scan
2–25
3972
3958
KAPPA CCD
293
x-Scan
2–27.5
Smart-CCD
100
x-Scan
3.5–70.5
6603
Scan type
h limits/◦
Total data
Unique total data
Unique data used
Ra
23625
4050
10840
3407
3894 [(Fo)2 > 2r(Fo)2]
0.024
3340 [(Fo)2 > 2r(Fo)2]
0.029
6107 [(Fo)2 > 3r(Fo)2]
3240 [(Fo)2 > 2r(Fo)2]
0.031
0.068
Rw
0.060c
1.066
0.064d
1.026
0.081e
0.083f
1.049
b
GOF
0.8811
Octants collected
No. of variables
Dr (mean, max.)
−41,41; −9,10; −19,19
0,10; −12,11; −16,15
−22,26; −12,12; −25,35
−10,9; −11,11; −16,16
363
335
576
318
0.003, 0.079
0.64, −0.75
0.006, 0.087
0.34, −0.28
0.001, 0.032
0.54, −0.99
−3
˚
Dq (max., min.)/e A
1.94, −0.97
ꢀ
ꢀ
ꢀ
ꢀ
a R = ꢁFo| − |Fcꢁ/ |Fo|. b Rw = [ w(|Fo| − |Fc| )2/ w|Fo| ]1/2
.
c w = 1/[r2(Fo2) + (0.0276P)2 + 7.5812P] where P = (Fo + 2Fc2)/3. d w =
2
2
2
2
ꢀ
ꢀ
1/[r2(Fo2) + (0.0285P)2 + 0.3818P] where P = (Fo + 2Fc2)/3. e Rw = [ w(|Fo| − |Fc|)2/ w|Fo| ]1/2; w = wꢀ [1 − (ꢁFo| − |Fcꢁ/6rFo)2]2, with wꢀ =
2
2
ꢀ
1/ rArTr(X) with coefficients 0.484, 0.463 and 0.205 for a Chebyshev series, for which X is Fc/Fc (max). f w = 1/[r2(Fo2) + (0.0385P)2 + 0.2486P]
where P = (Fo + 2Fc2)/3.
2
positions of the H atoms, and with the isotropic displacement
parameters of the H atoms set to 1.5 times the equivalent
isotropic displacement parameters of their respective parent
atoms. The bridging hydride was located in a difference map and
refined independently with an isotropic displacement parameter.
The hydride scattering factor from International Tables, Volume
C was used for this site.47 One of the Cp ligands was disordered
over two positions, related by rotation about the Mo–Cg vector,
in which Cg represents the centre of the Cp ring. The populations
of the two congeners refined stably to a 2 : 1 ratio, and were
fixed for the final refinement. Similarity restraints were used
for the two components, and rigid-bond conformity restraints48
were placed on the anisotropic displacement parameters of the
minor component. This treatment was intended to produce as
accurate as possible a representation of the average electron
density in the disordered region. The refinement converged with
the residuals shown in Table 6. A final difference map showed no
important features. An ORTEP view of the molecule showing
the more abundant position of the disordered Cp ring is shown in
Fig. 2.
absorption correction was applied by using DIFABS.50 All non-
hydrogen atoms were refined anisotropically. Because of the low
quality of the diffraction data, all hydrogen atoms were placed
in calculated positions and refined isotropically.
CCDC reference numbers 248103–248106.
See http://www.rsc.org/suppdata/dt/b4/b412875c/ for cry-
stallographic data in CIF or other electronic format.
Acknowledgements
We thank the Ministerio de Ciencia y Tecnolog´ıa of Spain
for financial support (Projects BQU2000-0944, BQU2002-00554
and BQU2003-05471) and for a grant to D. S. We also thank
the Ministerio de Educacio´n, Cultura y Deporte for a grant to
D. G.
References
1 M. E. Garc´ıa, V. Riera, M. A. Ruiz, D. Sa´ez, J. Vaissermann and J. C.
Jeffery, J. Am. Chem. Soc., 2002, 124, 14304.
2 A. M. Arif, A. H. Cowley, N. C. Norman, A. G. Orpen and M.
Pakulski, Organometallics, 1988, 7, 309.
X-Ray structure determination for compound 7cꢀ
Data were recorded at room temperature on a Bruker Kappa-
CCD diffractometer with graphite monochromated Mo-Ka
radiation. The orientation matrix and lattice parameters were
obtained by least-squares refinement of the diffraction data of
70 reflections within the range of 6 < h < 15◦. The index and h
ranges of data collection are given in Table 6. All the measured
independent reflections were used in the analysis. The structure
was solved by direct methods and refined with full-matrix least-
squares technique on F using the CRYSTALS49 program. An
3 G. Hu¨ttner and K. Evertz, Acc. Chem. Res., 1986, 19, 406.
4 G. K. Kubas, Metal Dihydrogen and r-Bond Complexes, Kluwer
Academic/Plenum, New York, 2001, ch. 3.
5 M. E. Garc´ıa, V. Riera, M. A. Ruiz, M. T. Rueda and D. Sa´ez,
Organometallics, 2002, 21, 5515.
6 (a) E. A. V. Ebsworth, A. P. McIntosh and M. Schro¨der,
J. Organomet. Chem., 1986, 312, C41; (b) A. J. Arce, R. Machado, Y.
De Sanctis, T. Gonza´lez, R. Atencio and A. J. Deeming, Inorg. Chim.
Acta, 2003, 344, 123.
7 D. Mani and H. Vahrenkamp, Chem. Ber., 1986, 119, 3639.
4 1 7 8
D a l t o n T r a n s . , 2 0 0 4 , 4 1 6 8 – 4 1 7 9

View Article Online
8 (a) S. B. Colbran, B. F. G. Johnson, J. Lewis and R. M. Sorrell,
J. Organomet. Chem., 1985, 296, C1; (b) A. J. Deeming, S. Doherty,
M. W. Day, K. I. Hardcastle and H. Minassian, J. Chem. Soc., Dalton
Trans., 1991, 1273.
9 H. J. Haupt, M. Schwefer, H. Egold and U. Flo¨rke, Inorg. Chem.,
1995, 34, 5461.
10 M. E. Garc´ıa, V. Riera, M. A. Ruiz, D. Sa´ez, H. Hamidov, J. C. Jeffery
and T. Riis-Johannessen, J. Am. Chem. Soc., 2003, 125, 13044.
11 (a) For recent reviews on phosphinidene chemistry, see for example:
K. Lammertsma and M. J. M. Vlaar, Eur. J. Org. Chem., 2002, 1127;
(b) F. Mathey, N. H. Tran Huy and A. Marinetti, Helv. Chim. Acta,
2001, 84, 2938; (c) R. R. Schrock, Acc. Chem. Res., 1997, 30, 9;
(d) A. H. Cowley, Acc. Chem. Res., 1997, 30, 445.
12 (a) M. A. Alvarez, G. Garc´ıa, M. E. Garc´ıa, V. Riera, M. A. Ruiz,
M. Lanfranchi and A. Tiripicchio, Organometallics, 1999, 18, 4509;
(b) M. A. Alvarez, Y. Anaya, M. E. Garc´ıa, V. Riera, M. A. Ruiz and
J. Vaissermann, Organometallics, 2003, 22, 456; (c) M. A. Alvarez,
Y. Anaya, M. E. Garc´ıa, V. Riera and M. A. Ruiz, Organometallics,
2004, 23, 433.
13 (a) H. Nishida, N. Takada, M. Yoshimura, T. Sonoda and H.
Kobayashi, Bull. Chem. Soc. Jpn., 1984, 57, 2600; (b) M. Brookhart,
B. Grant and A. F. Volpe Jr., Organometallics, 1992, 11, 3920.
14 (a) W. Beck and K. Su¨nkel, Chem. Rev., 1988, 88, 1405; (b) S. H.
Strauss, Chem. Rev., 1993, 93, 927; (c) K. Seppelt, Angew. Chem.,
Int. Ed. Engl., 1993, 32, 1025.
15 (a) For some recent work, see for example: M. Jime´nez-Tenorio,
M. C. Puerta, I. Salcedo, P. Valerga, S. I. Costa, P. T. Gomes and
K. Mereiter, Chem. Commun., 2003, 1168; (b) M. D. Leatherman,
S. A. Svejda, L. K. Johnson and M. Brookhart, J. Am. Chem. Soc.,
2003, 125, 3068; (c) C. M. Norris, S. Reinartz, P. S. White and
J. L. Templeton, Organometallics, 2002, 21, 5649; (d) F. L. Taw, H.
Mellows, P. S. White, F. J. Hollander, R. G. Bergman, M. Brookhart
and D. M. Heinekey, J. Am. Chem. Soc., 2002, 124, 5100; (e) M. H.
Voges and R. M. Bullock, J. Chem. Soc., Dalton Trans., 2002, 759.
16 K. Henrick, M. McPartlin, A. D. Horton and M. J. Mays, J. Chem.
Soc., Dalton Trans., 1988, 1083.
26 B. Wrackmeyer, H. G. Alt and H. E. Maisel, J. Organomet. Chem.,
1990, 399, 125.
27 C. J. Jameson, in Multinuclear NMR, ed. J. Mason, Plenum Press,
New York, USA, 1987, ch. 4.
28 M. A. Alvarez, M. E. Garc´ıa, V. Riera, M. A. Ruiz, C. Bois and Y.
Jeannin, Angew. Chem., Int. Ed. Engl., 1993, 32, 1156.
29 (a) R. L. Dekock and H. B. Gray, Chemical Structure and Bonding,
Benjamin/Cummings, Menlo Park, CA, 1980, ch. 7; (b) N. N. Green-
wood and A. Earnshaw, Chemistry of the Elements, Butterworth-
Heinemann, Oxford, UK, 1997, ch. 3.
30 (a) G. R. Desiraju and T. Steiner, The Weak Hydrogen Bond, Oxford
University Press, Oxford, UK, 1999, ch. 3; (b) T. Steiner, Angew.
Chem., Int. Ed., 2002, 41, 48.
31 P. S. Braterman, Metal Carbonyl Spectra, Academic Press, London,
UK, 1975.
32 J. W. Faller and A. S. Anderson, J. Am. Chem. Soc., 1970, 92, 5852.
33 G. A. Acum, M. J. Mays, P. R. Raithby and G. A. Solan, J. Organomet.
Chem., 1995, 492, 65.
34 J. A. Beck, S. A. R. Knox, G. H. Riding, G. E. Taylor and M. J.
Winter, J. Organomet. Chem., 1980, 202, C49.
35 E. P. Kyba, R. E. Davis, C. N. Club, S. T. Liu, H. O. A. Palacios and
J. S. McKennir, Organometallics, 1986, 5, 869.
36 D. D. Perrin and W. L. F. Armarego, Purification of Laboratory
Chemicals, Pergamon Press, Oxford, UK, 1988.
37 M. D. Curtis and M. S. Hay, Inorg. Synth., 1990, 28, 152.
38 (a) A. H. Cowley, J. E. Kilduff, T. H. Newman and M. Pakulski,
J. Am. Chem. Soc., 1982, 104, 5820; (b) Z. Hou, T. L. Breen and
D. W. Stephan, Organometallics, 1993, 12, 3158.
39 R. A. Barlett, M. M. Olmstead, P. P. Power and G. A. Sigel, Inorg.
Chem., 1987, 26, 1941.
40 SMART & SAINT Software Reference Manuals, Version 5.051 (Win-
dows NT Version), Bruker Analytical X-ray Instruments, Madison
WI, 1998.
41 G. M. Sheldrick, SADABS, Program for Empirical Absorption
Correction, University of Go¨ttingen, Go¨ttingen, Germany, 1996.
42 P. T. Beurkens, G. Admiraal, G. Beurkens, W. P. Bosman, S. Garc´ıa-
Granda, R. O. Gould, J. M. M. Smits and C. Smykalla, The
DIRDIF Program System, Technical Report of the Crystallographic
Laboratory, University of Nijmegen, Nijmegen, The Netherlands,
1999.
17 S. Woodward and M. D. Curtis, J. Organomet. Chem., 1992, 439, 319.
18 (a) J. L. Petersen, L. F. Dahl and J. M. Williams, J. Am. Chem. Soc.,
1974, 96, 6610; (b) R. A. Jones, S. T. Schwab, A. L. Stuart, B. R.
Whittlesey and T. C. Wright, Polyhedron, 1985, 4, 1689; (c) A. J.
Bridgeman, M. J. Mays and A. D. Woods, Organometallics, 2001, 20,
2076.
43 G. M. Sheldrick, SHELXL97: Program for the Refinement of Crystal
Structures, University of Go¨ttingen, Go¨ttingen, Germany, 1997.
ꢂ
C
44 CAD-4 Diffractometer Control Program, CAD4/PC Version 2.0,
19 J. Heck, J. Organomet. Chem., 1986, 311, C5.
1996 Nonius by, Delft, The Netherlands.
20 K. W. Muir, S. E. Guirdwood, F. Y. Pe`tillon, R. Pichon, S. Poder-
Guillou, P. Schollhammer and J. Talarmin, J. Organomet. Chem.,
1995, 486, 183.
45 Data were processed on an AlphaStation 200 4/166 (Open-
VMS/Alpha V6.2), with the program XCAD4B (K. Harms, 1996)
ꢂ
C
and with the commercial package SHELXTL Rel. 5.05/VMS,
21 H. Hartung, B. Walther, U. Baumeister, H. C. Bo¨ttcher, A. Krug, F.
Rosche and P. G. Jones, Polyhedron, 1992, 11, 1563.
22 J. L. Petersen and R. P. Stewart Jr., Inorg. Chem., 1980, 19, 186.
23 M. J. Bennet, W. L. Hutcheon and B. M. Foxman, Acta Crystallogr.,
Sect. A, 1975, 31, 488.
24 (a) J. G. Verkade and J. A. Mosbo, in Phosphorus-31 NMR
Spectroscopy in Stereochemical Analysis, ed. J. G. Verkade and
L. D. Quin, VCH, Florida, USA, 1987, ch. 13; (b) C. J. Jameson,
in Multinuclear NMR, ed. J. Mason, Plenum Press, New York, USA,
1987, ch. 16.
25 L. J. Todd, J. R. Wilkinson, J. P. Hickley, D. L. Beach and K. W.
Barnett, J. Organomet. Chem., 1978, 154, 151.
1996, Siemens Analytical X-ray Instruments, Inc., Madison, WI.
46 G. M. Sheldrick, SHELXS97: Program for the Solution of Crystal
Structures, University of Go¨ttingen, Go¨ttingen, Germany, 1997.
47 International Tables for Crystallography, Volume C: Mathematical,
Physical and Chemical Tables, ed. A. J. C. Wilson, Kluwer Academic
Publishers, Dordrecht, 1995, Tables 6.1.1.4 and 4.2.4.2.
48 F. L. Hirshfeld, Acta Crystallogr., Sect. A, 1976, 32, 239.
49 D. J. Watkin, C. K. Prout, J. R. Carruthers, P. W. Betteridge
and R. I. Cooper, CRYSTALS, Issue 11, Chemical Crystallography
Laboratory, University of Oxford, Oxford, UK, 2001.
50 D. J. Walter and D. Stuart, Acta Crystallogr., Sect. A, 1983, 39, 158.
D a l t o n T r a n s . , 2 0 0 4 , 4 1 6 8 – 4 1 7 9
4 1 7 9