Some reactions of Ni-Mo and Ni-W propargylic cations with nucleophiles

Article abstract of DOI:10.1016/j.jorganchem.2004.03.011

Preliminary reactions of the metal stabilized carbocationic species [(η-C₅H₅)Ni(μ-η²(Ni), η³(Mo)-HC₂CMe₂)Mo (CO)₂(η-C₅H₄Me)]⁺ BF₄

Full text of DOI:10.1016/j.jorganchem.2004.03.011

Journal of Organometallic Chemistry 689 (2004) 1882–1889
www.elsevier.com/locate/jorganchem
Some reactions of Ni–Mo and Ni–W propargylic cations with
nucleophiles
*
Michael J. Chetcuti , Steven R. McDonald
Department of Chemistry, ECPM, universitꢀe Louis Pasteur, 25 rue Bequerel, 67087 Strasbourg Cedex 2, France
Received 29 October 2003; accepted 11 March 2004
Abstract
Preliminary reactions of the metal stabilized carbocationic species [(g-C5H5)Ni(l-g2(Ni),g3(Mo)-HC2CMe2)Mo(CO)2(g-
C5H4Me)]þ BFꢀ4 (Ni–Mo) with nucleophiles are reported. The Ni–Mo cationic propargylic complex undergoes nucleophilic attack
by sodium methoxide to regenerate the neutral l-alkyne complex [(g-C5H5)Ni{l-g2,g2-HC2CMe2(OMe)}Mo(CO)2(g-C5H4Me)]
(Ni–Mo), from which the stabilized carbocation was originally derived by protonation. The new complexes [(g-C5H5)Ni{l-g2,g2-
HC2CMe2(C5H5)}Mo(CO)2(g-C5H4Me)] (Ni–Mo), which exist as an inseparable mixture of 1(c)-1,3- and 2(c)-1,3-cyclopentadienyl
isomers, were also obtained. When the Ni–Mo cations were treated with potassium t-butoxide, the alkyne isomers with pendant 1(c)-
1,3- and 2(c)-1,3-cyclopentadienyl groups are also formed. The l-hydroxyalkyne complex [(g-C5H5)Ni{l-g2,g2-HC2CMe2(OH)}-
Mo(CO)(g-C5H4Me)] (Ni–Mo) was also isolated concurrently, and presumably arises from nucleophilic attack of fortuitously
present hydroxide ions in the BuOꢀ reagent on the Ni–Mo cation. When NaBH4 was added to the Ni–Mo propargylic, nucleophilic
attack by hydride resulted and the l-iPrC2H heterobimetallic complex [(g-C5H5)Ni{l-g2,g2-HC2Pri}Mo(CO)2(g-C5H4Me)] (Ni–
Mo) was recovered in good yield. Small quantities of other side-products were isolated and characterized spectroscopically. Some
tantalizing differences in reactivity were observed when the corresponding Ni–W stabilized carbocation was reacted with methoxide
ions. When the not fully characterized solid formed by protonating [(g-C5H5)Ni(l-g2,g2-{HC2CMe2)(OMe)}W(CO)2(g-C5H4Me)]
(Ni–W) was treated with methoxide ions, regioisomers (1(c)-1,3- and 2(c)-1,3-cyclopentadienyl species) of composition
[(g-C5H5)Ni{l-g2,g2-HC2CMe2(C5H5)}W(CO)2(g-C5H4Me)] (Ni–W) were formed. Direct reaction of the pure cation [(g-
C5H5Nil-g2,g3-HC2CMe2)W(CO)2(g-C5H4Me)]þ (Ni–W) with methoxide also generated the same 1(c)-1,3- and 2(c)-1,3-cyclo-
pentadiene-substituted alkyne complexes. Unlike the case with the Ni–Mo complexes, the initial l-HC2CMe2(OMe) species was not
regenerated.
Ó 2004 Elsevier B.V. All rights reserved.
Keywords: Nickel; Molybdenum; Carbocation; Alkyne; Heterobimetallic; Nucleophilic attack; Metal-metal bonds
1. Introduction
bilized over Mo–Mo, and W–W centers. Heterobime-
tallic examples are also known [6–9], but prior to our
recent report [10], all such bimetallic l-HC2CR02þ species
had been anchored onto Co–Mo or Co–W centers. To
our knowledge no other heterobimetallic combinations
have yet been reported.
There has been increasing interest in the synthesis and
reactivity of metal-stabilized carbocations in which a C–
CRþ2 is formally bound to, and stabilized on, a dimetal
center and the field has been reviewed [1]. Many exam-
ples of such species ligated onto Co–Co bonds have been
recognized and they are synthetically useful [2–5].
Carbocations (propargylic cations) have also been sta-
Our recent results described the synthesis and
characterization of the di-metal stabilized carboca-
tions [CpNi(l-g2(Nig3(M)-HC2CMe2)M(CO)2Cp0]þ BFꢀ4
1
(Ni–M, 1a, M ¼ Mo; 1b, M ¼ W). The bridging
* Corresponding author. Tel.: +33-3-9024-2631; fax: +33-3-9024-
2787.
E-mail address: chetcuti@chimie.u-strasbg.fr (M.J. Chetcuti).
1
Cp ¼ g5-C5H5; Cp0 ¼ g5-C5H4Me.
0022-328X/$ - see front matter Ó 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2004.03.011

M.J. Chetcuti, S.R. McDonald / Journal of Organometallic Chemistry 689 (2004) 1882–1889
1883
organic ligand can also be considered to be a cationic
l-propargylic (allenyl) species. In these complexes, the
carbocation is spanning a Ni–Mo or a Ni–W bond
[10]. Complex 1a was isolated by protonation of
the l-alkyne complex 2a [CpNi{l-g2,g2-HC2CMe2-
(OMe)} Mo(CO)2Cp0] (Ni–Mo) with HBF4 ꢁ Et2O,
followed by MeOH loss, as shown in Eq. (1)
Complexes 3a and 4a are more intriguing. These two
species could not be separated by chromatography, and
were isolated as a together as orange-brown oil. They
are believed to be the 1(c)-1,3- and 2(c)-1,3-cyclopent-
adiene isomers of formula [CpNi{l-g2,g2-HC2-
CMe2(C5H5)} Mo(CO)2Cp0] (Ni–Mo) (see Chart 1).
The MS for the mixture showed a parent ion that cor-
½CpNifl-g2; g2-HC2CMe2ðOMeÞgMoðCOÞ Cp0ꢂðNi–Mo; 2aÞ þ HBF4
2
) ½CpNiðl-g2ðNiÞg3ðMoÞ-HC2CMe2ÞMoðCOÞ Cp0ꢂþBFꢀ4 ðNi–Mo; 1aÞ þ MeOH
ð1Þ
2
Curiously, the corresponding Ni–W l-propargylic
complex could not be obtained by direct protonation of
the Ni–W alkyne complex [CpNi{l-g2,g2-HC2
CMe2(OMe)}W(CO)2Cp0] (Ni–W), 2b.
responded to a m=e ratio of (1a + 65) mass/charge units,
i.e., the formal addition of a C5H5 unit to the cation
[CpNi{l-g2,g3-HC2CMe2}Mo(CO)2Cp0]þ
The mixture gave solution IR spectra that were consis-
(Ni–Mo).
However it, and the Ni–Mo cation 1a, could be iso-
lated by treating the metallacycles [CpNi{l-
(Ni–M,
M ¼ Mo, W) with tetrafluoroboric acid (Eq. (2))
tent with l-alkyne complexes (experimental section) and
the H NMR spectra (Table 1) of the mixture could be
rationalized with their proposed structures. Resonances
for two g-Cp and two g-Cp0 groups as well as over-
1
g3(Ni),g1(M)-C(O)-C(R)-C(H)}M(CO)2Cp0]
½CpNifl-g3ðNiÞ; g1ðMÞ-CðOÞ-CðRÞ-CðHÞgMðCOÞ Cp0ꢂðNi–M; M ¼ Mo; WÞ þ HBF4
2
) ½CpNiððl-g2ðNiÞg3ðMÞ-HC2CMe2ÞMðCOÞ Cp0ꢂþBFꢀ4 ðNi–M; M ¼ Mo; 1a; M ¼ W; 1bÞ þ MeOH þ CO
2
ð2Þ
Our previous study reported the synthesis and char-
acterization of cations 1a and 1b, in which the C–CMe2þ
group is believed to be coordinated to, and stabilized by,
the group 6 metal atom. We reported none of their
chemistry. Here we report some of the preliminary
chemistry of these carbocationic species towards nucle-
ophiles, specifically with sodium methoxide, with po-
tassium butoxide and with sodium borohydride. These
reagents are sources, respectively, of MeOꢀ, BuOꢀ and
Hꢀ. Minor impurities and co-generated reactants af-
forded additional, unexpected products.
lapping multiplets that may be assigned to two sets of
C5H5 vinylic protons were observed. Isomers 3a and 4a
are always found in a 3:2 ratio in solution at ambient
temperature (even when formed in other reactions dis-
cussed later), and are likely in equilibrium, but this has
not been definitely established. In addition, there is no
NMR evidence of the symmetric 5(c)-1,3-cyclopentadi-
ene isomer being formed.
While cyclopentadiene groups attached to metals or
metalloids are well known to isomerize and interconvert
via metallotropic shifts [11], the equivalent C–C5H5
shifts seem less likely here, as a stronger C5H5–C bond is
present. A similar mixture of cyclopentadienyl isomers is
found in a series of indole substituted cyclopentadienyl
compounds [12]. The isomers are found in a 1:1 ratio in
this work, and as is seen in our work, there is no evi-
dence of the symmetric 2,4-cyclopentadiene species.
It is not obvious how 3a and 4a form. Clearly the
source of the cyclopentadienyl ligand is the nickel atom.
Only cyclopentadienyl addition is seen and no addition
of methylcyclopentadienyl groups (bound to the group 6
metal) is ever observed. It is possible that the decom-
position of 1a which is observed concurrently during the
course of the reaction, liberates a C5Hꢀ5 ligand which
then nucleophillically attacks complex 1a. The apparent
lability of a Cp ligand bound to nickel is not unprece-
dented. We have reported several instances of both Cp
and Cp* ligand transfer off a nickel atom [13–16]. Al-
ternatively, the cationic complex 1a might directly at-
tack a CpNi moiety, present in either another molecule
2. Results and discussion
2.1. Reaction of methoxide ions with 1a
Complex 1a, formed in situ by the protonation of
the methoxyalkyne species 2a, reacted slowly with
methoxide ions to generate a mixture of complexes.
The reaction is unfortunately a race against time as 1a
spontaneously decomposes in solution at ambient
temperatures. Three products (2a, 3a and 4a, Chart 1)
were isolated from this reaction but the yields of all
three were low owing to significant decomposition. The
l-alkyne complex 2a was readily characterized by
comparison of its spectroscopic data with those of an
authentic sample [10]. The addition of methoxide to 1a
to give the l-alkyne complex 2a is thus, at least partly,
reversible. The reaction is depicted in Scheme 1.

1884
M.J. Chetcuti, S.R. McDonald / Journal of Organometallic Chemistry 689 (2004) 1882–1889
2.2. Reaction of potassium butoxide with 1a
Me
Me
OMe
Me
Me
t
The reaction of 1a with potassium butoxide also
+
afforded a mixture of the two cyclopentadiene substi-
tuted alkyne complexes 3a and 4a (in the same equilib-
rium 3:2 ratio that was observed in the previously
Me
M
H
H
M
Ni
C
Ni
C
O
O
C
Me
C
O
t
described reaction). Addition of BuOꢀ to the cationic
- O
BF4
complex did not take place. However, small quantities
of another species, subsequently identified as the hy-
droxyalkyne complex [CpNi(l-g2,g2-HC2CMe2OH)
Mo(CO)2Cp0] (Ni–Mo, 5a), were also obtained, as
shown in Scheme 2.
1a, M = Mo; 1b, M = W
2a, M = Mo; 2b, M = W
The more sterically hindered OBuꢀ anion is appar-
ently too bulky to add on to complex 1a. The isolation
of 5a was surprising, but may be attributed to contam-
ination of the KOtBu with KOH. Nucleophilic attack of
OHꢀ ions on 1a would generate 5a. It is noteworthy that
Me
Me
M
Me
Me
H
H
H
H
Mo
Ni
C
Ni
C
O
O
Me
Me
C
C
t
neither BuOꢀ nor MeOꢀ lead to isolable elimination
O
O
products. Instead, products that result from direct nu-
cleophilic addition are produced in each case.
4a, M = Mo; 4b, M = W
3a, M = Mo; 3b, M = W
OH
H
2.3. Reaction of sodium borohydride with 1a
Me
Me
Me
Me
Nucleophilic attack of 1a by hydride ions could in
principle lead to attack at the carbocationic center to
afford the bimetallic l-isopropyl acetylene complex.
Other possible products are complexes which contain
1,1-dimethylallene or its derivatives bound to the di-
metal center. Such species have been observed in the
reactions of Ni–Mo complexes with 1,1-dimethylallene
[14,20].
Mo
H
Mo
Ni
Ni
C
C
O
O
Me
Me
C
O
C
O
5a
6a
OEt
When the reaction of 1a with sodium borohydride
was attempted in ethanol, the major product obtained
was the heterobimetallic l-PrC2H species [CpNi(l-
g2,g2-PriC2H)Mo(CO)2Cp0] (Ni–Mo, 6a). This species
Me
Me
Mo
1
Ni
C
was fully characterized via MS, HRMS, H NMR and
IR spectroscopy. Two other complexes were also iso-
lated in minor yields.
O
Me
C
O
7a
One of these species was the previously described l-
hydroxyalkyne complex 5a. The other complex was
spectroscopically characterized as the ethoxyalkyne
species [CpNi{l-g2,g2-HC2CMe2(OEt)}Mo(CO)2Cp0]
(Ni–Mo, 7a). All these products result from the addition
of various nucleophiles (OHꢀ, Hꢀ or EtOꢀ, respectively,
for 5a, 6a and 7a) to 1a. Hydroxide and ethoxide ions
likely form via attack of the sodium borohydride on the
(wet) ethanol solvent used in this reaction. This reaction
is summarized in Scheme 3.
Chart 1.
of 1a or in one of its decomposition products. Addition
of electrophiles such as Hþ or Ph3Cþ to a Cp group in
nickelocene has long been known and would not be
novel [17–19].
Following the suggestions of reviewers, we have
attempted to prepare complexes 3a and 4a by letting a
sample of 1a decompose on its own without the presence
of added nucleophiles. This did not lead to detectable
quantities of complexes 3a and 4a. However, significant
yields of 2 were formed when suspensions of the car-
bocation 1a were reacted directly with thf solutions of
NaþC5Hꢀ5 indicating that 3a and 4a can be obtained
directly by nucleophilic attack of cyclopentadienide
anions.
2.4. Reactions with Ni–W complexes
The reaction of the alkyne complex 2b with
HBF4 ꢁ Et2O has never been shown to produce the cat-
ionic complex [CpNi(l-g2,g3-HC2CMe2)W(CO)2Cp0]
þBFꢀ4 (Ni–W, 1b). Nevertheless, the insoluble tan solid
that formed in this reaction [10] did react slowly with

M.J. Chetcuti, S.R. McDonald / Journal of Organometallic Chemistry 689 (2004) 1882–1889
1885
Me
OMe
Me
Me
Me
–
MeO
+
Mo
Me
H
H
Mo
HBF .Et O
Ni
4
2
C
O
Ni
C
2a
O
C
Me
( – MeOH)
C
O
O
-
BF
4
1a
Me
Me
Me
Me
...
Mo
+
+
+
H
Mo
H
Ni
Ni
C
C
O
3a
O
4a
Me
Me
C
O
C
O
Scheme 1. The reaction of complex 1a with methoxide ions (only 2a regenerates 1a when treated with methoxide).
Table 1
1H NMR data for the complexes [CpNi{l-g2,g2-HC2CMe2(R)}M(CO)2Cp0] (Ni–M, M ¼ Mo, W)a
Complex
C5H5 C5H4Me
CH
Me
Me
C5H4Me
R
3a (R ¼ C5H5) 5.10
3b (R ¼ C5H5) 5.14
4a (R ¼ C5H5) 5.11
4b (R ¼ C5H5) 5.14
5.20–4.98 6.08
5.30–4.98 5.85b
5.20–4.98 5.96
5.30–4.98 5.75b
5.33, 5.30, 5.93
5.28, 5.27
1.48
1.48
1.48
1.48
1.45
1.47
1.46
1.46
1.46
1.30
1.95
2.08
1.97
2.11
2.02
6.51(m), 6.32(m), 6.02(m), 2.90(m, CH2)
6.50(m), 6.33(m), 6.02(m), 2.90(m, CH2)
6.36(m), 6.22(m), 6.19(m), 2.98(m, CHH), 2.68(m, CHH)c
6.38(m), 6.24(m), 6.18(m), 3.00(m, CHH), 2.72(m, CHHc)
1.85
5a (R ¼ OH)
5.20
6a (R ¼ H)
5.15
5.14
5.24–5.18 5.91(d)d 1.24(d)e 1.07(d)e 1.99
2.75(m)
3.77 (OCHH), d of q, JHH ¼ 8:6, 7.0; 3.44 (OCHH), d of q,
JHH ¼ 8:6, 7.0; 1.15 (OCH2Me), t, JHH ¼ 7:0
7a (R ¼ OEt)
5.37, 5.35, 5.85
5.33, 5.23
1.50
1.27
2.04
a Spectra recorded on a GE GN-300 spectrometer in chloroform-d; @, ppm; J in Hz. All C5H4Me signals are (occasionally overlapping) ABCD
type multiplets.
b JWH ¼ 2:1.
c JHHðgemÞ ¼ 23.
d JHH ¼ 0:8.
e JHH ¼ 6:7.
Me
Me
Me
Me
Me
Me
-
–
OBu (+ OH )
+
Mo
Me
H
+
Ni
H
Mo
H
Mo
C
1a
O
C
Ni
Ni
C
C
O
-
3a
4a
O
BF
O
Me
Me
4
C
C
O
O
OH
Me
Me
+
O
Ni
Mo
C
H
C
5a
O
Me
t
Scheme 2. The reaction of 1a with wet butoxide ions.

1886
M.J. Chetcuti, S.R. McDonald / Journal of Organometallic Chemistry 689 (2004) 1882–1889
H
Me
Me
H
Mo
Ni
C
O
Me
C
O
6a
+
Me
Me
OH
Me
Me
+
NaBH
4
Me
H
Mo
H
Mo
wet EtOH
Ni
C
Ni
C
O
Me
O
C
C
O
O
-
BF
4
5a
1a
+
OEt
Me
Me
H
Mo
Ni
C
O
Me
C
O
7a
Scheme 3. The reaction of 1a with ethanolic sodium borohydride.
Me
Me
Me
Me
Me
Me
+
Me
–
W
H
MeO
+
Ni
H
W
W
C
O
H
C
C
Ni
Ni
1b
3b
C
4b
O
O
Me
O
Me
C
C
-
BF
4
O
O
Scheme 4. The reaction of the Ni–W carbocationic complex 1b with methoxide.
sodium methoxide to give the Ni–W equivalents of
complexes 3a and 4a [CpNi{l-g2,g2-HC2CMe2
(C5H5)}W(CO)2Cp0] (Ni–W) 3b and 4b. This mixture of
isomers was isolated and characterized spectroscopi-
a thf solution of 1b prepared this way was treated with
NaOMe, the only recovered products were the cyclo-
pentadiene-substituted l-alkyne complexes 3b and 4b
(see Scheme 4). Unlike the corresponding reaction with
cation 1a, the Ni–W l-alkyne complex that would
correspond to 2a was not formed in this reaction.
There is thus an interesting difference in behavior be-
tween the Ni–Mo and the Ni–W propargylic cations.
1
cally. Their H NMR data are very similar to those of
the 3a/4a mixture and indeed the 3b:4b isomer ratio is
also 3:2 at ambient temperatures. However, no l-alkyne
species that would have resulted from methoxide addi-
tion was ever observed.
The metallacycle [CpNi{l-g3(Ni),g1(W )-C(O)-C(R)-
C(H)}W(CO)2Cp0] [Ni–W, R ¼ CMe2(OMe)] undergoes
elimination of methanol and decarbonylation over a 2
h period when treated with HBF4 ꢁ Et2O to afford the
Ni–W propargylic cation [CpNi{l-g2(Ni),g3(W )-
3. Conclusions
Cationic l-HC2CMeþ2 ions, stabilized on a heterobi-
metallic Ni–Mo framework, undergo nucleophilic attack
to generate l-HC2CMe2(Nu) (Nu ¼ nucleophile)
HC2CMe }-W(CO)2Cp0]þ BFꢀ4 (Ni–W, 1b) [10]. When
2

M.J. Chetcuti, S.R. McDonald / Journal of Organometallic Chemistry 689 (2004) 1882–1889
1887
complexes with a variety of (sometimes unexpected)
nucleophiles. The reaction is slow, and in most cases
competes unfavorably with the spontaneous decompo-
sition of the heterobimetallic cations in solution. When
the nucleophile is methoxide (MeOꢀ), the l-HC2CMe2
(OMe) complex [CpNi{l-g2,g2-HC2CMe2(OMe)}-
Mo(CO)2Cp0] (Ni–Mo) was isolated, but the Ni–W
propargyl ether complex was not obtained in corre-
sponding Ni–W reaction with sodium methoxide. An
unexpected side-product in reactions of both the Ni–Mo
and Ni–W cations was the formation of two probably
equilibrating isomers, in each case, of the complexes
[CpNi{l-g2,g2-HC2CMe2(C5H5)}M(CO)2Cp0] (Ni–M,
M ¼ Mo, W). Decomposition of the bimetallic cations
may generate Cpꢀ ions which then nucleophilically at-
tack the cation. Direct electrophilic attack by the cation
cations 1a and 1b, and that of the ligand
HC2CMe2(OMe) have been described [10]. The reagents
NaOMe, KOBut and NaBH4 were purchased from
Aldrich and used as received.
IR data were collected on an IBM IR-32 spectrome-
ter in hexanes (unless otherwise noted). 1H NMR
spectra were recorded on a GE GN-300 spectrometer in
chloroform-d1, and referenced with respect to the re-
sidual protons in this solvent. Values are in ppm with
coupling constants J in Hz. All C5H4Me signals are
(occasionally overlapping) ABCD type multiplets. Mass
spectra were obtained on a Finnegan Matt 8430 spec-
trometer using electron impact (EI) or chemical ioniza-
tion (CI) techniques. The appropriate isotopic envelope
patterns were observed for the Ni–Mo and Ni–W
complexes. HRMS were obtained for all complexes).
Microanalytical data for 5a were obtained by M-H-W
labs of Phoenix, AZ. These data were not recorded for
the other complexes since they were isolated as oils).
on
a CpNi group is also conceivable. The l-
HC2CMe2(C5H5) species were also obtained when the
Ni–Mo propargylic cation was treated with BuOꢀ.
Traces of KOH in the KOBut reagent led to the isola-
tion of small quantities of the l-HC2CMe2OH complex.
When the Ni–Mo propargylic cation was treated with
NaBH4, the principal product was the l-PriC2H com-
plex. Smaller quantities of both l-HC2CMe2OEt and l-
HC2CMe2OH Ni–Mo species, presumably formed by
concurrent attack of OHꢀ and EtOꢀ ions generated in
the reaction solvent, were also isolated.
4.2. (b) Reaction of 1a, [CpNi{l-g2(Ni),g3(Mo)-
HC2CMe2}Mo(CO)2Cp0]þBFꢀ4 (Ni–Mo) with NaOMe
to afford the 3a/4a isomeric mixture [CpNi{l-g2,g2-
HC2CMe2(C5H5)}Mo(CO)2Cp0] (Ni–Mo)
The cationic complex 1a was generated and reacted
immediately.
A
solution of 2a, [CpNi{l-g2,g2-
Thus, we can conclude that the nickel-molybdenum
stabilized carbocation 1a undergoes nucleophilic addi-
tion at the carbocationic carbon atom with small or
medium sized nucleophiles such as Hꢀ, OHꢀ, MeOꢀ
and even (effectively) C5Hꢀ5 . The larger BuOꢀ anion
does not add to the complex. The Ni–W cations
are more reluctant to undergo nucleophilic addition
reactions.
The stability of these cationic l-HC2CMe2 Ni–Mo
and Ni–W complexes is greater than that of related Co–
Co propargylic cations, but less than that of Co–Mo and
Co–W allenyl cations of this type. As cation decompo-
sition is a problem, further studies are underway that are
aimed at stabilizing these heterobimetallic Ni–M and
Ni–W propargylic cations by using g-C5Me5 and not g-
C5H5 ligands ligated to the nickel. The reaction of such
species with other nucleophiles will also be explored.
HC2CMe2(OMe)}Mo(CO)2Cp0] (Ni–Mo) (100 mg, 0.22
mmol) in Et2O (30 mL) was treated with a 1% solution
of HBF4 ꢁ Et2O. After precipitation of 1a was complete
(ca. 8 mL HBF4 ꢁ Et2O was used), the solid was rinsed
with 3 ꢃ 5 mL Et2O and dried under vacuum. Sodium
methoxide (30 mg, 0.56 mmol) was added under and the
mixture was dissolved in thf (30 mL). The yellowish
solution slowly reddened over a 16 h period. The thf was
then removed in vacuo, the residue extracted with a
hexane:diethylether mixture (5:1), filtered through a
small alumina pad, and then subjected to silica-gel
chromatography. The isomeric mixture of [CpNi{l-
g2,g2-HC2CMe2(C5H5)}Mo(CO)2Cp0] (Ni–Mo) (3a
and 4a) eluted first, using a hexane:diethylether mixture
(200:1), as an orange band which could be concentrated
to a brown oil (10 mg, 9.3%). Complex 2a followed as an
orange brown band using a more polar hexane:diethy-
lether mixture (1:5) and was subsequently recrystallized
from this mixture. Yield: 8 mg (8%). IR [m(CO), cmꢀ1]:
1983(s), 1949(m), 1925(s), 1855(w). HRMS for the 3a:4a
mixture(60Ni, 100Mo; m=e): 488.017 C23H24MoNiO2 re-
quires 488.018.
4. Experimental
4.1. (a) General
All manipulations were performed under an atmo-
sphere of pre-purified nitrogen using standard Schlenk-
ware techniques. Solvents were pre-dried and distilled
from sodium benzophenone ketyl solutions (diethyle-
ther, thf, hexane). The syntheses of the l-alkyne com-
plexes 2a and 2b, the heterobimetallic propargylic
4.3. (c)
A suspension of 1a was prepared as described (Sec-
tion 4.2) starting with 2a (74 mg, mmol). The suspension
of 1a was dissolved in thf and allowed to stand at
room temperature overnight. The solution darkened

1888
M.J. Chetcuti, S.R. McDonald / Journal of Organometallic Chemistry 689 (2004) 1882–1889
progressively and insoluble solids precipitated. The
solvent was then removed in vacuo and the residue ex-
tracted with toluene to give a light green-brown solution
which was concentrated. Chromatography of this tolu-
ene solution on a silica-gel column with toluene eluted
small quantities of a green eluate. This green product is
volatile and paramagnetic, and is probably nickelocene.
A slightly more polar solvent mixture of hexane: dieth-
ylether (50:1) eluted the brown ethoxyalkyne complex
7a [CpNi{l-g2,g2-HC2CMe2(OEt)}Mo(CO)2Cp0] (Ni–
Mo), while 5a was eluted using a 10:1 ether:hexane
mixture. All three complexes were isolated as oils.
Yields: 5a; 8 mg, 8%. 6a; 50 mg, 52%. 7a; 12 mg, 11%.
IR (6a), [m(CO), cmꢀ1]: 1970(s), 1947(m), 1909(s),
1874(m), 1858(m). HRMS for 6a, (60Ni, 100Mo; m=e):
423.985 C18H20MoNiO2 requires 423.986.
4.4. (d)
IR (7a), [m(CO), cmꢀ1]: 1983(s), 1961(m), 1925(s),
1853(m). HRMS for 7a, (60Ni, 100Mo; m=e): 468.011.
C20H24MoNiO3 requires 468.012.
A suspension of 1a was prepared as noted in Section
4.2, by starting with 2a (124 mg, 0.28 mmol). Excess
NaC5H5 (0.85 mL of a 0.72 M solution in thf, 0.61
mmol) was then added to the solid 1a and the solution
stirred overnight. The thf was removed from the reac-
tion mixture in vacuo, and the residue was dissolved in
toluene and subjected to chromatography on a silica gel
column. Elution with toluene afforded a brown band
that was found to contain ꢄ18 mg of the 3a:4a mixture.
4.7. (g) Reaction of 2b with HBF4 ꢁ Et2O followed by
NaOMe to afford the 3b, 4b isomeric mixture [CpNi{l-
g2,g2-HC2CMe2(C5H5)}W(CO)2Cp0] (Ni–W)
The reaction of the Ni–W complex mirrored that of
the Ni–Mo (Section 4.2). Complex 2b (100 mg, 0.18
mmol) was reacted with 1% HBF4 ꢁ Et2O and then
treated with NaOMe (22 mg, 0.41 mmol) in thf (20 mL)
for 18h. The mixture turned from yellow-brown to red-
brown. After solvents were removed in vacuum, the
residue was dissolved in a hexane–ether mixture and
subjected to chromatography on a silica-gel. A single
brown band eluted and yielded the 3b:4b as an oily
mixture when concentrated. Yield: 10 mg (9%). HRMS
for the 3b:4b mixture(60Ni, 184W; m=e): 574.062.
C23H24NiO2W requires 574.063.
4.5. (e) Reaction of 1a with wet KOBut to afford the 3a/
4a mixture and complex 5a, [CpNi{l-g2,g2-HC2
CMe2(OH)}Mo(CO)2Cp0] (Ni–Mo)
KOBut (44 mg, 0.39 mmol) was added to a sample of
1a [prepared from 2a (100 mg, 0.22 mmol) as described
in Section 4.2] in thf (15 mL). The solution was stirred
for 19 h, after which time the solvent was removed under
vacuo and the residue extracted with hexanes and fil-
tered through a Celite pad. Chromatography (using
silica gel) led to the isolation of two bands: the 3a/4a
mixture eluted with a hexane:diethylether (200:1), fol-
lowed by 5a, [CpNi{l-g2,g2-HC2CMe2(OH)}Mo(CO)2
Cp0] (Ni–Mo). This complex eluted with a 1:1 mixture of
hexane:diethylether. Yields: 3a/4a, 15 mg (14%). 5a,
17 mg (18%). Anal. Calc. for C18H20MoNiO3, 5a: C,
49.25; H, 4.59. Found: C, 49.32, H. 4.74. IR [m(CO),
cmꢀ1]: 1986(s), 1928(s), 1861(w); m(CO) (in mineral oil,
cmꢀ1): 1980(s), 1926(s). HRMS for 5a, (60Ni, 100Mo;
m=e): 439.982. C18H20MoNiO3 requires 439.981.
4.8. (h) Reaction of [CpNi(l-g2(Ni),g3(W)-
HC2CMe2)W(CO)2Cp0]þBFꢀ4 (Ni–Mo, 1b) with Na-
OMe to afford the 3b/4b mixture
The cationic metallacyclic complex [CpNi{l-l-
g3(Ni),g1(M)-C(OH)-C(Me2(OMe)-C(H))}W(CO)2Cp0]
þBFꢀ4 (Ni–W) (48 mg, 0.073 mmol) was stirred in thf (10
mL) and monitored by IR till the complex had com-
pletely transformed into 1b (2.5 h). NaOMe (17 mg, 0.31
mmol) was then added and the solution stirred for 18h.
The solution was then passed through an alumina pad
and concentrated to give a dark brown oil of the 3b:4b
mixture. Yield: 17 mg (40%).
4.6. (f) Reaction of 1a with NaBH4 in wet ethanol to
afford 5a, complex 6a [CpNi(l-g2,g2-HC2Pri)Mo(CO)2
Cp0] (Ni–Mo), and complex 7a, [CpNi{l-g2, g2-HC2
CMe2(OEt)}Mo(CO)2Cp0] (Ni–Mo)
Acknowledgements
NaBH4 (18 mg, mmol) was added to a 1a, [prepared
from 2a (103 mg, 0.23 mmol) as described in Section
4.2], in ethanol (20 mL). The cation was slightly soluble
in this solvent and dissolved with effervescence as it
reacted. After 1 h, the now reddish brown solution was
pumped down to dryness in vacuum, and subjected to
chromatography on a silica-gel column. The major
product was l-PrC2H complex 6a [CpNi(l-g2,g2-
PriC2H)Mo(CO)2Cp0] (Ni–Mo), which eluted as a red-
brown band with a hexane: diethylether mixture (200:1).
This work was started and mainly carried out at the
University of Notre Dame, but was completed at the
ꢀ
ECPM, universite Louis Pasteur. Funding by the A.C.S.
administered PRF fund, and by universite Louis Pasteur
ꢀ
is gratefully acknowledged.
References
[1] M.J. McGlinchey, L. Girard, R. Ruffolo, Coord. Chem. Rev. 143
(1995) 331–381.

M.J. Chetcuti, S.R. McDonald / Journal of Organometallic Chemistry 689 (2004) 1882–1889
1889
[2] K.M. Nicholas, Acc. Chem. Res. 20 (1987) 207–214.
[3] S. Padmanabhan, K.M. Nicholas, J. Organomet. Chem. 268
(1983) C23–C27.
[12] M. Scherer, J.L. Sessler, A. Gebauer, V. Lynch, J. Org. Chem. 62
(1997) 7877–7881.
[13] M.J. Chetcuti, P.E. Fanwick, B.E. Grant, J. Organometal. Chem.
630 (2001) 215–220.
[4] T.J.J. Muller, Eur. J. Org. Chem. 11 (2001) 2021–2033.
[5] D. Osella, R. Mauro, N. Carlo, G. Cavigiolio, M. Vincenti, A.
[14] M.J. Chetcuti, P.E. Fanwick, S.R. McDonald, N.P. Rath,
Organometallics 10 (1991) 1551–1560.
Vessieres, G. Jaouen, Eur. J. Inorg. Chem.
497.
3 (2000) 491–
[15] M.J. Chetcuti, L. DeLiberato, P.E. Fanwick, B.E. Grant, Inorg.
Chem. 29 (1990) 1295–1298.
[6] M. Gruselle, H. El Hafa, M. Nikolski, G. Jaouen, J. Vaissermann,
L. Li, M.J. McGlinchey, Organometallics 12 (1993) 4917–
4925.
[16] M.J. Chetcuti, B.E. Grant, P.E.J. Fanwick, Am. Chem. Soc. 111
(1989) 2743–2744.
[7] M. Gruselle, M.A. Kondratenko, H. El Amouri, J. Vaissermann,
Organometallics 14 (1995) 5242–5250.
[17] G.K. Turner, W. Klaeui, M. Scotti, H. Werner, J. Organomet.
Chem. 102 (1975) C9.
[8] M.F. D’Agostino, C.S. Frampton, M.J. McGlinchey, J. Organo-
met. Chem. 394 (1990) 145–166.
[18] H. Werner, G. Mattmann, A. Salzer, T. Winkler, J. Organomet.
Chem. 25 (1970) 461.
[9] M.A. Kondratenko, M.-N. Rager, J. Vaissermann, M. Gruselle,
Organometallics 14 (1995) 3802–3809.
[19] M.J. Chetcuti, in: E.W. Abel, F.G.A. Stone, G. Wilkinson (Eds.),
Comprehensive Organometallic Chemistry II, vol. 9, Pergamon,
Oxford, 1995, pp. 108–191.
[10] M.J. Chetcuti, S.R. McDonald, Organometallics 21 (2002) 3162–
3168.
[11] A.J.I. Ashe, J. Am. Chem. Soc. 93 (1970) 1233.
[20] M.J. Chetcuti, S.R. McDonald, N.P. Rath, Organometallics 8
(1989) 2077–2079.