Hydrogen-halogen exchange reactions of (η5:η5-C10H8)W 2(CO)4(PPh2Me)2H2 with CCl4 and CHI3; Characterization of (η5:η5-C10H8)W 2(CO)4(PPh2Me)2X2 (X=Cl, I)

Article abstract of DOI:10.1016/S0022-328X(99)00174-6

Hydrogen-halogen exchange reactions of FvW₂(CO)₄(PPh₂Me)₂H₂ (1) with CCl₄ and CHI₃ gave FvW₂(CO)₄(PPh₂Me)₂Cl₂ (2) and FvW₂(CO)₄(PPh₂Me)₂I₂ (3), respectively, in nearly quantitative isolated yield (Fv=η⁵:η⁵-C₁₀H₈, fulvalene). Both reactions were shown by ¹H- and ³¹P{¹H}-NMR spectroscopy to proceed via FvW₂(CO)₄(PPh₂Me)₂HX (X=Cl, I) intermediates. Two isomers of the hydrido-halide complexes, containing a cis or trans halide half, were identified in the reaction mixtures. When the reactions were complete, 3 was found to maintain an equilibrium mixture of cis,cis (D,L-/meso-=1:1), cis,trans and trans,trans isomers in solution. The equilibrium was strongly solvent dependent. While minor amounts of the cis,trans (but no trans,trans) isomer were also detected during the formation of 2, this compound was found to exist solely as a 1:1 mixture of D,L- and meso-cis,cis isomers in equilibrated solutions. Crystallization of 3 gave the meso-cis,cis isomer, which was characterized by X-ray diffraction.

Full text of DOI:10.1016/S0022-328X(99)00174-6

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 584 (1999) 347–355
Hydrogen–halogen exchange reactions of
(h5:h5-C10H8)W2(CO)4(PPh2Me)2H2 with CCl4 and CHI3;
characterization of (h5:h5-C10H8)W2(CO)4(PPh2Me)2X2 (X=Cl, I)
and crystal structure determination of the iodide
Istvan Kovacs *, Celine Pearson, Alan Shaver
Department of Chemistry, McGill Uni6ersity, 801 Sherbrooke St. W., Montreal, Quebec, H3A 2K6, Canada
Received 15 January 1999
Abstract
Hydrogen–halogen exchange reactions of FvW2(CO)4(PPh2Me)2H2 (1) with CCl4 and CHI3 gave FvW2(CO)4(PPh2Me)2Cl2 (2)
and FvW2(CO)4(PPh2Me)2I2 (3), respectively, in nearly quantitative isolated yield (Fv=h5:h5-C10H8, fulvalene). Both reactions
were shown by 1H- and 31P{1H}-NMR spectroscopy to proceed via FvW2(CO)4(PPh2Me)2HX (X=Cl, I) intermediates. Two
isomers of the hydrido–halide complexes, containing a cis or trans halide half, were identified in the reaction mixtures. When the
reactions were complete, 3 was found to maintain an equilibrium mixture of cis,cis (D,L-/meso-=1:1), cis,trans and trans,trans
isomers in solution. The equilibrium was strongly solvent dependent. While minor amounts of the cis,trans (but no trans,trans)
isomer were also detected during the formation of 2, this compound was found to exist solely as a 1:1 mixture of D,L- and
meso-cis,cis isomers in equilibrated solutions. Crystallization of 3 gave the meso-cis,cis isomer, which was characterized by X-ray
diffraction. © 1999 Elsevier Science S.A. All rights reserved.
Keywords: Metal hydrides; Metal halides; Hydrogen–halogen exchange; Fulvalene complexes; Tungsten complexes; Carbonyl complexes;
Phosphine substitution; Isomerization; NMR spectroscopy; Crystal structure
1. Introduction
number of cases, this method was not only more conve-
nient and economic than the alternative syntheses uti-
lizing the elemental halogen and/or phosphine
substitution but is the only way to obtain the target
compounds. Such dihalides are versatile intermediates
in the synthesis of other fulvalene derivatives and also
are interesting in their own right [2,3]. The chemistry of
dinuclear (fulvalene)metal carbonyl and related com-
plexes has attracted considerable interest in recent years
due to their unique reactivity, different from that of
analogous cyclopentadienyl complexes, which is a result
of the close proximity and electronic communication of
neighbouring metal centres anchored to the fulvalene
ligand [4]. As for the hydrogen–halogen exchange reac-
tion, it has been reported that (fulvalene)chromium
and molybdenum dihydrides react in two distinctive
steps, via hydrido–halide intermediates of the type
The hydrogen–halogen exchange reaction of metal
carbonyl hydrides (HM(CO)nLm; L=phosphine) with
sufficiently activated alkyl halides (RX; X=Cl, Br, I) is
a mild and straightforward method to produce the
corresponding metal halide derivatives in practically
quantitative yield [1]. One of the most recent applica-
tions of this reaction has been the synthesis of
FvM2(CO)4L2X2 (Fv=h5:h5-C10H8 (fulvalene); M=
Cr, Mo, W; L=CO, phosphine; X=Cl, Br, I) type
dihalides from the corresponding dihydrides [2]. In a
* Corresponding author.
E-mail address: ezc4@musica.mcgill.ca (I. Kovacs)
0022-328X/99/$ - see front matter © 1999 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(99)00174-6

348
I. Ko6acs et al. / Journal of Organometallic Chemistry 584 (1999) 347–355
1
FvM2(CO)4L2HX, to give the corresponding dihalides
[2b,c]. Although the complexes FvW2(CO)6X2 (X=Cl,
Br) were also prepared by this method, no comment
was made about the course of the reaction [2a].
FvW2(CO)6I2 was obtained by treatment of the corre-
sponding dihydrides with iodine [2a]. Its ligand-substi-
tuted derivatives were obtained by Me3NO assisted
ligand substitution of the parent hexacarbonyl [3e].
One of the most interesting features of ligand-substi-
tuted fulvalene complexes of the general formula
FvM2(CO)4L2X2 (M=Cr, Mo, W; L=phosphine,
phosphite; X=H, halide, SH, SR) type complexes is
their isomerization behaviour. Four isomers (not in-
cluding enantiomers) are possible: cis,cis (D,L- and
meso-), cis,trans and trans,trans (Scheme 1). Each iso-
mer can easily be identified by its complex NMR
pattern reflecting its symmetry [2,3e]. In particular, the
dihalide isomers, unlike the dihydrides, undergo slow
exchange on the NMR time scale at room temperature
and thus exhibit well resolved spectra. NMR studies
indicate that sometimes the equilibrium lies completely
in favour of the cis,cis isomers. Although this cannot be
predicted on the basis of present knowledge of the
combined steric and electronic effects of the metal,
ligand and halide, it suggests that FvM2(CO)4L2X2
dihalides might have a general tendency to assume the
cis,cis arrangement under appropriate conditions. This
is supported by our most recent observation that
FvW2(CO)4(PPh2Me)2(SH)2, which shows an isomeriza-
tion behaviour similar to FvW2(CO)4(PPh2Me)2I2 in
solution (four isomers), crystallized solely as the cis,cis
shifted upon crystallization. The H-NMR spectrum of
freshly dissolved cis,cis-FvMo2(CO)4(PCy3)2Cl2 in
CD2Cl2 exhibited a ca. 2:1 ratio of D,L and meso forms
and the final 1:1 ratio was achieved quite slowly [2b]. In
order to determine which one of the two cis,cis forms of
a particular complex is present in the crystalline state,
an X-ray diffractional analysis has to be carried out.
However, no crystallographic data at all are available
in the literature for substituted fulvalene complexes of
the general formula FvM2(CO)4L2X2 (M=Cr, Mo, W;
L=phosphine, phosphite; X=H, halide, SH, SR,
alkyl, acyl, etc.).
Here we report a multinuclear NMR study of the
hydrogen–halogen
exchange
reaction
of
FvW2(CO)4(PPh2Me)2H2 (1) with CCl4 and CHI3, char-
acterization of the corresponding dihalide products
FvW2(CO)4(PPh2Me)2X2 (X=Cl (2), I (3)), as well as
the crystal structure of meso-cis,cis-3. A remarkable
solvent and phase dependency of the isomeric composi-
tion of 3 was observed.
2. Experimental
2.1. General procedures
All manipulations were carried out under an inert
atmosphere (N2 or Ar) at room temperature, using
standard Schlenk technique and a drybox. Solvents
were dried and freshly distilled under nitrogen prior to
use. FvW2(CO)4(PPh2Me)2H2 (1) was prepared as re-
ported [2d]. IR spectra were recorded on a Bruker IFS
48 spectrometer in THF solutions using a 0.1 mm NaCl
isomer(s) [2d]. Upon dissolution, a 1:1 mixture of D,L
-
and meso-cis,cis isomers slowly transformed into the
equilibrium mixture of four isomers. The reverse pro-
cess, however, has been demonstrated previously for
cis,cis-FvMo2(CO)4(PPh3)2I2 [2b].
Furthermore, the cis,cis isomers are always present in
solution as a 1:1 mixture of D,L and meso forms. In
only one instance has this equilibrium, which is gener-
ally much faster than the cisltrans equilibrium, been
cell. H-, 13C{1H}-, and 31P{1H}-NMR measurements
1
were performed on Jeol CPF 270 and Varian Unity 500
spectrometers at operating frequencies 270.167
(499.843), 67.940 and 109.376 MHz, respectively.
COSY experiments were performed on the 500 MHz
instrument. Elemental analyses were carried out by the
‘Laboratoire d’Analyse Elementaire’ at the University
of Montreal.
2.2. F6W2(CO)4(PPh2Me)2Cl2 (2)
To a 50 ml CH2Cl2 solution of 1 (1.0 g, 0.99 mmol)
5 ml of CCl4 was injected at room temperature. The
originally yellow solution rapidly changed colour to red
and after stirring for 1 day a red precipitate had
formed. The volume of the solution was reduced to
about 10 ml under vacuum and hexane was added to
complete the precipitation of 2. The solid product was
filtered, washed with hexane and dried under reduced
pressure. Yield: 1.08 g (1.00 mmol, 99%). Complex 2 is
very poorly soluble in most commonly used solvents.
Crystalline samples were obtained by performing the
reaction in a filtered and concentrated acetone-d6 solu-
Scheme 1.

I. Ko6acs et al. / Journal of Organometallic Chemistry 584 (1999) 347–355
349
tion of 1 without agitation. Combustion analysis sug-
gested that two solvent molecules per complex were
retained in the lattice upon crystal formation, similar to
3 (see below). Anal. Calc. for C46H34D12Cl2O6P2W2: C,
45.76; H(D), 3.84. Found: C, 45.56; H(D), 3.84. IR
2.3.2. NMR data in DMSO-d6 solution (500 MHz)
1
cis,cis isomers (D,L-/meso-=1:1, 14%); H-NMR: l
2.57 (d, J(P–H)=9 Hz, 12H, PMe), 5.07, 5.23, 5.28,
5.70, 5.94 (all m, 2H, Fv), 6.35 (m, 4H, Fv), :7.3–7.5
(m, 40H, PPh). 31P{1H}-NMR: l −1.9, −1.3 (both s,
J(W–P)=240 Hz, 2P).
(THF): w(CO) 1949 (vs), 1856 (m) cm−1
.
cis,trans isomer (53%); 1H-NMR: l 2.35 (d, J(P–
H)=9 Hz, 3H, PMe, trans), 2.52 (d, J(P–H)=9 Hz,
3H, PMe, cis), 5.00, 5.20, 5.28, 5.33, 5.38, 5.43, 5.70,
6.17 (all m, 1H, Fv), :7.3–7.5 (m, 20H, PPh).
31P{1H}-NMR: l −1.9 (s, J(W–P)=240 Hz, 1P, cis),
9.1 (s, J(W–P)=240 Hz, 1P, trans).
2.2.1. NMR data in acetone-d6 solution (270 MHz)
cis,cis isomers (D,L-/meso-=1:1, 100%); H-NMR: l
1
2.28, 2.29 (both d, J(P–H)=9 Hz, 6H, PMe), 5.17 (m,
6H, Fv), 5.31, 5.64, 5.76, 5.83, 6.10 (all m, 2H, Fv),
7.45 (m, 24H, m,p-Ph), 7.52, 7.63 (both m, 8H, o-Ph).
13C{1H}-NMR: l 84.5, 84.6, 86.1, 87.0, 90.2, 92.2, 96.9,
97.3 (all Fv), 128.3, 128.4 (both d, J(P–C)=8 Hz,
m-Ph), 130.1, 130.4 (both p-Ph), 132.3, 132.4 (both d,
J(P–C)=9 Hz, o-Ph). The weak PMe, C-1 fulvalene
and ipso-Ph resonances could not be distinguished un-
equivocally due to low concentration of the samples.
31P{1H}-NMR: l 7.9, 8.1 (both s, J(W–P)=260 Hz,
2P).
1
trans,trans isomer (33%); H-NMR: l 2.33 (d, J(P–
H)=9 Hz, 6H, PMe), 5.20 (m, 8H, Fv), :7.3–7.5 (m,
20H, PPh). 31P{1H}-NMR: l 9.3 (s, J(W–P)=240 Hz,
2P).
2.3.3. NMR data in CDCl3 solution (270 MHz)
1
cis,cis isomers (D,L-/meso-=1:1, 24%); H-NMR: l
2.46, 2.48 (both d, J(P–H)=8 Hz, 6H, PMe), 4.81,
4.90, 5.10, 5.26, 5.67, 5.73 (all m, 2H, Fv), 5.19 (m, 4H,
Fv), :7.4–7.7 (m, 40H, PPh). 31P{1H}-NMR: l −4.4,
−4.3 (both s, 2P).
2.3. F6W2(CO)4(PPh2Me)2I2 (3)
Complex 1 (1.5 g, 1.48 mmol) was dissolved in 50 ml
of toluene and solid iodoform (1.2 g, 3.05 mmol) was
added to the stirred solution at room temperature. The
starting yellow colour quickly changed to orange and a
red–orange solid precipitated overnight. Following a
workup similar to that of 2, complex 3 was obtained as
an orange powder. Yield: 1.8 g (1.41 mmol, 95%). For
an alternative preparation see reference [3e]. Again, a
crystalline sample was obtained by performing the reac-
tion in a concentrated acetone-d6 solution of 1. Com-
bustion analysis suggested that the crystals consisted of
3 and solvent in the lattice in a 2:1 molar ratio. This
was verified by an X-ray crystal structure. Anal. Calc.
for C46H34D12I2O6P2W2: C, 39.74; H(D), 3.33. Found:
C, 39.39; H(D), 3.13. IR (THF): w(CO) 1949 (s), 1870
cis,trans isomer (59%); 1H-NMR: l 2.36 (d, J(P–
H)=9 Hz, PMe, trans), 2.43 (d, J(P–H)=8 Hz, PMe,
cis), 4.66 (m, 1H, Fv, trans), 4.75 (m, 1H, Fv, trans),
4.90 (m, 1H, Fv, cis), 4.95 (m, 1H, Fv, trans), 5.00 (m,
1H, Fv, trans), 5.14 (m, 1H, Fv, cis), 5.19 (m, 1H, Fv,
,
cis), 5.79 (m, 1H, Fv, cis), A7.4-7.7 (m, 20H, PPh).
31P{1H}-NMR: l −4.3 (s, 1P, cis), 10.3 (s, 1P, trans).
1
trans,trans isomer (17%); H-NMR: l 2.34 (d, J(P–
H)=9 Hz, 6H, PMe), 4.81, 5.00 (both m, 4H, Fv),
:7.4–7.7 (m, 20H, PPh). 31P{1H}-NMR: l 10.5 (s,
2P).
2.4. Characterization of the hydrido–halide
intermediates
(vs) cm−1
.
Reactions of 1 (30 mg, 0.03 mmol) with CCl4 (40 ml,
0.41 mmol) or CHI3 (24 mg, 0.06 mmol) in 0.7 ml
1
2.3.1. NMR data in acetone-d6 solution (270 MHz)
acetone-d6 solutions were monitored by H- (270 MHz)
1
cis,cis isomers (D,L-/meso-=1:1, 28%); H-NMR: l
and 31P{1H}-NMR spectroscopy at room temperature.
The following data were collected:
2.63 (d, J(P–H)=8 Hz, 12H, PMe), 5.16, 5.20, 5.26,
5.45, 5.56, 5.71 (all m, 2H, Fv), 6.14 (m, 4H, Fv),
:7.4–7.7 (m, 40H, PPh). 31P{1H}-NMR: l −3.5,
−3.2 (both s, J(W–P)=240 Hz, 2P).
cis-FvW2(CO)4(PPh2Me)2HCl; 1H-NMR: l −7.07
(br d, J(P–H)=55 Hz, J(W–H)=48 Hz, 1H, WH),
2.24 (d, J(P–H)=9 Hz, 3H, PMe of hydride), 2.27 (d,
J(P–H)=9 Hz, 3H, PMe of chloride), 5.04, 5.09, 5.21,
5.70, 5.92 (all m, 1H, Fv). The missing fulvalene reso-
nances were hidden under those of 1 and 2, probably in
the range of l 5.13–5.20. Phenyl resonances could not
be distinguished from those of 1 and 2. 31P{1H}-NMR:
l 8.2 (s, 1P, chloride), 18.9 (br s, 1P, hydride).
cis,trans isomer (48%); 1H-NMR: l 2.43 (d, J(P–
H)=9 Hz, 3H, PMe, trans), 2.57 (d, J(P–H)=9 Hz,
3H, PMe, cis), 4.96 (m, 1H, Fv), 5.16 (m, 2H, Fv), 5.20,
5.27, 5.42, 5.56, 6.03 (all m, 1H, Fv), :7.4–7.7 (m,
20H, PPh). 31P{1H}-NMR: l −3.5 (s, J(W–P)=240
Hz, 1P, cis), 9.4 (s, J(W–P)=240 Hz, 1P, trans).
1
1
trans,trans isomer (24%); H-NMR: l 2.40 (d, J(P–
trans-FvW2(CO)4(PPh2Me)2HCl; H-NMR: l −7.06
H)=9 Hz, 6H, PMe), 5.13 (m, 8H, Fv), :7.4–7.7 (m,
20H, PPh). 31P{1H}-NMR: l 9.7 (s, J(W–P)=240 Hz,
2P).
(br d, J(P–H)=55 Hz, J(W–H)=48 Hz, 1H, WH).
No other resonances of this minor intermediate could
be positively identified due to extensive overlap with

350
I. Ko6acs et al. / Journal of Organometallic Chemistry 584 (1999) 347–355
PPh). 31P{1H}-NMR: l 7.8 (s, J(W–P)=260 Hz, 1P,
cis), 13.8 (s, J(W–P)=220 Hz, 1P, trans).
cis-FvW2(CO)4(PPh2Me)2HI; H-NMR: l −7.03 (br
Table 1
Crystallographic
FvW2(CO)4(PPh2Me)2I2 · 2(CD3)2CO
data
and
structure
refinement
for
1
d, J(P–H)=55 Hz, J(W–H)=48 Hz, 1H, WH), 2.27
(d, J(P–H)=9 Hz, 3H, PMe of hydride), 2.58 (d,
J(P–H)=9 Hz, 3H, PMe of iodide). The fulvalene and
phenyl resonances could not be assigned due to overlap
with those of 1 and 3. 31P{1H}-NMR: l −3.2 (s, 1P,
iodide), 19.0 (br s, 1P, hydride).
Empirical formula
Formula weight
T (K)
C46H34D12I2O6P2W2
1390.29
299 (2)
Mo–Ka, 0.70930
Monoclinic
P21/c
13.2577 (2)
9.6743 (2)
20.2480 (2)
90
112.103 (1)
90
2406.13 (7)
2
1.919
0.626
,
Radiation, u (A)
Crystal system
Space group
,
a (A)
1
,
b (A)
trans-FvW2(CO)4(PPh2Me)2HI; H-NMR: l −7.06
,
c (A)
(br d, J(P–H)=55 Hz, J(W–H)=48 Hz, 1H, WH),
2.23 (d, J(P–H)=9 Hz, 3H, PMe of hydride), 2.39 (d,
J(P–H)=9 Hz, 3H, PMe of iodide). The fulvalene and
phenyl resonances could not be assigned due to overlap
with those of 1 and 3. 31P{1H}-NMR: l 9.9 (s, J(W–
P)=240 Hz, 1P, iodide), 19.0 (br s, 1P, hydride).
h (°)
i (°)
k (°)
3
,
V (A )
Z
DCalc (Mg m−3
)
v (mm−1
)
Transmission range
F(000)
0.67 to 0.91
1308
2.5. X-ray structure determination for 3
Crystal size (mm)
Limiting indices
0.24 × 0.20 × 0.02
−175h516
−125k510
−115l525
16 789
Orange crystals of 3 were obtained from a reaction of
1 with iodoform carried out overnight in a filtered and
saturated acetone-d6 solution at room temperature. A
well-shaped crystal was selected by optical microscopy.
Intensity data were collected on a Siemens P4 diffrac-
tometer equipped with a SMART 1K CCD area detec-
tor. Data collection and structure solution parameters
are listed in Table 1. The initial unit cell parameters
were determined by a least-squares fit of the angular
setting of 7231 strong reflections, collected by a 4.5°
scan in 15 frames over three different parts of recipro-
cal space (45 frames total). One complete hemisphere of
Reflections collected
Independent reflections (Rint
)
5408 (0.055)
3769 (I\2|(I))
1.058
Observed reflections (criterion)
Goodness-of-fit on F2
Final R indices (I=2|(I))
R indices (all data)
R1=0.0496, wR2=0.1032a
R1=0.0857, wR2=0.1123a
a R1=ꢀ(Fo−Fc)/ꢀ Fo;
W
R2 =[ꢀ w(Fo2−F2c)2/ꢀ w(F2o)2]1/2
.
those of major species. 31P{1H}-NMR: l 14.5 (s, 1P,
chloride), 19.5 (br s, 1P, hydride).
,
data was collected to better than 0.8 A resolution.
1
cis,trans-2; H-NMR: l 2.25 (d, J(P–H)=9 Hz, 3H,
Upon completion of data collection, the first 50 frames
were recollected to improve decay correction analysis.
A decay correction and an empirical absorption correc-
tion based on redundant reflections were performed by
the program SADABS. The program SAINT was applied
for Lorentz and polarization corrections.
PMe, cis), 2.42 (d, J(P–H)=9 Hz, 3H, PMe, trans),
4.81, 4.89 (both m, 1H, Fv), 5.10, :5.21 (both m, 2H,
Fv), 5.85, 6.03 (both m, 1H, Fv), :7.4–7.5 (m, 20H,
Table 2
,
Selected bond lengths (A) and angles (°) for meso-cis,cis-3
The structure was solved by the Patterson method
using SHELXS-96. All non-hydrogen atoms were refined
anisotropically, the hydrogen (deuterium) atoms were
introduced in calculated positions. Selected bond dis-
tances and angles are listed in Table 2. The crystal
structure of 3 is shown in Fig. 2. A full listing of bond
distances and angles, atomic coordinates, isotropic and
anisotropic displacement parameters are available as
Supplementary Material.
,
Bond lengths (A)
W–C(2)
W–C(3)
W–C(4)
W–C(5)
W–C(6)
W–C(7)
W–C(8)
W–(P)
2.0000(9)
1.9997(8)
2.426(8)
2.333(9)
2.285(9)
2.306(9)
2.387(8)
2.508(2)
2.8122(9)
C(2)–O(2)
C(3)–O(3)
C(4)–C(5)
C(5)–C(6)
C(6)–C(7)
C(7)–C(8)
C(4)–C(8)
C(4)–C(4%)
1.1603(9)
1.1599(9)
1.450(11)
1.431(13)
1.407(14)
1.443(13)
1.394(12)
1.45(2)
W–I(1)
Bond angles (°)
C(2)–W–P
C(2)–W–I(1)
C(3)–W–P
C(3)–W–I(1)
P–W–I(1)
O(2)–C(2)–W
O(3)–C(3)–W
3. Results and discussion
121.3(2)
78.6(3)
79.8(2)
125.2(2)
78.79(6)
176.8(8)
178.1(11)
C(6)–C(5)–C(4)
C(7)–C(6)–C(5)
C(8)–C(7)–C(6)
C(4)–C(8)–C(7)
C(8)–C(4)–C(5)
C(4%)–C(4)–C(5)
C(8)–C(4)–C(4%)
108.5(8)
107.4(8)
108.2(8)
109.0(9)
106.9(7)
125.3(10)
127.7(10)
3.1. Synthesis and characterization of F6W2(CO)4-
(PPh2Me)2X2 (X=Cl (2), I (3))
Reactions of FvW2(CO)4(PPh2Me)2H2 (1) with car-
bon tetrachloride and iodoform readily gave red

I. Ko6acs et al. / Journal of Organometallic Chemistry 584 (1999) 347–355
351
Fig. 1. COSY (500 MHz) spectrum of complex 2 in acetone-d6 solution.
FvW2(CO)4(PPh2Me)2Cl2 (2) and orange FvW2(CO)4-
(PPh2Me)2I2 (3), respectively, in excellent isolated
yields. The use of iodoform as hydrogen–iodine ex-
change reagent is more advantageous than ethyl
iodoacetate, iodoacetonitrile, t-butyl iodide, etc. [2b,c],
because it is an inexpensive, stable and easy-to-handle
solid with high reactivity toward tungsten hydrides. In
addition, the hydrogenated side-products (CH2I2 or
MeI) are volatile and exhibit very simple proton spec-
tra, which do not obscure those of the fulvalene com-
plexes. In general, hydrogen–iodine exchange reactions
give metal iodides of higher purity than those utilizing
I2. Both 2 and 3 were isolated as poorly soluble crys-
talline materials and characterized by combustion anal-
ysis, IR, 1H- and 31P{1H}-NMR spectroscopy. The
crystal structure of 3 was also determined. While this
work was in progress, an alternative synthesis of 3 via
triethylamine oxide assisted substitution of FvW2-
(CO)6I2 with PPh2Me was reported [3e].
protons, six different fulvalene multiplets at 5.17 (3×),
5.31, 5.64, 5.76, 5.83, 6.10, and phenyl resonances
consistent with two different PPh environments. The
observed resonance pattern is characteristic of a 1:1
equilibrium mixture of D,L- and meso-cis,cis isomers. It
is also supported by the 31P{1H}-NMR spectrum which
featured two closely spaced signals of equal intensity at
l 7.9 and 8.1. Evidence for four different Ca and Cb
fulvalene carbon atoms and two different PPh environ-
ments was obtained in the 13C{1H}-NMR spectrum.
The spectroscopic results also suggested that the cis,cis
isomers of 2 did not transform into the trans isomers to
any significant extent and their interconversion was
slow on the NMR time scale. The same isomeric com-
position has been established recently for FvMo2-
(CO)4(PR3)2X2 (X=Cl, Br) [2b], FvCr2(CO)4-
(PPh2Me)2X2 (X=Br, I) [2c] and FvW2(CO)4(PCy3)2I2
[3e].
Additional information about 2 was obtained by a
500 MHz COSY experiment. Fig. 1 demonstrates that
the resonances at l 5.19 (2×) are coupled to those at
5.66 and 6.13 and probably to each other, while those
at l 5.22 and 5.33 are coupled to those at 5.86 and 5.79,
respectively, as well as to each other. Each group of
four signals therefore belong to different cis,cis forms.
The coupling pattern also suggests that all four upfield
Complex 2 had two absorbances in the IR spectrum
at 1949 (vs), 1856 (m) cm−1, suggesting that both
hydridic sites had been chlorinated and the chloride
and phosphine ligands occupy mutually cis positions
1
about each metal centre. The 270 MHz H-NMR spec-
trum in acetone-d6 solution exhibited two doublets of
equal intensity at l 2.28 and 2.29 attributable to PMe

352
I. Ko6acs et al. / Journal of Organometallic Chemistry 584 (1999) 347–355
signals represent Hb protons and all four downfield
signals belong to Ha protons. This is in complete agree-
ment with previous heteronuclear NOE difference mea-
surements [5].
relative concentration of different isomers, which is
otherwise easily determined by integrating both the
phosphorus and methyl proton signals. As a COSY
experiment only determines the proton resonances at-
tributable to individual h5-ring environments, correct
identification of the three types of cis environments
Complex 3 shows markedly different isomerization
behaviour than 2. Both the 1H- and 31P{1H}-NMR
spectra indicated that this compound maintains a per-
manent equilibrium of all four isomers (Scheme 1), the
cis,trans being in the highest concentration. In acetone-
d6 solution, the 31P{1H}-NMR spectrum exhibited two
distant sets of signals, those at high-field (l −3.5 and
−3.2) being characteristic to cis and those at low-field
(l 9.4 and 9.7) to trans P–W–I arrangements. Peak
integrals suggested that the resonance of one of the
cis,cis forms was hidden under that of the cis half of the
cis,trans isomer at l −3.5. Accordingly, there were
four sharp doublets (l 2.40, 2.43, 2.57, 2.63) in the
(D,L- and meso-cis,cis as well as the cis half of cis,trans,
each having four proton resonances of equal intensity)
was not at all obvious. Since we were able to shift the
equilibrium by crystallization (see below), we recorded
1H-NMR spectra exhibiting an intensity ratio of fulva-
lene signals attributable to cis,cis- and cis,trans-3 differ-
ent from that found in the literature. After the major
cis,trans (59%) and minor cis,cis (2×12%) isomers had
been identified on the basis of their phosphorus and
methyl signals, it became evident that the fulvalene
resonances at l 5.79, 5.19, 5.14 and 4.90 belong to the
cis half of cis,trans-3, instead of one of the cis,cis
isomers as reported [3e]. Secondly, six PMe resonances
were reported for this mixture, four of them singlets
[3e]. However, it is obvious that only five h5-rings out
of ten (Scheme 1) are magnetically non-equivalent and
each must be represented by a PMe doublet due to
1
high-field region of the 270 MHz H-NMR spectrum
attributable to PMe protons, the one at l 2.63 being
considerably broader than the others. It was therefore
assigned to both cis,cis isomers, which were quite un-
usually not separated, and appeared at lower field than
those of trans isomers. The doublet at l 2.40 was
assigned to the trans,trans isomer and the others to the
cis,trans isomer by default. This spectrum also exhibited
at least 11 multiplets in the fulvalene ‘fingerprint’ region
(l 4.96–6.14) due to multiple overlaps. Full assignment
of this spectrum to the individual isomers was aided by
a 500 MHz NMR spectrum. The total cis/trans ratio
was found to be 52:48 (48% cis,trans, 24% trans,trans
and 28% cis,cis (D,L-/meso-=1:1)).
2
pronounced J(P–H) coupling. A complete set of re-
1
vised H-NMR data is listed in Section 2.3.3.
3.2. Crystal and molecular structure of 3
Analysis of the structural data for 3 revealed that this
complex assumes the meso-cis,cis form in the crystalline
1
When the H-NMR spectrum of 3 was recorded in
DMSO-d6 solution, a considerable solvent effect on
both the chemical shift values and the isomeric ratio
was evident. Four PMe doublets were observed at high
field, the two most shielded at l 2.33 (trans,trans) and
2.35 (cis,trans) being attributable to trans isomers and
the well separated other two doublets at l 2.52
(cis,trans) and 2.57 (both D,L- and meso-cis,cis) to cis
isomers. In the fulvalene region of the 270 MHz spec-
trum (l 5.00–6.35), at least 11 individual multiplets
were exhibited due to various overlaps. Full assignment
was made on the basis of a 500 MHz NMR spectrum.
The total cis/trans ratio was found to be 40:60 (53%
cis,trans, 33% trans,trans and 14% cis,cis).
NMR spectroscopic data recorded in CDCl3 solu-
tions for 3, including a COSY experiment, have been
1
reported recently [3e]. We repeated both the H- and
31P{1H}-NMR spectroscopic measurements in CDCl3
and found several discrepancies with the published
1H-NMR results. Firstly, the literature describes an
unfortunate situation where the equilibrium consisted
of a 1:4:4 ratio of trans,trans, cis,trans and cis,cis
Fig. 2. An ORTEP plot of complex 3 demonstrating the probability
of different isomers. P, I(1), C(2), C(3): 71%; P, C(1), I(2), C(3): 20%;
P, C(1), C(2), I(3): 9%. Thermal ellipsoids are drawn at the 40%
probability level. Hydrogen and PPh2Me carbon atoms are omitted
for clarity.
(D,L-/meso-=1:1) isomers, respectively, and therefore
all fulvalene resonances are of equal intensity. Conse-
quently, no assignments were made on the basis of the

I. Ko6acs et al. / Journal of Organometallic Chemistry 584 (1999) 347–355
353
Fig. 3. Hydridic proton region of the 1H-NMR (270 MHz) spectrum of a reaction mixture of 1 with CCl4 in acetone-d6. (I) Complex 1; (II)
cis-FvW2(CO)4(PPh2Me)2HCl; (III) trans-FvW2(CO)4(PPh2Me)2HCl.
state (Fig. 2). The two W(CO)2(PPh2Me)I units are
situated in the sterically least congested anti position
with respect to the perfectly planar fulvalene ligand,
which is typical for non-metal–metal bonded fulvalene
complexes in the solid state [2a,3e], and there is a
crystallographically imposed centre of inversion in the
molecule. The ligands about each tungsten centre define
a square pyramidal structure. The bridgehead C(4)–
its concentration was achieved upon crystal formation.
These results support our initial hypothesis that the cis
conformation is preferred by phosphine-substituted
FvM2(CO)4L2X2 dihalides and the D,Llmeso equi-
librium can be shifted in the solid state, apparently in
favour of the meso form. Moreover, due to a similarity
between the isomerization behaviour of 3 and that of
recently
characterized
FvW2(CO)4(PPh2Me)2(SR)2
(R=H, alkyl) hydrosulfide and thiolate complexes [2d],
it seems reasonable to assume that the latter probably
crystallize in the same way.
,
,
C(4%) (1.45(2) A) and the W–I (2.81(9) A) distances are
,
shorter, while the CꢀO (1.16 A) and the average W–
,
Cring (2.35 A) distances are longer than in the parent
hexacarbonyl, FvW2(CO)6I2 [3e], indicating an in-
creased electron density at the tungsten centre due to
the phosphine being a better donor than CO. Interest-
3.3. Mechanism of the formation of 2 and 3
,
ingly, there is a 0.14 A ring slippage from the centre of
By carefully monitoring the hydrogen–halogen ex-
change of 1 with CCl4 and CHI3 in acetone-d6 by H-
1
each C5H4-ring. The structural analysis also determined
the position of the two solvent molecules in the lattice,
one acetone-d6 per tungsten centre.
and 31P{1H}-NMR spectroscopy, the formation of the
corresponding hydrido–halide intermediates, FvW2-
(CO)4(PPh2Me)2HX (X=Cl, I), was unambiguously
established. These intermediates were detectable in sig-
nificant concentration throughout the reactions and
slowly transformed into 2 and 3, indicating that the
hydrogen-for-halogen exchange took place in two dis-
tinctive steps. Similar stepwise hydrogen–halogen ex-
changes, which probably proceed by a radical-chain
mechanism [1], have recently been demonstrated for
Initial data processing suggested severe distortion of
one of the CꢀO ligands and that it might have been
replaced by a heavy iodine atom. Full refinement re-
vealed that the iodine atom occupied position I(1) with
71%, position I(2) with 20% and I(3) with 9% probabil-
ity (Fig. 2). Nevertheless, considering that the meso-
cis,cis form constituted only 7% of the equilibrium
mixture in the starting solution, a ca. tenfold increase in

354
I. Ko6acs et al. / Journal of Organometallic Chemistry 584 (1999) 347–355
analogous (fulvalene)chromium [2c] and molybdenum
dihydrides [2b]. The formation and persistence of
such intermediates suggest lower reactivity of the hy-
dride function of hydrido-halides than that of the
starting dihydrides. Decreased reactivity could be due
to an electronic effect of the distant halogen atom on
the hydride moiety, which would suggest efficient
electronic communication between the two halves via
the fulvalene p-system, consistent with electrochemical
studies [3d,6].
species were also detected by their phosphorus reso-
nances being different from those of 2 (l 8.2 (cis
chloride), 14.5 (trans chloride)). However, only frag-
mented information could be obtained from the PMe
and fulvalene proton regions due to excessive over-
laps with the resonances attributable to 1 and 2. The
available data indicate that trans-FvW2(CO)4-
(PPh2Me)2HCl is only a minor intermediate in this
reaction.
Fig. 4 shows nine methyl doublets clearly dis-
tributed into three groups; the PMe protons in three
different hydridic environments being the most
shielded, while those in three different cis iodide envi-
ronments being the most deshielded (D,L- and meso-
cis,cis-3 completely overlap). In addition to this, the
presence of cis- and trans-FvW2(CO)4(PPh2Me)2HI
isomers was confirmed by two new hydridic doublets
in the WH proton region, a number of new multiplets
in the fulvalene region and phosphorus resonances at
l −3.2 (cis iodide) and 10.0 (trans iodide). The two
isomers were detected in nearly equal concentrations.
When the exchange was complete, all four isomers
of 3 were present in substantial amounts and no
other species could be detected in the reaction mix-
ture. The isomeric composition did not change subse-
Each FvW2(CO)4(PPh2Me)2HX intermediate was
present in the reaction mixtures in two isomeric
forms, containing a cis and trans halide half, while
the hydride halves remained stereochemically non-
rigid on the NMR time scale. Both forms were most
conveniently identified in the high-field hydridic and
1
PMe proton regions of the H-NMR spectra, as well
as by their 31P{1H}-NMR spectra. Figs. 3 and 4 show
the WH and PMe regions of 1H-NMR spectra
recorded during the formation of 2 and 3, respec-
tively. In Fig. 3, there are two new, asymmetric dou-
blets in addition to that of 1, demonstrating the
extent of deshielding on the hydridic proton brought
about by the electron-withdrawing chlorine atom oc-
cupying cis and trans positions relative to the phos-
phine ligand in the halide half of the molecule. Both
Fig. 4. Methyl proton region of the 1H-NMR (270 MHz) spectrum of
a reaction mixture of 1 with CHI3 in acetone-d6. * cis-
FvW2(CO)4(PPh2Me)2HI; ** trans-FvW2(CO)4(PPh2Me)2HI.

I. Ko6acs et al. / Journal of Organometallic Chemistry 584 (1999) 347–355
355
quently, as long as the homogeneity of the solution
Acknowledgements
was maintained. However, when the formation of 2
was complete, only small amounts of its cis,trans and
no trans,trans isomers were present along with the
major cis,cis isomers. After several hours, when the
final equilibrium was achieved, only the cis,cis isomers
could be detected. Thus, cis,trans-2 can also be con-
sidered as an intermediate in the formation of the
end-product cis,cis-2. Interestingly, neither the
cis,trans isomer nor the trans-FvM2(CO)4L2HX inter-
mediate was previously reported in reaction mixtures
giving FvM2(CO)4L2X2 (M=Cr, Mo) which exist ex-
clusively as the cis,cis isomer [2b,c]. We believe that in
the particular reaction of 1 with CCl4 the hydrogen–
chlorine exchange must be successfully competing with
the isomerization of 1, probably because the latter is
slow. It also implies that the cis hydride isomer is
more reactive than the trans isomer, that is, the W–H
bond is weaker in the cis form. Since no trans,trans-2
was detected and the hydride function has lower reac-
tivity in the mixed intermediates, cis,trans-2 was prob-
ably generated from trans-ClW(CO)2(PPh2Me)(m-
Fv)cis-HW(CO)2(PPh2Me).
We thank the Natural Sciences and Engineering
Research Council of Canada (NSERC) and the Que-
bec Department of Education for financial support.
References
[1] T.L. Brown, in: W.C. Trogler (Ed.), Organometallic Radical Pro-
cesses, J. Organomet. Chem. Library 22, Elsevier, New York, 1990,
p. 67 and references therein.
[2] (a) M. Tilset, K.P.C. Vollhardt, R. Boese, Organometallics 13
(1994) 3146. (b) I. Kovacs, M.C. Baird, Organometallics 14 (1995)
4074. (c) I. Kovacs, M.C. Baird, Organometallics 15 (1996) 3588. (d)
I. Kovacs, A. Shaver, J. Organomet. Chem. 586 (1999) 31.
[3] (a) R. Moulton, T.W. Weidman, K.P.C. Vollhardt, A.J. Bard,
Inorg. Chem. 25 (1986) 1846. (b) I. Kovacs, M.C. Baird,
Organometallics 14 (1995) 4084. (c) C. Moreno, M.L. Marcos, M.J.
Macazaga, R.M. Medina, D.H. Farrar, J. Gonzalez-Velasco, S.
Delgado, Organometallics 17 (1998) 4657. (d) M.L. Marcos, C.
Moreno, R.M. Medina, M.-J. Macazaga, S. Delgado, J. Gonzalez-
Velasco, J. Organomet. Chem. 568 (1998) 185. (e) C. Moreno, M.-J.
Macazaga, R.-M. Medina, D.H. Farrar, S. Delgado, Organometal-
lics 17 (1998) 3733.
[4] Recent reviews relevant to fulvalene chemistry: (a) P.A. McGovern,
K.P.C. Vollhardt, Synlett (1990) 493. (b) N.J. Coville, K.E.
duPlooy, W. Pickl, Coord. Chem. Rev. 116 (1992) 1. (c) M.D.
Ward, Chem. Soc. Rev. 24 (1995) 121. (d) H. El Amouri, M.
Gruselle, Chem. Rev. 96 (1996) 1077. (e) D. Astruc, Acc. Chem.
Res. 30 (1997) 383. (f) S. Barlow, D. O’Hare, Chem. Rev. 97 (1997)
637.
4. Supplementary material
Crystallographic data for the structure in this paper
have been deposited with the Cambridge Crystallo-
graphic Data Centre as supplementary publication
No. CCDC 115338. Copies of the data can be ob-
tained, free of charge, on application to CCDC, 12
Union Road, Cambridge CB2 1EZ, UK (fax: +44
1223 336033 or e-mail: deposit@ccdc.cam.ac.uk).
[5] D.J. Meyerhoff, R. Nunlist, M. Tilset, K.P.C. Vollhardt, Magn.
Res. Chem. 24 (1986) 709.
[6] (a) T.T. Chin, W.E. Geiger, Organometallics 14 (1995) 1316. (b)
D.L. Lichtenberger, N.E. Gruhn, M.E. Rempe, W.E. Geiger, T.T.
Chin, Inorg. Chim. Acta 240 (1995) 623. (c) R. Winter, D.T. Pierce,
W.E. Geiger, T.J. Lynch, J. Chem. Soc. Chem. Commun. (1994)
1949.
.