Linking metal centres with diimido ligands: Synthesis, electronic and molecular structure and electrochemistry of organometallic ditungsten complexes [{WCl2(Ph2PMe)2(CO)}2(N-X-N)] (X = π-conjugated organic)

View Article Online / Journal Homepage / Table of Contents for this issue

Linking metal centres with diimido ligands: synthesis, electronic

and molecular structure and electrochemistry of organometallic

ditungsten complexes [{WCl₂(Ph₂PMe)₂(CO)}₂(N–X–N)]

(X

؍

ꢀ-conjugated organic)

Graeme Hogarth,* David G. Humphrey, Nikolas Kaltsoyannis, Woo-Sung Kim,

Mo-yin (Venus) Lee, Tim Norman and Simon P. Redmond

Chemistry Department and Centre for Theoretical and Computational Chemistry,

University College London, 20 Gordon Street, London, UK WC1H 0AJ

Received 19th May 1999, Accepted 21st June 1999

Tungsten() diimido-bridged complexes [{WCl₂(Ph₂PMe)₂(CO)}₂(µ-N–X–N)] have been prepared via oxidative

addition of diisocyanates to two equivalents of [WCl₂(Ph₂PMe)₄]. para-Substituted monoimido complexes

᎐

[WCl (Ph PMe) (CO)(NC H X-p)] (X = I, Br or C᎐CPh) have also been prepared but attempts to couple the X = I

᎐

2

6

4

complex as a route to diimido-bridged complexes were unsuccessful. All complexes are air-stable crystalline solids

and ﬁve diimido (N–X–N = p-NC₆H₄N, p-N-o-MeC₆H₃N, p-N(o-MeOC₆H₃C₆H₃OMe-o)N, 1,5-NC₁₀H₆N or

m-NC₆H₄N) and one monoimido complex (X = I) have been characterised crystallographically. All show the same

gross structural features, namely a trans arrangement of phosphines and cis chlorides. The central aryl ring generally

lies approximately in the Cl₂(CO) plane (torsional angles 4.1–26.1Њ) except for one complex in which the ring lies

almost perpendicular to this (torsional angle 80.2Њ). A series of density functional calculations conducted on

model mono- and di-imido tungsten-() and -() compounds indicated that the most stable aryl ring orientation is

perpendicular to the plane containing the trans phosphines, i.e. as found in all cases except one (N–X–N = p-NC₆H₄N).

The anomaly in the latter may be due to cocrystallisation with chlorobenzene. In order to assess the degree of

communication between the tungsten() centres through the highly π-conjugated diimido linkages, electrochemical

studies have been carried out. All diimido-bridged complexes show two closely spaced oxidative processes at low

temperature indicative of weak electronic communication. The reductive chemistry of the para-phenylene bridged

complexes is diﬀerent from other diimido complexes, displaying two closely spaced reductive processes. Spectro-

electrochemical studies have also been carried out on N–X–N = p-NC₆H₄N, oxidation at ϩ1.2 V leading to CO

loss. In order to gain further insight into the nature of the electronic communication between metal centres density

functional calculations were carried out and were generally in agreement with the electrochemical results, suggesting

that there is at best a weak interaction between the metal centres in these π-conjugated diimido-bridged complexes.

via a conjugated unsaturated organic moiety should allow them

to communicate with one another. To date only a few examples

Introduction

Developing, understanding and exploiting electronic communi-

cation between metal centres is a continuing area of interest

to both the synthetic and theoretical chemist¹and since the

pioneering days of Creutz and Taube²many exciting and novel

developments have been made. The focus of most synthetic

studies in this area are complexes in which the metals are linked

via a π-conjugated organic backbone. For example, Gladysz

and co-workers³and others⁴have shown that when metal

centres are linked via π-conjugated acetylide based groups a

high degree of electronic communication between them can be

achieved. Since multiple bonding between metal atoms and

main group elements such as carbon and nitrogen is highly

delocalised in nature⁵this allows for strong conjugation

between the metal and heteroatom, and related studies have

focused on the use of metal complexes with multiply bonded

ligands such as carbon-based alkylidyne⁶and alkylidene⁷

complexes.

The binding of imido ligands to transition metal centres is

well known and occurs for metals in a variety of oxidation

states, although high-valent centres in which the metal has a

d⁰–d²electronic conﬁguration are particularly stable.⁵Such

ligation is particularly desirable for the design of π-conjugated

systems since strong interactions occur between the metal

d orbitals and p orbitals on nitrogen. This interaction is

well developed such that the imido ligand is a strong π-donor

moiety.⁸Joining metal centres with imido functionalities linked

of this type of ligation have been reported. Thus, tungsten,^9–11

molybdenum,^12–14rhenium,^15,16niobium,¹⁷vanadium¹⁸and

uranium¹⁹centres have been linked via a para-phenylene

diimido ligand, however other π-conjugated organic backbones

have only been utilised to couple dirhodium centres.²⁰We

recently reported the facile synthesis of two such phenylene

linked tungsten() complexes, namely the isomeric pair

[{WCl₂(Ph₂PMe)₂(CO)}₂(N–X–N)] (X = p-C₆H₄, or m-C₆H₄).⁹

Since such complexes are easily prepared in high yield via

oxidative addition of diisocyanates to [WCl₂(Ph₂PMe)₄]^21,22we

decided to exploit this synthetic procedure in order to prepare

a range of such complexes and assess the degree of com-

munication between the metal centres as a function of the

π-conjugated organic linking unit. Herein we report the syn-

thesis of a wide range of such complexes together with electro-

chemical studies and molecular orbital calculations in an

attempt to assess and understand the size and nature of

the metal–metal interaction. Further, we have attempted to

prepare other linked complexes via coupling of monoimido

fragments. A preliminary account of an aspect of this work

has appeared.⁹

Results

During the course of these studies a number of amines I, iso-

cyanates II and tungsten() imido complexes 1 were utilised.

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

2705

View Article Online

N

a

c

d

b

e

N

MeO

N

OMe

f

g

N

i

h

N

C

N

k

l

j

N

C

N

m

I

n

o

N

Br

N

C

r

p

q

Chart 1

Chart 1 shows the backbones of these (together with the

labelling which is used throughout), while Scheme 1 shows the

general reaction conditions employed.

the ethynyl moieties. As expected, no such absorption was seen

for Il. para-Aminodiphenylethyne Ir was prepared in a similar

manner upon coupling of para-iodoaniline and phenylethyne;

conditions being optimised to give a 72% yield.

While a few of the diisocyanates were commercially available,

in order to extend the scope of the diimido linkage reaction it

proved necessary to prepare a number of new such compounds

(Scheme 1). Conversion of the diamines into diisocyanates was

achieved using triphosgene (COCl₂)₃.²⁵Yields varied enor-

mously (17–90%), low amounts normally being associated with

poor solubility of the diamine and low volatility of the diiso-

cyanate, the latter making puriﬁcation by sublimation diﬃcult.

Most diisocyanates were white solids and all were characterised

by bands between 2295 and 2250 cm^Ϫ1in the IR spectrum

associated with the NCO group. With ethynylamines an excess

of triethylamine was necessary in order to prevent attack of the

released HCl at the carbon–carbon triple bond. para-Iodo-

phenyl isocyanate IIp could not be prepared from para-

iodoaniline Ip and triphosgene, the former being recovered even

under forcing conditions. Use of phosgene in toluene did give

IIp (55%) being easily puriﬁed by sublimation. Halide deactiv-

ation of anilines towards phosgene attack and isocyanate

formation is not noted in many current texts, despite being

made clear in the early literature.²⁶

(i)

H₂N

NH₂

OCN

NCO

II

I

(ii)

PPh₂Me

CO

PPh₂Me

Cl

W

N

OC

W

Cl

PPh₂Me

1

Scheme 1 (i) (COCl₂)₃, 180 ЊC; (ii) 2[WCl₂(Ph₂PMe)₄], RT, Ϫ2Ph₂-

PMe.

(i) Synthesis of amines and isocyanates

The majority of the amines and diamines used in this study

were obtained commercially, however those containing the

ethynyl group were prepared (Scheme 2). The alkyne para-

H₂N

(i)

C

NH₂

I

(ii) Synthesis and characterisation of tungsten(IV) imido

complexes

H

C

NH₂

(ii)

Room temperature reaction of 2 equivalents of [WCl₂(Ph₂-

PMe)₄] and diisocyanates II resulted in the slow formation

(2–48 h) of the diimido-bridged tungsten() complexes 1a–1o

in yields of 40–93%, while analogous monoimido complexes

1p–1r were prepared similarly. Reactions were easily monitored

by a progressive change from orange to either green or purple

and also by IR spectroscopy, the isocyanate band at around

2200–2150 cm^Ϫ1disappearing to be replaced by a band assigned

to the metal-bound carbonyl at around 1975–1955 cm^Ϫ1. High

molecular weight isocyanates were not very soluble in toluene

and these reactions could also be followed by their slow dis-

solution. Yields were generally quantitative by ³¹P NMR spec-

troscopy, with no evidence of other products. Isolation was

generally straightforward being achieved by precipitation from

the toluene solution (after concentration) and washing with

H₂N

C

NH₂

II

Scheme 2 (i) para-Iodoaniline, [PdCl₂(PPh₃)₂]/CuI; (ii) 0.5 para-

iodobenzene, [PdCl₂(PPh₃)]/CuI.

ethynylaniline²³proved to be a pivotal intermediate in the syn-

thesis of diamines Il and Im. Palladium-catalysed coupling

reactions were carried out using the standard literature pro-

cedures²⁴and gave moderate unoptimised yields (Il, 28; Im,

55%). Both showed NH stretches in the IR spectrum, while a

band at 2206 cm^Ϫ1of Im is assigned to the asymmetric mode of

2706

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

View Article Online

light petroleum to remove released phosphine. All tungsten

imido complexes 1 are stable in air, and recrystallisation upon

slow diﬀusion of methanol into a saturated dichloromethane

solution gave green or purple crystalline solids. In a few

instances released phosphine proved diﬃcult to remove from

the small amounts of tungsten complex formed and satisfac-

tory analytical results could not be obtained.

suggested a complex reaction mixture and thus this strategy was

not pursued further.

(iii) X-Ray crystallographic studies

Single-crystal X-ray crystallographic studies were carried out

on six of the tungsten() complexes, namely diimido-bridged

1a, 1b, 1g, 1h and 1n, and mono-imido 1p. All were crystallised

upon slow diﬀusion of methanol into saturated dichloro-

methane solutions, except 1a which was crystallised upon slow

diﬀusion of methanol into a saturated chlorobenzene solution.

Selected bond lengths and angles for these and related tungsten

imido complexes are given in Table 1, while views of the

molecules are shown in Figs. 1–6. For both 1g and 1h the

asymmetric unit contains two half molecules and there are no

intermolecular contacts between them. Further, 1a contains a

disordered molecule of chlorobenzene in the asymmetric unit

while 1h contains a molecule of dichloromethane. Again there

are no intermolecular contacts. In all cases except for 1g all

non-hydrogen atoms were reﬁned anisotropically. For 1g a

number of atoms were disordered over two sites (50% occupan-

cies) and these and the phenyl rings were reﬁned only isotropic-

ally. The representations of the molecules of 1g shown in Fig. 3

show the average of the disordered sites. In 1b (Fig. 2) the

methyl carbon on the aryl ring, C(5), is disordered over two

sites. It is given only a 50% occupancy as the inversion centre

generates the second carbon atom. This renders crystallo-

graphically equivalent the two non-equivalent tungsten centres

in the molecule.

All molecules contain an approximately octahedral tung-

sten() centre(s), the co-ordination sphere consisting of trans-

phosphines [P–W–P 166.4 2.7Њ], cis-chlorides [Cl–W–Cl

88.9 2.1Њ] and a cis arrangement of imido and carbonyl

ligands [N–W–C 89.9 2.9Њ]. Tungsten–phosphorus bonds are

within the expected range [W–P 2.529 0.021 Å], while the

tungsten–chlorine bond trans to the imido [W–Cl 2.462 0.016

Å] is not signiﬁcantly diﬀerent to that trans to the carbonyl

[W–Cl 2.458 0.013 Å], the absence of a trans inﬂuence at the

tungsten() centre being expected.²⁹The tungsten–carbon

bond to the carbonyl is as expected [1.144 0.014 Å] and the

W–C–O vector is essentially linear. Since most of the molecules

contain an inversion centre, then the carbonyls on diﬀerent

tungsten centres are constrained to lie on opposite sides of the

molecule. This is not the case with meta complex 1n, the asym-

metric unit containing a full molecule with carbonyls lying on

the same side of the molecule. The short tungsten–nitrogen

distance [W–N 1.755 0.012 Å] is within the range expected

from studies on other tungsten imido complexes^5,8and the near

linearity at nitrogen [W–N–C 171.1 2.2Њ] suggests that it is

acting as a 4-electron donor to the tungsten centre.

The most striking feature of these structures is the relative

orientation of the aryl ring(s) of the imido ligands to the tung-

sten centre as illustrated in Figs. 1–6. This is represented in

Table 1 by a torsion angle, which is that between the plane of

the aryl ring(s) and the WCl₂(CO) plane, ranging between 4.1Њ

in 1p to 80.2Њ in 1a. Thus, in the latter the phenylene ring lies

approximately perpendicular to this plane lying almost parallel

to the WP₂(CO) plane. This situation is, however, the only one

of its type. Thus, for all other diimido-bridged molecules tor-

sion angles vary between 5.8 and 26.1Њ, that is the aryl ring(s) lie

approximately in the WCl₂(CO) plane. It is noteworthy that

there are no intermediate values i.e. 35–75Њ, that is the aryl

ring(s) either lie close to one plane or the other. The two

independent molecules in the asymmetric units of 1g and 1h

vary by virtue of the diﬀerent torsional angles they adopt; 1g

(A, 26.1; B, 17.3Њ) and 1h (A, 10.7; B, 6.2Њ). Otherwise these

independent molecules of 1g and 1h are very similar, minor

diﬀerences being the carbonyl C–O distances in 1g [1.139(13)

and 1.153(13) Å] and nitrogen–carbon [1.44(2) and 1.41(2) Å]

and tungsten–nitrogen [1.746(11) and 1.757(13) Å] in 1h.

Characterisation was generally straightforward being based

1

on elemental analysis, IR, H and ³¹P NMR spectroscopy and

positive-ion FAB mass spectra. In the IR spectra the metal-

bound carbonyl appears as a broad band at 1975–1955 cm^Ϫ1,

conﬁrming the oxidative addition of the isocyanate. All

positive-ion FAB spectra showed a peak corresponding to the

loss of a single carbonyl group, together with other peaks due

to sequential loss of the second carbonyl and the phosphines.

The ³¹P NMR spectra generally consist of a single resonance at

around δ 3, indicating the equivalence of all phosphorus atoms,

with satellites due to coupling to tungsten (J_PW= 290–292 Hz).

For asymmetric complexes 1b and 1o two phosphorus reson-

ances were observed in both cases, being separated by about 1

ppm. Proton NMR spectra showed complex signals in the aro-

matic region of the spectrum due to the phenyl groups of the

phosphine ligands. Generally, the aryl protons of the imido

ligands appeared further upﬁeld of these, but in a few cases

it was diﬃcult to deconvolute resonances associated with the

arylimido group and the phenyl rings on the phosphines. For

example, in the 3,3Ј-dimethoxybenzidene complex 1g the

arylimido protons appeared between δ 6.57 and 6.53 as two

doublets (J 8.4 Hz) and a singlet, while the phenyl protons

appeared as a multiplet between δ 7.69 and 7.24. Further, all

spectra also showed a triplet at around δ 2.3 assigned to the

phosphorus-bound methyl group. All ¹³C NMR spectra showed

a low-ﬁeld triplet at around δ 242 assigned to the carbonyl

ligand, a complex set of peaks between δ 155 and 105 due to

aromatics and a triplet at around δ 13 assigned to the methyl

groups of the phosphines.

Reaction of compound IIa with one equivalent of [WCl₂-

(Ph₂PMe)₄] resulted only in the formation of 1a and unchanged

tungsten() complex with no evidence for an intermediate

mononuclear complex, the latter being the case in all other

reactions except one. This exception was the reaction with 1,5-

diisocyanatonaphthalene IIh. Here, long reaction times were

needed as the isocyanate was only sparingly soluble in toluene.

Monitoring the reaction by IR spectroscopy showed the

expected decrease in intensity of the isocyanate stretch (2270

cm^Ϫ1), together with the anticipated growth of the new metal–

carbonyl absorption (1964 cm^Ϫ1). However, a further absorp-

tion at 2094 cm^Ϫ1also grew in concomitant with that of the

tungsten carbonyl which latter decreased in intensity. In an

attempt to isolate this intermediate, addition of a 1:1 ratio of

reactants was carried out but again only 1h (albeit in lower

yields) and [WCl₂(Ph₂PMe)₄] were isolated after work-up.

In a second approach towards the synthesis of tungsten

bridged diimido complexes, we attempted to prepare a biphenyl

diimido complex via the Ullman coupling^27,28of the mono-

nuclear iodo-complex 1p. Dissolution of 1p in dimethylform-

amide followed by treatment with activated copper at high

temperature resulted in the formation of a dark solution from

which a brown intractable oil was isolated after removal of pre-

cipitated copper iodide. Proton NMR spectroscopy indicated

the presence of some unchanged 1p, together with other

unidentiﬁed species, while in the ³¹P NMR spectrum a number

of species were also observed. The mass spectrum showed

peaks centred at m/z 1565 and 1173 which could be due to a

ditungsten complex, but the IR spectrum showed the complete

absence of any carbonyl absorptions. Clearly, if coupling is

taking place, then it is not clean and other secondary processes

are also operating. Repeating the reaction using pyridine as the

solvent gave a similar result, although a carbonyl resonance was

apparent in the IR spectrum. Here again, however, NMR data

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

2707

View Article Online

Fig. 1 Molecular structure of complex 1a; (a) full molecule, (b) end-on (phenyl rings omitted for clarity).

Further, a number of the bond angles around tungsten vary

W(1)Cl₂(CO) and W(2)Cl₂(CO) planes respectively. Such a

in 1h, representative examples being the Cl–W–Cl [90.9(2)

and 86.7(2)Њ] and P–W–P [163.65(12) and 169.12(13)Њ] angles.

Generally, however, little conformational change is required

about the tungsten centre for the diimido-linkage to take up

diﬀerent conformations.

While diﬀerences in individual bond lengths and angles with-

in isomeric 1a and 1n are not signiﬁcant, the gross structures are

quite diﬀerent. In para-1a the tungsten centres, related by the

crystallographic inversion centre, are necessarily eclipsed; in

meta-1n each metal centre is crystallographically unique being

twisted by approximately 7.4Њ with respect to one another. It is

probable that this arises to reduce adverse steric interactions

between the phenyl rings of the phosphine ligands on adjacent

metal centres, being minimised by the twist. Thus, the central

aryl ring makes torsional angles of 5.8 and 23.9 to the

twist renders the two metal centres crystallographically in-

equivalent. However, in solution they are equivalent suggest-

ing that the energy barrier to this twist about the W–N

bonds is small. Again, it should be noted that this has little

eﬀect on the conformation at the metal centre with maximum

diﬀerences in bond angles and lengths at the two independent

tungsten centres being within the 3σ range, that is not

signiﬁcant.

The mononuclear para-iodoimido complex 1p (Fig. 6) is

closely related to the para-tolyl complex [WCl₂(Ph₂PMe)₂(CO)-

(NC₆H₄Me-p)] 1t structurally characterised by Mayer and co-

workers^22a(Table 1). Again, there are only small diﬀerences

between these two molecules which are characterised by torsion

angles of 4.1 (1p, X = I) and 4.3Њ (X = Me), and suggesting

that exchange of electron-releasing and electron-withdrawing

2708

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

View Article Online

Fig. 2 Molecular structure of complex 1b as in Fig. 1. Showing the Me group of the central ring disordered over two sites, C(5) and C(5a).

groups at the para position has little eﬀect upon the binding of

(CO)(NPh)] 1sЈ, [WCl₃(PH₃)₂(NPh)] 3Ј, [{WCl₂(PH₃)₂(CO)}₂-

the imido ligand or the tungsten centre in general.

(µ-p-NC₆H₄)] 1aЈ and [{WCl₃(PH₃)₂}₂(µ-p-NC₆H₄)] 2Ј, in which

the substituted phosphine ligands of the actual compounds are

replaced by PH₃groups. X-Ray crystallography has shown that

in the unsubstituted phosphine analogues of both of the

mononuclear systems 3Ј and 1sЈ and the dinuclear compound 2Ј

the aryl ring lies close to the Cl₃/Cl₂(CO) plane.^11,22,30

The calculations reveal the anticipated electronic structures,

with a single electron in the M–Cl π antibonding highest occu-

pied molecular orbital (HOMO) of the formally tungsten() 3Ј

and two electrons in the equivalent MO of 1sЈ (W–Cl anti-

bonding/W–CO back bonding). The ditungsten compounds 2Ј

and 1aЈ have twice the number of metal-based electrons as 3Ј

and 1sЈ respectively. Compound 1aЈ is predicted to have no

unpaired electrons, while the two metal-based electrons of 2Ј

give rise to a high spin triplet state 27.5 kJ mol^Ϫ1more stable

than the spin paired singlet.

(vi) Electronic structure calculations: central ring rotation

As discussed above, a striking feature of the crystal structures

of complexes 1a, 1b, 1g, 1h, 1n and 1p is the relative orientation

of the aryl ring(s) of the imido ligands and the tungsten centre.

In 1a the aryl ring is almost coplanar with the two phosphorus

atoms and the tungsten centre, in contrast to other compounds

in which the ring is rotated by virtually 90Њ to this plane.

Further, this ring orientation is also found in [WCl₂(Ph₂PMe)₂-

(CO)(NC₆H₄Me-p)] 1t^22aand the tungsten() diimido-bridged

complex [{WCl₃(PMe₂Ph)₂}₂(µ-p-NC₆H₄N)] 2.¹¹In order to

establish if there is an electronic explanation for the atypical

geometry of 1a, we have conducted a series of density func-

tional calculations on the model compounds [WCl₂(PH₃)₂-

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

2709

View Article Online

Fig. 3 Molecular structure of complex 1g; (a) full picture of molecule A, (b) molecule A end-on (phenyl rings omitted for clarity), (c) molecule B

end-on (phenyl rings omitted for clarity).

For each of the molecules the aryl ring was rotated from

coplanarity with the Cl₃/Cl₂(CO) groups to perpendicular in

steps of 10Њ, and Fig. 7 presents the variation in the total

molecular binding energies (kJ mol^Ϫ1) as this distortion is per-

formed. Our calculations clearly indicate that the most stable

orientation of the aryl ring is similar in all four cases, i.e. that

the preferred geometry has the ring almost coplanar with the

Cl₃/Cl₂(CO) groups. This is in agreement with the crystal struc-

tures of the substituted phosphine complexes 1s, 2 and 3, but

contrary to 1a. We therefore conclude that the orientation of

the central aryl ring in 1a does not have an intramolecular

cause, and arises as a result of external eﬀects, probably crystal

packing forces. In this respect it is noteworthy that 1a is co-

crystallised with chlorobenzene and it may be the interaction

between the cocrystallised molecules which leads to 1a adopt-

ing an unfavourable central ring geometry. The relatively small

energy diﬀerences between the least stable (80–90Њ ring rotation

vs. Cl₃/Cl₂(CO) planes) and the most stable (10–20Њ) ring

orientation is consistent with this suggestion.

bonding energy, steric repulsion and orbital interaction (elec-

tronic component) are all arbitrarily set to zero at the geometry

in which the ring is in the plane of the three chlorine atoms (0Њ

rotation), and their values at the other orientations are plotted

relative to this point. It is immediately clear that the small

diﬀerences in the total bonding energy from 0 to 90Њ arise

from the virtual self-cancellation of the steric and electronic

components. Up to a ring rotation of 50Њ neither of these

components is greatly diﬀerent from their value at 0Њ, but this

changes signiﬁcantly as the ring approaches coplanarity with

the phosphine groups. The latter geometry is clearly favoured

on electronic grounds but disfavoured sterically, and it is the

steric repulsion that causes the 0Њ geometry to be favoured over

the 90Њ.

(v) Electrochemical and spectroelectrochemical studies

In order to assess the degree of communication between metal

centres in diimido-bridged complexes, the solution redox

behaviour of ditungsten complexes 1a, 1b, 1g and 1n, 1o and

monomeric 1s has been examined by cyclic voltammetry, diﬀer-

ential pulse voltammetry and rotating disk voltammetry. At

room temperature 0.5 mol dm^Ϫ3[ⁿBu₄N][PF₆]–dichloromethane

solutions of the ditungsten complexes each display an oxidation

The Amsterdam Density Functional (ADF) package allows

the total bonding energy to be broken down into steric

and electronic components using the transition state pro-

cedure developed by Ziegler and others.^31–33Fig. 8 presents the

results of such an analysis for [WCl₃(PH₃)₂(NPh)] 3Ј. The total

2710

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

View Article Online

Fig. 4 Molecular structure of complex 1h as in Fig. 3.

and a reduction, both of which are irreversible at scan rates up

shoulder on the leading edge of the forward wave, a feature

to 1000 mV s^Ϫ1. Similar behaviour is found with other solvents,

e.g. 1,2-dichlorobenzene and propylene carbonate. Upon cool-

ing the dichloromethane solutions to 213 K the oxidative pro-

cesses for all ditungsten complexes and 1s become chemically

reversible thus permitting more detailed analysis of the voltam-

metric response. The reductive processes remain irreversible

at low temperature for all but two of the complexes, namely

1a, 1b. For those, cooling of the solutions leads to the appear-

ance of two waves which have some degree of chemical revers-

ibility. Further discussion is limited to the results obtained at

low temperature and these are summarised in Table 2.

Low temperature cyclic voltammograms of complex 1a dis-

solved in 0.5 mol dm^Ϫ3[ⁿBu₄N][PF₆]–dichloromethane are

shown in Fig. 9. The scan to positive potentials reveals an oxid-

ative response near ϩ1.0 V, on which there is evidence for split-

ting of both the forward and return waves by the presence of a

mirrored in the return wave. By comparing the limiting current

in a steady state voltammogram (obtained with a rotating disk

electrode) with that for the oxidation of the monomeric com-

plex 1s, it has been established that the oxidation of 1a (and all

other bimetallic complexes) overall involves two electrons. The

oxidative response is consistent with two closely spaced, one-

electron oxidative processes, both of which have a high degree

of chemical reversibility. Variable scan rate studies indicate that

these processes are quasi-reversible in the electrochemical sense,

in that the separation of the peaks in the forward and reverse

scan increases with increasing scan rate. The small separation

of the individual couples does not permit evaluation of peak-

to-peak separation (∆E_p) for each individual one-electron

process. Similarly, the limiting currents for the oxidation and

reduction, which is clearly resolved into two sequential one-

electron processes (see later), are equal, such that the oxidation

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

2711

View Article Online

Fig. 5 Molecular structure of complex 1n as in Fig. 1.

and reduction of the bimetallic complexes overall involves the

bridged complexes 1a, 1b display two well resolved reductions in

same number of electrons. The results from diﬀerential pulse

voltammetry are also consistent with the existence of two

closely spaced oxidative waves, as shown by the asymmetry in

the peak maximum. The individual oxidation potentials for the

36/35e and 35/34e couples have been estimated from diﬀeren-

tial pulse voltammograms, using the procedure of Taube and

Richardson³⁴and are listed in Table 2. The separation of the

couples in 1n is somewhat greater than that in isomeric 1a as

is evident from both the cyclic and diﬀerential pulse voltam-

mograms, shown in Fig. 10. The oxidative response for the

other bimetallic complexes are in general similar, with the

exception of 1o, which displays two couples that are clearly

resolved.

the region Ϫ1.1 to Ϫ1.3 V, the ﬁrst couple being more reversible

than the second in the chemical sense. For the remainder of the

complexes only a broad irreversible reduction in the region of

Ϫ1.5 V is found. Steady-state voltammetry indicates that over-

all reduction of the bimetallic complexes involves two electrons.

In situ infrared spectroelectrochemical (IRSEC) studies have

been conducted on complex 1a, since its voltammetry oﬀered

the greatest promise of observing chemically reversible redox

processes on the timescale of the IRSEC experiment. The oxid-

ation of 1a at ϩ1.2 V in an infrared reﬂection absorption

spectroscopic (IRRAS) cell at low temperature resulted in the

collapse of the ν(CO) band due to the neutral complex at 1969

cm^Ϫ1, and the growth of a weak band at 2139 cm^Ϫ1which can

be attributed to the formation of free CO. No other bands were

observed that can be attributed to the formation of new tung-

sten carbonyl complexes. Returning the potential of the work-

Whilst the oxidative chemistry of the diimido-bridged

ditungsten complexes is in general quite similar, the reductive

chemistry is considerably diﬀerent. At 213 K the phenylene-

2712

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

View Article Online

Fig. 6 Molecular structure of complex 1p.

Fig. 7 Walsh diagrams for rotation of the aryl ring in (a) [WCl₂(PH₃)₂(CO)(NPh)] 1sЈ, (b) [WCl₃(PH₃)₂(NPh)] 3Ј, (c) [{WCl₂(PH₃)₂(CO)}₂-

(µ-p-NC₆H₄)] 1aЈ, (d) [{WCl₃(PH₃)₂}₂(µ-p-C₆H₄)] 2Ј.

ing electrode to its initial rest value, i.e. 0.0 V, did not regenerate

any of the neutral parent complex. Although the oxidation

of 1a appears chemically reversible on the timescale of the

voltammetric experiment, on the timescale of an in situ infrared

spectroelectrochemical experiment this process is chemically

irreversible, with the decomposition of [1a]^ϩ(or [1a]^2ϩ) occur-

ring through loss of CO. The species produced upon reduction

of 1a is equally unstable, in that the reduction is also chemically

irreversible. Reduction at Ϫ1.1 V led to the collapse of the

ν(CO) band and the growth of bands at 1729, 1685, 1627, 1581

and 1518 cm^Ϫ1. The higher wavenumber bands are more typical

of bridging carbonyls, although the sharpness of some of the

observed bands may indicate the formation of ketonic organic

species. Given the instability of the redox products from 1a,

spectroelectrochemical studies of the other complexes were not

undertaken.

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

2713

View Article Online

2714

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

View Article Online

Table 2 Summary of electrochemical data

E^a_1/2/V

Complex

W^IV–W^V

W^V–W^V

W^IV–W^III

W^III–W^III

1a

ϩ1.07

ϩ1.13

Ϫ1.08

Ϫ1.18

Ϫ1.20

Ϫ1.30

1b

ϩ1.00

ϩ1.02

ϩ1.05

ϩ1.02

ϩ1.09

ϩ1.08

ϩ1.16

ϩ1.19

1g^b

1n^b

1o^b

1s^b,c

≈Ϫ1.6

≈Ϫ1.5

≈Ϫ1.6

≈Ϫ1.7

ϩ1.00

^aRecorded in 0.5 mol dm^Ϫ3[ⁿBu₄N][PF₆]–dichloromethane at 213 K vs.

Ag–AgCl against which ferrocenium–ferrocene is measured at ϩ0.55 V.

The E_1/2potentials are estimated from diﬀerential pulse voltammo-

grams, where E_1/2= E_p(MA/2) (MA = modulation amplitude),

assuming that the couples are reversible. For the W^IV–W^Vand W^V–W^V

couples the individual E_1/2potentials were estimated by the procedure

of Taube and Richardson.^{34 b}Irreversible, E_pcquoted. ^cThe mono-

meric complex 1s displays one-electron processes; E_1/2= (E_paϩ E_pc)/2

taken from cyclic voltammetry.

Fig. 10 Cyclic voltammetry (a,c) and diﬀerential pulse voltammetry

(b,d) for complex 1n at 213 K in dichloromethane.

conducted a series of density functional calculations on models

for 1a, 1b and 1o in which the Ph₂PMe ligands are replaced by

PH₃groups. In these calculations the atomic positions were

taken from X-ray crystallographic data and idealised to the

highest possible symmetry. Generally, this enforces symmetry

equivalence of the metal atoms, thereby imposing delocalised

bonding. This is anticipated to be a realistic approach to our

studies, in that previous work, for example by McGrady et al.,³⁵

has shown that there is an increased tendency toward delocalis-

ation with increasing atomic number in bimetallic systems such

as [M₂Cl₉]^3Ϫ(M = Cr, Mo or W) and [M₂Cl₉]^Ϫ(M = Mn, Tc or

Re).

Although there is not necessarily direct correlation between

the calculated electronic structures of isolated molecules and

solution electrochemistry data, it was our hope that analysis

of the compositions and energies of the highest occupied and

lowest unoccupied molecular orbitals (MOs) of the model

compounds would aid our understanding of the experimental

results. The electrochemical data reveal that all of the com-

pounds studied display similar oxidation behaviour. Our calcu-

lations are consistent with this, in that the composition and

energies of the highest occupied molecular orbital, from which

the electrons are presumably removed in the oxidation process,

are very similar in each of the model compounds. Fig. 11 pres-

ents plots of the highest occupied and lowest unoccupied MOs

of 1aЈ and 1oЈ, from which it may be seen that the HOMOs are

W–C bonding but C–O and W–Cl antibonding.† The W–C

bonding character of the HOMO is consistent with the spectro-

electrochemical observation that free CO is generated upon

oxidation of 1a. Removal of electron(s) from the HOMO will

result in reduced W–C bonding, potentially leading to expul-

sion of CO from the complex.

Fig. 8 Total binding energy, steric repulsion and orbital interaction

upon phenyl ring rotation in complex 3Ј.

Compounds 1a and 1b display reductive electrochemistry

diﬀerent from that of the other compounds studied. Again,

this is consistent with our calculations, which reveal that the

lowest unoccupied molecular orbital (LUMO) composition and

energy relative to the HOMO is diﬀerent in 1aЈ and 1bЈ in com-

parison with the other molecules. Thus, while the LUMO in 1aЈ

and 1bЈ is 1.01 eV less stable than the HOMO, the HOMO–

LUMO gap in 1oЈ is 1.43 eV, consistent with the greater separ-

ation between the oxidation and reduction waves in 1o. Fig. 11

clearly demonstrates that the composition of the LUMOs in 1aЈ

and 1oЈ are diﬀerent from one another, with no CO contribu-

tion to the LUMO of 1oЈ and no C–C π interaction in the

LUMO of 1aЈ.

Fig. 9 Cyclic voltammetry (a,c) and diﬀerential pulse voltammetry

(b,d) for complex 1a at 213 K in dichloromethane.

(v) Electronic structure calculations: probing redox potentials

† Note that the higher symmetry (C_2h) of complex 1aЈ forces equivalent

contributions from both metal centres, while the HOMO and LUMO

of the C_ssymmetric 1oЈ are concentrated on diﬀerent ends of the

molecule. The bonding characteristics of the HOMOs and LUMOs

of both compounds are, however, very similar.

In order to probe further the electronic structure of diimido-

bridged complexes and to gain additional insight into the

electrochemical and spectroelectrochemical studies, we have

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

2715

View Article Online

Fig. 11 MOLDEN plots of the highest occupied molecular orbital and lowest unoccupied molecular orbital for (a) complex 1aЈ (HOMO), (b) 1aЈ

(LUMO), (c) 1oЈ (HOMO), (d) 1oЈ (LUMO).

The spectroelectrochemical behaviour of complex 1a under

reduction is more complicated than under oxidation, with the

growth of several new bands and the collapse of ν(CO).

Although our calculations do not fully explain this behaviour, it

is noticeable that the LUMO of 1aЈ is strongly W–N antibond-

ing, and hence population of this orbital may be expected to

result in the breaking of the W–N bonds. This is consistent with

the observation that 1a is unstable with respect to reduction in

the spectroelectrochemical experiment.

related ketone addition to I Bryan and Mayer³⁶postulated that

reactions proceed via a side-bound isocyanate, which sub-

sequently undergoes oxidative addition. It is diﬃcult to see how

co-ordination of the tungsten() centre can substantially eﬀect

the binding constant of the isocyanate. However, it may aﬀect

the rate of oxidative addition by virtue of weakening the

nitrogen–carbon and strengthening the carbon–oxygen bond.

Reaction of 1,5-diisocyanatonaphthalene IIh with com-

pound I proceeded somewhat diﬀerently to other diisocyanates.

When followed by IR spectroscopy the isocyanate stretch at

2270 cm^Ϫ1decreased with the growth of two new absorptions;

at 1964 cm^Ϫ1associated with the metal carbonyl moiety in 1h

and an unidentiﬁed band at 2094 cm^Ϫ1, which later disappeared.

Attempts to isolate this intermediate were unsuccessful but it

is tempting to speculate that it is the result of a side-bound

isocyanate and indeed such species are known.³⁷Again,

the unusual behaviour of the 1,5-naphthalene linking unit

seen here mirrors that found in the molybdenum() diamine

chemistry. Thus, deactivation of the second amine upon co-

ordination of the ﬁrst is not pronounced and allows the

synthesis and characterisation of both monoimido [MoCl₂-

(S₂CNEt₂)₂(1,5-NC₁₀H₆NH₂)] and diimido-bridged [{MoCl₂-

(S₂CNEt₂)₂}₂(1,5-NC₁₀H₆N)] products.¹⁴

Discussion

In a manner analogous to that developed by Mayer and co-

workers,²²oxidative addition of a range of π-conjugated diiso-

cyanates IIa–IIo to two equivalents of [WCl₂(Ph₂PMe)₄]

aﬀorded a range of diimido-bridged tungsten() complexes

[{WCl₂(Ph₂PMe)₂(CO)}₂(µ-N–X–N)] 1a–1o in good to moder-

ate yield. Attempts to add a single tungsten centre to para-

phenylene diisocyanate were unsuccessful resulting only in a

50:50 mixture of the double addition product 1a and un-

changed isocyanate. This suggests that co-ordination of one

tungsten() group to the isocyanate activates the second NCO

group towards oxidative addition and provides some evidence

of communication through the π-conjugated backbone. In

related work concerning attempts to link molybdenum()

centres together via reaction of [MoOCl₂(S₂CNEt₂)₂] with para-

phenylene diamines we found that co-ordination of the one

molybdenum() centre deactivated the second towards further

metal co-ordination.¹⁴Both results provide evidence for com-

munication through the π-conjugated framework, the deactiv-

ation of the amine resulting from delocalisation of the lone pair

towards the electron-deﬁcient molybdenum() centre. The

nature of the activation of the isocyanate by the tungsten()

centre is less clear. On the basis primarily of observations in

A second synthetic strategy towards diimido-bridged com-

plexes via the attempted Ullman coupling of the mononuclear

iodo-complex [WCl₂(Ph₂PMe)₂(CO)(p-NC₆H₄I)] 1p was unsuc-

cessful as was the attempted coupling of 1p to para-amino-

phenylethyne. The latter is somewhat surprising as Mayr and

co-workers²⁸recently showed that this is a useful strategy to

derivatise tungsten alkylidyne complexes and possibly reﬂects

the ease of reduction of the tungsten() imido centre under the

coupling conditions.

Characterisation of new tungsten() imido complexes 1 was

straightforward, six being analysed by X-ray crystallography.

2716

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

View Article Online

Each tungsten centre consists of an approximately octahedral

co-ordination sphere with cis-chlorides and trans-phosphines.

The carbonyl and the imido ligands lie cis to one another, as

expected for strong π-acceptor and donor ligands respectively,

and as discussed fully in earlier papers by Mayer.²²Structural

diﬀerences between isomeric para- and meta-phenylene diimido-

bridged complexes 1a and 1n have been discussed previously.⁹

The major feature is a twisting out of the plane by 7.4Њ of

the two tungsten centres in the meta isomer 1n, while in

para-1a they are related by a crystallographic inversion

centre. The diﬀerence is probably due to increased steric crowd-

ing in 1n as a result of the close proximity of the bulky tungsten

centres.

Å is almost identical to that found in neutral biphenyl [1.489 Å]

suggesting that it is a simple single carbon–carbon bond.

A key aim of the work was to access the degree of communi-

cation between π-conjugated metal centres and this was under-

taken by electrochemical studies and supported by electronic

structure calculations. All complexes investigated showed simi-

lar oxidative behaviour, being reversible at low temperature by

cyclic voltammetry. The oxidative response is consistent with

two closely spaced, one-electron oxidations suggesting that

there is some communication between the metal centres. Thus,

peak-to-peak separations have been estimated from diﬀerential

pulse voltammograms as 60 mV in 1a, 90 mV in 1b, 40 mV in

1g, 110 mV in 1n and 170 mV in 1o, all of which are greater

than the 36 mV separation expected on statistical grounds if

both redox centres have identical E_1/2values.⁴¹That the diﬀer-

ence is greater in isomeric meta complexes 1n and 1o versus para

complexes 1a and 1b is somewhat unexpected as communi-

cation between metal centres in the former is not allowed via a

simple superexchange pathway. The slightly greater coupling in

1b versus 1a results from the inequivalence of the metal centres

in the former, while as the chain length increases the interaction

decreases as shown by the small value in 1g. Spectroelectro-

chemical studies on 1a reveal that oxidation results in CO loss

and this is supported by calculations which show that the

HOMO is W–C bonding in character.

For all complexes the major feature of interest is the orient-

ation of the aryl ring(s) with respect to the other tungsten-bound

ligands. Mayer has shown that in the monoimido complex

[WCl₂(Ph₂PMe)₂(CO)(NC₆H₄Me-p)] 1t^22athe ring lies approxi-

mately in the WCl₂(CO) plane, an almost identical orientation

to that in the para-iodo complex 1p. Contrast this with the para-

phenylene imido-bridged 1a in which the aryl ring lies almost

at right angles to the WCl₂(CO) plane, that is almost coplanar

with the two phosphorus atoms and the tungsten centre. The

initial surprise of this result is tempered by the observation

that simple methyl substitution of the ring as in 1b results in a

reversion to the WCl₂(CO) plane. From these and other results

(Table 1) it is clear that the aryl ring(s) can lie either in the

WCl₂(CO) plane or at right angles to it, with intermediate

orientations being disfavoured. For both biaryl-bridged 1g and

1,5-naphthalene-bridged 1h two crystallographically independ-

ent molecules result, the major diﬀerences between them being

the aryl ring orientation. Thus in 1g independent molecules are

characterised by twists of 26.1 and 17.3Њ out of the WCl₂(CO)

planes, with the former being the biggest deviation seen in all

complexes. It thus appears from experimental data that there is

a strong preference for the aryl ring(s) to lie in the WCl₂(CO)

plane. As discussed previously by Mayer^22ain this orientation

the two ﬁlled W–N π orbitals are not of the same energy. That

which lies in the W(CO) plane is stabilised via overlap of the p

orbitals on nitrogen and the carbonyl carbon and thus it is the

higher energy ﬁlled W–N π orbital parallel to this which over-

laps better with the empty aromatic π* orbital(s) of the aryl

ring. The diﬀerence in energy of the ﬁlled W–N π orbitals is

expected to be small, being estimated at <9 kcal mol^Ϫ1from VT

NMR measurements.²³In order to probe this further we carried

out a series of density functional calculations on model com-

pounds with PH₃ligands. From these it is clear that the most

stable orientation of the aryl ring is that in which it lies in the

WCl₂(CO) plane. It thus seems that the ring orientation found

in 1a is anomalous and may be the result of crystal packing

forces. Complex 1a cocrystallises with a molecule of chloro-

benzene and although there are no intermolecular contacts

between the two their preferred relative orientation may be

enough to counter the relatively small energy diﬀerence

between the most stable (in the Cl₂(CO)W plane) and least

stable (at right angles to this) ring orientation. A closer look at

the energy changes upon ring rotation reveals that between

0 and 30Њ neither steric nor electronic eﬀects vary signiﬁcantly,

thus accounting for the diﬀerences seen within independent

molecules of the same compound.

While the oxidative chemistry is very similar for all com-

pounds studied, the reductive chemistry is signiﬁcantly diﬀer-

ent. para-Phenylene-bridged diimido complexes 1a and 1b both

show two well resolved reductions, peak-to-peak separations

being 120 mV for both, while in contrast other complexes

showed only a broad irreversible reduction at high potential.

This diﬀerence is supported by calculations, the HOMO–

LUMO gap for 1aЈ and 1bЈ being signiﬁcantly smaller than in

the other complexes.

It is not easy to compare our results with those of others

as only two preliminary reports of electrochemical studies

on similar para-phenylene diimido bridged complexes have

appeared.^10,13In the Creutz–Taube ion the diﬀerence between

the two oxidation potentials observed for the reduced isovalent

system is 390 mV.²McCleverty, Jones and co-workers⁴²have

carried out detailed studies on 16-electron molybdenum

nitrosyl systems bridged by a wide range of ligands; in 4,4Ј-

bipyridyl bridged [{Mo(tp*)(NO)Cl}₂(4,4Ј-bipy)] [tp* = tris-

(3,5-dimethylpyrazol-1-yl)hydroborate] reduction waves are

separated by 765 mV.⁴²Most closely related to our studies and

speciﬁcally 1a are their amido-bridged complexes [{M(tp*)-

(NO)}₂(µ-p-NHC₆H₄NH)] in which redox processes are separ-

ated by 1.04 V (M = Mo) and 560 mV (M = W).⁴³Clearly, this

latter value is much greater than that of 120 mV found for 1a

and we conclude that at best the coupling between metal centres

in imido-bridged complexes 1 is small. While the degree of

communication between metal centres is clearly dependent

upon the nature of the bridging ligand,^43,44it is also a function

of the distance between metal centres such that the metal–metal

interaction decreases with increasing distance.⁴⁵The diﬀerence

in the oxidative responses of 60 mV for phenylene-bridged 1a

and 40 mV for biphenylene-bridged 1g is in line with this, and

as a result we did not try and assess the communication in

complexes with longer bridges. Indeed, the value of 40 mV for

1g is not signiﬁcantly diﬀerent to that expected if there were no

communication.⁴¹

For the biaryl-bridged diimido complex 1g a notable feature

is the coplanarity of the aryl rings in both independent mole-

cules. This is in accord with the situation in biphenyl in the

solid-state.³⁸In solution, however, biphenyl is known to prefer

an orientation in which the dihedral angle between the two

rings is 41.6Њ,³⁹although the potential energy surface is very ﬂat

and the rotation barrier small. While we have no direct evidence

of rotation about the linking carbon–carbon bond in 1g we

suppose that this is facile in solution. Studies on biphenyl have

shown that the central carbon–carbon bond shortens upon

both oxidation or reduction.⁴⁰In 1g this bond length of 1.49(2)

Experimental

General

Unless otherwise stated, all manipulations were carried out

under nitrogen using standard vacuum line and glove-box

techniques. Solvents were dried and degassed prior to use.

Most diisocyanates and diamines were purchased from Aldrich,

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

2717

View Article Online

Lancaster and Carbolabs and used without further puriﬁcation.

Phenylene diisocyanates and phenylenediamines were puriﬁed

prior to use by vacuum sublimation using a tube furnace.

Triphosgene, phosgene in toluene and [PdCl₂(PPh₃)₂] were

purchased from Aldrich and used as supplied while [WCl₂-

(PPh₂Me)₄],²¹[WCl₂(Ph₂PMe)₂(CO)(NPh)] 1s^22aand para-

ethynylaniline²³were prepared via the literature methods.

Chromatography was carried out on deactivated alumina (6%

w/w distilled water) wet packed with light petroleum (bp 40–

60 ЊC) unless otherwise stated. The solution to be separated was

added to alumina (3–5 g) and the solvent removed under

reduced pressure. The resulting solids were then deposited on

top of the prepared column and separation eﬀected by elution

with progressively more polar solvents. Great caution was

exercised when handling triphosgene, all manipulations being

carried out in a well ventilated fume-hood and exit gases from

the reaction and work-up were passed through a strong

sodium hydroxide solution. The experimental procedure used

is adapted from that described by Eckert and Forster²⁵and

solutions were routinely purged with nitrogen for at least 1 h

after the reaction.

MOLDEN format using the program ADFrom written by

F. Mariotti of the University of Florence.⁵¹

Syntheses

Bis(4-aminophenyl)ethyne Il. Dried triethylamine (40 cm³) was

added to a mixture of para-iodoaniline (0.78 g, 3.56 mmol),

copper() iodide (0.01 g, 0.06 mmol) and para-ethynylaniline

(0.50 g, 4.27 mmol). The mixture was thoroughly degassed and

heated to 50 ЊC in an oil-bath while stirring. After 10 min

[PdCl₂(PPh₃)₂] (0.05 g, 0.07 mmol) was added and the mixture

left at 50 ЊC for 20 min then brought to reﬂux for 20 min.

Cooling to room temperature aﬀorded a yellow solution with

copious precipitate. Diethyl ether (75 cm³) was added and the

mixture was ﬁltered in air to leave a white solid. This was redis-

solved in dichloromethane and prepared for chromatography

on a column packed with hexane–benzene (3:2). Elution with

dichloromethane gave a pale yellow band which aﬀorded com-

pound Il (0.21 g, 28%) as a pale yellow powder. Recrystallis-

ation from hot benzene aﬀorded yellow needles. IR (KBr):

1

3405, 3295 and 3195 cm^Ϫ1(NH). H NMR (CDCl₃): δ 7.31

(d, J 8.0, 4 H, Ar), 6.63 (d, J 8.0 Hz, 4 H, Ar) and 3.78 (br,

4 H, NH₂). Mass spectrum (EI): m/z 208 (M^ϩ). Calc. for

C₇H₆N: C, 80.77; H, 5.77; N, 13.46. Found: C, 79.80; H, 5.75;

N, 13.06%.

The IR spectra were recorded on a Nicolet 205 FTIR spec-

trometer, solid-state spectra being recorded as KBr discs, NMR

spectra on Varian XL200 (¹H) or VXR400 (¹H, ¹³C, ³¹P) spec-

trometers and internally referenced to residual solvent peaks

(¹H, ¹³C) or externally to P(OMe)₃(³¹P) and mass spectra on

VG 7070 high resolution and VG Analytical ZAB2F spectro-

meters. Elemental analyses were performed in house.

᎐

4-H NC H C᎐CC H C᎐CC H NH -4 Im. An analogous pro-

᎐

2

6

4

6

4

6

4

2

cedure was adopted to that described above using para-

ethynylaniline (0.58 g, 4.96 mmol), 1,4-diiodobenzene (0.68 g,

2.06 mmol), copper() iodide (0.01 g, 0.06 mmol), [PdCl₂-

(PPh₃)₂] (0.058 g, 0.08 mmol) and triethylamine (40 cm³) and

yielded a cream solid. Chromatography gave a very slow mov-

ing yellow band with dichloromethane and thus acetone was

added to hasten its removal. The yellow fraction eluted was

evaporated to dryness to give orange compound Im (0.70 g,

Electrochemistry

Voltammetric experiments utilised a single-compartment, three

electrode cell, the electrodes consisting of a platinum bead

working electrode, a platinum bar auxiliary electrode and a

double-fritted Ag–AgCl reference electrode.⁴⁶A PAR 174A

Polarographic Analyser was used in conjunction with a PAR

175 Waveform Generator, without compensation for solution

resistance. Rotating disk voltammetry employed a Metrohm

Model 628-50 Pt rotating disk electrode. In 0.5 mol dm^Ϫ3

[ⁿBu₄N][PF₆]–dichloromethane solution the oxidation of ferro-

cene (Fc^0/ϩ) at ϩ0.55 V was used as an internal reference. Infra-

red spectroelectrochemical experiments utilised an IRRAS

cell,⁴⁷mounted in the sample compartment of a Nicolet Magna

750 FTIR spectrometer. Dichloromethane for electrochemical

and spectroelectrochemical experiments was predried over

KOH and distilled from CaH₂just prior to use. The electrolyte

was prepared by neutralisation of [ⁿBu₄N]OH with HPF₆and

the crude product recrystallised twice from hot absolute ethanol

before drying for 24 h in vacuo at 373 K.

55%). Recrystallisation from hot benzene aﬀorded orange

1

needles. IR (KBr): 3361 (NH), 2206 cm^Ϫ1(C᎐C). H NMR

᎐

(CDCl₃): δ 7.45 (s, 4 H, Ar), 7.35 (d, J 8.0 Hz, 4 H, Ar), 6.65 (d,

J 8.0 Hz, 4 H, Ar) and 3.80 (br, 4 H, NH₂). Mass spectrum (EI):

m/z 308 (M^ϩ). Calc. for C₂₂H₁₆N: C, 85.71; H, 5.19; N, 9.09.

Found: C, 85.12; H, 5.65; N, 6.82%.

para-Aminodiphenylethyne Ir. para-Iodoaniline (2.00 g, 9.13

mmol), copper() iodide (0.02 g, 0.09 mmol) and phenylethyne

(1.17 g, 0.011 mmol) were placed in a Schlenk tube. Dried

triethylamine (40 cm³) was added and the contents were thor-

oughly degassed. To this yellow solution was added [PdCl₂-

(PPh₃)₂] (0.128 g, 0.183 mmol) and the resulting solution stirred

overnight at room temperature. The solvent was removed under

reduced pressure. Benzene (100 cm³) extraction gave a yellow-

brown solid which was chromatographed on a column made

up from light petroleum–dichloromethane (1:1). Elution with

dichloromethane gave an orange band which aﬀorded com-

Computational

All calculations were performed with the Amsterdam Density

Functional program suite.^48,49A triple-zeta, uncontracted

Slater-type orbital valence basis set was employed for tungsten

(ADF Type IV). Double-zeta bases were used for all the other

atoms, together with a single polarisation function (ADF Type

III). Quasi-relativistic scalar frozen cores were used for all

the elements; C (1s), N (1s), O (1s), P (2p), Cl (2p) and W (5p).

Relativistic core potentials were computed using the ADF

auxiliary program “Dirac”. The local density functional of

Vosko, Wilk and Nusair⁵⁰was used in all calculations. Both spin

restricted and spin unrestricted (i.e. when α and β spin electrons

of the same MO number and symmetry are not constrained to

have the same spacial wavefunction) calculations were per-

formed where appropriate. Bond lengths and angles were taken

from the appropriate crystallographic data and the molecular

geometry idealised to the highest possible symmetry. Molecular

orbital plots were generated using the program MOLDEN,

written by G. Schaftenaar of the COAS/CAMM Centre,

Nijmegen. The ADF output ﬁles (TAPE21) were converted into

pound Ir (1.27 g, 72%) as a yellow-orange powder. IR (KBr):

1

3476, 3381 (NH), 2211 cm^Ϫ1(C᎐C). H NMR (CDCl ): δ 7.50

᎐

3

(d, J 8.0, 2 H, Ar), 7.34 (m, 5 H, Ar), 6.65 (d, J 8.0, 2 H, Ar) and

3.82 (br, 2 H, NH₂). Mass spectrum (EI): m/z 193 (M^ϩ). Calc.

for C₁₄H₁₁N: C, 87.05; H, 5.70; N, 7.25. Found: C, 86.75; H,

5.63; N, 7.16%.

2,3-Dimethylphenylene-1,4-diisocyanate IIc. 2,3-Dimethyl-

phenylene-1,4-diamine Ic (0.25 g, 1.84 mmol) was dissolved in

1,2-dichlorobenzene (30 cm³) to give a red solution which upon

brisk addition of triphosgene (0.36 g, 1.22 mmol) became

cloudy. A reﬂux condenser ﬁtted with a bubbler was attached

and on warming to reﬂux the solution became clear and purple.

Reﬂux was maintained for 30 min and then cooling to room

temperature yielded a yellow solution with some ﬂocculent

brown solid. After decanting the solution was purged vigor-

ously with nitrogen for 45 min. Removal of the solvent under

2718

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

View Article Online

reduced pressure at 40 ЊC gave an oﬀ-white solid. This was

placed into a Schlenk ﬂask ﬁtted with a cold-ﬁnger and sublim-

ation at 0.05 mmHg aﬀorded compound IIc (0.15 g, 43%) as a

reduced pressure to yield oﬀ-white compound IIi (0.08 g, 20%).

No further attempt was made at puriﬁcation. Attempts were

made to extract more product from the remaining solid using

hot toluene, however, these were unsuccessful. IR (KBr): 2265

1

white solid. IR (KBr): 2285 cm^Ϫ1(NCO). H NMR (CDCl₃):

1

δ 6.91 (s, 2 H, Ar) and 2.25 (s, 6 H, Me). Mass spectrum (EI):

m/z 188 (M^ϩ). Calc. for C₅H₄NO: C, 63.83; H, 4.26; N, 14.89.

Found: C, 63.74; H, 4.27; N, 14.51%.

cm^Ϫ1(NCO). H NMR (CDCl₃): δ 7.63 (s, 4 H, Ar), 7.58 (d,

J 8.0, 4 H, Ar), 7.26 (d, J 8.0, 4 H, Ar). Mass spectrum (EI):

m/z 312 (M^ϩ).

2,5-Dimethylphenylene-1,4-diisocyanate 2d. The procedure

described above was followed using 2,5-dimethylphenylene-1,4-

diamine Id (0.50 g, 3.97 mmol) in 1,2-dichlorobenzene (40 cm³).

The amine did not completely dissolve before triphosgene (0.73

g, 2.45 mmol) addition. Brilliant white compound IId (0.44 g,

59%) was sublimed from a green-brown sludge. IR (KBr): 2264

Bis(4-isocyanatophenyl)ethyne IIl. A mixture of bis(4-amino-

phenyl)ethyne Il (0.29 g, 1.39 mmol), triethylamine (0.42 g, 4.16

mmol) and toluene (40 cm³) was gently warmed to aid diamine

dissolution. Triphosgene (0.28 g, 0.94 mmol) was added and

standard reaction procedures were followed. Sublimation

aﬀorded compound IIl (0.16 g, 80%) as a pale yellow solid. IR

1

cm^Ϫ1(NCO). H NMR (CDCl₃): δ 6.88 (s, 2 H, Ar) and 2.24

(KBr): 2266 cm^Ϫ1(NCO). H NMR (CDCl₃): δ 7.47 (d, J 8.0,

(s, 6 H, Me). Mass spectrum (EI): m/z 188 (M^ϩ). Calc. for

C₅H₄NO: C, 63.83; H, 4.26; N, 14.89. Found: C, 63.76; H, 4.19;

N, 14.86%.

4 H, Ar) and 7.07 (d, J 8.0 Hz, 4 H, Ar). Mass spectrum (EI):

m/z 260 (M^ϩ).

᎐

4-OCNC H C᎐CC H C᎐CC H NCO-4 IIm. Diamine Im

᎐

6

4

6

4

6

4

(0.20 g, 0.65 mmol), triethylamine (0.59 g, 5.84 mmol) and

toluene (40 cm³) were warmed gently to aid dissolution of

the diamine. Triphosgene (0.13 g, 0.43 mmol) was added and

the mixture reﬂuxed for 2.5 h. After ﬁltration, removal of

solvent aﬀorded compound IIm (0.04 g, 17%) as an orange

solid. This did not sublime and no attempts were made at

Tetramethylphenylene-1,4-diisocyanate IIe. A similar method

was followed and from tetramethylphenylene-1,4-diamine Ie

(0.50 g, 3.04 mmol) and triphosgene (0.60 g, 2.03 mmol) in 1,2-

dichlorobenzene (30 cm³), white compound IIe (0.50 g, 75%)

1

was isolated. IR (KBr): 2285 cm^Ϫ1(NCO). H NMR (CDCl₃):

δ 2.26 (s, Me). Mass spectrum (EI): m/z 216 (M^ϩ). Calc. for

C₆H₆NO: C, 66.67; H, 5.56; N, 12.96. Found: C, 66.57; H, 5.54;

N, 12.91%.

1

further puriﬁcation. IR (KBr): 2284, 2253 cm^Ϫ1(NCO). H

NMR (CDCl₃): δ 7.51 (s, 4 H, Ar), 7.45 (d, J 8.0, 4 H, Ar) and

7.11 (d, J 8.0 Hz, 4 H, Ar). Mass spectrum (EI): m/z 360 (M^ϩ).

3,3Ј,5,5Ј-Tetramethylbiphenyl-4,4Ј-diisocyanate IIf. A similar

method was followed and from 3,3Ј,5,5Ј-tetramethylbiphenyl-

4,4Ј-diamine If (0.23 g, 0.96 mmol) and triphosgene (0.19 g,

0.64 mmol) in 1,2-dichlorobenzene (30 cm³), white compound

IIf (0.25 g, 90%) was isolated. IR (KBr): 2274 cm^Ϫ1(NCO). ¹H

NMR (CDCl₃): δ 7.24 (s, 4 H, Ar) and 2.38 (s, 12 H, Me). Mass

spectrum (EI): m/z 292 (M^ϩ). Calc. for C₉H₈NO: C, 74.00; H,

5.48; N, 9.59. Found: C, 73.67; H, 5.36; N, 9.50%.

para-Iodophenyl isocyanate IIp. para-Iodoaniline (1.00 g,

4.57 mmol) was dissolved in toluene (30 cm³) and triethylamine

(1 cm³) added. To this was added phosgene in toluene (24

cm³, 46 mmol) via a syringe. Some cloudiness was seen and the

mixture was heated to reﬂux for 30 min. The solvent was

removed under reduced pressure and white compound IIp (0.62

g, 55%) isolated by sublimation. IR (KBr): 2283 and 2258 cm^Ϫ1

1

(NCO). H NMR (CDCl₃): δ 7.63 (d, J 8.4, 2 H, Ar) and 6.85

(d, J 8.4 Hz, 2 H, Ar). Mass spectrum (EI): m/z 245 (M^ϩ).

3,3Ј-Dimethoxybiphenyl-4,4Ј-diisocyanate IIg. Triphosgene

(0.20 g, 0.68 mmol) was swiftly added to a red solution of

3,3Ј-dimethoxybiphenyl-4,4Ј-diamine Ig (0.25 g, 1.03 mmol)

in 1,2-dichlorobenzene (30 cm³). A reﬂux condenser with a

bubbler ﬁtted was attached. After a few minutes the solution

became clear and lightened, then a blue hue formed. The mix-

ture was brought to reﬂux over 40 min and maintained for 90

min. Cooling to room temperature aﬀorded a colourless solu-

tion with some brown solid. The solution was decanted from

the small amount of solid and purged vigorously with nitrogen

for 1 h. The solvent was removed under reduced pressure to

yield compound IIg (0.26 g, 85%) as a white solid. No further

᎐

para-OCNC H C᎐CC H IIr. para-Aminodiphenylethyne Ir

᎐

6

4

6

5

(0.33 g, 1.71 mmol), triethylamine (0.52 g, 5.13 mmol) and tolu-

ene (40 cm³) were warmed in order to dissolve the amine and

triphosgene (0.17 g, 0.57 mmol) was added. The solution was

reﬂuxed for 2 h. After cooling to room temperature and purging

with nitrogen for 1 h, ﬁltration gave an orange solution and a

solid which was washed further with warm toluene (40 cm³).

The combined ﬁltrates were evaporated to dryness to give an

orange-brown solid from which sublimation aﬀorded com-

pound IIr (0.29 g, 77%) as a white powder. IR (KBr): 2263

(NCO), 2124 cm^Ϫ1(C᎐C). H NMR (d -acetone): δ 7.6–7.2 (m,

1

6

᎐

1

puriﬁcation was attempted. IR (KBr): 2280 cm^Ϫ1(NCO). H

Ar). Mass spectrum (EI): m/z 219 (M^ϩ).

NMR (CDCl₃): δ 7.14 (m, 2 H, Ar), 7.04 (m, 4 H, Ar) and 3.99

(s, 6 H, OMe).

[{WCl₂(Ph₂PMe)₂(CO)}₂(ꢁ-p-NC₆H₄N)] Ia. In a glove-box,

[WCl₂(Ph₂PMe)₄] (2.00 g, 1.90 mmol) and p-phenylene-1,4-

diisocyanate (0.15 g, 0.95 mmol) were placed in a Schlenk tube.

After removal from the box, toluene (30 cm³) was added via a

cannula and the mixture stirred at room temperature for 3 h.

The initially orange solution progressively darkened over this

period becoming green-brown and all solids dissolved. The

volume was reduced under reduced pressure to approximately 5

cm³and light petroleum (30 cm³) added. A green solid precipi-

tated which was ﬁltered oﬀ and washed with light petroleum

(10 cm³). Elemental analysis was obtained on this solid. Dis-

solution in dichloromethane (5 cm³) in air gave a dark green

solution. This was layered with methanol (10 cm³), slow diﬀu-

sion of the two solids depositing complex 1a (1.50 g, 93%) as

a green powder. Attempts to grow good quality single crystals

from a variety of organic solvents led only to the formation of

ﬁne and brittle needles. However, slow evaporation of a 1,4-

dichlorobenzene solution aﬀorded large green blocks suitable

for X-ray crystallography. IR (KBr): 1960vs (CO), 1486m,

1,5-Diisocyanatonaphthalene IIh. A similar method was fol-

lowed from naphthalene-1,5-diamine Ih (0.40 g, 2.53 mmol)

and triphosgene (0.50 g, 1.69 mmol) in 1,2-dichlorobenzene (40

cm³). Filtration was carried out while the reaction mixture was

still warm and white compound IIh (0.12 g, 23%) was sublimed

with some diﬃculty from an orange solid. IR (KBr): 2270 cm^Ϫ1

(NCO). ¹H NMR (CDCl₃): δ 7.96 (d, J 8.6, 2 H, H²of Ar), 7.49

(t, J 8.0, 2 H, H³of Ar) and 7.36 (d, J 8.2 Hz, 2 H, H⁴of Ar).

Mass spectrum (EI): m/z 210 (M^ϩ). Calc. for C₆H₃NO, C, 68.57;

H, 2.86; N, 13.33. Found: C, 67.02; H, 2.68; N, 12.87%.

Terphenyl-4,4Љ-diisocyanate IIi. A similar procedure was fol-

lowed using terphenyl-4,4Љ-diamine 1i (0.33 g, 1.27 mmol) in

1,2-dichlorobenzene (40 cm³). Some dissolution was observed

forming an ochre suspension. Triphosgene (0.25 g, 0.85 mmol)

was added and the mixture heated at reﬂux for 3h. The solution

was decanted after cooling and the solvent removed under

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

2719

View Article Online

1

1438m, 1432m, 1346w, 1100w, 890s, 846w, 742s, 693s, 504m and

455w cm^Ϫ1. ¹H NMR (CDCl₃): δ 7.67–7.17 (m, 40 H, Ph), 5.95

(s, 4 H, Ar) and 2.28 (t, J 3.9 Hz, 12 H, Me). ¹³C NMR (CDCl₃):

δ 239.8 (s, CO), 151.7 (s), 135.4 (t, 22.8), 134.0 (t, J 20.5), 133.0

(t, J 5.5), 131.9 (t, J 5.2), 130.0 (s), 129.9 (s), 128.9 (s), 128.2 (t,

J 4.8), 128.1 (t, J 4.8 Hz) and 124.7 (s). ³¹P NMR (CDCl₃):

δ 2.08 (s, J_PW290 Hz). Mass spectrum (FAB): m/z 1442

(M^ϩϪ CO) and 1414 (M^ϩϪ 2CO). Calc. for C₆₀H₅₆Cl₄N₂O₂-

P₄W₂ؒC₇H₈: C, 51.47; H, 4.10; N, 1.79; P, 7.93. Found: C, 51.56;

H, 4.07; N, 1.80; P, 7.97%.

1953vs cm^Ϫ1(CO). H NMR (CDCl₃): δ 8.00–7.08 (m, 40 H,

Ph), 6.74 (s, 4 H, Ar), 2.31 (t, J 4.0, 12 H, Me) and 1.64 (s, 12 H,

Me). ¹³C NMR (CDCl₃): δ 241.2 (s, CO), 155.5 (s), 135.8 (t,

J 22.8), 134.6 (t, J 21.7), 133.1 (t, J 5.3), 132.3 (t, J 5.3), 129.8 (t,

J 4.5), 125.9 (t, J 4.7), 126.2 (s), 19.2 (s, Me), 17.9 (s, Me) and

13.5 (t, J 15.5, Me). ³¹P NMR (CDCl₃): δ 2.20 (s, J_PW293 Hz).

Mass spectrum (FAB): m/z 1574 (M^ϩϪ CO). Calc. for C₃₅H₃₄-

Cl₂NOP₂WؒCH₂Cl₂: C, 48.76; H, 4.06; N, 1.58. Found: C,

48.66; H, 3.91; N, 1.52%.

[{WCl₂(Ph₂PMe)₂(CO)}₂{ꢁ-p-N-3,3Ј-(MeO)₂C₁₂H₆N}]

1g.

For other complexes 1 a similar procedure was adopted

Reaction of [WCl₂(Ph₂PMe)₄] (0.71 g, 0.68 mmol) and 3,3Ј-

dimethoxybiphenyl-4,4Ј-diisocyanate IIg (0.10 g, 0.33 mmol)

in toluene (40 cm³) gave complex 1g (0.48 g, 88%). Crystals

suitable for X-ray analysis were grown upon slow diﬀusion

of methanol into a saturated dichloromethane solution. IR

(KBr): 1975vs (CO), 1591m, 1545w, 1483m, 1435s, 1392w,

1283w, 1241w, 1100m, 1021m, 889s, 805m, 742m, 692s and

506m cm^Ϫ1. ¹H NMR (CDCl₃): δ 7.69–7.24 (m, 40 H, Ph), 6.57

(d, J 8.4, 2 H, Ar), 6.54 (s, 2 H, Ar), 6.53 (d, J 8.4, 2 H, Ar), 3.56

(s, 6 H, MeO) and 2.31 (t, J 4.0 Hz, 12 H, Me). ¹³C NMR

(CDCl₃): δ 238.1 (s), 155.5 (s), 143.5 (s), 139.5 (s), 135.7 (t, J

20.8), 134.5 (t, J 21.1), 133.1 (t, J 5.3), 132.3 (t, J 5.2), 129.7 (d,

J 11.5), 128.0 (m), 125.9 (s), 117.9 (s), 108.7 (s), 54.5 (s, MeO)

and 13.5 (t, J 15.2 Hz, Me). ³¹P NMR (CDCl₃): δ 3.20 (s, J_PW

292 Hz). Mass spectrum (FAB): m/z 1578 (M^ϩϪ CO). Calc. for

C₃₄H₃₂Cl₂NO₂P₂W: C, 50.81; H, 3.99; N, 1.74. Found: C, 50.87;

H, 4.18; N, 1.85%.

unless otherwise stated.

[{WCl₂(Ph₂PMe)₂(CO)}₂(ꢁ-p-N-o-MeC₆H₃N)] 1b. Reaction

of [WCl₂(Ph₂PMe)₄] (1.10 g, 1.16 mmol) and 2-methyl-

phenylene-1,4-diisocyanate IIb (0.1 g, 0.58 mmol) in toluene (35

cm³) gave complex 1b (0.73 g, 85%). Crystals suitable for X-ray

analysis were grown from dichloromethane–methanol mixtures

in air. IR (KBr): 1968vs (CO), 1483w, 1435s, 1098w, 891s, 738m,

693s and 507m cm^Ϫ1; ¹H NMR (CDCl₃): δ 7.78–6.94 (m, 40 H,

Ph), 6.14 (d, J 8.2, 1 H, Ar), 5.94 (d, J 8.2, 1 H, Ar), 5.74 (s, 1 H,

Ar), 2.33 (t, J 4.2, 3 H, Me) and 2.29 (t, J 4.3 Hz, 12 H, Me). ¹³

C

NMR (CDCl₃): δ 243.2 (s, CO), 239.4 (s, CO), 151.6 (s), 150.5

(s), 135.5 (t, J 27.2), 134.2 (t, J 22.7), 133.4 (t, J 5.1), 132.4 (t,

J 5.1), 132.1 (t, J 4.7), 131.9 (t, J 4.7), 130.1 (s), 130.0 (s), 129.5

(s), 128.3 (s), 127.8 (t, J 4.2), 126.0 (m), 122.5 (s), 17.0 (s, Me),

13.5 (t, J 15.3, Me) and 13.3 (t, J 15.7 Hz). ³¹P NMR (CDCl₃):

δ 3.43 (s, J_PW291) and 2.06 (s, J_PW291 Hz). Mass spectrum

(FAB): m/z 1455 (M^ϩϪ CO). Calc. for C₆₁H₅₈Cl₄P₄N₂O₂W₂ؒ

CH₃OH: C, 49.08; H, 3.83; N, 1.85. Found: C, 49.10; H, 3.90;

N, 1.86%.

[{WCl₂(Ph₂PMe)₂(CO)}₂(ꢁ-1,5-NC₁₀H₆N)] 1h. Reaction of

[WCl₂(Ph₂PMe)₄] (0.25 g, 0.24 mmol) and 1,5-diisocyanato-

naphthalene Ih (0.025 g, 0.12 mmol) in toluene (20 cm³)

gave complex 1h (0.10 g, 55%) as a green powder after several

recrystallisations from methanol–dichloromethane mixtures.

[{WCl₂(Ph₂PMe)₂(CO)}₂(ꢁ-p-N-2,3-Me₂C₆H₂N)] 1c. Reac-

tion of [WCl₂(Ph₂PMe)₄] (0.12 g, 0.11 mmol) and 2,3-dimethyl-

phenylene-1,4-diisocyanate IIc (0.011 g, 0.06 mmol) in toluene

(10 cm³) gave complex 1c (0.035 g, 40%). IR (KBr): 1965vs cm^Ϫ1

(CO). ¹H NMR (CDCl₃): δ 7.67–6.80 (m, 42 H, Ph ϩ Ar), 2.30

(t, J 4.0 Hz, 12 H, Me) and 1.24 (s, 6 H, Me). ³¹P NMR

(CDCl₃): δ 2.30 (s, J_PW292 Hz). Mass spectrum (FAB): m/z 1499

(M^ϩ). Calc. for C₃₁H₃₀Cl₂NOP₂W: C, 49.73; H, 4.01; N, 1.87.

Found: C, 49.42; H, 3.89; N, 1.86%.

1

IR (KBr): 1964vs cm^Ϫ1(CO). H NMR (CDCl₃): δ 7.83–6.94

(m, 46 H, Ph ϩ Ar) and 2.31 (t, J 4.0 Hz, 12 H, Me). ³¹P NMR

(CDCl₃): δ 3.00 (s, J_PW292 Hz). Mass spectrum (FAB): m/z

1492 (M^ϩϪ CO). Calc. for C₆₄H₅₈Cl₄N₂O₄P₄W₂ؒ0.5CH₂Cl₂: C,

49.54; H, 3.78; N, 1.79. Found: C, 49.22; H, 3.83; N, 1.76%.

[{WCl₂(Ph₂PMe)₂(CO)}₂{ꢁ-p-N(C₆H₄)₃N}] 1i. Reaction of

[WCl₂(Ph₂PMe)₄] (0.41 g, 0.39 mmol) and terphenylene-4,4Љ-

diisocyanate IIi (0.06 g, 0.19 mmol) in toluene (30 cm³) gave

[{WCl₂(Ph₂PMe)₂(CO)}₂(ꢁ-p-N-2,5-Me₂C₆H₂N)] 1d. Reac-

tion of [WCl₂(Ph₂PMe)₄] (0.45 g, 0.43 mmol) and 2,5-dimethyl-

phenylene-1,4-diisocyanate IId (0.04 g, 0.21 mmol) in toluene

(10 cm³) gave complex 1d (0.23 g, 74%). IR (KBr): 1954vs cm^Ϫ1

1

complex 1i (0.10 g, 55%). IR (KBr): 1954vs cm^Ϫ1(CO). H

NMR (CDCl₃): δ 7.78–7.31 (m, 44 H, Ph ϩ Ar), 7.17 (d, J 8.5, 4

H, Ar), 6.51 (d, J 8.5, 4 H, Ar) and 2.34 (t, J 4.0, 12 H, Me). ¹³

C

NMR (CDCl₃): δ 240.1 (s, CO), 153.9 (s), 139.2 (s), 137.8 (s),

135.5 (t, J 22.9), 134.2 (t, J 21.0), 133.2 (t, J 5.0), 132.1 (t, J 5.0),

130.0 (s), 129.7 (s), 128.3 (d, J 4.8), 127.2 (s), 126.6 (s), 126.5 (s),

125.6 (s) and 13.6 (t, J 14.1 Hz, Me). ³¹P NMR (CDCl₃): δ 3.10

(s, J_PW292 Hz). Mass spectrum (FAB): m/z 1594 (M^ϩϪ CO).

Calc. for C₇₂H₆₄Cl₄N₂O₄P₄W₂ؒCH₂Cl₂: C, 51.30; H, 3.87; N,

1.64. Found: C, 50.77; H, 3.73; N, 1.70%.

1

(CO). H NMR (CDCl₃): δ 7.57–7.00 (m, 40 H, Ph), 6.02 (s,

2 H, Ar), 2.28 (t, J 4.0 Hz, 12 H, Me) and 1.27 (s, 6 H, Me). ¹³

C

NMR (CDCl₃): δ 243.0 (s, CO), 150.4 (s), 135.4 (t, J 22.8), 134.2

(t, J 23.9), 132.7 (s), 132.4 (t, J 5.3), 132.2 (t, J 5.3), 130.1 (s),

129.7 (s), 128.2 (t, J 4.8), 127.9 (t, J 4.7), 127.8 (s), 16.5 (s, Me)

and 13.3 (t, J 15.5 Hz, Me). ³¹P NMR (CDCl₃): δ 2.20 (s, J_PW

292 Hz). Mass spectrum (FAB): m/z 1470 (M^ϩϪ CO). Calc. for

C₃₁H₃₀Cl₂NOP₂W: C, 49.73; H, 4.01; N, 1.87. Found: C, 49.42;

H, 3.84; N, 1.89%.

[{WCl₂(Ph₂PMe)₂(CO)}₂(ꢁ-p-NC₆H₄CH₂C₆H₄N)] 1j. Reac-

tion of [WCl₂(Ph₂PMe)₄] (0.25 g, 0.24 mmol) and methylene-

diphenylene-4,4Ј-diisocyanate IIj (0.03 g, 0.12 mmol) in toluene

(20 cm³) gave purple complex 1j (0.16 g, 85%). IR (KBr): 1962vs

cm^Ϫ1(CO). ¹H NMR (CDCl₃): δ 7.72–7.22 (m, 40 H, Ph), 6.58

(d, J 8.2, 4 H, Ar), 6.35 (d, J 8.2, 4 H, Ar), 3.57 (s, 2 H, CH₂)

and 2.31 (t, J 4.0, 12 H, Me). ¹³C NMR (CDCl₃): δ 240.2 (s,

CO), 153.6 (s), 139.3 (s), 136.2 (t, J 20.9), 134.9 (t, J 20.5), 133.9

(t, J 5.1), 132.8 (t, J 5.2), 130.7 (t, J 3.4), 129.3 (s), 129.0 (d,

J 4.5), 128.9 (d, J 5.0), 125.9 (s), 41.4 (s, CH₂) and 13.5 (t, J 15.2

Hz, Me). ³¹P NMR (CDCl₃): δ 3.20 (s, J_PW292 Hz). Mass spec-

trum (FAB): m/z 1532 (M^ϩϪ CO). Calc. for C₆₇H₆₂Cl₄N₂O₄-

P₄W₂: C, 51.54; H, 3.97; N, 1.80. Found: C, 51.60; H, 4.16; N,

1.86%.

[{WCl₂(Ph₂PMe)₂(CO)}₂(ꢁ-p-NMe₄C₆N)] 1e. Reaction of

[WCl₂(Ph₂PMe)₄] (0.43 g, 0.40 mmol) and tetramethyl-

phenylene-1,4-diisocyanate IIe (0.034 g, 0.20 mmol) in toluene

(40 cm³) after overnight stirring gave complex 1e (0.20 g, 66%).

IR (KBr): 1956vs cm^Ϫ1(CO). H NMR (CDCl₃): δ 7.95–6.85

1

(m, 40 H, Ph), 2.29 (t, J 4.0 Hz, 12 H, Me) and 1.54 (s, 12 H,

Me). ³¹P NMR (CDCl₃): δ 2.80 (s, J_PW292 Hz). Mass spectrum

(FAB): m/z 1498 (M^ϩϪ CO). Calc. for C₃₂H₃₂Cl₂NOP₂W: C,

50.33; H, 4.32; N, 1.83. Found: C, 51.52; H, 4.51; N, 1.68%.

[{WCl₂(Ph₂PMe)₂(CO)}₂(ꢁ-p-N-3,3Ј,5,5Ј-Me₄C₁₂H₄N)]

1f.

Reaction of [WCl₂(Ph₂PMe)₄] (0.43 g, 0.40 mmol) and 3,3Ј,5,5Ј-

tetramethylbiphenyl-4,4Ј-diisocyanate IIf (0.06 g, 0.21 mmol)

in toluene (40 cm³) gave complex 1f (0.19 g, 56%). IR (KBr):

[{WCl₂(Ph₂PMe)₂(CO)}₂(ꢁ-p-NC₆H₁₀CH₂C₆H₁₀N)] 1k. Reac-

tion of [WCl₂(Ph₂PMe)₄] (0.32 g, 0.31 mmol) and bis(4-iso-

2720

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

View Article Online

cyanatocyclohexyl)methane IIk (0.04 g, 0.15 mmol) in toluene

(20 cm³) gave purple complex 1k (0.21 g, 90%). IR (KBr):

0.42 mmol) were placed in a Schlenk tube to which toluene (40

cm³) was added. After 3 h volatiles were removed under reduced

pressure aﬀording an oily purple solid. This was redissolved in

diethyl ether (2 cm³) and light petroleum added to give pink

complex 1p (0.37 g, 91%) which was further washed with light

petroleum. Crystals suitable for X-ray analysis were grown

upon slow diﬀusion of methanol into a saturated dichloro-

1

1961vs cm^Ϫ1(CO). H NMR (CDCl₃): δ 7.88–7.38 (m, 40 H,

Ph), 3.46 (q, J 8.0, 2 H, CH₂), 2.31 (t, J 4.0 Hz, 12 H, Me) and

1.20–0.52 (m, 20 H, Cy). ³¹P NMR (CDCl₃): δ 3.20 (s, J_PW292).

Mass spectrum (FAB): m/z 1543 (M^ϩϪ CO). Calc. for

C₇₇H₆₂Cl₄N₂O₄P₄W₂: C, 51.15; H, 4.71; N, 1.78. Found: C,

50.59; H, 4.63; N, 1.76%.

1

methane solution. IR (CH₂Cl₂): 1967s cm^Ϫ1(CO). H NMR

(CDCl₃): δ 7.73–7.31 (m, 20 H, Ph), 7.23 (d, J 8.0, 2 H, Ar), 6.13

(d, J 8.0, 2 H, Ar) and 2.32 (t, J 3.9 Hz, 6 H, Me). ³¹P NMR

(CDCl₃): δ 2.80 (s, J_PW290 Hz). Mass spectrum (FAB): m/z 873

(M^ϩϪ CO). Calc. for C₃₃H₃₀Cl₂INOP₂W: C, 44.00; H, 3.33;

N, 1.55. Found: C, 43.14; H, 3.22; N, 1.44%.

᎐

[{WCl (Ph PMe) (CO)} (ꢁ-p-NC H C᎐CC H N)] 1l. Reac-

᎐

2

6

4

6

4

tion of [WCl₂(Ph₂PMe)₄] (0.40 g, 0.38 mmol) and bis(4-

isocyanatophenyl)ethyne IIl (0.05 g, 0.19 mmol) in toluene

(20 cm³) for 2d at room temperature gave a purple solution and

an oﬀ-white solid. After ﬁltration and removal of most of the

solvent under reduced pressure light petroleum (10 cm³) was

added to give a green precipitate. This was redissolved in

dichloromethane (ca. 1 cm³) and addition of methanol gave

green complex 1l (0.17 g, 57%). IR (KBr): 1965vs cm^Ϫ1(CO).

¹H NMR (CDCl₃): δ 7.67–6.80 (m, 48 H, Ph ϩ Ar) and 2.31 (t,

J 4.0 Hz, 12 H, Me). ³¹P NMR (CDCl₃): δ 3.00 (s, J_PW292 Hz).

Mass spectrum (FAB): m/z 1540 (M^ϩϪ CO).

[WCl₂(Ph₂PMe)₂(CO)(NC₆H₄Br-p)] 1q. para-Bromophenyl

isocyanate IIq (0.05 g, 0.25 mmol) and [WCl₂(Ph₂PMe)₄] (0.27

g, 0.26 mmol) were placed in a Schlenk tube to which toluene

(30 cm³) was added. After 3 h volatiles were removed under

reduced pressure aﬀording an oily purple solid. This was redis-

solved in diethyl ether (2 cm³) and light petroleum added to give

pink complex 1q (0.19 g, 88%) which was further washed with

light petroleum. IR (CH₂Cl₂): 1961s cm^Ϫ1(CO). ¹H NMR

(CDCl₃): δ 7.71–7.32 (m, 20 H, Ph), 7.02 (d, J 8.2, 2 H, Ar), 6.24

(d, J 8.2, 2 H, Ar) and 2.31 (t, J 4.0, 6 H, Me). ³¹P NMR

(CDCl₃): δ 3.10 (s, J_PW290 Hz). Mass spectrum (FAB): m/z 825

(M^ϩϪ CO). Calc. for C₃₃H₃₀BrCl₂NOP₂W: C, 46.42; H, 3.52;

N, 1.64. Found: C, 47.00; H, 3.50; N, 1.54%.

᎐

[{WCl (Ph PMe) (CO)} (ꢁ-p-NC H C᎐CC H C᎐CC H N)]

᎐

2

6

4

6

4

6

4

1m. Reaction of [WCl₂(Ph₂PMe)₄] (0.40 g, 0.38 mmol) and

diamine IIm (0.04 g, 0.15 mmol) in toluene (40 cm³) for 3d

resulted in the slow dissolution of most solids. After ﬁltration

the solvent was reduced to ca. 2 cm³and light petroleum (10

cm³) added to eﬀect precipitation of green complex 1m (0.14 g,

44%). IR (CH₂Cl₂): 1966vs cm^Ϫ1(CO). H NMR (CDCl₃):

1

᎐

[WCl (Ph PMe) (CO)(p-NC H C᎐CC H )] 1r. para-Phenyl-

᎐

2

6

4

6

5

δ 7.73–6.87 (m, 52 H, Ph ϩ Ar) and 2.31 (t, J 4.0 Hz, 12 H, Me).

³¹P NMR (CDCl₃): δ 3.00 (s, J_PW292). Mass spectrum (FAB):

m/z 1642 (M^ϩϪ CO).

ethynylphenyl isocyanate IIr (0.066 g, 0.30 mmol) and

[WCl₂(Ph₂PMe)₄] (0.31 g, 0.30 mmol) were placed in a Schlenk

tube to which toluene (30 cm³) was added. After 16 h, the green

solution was ﬁltered and volatiles were removed under reduced

pressure aﬀording a viscous green paste. This was redissolved in

dichloromethane (2 cm³) and layered with methanol to give a

green powder 1r (0.15 g, 57%). Some decomposition occurred

while standing in air and an analytically pure sample could not

be obtained. IR (CH₂Cl₂): 1966s cm^Ϫ1(CO). ¹H NMR (CDCl₃):

δ 7.64–6.81 (m, 29 H, Ph ϩ Ar) and 2.31 (t, J 4.0 Hz, 6 H, Me).

³¹P NMR (CDCl₃): δ 3.00 (s, J_PW292 Hz). Mass spectrum

(FAB): m/z 845 (M^ϩϪ CO).

[{WCl₂(Ph₂PMe)₂(CO)}₂(ꢁ-m-NC₆H₄N)] 1n. Reaction of

[WCl₂(Ph₂PMe)₄] (0.53 g, 0.50 mmol) and meta-phenylene-

diisocyanate IIn (0.04 g, 0.25 mmol) in toluene (35 cm³) for

1d gave a purple solution. Removal of solvent under reduced

pressure aﬀorded an oily solid which was washed with light

petroleum. Redissolution in dichloromethane and addition of

methanol aﬀorded purple complex 1n (0.33 g, 89%). Cooling

a saturated THF solution to Ϫ40 ЊC gave crystals suitable for

X-ray analysis. IR (CH₂Cl₂): 1966vs cm^Ϫ1(CO). ¹H NMR

(CDCl₃): δ 7.60–6.95 (m, 40 H, Ph), 6.35 (t, J 8.1, 1 H, Ar), 6.10

(d, J 8.1, 2 H, Ar), 6.00 (s, 1 H, Ar) and 2.24 (t, J 3.9 Hz, 12 H,

Me). ¹³C NMR (CDCl₃): δ 239.5 (s, CO), 153.9 (s), 135.3 (t,

J 22.9), 133.8 (t, J 20.7), 132.9 (t, J 5.1), 132.0 (t, J 5.1), 130.1

(s), 128.3 (t, J 4.5), 128.1 (t, J 4.7), 123.5 (s), 119.5 (s) and 13.5

(t, J 15.5 Hz, Me). ³¹P NMR (CDCl₃): δ 2.14 (s, J_PW290 Hz).

Mass spectrum (FAB): m/z 1442 (M^ϩϪ CO). Calc. for C₆₀H₅₆-

Cl₄N₂O₂P₄W₂ؒ0.5CH₂Cl₂: C, 48.00; H, 3.77; P, 8.20; Cl, 11.73;

N, 1.85. Found: C, 47.81; H, 3.85; Cl, 11.16; N, 1.84, P, 8.55%.

Attempted coupling of [WCl₂(Ph₂PMe)₂(CO)(NC₆H₄I-p)] 1p.

Activated copper (0.011 g, 0.16 mmol) was added to a purple

solution of complex 1p (0.50 g, 0.06 mmol) in dmf (10 cm³).

Upon heating to 110 ЊC the solution became red-orange and

this was maintained for 15 min. After cooling to room temper-

ature, grey CuI was removed via ﬁltration to leave a red-brown

solution from which volatiles were removed under reduced

pressure to give an oily brown solid. A number of attempts were

made to triturate this material from a range of solvents but with

little success. Some material was soluble in dichloromethane

and a mass spectrum (FAB) was obtained from this: m/z 1565,

1173, 1144 and 939.

[{WCl₂(Ph₂PMe)₂(CO)}₂(ꢁ-m-N-4-MeC₆H₃N)] 1o. Reaction

of [WCl₂(Ph₂PMe)₄] (0.20 g, 0.19 mmol) and 4-methyl-meta-

phenylenediisocyanate IIo (0.017 g, 0.09 mmol) in toluene (35

cm³) for 1 d gave a claret solution. Removal of solvent under

reduced pressure aﬀorded an oily solid which was washed with

light petroleum. Redissolution in dichloromethane and addi-

tion of methanol gave purple complex 1o (0.13 g, 92%). IR

(KBr): 1966s (CO), 1961s (CO), 1481m, 1432m, 1340w, 1315w,

1271w, 1095m, 895s, 891s, 734m, 730m, 693s, 498m, 488w and

450m cm^Ϫ1. ¹H NMR (CDCl₃): δ 7.78–6.75 (m, 40 H, Ph), 6.14

(s, 2 H, Ar), 6.09 (s, 1 H, Ar), 2.26 (t, J 3.9, 6 H, Me), 2.24 (t,

J 4.0 Hz, 6 H) and 1.55 (s, 3 H, Me). ³¹P NMR (CDCl₃): δ 3.30

(s, J_PW291) and 2.20 (s, J_PW291 Hz). Mass spectrum (FAB): m/z

1456 (M^ϩϪ CO). Calc. for C₆₁H₅₈Cl₄N₂O₂P₄W₂ؒ0.5CH₂Cl₂:

C, 48.37; H, 3.89; Cl, 11.61; N, 1.83; P, 8.11. Found: C, 47.98;

H, 3.81; Cl, 11.67; N, 1.76; P, 8.44%.

Attempted coupling of complex 1p p-diiodobenzene. Activated

copper (0.04 g, 0.61 mmol) was added to a pyridine solution

(30 cm³) of complex 1p (0.10 g, 0.019 mmol) and para-

diiodobenzene (0.055 g, 0.01 mmol). After heating to 100 ЊC

a red solution was formed, which after removal of CuI by

ﬁltration and removal of volatiles under reduced pressure, gave

a red oil. Mass spectrum (FAB^ϩ): m/z 1634, 1435, 1246, 1046,

875, 855, 655 and 463.

X-Ray data collection and solution

Details of the crystal structure analyses of complexes 1a and 1n

have been communicated previously.⁹For all structures a single

crystal was mounted on a glass ﬁbre and all geometric and

intensity data were taken from this sample using an automated

[WCl₂(Ph₂PMe)₂(CO)(NC₆H₄I-p)] 1p. para-Iodophenyl iso-

cyanate IIp (0.10 g, 0.45 mmol) and [WCl₂(Ph₂PMe)₄] (0.43 g,

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

2721

View Article Online

Table 3 Crystallographic data for complexes 1a, 1b, 1g, 1h, 1n and 1p

1aؒC₆H₅Cl

1b

1g

1hؒCH₂Cl₂

1n

1p

Formula

C₆₆H₆₁Cl₅N₂O₂-

P₄W₂

C₆₁H₆₀Cl₄N₂O₂-

P₄W₂

C₆₈H₆₄Cl₄N₂O₄-

P₄W₂

C₆₅H₆₀Cl₆N₂O₂-

P₄W₂

C₆₀H₅₆Cl₄N₂O₂-

P₄W₂

C₃₃H₃₀Cl₂INOP₂W

Formula weight

Crystal system

Space group

a/Å

b/Å

c/Å

1583.12

1486.49

Monoclinic

P2₁/n

18.257(2)

8.855(2)

19.756(4)

1606.59

1605.43

1470.56

Monoclinic

P2₁/a

15.6810(28)

22.1749(50)

17.0619(50)

900.17

Orthorhombic

Pcan

15.8427(13)

17.674(2)

24.198(5)

Triclinic

¯

P1

10.1257(12)

11.5548(30)

14.7543(32)

102.280(19)

97.360(14)

91.770(16)

1669.5(6)

1

38.53

6204

4336

338

11.3261(2)

17.760(3)

18.226(3)

99.320(12)

102.220(13)

98.980(14)

3466.2(10)

2

36.09

8997

8994

510

11.865(3)

16.051(5)

17.561(8)

99.56(3)

97.73(3)

98.21(2)

3221(2)

2

39.61

9406

8871

715

α/Њ

β/Њ

107.130(13)

92.060(20)

γ/Њ

V/Å³

3052.0(10)

2

40.89

556.7

5382

343

0.031 (0.041)

0.064 (0.069)

5929(2)

4

42.91

11106

7353

667

6776(2)

8

46.00

5904

5883

370

0.045 (0.065)

0.110 (0.126)

Z

µ(Mo-Kα)/cm^Ϫ1

Data measured

Unique data used

No. parameters

R (all data)

RЈ (all data)

0.039

0.044

0.53 (0.067)

0.126 (0.136)

0.080 (0.099)

0.198 (0.209)

0.041

0.038

four-circle diﬀractometer (Nicolet R3mV) equipped with

Mo-Kα radiation (λ = 0.710 73 Å) at 293 2 K. The lattice

parameters were identiﬁed by application of the automatic

indexing routine of the diﬀractometer to the positions of a

number of reﬂections taken from a rotation photograph and

centred by the diﬀractometer. The ω–2θ or ω technique was

used to measure reﬂections and three standard reﬂections

(remeasured every 97 scans) showed no signiﬁcant loss in

intensity during data collection. The data were corrected for

Lorenz-polarisation eﬀects and empirically for absorption. The

unique data with I ≥ nσ(I) (n = 2 or 3) were used to solve and

reﬁne the structures.

further thank Drs G. D. Forster and D. A. Tocher for help

with crystallography and Bob Collins (Bayer UK Ltd.) for

generous provision of some isocyanates.

References

1 M. D. Ward, Chem. Soc. Rev., 1995, 121.

2 C. Creutz, Prog. Inorg. Chem., 1983, 30, 1.

3 W. Weng, T. Bartik and J. A. Gladysz, Angew. Chem., Int. Ed. Engl.,

1994, 33, 2199; M. Brady, W. Weng and J. A. Gladysz, J. Chem. Soc.,

Chem. Commun., 1994, 2655.

4 F. Coat and C. Lapinte, Organometallics, 1996, 15, 477 and refs.

therein.

Structures were solved by either heavy atom methods (1a) or

direct methods and developed by using alternating cycles of

least-squares reﬁnement and Fourier-diﬀerence synthesis. All

non-hydrogen atoms were reﬁned anisotropically. Hydrogens

were placed in idealised positions (C–H 0.96 Å) and assigned

a common isotropic thermal parameter (U = 0.08 Å²). Final

Fourier-diﬀerence maps were featureless and contained no

peaks greater than 1.00 e Å^Ϫ3. Structure solution used either the

SHELXTL⁵²or SHELXTL PLUS⁵³program packages on a

microVax or IBM PC respectively. Crystallographic data are

given in Table 3.

The structure of complex 1a contains a molecule of chloro-

benzene, which was reﬁned only isotropically. A disorder in the

solvent was successfully modelled by ﬁxing the occupancy of

the chloride [Cl(3)] to 50%. In 1b the methyl group on the aryl

ring was disordered over two sites [C(5) and C(5a)] the occu-

pancies of each being ﬁxed at 50%. In 1g the carbons of the

phenyl rings on phosphorus and the methoxy substituents on

the central diaryl moiety were reﬁned only isotropically, while a

number of atoms [O(2), C(8), C(9), O(52), C(53), C(54), C(58)]

were also disordered over two sites, occupancies being ﬁxed at

50%. In 1h a dichloromethane solvate molecule was reﬁned only

isotropically and a number of atoms [C(2), C(4), C(6), C(52),

C(54), C(56)] were disordered over two sites with a ﬁxed 50%

occupancy.

5 W. A. Nugent and J. M. Mayer, Metal–Ligand Multiple Bonds,

Wiley Interscience, New York, 1988.

6 T. P. Pollagi, S. J. Geib and M. D. Hopkins, J. Am. Chem. Soc., 1994,

116, 6051; H. A. Brison, T. P. Pollagi, T. C. Stoner, S. J. Geib and

M. D. Hopkins, Chem. Commun., 1997, 1263; B. E. Woodworth,

P. S. White and J. S. Templeton, J. Am. Chem. Soc., 1997, 119, 828;

B. E. Woodworth and J. L. Templeton, J. Am. Chem. Soc., 1996,

118, 7418; N. A. Ustynyuk, V. N. Vinogradova, V. G. Andrianov and

Yu. T. Struchkov, J. Organomet. Chem., 1984, 268, 73; K.-Y. Shih,

R. R. Schrock and J. Kempe, J. Am. Chem. Soc., 1994, 116, 8804;

S. A. Krouse and R. R. Schrock, J. Organomet. Chem., 1988, 335,

257.

7 E. O. Fischer, D. Wittmann, D. Himmelreich and D. Neugebauer,

Angew. Chem., Int. Ed. Engl., 1982, 21, 454; G. Jia, W. F. Wu, R. C.

Y. Yeung and H. P. Xia, J. Organomet. Chem., 1997, 539, 53;

A. Rabier, N. Lugan, R. Mathieu and G. L. Geoﬀroy,

Organometallics, 1994, 13, 4676; N. Le Narvor, L. Toupet and

C. J. Lapinte, J. Am. Chem. Soc., 1995, 117, 7129; H. Werner,

T. Rappert and J. Wolf, Isr. J. Chem., 1990, 30, 377; D. Unseld,

V. V. Krivykh, K. Heinze, F. Wild, G. Artus, H. Schmalle and

H. Berke, Organometallics, 1999, 18, 1525; R. S. Iyer and J. P.

Selegue, J. Am. Chem. Soc., 1987, 109, 910; R. L. Beddoes,

C. Bitcon, A. Ricalton and M. W. Whitley, J. Organomet. Chem.,

1989, 367, C21; R. L. Beddoes, C. Bitcon, R. W. Grime, A. Ricalton

and M. W. Whitley, J. Chem. Soc., Dalton Trans., 1995, 2873;

M. I. Bruce, M. P. Cifuentes, M. R. Snow and R. T. Tiekink,

J. Organomet. Chem., 1989, 359, 379; C. Löwe, H.-U. Hund and

H. Berke, J. Organomet. Chem., 1989, 372, 295.

8 D. E. Wigley, Prog. Inorg. Chem., 1994, 42, 239.

9 G. Hogarth, R. L. Mallors and T. Norman, J. Chem. Soc., Chem.

Commun., 1993, 1721.

CCDC reference number 186/1529.

10 M. Liang and E. A. Maatta, Inorg. Chem., 1992, 31, 953.

11 W. Clegg, R. J. Errington, D. C. R. Hockless, J. M. Kirk and

C. Redshaw, Polyhedron, 1992, 11, 381.

12 E. A. Maatta and D. D. Devore, Angew. Chem., Int. Ed. Engl., 1988,

27, 569; J. L. Stark, A. L. Rheingold and E. A. Maatta, J. Chem.

Soc., Chem. Commun., 1995, 1165.

13 W. Clegg, R. J. Errington, K. A. Fraser, S. A. Holmes and

A. Schäfer, J. Chem. Soc., Chem. Commun., 1995, 455.

14 G. Hogarth, T. Norman and S. P. Redmond, Polyhedron, 1999, 18,

1221.

Acknowledgements

We thank the EPSRC for funding of this work in the form of

an earmarked studentship (S. P. R.), a studentship (M.-Y. Lee)

and electrochemical equipment. We also thank University

College London for partial-funding (T. N.). D. G. H. thanks

the Royal Society for an Endeavour Fellowship and the

Ramsay Memorial Fund for a British Ramsay Fellowship.

N. K. thanks the Royal Society for equipment grants and we

15 E. A. Maatta and C. Kim, Inorg. Chem., 1989, 28, 624.

2722

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

View Article Online

16 G. Hogarth and S. P. Redmond, unpublished results.

17 A. Antiñolo, F. Carrillo-Hermosilla, A. Otero, M. Fajardo,

A. Garcés, P. Gómez-Sal, C. López-Mardomingo, A. Martín and

C. Miranda, J. Chem. Soc., Dalton Trans., 1998, 59.

18 D. E. Wheeler, J.-F. Wu and E. A. Maatta, Polyhedron, 1998, 17,

969.

19 R. K. Rosen, R. A. Anderson and N. M. Edelstein, J. Am. Chem.

Soc., 1990, 112, 4588.

20 Y.-W. Ge, Y. Ye and P. R. Sharp, J. Am. Chem. Soc., 1994, 116, 8384.

21 P. R. Sharp, Organometallics, 1984, 3, 1217.

22 (a) F.-M. Su, J. C. Bryan, S. Jang and J. M. Mayer, Polyhedron, 1989,

8, 1261; (b) K. A. Hall and J. M. Mayer, J. Am. Chem. Soc., 1992,

114, 10402; (c) J. C. Bryan, S. J. Geib, A. L. Rheingold and J. M.

Mayer, J. Am. Chem. Soc., 1987, 109, 2826.

37 S. Cenini, M. Pizzotti, F. Porta and G. La Monica, J. Organomet.

Chem., 1975, 88, 237.

38 J. Trotter, Acta Crystallogr., 1961, 14, 1135.

39 J. Almlö, Chem. Phys., 1974, 6, 135; O. Bastiansen and M.

Traettenberg, Tetrahedron, 1962, 17, 147.

40 J. L. Courtneidge, A. G. Davies, T. Clark and D. Wilhelm, J. Chem.

Soc., Perkin Trans. 2, 1984, 1197.

41 J. B. Flanagan, S. Margal, A. J. Bond and F. C. Anson, J. Am. Chem.

Soc., 1978, 100, 4248.

42 S. L. W. McWhinnie, C. J. Jones, J. A. McCleverty, D. Collison and

F. E. Mabbs, J. Chem. Soc., Chem. Commun., 1990, 940.

43 J. A. McCleverty, Chem. Soc. Rev., 1983, 12, 331; G. Denti,

C. J. Jones, J. A. McCleverty, B. D. Neaves and S. J. Reynolds,

J. Chem. Soc., Chem. Commun., 1983, 474.

23 A. P. Melissaris and M. H. Litt, J. Org. Chem., 1994, 59, 5918.

24 S. Takahashi, Y. Kuroyama, K. Sonogashira and N. Hagirara,

Synthesis, 1974, 424.

44 N. AlObaidi, D. Clague, M. Chaudhury, C. J. Jones, J. A.

McCleverty, J. C. Pearson and S. S. Salam, J. Chem. Soc., Dalton

Trans., 1987, 1733; T. E. Berridge, F. S. McQuillan, H. Chen, T. A.

Hamor and C. J. Jones, Inorg. Chem., 1998, 37, 4959.

45 J. A. Thomas, C. J. Jones, J. A. McCleverty, D. Collison, F. E.

Mabbs, C. J. Harding and M. G. Hutchings, J. Chem. Soc., Chem.

Commun., 1992, 1796.

46 R. J. H. Clark and D. G. Humphrey, Inorg. Chem., 1996, 35, 2053.

47 S. P. Best, S. A. Ciniawsky and D. G. Humphrey, J. Chem. Soc.,

Dalton Trans., 1996, 2945.

25 H. Eckert and B. Forster, Angew. Chem., Int. Ed. Engl., 1987, 26,

894.

26 H. Vittenet, Bull. Soc. Chim. Fr., 1899, 3, 956.

27 P. E. Fanta, Synthesis, 1974, 9; H. Tanaka, S. Sumida, N. Kobayashi,

N. Komatsu and S. Torii, Inorg. Chim. Acta, 1994, 222, 323.

28 M. P. Y. Yu, K.-K. Cheung and A. Mayr, J. Chem. Soc., Dalton

Trans., 1998, 2373.

29 N. Kaltsoyannis and P. Mountford, J. Chem. Soc., Dalton Trans.,

1999, 781 and refs. therein.

48 G. te Velde and E. J. Baerends, J. Comput. Phys., 1992, 99, 84.

49 ADF〈2.3〉, Department of Theoretical Chemistry, Vrije

Universiteit, Amsterdam, 1997.

50 S. H. Vosko, L. Wilk and M. Nusair, Can. J. Phys., 1980, 58, 1200.

51 For details of both MOLDEN and ADFrom, the reader is directed

to http://ww w .caos.kun.nl/~schaft/molden/molden.html

52 G. M. Sheldrick, SHELXTL, University of Göttingen, 1985.

53 G. M. Sheldrick, SHELXTL PLUS, Program package for structure

solution and reﬁnement, version 4.2, Siemens Analytical Instru-

ments Inc., Madison, WI, 1990.

30 D. C. Bradley, M. B. Hursthouse, K. M. A. Malik, A. J. Nielson

and R. L. Short, J. Chem. Soc., Dalton Trans., 1983, 2651.

31 T. Ziegler and A. Rauk, Theor. Chim. Acta, 1977, 46, 1.

32 T. Ziegler and A. Rauk, Inorg. Chem., 1979, 18, 1558.

33 P. D. Lyne and D. M. P. Mingos, J. Chem. Soc., Dalton Trans., 1995,

1635.

34 H. Taube and D. E. Richardson, Inorg. Chem., 1981, 20, 1278.

35 J. E. McGrady, T. Lovell and R. Stranger, Inorg. Chem., 1997, 36,

3242.

Paper 9/04001C

36 J. C. Bryan and J. M. Mayer, J. Am. Chem. Soc., 1990, 112, 2298.

J. Chem. Soc., Dalton Trans., 1999, 2705–2723

2723

Article Doi

Linking metal centres with diimido ligands: Synthesis, electronic and molecular structure and electrochemistry of organometallic ditungsten complexes [{WCl2(Ph2PMe)2(CO)}2(N-X-N)] (X = π-conjugated organic)

DOI: 10.1039/a904001c

Source and publish data:

Authors:

Article abstract of DOI:10.1039/a904001c

Full text of DOI:10.1039/a904001c

Products guided by the article

R&D Labs maybe for 76806-38-7

Relevant to this article

Hot Product