Complexes with phosphorus analogues of imidazoyl carbenes: Unprecedented formation of phosphenium complexes by coordination induced P-Cl bond heterolysis

Article abstract of DOI:10.1016/S0022-328X(00)00624-0

Diazaphospholenium complexes are readily accessible from reactions of complexes [M(bipy)(CO)₃(L)] (L=CO, MeCN, M=Mo, W) with both the 1,3-dimesityl-4-chloro-1,3,2-diazaphospholenium triflate (1) [OTf] and the corresponding p-chloro-diazaphospholene (8). The latter reaction proceeds via an unprecedented coordination induced ionisation of P-Cl bonds, which requires no further assistance by an external electrophile. The complexes were found to be configurationally stable, but may undergo selective substitution of trans-ligands with retention of the phosphenium moiety. All compounds were characterised by analytical and spectroscopic techniques, and two of the complexes were investigated by single-crystal X-ray diffraction. The spectroscopic and structural data provide evidence for considerable M-P double bond character which, leads to a marked reduction of π-delocalization in the diazaphospholenium unit. Studies of metal NMR spectra of tungsten complexes revealed further, a linear correlation between δ¹⁸³W and ¹J_WP which allows monitoring of trends in metal-phosphorus multiple bonding. Surprisingly, spectroscopic and structural data suggest that the cation 1 displays a higher π-acceptor ability than an analogous CC-saturated heterocyclic phosphenium ligand, which contrasts with the lower electrophilicity of 1 with respect to unconjugated diaminophosphenium species.

Full text of DOI:10.1016/S0022-328X(00)00624-0

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 617–618 (2001) 383–394
Complexes with phosphorus analogues of imidazoyl carbenes:
unprecedented formation of phosphenium complexes by
coordination induced PꢀCl bond heterolysis
Dietrich Gudat *, Asadollah Haghverdi, Martin Nieger
Institut fu¨r Anorganische Chemie der Uni6ersita¨t Bonn, Gerhard-Domagk-Strasse 1, 53121 Bonn, Germany
Received 26 July 2000; accepted 1 September 2000
Dedicated to Professor Manfred Regitz on the occasion of his 65th birthday
Abstract
Diazaphospholenium complexes are readily accessible from reactions of complexes [M(bipy)(CO)3(L)] (L=CO, MeCN,
M=Mo, W) with both the 1,3-dimesityl-4-chloro-1,3,2-diazaphospholenium triflate (1) [OTf] and the corresponding p-chloro-di-
azaphospholene (8). The latter reaction proceeds via an unprecedented coordination induced ionisation of PꢀCl bonds, which
requires no further assistance by an external electrophile. The complexes were found to be configurationally stable, but may
undergo selective substitution of trans-ligands with retention of the phosphenium moiety. All compounds were characterised by
analytical and spectroscopic techniques, and two of the complexes were investigated by single-crystal X-ray diffraction. The
spectroscopic and structural data provide evidence for considerable MꢀP double bond character which, leads to a marked
reduction of p-delocalization in the diazaphospholenium unit. Studies of metal NMR spectra of tungsten complexes revealed
1
further, a linear correlation between l183W and JWP which allows monitoring of trends in metalꢀphosphorus multiple bonding.
Surprisingly, spectroscopic and structural data suggest that the cation 1 displays a higher p-acceptor ability than an analogous
CC-saturated heterocyclic phosphenium ligand, which contrasts with the lower electrophilicity of 1 with respect to unconjugated
diaminophosphenium species. © 2001 Elsevier Science B.V. All rights reserved.
Keywords: Carbene analogues; Phosphenium complexes; p-Acceptor ligands; Metalꢀligand multiple bonding; Bond heterolysis
plexes have been devised [2], one of the most funda-
mental routes, viz. direct reaction of a transition metal
1. Introduction
precursor complex with a free carbene, has long re-
Transition metal carbene complexes [1] are known
mained unexplored owing to the lack of sufficiently
for some 40 years, and their unique reactivity have
stable carbenes. The situation changed merely some 10
made them indispensable reagents in organometallic
years ago when the groups of Bertrand [3] and Ar-
chemistry, organic synthesis, and catalysis [2]. Even
duengo [4] described the first carbene derivatives of the
though various synthetic approaches to carbene com-
type I and II which could be isolated as stable com-
pounds and were accessible on a preparative scale. This
discovery triggered extensive studies of the synthesis of
new stable carbenes and the exploration of their elec-
tronic structures and chemical reactivities, including
their potential for complex formation. The major
achievements in this area during the last years have
been recently comprehensively reviewed [2,5] (Scheme
1).
The renaissance of carbene chemistry has also en-
couraged studies of carbene analogues which are ob-
Scheme 1.
* Corresponding author. Tel.: +49-228-735334; fax: +49-228-
735327.
E-mail address: dgudat@uni-bonn.de (D. Gudat).
0022-328X/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(00)00624-0

384
D. Gudat et al. / Journal of Organometallic Chemistry 617–618 (2001) 383–394
2. Results and discussion
2.1. Reactions
Known synthetic routes to phosphenium complexes
[12,13] include direct synthesis from a stable phosphe-
nium ion and a suitable transition metal fragment,
metathesis of a phosphanide with a metal halide, or
electrophile induced cleavage of an alkoxide from a
co-ordinated phosphinite (Scheme 2(a–c)). Base in-
Scheme 2. Known synthetic pathways to phosphenium complexes.
duced dehydrochlorination of complexes with
a
ClꢀPꢀMꢀH moiety has been utilised for the generation
of neutral phosphenium complexes (Scheme 2(d)) [13].
The formation of cationic complexes by electrophile
promoted halide abstraction from coordinated
halophosphines (Scheme 2(e), X=F, Cl) which works
well in the synthesis of free phosphenium ions [12,13]
was successfully demonstrated in the case of Fe(CO)4-
complexes [14] but reportedly failed for W(CO)5-com-
plexes [15]. Regarding that the back-donation
capability of a [M(CO3)(bipy)] is larger than that of a
[W(CO)5] fragment and that the p-chloro-1,3,2-diaza-
phospholenes exhibit extraordinarily labile PꢀCl bonds
[7c,9b,c], we nonetheless thought this approach worth
considering, and thus pursued both strategies (a) and
(e) for the synthesis of the target complexes.
A survey of the reactions of [M(bipy)(CO)3(L)] with
different available diazaphospholenium salts and p-
chloro-diazaphospholenes [9c] revealed that complexa-
tion occurred only for compounds with N-aryl
substituents whereas N-tert-butyl substituted deriva-
tives failed to react. A similar behaviour had been
observed previously for the reactions of diazaphospho-
lenium ions with silver salts [10]. Regarding, that varia-
tion of the N-substituents has only limited effects on
structure and bonding in the free cations [9c], this
difference in reactivities is presumably of steric rather
than of electronic origin.
Successful direct syntheses of diazaphospholenium
complexes was achieved by heating toluene suspensions
of equimolar amounts of the triflate (1) [OTf] [16] and
[M(bipy)(CO)4] (M=Mo (2a), W (2b). Both reactions
proceeded with eventual dissolution of the starting ma-
terials and formation of red crystalline precipitates of
the triflato complexes 3a,b which were isolated by filtra-
tion and characterised by spectroscopic and analytical
techniques (see below). Realising that the formation of
3a,b required elimination of two carbonyl ligands from
the starting materials, we anticipated that the reaction
might proceed in a two step mechanism. To corrobo-
rate this hypothesis, we carried out the complexation
under milder conditions by reacting 1 [OTf] with
[W(bipy)(CO)3(MeCN)] (4b) in CH2Cl2 at ambient tem-
perature and obtained a single product which was
isolated after evaporation of volatiles and identified
spectroscopically as the fac-isomer of the tricarbonyl
tained by formal isoelectronic replacement of the diva-
lent carbon by another p-block element. Most of these
investigations focused on the variation of the structural
motif of imidazoyl carbenes, II, and a wealth of subva-
lent main group compounds III are now known which
include both neutral compounds with heavier Group 14
elements (E=Si, Ge) [6] and ionic species containing a
positively or negatively charged element of an adjacent
main group (E=Ga−, N+, P+) [7]. A common feature
of all these derivatives which stimulated intensive re-
search is the presence of six p-electrons in the ring
which offer the possibility of aromatic stabilisation. It
has, meanwhile, been established that (i) aromaticity
contributes far less to the overall stability than N“E
p-donation into the empty p-orbital at the divalent
element [8]; and (ii) the extent of cyclic p-delocalisation
in phosphenium cations (III, E=P) is larger than in
neutral carbene analogues and renders these cations
more stable and less electrophilic than acyclic or satu-
rated cyclic diamino phosphenium ions [7c,9].
As compared to the extensive knowledge on carbene
complexes with ligands of type II [5], the coordination
chemistry of the analogous main group compounds III
is far underdeveloped. Some initial reports on silylene
and germylene complexes have been published [10], but
the coordination chemistry of cationic nitrogen and
phosphorus compounds is, apart from a study of the
reactions of diazaphospholenium ions (III, E=P+)
with Ag+ [11], largely unexplored.
In this work we report on a study of the reactivity of
diazaphospholenium cations towards molybdenum and
tungsten carbonyls [M(bipy)(CO)3L] (M=Mo, W;
bipy=2,2%-bipyridine, L=CO, MeCN). Regarding the
importance of metal-to-ligand back bonding for the
stability and physical and chemical properties of pho-
sphenium complexes [12], the selected metal complexes
were chosen in order to provide sufficient back-dona-
tion ability. A main focus of the presented work will
deal with the characterisation of the interplay between
(aromatic) intra-ligand and metal-to-ligand p-donor–
acceptor interactions for the stabilisation of the for-
mally electron-deficient phosphenium centre.

D. Gudat et al. / Journal of Organometallic Chemistry 617–618 (2001) 383–394
385
and characterised in situ by means of 31P- and 183W-
NMR spectroscopy when 8 was reacted with 4b at
ambient temperature in CH2Cl2. In contrast to 5b
[OTf], the intermediate was in this case unstable and
decomposed slowly to the final product, 9b, thus pre-
cluding all attempts towards its isolation. Based on the
similarity of the 31P- and 183W-NMR data with those of
5b [OTf] (see Table 1), the transient species was as-
signed as the tricarbonyl complex 5b [Cl]. The slight
complex 5b [OTf]. Solutions of 5b [OTf] in CH2Cl2 were
stable towards CO elimination up to 40°C, but clean
conversion into 3b was observed upon heating a sus-
pension of the complex in boiling toluene. Refluxing 5b
[OTf] in acetonitrile lead to the quantitative formation
of a new product which could not be isolated in analyt-
ically pure form, but whose constitution as the cationic
nitrile complex 6b [OTf] was unequivocally established
from in situ IR and multinuclear 1D and 2D NMR
investigations. A further chemical proof for the postu-
lated constitution was obtained from the reaction with
triphenylphosphine which gave the corresponding com-
plex 7b via selective displacement of the coordinated
nitrile and retention of the phosphenium moiety. The
same products as well as its molybdenum congener 7a
were alternatively accessible from the reaction of PPh3
with 3a,b (Scheme 3).
1
differences in l31P and JWP between both complexes
are due presumably to the stronger nucleophilicity of a
chloride as compared to a triflate anion, which leads to
ion pairing in solution. This hypothesis was corrobo-
rated by the observation that portionwise addition of
[Et4N][Cl] to a CH2Cl2 solution of 5b [OTf] led to a
continuous shift of the 31P-NMR signal. Reaction of
9a,b with trimethylsilyl triflate afforded further a quan-
titative yield of trimethylchlorosilane and the triflate
complexes 3a,b.
According to the spectroscopic investigations, all di-
azaphospholenium complexes exist in solution as a
single stereoisomer (the fac-isomer for 5b, and the
trans-isomers for 3a,b, 6b, 7a,b, and 9a,b) and showed
Complexes 2a,b were found to react likewise with the
p-chloro-diazaphospholene (8) in boiling toluene to af-
ford the phosphenium complexes 9a,b with
metalꢀchlorine bonds. Both products precipitated from
the reaction mixture and were isolated in good yields
after filtration. Again, a transient species was detected
Scheme 3. Synthesis of diazaphospholenium complexes (L=MeCN).

386
D. Gudat et al. / Journal of Organometallic Chemistry 617–618 (2001) 383–394
Table 1
31P,183W, and IR (wCO) data of complexes [M(1)(bipy)(CO)2(L)] in CH2Cl2 solution
a
M
L
l
31P
Dl31P
l
183W
1JWP (Hz)
w(CO) (cm−1
)
1
205.5
3a
3b
Mo
W
W
W
W
Mo
W
Mo
W
OTf
OTf
CO
167.0
137.3
161.8
160.1
−38.5
−68.2
−43.7
−45.4
−47.4
−33.1
−55.5
−60.6
−83.2
1932, 1853
1921, 1841
2015, 1932, 1877
−1440
−2105
−2100
−1580
770
497
494
707
5b [OTf]
5b [Cl]
6b [OTf]
7a [OTf]
7b [OTf]
9a
CO
MeCN
PPh3
PPh3
Cl
148.1
1911, 1845
1917, 1846
1911, 1838
1923, 1844
1913, 1834
172.4 33.7 b
150.0 24.3 c
145.1
−1860
−1680
598, 211
703
9b
Cl
122.3
a Coordination shifts: Dl31P=l31P(complex)−l31P(1).
b 2JPP=207 Hz.
c 2JPP=208 Hz.
no tendency towards isomerisation to the correspond-
ing mer- or cis-isomers which had been observed for
the complexes 10 containing phosphenium ligands with
CC-saturated 5-membered rings (Scheme 4) [17]. For
these compounds, the relative stabilities of both
stereoisomers were proven to be controlled by a bal-
ance of electronic and steric factors [17]: the mer/cis-
isomers gain thermodynamic stability over the
corresponding fac/trans isomers from the tendency of
the highly p-acidic phosphenium ligand to occupy a
position trans to the bipy ligand which exhibits the
lowest p-acceptor ability, whereas steric factors enhance
the preference of trans over cis isomers. In view of these
findings, the configurational stability of 3, 5–7, and 9 is
presumably attributable to the increased size of the
N-mesityl-diazaphospholenium units as compared to
the N-methylated heterocycles in 10.
The formation of the chloro complexes 9a,b from the
p-chloro-diazaphospholene (8) is remarkable since the
PꢀCl heterolysis (which was directly observed during
the formation of 5b [Cl] from 8 and 4b) proceeds
spontaneously without support of a Lewis-acid. We
relate this unusual reactivity to the combined effects of
‘preorganisation’ of the ligand 8 due to the presence of
an intrinsically weak PꢀCl bond, and the strong p-ba-
sicity of the metal fragment 4b. The directing influence
of phosphenium substituents to activate a carbonyl
ligand in the trans-position for substitution by donor
ligands of lower p-acidity has been noted previously
[17,18].
bonyl complexes displayed only a single resonance,
indicating both carbonyl ligands being trans to a
bipyridine nitrogen atom and thus necessarily cis with
respect to each other. The trans-arrangement of the
remaining ligands was in the case of 7a,b further sup-
2
ported by characteristically large values of JPP (207–
208 Hz) between the phosphorus atoms of the
diazaphospholenium and triphenylphosphine ligands
which are of similar magnitude as the coupling in the
trans-isomers of 10c–e [17]. Some compounds dis-
played a slight broadening of the resonances of the
carbonyl as well as the 1H and 13C signals of the
bipyridine ligand. Even though the origin of this effect
was not elucidated in detail, we take it as indication for
the onset of hindered rotation of the diazaphosphole-
nium moiety around the PꢀM bond which slows down
the dynamic interconversion between conformers with
different angular orientation of the unsymmetric diaza-
phospholenium moiety with respect to the M(bipy)-
(CO)2 set of ligands.
The spectrum of the tricarbonyl complex 5b displays
two carbonyl signals with relative intensities of 1:2. A
distinction between fac- and mer-arrangement of the
CO ligands was feasible from the observation of a
2
substantially larger value of JPC (84.3 vs. 12.2 Hz) for
the less intense signal. This suggested the unique car-
bonyl ligand to be trans to the phosphenium moiety
2.2. Spectroscopic in6estigations
The constitution of all diazaphospholenium com-
plexes was established from analytical and spectro-
scopic data. Assignment of the metal stereochemistry
was derived in all cases from the 13C-NMR spectra,
with the important information being provided by the
signals attributable to the CO ligands. Thus, all dicar-
Scheme 4. Fac–mer/cis–trans stereoisomerization of phosphenium
complexes [17]. (L=CO, M=Mo (10a), W (10b); L=PPh3, M=
Mo (10c); L=2-methoxy-1,3-dimethyl-1,3,2-diazaphospholidine,
M=Mo (10d), W (10e)).

D. Gudat et al. / Journal of Organometallic Chemistry 617–618 (2001) 383–394
387
dominates the trends in shielding for heavy nuclei) and
1
the reduced spin-spin coupling constants KMP (1KMP
=
4y2/(hkMkP)1JMP) according to [19b,c]
|
para8(1DE−1)(Br−3
\ \mdDu)
npPu+ Br−3
1KMP8(3DE−1)s2(0)Ps2(0)MPM2 P
where [1,3]DE denotes an average electronic excitation
energy, Br−3\ an orbital expansion term which is
related to the nephelauxetic effects, s(0)X the s-electron
density at a nucleus X, and P, D, p- and d-orbital
population terms and bond orders (often summoned to
represent the ‘imbalance of electron density’), respec-
tively. Even if a concise theoretical analysis is yet
lacking, qualitative arguments [17,20] support the as-
sumption that changes in the electronic situation of the
phosphorusꢀmetal p-bonding exert a strong influence
on both DE and the orbital/bond order terms P, D
whose variations dominate presumably the shifts in
spara and K in the structurally closely related molecules
considered here. Assuming the validity of these argu-
1
Fig. 1. Correlation between l183W and JWP for complexes [M(diaza-
phospholenium)(bipy)(CO)2(L)]. Diamonds denote experimental data;
the straight line corresponds to the result of a least-squares fit which
gave 1JWP=0.426, l183W+1393 with a correlation coefficient r=
0.987.
and thus permitted identification of the complex as the
fac-isomer. This assignment was confirmed by the ob-
1
ments, the observed magnitudes of JWP suggest that
despite the lower electrophilicity of free 1 with respect
to other aminophosphenium ions [9c] the degree of
M“L back donation, and thus the phosphorusꢀmetal
p-bonding character, in the studied diazaphosphole-
nium complexes is still significant.
1
servation of only one set of resonances for the H and
13C nuclei in both rings of the bipyridine unit which is
incompatible with the constitution of a mer-isomer
featuring one pyridine nitrogen cis and the other one
trans to the phosphenium ligand.
In addition to the interpretation of NMR data, com-
parison of CO stretching frequencies is a well estab-
lished alternative to assess the p-acidity of co-ligands in
mixed carbonyl complexes. Although quantitative anal-
ysis is precluded by the unavailability of a complete set
of data which would allow concise evaluation of force
constants and the inapplicability of concepts such as
Tolman’s electronic parameter [21], comparison of the
wCO frequencies of complexes [M(bipy)(CO)2(L)-
(PR2)]+ which differ only in the R2P ligand should still
provide for a qualitative comparison of the properties
of different phosphenium moieties. In fact, comparison
of the spectroscopic data of 5b [OTf] (wCO=2015,
1932, 1877 cm−1) with those of [17] [W(bipy)(CO)3(1,3-
dimethyl-1,3,2-diazaphospholidinium)]+ (10b, wCO=
The 31P-NMR data listed in Table 1 reveal negative
coordination shifts whose magnitudes (Dl= −33 to
−60 for Mo and −44 to −83 for W complexes)
exceed those of the phosphenium complexes 10 (Dl=
−12 to −20 for Mo and −30 to −42 for W com-
1
plexes [17]). The J183
coupling constants are
31
significantly larger than tWhose typically found for phos-
phine complexes (1JWP=200–300 Hz; cf. 1JWP=211
Hz for the PPh3 ligand in 7b) and match the character-
istically large couplings of known phosphenium com-
plexes (1JWP=400–620 Hz [12,13,17,18]). The coupling
P
1
in 5b is larger than in 10b (M=W, L=CO, JWP=442
Hz [17]) and increases further with decreasing p-acidity
of the ligand in the trans-position in the order of
5bB7bB6b, 9bB3b. Comparison of the values of
1JWP and l183W which have been obtained from 31P
detected 31P, 183W 2D-HMQC spectra [19] reveal a
linear correlation between both quantities, with increas-
2006, 1920, 1868 cm−1
)
and [W(bipy)(CO)3
(2 - methoxy - 1,3 - dimethyl - 1,3,2 - diazaphospholidine)]
(11, wCO=1909, 1814, 1790 cm−1) reveals a strong
blue shift of wCO for both phosphenium complexes
with respect to 11 which reflects the much higher p-
acidity of phosphenium as compared to phosphite lig-
ands. The blue shift is slightly larger for 5b than for
10b, suggesting that the diazaphospholene 1 is a
stronger p-acceptor than the CꢀC saturated diazaphos-
pholidine ligand in 10b. The same trend becomes ap-
parent from the comparison of wCO in 7a [OTf]
(wCO=1917, 1831 cm−1) and [Mo(bipy)(CO)2(PPh3)-
(1,3 - dimethyl - 1,3,2 - diazaphospholidinium)]+ (10c,
wCO=1909, 1831 cm−1 [17]). The higher p-acidity
1
ing values of JWP being accompanied by a deshielding
of the metal nucleus. A similar qualitative relation
between 1JMoP and l95Mo has been reported for a series
of complexes [Mo(bipy)(phosphenium)(CO)2(L)]+ (L=
phosphites, CO) [20] (Fig. 1).
1
The observed trends in JMP (and l M) are generally
attributed to an increasing degree in MꢀP double bond
character [12,13,17,20]. A rationale for this behaviour
can be given by representing the paramagnetic contri-
bution to the magnetic shielding |para (whose variation

388
D. Gudat et al. / Journal of Organometallic Chemistry 617–618 (2001) 383–394
Table 2
of the cation 1 as compared to the CꢀC saturated
analogue, [(CH2NMe)2P]+, is at a first glance surprising
in the light of the fact that diazaphospholenium ions
are generally much weaker electrophiles than acyclic or
saturated cyclic diaminophosphenium ions [7c,9]. How-
ever, regarding that introduction of an electron with-
drawing Cl substituent at the diazaphospholene ring
has been shown to enhance the Lewis acidity of the
,
Relevant bond lengths (A) and bond angles (°) of the phosphenium
complexes 3a, 7b [OTf], 10d, and the 1,3,2-diazaphospholenium
cation 1a a
3a
7b (OTf)
1 [9c]
10d (OTf)
[16b]
Bond lengths
PꢀN
1.697(4)
1.682(4)
1.397(6)
1.404(6)
1.318(7)
1.436(6)
1.435(5)
1.671(7)
1.678(7)
1.389(10)
1.366(9)
1.303(11)
1.450(9)
1.438(10)
1.6662(15)
1.6708(15)
1.374(2)
1.378(2)
1.351(2)
1.460(2)
1.464(2)
1.642(5),
1.644(5)
NꢀC
CꢁC
NꢀCexo
Bond angles
MꢀP
2.2016(12)
2.247(2)
2.254(1)
MꢀC(CO)
1.969(5)
1.964(5)
2.230(4)
2.231(4)
2.250(3)
176.88(8)
1.956(9)
1.973(9)
2.219(6)
2.224(6)
2.543(2)
167.80(7)
1.953(6)
1.966(6)
2.244(4)
2.255(5)
2.495(1)
166.85(5)
MꢀN(bipy)
MꢀX
PꢀMꢀX
a NꢀCexo denotes the exocyclic NꢀC(mesityl) bond and X the
ligating atom of the ligand trans to the phosphenium fragment.
phosphenium centre [9c], the observed sequence of lig-
and p-acidities is presumably a consequence of different
inductive substituent effects.
Fig. 2. Molecular structure of 3a in the crystal, ORTEP view thermal
ellipsoids are at the 50% probability level, H atoms omitted for
clarity; selected bond distances and angles are given in Table 2.
2.3. Crystal structure studies
Single crystal structure determinations were carried
out for 3a and 7b [OTf]. Both crystalline phases contain
isolated neutral or cationic complexes, which displayed
no unusual short intermolecular distances. In both
compounds the Cl atom at the diazaphospholene ring
was found to be disordered between two positions with
relative occupancy factors of 70:30 (3a) and 76:24 (7b
[OTf]), respectively. We interpret this in terms of a
superposition of two different molecules, which differ in
the rotational orientation of the diazaphospholene moi-
ety around the PꢀM bond. Inspection of the final
atomic coordinates and anisotropic displacement
parameters indicated that the positions of all ring
atoms are sufficiently well defined to extract reliable
bonding parameters. Drawings of the complexes are
depicted in Figs. 2 and 3, respectively, and a reduced
representation of the metal coordination environments
is given in Fig. 4. Important bond distances are com-
pared with those of the free ligand 1 [9c] and the
complex trans-10d [17b] in Table 2.
Fig. 3. Molecular structure of the cation 7b in the crystal, ORTEP view
thermal ellipsoids are at the 50% probability level, H atoms omitted
for clarity; selected bond distances and angles are given in Table 2.
Both 3a and 7b have distorted octrahedral coordina-
tion spheres around the metal atom. The metal lies in
one plane with the ligating carbon and nitrogen atoms
of the CO and bipy ligands and the remaining MꢀP and
MꢀO bonds are close to perpendicular to this plane.
Fig. 4. Representation of the metal coordination spheres in complex
3a (left) and 7b (right). C1A, C1B and N1C, N12C denote the ligating
carbon and nitrogen atoms of the carbonyl and bipyridine ligands,
respectively.

D. Gudat et al. / Journal of Organometallic Chemistry 617–618 (2001) 383–394
389
The MꢀC and MꢀN distances in 3a, 7b, and 10d are
indistinguishable within experimental error, in accor-
dance with the assumption of practically identical cova-
lence radii for both Mo and W in the present
coordination enviroment. The diazaphospholene rings
in both complexes are planar (mean deviations of ring
these effects can be interpreted by assuming that the
strong MꢀP p-interaction introduces a massive reduc-
tion in cyclic intraligand p-delocalisation. Structure
comparisons involving the complexes 7b and 10d and
the corresponding free phosphenium ligands [23] sug-
gest that the PꢀN bond weakening which accompanies
the development of MꢀP p-bonding in phosphenium
complexes [12,17b] is more pronounced in 3a and 7b
than in 10d, again indicating a greater p-accepting
power of 1 as compared to the CC-saturated phosphe-
nium ligand in 10d.
,
atoms from the best-fit planes are 0.03(0.01) A) and the
sum of valence angles around the P1 atom amounts to
360 (359)° for 3a (7b). The angle between the diaza-
phospholene ring plane and a line bisecting the
C1AꢀMꢀC1B angle enclosing the two metalꢀcarbonyl
bonds is 21° for 3a and 102° for 7b. As a consequence,
the relative orientations of the rings, with respect to the
remaining ligands in both complexes, are close to or-
thogonal, reflecting no particular preference for a cer-
tain conformation and suggesting that the rotational
barrier around the PꢀM bond may be similarly low as
in carbene complexes. The MꢀP1(phosphenium) dis-
3. Conclusions
Diazaphospholenium complexes are readily accessi-
ble from reactions of complexes 2a,b and 4b with both
the cation 1 and the p-chloro-diazaphospholene (8).
The latter reaction involves an unprecedented coordina-
tion induced ionisation of PꢀCl bonds. In contrast to
known phosphenium syntheses, this reaction is pro-
moted by strong M“P charge transfer from the highly
p-basic transition metal fragment in combination with a
suitable ‘preorganisation’ (due to intrinsic, electroni-
cally induced PꢀCl bond weakening) of the ligand and
requires no further assistance by an external
electrophile.
Spectroscopic and structural studies give evidence for
the presence of considerable MꢀP double bond charac-
ter in the formed complexes which leads to a marked
reduction of p-delocalisation in the diazaphospholene
ring of the ligand. This suggests a low ‘hardness’ of the
diazaphosphenium moiety which is in contradiction
with a high degree of aromatic character [24] and gives
further support to the results of our previous investiga-
tion of this issue [9c]. The structural studies indicate
further that, as in the case of carbene complexes, no
peculiar preference for a fixed configuration of the
formal MꢀP double bond prevails. Spectroscopic and
structural data indicate a higher p-acidity for the diaza-
phospholenium as compared to a CC-saturated hetero-
cyclic phosphenium ligand which contrasts with the
lower electrophilicity of 1 with respect to non-conju-
gated diaminophosphenium species [9c]. Even if a con-
cise investigation of the origin of this discrepancy has
to be postponed to future studies, it is presumably
explained by tuning of the ligand electrophilicities by
subtle inductive substituent effects [9c].
,
tances increase in the order 3a (2.2016(12) A)ꢀ7b
,
,
(2.247(2) A)B10d (2.254(1) A [17b]) and the observed
shortening of the P1ꢀM with respect to the
P2(phosphine)ꢀM bond is more pronounced in 7b
(12%) than in 10d (10% [17b]). These trends are consid-
ered to reflect mainly the varying electronic influences
of the trans-ligands: the presence of a triflate, which is
both a weak p-donor and a negligible p-acceptor, leads
to distinct contraction of the MꢀP(phosphenium) bond,
and the order of MꢀP distances in 7b and 10d is in
accordance with a greater p-accepting power of the
phosphite type ligand in 10d as compared to PPh3. The
pronounced MꢀP1 bond shortening in combination
with the planar phosphorus coordination environment
is certainly indicative of considerable MꢀP double bond
character, and, assuming that the structural parameters
of 3a and the analogous tungsten complex 3b should be
similar, corroborates the conclusions derived from the
discussion of spectroscopic data.
The PꢀN, CꢀN, and CꢀC bonds in the diazaphospho-
lene rings of 3a and 7b are indistinguishable within
experimental error. Comparison with the corresponding
bond distances of free 1 [9c] (Table 2) reveals a length-
ening of PꢀN and CꢀN bonds and a slight shortening of
the CꢀC bond, and the observed distances come close
to those found for the corresponding 4-chloro-1,3-
dimesityl-2-hydrido-1,3,2-diazaphospholene [22] (PꢀN
1.709(3), 1.722(3), CꢀN 1.407(5), 1.410(5), C=C
,
1.327(5) A) with a pyramidal phosphorus atom. The
exocyclic NꢀC(mesityl) bonds in 3a and 7b are slightly
shorter than that in 1 so that on the whole, the marked
deviation between endocyclic and exocyclic NꢀC bonds
emanating from the same nitrogen atom which was
observed for 1 [9c] becomes less pronounced. Compari-
son of the PꢀN distances in 3a, 7b and 10d reveals that
the bond lengthening associated with the introduction
of CC unsaturation in the ring of the free ligand 1 [9c]
is further reinforced upon complexation. Altogether,
Last, but not least, the high p-accepting power of a
coordinated diazaphospholenium unit facilitates the
substitution of ligands in the trans-position and allows
easy access to reactive phosphenium complexes contain-
ing weakly coordinated nitrile or triflate ligands which
promise further interesting reactivities.

390
D. Gudat et al. / Journal of Organometallic Chemistry 617–618 (2001) 383–394
1
4. Experimental
4.1. General remarks
yield 86%, m.p.\260°C. H-NMR (CD2Cl2): 1.96 [s,
6H, o-Me], 2.11 [s, 6H, o-Me], 2.22 [s, 3H, p-Me], 2.26
[s, 3H, p-Me], 6.39 [br, 1H, m-H(bipy)], 7.75 [br, 4H,
p-H(bipy) and m-H(bipy)], 8.31 [br, 2H, o-H(bipy)].
13C{1H}-NMR (CD2Cl2): 18.1 [br, o-Me], 21.0 [s, p-
Me], 21.1 [s, p-Me], 118.8 [s, 4-C], 119.4 [s, 5-C], 123.0
[s, m-CH(bipy)], 125.6 [s, m-CH(bipy)], 128.9 [s, m-
CH], 129.1 [s, m-CH], 131.4 [d, 2JPC=6.0 Hz, i-C],
All manipulations were carried out under dry argon.
Solvents were dried by standard procedures. Com-
pounds 1 [OTf], 8 [9c] 2a,b, 4b [17], were prepared as
described. NMR spectra: Bruker AMX300 (1H, 300.1
MHz; 31P, 121.5 MHz; 13C, 75.4 MHz; 183W, 12.5
MHz) and Bruker AMX500 (1H, 500.1 MHz; 13C, 125.7
MHz); chemical shifts referenced to ext. TMS (1H, 13C),
85% H3PO4 (]=40.480747 MHz), aq. WO24− (]=
4.166388 MHz); 183W chemical shifts were determined
from 31P detected 31P-, 183W{1H}-HMQC spectra [19];
positive signs of chemical shifts denote shifts to lower
frequencies, and coupling constants are given as abso-
lute values; prefixes i-, o-, m-, p- denote atoms of
mesityl and phenyl substituents, and atoms in the di-
azaphospholene ring are denoted as 4-C, 5-H, etc.;
assignment of resonances were derived in ambiguous
2
3
134.0 [d, JPC=6.5 Hz, i-C], 137.1 [d, JPC=2.3 Hz,
o-C], 137.2 [s, p-CH(bipy)], 138.0 [d, 3JPC=2.3 Hz,
5
5
o-C], 138.3 [d, JPC=1 Hz, p-C], 138.5 [d, JPC=1 Hz,
p-C], 151.9 [s, o-CH(bipy)], 155.3 [s, o-C(bipy)], 223.1
[d, 2JPC=11.4 Hz, CO]. FAB-MS: m/z (%): 790 (70)
[M+]. Anal. Calc. for C32H31Cl2N4O2PW (789.4): C,
48.69; H, 3.96; N, 7.10. Found: C, 46.99; H, 4.08; N,
7.11%.
4.3. Preparation of complexes 3a,b
cases from analysis of 2D H,13C-HMQC and HMBC,
(a) A solution of 1 [OTf] (1 mmol) and the appropri-
ate complex 2a or 2b (1 mmol) in 30 ml of toluene was
refluxed for 24 h. After cooling, the formed precipitate
was filtered over a glass filter frit and washed with cold
toluene.
(b) Trimethylsilyltriflate (2 mmol) was added drop-
wise to a cooled (4°C) solution of 9a,b (1 mmol) in 50
ml of CH2Cl2. The solution was stirred for 1 h and then
allowed to warm to ambient temperature. Volatiles
were removed in vacuo, and the remaining residue was
washed twice with toluene–hexane (2:1). and added
dropwise to 4°C for 24 h. The crude products were
purified by recrystallisation from THF at −20°C.
trans-[(2,2%-Bipyridine)-dicarbonyl-(1,3-dimesityl-4-
chlorodiazaphospholenium) - triflato - molybdenum(0)]
1
1
and H-,31P-HMQC NMR spectra. MS: Kratos Con-
cept 1H, Xe-FAB, m-NBA matrix-FT-IR spectra:
Bruker IFS 113V, in CH2Cl2 solution. Elemental analy-
ses: Heraeus CHNO-Rapid. Melting points were deter-
mined in sealed capillaries. 31P- and 183W-NMR data as
well as IR data are listed in Table 1.
4.2. Preparation of complexes 9a,b
A solution of 8 (5 mmol) and the appropriate metal
complex 2a or 2b (5 mmol) in 150 ml of toluene was
refluxed for 24 h. After cooling, the formed precipitate
was filtered over a glass filter frit and washed with cold
toluene. The crude products were purified by recrystalli-
sation from boiling toluene.
1
(3a): yield 76%, m.p. 242–246°C. H-NMR (CDCl3):
trans-[(2,2%-Bipyridine)-dicarbonyl-chloro-(1,3-dime-
sityl - 4 - chlorodiazaphospholenium) - molybdenum(0)]
(9a): yield 95%, m.p.\260°C. H-NMR (CDCl3): 1.94
1.94 [s, 6H, o-Me], 2.09 [s, 6H, o-Me], 2.21 [s, 3H,
p-Me], 2.25 [s, 3H, p-Me], 6.43 [d, JPH=6.6 Hz, 1H,
3
1
5-H], 6.61 [s, 2H, m-H], 6.74 [s, 2H, m-H], 7.17 [br, 2H,
m-H(bipy)], 7.75 [br, 4H, p-H(bipy) and m-H(bipy)],
8.62 [s, 2H, o-H(bipy)]. 13C{1H}-NMR (CDCl3): 18.5
[br, o-Me], 21.6 [s, p-Me], 119.9 [q, 1JFC=321 Hz,
CF3], 120.3 [s, 4-C], 121.5 [s, 5-C], 123.2 [s, m-
CH(bipy)], 125.9 [s, m-CH(bipy)], 129.6 [br, m-CH],
[s, 6H, o-Me], 2.12 [s, 6H, o-Me], 2.21 [s, 3H, p-Me],
3
2.26 [s, 3H, p-Me], 6.35 [d, JPH=6.4 Hz, 1H, 5-H],
6.58 [s, 2H, m-H], 6.74 [s, 2H, m-H], 7.11 [br, 2H,
m-H(bipy)], 7.64 [br, 2H, p-H(bipy)], 7.69 [br, 2H,
m-H(bipy)], 8.60 [d, 3JHH=4.2 Hz, 2H, o-H(bipy)].
13C{1H}-NMR (CDCl3): 18.7 [br, o-Me], 21.72 [s, p-
Me], 21.77 [s, p-Me], 119.14 [s, 4-C], 120.4 [s, 5-C],
123.2 [s, m-CH(bipy)], 125.3 [s, m-CH(bipy)], 129.5 [br,
m-CH], 129.8 [br, m-CH], 131.6 [br, i-C], 134.6 [br,
i-C], 137.4 [br, o-C], 137.5 [s, p-CH(bipy)], 138.2 [br,
o-C], 139.0 [s, p-C], 139.1 [s, p-C], 152.6 [s, o-
CH(bipy)], 155.1 [s, o-C(bipy)], 228.2 [d, 2JPC=19.4
Hz, CO]. FAB-MS: m/z (%): 702 (20) [M+]. Anal. Calc.
for C32H31Cl2MoN4O2P (701.4): C, 54.79; H, 4.45; N,
7.99. Found: C, 55.59; H, 4.58; N, 8.11%.
2
129.8 [br, m-CH], 130.9 [d, JPC=5.0 Hz, i-C], 133.9
2
[d, JPC=5.7 Hz, i-C], 137.0 [br, o-C], 137.9 [br, o-C],
138.9 [s, p-CH(bipy)], 139.4 [s, p-C], 139.6 [s, p-C],
152.9 [s, o-CH(bipy)], 155.4 [s, o-C(bipy)], 228.2 [d,
2JPC=21.7 Hz, CO]. FAB-MS: m/z (%): 816 (1) [M+],
667 (65) [M+−OTf]. Anal. Calc. for C33H31ClF3Mo-
N4O5PS (815.0): C, 48.63; H, 3.83; N, 6.87. Found: C,
47.9; H, 3.9; N, 6.8%.
trans-[(2,2%-Bipyridine)-dicarbonyl-(1,3-dimesityl-4-
chlorodiazaphospholenium)-triflato-tungsten(0)] (3b):
1
trans-[(2,2%-Bipyridine)-dicarbonyl-chloro-(1,3-dime-
yield 72%, m.p. 222–224°C. H-NMR (CDCl3): 1.95 [s,
sityl-4-chlorodiazaphospholenium)-tungsten(0)]
(9b):
6H, o-Me], 2.11 [s, 6H, o-Me], 2.21 [s, 3H, p-Me], 2.24

D. Gudat et al. / Journal of Organometallic Chemistry 617–618 (2001) 383–394
391
3
[s, 3H, p-Me], 6.43 [d, JPH=7.9 Hz, 1H, 5-H], 6.61 [s,
resulting mixture was stirred for 2 h. After removal of
volatiles in vacuo, the remaining residue was triturated
three times with hexanes under ultrasonic irradiation.
The crude products were purified by recrystallisation
from THF–CH2Cl2 (1:1) at −20°C.
2H, m-H], 6.74 [s, 2H, m-H], 7.23 [br, 2H, m-H(bipy)],
7.83 [br, 4H, p-H(bipy) and m-H(bipy)], 8.70 [br, 2H,
o-H(bipy)]. 13C{1H}-NMR (CDCl3): 18.7 [br, o-Me],
1
21.8 [s, p-Me], 21.7 [s, p-Me], 119.5 [q, JFC=322 Hz,
CF3], 120.3 [s, 4-C], 120.7 [s, 5-C], 123.4 [s, m-
CH(bipy)], 126.6 [s, m-CH(bipy)], 129.6 [s, m-CH],
129.8 [s, m-CH], 130.9 [br, i-C], 133.9 [br, i-C], 137.4
[br, o-C], 138.2 [br, o-C], 139.0 [s, p-CH(bipy)], 139.2
[s, p-C], 139.4 [s, p-C], 153.1 [s, o-CH(bipy)], 156.4 [s,
Table 4
Atomic coordinates (×104) and equivalent isotropic displacement
2
3
,
parameters (A ×10 ) for 3b (estimated S.D. in parentheses; Ueq is
defined as one third of the trace of the orthogonalized Uij tensor)
2
o-C(bipy)], 222.0 [d, JPC=9.5 Hz, CO]. FAB-MS: m/z
x
y
z
Ueq
(%): 753 (40) [M+−OTf]. Anal. Calc. for
C33H31ClF3N4O5PSW (903.0): C, 43.90; H, 3.46; N,
6.20. Found: C, 42.9; H, 3.5; N 6.1% (Table 4).
S(1)
O(1)
O(2)
O(3)
C(1)
F(1)
F(2)
F(3)
Mo(1)
C(1A)
O(1A)
C(1B)
O(1B)
N(1%)
C(2%)
C(3%)
C(4%)
C(5%)
C(6%)
C(7%)
C(8%)
C(9%)
C(10%)
C(11%)
N(12%)
P(1)
N(2)
C(3)
Cl(1)
C(4)
Cl(1%)
N(5)
C(6)
C(7)
C(8)
C(9)
C(10)
C(11)
C(12)
C(13)
C(14)
C(15)
C(16)
C(17)
C(18)
C(19)
C(20)
C(21)
C(22)
C(23)
2956(1)
3312(3)
1987(4)
4205(4)
1745(5)
2464(3)
1236(3)
562(3)
7828(1)
6755(3)
7483(4)
8680(3)
8764(4)
9215(3)
9751(3)
8074(3)
6302(1)
8097(5)
9156(3)
6166(4)
6072(3)
4386(3)
3521(4)
2384(5)
2107(5)
2990(4)
4120(4)
5118(4)
4951(5)
5918(5)
7036(5)
7143(4)
6217(3)
5793(1)
4546(3)
4641(5)
3388(1)
5686(5)
6521(8)
6425(3)
3529(4)
3660(4)
2685(4)
1625(4)
1538(4)
2474(4)
4830(5)
607(4)
1238(1)
1655(2)
621(2)
38(1)
36(1)
58(1)
62(1)
43(1)
68(1)
72(1)
66(1)
1(1)
35(1)
43(1)
38(1)
51(1)
38(1)
45(1)
55(1)
56(2)
48(1)
39(1)
38(1)
48(1)
52(1)
43(1)
37(1)
34(1)
33(1)
38(1)
47(1)
55(1)
48(1)
1151(2)
1805(3)
2423(2)
1476(2)
1961(2)
2429(1)
2796(2)
3010(2)
3229(3)
3690(2)
1852(2)
2078(3)
1691(3)
1044(3)
795(3)
4.4. Preparation of complexes 7a,b [OTf ]
Triphenylphosphine (1 mmol) was added to a solu-
tion of 3a or 3b (1 mmol) in 50 ml of CH2Cl2 and the
5206(1)
5543(5)
5679(4)
3878(5)
3071(4)
4679(4)
3711(5)
3333(6)
3928(6)
4917(6)
5264(5)
6308(5)
7135(6)
8097(6)
8202(5)
7359(5)
6438(4)
7109(1)
8281(4)
9370(6)
10 569(2)
9265(6)
10 571(10)
8076(4)
8206(5)
9062(5)
8945(5)
7996(5)
7202(5)
7286(5)
10 052(5)
7840(6)
6389(6)
7766(5)
8436(5)
8196(6)
7325(6)
6667(6)
6855(5)
9372(6)
7053(7)
6120(6)
Table 3
Crystallographic data, structure solution and refinement of 3a and 7b
(OTf)
3a
7b [OTf]
Molecular formula C22H31ClF3MoN4- C55H54ClF3N4O6P2SW
O5PS
1202(3)
982(2)
373(2)
205(3)
633(3)
Molecular weight
Dimensions (mm)
Crystal system
Space group
815.04
0.20×0.20×0.05
Triclinic
1237.32
0.35×0.20×0.10
Orthorhombic
Pbca (no. 61)
17.9292(9)
16.1206(14)
39.207(4)
(
P1 (no. 2)
,
a (A)
8.9238(6)
10.5219(6)
18.6742(10)
94.931(3)
93.141(3)
91.417(3)
1743.54(18)
2
1232(2)
1419(2)
3138(1)
3120(2)
3690(2)
3788(1)
4124(3)
4777(4)
3889(2)
2554(2)
1958(2)
1415(2)
1448(2)
2056(2)
2615(2)
1911(3)
828(2)
3276(2)
4284(2)
4094(2)
4505(3)
5094(3)
5268(3)
4876(2)
3452(3)
5530(4)
5062(3)
,
b (A)
,
c (A)
h (°)
i (°)
k (°)
a
3
,
V (A )
Z
11332.1(15)
8
1.450
2.24
51(2) b
37(1)
35(1)
37(1)
38(1)
37(1)
40(1)
39(1)
47(1)
48(1)
51(1)
35(1)
40(1)
50(1)
53(1)
51(1)
41(1)
56(1)
84(2)
55(1)
z (g cm−3
)
1.552
0.62
828
v (mm−1
F(000)
)
4992
Diffractometer
Radiation
Nonius KappaCCD Nonius KappaCCD
Mo–Ka
0.71073
123(2)
Mo–Ka
0.71073
123(2)
,
u (A)
Temperature (K)
2q max. (°)
50.0
50.0
Index ranges
−95h59,
−125k512,
−225l522
12815
−205h516,
−175k519,
−465l546
45435
2361(5)
7610(4)
8727(5)
9850(5)
9889(5)
8752(5)
7589(5)
8751(5)
1123(6)
6379(5)
Measured data
Unique data
Rint
5006
8962
0.051
0.075
Refinement
F2
F2
Parameters/restraints 458/7
657/656
R for [I\2|(I)]
wR2 (all data)
Max./min.
0.047
0.120
−0.77/0.73
0.049
0.124
−0.871/1.564 (near
difference peak
W1)
a s.o.f.=0.699(3).
b s.o.f.=0.301(3).
−3
,
(e A
)

392
D. Gudat et al. / Journal of Organometallic Chemistry 617–618 (2001) 383–394
Table 5 (Continued)
Table 5
Atomic coordinates (×104) and equivalent isotropic displacement
2
3
,
parameters (A ×10 ) for 7b [OTf] (estimated S.D in parentheses; Ueq
is defined as one third of the trace of the orthogonalized Uij tensor)
x
y
z
Ueq
43(2)
86(1)
84(2)
C(26C)
S(1)
O(1)
O(2)
O(3)
C(1)
F(1)
F(2)
F(3)
O(1T)
C(1T)
C(2T)
C(3T)
C(4T)
O(2T)
C(5T)
C(6T)
C(7T)
C(8T)
4228(4)
5814(2)
6356(3)
5853(4)
5689(5)
4978(5)
4979(3)
4829(3)
4389(3)
2097(6)
1579(8)
1026(8)
965(8)
1653(9)
2263(17)
1950(2)
1720(2)
1610(2)
2220(2)
4875(5)
9003(2)
8575(4)
8890(6)
9873(5)
8492(8)
7704(4)
8618(5)
8775(4)
4727(6)
4524(9)
5136(10)
5178(10)
5188(10)
1982(19)
2750(2)
3100(2)
2450(3)
1940(2)
2780(2)
1124(1)
1332(2)
775(2)
x
y
z
Ueq
W(1)
4780(1)
3933(4)
3448(3)
4237(4)
3960(3)
5614(4)
5460(5)
5972(5)
6700(6)
6872(5)
6336(5)
6477(5)
7178(5)
7248(5)
6631(5)
5942(5)
5855(4)
4272(1)
3786(4)
3482(6)
3006(2)
3650(5)
3107(6)
4090(4)
3670(5)
3015(5)
2916(5)
3425(5)
4077(5)
4227(5)
2400(4)
3260(5)
4933(4)
4385(6)
5117(6)
5360(6)
4898(8)
4197(7)
3894(6)
5591(5)
5128(7)
3120(5)
5071(1)
5664(4)
6416(4)
6860(5)
6557(5)
5837(5)
5359(4)
5577(4)
5781(4)
6192(4)
6366(4)
6125(4)
5716(4)
4256(4)
3659(4)
3041(4)
3010(4)
3604(5)
4530(1)
4003(5)
3666(3)
5581(5)
6221(3)
3527(4)
2788(5)
2170(6)
2338(6)
3094(6)
3694(5)
4529(6)
4803(6)
5542(6)
6044(6)
5735(5)
5006(4)
4089(1)
3251(4)
3359(6)
2566(2)
4059(6)
4337(7)
4528(4)
2510(5)
2403(6)
1645(6)
1038(6)
1157(5)
1890(6)
3061(5)
222(5)
1978(6)
5313(6)
5358(7)
6132(7)
6820(8)
6758(6)
6010(6)
4606(6)
7681(6)
5937(6)
5012(1)
4260(4)
4383(4)
3812(5)
3076(5)
2928(5)
3499(5)
5986(5)
6218(4)
6956(5)
7454(5)
7238(5)
6517(5)
5145(4)
5587(4)
5755(4)
5474(5)
5011(5)
1564(1)
1793(2)
1939(1)
1493(2)
1461(1)
1634(2)
1757(2)
1794(2)
1715(3)
1596(2)
1554(2)
1411(2)
1303(2)
159(2)
1103(2)
1200(2)
1360(2)
1073(1)
956(2)
633(2)
441(1)
485(2)
147(3)
35(1)
148(3)
C(1A)
O(1A)
C(1B)
O(1B)
N(1C)
C(2C)
C(3C)
C(4C)
C(5C)
C(6C)
C(7C)
C(8C)
C(9C)
C(10C)
C(11C)
N(12C)
P(1)
N(2)
C(3)
Cl(1)
C(4)
Cl(1%)
N(5)
C(6)
C(7)
C(8)
34(2)
43(2)
39(2)
51(2)
38(2)
55(2)
70(3)
81(3)
58(3)
39(2)
43(2)
54(2)
63(3)
58(2)
45(2)
48(2)
45(1)
56(2)
78(3)
96(1)
68(3)
1226(2)
1242(3)
1190(2)
1567(2)
1071(2)
153(3)
433(4)
405(4)
48(4)
−94(4)
−689(9)
−792(9)
−495(11)
−275(9)
−317(9)
145(3)
72(3)
120(2)
116(2)
114(2)
134(3) a
132(4) a
133(4) a
146(4) a
139(4) a
134(3) b
126(4) b
129(4) b
131(4) b
131(4) b
a s.o.f.=0.756(3).
b s.o.f.=0.244(3).
a
trans-[(2,2%-Bipyridine)-dicarbonyl-(1,3-dimesityl-4-
chlorodiazaphospholenium)-triphenylphosphine-molyb-
denum(0)] triflate (7a): yield 82%, m.p. 226–228°C.
1H-NMR (CDCl3): 1.92 [s, 6H, o-Me], 2.01 [s, 6H,
96(1) b
50(2)
43(2)
52(2)
54(2)
52(2)
53(2)
54(2)
71(3)
73(3)
79(3)
56(2)
63(2)
5(3)
96(4)
83(3)
61(2)
91(3)
39(5)
89(3)
34(1)
31(2)
36(2)
50(2)
51(2)
58(3)
52(2)
30(2)
37(2)
38(2)
45(2)
45(2)
44(2)
30(2)
39(2)
41(2)
47(2)
54(2)
694(2)
1159(2)
1328(2)
1498(2)
1505(2)
1334(2)
1150(2)
1308(2)
1679(2)
946(3)
585(2)
464(2)
352(2)
356(3)
3
o-Me], 2.3 [br, 6H, p-Me], 6.64 [d, JPH=5.3 Hz, 1H,
5-H], 6.81 [s, 2H, m-H], 6.86 [s, 2H, m-H], 6.9-7.2 [m,
15H, PPh3], 7.22 [br, 2H, m-H(bipy)], 7.75 [br, 2H,
p-H(bipy)], 7.84[m, 2H, m-H(bipy)], 8.31 [br, 2H, o-
H(bipy)]. 13C{1H}-NMR (CDCl3): 18.3 [br, o-Me], 18.4
[br, o-Me], 21.6 [s, p-Me], 21.7 [s, p-Me], 121.5 [q,
1JFC=321 Hz, CF3], 122.2 [s, 4-C], 123.2 [s, 5-C], 123.3
[s, m-CH(bipy)], 125.7 [s, m-CH(bipy)], 129.0 [s, m-
CH], 129.1 [s, m-CH], 130.1 [d, 3JPC=6.5 Hz, m-
C(9)
C(10)
C(11)
C(12)
C(13)
C(14)
C(15)
C(16)
C(17)
C(18)
C(19)
C(20)
C(21)
C(22)
C(23)
P(2)
C(21A)
C(22A)
C(23A)
C(24A)
C(25A)
C(26A)
C(21B)
C(22B)
C(23B)
C(24B)
C(25B)
C(26B)
C(21C)
C(22C)
C(23C)
C(24C)
C(25C)
4
CH(PPh3)], 130.4 [d, JPC=1.9 Hz, p-CH(PPh3)], 130.4
486(2)
600(2)
448(2)
193(3)
1
[br, i-C], 131.1 [d, JPC=34.5 Hz, i-C(PPh3)], 133.1 [d,
2JPC=11 Hz, o-CH(PPh3)], 133.4 [d, 2JPC=6.1 Hz,
2
2
i-C], 136.6 [d, JPC=3.1 Hz, o-C], 137.5 [d, JPC=3.1
Hz, o-C], 139.1 [s, p-CH(bipy)], 140.2 [s, p-C], 140.5 [s,
p-C], 152.6 [s, o-CH(bipy)], 154.3 [s, o-C(bipy)], 225.0
723(2)
2168(1)
2392(2)
2425(2)
2578(2)
2695(2)
2656(2)
2501(2)
2241(2)
2573(2)
2623(2)
2356(2)
2034(2)
1974(2)
2446(2)
2328(2)
2526(2)
2847(2)
2981(2)
2
[dd, JPC=12.8 Hz, 18.9 Hz, CO]. FAB-MS: m/z (%):
929 (35) [M+−OTf], 667 (75) [M+−OTf−PPh3].
Anal. Calc. for C51H46ClF3MoN4PO5S (1046.4)·2THF:
C 58.01 H 5.12 N 4.59; Found C 57.50 H 5.0 N 4.55%.
trans - [(2,2% - Bipyridine) - dicarbonyl - (1,3 - dimesityl-
4 - chlorodiazaphospholenium) - triphenylphosphine-
tungsten(0)] triflate (7a): yield 71%, m.p. 212–214°C.
1H-NMR (CD2Cl2): 1.93 [s, 6H, o-Me], 2.02 [s, 6H,
3
o-Me], 2.33 [br, 6H, p-Me], 6.71 [d, JPH=5.9 Hz, 1H,
5-H], 6.83 [s, 2H, m-H], 6.86 [s, 2H, m-H], 6.9-7.2 [m,
15H, PPh3], 7.3-8.2 [8H (bipy)]. 13C{1H}-NMR
(CD2Cl2): 17.9 [br, o-Me], 18.1 [br, o-Me], 21.1 [s,
p-Me], 21.2 [s, p-Me], 120.9 [q, JFC=320 Hz, CF3],
122.0 [s, 5-C], 122.7 [s, 4-C], 123.8 [s, m-CH(bipy)],
1

D. Gudat et al. / Journal of Organometallic Chemistry 617–618 (2001) 383–394
393
125.5 [s, m-CH(bipy)], 128.8 [s, m-CH], 128.9 [s, m-
[m, 4H, o-H(bipy)]. 13C{1H}-NMR (CD3CN): 17.8 [s,
o-Me], 17.9 [s, o-Me], 20.4 [s, p-Me], 20.6 [s, p-Me],
120.5 [s, 4-C], 121.4 [q, JFC=321 Hz, CF3], 122.3 [s,
3
CH], 129.8 [d, JPC=5.3 Hz, m-CH(PPh3)], 130.1 [br,
i-C], 130.4 [d, 4JPC=1.9 Hz, p-CH(PPh3)], 131.3 [d,
1
2
1JPC=42 Hz, i-C(PPh3)], 133.1 [d, JPC=10.7 Hz, o-
5-CH], 124.0 [s, m-CH(bipy)], 125.3 [s, CN], 127.3 [s,
m-CH(bipy)], 129.4 [s, m-CH], 129.5 [s, m-CH], 130.4
CH(PPh3)], 135.1 [d, 3JPC=3.4 Hz, o-C], 135.2 [br,
i-C], 137.7 [d, 3JPC=3.1 Hz, o-C], 138.1 [s, p-
CH(bipy)], 139.9 [d, 5JPC=1.1 Hz p-C], 140.2 [d,
5JPC=1.1 Hz, p-C], 152.7 [s, o-CH(bipy)], 154.9 [s,
o-C(bipy)], 218.1 [dd, 2JPC=12.4 Hz, 9.3 Hz, CO].
FAB-MS: m/z (%): 929 (35) [M+−OTf], 667 (75)
2
2
[d, JPC=5 Hz, i-C], 133.3 [d, JPC=6 Hz, i-C], 137.1
3
3
[d, JPC=2 Hz, o-C], 137.9 [d, JPC=2 Hz, o-C], 138.0
[s, p-C], 139.4 [s, p-CH(bipy)], 140.4 [s, p-C], 152.9 [s,
2
o-CH(bipy)], 156.0 [s, o-C(bipy)], 219.7 [d, JPC=10.2
Hz, CO].
[M+−OTf,
ꢀPPh3].
Anal.
Calc.
for
C51H46ClF3N4PO5SW (1134.3)·2THF: C, 55.43; H,
4.89; N, 4.38. Found: C, 54.56; H, 4.95; N, 4.31%
(Table 5).
4.7. Reaction of the p-chloro-diazaphospholene (8) with
4b
A solution containing 1 [OTf] (1 mmol) and 4b (1
mmol) in 20 ml of CH2Cl2 was stirred for 30 min at
ambient temperature. An 31P-NMR spectroscopic assay
revealed the presence of a mixture of 5b [Cl] and 9b
beside unreacted 8. For longer reaction times, increas-
ing conversion of 5b and 8 into 9b was observed.
Attempts towards isolation of 5b [Cl] were unsuccessful.
4.5. Preparation of fac-[(2,2%-bipyridine)-tricarbonyl-
(1,3-dimesityl-4-chlorodiazaphospholenium)-
tungsten(0)] triflate (5b) [OTf ]
A solution containing 1 [OTf] (1 mmol) and 4b (1
mmol) in 20 ml of CH2Cl2 was stirred for 1 h at
ambient temperature. After removal of volatiles in
vacuo, the residue was digerated with 10 ml of toluene,
and filtered. The resulting crude product was washed
several times with toluene and ether. Yield 65%, m.p.
4.8. Crystal structure determinations of compounds 3a
and 7b [OTf ]
145–147°C. 1H-NMR (CD2Cl2): 1.90 [s, 6H, o-Me],
3
1.98 [s, 6H, o-Me], 2.37 [br, 6H, p-Me], 6.63 [d, JPH
=
7 Hz, 1H, 5-H], 6.91 [s, 2H, m-H], 6.93 [s, 2H, m-H],
7.34 [m, 2H, m-H(bipy)], 8.16 [m, 2H, p-H(bipy)], 8.43
[m, 4H, o-H(bipy)]. 13C{1H}-NMR (CD2Cl2): 17.7 [br,
o-Me], 17.9 [br, o-Me], 21.1 [s, p-Me], 21.2 [s, p-Me],
121.2 [q, JFC=321 Hz, CF3], 124.3 [s, 4-C], 124.8 [s,
m-CH(bipy)], 125.1 [s, 5-C], 127.3 [s, m-CH(bipy)],
128.9 [br, i-C], 130.1 [s, m-CH], 130.2 [s, m-CH], 131.9
The data were collected on a Nonius KappaCCD
diffractometer at −150°C using Mo–Ka radiation
,
(u=0.71073 A). The structures were solved by direct
methods (SHELXS-97) [25a]. The non-hydrogen atoms
were refined anisotropically, H atoms were refined us-
ing a riding model (full-matrix least-squares refinement
1
on F2
(SHELXL-97) [25b]. In both compounds, the
2
3
[d, JPC=4.6 Hz, i-C], 137.0 [d, JPC=3.8 Hz, o-C],
chlorine atom attached to the diazaphospholene ring
was disordered between two positions with relative
occupancy factors of 85:15 (3a) and 76:24 (7b [OTf]),
respectively. Crystals of 7b [OTf] contained two addi-
tional solvent molecules (THF). Details of data collec-
tion and refinement are given in Table 3.
138.0 [d, 3JPC=4.2 Hz, o-C], 139.9 [s, p-CH(bipy)],
5
5
141.0 [d, JPC=1.5 Hz, p-C], 141.1 [d, JPC=1.5 Hz,
p-C], 152.9 [s, o-CH(bipy)], 155.4 [s, o-C(bipy)], 198.5
2
2
[d, JPC=84.3 Hz, CO], 209.2 [d, JPC=12.2 Hz, CO].
FAB-MS: m/z (%): 781 (15) [M+−OTf], 753 (30)
[M+−OTf, ꢀCO].
4.6. Preparation of trans-[(acetonitrile)-(2,2%-bipyridine)-
dicarbonyl-(1,3-dimesityl-4-chlorodiazaphospholenium)-
tungsten(0)] triflate (6b) [OTf ]
5. Supplementary material
Refluxing a solution of 5b [OTf] for 30 min in
[D3]-MeCN resulted in quantitative formation (31P-
NMR control) of a new product identified in situ as 6b
[OTf]. Attempts towards isolating the product by re-
moval of the solvent in vacuo resulted in partial decom-
Crystallographic data (excluding structure factors)
for the structures reported in this work have been
deposited with the Cambridge Crystallographic Data
Centre as supplementary publication no. CCDC 148149
(3a) and CCDC 148150 (7b [OTf]), respectively. Copies
of the data can be obtained free of charge on applica-
tion to The Director, CCDC, 12 Union Road, Cam-
bridge CB2 1EZ, UK (fax: +44-1223-336033; e-mail:
deposit@ccdc.cam.ac.uk or www: http://www.ccdc.
cam.ac.uk).
1
position, presumably with formation of 3b. H-NMR
(CD3CN): 1.98 [s, 6H, o-Me], 2.10 [s, 6H, o-Me], 2.30
[br, 6H, p-Me], 6.74 [s, 2H, m-H], 6.81 [s, 2H, m-H],
6.83 [d, 3JPH=8 Hz, 1H, 5-H], 7.50 [m, 2H, m-
H(bipy)], 8.13 [br, 4H, p-H(bipy) and m-H(bipy)], 8.75

394
D. Gudat et al. / Journal of Organometallic Chemistry 617–618 (2001) 383–394
Gudat, A. Haghverdi, H. Hupfer, M. Nieger, Chem. Eur. J.
Acknowledgements
(2000) 3414.
[10] M. Denk, R.K. Hayashi, R. West, J. Chem. Soc. Chem. Com-
mun. (1994) 33.
[11] D. Gudat, A.W. Holderberg, S. Kotila, M. Nieger, Chem. Ber.
129 (1996) 465.
This work was financially supported by the Deutsche
Forschungsgemeinschaft and the Fonds der Chemis-
chen Industrie.
[12] D. Gudat, Coord. Chem. Rev. 173 (1997) 71.
[13] (a) A.H. Cowley, R. Kemp, Chem. Rev. 85 (1985) 367. (b) M.
Sanchez, M.R. Mazie`res, L. Lamande´, R. Wolf, in: M. Regitz,
O. Scherer (Eds.), Multiple Bonds and Low Coordination Chem-
istry in Phosphorus Chemistry, Georg Thieme Verlag, Stuttgart,
1990, p. 129.
[14] (a) R.G. Montemayor, D.T. Sauer, S. Fleming, D.W. Bennett,
M.G. Thomas, R.W. Parry, J. Am. Chem. Soc. 100 (1978) 2231.
(b) A.H. Cowley, R.A. Kemp, J.C. Wilburn, Inorg. Chem. 20
(1981) 4289.
[15] D.W. Bennett, J.L. Coffer, J. Coord. Chem. 33 (1994) 271.
[16] An analogous reactivity for 1 [OTf] and 8 was observed as well
as for the 1,3-dimesityl-1,3,2-diazaphospholenium triflate and the
corresponding p-chloro-compound which both lack the addi-
tional Cl substituent at the ring, however, due to the limited
amount of available material the nature of the formed complexes
was only elucidated by in situ NMR spectroscopic studies.
[17] (a) H. Nakazawa, Y. Yamaguchi, K. Miyoshi, J. Organomet.
Chem. 465 (1994) 193. (b) H. Nakazawa, Y. Yamaguchi, T.
Mizuta, K. Miyoshi, Organometallics 14 (1995) 4173.
[18] D. Gudat, M. Nieger, E. Niecke, J. Chem. Soc., Dalton Trans.
(1989) 693.
[19] For recent reviews on the use of inverse detection NMR tech-
niques for the characterisation of metal complexes see: (a) D.
Gudat, Ann. Rep. NMR Spect. 38 (1999) 139. (b) W.v. Philips-
born, Chem. Soc. Rev. 28 (1999) 95. (c) J. Mason (Ed.), Multin-
uclear NMR, Plenum, New York, 1987.
[20] H. Nakazawa, Y. Yamaguchi, K. Miyoshi, A. Nagasawa,
Organometallics 15 (1996) 2517.
[21] C.A. Tolman, Chem. Rev. 77 (1977) 313.
[22] D. Gudat, A. Haghverdi, M. Nieger, Angew. Chem. Int. Ed.
Engl. 39 (2000) 3084.
[23] Since no structural data for the 1,3-dimethyl-1,3,2-diazaphos-
pholidinium cation are available, the bonding parameters of the
N-t-Bu substituted cation from Ref. [9b] were considered in-
stead.
[24] (a) Z. Zhou, R.G. Parr, J.F. Garst, Tetrahedron Lett. (1988)
4843. (b) Z. Zhou, R.G. Parr, J. Am. Chem. Soc 111 (1989)
7371.
[25] (a) G.M. Sheldrick, SHELXS-97, Acta Crystallogr. Sect. A 46
(1990) 467. (b) G.M. Sheldrick, SHELXL-97, University of Go¨t-
tingen, Germany, 1997.
References
[1] E.O. Fischer, A. Maasbo¨l, Angew. Chem. Int. Ed. Engl. 3 (1964)
580.
[2] (a) K.H. Do¨tz, in: A. de Meijere, H. tom Dieck (Ed.),
Organometallics in Organic Syntheses, Springer, Berlin, 1988. (b)
R.H. Grubbs, in: G. Wilkinson, F.G.A Stone, E.W. Abel (Eds.),
Comprehensive Organometallic Chemistry, vol. 8, Pergamon,
Oxford, 1982, p. 499. (c) W. Herrmann, C. Ko¨cher, Angew.
Chem. Int. Ed. Engl. 36 (1997) 2162.
[3] A. Igau, H. Gru¨tzmacher, A. Baceiredo, G. Bertrand, J. Am.
Chem. Soc. 110 (1988) 6463.
[4] A.J. Arduengo, R.L. Harlow, M. Kline, J. Am. Chem. Soc. 113
(1991) 361.
[5] (a) G. Bertrand, R. Reed, Coord. Chem. Rev. 137 (1994) 323. (b)
A.J. Arduengo, Acc. Chem. Res. 32 (1999) 913. (c) D. Bourissou,
O. Guerret, F.P. Gabbai, G. Bertrand, Chem. Rev. 100 (2000)
39.
[6] (a) (E=Si): M. Denk, R. Lennon, R. Hayashi, R. West, A.V.
Belyakov, H.P. Verne, A. Haaland, M. Wagner, N. Metzler, J.
Am. Chem. Soc. 116 (1994) 2691. (b) B. Gehrhus, M.F. Lappert,
J. Heinicke, R. Boese, D. Bla¨ser, J. Chem. Soc. Chem. Commun.
(1995) 1931. (c) (E=Ge): W.A. Herrmann, M. Denk, J. Behm,
W. Scherer, F.R. Klingan, H. Bock, B. Solouki, M. Wagner,
Angew. Chem. 104 (1992) 1489; Angew. Chem. Int. Ed. Engl.
31(1992) 1485.
[7] (a) (E=B-): E.S. Schmidt, A. Jockisch, H. Schmidbaur, J. Am.
Chem. Soc. 121 (1999) 9758. (b) (E=N+): G. Boche, P. An-
drews, K. Harms, M. Marsch, K.S. Rangappa, M. Schimeczek,
C. Willeke, J. Am. Chem. Soc. 118 (1996) 4925. (c) (E=P+):
I.A. Litvinov, V.A. Naumov, T.V. Gryaznova, A.N. Pudovik,
A.M. Kibardin, Dokl. Akad. Nauk SSSR 312 (1990) 623. (d)
(E=P+) M.K. Denk, S. Gupta, A.J. Lough, Eur. J. Inorg.
Chem. (1999) 41. (e) (E=As+): C.J. Carmalt, V. Lomeli, B.G.
McBurnett, A.H. Cowley, J. Chem. Soc. Chem. Commun. (1997)
2095.
[8] (a) C. Heinemann, T. Mu¨ller, Y. Apeloig, H. Schwarz, J. Am.
Chem. Soc. 118 (1996) 2023. (b) C. Boehme, G. Frenking, J. Am.
Chem. Soc. 118 (1996) 3039.
[9] (a) D. Gudat, Eur. J. Inorg. Chem. (1998) 1087. (b) M.K. Denk,
S. Gupta, A.J. Lough, Eur. J. Inorg. Chem. (1999) 41. (c) D.
.