Chelate-stabilized alkylidene complexes of molybdenum and tungsten


Material Information

Chelate-stabilized alkylidene complexes of molybdenum and tungsten
Physical Description:
vii, 183 leaves : ill. ; 29 cm.
Vaughan, William M ( William Michael )
Publication Date:


Subjects / Keywords:
Organomolybdenum compounds   ( lcsh )
Organotransition metal compounds   ( lcsh )
Organotungsten compounds   ( lcsh )
Chemistry thesis, Ph. D
Dissertations, Academic -- Chemistry -- UF
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1995.
Includes bibliographical references (leaves 176-182).
Statement of Responsibility:
by William M. Vaughan.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002051792
oclc - 33436624
notis - AKN9751
System ID:

Full Text







Enter to think God's thoughts after Him.

Go forth to apply His thoughts in

Townes Randolph Leigh

That statement on the facade of Leigh Hall, along with the names of great scientists and

their gargoyles, has done much to inspire and humble over the past few years. The

chemists of Leigh's time would be amazed at the strides that have been made, but they

would have expected no less.

I could only hope that I too have made a small contribution

to the quality reputation of the Department of Chemistry at the University of Florida.

I have great appreciation for the guidance and support of my research director, Dr.

James M. Boncella. From Jim, I have not only begun to learn the art of science, but to

understand the true excitement in learning and discovery.

I thank the members of my committee, Professors David Richardson, Bill Jones,

Ken Wagener and Guerry McClellan, for their time, advice and encouragement

Discussions with Profs. Richardson and Jones were instrumental in developing the

mechanisms in Chapter Three. A special thanks is extended to Prof. Wagener.

I often

pass along "the best advice I ever got about graduate school" to anyone that will listen.

The research of this dissertation is greatly dependent on the assistance and many

helpful conversations with Drs. Khalil Abboud, Roy King and David Powell.


Ospina with Dr. Powell collected the CIMS data in Chapter Three.

One can learn so much from the people around them, and my fellow group

members over the years have made that all too evident.

Those people that have made

has been an inspiration over the last year by the consuming interest she takes with research.

Dan VanderLende has been a perennial source of motivation and friendship.

Dennis W

Smith, Jr. has been a constant best friend throughout graduate school,

and anyone who knows him will know why.

With whom else one would attempt to

conquer the continent of Europe in a fortnight, and truly believe that one would? Kathy

Novak will always remain a special friend, and it is ironic that in thinking of someone who

is so eloquent with words, I cannot think of how to describe how she has benefited my

experience at Florida.

My parents, Michael and Kay, and my brother Steven define encouragement and

support more than anyone.

Through Sunday evening phone conversations, the demands of

research never seemed quite so bad.

I finally thank my fiance, Pamela Pippin.

One day it was extremely obvious why

and for whom I work so hard, and Pam's understanding and commitment have been

unfaltering during the writing of this dissertation.


AC KNOWLEDGMENTS................... ................................................... ii

ABSTRACT............................................. ...... ............................... v




BACKGROUND ................... ................... ........ 1

Transition Metal-Ligand Multiple Bonds .........................
Transition Metal Alkylidene Complexes..........................
The Nature of Metal-Carbon Double Bonds.................
Synthesis and Characterization of Metal Alkylidenes.
Reactivity of Metal Alkylidene Ligands ......................
Olefin Metathesis....................................................
Chelated Transition Metal Alkylidene Complexes...............

.......... 2

HYDRIDOTRIS(PYRAZOLYL)BORATE......... ............... ...............

Neutral Molybdenum Imido Alkylidene Complexes..................................
Synthesis of TpMo(CHC(Me)2Ph)(NAr)(OTf)................................
Crystal Structure of TpMo(CHC(Me)2Ph)(NAr)(OTf) ......................
Synthesis of TpMo(CH2C(Me)2Ph)(NAr)(0) ...................................
Crystal Structure of TpMo(CH2C(Me)2Ph)(NAr)(O) ..........................
Crystal Structure of [TpMo(NAr)()]20. .............. ..... .............. ........
Synthesis of TpMo(CHC(CH3)2Ph)(NAr)(OCH3) ..............................
Synthesis of TpMo(CC(CH3)2Ph)(NHAr)(OCH3) ..............................
Crystal Structure Determination of TpMo(CHC(CH3)2Ph)(NAr)(OCH3)

and TpMo(CC(CH3)2Ph)(NHAr)(OCH3).....................
Cationic Molybdenum Imido Alkylidene Complexes ..................
Synthesis of TpMo(CHC(CH3)2Ph)(NAr)(CH3) .................
Synthetic Routes to Cationic Molybdenum Imido Alkylidene
Complexes .........................................................
The Trans Influence of Ligands in Tp Molybdenum Complexes.....
Rotational Isomerism of the Molybdenum-Alkylidene Bond .........
Polymerization Studies....................................................




. .. ... .. ..5049
a. ..... ". """ 5

.. d
,f... f.. S

S. . .64
ft.t. ..*tf 64

Kinetics of the Formation of the Alkylidene Complex... ............................76
A Closer Look at the Role of Phosphine ................... .............................77
Interconversion of the Metallacycle and the Alkylidene Complexes..................81
Deuterium-Labeling Studies............................................................... ..83
Chemical Ionization Mass Spectrometry Study..................................84
Deuterium NMR Spectroscopy .................................... ...................87
Kinetic Isotope Effects in the Formation of the Metallacycle and the
Alkylidene Complexes ................. ............................ ............. ..............89
Proposed Mechanistic Scheme................................. .........................91


Materials and Methods............................ .............. ...........................94
Syntheses......... ..... .. ................... ............................................95
TpMo(CHC(Me)2Ph) (NAr)(OTf) (1) .. ................ ................. ....... 95
TpMo(CH2C(Me)2Ph)(NAr)(0) (2) ................................................95
TpMo(CHC(CH3)2Ph)(NAr)(OCH3) (4).........................................96
TpMo(CC(CH3)2Ph)(NHAr)(OCH3) (5)............ ..........................97
TpMo(CHC(CH3)2Ph)(NAr)(CH3) (6)............................................98
[TpMo(CHC(CH3)2Ph)(NAr)(Et20)] [BAr'4] (7)...............................99
[TpMo(CHC(CH3)2Ph)(NAr)(NCCH3)][B(C6F5)4] (8)....................... 100
[TpMo(CHC(CH3)2Ph)(NAr)(NCCH3)][B(CH3)(C6F5)3] (9)............... 100
[TpMo(CHC(CH3)2Ph)(NAr)(THF)][B(CH3)(C6F5)3] (10) ................. 101
[TpMo(CHC(CH3)2Ph)(NAr)(P(CH3)3)][OS 02CF3] (11).................... 102
[(H3CO)2C6H2(NSi(CH3)3)2]JW(NPh)(C1)2 (17) .............................. 103
[(H3CO)2C6IH2(NSi(CH3)3)2]W(NPh)(Cl)2(P(CH3)3) (19).................. 104
Crystallographic Studies..................... .............................................. 104
X-ray data collection and structure refinement for compounds 2 and 3....... 104
X-ray data collection and structure refinement for compounds 4 and 5....... 106
X-ray data collection and structure refinement for compounds 16 ............ 107
Alkylidene Rotational Studies.................................................... ....... 108

Polymerizations. ............... ................................... ... ............
ROMP reactions by 1 and 6 with A1C13 .......................
Oligomerization of 1,9-decadiene by 1 and 6 with AIC13...

Thermolysis Kinetics...................
UV-VIS Spectroscopy Studies.........
Mass Spectrometry of Neopentane-dn.
Deuterium NMR Spectroscopy .........

."C 6 C 0" "'' 0" ""

..** *********- C
*SS)t1c(eC 5CC *1

. ... ."**** ******""** "' ''" "C'" **" ***S "
* C*C CS CCCS) l )* C C** C*a S*t* SCt(*SS

A PPEN D ICES......................... C............................ .. ... .................. ...... 113
A TABLES OF CRYSTALLOGRAPHIC DATA....................................... 114
B TABLES OF KINETIC AND CIMS DATA............... ................ ......... 171

BIOGRAPHICAL SKETCH................................................................... 183

LIST OF REFERENCES ........................................................ ................... 176

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


William M. Vaughan

May, 1995

Chairman: James M. Boncella
Major Department: Chemistry

Two chelating ligands, hydridotris(pyrazolyl)borate (Tp) and N,N'-

bis(trimethylsilyl)phenylenediamide (TMS2pda), were used to stabilize imido alkylidene

complexes of molybdenum and tungsten.

The tridentate Tp ligand successfully stabilized a

number of molybdenum complexes of the formula TpMo(CHC(CH3)2Ph)(N(2,6-

iPr2C6H3)(X) where X is triflate, methoxy or methyl substituents.

The stability imparted

by the ligand system allowed for numerous synthetic transformations not commonly

observed for high oxidation state metal alkylidene complexes.

These include the isolation

of hydrolysis products, base-catalyzed proton transfer from an alkylidene ligand to an

imido ligand, and the preparation of cationic molybdenum alkylidenes by several methods.

Differing rates of rotation of the molybdenum-alkylidene double bond are correlated to the

electronic properties of the X substituent. Polymerization studies indicate that the

The reactions of tungsten imido complexes stabilized by TMS2pda ligand have been


Thermolysis of complex (TMS2pda)W(NPh)(CH2C(CH3)3)2 (13) gives a

metallasilylcycle complex [(Me3SiN)C6H4(N'SiMe2CH2)]W(NPh)(CH2C(CH3)3) (16),

and in the presence of trimethylphosphine, the alkylidene complex

(TMS2pda)W(NPh)(CHC(CH3)3)(P(CH3)3) (14), is isolated.
be interconverted by the addition or abstraction of PMe3. The i

Complexes 16 and 14 can

mechanisms of these

transformations were studied by 1H NMR kinetics and deuterium-labeling studies. A
mechanistic scheme is proposed based on the observations that two competing abstraction

pathways are present and that PMe3 is not involved in the rate determining step.


Our interest in the synthesis of novel transition metal alkylidene complexes is linked

to the birth of coordination polymerizations in the early 1950s.1

The chemistry world was

inundated by a flurry of research centered upon the discoveries of Ziegler and Natta that

transition metal halides and aluminum alkyls polymerize ethylene and propylene to give

high molecular weight polyethylene and stereoregular polypropylene under mild


Naturally, their Nobel Prize-winning discovery prompted research extending

this new concept to all metals in the periodic table. However, when molybdenum and

tungsten were employed, researchers stumbled upon olefin metathesis and the ring-opening

polymerization of cyclic olefins.

The nature of these early, heterogeneous catalysts made the characterization of the

catalytic active species extremely difficult, but this "blissful ignorance" did not hinder the

application of this new reaction to numerous industrial processes.2

More than 10 years

after the first reports of olefin metathesis, H6rrison and Chauvin proposed that metal-

carbon double bonds were the reactive moiety as a result of detailed kinetic and mechanistic

studies by numerous researchers.3

Eventually, the preparation of isolable transition metal

alkylidene complexes that catalyzed olefin metathesis solidified support for the Chauvin

mechanism, and these well-defined alkylidene complexes were more amenable to

systematic variations of the metal and ligand environment.2

As a result, the in-depth

* fl S* S d* ** S S S S 4 S


synthesizing and characterizing novel high oxidation state transition metal complexes for

their application to olefin metathesis.

Transition Metal-Ligand Multiple Bonds

Highly electrophilic do metal centers are capable of forming multiple bonds with
carbon, nitrogen, and oxygen by a o bond and one or two nt bonds between the metal dn

and ligand pn orbitals.4,5 The formal oxidation state of the ligands is that of their closed-

shell anions, thus the triply-bonded nitrido6'7 and alkylidyne ligands are N3- and CR3-, ai

the doubly-bonded oxo, imido, and alkylidene ligands are 02-


, and CR22-


This formalized bonding description of a dO metal center and anionic ligands

reflects the observed electrophilicity of the metal center and the nucleophilicity of the ligand


The bonding orbitals involved in forming the metal-ligand multiple bonds are

similar for the above ligands since oxo, imido, nitrido, and alkylidyne ligands are

isoelectronic, involving empty metal d7r orbitals and filled ligand pt orbitals.

in Figure 1.1, if the metal-ligand bond is taken to lie along the

z nr

As indicated

z axis of an xyz coordinate






Metal-ligand bonding orbitals.

The do-po bond; b) A dn-pt

bond in xz plane; c) A din-pn bond in yz plane orthogonal to (b).



metal dyz and ligand py orbitals.

In simple valence bonding terms, the trianionic nitrido and

alkylidyne8 ligands form three bonds, and the dianionic oxo, imido, and alkylidene ligands

form double bonds.

The oxo9 and imidol011 ligands can be viewed as isoelectronic with the nitrido and

alkylidyne ligands because of their ability to donate a lone pair of electrons to an empty dn

orbital (Figure 1.2).

Metal-oxo and metal-imido groups are considered to be triply bonded

whenever the formal electron count at the metal center is less than eighteen. Experimental

evidence for triply bonded oxo and imido ligands comes from vMN and VMO determined by

infrared spectroscopy and structural data determined by x-ray crystallography.4"10 Higher

frequency stretching energies and shorter metal-ligand distances are observed for triply

bonded ligands, and in the case of imido ligands, the rehybridization of the nitrogen atom

results in M-N-R bond angles approaching linearity.

This additional t stabilization

provided by oxo and imido ligands makes the use of such ligands highly desirable in

stabilizing electropositive d metal centers, and the ligands are often employed as inert

"spectator" ligands in the synthesis of stable transition metal complexes.12



Bond orders of oxo and imido ligands. a) A oxo triple bond; b) A
linear imido triple bond; c) A bent imido double bond.

With this understanding of the nature of multiple bonds between transition

metals and ligands, the consequence of more than one metal-ligand multiple bond in a

complex will be considered. The dO metal center has only three unfilled orbitals of nt

symmetry available for bonding. For octahedral complexes, only three dnt-pTt bonds can

a a *

a 1. 1 --1 1_ -- _1- 1 1 -.. -1 .



thus destabilizing the t interactions.

The advantage of cis-oriented ligands is demonstrated

in the orbital diagram shown (Figure 1.3), in which the x interactions of a triply bonded

oxo ligand and a doubly bonded alkylidene ligand each involve different, mutually

orthogonal di orbitals.


- C


Metal-ligand interaction with two multiply bonded ligands. a) A

metal-alkylidene x bond; b) A metal-oxo x bond; c) A metal-oxo K
bond orthogonal to (b).

The limit of three x bonds due to orbital competition is seen in the structure of

Mo(NPh)2(S2CNEt2)2, in which the possibility of two triply bonded imido ligands is

precluded (Figure 1.4).

The x-ray crystal structure of Mo(NPh)2(S2CNEt2)2 shows one of

the imido ligands to be nearly linear (169.4(4)0, triply bonded) and the other imido ligand to

be bent (139.4(4), doubly bonded).17'18


s1 /


; N"

SA 1 1 I ---- ----- -------t _-

1- 4


concepts will be used to explain observations and results concerning the metal imido

alkylidene complexes throughout this dissertation.

Transition Metal Alkvlidene Comolexes

Metal-carbon double bonds are the focus of this research, and in this section, the

nature of metal alkylidenes will be reviewed more extensively.

Also, the synthesis,

characterization, and reactivity of metal alkylidenes will be briefly surveyed.

The Nature of Metal-Carbon Double Bonds

Transition metal alkylidenes or carbenes are capable of forming only one it bond

due to the sp2 hybridization at the carbon atom.

In classifying metal-carbon doubly bonded

complexes, there is a distinction drawn between Schrock-type alkylidene and Fischer-type

carbene complexes.4'5 The difference between these two types of ligands is determined by

the oxidation state of the metal, the electronic configuration of the ligand, and the
substituents on the a-carbon of the ligand.

In Fischer-type carbenes, the carbon atom is considered to be electrophilic.


complexes characteristically contain transition metals (groups 6-9) in low oxidation states

and have Kx-donor substituents (heteroatoms, phenyl ring) on the a-carbon.19 E.O. Fischer

reported the first example, W(CO)5(=C(CH3)(OCH3)), in 1964 in which the tungsten atom

is in its zero oxidation state and the carbene ligand is considered to be a neutral donor.Y

Shown in Figure 1.5, the carbene fragment in Fischer complexes is considered to be a

singlet carbene donating its electron pair to an empty do orbital.4 The nt bond results from

overlap of a filled dn metal orbital with the empty pnt carbon orbital. For the later transition

metals, the dcl orbital is lower in energy than the p7t orbital, and the bond is polarized in the


Figure 1.5.

Electronic structure of a Fischer-type carbene.

The Schrock-type alkylidene, however, is considered to be nucleophilic.

Alkylidene complexes contain high-valent, early transition metals, most commonly in the

d oxidation state.

The a-carbon is substituted by hydrogen or alkyl groups, and the

highly electrophilic metal center is often stabilized by x-donating ancillary ligands, such as

oxo, imido, alkoxide, or halide ligands. R.R. Schrock reported the first example,

Ta(CHC(CH3)3)(CH2C(CH3)3)3 in 1973.21

The bonding in metal alkylidenes can be

viewed in different ways (Figure 1.6), and a number of theoretical investigations suggest

the bond descriptions vary depending on the complex in question. An ionic valence bond

description considers the alkylidene fragment as a CR22- carbanion interacting with two

empty metal d orbitals, such as was described in the previous section for oxo and imido


Other descriptions suggest a more covalent bond consisting of a triplet carbene

fragment interacting with a triplet metal fragment.22,23

Reality lies somewhere between

these two extremes.

i\t '"

(a) ionic model

(b) covalent model



The orbitals for early transition metals are higher in energy than the carbon orbitals

resulting in the metal alkylidene bond being polarized in the M+-C- direction.

The higher

this energy difference is the more ionic the nature of the bond is, and the ionic valence

description is of more importance. However, this energy difference for carbon ligands is

not as great as the difference for the more electronegative oxygen or nitrogen ligands, and

thus metal alkylidene bonds are increasingly covalent compared to oxo and imido bonds.

Such an "ethylene-like" description is evidenced by the high barriers to rotation of metal-

carbon double bonds.

The mechanism of metal-carbon double bond rotation is described

later in this section.

The transition metal involved in the existence of alkylidene complexes is of strict


Schrock-type terminal alkylidenes complexes have been reported for

zirconium,24 titanium,2 tantalum, molybdenum, tungsten, rhenium,26 ruthenium27 and


However, most dO alkylidene complexes are found in groups 5, 6 and 7,

alkylidene complexes found outside these groups often have ambiguous oxidation states.8

Without a doubt, molybdenum and tungsten are the most common and successful metals in

stabilizing metal-carbon double bonds.

Groups 3, 4, and the late transition metals tend to

form metal-carbon single bonds rather than double bonds. Examples in Figure 1.7 are the

Tebbe-Grubbs reagent, Cp2Ti(L-CH2)(4-C1)Al(CH3)2 in which the nucleophilic

methylidene ligand bridges the titanium and Lewis acidic aluminum atoms.29 However, in

the presence of phosphines, the methylidene complex is stable.25

Also, platinum,30

iridium,30 ruthenium,31 rhodium32 and thorium33 complexes exist as metallacycles (two

M-C single bonds) rather than as alkylidenes (one M-C double bond).

The reason for this periodic trend is not well understood.

However, it has been

suggested that the d orbitals of the early transition metals are very high in energy, resulting



+ PMe3

- ACi3


Et3P, _


Figure 1.7.


Examples of groups 3, 4, and late transition metals favoring single
bonds rather multiple bonds. a) Tebbe-Grubbs reagent; b) A
platinum metallacycle; c) A thorium metallacycle.

overlap, and bridged structures are again favored.

Thus, the considerations of bond

polarity and orbital overlap result in the preponderance of metal alkylidene complexes being

found with groups 5, 6 and 7.

Schrock-type alkylidene complexes are often stabilized by oxo and imido ligands.
These metal oxo/imido alkylidene complexes invariably maximize dxt-p7t bonding by

adapting a mutually cis orientation.14 This allows each ligand pit orbital to overlap with a

separate dA orbital. As a result, the substituents on the alkylidene ligand lie in the Ca-M-Y

plane, where Y is O or N of the spectator ligand.

For mono-substituted alkylidenes, the

syn or s-cis orientation has the hydrocarbyl substituent situated towards the spectator

ligand, and the anti or s-trans orientation has the substituent situated away from the

spectator ligand (Figure 1.8).

The syn and anti rotational isomers are referred to as

rotamers in the literature.34 Though the syn rotamer may seem to be not favored

thermodynamically due to steric arguments, it is observed in most compounds to be the

mnre. ctahle imnmer

This m .~nlt ic rhtinnali7MP hv the. nnAcihilitv nf nrntiC interntihnn


syn rotamer

Figure 1.8.

annt rotamer

Depiction of syn and anti rotamers for oxo/imido alkylidenes.
(E = O, NR)

Recently, in-depth studies have dealt with relative stabilities of the two alkylidene

rotamers and the energetic barriers that separate them.3436 The syn and anti rotamers can

be interconverted either thermally of photochemically, and rotational barriers have been

reported ranging from 12 to >21 kcal/mol.

For transition metal imido alkylidene

complexes, a mechanism for alkylidene rotation has been suggested and theoretical

calculations support such a barrier (Figure 1.9).14,37

The mechanism suggests that the

barrier to rotation is stabilized by the reforming of a metal-alkylidene nx bond in the


90 o M=C



,,.." R
^\i *


transition state.

nitrogen atom.


Mechanism for the rotation of a metal-carbon double bond.

In the ground state, the imido is triply bonded with an sp hybridized

Upon a 900 rotation of the metal-carbon bond, the sp nitrogen can

rehybridize to sp2

, giving a lone pair on the nitrogen and a bent imido bond.

This allows


accommodates the excess electron density rather than the carbon.

Alternatively, the

bonding in the transition state could be described as a three-center, four electron bond, and

rehybridization of the nitrogen atom would not be necessary.

Since its discovery, the metal-carbon double bond has received much study, and

thus much can be said about its nature. However, considerable strides have been made to

improve the synthesis and stability of metal-carbon double bonds.

The following section

will describe synthetic routes to metal alkylidene complexes as well as the effects of

ancillary ligands on their stability.

Synthesis and Characterization of Metal Alkylidenes

The synthetic routes to transition metal alkylidenes parallel organic reactions that

result in the formation of carbon-carbon double bonds.

These parallels include elimination

across a metal-carbon single bond, addition across a metal-carbon triple bond, and carbene

transfer by phosphonium ylides, or Wittig reagents.
The most widespread example of alkylidene bond formation is abstraction of an a-

proton of a metal alkyl.4 The first example was Schrock's compound,

Ta(CHC(CH3)3)(CH2C(CH3)3)3, from the reaction of two equivalents of neopentyl
lithium with Ta(CH2C(CH3)3)3C12 (Figure 1.10).38 The a-abstraction results during

dehydrohalogenation of the tantalum-carbon bond promoted by a strong base, the

neopentyl carbanion.

Dehydrohalogenation is also seen in the deprotonation of

Re(NPh)2(CH2C(CH3)3)C12 to give an alkylidene complex.39

[ qp



a .

Np3Ta C


Intramolecular abstraction of a-proton by another alkyl ligand is seen in a number

of systems, and this mechanism depends greatly on the steric environment of the metal
center. Heating or adding a coordinating ligand promotes a-abstracion in bisalkyl

complexes that do not spontaneously give alkylidene. Because of their steric bulk, the

neopentyl ligand and the related neophyl ligand (-CH2CMe2Ph) have a particular synthetic

advantage. In fact, most syntheses of stable, terminal alkylidene complexes employ these

bulky alkyl ligands.

Also, the neopentyl and neophyl ligands have no P-protons, and thus

3-hydride reductive elimination pathways are precluded.

Examples of intramolecular

abstraction included the syntheses of M(CHC(CH3)3)(NAr)(CH2C(CH3)3)24 and

M(CHC(CH3)3)(NAr)(DME)(X)241,42 (M

= W, Mo; NAr

= 2,6-diisopropylphenylimido;

DME = dimethoxyethane; X

= Cl, OSO2CF3).

In the case of

M(CHC(CH3)3)(NAr)(CH2C(CH3)3)2 (Figure 1.1 la), the formation of the alkylidene is
spontaneous with the increased steric bulk at the metal center. For

M(CHC(CH3)3)(NAr)(DME)(X)2 (Figure 1.1llb), the increased steric congestion caused
by the chelating ligand DME promotes alkylidene formation. A ligand-induced a-

abstraction process is also seen in the addition of trimethylphosphine to

W(NPh)(CH2C(CH3)3)2(PMe3)(C1)2 to give W(NPh)(CHC(CH3)3)(PMe3)2(C1)2 (Figure


Metal alkylidenes can also be formed by the protonation of metal alkylidynes. An

intermolecular example is the addition of two equivalents of HC1 to

W(CC(CH3)3)(OC(CH3)3)3 to give W(CHC(CH3)3)(OC(CH3)3)2(C1)2 (Figure 1.12a).44

Related examples include intramolecular and base-catalyzed proton transfer from amido

ligands to alkylidynes to give the metal imido alkylidene (Figure 1.12b).45

(a) II /Bu
tBu /--M r%
'Buj | ^^BU


+ Li

, -NpH













/ CI

Figure 1.11.

Examples of steric-induced (a) and ligand-induced (b, c) a-
abstraction reactions.


tBuO ...-W

- 'M







I ,,cl
M C -'Bu



-E -a *

Phosphonium ylides, or Wittig reagents, have been used to transfer their carbene
moiety to metal complexes. However, ylides do not react with metal oxo complexes to
give phosphine oxide and an alkylidene as ylides do with organic carbonyl compounds.
Rather, phosphonium ylides react with reduced metal centers to give oxidized metal
alkylidenes and free phosphine. An advantage of using ylides is that alkylidene ligands not
normally accessible by a-abstraction processes are possible, such as alkylidenes without

considerable steric bulk. Examples employing ylides are shown in Figure 1.13 for (a) the
formation of the first group 4 metal alkylidene46 and (b) the reduction of a tungsten (VI)
complex in situ to give a chelate-stabilized tungsten (VI) alkylidene complex.47,48

" "z.ZPMePh2






2. Ph3P=CHAr'

= (CF3)2(CH3)C-

Figure 1.13.

Alkylidene complexes via phosphoranes. a) The first stable
zirconium alkylidene complex; b) Tungsten alkylidene with unique

Other methods have resulted in the synthesis of metal alkylidene complexes,4 but
the routes described above are the most applicable to a number of metals and ancillary
ligand environments.

RF60 III .,tC

RF60 W CIl

field ranging from 200-350 ppm.

Coupling constants,

1JCH, range from 70 Hz to 160 Hz.

These large differences in chemical shifts indicate the variety of electronic and structural

environments in which alkylidene ligands may be found.

These differences in NMR data

are often correlated to the structure of the metal alkylidene complexes.

X-ray crystallography provides information concerning the metal-carbon bond

distances in metal alkylidenes. These distances range from 1.83(2) to

14(2) A for

tungsten complexes and from 1.917(5) to 1.989(3) A for molybdenum complexes.4

average, these bond distances are shorter for electronically unsaturated complexes and

longer when other x donor ligands are present.

ligands ranges from 1250 to 170

The M-Ca-Cp angle of the alkylidene

. Bond angles less than 150" are considered to be normal

when no agostic interaction of the a-CH bond is present. For bond angles greater than

1500, a agostic interactions are usually responsible and a corresponding decrease in 1JCH

(less than 120 Hz) is often observed.

Carbon-hydrogen coupling constants have been used to assign syn or anti

orientations for alkylidene ligands when oxo or imido ligands are present.3436 For

electronically and coordinatively unsaturated Mo(NR)(CHR')(OR")2 complexes,

1JCH for

syn rotamers range from 114.0 to 122.5 Hz, and 1JCH for anti rotamers range from 142.8

to 155.7 Hz.

Such variations in coupling constants are not observed for electronically and

coordinatively saturated complexes.35

Reactivity of Metal Alkylidene Ligands

The nucleophilicity of the a-carbon in Schrock-type alkylidene complexes defines

their reactivity.38

Alkylidenes react with a range of electrophiles from protons to

unsaturated, polar carbonyl-containing compounds.

In addition, the heightened reactivity

As metal alkylidynes are protonated by acids, alcohols, and amines to give
alkylidenes, alkylidenes are protonated to give metal alkyl complexes. For
Ta(CHC(CH3)3)(CH2C(CH3)3)3 in Figure 1.14a, the alkylidene ligand is protonated by
HC1 more readily than the less basic alkyl substituents.38 Legzdins has reported a low-
valent, Schrock-type alkylidene complex, CpMo(NO)(CHC(CH3)3)(pyridine) (Figure


that is protonated by alcohols and amines to give CpMo(NO)(CH2C(CH3)3)(ER)

= 0, NH).50 Protonation of an alkylidene is normally not desired synthetically, but in

the Legzdins work, the mixed alkyl-amido or -alkoxo complexes are not readily accessible
by other routes.

Np3Ta= C-







M --


Figure 1.14.

Protonation of alkylidene ligands. a) Protonation suggesting
greater basicity of alkylidene versus alkyl ligands; b) Protonation
with alcohols and amines as a synthetic route to mixed alkyl-ER

Carbon-hydrogen bonds can also be activated by metal alkylidenes.41

interesting example that also demonstrates the greater nucleophilicity of an alkylidene
versus alkyl ligands is the intramolecular e-C-H activation of an ancillary aryloxide ligand
hv a tnntnallm methvl mbthvlildene rmn1enlY (1irntrp 1 1 5~ 51

' CH2

/ O0


Figure 1.15.




Intramolecular e-C-H activation by a methylidene ligand.

Metal alkylidenes react with carbonyl compounds in a manner similar to Wittig

reagants to give metal oxo complexes and olefins.

The high electrophilicity of the metal

center makes alkylidene transfer possible for substrates not normally activated by

phosphonium ylides, such as esters, amides and carbon dioxide.

Thus, the Tebbe-Grubbs

reagent can be applied to organic synthesis for the methylenation of substrates not amenable

to Wittig reactions (Figure 1.16).


- (Cp2TiO).

Figure 1.16. Methylenation of a ketone that is susceptible to enolization.

Olefin Metathesis

The considerable interest in metal alkylidene chemistry has been primarily focused

on their reactions with olefins.

The topics of the olefin metathesis reaction, its use in

polymerizations, and the requirements for highly active olefin metathesis catalysts are

covered in detail in this section.

The olefin metathesis reaction interchanges the carbon


metathesis is unobserved as a purely organic transformation, and thus a transition metal

catalyst is required (Figure 1.17).53


metal catalyst



Figure 1.17.

Transition metal-catalyzed olefin metathesis.

Olefin metathesis catalysts were discovered as a result of intense research efforts to

polymerize a-olefins by Ziegler-Natta catalysts.1

Early Ziegler-Natta catalysts that gave

stereoregular, unsaturated polymers from a-olefins were composed of transition metal

halide compounds and Lewis acid cocatalysts, such as alkyl aluminum compounds.

Attempts to extend Ziegler -Natta chemistry to group 6 metals resulted in the first olefin
metathesis catalysts.
The nature of the catalytic species in these early catalysts and the mechanism of the
reaction were unknown for years even though olefin metathesis was highly exploited by
industry. After years of observations of product ratios, reaction kinetics, and model

studies from several researchers,54 H6rison and Chauvin proposed that the active catalytic

species consisted of a metal alkylidene and that the mechanism involved the fragmentation

of reactant olefins to give product olefins.3

This non-pairwise mechanism is shown below.

The non-pairwise mechanism, shown in Figure 1.18, consists of a series of

equilibria involving (a) coordination of an olefin to a metal alkylidene complex, the (b)
formation and (c) constructive decomposition of a metallacyclobutane, and (d) the

dissociation of the new olefin from a new metal alkylidene complex.

The formation of a

reactive metal alkvlidene complex in some classical catalyst mixtures results from the

[M] C







Figure 1.18.

The non-pairwise mechanism for olefin metathesis.

Polymers can be synthesized by olefin metathesis in two distinct ways: the ring-
opening metathesis polymerization (ROMP) of cyclic olefins (Figure 1.19)1 and acyclic

diene metathesis (ADMET) polymerization (Figure 1.20).

Both reactions result in linear,

unsaturated polymers, but they differ greatly from a synthetic viewpoint.
The ROMP reaction involves the metathesis of strained cyclic olefins, and it is the

relief of ring strain that drives the reaction to completion.

The advantages of ROMP

include a chain-growth polymerization mechanism in which the growing polymer chain

never dissociates from the metal center in well-behaved systems.

This results in rapid

increases in molecular weight and narrow molecular weight distributions.

n -..--

Acyclic diene metathesis differs from ROMP in that ADMET is a step-growth,

condensation polymerization.

The equilibrium of this condensation polymerization is

driven to polymer by the removal of small, low-boiling by-products, such as ethylene. A

consequence of the step-growth mechanism is that high molecular weights are only

achieved at quantitative conversion and polydispersities are statistically broad.

side reactions must be eliminated and monomer purity is imperative.56

As a result,

Figure 1.20. Acyclic diene metathesis (ADMET) polymerization.

ADMET polymerization was not viable in the classical catalyst systems because of

side reactions such as vinyl addition, but by employing the highly active, Lewis acid-free

metathesis catalysts reported by Schrock, the quantitative conversion of monomer to

polymer was first achieved by the Wagener research group at the University of

Florida.57'58 The major advantage of ADMET over its ROMP counterpart is that strained

cyclic monomers are not necessary, and a wider array of monomers with a variety of

functionalities are suitable candidates for polymerizations. Also, it has been shown that

depolymerization of unsaturated polymers in the presence of excess olefins is possible by

nature of the reversible equilibrium.59,60

As mentioned, the success of ADMET depended greatly on the use of highly active,

Lewis acid-free metathesis catalysts of the form M(CHR)(NAr)(ORF6)2 (M

= Mo, W; R

t-Bu; Ar

= 2,6-(i-Pr)2C6H3; ORF6

= OC(CH3)(CF3)2) (Figure 1.21).

These new

metathesis catalysts were the dawn in a new era in catalyst activity and design, and a

discussion of their development provides considerable insight into some essential elements
nf native mettatheic reatalvete 61



catalysts were five or six coordinate to avoid dimerization of the complexes and required

Lewis acids to open a coordination sight62'63 The use of dianionic oxo or imido ligands

would lower the coordination number to four, and with imido ligands, a sterically

demanding substituent could deter dimerization.

Thus, Schrock employed the bulky 2,6-

diisopropyl imido ligand to shield the metal center and maintain a low coordination number

and a high oxidation state.61

The use of the K-donating imido ligand is also advantageous

because of its ability to stabilize the catalytic intermediates as a 'spectator ligand'.

CF3 ..7mhn II



Figure 1.21.

The Lewis acid-free metathesis catalyst.

Active metathesis catalysts seem to require K-donation from the other ancillary

ligands. Alkoxide ligands have an advantage over halide ligands due to their enhanced
stabilization of intermediates by it-donation.64 On the other hand, it is necessary to

maintain an electrophilic metal center to enhance olefin binding.

Schrock employed a series

of relatively bulky t-butoxides that were fluorinated to increase the electrophilicity of the

metal.65 The metathesis activities depend dramatically on the balance of electronic and

steric properties of the alkoxide ligand employed, following the order OC(CH3)(CF3)2 :


> OC(CF3)3

> OC(CF3)2(CF2CF2CF3)

>> OC(CH3)3.

= W

differences in their preference for terminal versus internal olefins.

The tolerance of

functionalities for molybdenum is attributed to the molybdenum-carbon double bond being

less polar than that of tungsten.42 For ADMET, these functional groups include amines,

ethers,66 thioethers,67 ketones, esters,68 carbonates,69 carbosilanes70 and


Metathesis activity for internal olefins is greater for tungsten while

metathesis activity of terminal olefins is greater for molybdenum.42

As a result, ADMET

polymerizations utilize the molybdenum catalyst and depolymerizations utilize the tungsten

catalyst.6'68 Elimination of the non-productive, internal olefin metathesis reactions would

greatly enhance the rate of ADMET.

The success of Schrock's catalysts in ADMET polymerizations has led to a wide

range of new polymers not previously attainable with ROMP monomers.


drawbacks still exist, and catalysts with greater ADMET reactivity and applications are


The drawbacks most importantly are due to the thermal sensitivity of the catalyst

during polymerizations. Since ADMET is a step polymerization, the polymerization is

preferably performed in neat monomer in order to maximize the concentration of reactive

end groups. As a result of this condition, solubility problems arise for the growing

polymer. Neat poly(octenamer) solidifies at ambient temperature after only reaching a

degree of polymerization of 11.68 Heating the reaction would increase solubility or

essentially allow the reaction to take place in the melted phase. In addition, step

polymerizations are inherently slower than chain polymerizations, and increasing the

temperature would accelerate the overall reaction.

Chelated Transition Metal Alkvlidene Comolexes

The problems associated with catalyst sensitivity can be addressed by the use of


dimerization, and thus increases the stability of a complex over a wider range of

temperatures and conditions. As a consequence of the increased stability, catalytic activity

often suffers.

The increased hapticity, or coordination number, of the ligand can occupy

vacancies necessary for substrate coordination. Also, the chelating ligand decreases the

fluctionality and alters the coordination geometry of the complex, and certain

rearrangements or related processes may be critical to the mechanism of catalysis. Because

a balance between stability and reactivity is necessary in the design of d transition metal

alkylidenes, it is desirable to have chelating ligands of high charge and low coordination


In varying the hapticity and charge of the ligand system in neutral group 6 metal

imido alkylidene complexes, the ancillary ligands must have a combined charge of

Ideally, a bidentate, dianionic chelating ligand would satisfy the dO requirement and result

in the lowest possible coordination number of four.

Chelating ligands that are neutral or

monoanionic only raise the total coordination number and the electron count at the metal

center. Examples of a variety of hapticity/charge combinations are reported in the literature

and are shown in Figure 1.22.72-79

Of particular note are the examples using hydridotris(pyrazolyl)borate (Tp, Figure

1.22e)76-78 and N,N'-bis(trimethylsilyl)-o-phenylenediamine (TMS2pda, Figure 1.22f)

ligands.79 Both ligands are the focus of study in this dissertation, and the previous work

from our research group using these ligands is described below.

Polypyrazolylborates have found wide application as chelating ligands since their

synthesis was described in 1967.80*81

This class of ligand has been used with virtually

every metal and metalloid in the periodic table. A recent review by their discoverer

Swiatoslaw Trofimenko references 460 papers from 1984 to 1993!82

Mostly the tridentate,







= Si(CH3)2Ph

a. bidentate, dianionic

~- cm"s-r^sr^



" H

c. pentadentate, monoanionic


= O, NAr
= Me, Ph
= CI, Br, OTf,
alkyl, solvent

bidentate, monoanionic

\ /H

, 2r-

e. tridentate, monoanionic



Figure 1.22.

Multidentate ligands used in the stabilization of metal-carbon
multiple bonds (hapticity/charge).
a) Binaphtholate (2/-2)
b) Tris(amidoethyl)amine (3/-3)
c) Bis(phosphino)cyclopentadienide (5/-1)

to give Tp'W(CC(CH3)3)C12 and Tp'W(CHC(CH3)3)(0)Cl (Tp'

= hydridotris(


The Tp and Tp' tungsten alkylidene complexes synthesized by the Boncella group

have a generalized formula Tp*W(CHR)(E)(X) (Tp*

= Tp, Tp'; R

C(CH3)2Ph; E

= 0, NPh, N(2,6-i-Pr2C6H3); X

= Cl, Br, OSO2CF3, alkyl, donor


These complexes exhibit exceptional air, moisture, and thermal stabilities in

comparison to the Schrock catalysts.

Their high stability is a result of electronically

saturated, six-coordinate metal centers imparted by the facially bound, tridentate ligand.

The increased stability allows for synthetic manipulations of the complexes not available to

the more reactive Schrock catalysts. Highly electrophilic, cationic compounds can be

synthesized,78 and detailed rotamer isomerization studies were possible.

However, the

saturated metal centers did not result in the complexes being active metathesis catalysts.83

Catalytic activity for the ROMP of cyclooctene and norbornylene and ADMET

oligomerizations was induced by the addition of the Lewis acid A1C13, but the active

catalytic species remains unknown.

In order to allow metathesis activity without Lewis acids while still gaining

advantages from chelating ligands,

VanderLende and Boncella used the sterically

demanding N,N'-bis(trimethylsilyl)-o-phenylenediamine (TMS2pda) ligand to stabilize the

metal alkylidene complex (Figure 1.23).79.84 The five-coordinate TMS2pda tungsten



A, 24 h





r?:..... i it


Q ,n 4ltAfn an nL at L a 11 AC nA nh** n1 i.1^^]j.. Z A ana I

= C(CH3)3,


in toluene at 110 OC. However, due to the nature of the silicon-nitrogen, tungsten-amido,

and tungsten-alkylidene bonds, the complex is extremely air and moisture sensitive.

In continuing research in the area of chelate-stabilized transition metal alkylidene

complexes, two projects were undertaken using the Tp and the TMS2pda ligand systems.

Molybdenum analogues of the Tp tungsten compounds were prepared.

Tungsten and

molybdenum complexes often behave quite similarly with the exceptions mentioned above.

Because the Tp tungsten compounds were active metathesis catalysts in the presence of

Lewis acids, it was hoped that the advantages of molybdenum might be gained pertaining

to the higher affinity of terminal olefins and tolerance of functionality.

molybdenum imido alkylidene complexes is covered in Chapter Two.

The chemistry of Tp

The chapter contents

have either previously appeared in refereed journals or are currently in press.8587

Attempts were also made to extend the use of TMS2pda to molybdenum.

Regrettably, the lower oxidation potentials for molybdenum versus tungsten resulted in the

reduction of molybdenum to presumably d1 decomposition products. However, the

TMS2pda tungsten alkylidene complex and its precursors provided rich opportunities for

the synthesis of chelated tungsten alkyl and hydride complexes.84

As a result of the

synthetic observations, it became necessary to study the kinetics and mechanisms of the

formation of these new TMS2pda tungsten complexes.

Close examination of the system

has revealed that the TMS2pda ligand is not merely ancillary and innocuous.

Chapter Three

details the observations gathered to date and mechanistic relationships inferred between the

TMS2pda tungsten complexes.


The success of using hydridotris(pyrazolyl)borate, Tp, to stabilize tungsten

complexes led to the extension of this chemistry to molybdenum.

Reasons for exploring

the use of molybdenum are more than just its similarity to tungsten. As previously

mentioned, molybdenum alkylidene complexes are typically more tolerant of functionalized

substrates and show higher metathesis activity for terminal olefins than tungsten


Though the reaction chemistry of molybdenum and tungsten parallels one

another in most cases, do molybdenum complexes are often observed to be less stable to

reduction than tungsten. As a result, some starting materials available for tungsten are not

as readily accessible for molybdenum.

The molybdenum imido alkylidene complex,

Mo(CHC(Me)2Ph)(NAr)(OTf)2(DME), reported by Schrock is a suitable starting material

to enter into the chemistry of chelated hydridotris(pyrazolyl)borates.


molybdenum imido alkylidene complexes with Tp has led to chemistry previously seen

with tungsten, including the synthesis of cationic complexes and ROMP catalyst


In addition, several new complexes and transformations not seen with

tungsten have been observed, and analysis of the series of new compounds has offered

further insight to the nature of metal imido alkylidenes in the octahedral environment of the

Tp ligand.

The contents of this chapter are subdivided into new neutral Tp molybdenum


Neutral Molvbdenum Imido Alkvlidene Complexes

The chelating Tp ligand provides an excellent opportunity to substitute alkylidene

complexes with different ancillary ligands.

This section describes the synthesis,

characterization and reactivity of neutral molybdenum imido alkylidene complexes

beginning with TpMo(CHC(Me)2Ph)(NAr)(OTf) (1) which proved to be a useful starting

material that is readily available.

Synthesis of TpMo(CHC(Me)2Ph)(NAr)(OTf) (1)

The reaction of one equivalent of KTp with

Mo(CHC(Me)2Ph)(NAr)(OTf)2(DME)42 in THF at ambient temperature results in the rapid

formation of TpMo(CHC(Me)2Ph)(NAr)(OTf) (1, eq 1) which is obtained as bright yellow

needles from pentane.

Compound 1 is stable in air as a solid for one month with no

evidence of decomposition by 1H NMR and it decomposes in air at 155 OC.

The chelating

ability of the Tp ligand and the increased steric bulk at the metal,88 as well as the formal 18

electron count at the metal, are responsible for the robust nature of compound 1.

KTp +




Figure 2.1.



i /

N Ph

Synthesis of TpMo(CHC(Me)2Ph)(NAr)(OTf) (1).







* C)4

S S\

* 0

of the aryl imido ligand.

The proton signals of the isopropyl groups of the aryl imido

ligand are broad, suggesting hindered rotation about the nitrogen-aryl bond.


temperature NMR measurements were used to measure a barrier to rotation of 15 kcal/mole

for the nitrogen-carbon bond.

The solution structure was determined by 1H difference nOe

experiments, and found to be the syn isomer in which the alkylidene C(Me)2Ph group is

oriented towards the imido ligand.34 The triflate ligand is covalently bound to the metal

center, demonstrated by strong characteristic IR absorptions at 1202 and 633 cm-1 assigned

to the S--= stretches of the triflate ligand.89

Crystals of 1 suitable for x-ray diffraction

studies were obtained by slowly cooling a saturated hexane solution of 1 from 40 OC to

ambient temperature.

Crystal Structure of TpMo(CHC(Me)2Ph)(NAr)(OTf) (1)

There are several pertinent features of the structure of 1 (Figure

3) that are

analogous to features of other structurally characterized Tp or imido-alkylidene


Selected bond distances and angles are given in Table 2.1.

The coordination

geometry around the molybdenum atom in 1 is that of a distorted octahedron.

geometric constraints of the facially-bound Tp ligand force the neophylidene, aryl imido,

and triflate ligands to be mutually cis.

The N-Mo bond lengths of the chelating pyrazolyl

rings [Mo-N1, 2.311(8)A; Mo-N3, 2.167(8)A; Mo-N5, 2.311(9)A] are consistent with the

decreasing trans influence of the ligands imido

= alkylidene

> triflate.

The cis orientation of the aryl imido ligand and alkylidene ligand allows for
maximum dx-ps bonding between the molybdenum and the multiply bonded ligands. The

Mo-N bond length of 1.753(8) A and Mo-N-C11 bond angle of 170.9(7)0 are within the

normal ranges for molybdenum imido complexes in which the molybdenum-nitrogen bond



o -,

-- -)

0z o 0
a 2 ma

cu a

r\ ag

0 O Ln

o> Z




0 Z

Table 2.1:


Selected Bond Lengths (A) and Angles (0) for compound 1.










The Mo-Cl bond length, 1.949(10) A, lies within the expected range for

molybdenum-carbon double bonds. The -CMe2Ph group of the alkylidene is in the syn or

s-cis orientation with respect to the imido ligand. This syn orientation for the neophylidene

ligand is also present in the solution structure as determined by 1H difference nOe

difference spectra

The steric congestion imposed by the syn orientation of the

neophylidene ligand results in the aryl ring of the imido ligand being situated perpendicular

to the plane containing atoms N, Mo, C1 and C2.

with molybdenum [Mo-C1-H1

The acute angle of the alkylidene proton

= 104.(5)] is not likely a result of any agostic interaction

since the molybdenum atom is coordinatively and electronically saturated, but rather due to

the opening of the Mo-C1-C2 angle caused by the steric constraints of the coordination


The Mo-OTf bond distance of 2.121(7)A is long compared to the Mo-OMe bond


ligands.90 The presence of a bound triflate ligand was also supported by the presence of

characteristic S=0O stretches in the infrared spectrum of 1.

Though the triflate anion is often

thought of as being non-coordinating, the high electrophilicity of the metal center and the

ability of the Tp ligand to efficiently polarize metal orbitals into an octahedral array 88 lead

to the interaction between the metal center and the relatively ionic, poor x-donating triflate


The triflate ligand is slightly disordered which can be described as a distortion of

the Mo-O1-S angle propagating into increasing thermal parameters for 02, 03,

C32, Fl,

F2 and F3 atoms.

Synthesis of TpMo(CH2C(Me)bPh)(NAr)(0) (2)

Compound 1 in THF is stable in the presence of H1120 over a period of 24 hours.

However, when a wet Et20 solution of 1 is passed over a column of alumina, a small

amount of TpMo(CH2C(Me)2Ph)(NAr)(0) (2) was observed by 1H NMR.


compound 1 in a slurry of Florisil in Et20O and H20 for 24 hours afforded an orange

solution of compound 2 which can be recrystallized from pentane to give orange crystals.

This result compares with a previously reported reaction by our group in which a tungsten

alkylidyne complex, TpW(CCMe3)C12, is converted to the tungsten alkylidene complex,

Tp'W(CHCMe3)OC1, by stirring over alumina.76

Also, the reaction of compound 1 with

one equivalent of CsOH in THF gave a quantitative yield of 2. Heating 2 in



for 5 days at 65 OC showed no reaction by 1H NMR and no evidence of phosphine oxide or


Tp Mo

HzO. Et2O. Florisil
or CsOH-H2, THF



0 I




The distinctive features of the 1H NMR spectrum (Figure 2.5) of compound 2 are

the presence of an AB quartet corresponding to the diastereotopic protons of the

CH2C(Me)2Ph ligand and the loss of the downfield alkylidene resonance of the alkylidene

proton of 1.

The isopropyl groups of the aryl imido ligand are responsible for a broad

multiple due to the methine protons and two slightly broadened doublets due to the methyl

protons indicating that the N-Ar bond rotates more freely than that of compound 1. The

molybdenum-oxygen bond appears as a strong absorption at 896 cm-1 in the IR spectrum

which is within the normal range for molybdenum oxo compounds containing additional

multiply bonded ligands.4

One possible mechanism for the formation of 2 is the displacement of the triflate

ligand by hydroxide to form an unobserved hydroxy-alkylidene species (Figure 2.4).

Subsequent proton transfer from the oxygen to the a-carbon of the alkylidene results in the

favorable formation of the strong molybdenum-oxygen triple bond.

The previously

reported tungsten analogue of 2 was prepared by the protonation of TpW(CHC(Me)2Ph)

(NAr)(CH2C(Me)2Ph) by HBF4 in the presence of H20 to give [TpW(NHAr)(0)

(CH2C(Me)2Ph)][BF4] followed by deprotonation by NEt3.78 This result suggests an

acid-catalyzed mechanism for the formation of 2 in the Florisil preparation.

Crystal Structure of TpMo(CH2C(Me)2Ph)(NAr)(O) (2)

The structure of compound 2 was determined by X-ray diffraction methods and a

thermal ellipsoid plot is found in Figure 2.6.

in Table 2.2.

Selected bond distances and angles are given

The structure consists of well-separated molecules with a pseudooctahedral

coordination geometry about the metal.

The geometric constraints of the facially-bound Tp

I ~~ --- -






k^ ^


X ---2








2.379(2)A to 2.207(2)A and are consistent with the decreasing trans influence of the

ligands oxo

> imido

> alkyl.4

Table 2.2:


Selected Bond Lengths (A) and Angles (0) for compound 2.









In transition metal-oxo and -imido compounds in which the formal electron count of

the metal is less than 18 electrons, triply bonded oxo and imido ligands are the preferred

valence bond description by donation of a lone pair of electrons of the ligand to an empty

metal d orbital of appropriate a-symmetry.

In compound 2, however, the metal has an

electron count of 20 if both the oxo and the imido are triply bonded.

The crystal structure

of 2 shows the oxo and the imido ligands to have bond orders between two and three with

metal bond lengths of 1.706(2)A and 1.760(2)A, respectively.4,17'18

bond angle, 168.07(13)0

Also, the Mo-N-Ar

, approaches linearity, suggesting to a first approximation that the

molybdenum-nitrogen bond order is greater than two.

However, the steric bulk at the

metal center may also contribute to the linearity of the sterically demanding aryl imido


Crystal Structure of [TpMo(NAr)(O)a0 (3)

A red single crystal suitable for x-ray diffraction studies was isolated from the

pentane mother liquor of 2 after two weeks.

shown in Figure 2.7, was obtained. Compc

The structure of [TpMo(NAr)(0)]20 (3),

)und 3 appears to be a hydrolysis product of

compound 2 in which the neophyl ligands of two equivalents of 2 are protonated and lost

and the metal centers are bridged by an oxygen atom, most likely from H20.

The synthesis

of compound 3 was attempted in two ways. H20 and 2,6-lutidine were added to

compound 2 dissolved in THF, and H20 and HBF4 were added to 2 dissolved in Et20.

Both attempts were unsuccessful with only unreacted 2 being recovered. [TpMoOCl]20, a

molybdenum (V) analogue of 3 has been prepared by the hydrolysis of TpMoC13.91

The crystal structure of compound 3 shows the molecule to possess a center of

inversion at the position of the bridging oxygen atom, with the oxo, imido, and other

ligands being trans to one another.

Selected bond distances and angles are given in Table

2.3. The crystal consists of well-separated molecules with the pseudooctahedral

coordination geometry and metal-ligand bond lengths about the molybdenum centers not

differing remarkably from that of compound 2.

The N-Mo bond lengths of the chelating

pyrazole rings range from 2.342(11)A to 2.200(12)A and are consistent with the

decreasing trans influence of terminal oxo > imido

> bridging oxo [1].

The Mo-O1-Moa

angle is linear as required by symmetry and the Mo-O1 bond length (1.879(1)A) is

shortened92 suggesting the bridging oxygen can participate in stabilizing the dO metal
centers through di-pit donation. However, the steric bulk about the molybdenum atoms

likely contribute to the linearity of the Mo-01-Moa angle.

It is concluded that the stability of the described molybdenum compounds arises
from the chelating ability of the Tp ligand and the presence of coordinatively and


'S p.1.4


demonstrate the effects of competition between the multiply bonded ligands for the limited
number of x-symmetry metal orbitals.

Table 2.3:


Selected Bond Lengths (A) and Angles (0) for compound 3.












Synthesis and Characterization of TpMo(CHC(CH3)TPh)(NAr)(OCH3) (4)

So far, the compounds of the general formula TpM(CHC(CH3)2R)(NAr)(X) [M =

Mo, W; R = CH3, C6H5;

Cl, Br, OTf, pyrazolide, alkyl] described herein and

elsewhere have X being electron withdrawing substituents and/or poor x donors.76-78'83

We were interested in replacing the triflate ligand in compound 1 with a better K donating

substituent, namely methoxide, to observe how this substitution would effect the chemistry

of these molecules.

Similar to the hydrolysis of 1 to give TpMo(CH2C(CH3)2Ph)(NAr)(O),


crystals. An unidentified insoluble white powder remained from the pentane extraction,

however a low resolution mass spectrum indicated the presence of molybdenum.


NAr H3


CH30OH, Florisil
Et20O, RT, 24 h

Tp --




Figure 2.8.

Synthesis of TpMo(CHC(Me)2Ph)(NAr)(OCH3) (4).

The 1H NMR spectrum (Figure 2.9) of compound 4 at 22 C is comparable to its

parent compound.

The spectrum has nine distinct pyrazole proton resonances and

diastereotopic methyl groups for the neophylidene ligand and isopropyl groups of the aryl

imido ligand.

The broadening of the isopropyl group proton resonances suggests hindered

rotation about the nitrogen-aryl bond due to the steric demands of the Tp ligand.

chemical shift of the alkylidene proton is 8 13.14.

This is one the highest field shifts for

these molybdenum alkylidene compounds indicating a more electron rich metal center. Tl
new resonance corresponding to the methyl protons of the methoxide ligand is found at 8

4.60. Absorptions due to the methoxide ligand are also observed in the IR spectrum at


and 1081.8 cm-1 corresponding to the methoxy C-H and C-O stretches,


Synthesis and Characterization of TpMo(CC(CHq2Ph)(NHAr)(OCH3) (5)

In a separate attempt to prepare TpMo(CHC(CH3)2Ph)(NAr)(OCH3), compound 1
was allowed to react with one equivalent of potassium methoxide in THF at room

9- a-=





< 0o.-0

singlets at 8 9.55 and 8 10.26 and two methoxide singlets at 8 5.11 and 8 4.98.

A 13C

{H} NMR spectrum confirmed the presence of sp-hybridized alkylidyne carbons at 8

301.7 and 6 298.7, and the structure was determined to be the major and minor rotamers of

TpMo(CC(CH3)2Ph)(NHAr)(OCH3) (5, Figure

This molybdenum amido

alkylidyne complex is the tautomer of 4.

Tp -





Tp -- Mo..


4- Ph

Figure 2.11.

Synthesis of TpMo(CC(Me)2Ph)(NHAr)(OCH3) (5).

The major and minor isomers arise from rotational isomers of the molybdenum-amide


Variable temperature 1H NMR experiments showed the two amide proton

resonances to coalesce at 65 oC.


This corresponds to an activation energy, AGt, of 15.7

Cooling the solution to room temperature reestablishes the same equilibrium ratio

of rotamners and shows that AGo for the two rotamers is 0.8 kcal/mol.

The IR spectrum of

5 reveals a single sharp absorption at 3308.7 cm-1 for the N-H stretch of the amido ligand.

Since compounds 4 and 5 are formally related by transfer of the alkylidene proton

to the nitrogen of the imido ligand, we tried to convert 4 to 5.

Compound 4 was stable

towards tautomerization for one week in dg-toluene. Photolysis, continued heating at 110

C, and addition of Lewis bases (NEt3, P(CH3)3) did not cause tautomerization of 4 to give

5. Only the addition of excess potassium methoxide caused the proton to transfer from the

alkylidene to the imido ligand.

To avoid adding excess methoxide, a THF solution of 0.9

equivalents of potassium methoxide was added dropwise to 1.

However, only 5 resulted


The experimental results concerning the formation of 5 and orbital considerations
suggest that the mechanism for the formation of 5 involves a methoxide mediated proton





II __


_ ^II
- OCH3


Figure 2.12.

Proposed mechanism of the formation of 5.

In Figure 2.12, a methoxide anion deprotonates the alkylidene ligand of either the
starting material, 1, or compound 4 forming an anionic molybdenum species with the
negative charge localized on the imido nitrogen. Following this step is a reprotonation of

the metal complex at the nitrogen by methanol.

The reverse reaction to deprotonate the

amido ligand and reform the alkylidene is less likely due to the stability of the strong metal-
carbon triple bond. Also, the amide would be expected to be less acidic in complexes with

good x-donors versus those with electron-withdrawing triflates.

The conversion of 4 to 5

is the reverse of the proton transfer reaction observed by Schrock and co-workers for the
NEt3-catalyzed conversion of M(NHAr)(C-t-Bu)(dme)C12 to M(NAr)(CH-t-Bu)(dme)C12

W, Mo]

They observed that the same reaction was impossible when the chlorides of

the metal amido alkylidyne complexes were replaced with hexafluoro-t-butoxide

Attempts to isolate derivatives of 4 and 5 with the stoichiometry

TpMo(CC(CH3)2Ph)(NHAr)(X) [X

= N(CH3)2, O-p-tol, OPh] were unsuccessful giving

complex mixtures of imido-alkylidene, amido-alkylidyne, and other species which have
proven impossible to separate.

Crystal Structure Determination of 4 and 5

The structure of compound 4 was determined by X-ray diffraction methods and a

thermal ellipsoid plot is found in Figure

given in Table 2.4.

Selected bond distances and angles are

The structure consists of well-separated molecules with the

coordination geometry around the molybdenum atom in 4 being a distorted octahedron.

The geometric constraints of the facially-bound Tp ligand force the alkylidene, imido, and

methoxide ligands to be mutually cis.

The N-Mo bond lengths of the chelating pyrazole

rings range from 2.327(10)A to 2.232(8)A and are consistent with the decreasing trans

influence of the ligands imido

Table 2.4:


> alkylidene > methoxide.1

Bond Lengths (A) and Angles (0) for compound 4.









mi- -* -P I _.1.?_ ^ -i _I_ 1 -~-f -i *-_. r _- -l __L- i 1 l -- _


It, ES

~" C,

e H n

3 '4408


normal ranges for molybdenum imido complexes in which the molybdenum-nitrogen bond

can be considered to be triply bound with the lone pair of the nitrogen donating to an empty

dx orbital of the dO molybdenum atom.

The Mo-C1 bond length, 1.963(10) A, lies within

the expected range for molybdenum-carbon double bonds, and the -C(CH3)2Ph group of

the alkylidene is in the syn, or s-cis, orientation with respect to the imido ligand.

The steric

requirements of the bulky aryl imido group force the Mo-C1-C4 angle to open to 136.8(10)

A. The syn orientation is preferred to the anti orientation due to the apparently unfavorable

steric interaction of wedging the -C(CH3)2Ph group between two pyrazole rings of the Tp

The crystal structure of compound 5 was also determined by X-ray diffraction

methods and a thermal ellipsoid plot is found in Figure

Selected bond distances and

angles are given in Table 2.5. Similar to 4, the geometry around the molybdenum is a

distorted octahedron dictated by the geometric and electronic constraints of the Tp ligand.

The N-Mo bond lengths of the chelating pyrazole rings range from 2.386(4)A to 2.235(4)A

and are also consistent with the decreasing trans influence of the ligands alkylidyne >

amido > methoxide.4 The Mo-C1 bond length and angle [1.765(4)A, 177.2(4)0] are within

accepted ranges for do molybdenum alkylidyne bonds, and the sp hybridized alkylidyne

carbon donates to two metal d orbitals of x symmetry.

Thus, compound 5 is formally a

sixteen electron complex with one dc metal orbital available for electron donation from the
aryl imido and/or methoxide ligands.

The molybdenum-aryl amido bond length and the Mo-N-C angle [1.952(3)A,

141.1(3)0] differ markedly from the triply bound and linear aryl imido ligand of 4. The

amido proton was located for compound 5, and the geometry about the amido nitrogen is

essentially planar.

This planar geometry is consistent with the lone pair of the nitrogen


0) (N
o -



,- )


both rotamers are present.

The syn rotamer is assumed be the major rotamer in solution.

The molybdenum-methoxide bond in compound 5 is slightly shorter than the Mo-O bond

of compound 4 by 0.032 A, and may indicate a higher degree of it-bonding between the

oxygen atom in 5 versus 4. However, as evidenced by the geometry about the nitrogen

and the respective trans influence of the arnido and the methoxide ligands, the amide is a

better x donor than the methoxide.

Table 2.5:


Selected Bond Lengths (A) and Angles (0) for compound









Cationic Molvbdenum Imido Alkvlidene Complexes

The stability imposed at the molybdenum metal center by the tridentate Tp ligand

enables the modification of the TpMo(CHC(CH3)2Ph)(NAr) template by varying the ligand

in~ ~ ~~~~~~~~~~~~, 44,a ori nnn..A, nnna f I, l: tan n A

*n -1r *~~l -<~~inrr t* *


for such a transformation was to abstract an alkyl group from a complex of the type

TpMo(CHC(CH3)2Ph)(NAr)(alkyl) in the absence of a coordinating anion to give the

desired cationic metal complex.

By employing large, non-coordinating counterions, the

open coordination site of the cationic complex should readily bind olefin substrates or

coordinating solvent molecules. Similar schemes have been successful in preparing and

isolating cationic catalysts for olefin-insertion polymerizations.94,95

Cationic tungsten

alkylidenes have been prepared by using Br0nsted acids of non-coordinating anions to

protonate the complex followed by loss of the alkyl ligand.78

SynthesisFor the puand Characterization of eventuallyMo(C preparC(CHing caPh)(NAr)CHkyidenes,) (6)

For the purposes of eventually preparing cationic alkylidenes, the alkylation of 1

with a methyl group proved to be the most facile route.

The transmetallation of 1 with a

moderate excess of methyl lithium in Et20 at ambient temperature gave

TpMo(CHC(CH3)2Ph)(NAr)(CH3) (6, Figure

15) in good yield.

Compound 6 exhibits

marked air and moisture stability.

stable as a solid in air indefinitely.

no reaction.

Compound 6 can be purified over neutral alumina and is

Stirring 6 in THF with an equivalent of H20 showed

In dg-toluene at 80 C over 3.5 hours, compound 6 decomposes, forming

unidentified products.



EtzO, RT

II "


~ Ka



The 1H NMR spectrum (Figure 2.16) of 6 at 22 OC is complicated by the

asymmetry of the molecule, resulting in nine distinct pyrazolyl ring proton resonances and

diastereotopic neophyl methyl resonances.

The alkylidene proton resonance is shifted

upfield from that of 1 at 14.73 ppm to 13.11 ppm, and the 1JCH observed in the 13C(H)

spectrum is reduced slightly to 115 Hz.

The methyl proton signal at 1.29 ppm is a sharp

singlet, and the methyl carbon resonance at 18.0 ppm in the 13C NMR has a 1JCH of 123

Hz. This 1H-13C coupling constant is normal for sp3 C centers96 and does not suggest any
interaction of the C-H o bond with the metal center as is expected for a coordinatively and

electronically saturated metal complex.5 The proton signals for the isopropyl groups of the

aryl imido ligand are broadened and occur in a 6:3:3 ratio, suggesting sterically hindered

rotation about the carbon-nitrogen bond.

Hindered rotation has been observed in other Tp

aryl imido alkylidenes.77

Protonation of 6 with one equivalent of triflic acid in C6D6 resulted in a quantitative

conversion to TpMo(CHC(CH3)2Ph)(NAr)(OTf) (1) and methane. Addition of a second

equivalent of triflic acid causes loss of the alkylidene proton resonance and formation of

unidentified decomposition products. In the presence of one equivalent of H20, treatment

of 6 with triflic acid in Et20 only yielded 1.

Synthetic Routes to Cationic Molybdenum Imido Alkylidene Complexes

Removing the methyl group from compound 6 led to several cationic molybdenum

alkylidene complexes.

Three methods were successfully employed to remove the methyl

ligand: protonation by the acid of a non-coordinating anion, abstraction by the trityl cation,

and abstraction by a boron-containing Lewis acid.

The anions used were fluorinated aryl

The addition of tetrakis(3,5-trifluoromethylphenyl)boric acid (HBAr'4

= 3,5-C6H3(CF3)2)97 to a Et20 solution of 6 at

[TpMo(CHC(CH3)2Ph)(NAr)(Et20)][BAr'4] (7, Figure


2 Et20, Ar'

o C gave the thermally unstable cation

17) as a brown oily solid and

Compound 7 is insoluble in hydrocarbons and benzene, and the cation

decomposes within hours at room temperature in CD2C12.

NAr Ha(
Tp Mo=<





-78 C

Tp -

NAr Hu

Mo -
\ H


+ CH4

Figure 2.17.

Protonation of 6 to give a cationic complex.

The 1H NMR spectrum (Figure

of this cationic complex.

18) of 7 indicates the highly electrophilic nature

The chemical shift of the alkylidene proton is at 6 14.79, and is

shifted significantly downfield from that of compound 6. A diethyl ether molecule is

tightly bound to the metal center as indicated by the ABX3 coupling pattern for the

methylene protons.

The ether molecule freely rotates on the NMR time scale interchanging

the two sets of diastereotopic methylene protons. At room temperature, four doublets and

two septets are observed for the isopropyl groups of the aryl imido ring indicating that the

C-N bond of the aryl-imido group does not rotate on the NMR time scale.

Using the trityl cation to abstract the methyl ligand,94 the reaction of 6 and trityl

tetrakis(pentafluorophenyl)borate in the presence of excess acetonitrile gave

[TpMo(CHC(CH3)2Ph)(NAr)(NCCH3)][B(C6F5)4] (8, Figure

19) and Ph3CCH3.

cationic product was obtained as a yellow-brown solid which was insoluble in benzene and



Ph3C+B(C6F5)4, NCCH3


-78 C to RT. 30 min.


Tp Mo=


+ Ph3CCH3

Figure 2.19.

Abstraction of the methyl ligand of 6 by the trityl cation to give a
cationic complex.

The 1H NMR spectrum of 8 at -40.0 OC shows two distinct alkylidene rotamers in

equilibrium which is ca 10 % in the minor isomer.

The chemical shifts of the alkylidene

protons of the major and minor isomers are shifted downfield relative to the neutral

complexes and occur at 8 14.47 and 15.28, respectively.

The presence of the

tetrakis(pentafluorophenyl)borate anion is confirmed by its broad doublets in the 13C NMR


The 1H and 13C NMR resonances of the Tp, aryl imido, and alkylidene ligands

are similar to those observed in the parent compounds, and the new resonance for the

acetonitrile ligand is prominent at 2.35 ppm in the proton spectrum.

The Lewis acid tris(pentafluorophenyl)boron, B(C6F5)3, has been shown to

abstract alkyl ligands forming cationic metal complexes.95 The addition of B(C6F5)3 to an

Et20 solution of 6 in the presence of acetonitrile gives the compound

[TpMo(CHC(CH3)2Ph)(NAr)(NCCH3)] [B(CH3)(C6F5)3] (9, Figure 2.20) in which the

Lewis acid abstracts the methyl ligand to give methyltris(pentafluorophenyl)borate as the

non-coordinating anion.

The cationic product was obtained as a yellow-brown solid which

was insoluble in hydrocarbons. Efforts to obtain crystals of 9 failed, but precipitation from

Et20 afforded an analytically pure solid.

Similar to 8, compound 9 is a mixture of two rotamers which is composed of ca 10

0.47 ppm.

The remaining NMR spectral data for 9 are similar to those data for 8.

This is

not surprising since the only difference between 8 and 9 is the substitution of a methyl

group for a perfluorophenyl group on the borate anion.

B(C6F5)3, NCCH3 or THF
Et20O, -78 OC to RT, 30 min.


Tp Mo-


9, 10

= NCCH3; 10: S =THF

Figure 2.20.

Abstraction of the methyl ligand of 6 by the Lewis acid B(C6F5)3
to give a cationic complex.

The THF-coordinated analogue of 9, [TpMo(CHC(CH3)2Ph)(NAr)(THF)]
[B(CH3)(C6F5)3] (10, Figure 2.20), was also prepared and isolated as was compound 9.
NMR spectral data of 10 are similar to those of 9, and the coordinated THF ligand is

observed to be rotating slowly on the NMR time scale showing two sets of diastereotopic
methylene proton triplets at 8 3.65 and 3.55.

We later found that a cationic complex,

[TpMo(CHC(CH3)2Ph)(NAr)(P(CH3)3)][OTf] (11, Figure 2.21), could be prepared
directly from compound 1 in quantitative yield by the addition of trimethylphosphine at

room temperature.

Compound 11 is isolated as a pale yellow powder, and it is soluble in

methylene chloride and insoluble in pentane and benzene.

The 1H NMR spectrum of 11 in CD2C12 at -20 OC is comparable to its parent

compound 1 and the other cationic complexes 7-10.

The distinctive feature of the 1H

NMR spectrum of 11 is the 3JPH coupling to the bound phosphine observed for the

alkylidene protons of the two rotamers.

The chemical shifts for the alkylidene proton for

31P NMR signals for the coordinated phosphine ligand are observed at -2.83 and

ppm for the major and minor rotamers, respectively.


The 19F signal for the ionic triflate (8

-78.8, vs. CFCl3) is slightly upfield from the signal for the covalently-bound triflate ligand

of 1 (8 -77.7, vs. CFCl3).98

Evidence for an ionic triflate species in 11 is best seen by a

comparison of its vibrational spectrum with that of 1.90 The S=O stretching frequency of
the free triflate anion of 11 (1268 cm-1) is split into two observable stretching modes in 1
(1332, 1202 cm-1). Also, a S=O stretch at 1048 cm-1 for 11 is shifted to 1011 cm-1 for

Tp -




CH2Cl2, RT

Tp -

\ H


Figure 2.21.

Synthesis of [TpMo(CHC(CH3)2Ph)(NAr)(P(CH3)3)][OTf] (11).

Obviously, the ability of the triflate ligand of 1 to be displaced by
trimethylphosphine suggests that the triflate ligand is not as tightly bound as originally
considered, and possibly an equilibrium exists between the bound triflate complex (1) and
a base-free, ionic triflate complex. Such an equilibrium would lie markedly towards the
coordinatively and electronically saturated complex 1, but in the presence of the strongly
coordinating phosphine ligand, the base-free cationic complex is consumed to give 11.
Attempts to observe adducts analogous to 11 in CD2C12 by 1H NMR by mixing 1 with 30
equivalents of ethylene or 50 equivalents of phenylacetylene were unsuccessful showing
only unreacted 1.

The Trans Influence of Ligands in Tp Molybdenum Complexes


ligand is its ability to weaken metal-ligand bonds trans to it by competition for metal a and

n bonding orbitals.

The most direct method to observing the trans influence is by

observing metal-ligand bond distances.100 Variation in the molybdenum-pyrazolyl bond

distances of compounds 1-5 is used to rank the trans influence of the ligands.

Thus, the

longer the molybdenum-pyrazolyl bond distances correspond to ligands possessing greater

trans influence.

Table 2.6 lists the molybdenum-pyrazolyl bond distances for pyrazolyl

rings trans to a number of ligands.

Table 2.6.

Trans influence of ligands in Tp molybdenum complexes.

This table suggests a trans influence series of alkylidyne


> imido


> amido

> alkoxo

> alkyl

-= -oxo

> triflate.

The trends in this series agree with

other series in that ligands of higher charge and it-donating abililities have greater trans


Rotational Isomerism of the Molvbdenum-Alkvlidene Bond

Rotational isomers are observed for many of the compounds reported in this


These isomers arise from restricted rotation about the molybdenum-alkylidene

Trans substituent Mo-pyralyl bond length, A Compound
Alkylidyne, CCPh(CH3)23- 2.386 5
Oxo, 02- 2.379(2) 2
Imido, NAr2- 2.327, 2.311 12
Alkylidene, CHCPh(CH3)22- 2.311, 2.295 1, 4
Amido, NHAr- 2.265 5
Alkoxide, CH-1.25, 2.232 4, 5
Alky, CH2CPh(CH3)21- 2.207(2) 2
p-Oxo, Mo-O-Mo, -1 2.200 3
Triflate, OSOCF3-1 2.167 1


considerations best describe the energetic differences between the related syn and anti


Interestingly, it has been reported that the different alkylidene

orientations of Mo(CHC(CH3)2Ph)(NAr)(OC(CH3)(CF3)2)2 can have profound effects on
the rates of olefin metathesis.3436 Due to the unique stabilizing ability of the Tp ligand

versus the more active alkylidene complexes reported, extensive kinetic studies have been

possible for a number of tungsten alkylidene complexes.35 Here we report preliminary

studies and evidence for rotational isomerism in several Tp molybdenum alkylidenes.

Irradiating TpMo(CHC(CH3)2Ph)(NAr)(OTf) (1) with a sun lamp at -50 C2 for one

hour generated a minor rotamer in a 1:4 ratio with the major rotamer.

The alkylidene

proton signal for this minor rotamer was observed at 8 14.82, which is 0.09 ppm

downfield from the major rotamer.

The 1JCH of the minor rotamer is 126 Hz compared to

120 Hz for the major rotamer. Large coupling constants (145-155 Hz) have been observed

for the anti rotamers of four-coordinate alkylidene complexes, and differences in 1JCH
values have been used to assign syn/anti orientations.3436 The electronically and

coordinatively saturated nature of 1 minimizes interaction of the alkylidene C-H bond with
the metal center that would give rise to large differences in coupling constants for the two

rotamers. The major rotamer was determined to be the syn rotamer by 1H nOe difference

spectroscopy. The rates of thermal conversion of the anti rotamer to the syn rotamer were

measured from +40.0 OC to +60.0 C

The activation parameters for this unimolecular

process were determined to be


= +21.5 0.3 kcal/mol and ASt

= -8.9 0.8 e.u.

The reduction of entropy in the transition state is consistent with restriction of ancillary

bond rotation during alkylidene rotation.

For TpMo(CHC(CH3)2Ph)(NAr)(CH3) (6), only one rotamer is observed at room
temperature. Irradiation of 6 in dg-toluene by a sun lamp at -50 C generated a new rotamer

rotamer is 3.7 x 104 sec-1 at -20.0 OC.

Similar rates were found for 1 at higher

temperatures of +40.0 and +50.0 OC.

The cationic molybdenum alkylidene complexes, compounds 8, 9, 10 and 11,

each have two rotamers present at equilibrium at ambient temperatures.

The irradiation of

TpMo(CHC(CH3)2Ph)(NAr)(NCCH3)][B(C6F5)4] (8) by sun lamp at -40 OC increased the
population of the minor rotamer to 32 % enabling 1H NMR peak assignments for both


The most notable difference between the spectra of the two rotamers is the 0.8

ppm downfield shift for the alkylidene proton resonance of the minor rotamer versus that of

the major rotamer. The rotamer distribution of 8 returned to equilibrium over time upon

warming the sample. Irradiation of compounds 7, 9, 10 and 11 was not performed.

Rotamers for TpMo(CHC(CH3)2Ph)(NAr)(OCH3) (4) were not observed at room

temperature. Photolysis at -50

o C for 90 minutes in dg-toluene using a sun lamp generated

a minor rotamer (8 13.44, 12.5%) which isomerizes thermally to the major rotamer very


The rate of thermal conversion of the minor rotamer to the major rotamer was

measured at -40.0 OC with a k

= 3.7 x 10-4 sec-1

From the crystal structure of 4 discussed

above, the major rotamer in solution is reasoned to be the syn rotamer.

This rate compares

with the anti to syn thermal isomerization rate of compound 1 at +40.0 OC where k

10-5 sec-1

=8.2 x

The extremely fast rates of the alkylidene rotation of 4 made repeated rate

measurements at low temperatures difficult, and thus activation parameters were not

determined for the compound.

The marked difference in the rates of alkylidene rotation for compounds 1, 4, and 6
is best understood by considering the competition for the empty dx orbitals of molybdenum

by the pt orbitals of the alkylidene, imido, and X, where X is a triflate, methyl, or

methoxide ligand, respectively. In the ground state, the alkylidene carbon py orbital


px orbital for the metal dxz orbital. Rehybridization of the nitrogen atom allows the filled
alkylidene pit orbital to fully donate to the empty dxz orbital, thus lowering the barrier to the

transition state.

Such a scheme has been proposed and supported by ab initio calculations

for four-coordinate tungsten and molybdenum imido alkylidenes.14'37

The Tp ligand system employed in compounds 1, 4, and 6 force the alkylidene,

imido and X ligands to be oriented mutually cis.

For ligands X with x-donor capabilities,

the bonding situation described above is further complicated.

Considering 4, when X is a

x-donating methoxide, a pnt orbital of oxygen can interact with either the dyz or the dxy

orbitals, but not the dxz orbital.


Figure 2.22.

Orbital diagram showing competitive donation in the ground state
to the metal dxy orbital by the alkylidene and methoxide n

Interaction of the oxygen pz orbital with the dyz orbital competes only with a metal-imido x-

bond, a bond that is unaffected during alkylidene rotation.

Shown in Figure


interaction of the oxygen px orbital with the dxy orbital competes directly with the metal-
alkylidene x-bond in the ground state. Such an interaction should raise the ground state

energy of the metal-alkylidene bond, resulting in a decrease in the barrier for alkylidene


The kinetic data for alkylidene rotation in 4 versus 1 and 6 support this


rates for molybdenum aryl imido neopentylidenes also show increased rotational rates as

alkoxide substituents become more electron donating, as in the series OR

= OC(CH3)3


> OC(CH3)(CF3)2

> OC(CF3)2(CF2CF2CF3).34,36

Such a relationship

is consistent with our finding for the relative rates of rotation for the series, X


= OCH3

> OS02CF3.

Polymerization Studies

The olefin metathesis activity of these Tp molybdenum complexes was of particular

interest considering their thermal stability. However, their electronically and coordinatively

saturated nature that imparts the thermal stability also negates their interaction with olefinic

substrates, and cocatalysts are required.

The metathesis activity of Tp molybdenum

alkylidene complexes is detailed below.

Compound 1 is inert toward the metathesis of neat cyclooctene or 1,9-decadiene,

and no polymerization was observed for 1 and 500 equivalents of norbornylene in o-

dichlorobenzene at 80 OC. However, in the presence of the Lewis acid AlCI3, 1

quantitatively catalyzed the ring-opening metathesis polymerization of cyclooctene.

1:6:500 mixture of 1, A1Cl3, and neat cyclooctene, polyoctenamer (Mn

= 1.30) was formed.

With a

= 57,000, Mw/Mn

This result compares with the tungsten analogue of 1 which

quantitatively polymerizes cyclooctene in the presence of a Lewis acid.77

A mixture of 1, AlC13, and 1,9-decadiene at 90 C under static vacuum only

dimerized a small fraction of the monomer to give a degree of polymerization of

the catalyst became inactive.

3 before

The stepwise fashion of ADMET condensation

polymerizations require quantitative conversion ( >

99 %) of end groups to achieve high


I __


Compound 6 in the presence of AlC13 polymerized norbornylene to give a

quantitative yield of poly(norbomrnylene) in toluene at room temperature.

With a 1:7:500

mixture of 6, ACl3, and neat norbornylene, poly(norbornylene),


,000 and Mw/Mn

= 1.8, was formed. A related compound, TpW(CHC(CH3)3)(NAr)(CH2C(CH3)3), also

catalyzes the ROMP of cyclooctene under similar conditions.78

As with 1, compound 6 and AlC13 did not successfully polymerize 1,9-decadiene


. With a 1

:500 mixture of 6, A1C13, and neat 1,9-decadiene, only 7

of the

monomer was converted to dimer during 8 hours of stirring under static vacuum.

Attempts to trap and isolate any unsaturated metal complex from the catalytic

mixture of 1 or 6 and A1C13 with P(CH3)3 were unsuccessful.

Observation of 1 or 6 with

A1C13 in C6D6 by 1H NMR initially gave a complex mixture, and after several days free

pyrazole was observed.

Lewis acids are proposed to play a role in generating a four-coordinate cationic

active catalyst in W(CHR)(OCH2R)2X2/Al2X6 metathesis systems.102-104 Also,

W(O)(CHC(CH3)3)(PEt3)2Cl2/AlCl3 metathesizes olefins while the five-coordinate, neutral
complex W(O)(CHC(CH3)3)(PEt3)C12 is an active catalyst in the absence of a Lewis

acid.62 Contrary to a mechanism involving the Lewis acid simply removing the triflate or

methyl ligand of 1 or 6 to generate a five-coordinate, cationic active catalyst, no metathesis

activity is seen for the solvent-bound, cationic complexes, 7 and 8.

Contributing to the

inactivity of the cationic complexes, the high electrophilicity of the metal center does not

allow dissociation of the bound solvent molecule to open a coordination site on these

electronically and coordinatively saturated compounds. Heating a CD2C12 solution of 8

pressurized with ethylene at 60 OC for 24 hours showed no reaction or loss of acetonitrile.

Though five-coordinate metathesis catalysts are known,62,102103 other reported


complex, the Lewis acid also reacts with the Tp ligand system and removes a pyrazole ring

from the coordination sphere of molybdenum.

This would result in a four-coordinate, 14-

electron, cationic complex in which the Tp ligand is bound to the metal center by only two

pyrazolyl rings.

Indirect evidence for this proposal is the production of free pyrazole from 1 and 6
in the presence of AC13 mentioned above and the observation that

[TpMo(CHC(CH3)2Ph)(NAr)(THF)] [B(CH3)(C6F5)3] (10) and two equivalents of AlC13
in toluene quantitatively polymerizes 500 equivalents of norbornylene. Further indirect

evidence includes the observation that the closely related complex

[TpW(NPh)(CHCMe3)(i-Pr20)][BAr'4] does not initiate the polymerization of
norbornylene even though the diisopropyl ether ligand readily dissociates and is displaced

by other Lewis bases such as diethyl ether and acetonitrile.78


Undoubtedly, the chelating hydridotris(pyrazolyl)borate ligand plays an important
role in the synthesis of air, moisture, and thermally stable molybdenum alkylidene

complexes. Since the Tp ligand saturates the complexes both electronically and

coordinatively, the metal centers are extremely robust, however they are unreactive towards

nonpolar molecules, namely olefms.

Though the complexes prepared and characterized in

this chapter are not metathesis catalysts in the absence of Lewis acids, insight has been

gained as to strategies for developing chelated olefin metathesis catalysts.


chelating ligands must be chosen such that the complexes are electronically and
coordinatively unsaturated under the conditions of catalysis.


Recently our research group has realized the utility of using phenylenediamido

ligands to stabilized tungsten complexes.

The most extensive work has been with N, N'-

bis(trimethylsilyl)-o-phenylenediamide (TMS2pda) which serves as a bidentate, dianionic


VanderLende has synthesized a large variety of alkyl, alkylidene, and hydride

complexes starting with (TMS2pda)W(NPh)(C1)2 (12).79,84 The tungsten alkylidene

complex, (TMS2pda)W(NPh)(CHC(CH3)3)(P(CH3)3), is thermally stable at 110 C and

successfully polymerizes norbornylene via a ring-opening metathesis polymerization

mechanism.79 These complexes do lack the air and moisture stabilities observed in the Tp

complexes of Chapter Two, and in the presence of trace protic species, N, N'-

bis(trimethylsilyl)-o-phenylenediamine (TMS2pdaH2) is observed.

This is not unexpected

considering the polar nature of tungsten-nitrogen single bonds and the sensitivity of the

silicon-nitrogen bond to hydrolysis.

Vanderlende reported the synthesis of both the alkylidene complex and a

metallacycle species from the bisalkyl complex (TMS2pda)W(NPh)(CH2C(CH3)3)2 (13,

Figure 3.1).84

The thermolysis of 13 at 70 OC in the presence of

equivalents of

trinmethylphosphine gave the ca-abstraction product

(TMS2pda)W(NPh)(CHC(CH3)3)(P(CH3)3) (14) in a 66% yield.
in the absence of trimethylphosphine gave a metallacycle complex.
1 1 1 1 1 1 a r.

The thermolysis of 13

This complex was not
4l 4 *


videe infra) revealed the structure of the metallacycle to be a tungstasilacycle (16) which
might result from y-CH-activation of a trimethylsilyl group of the ligand by either complex

13 or 14.

\ / nPh


\ i




\ /

Figure 3.1. Products of the


\/ NPh
--s W



Thermolysis of (TMS2pda)W(NPh)(CH2C(CH3)3)2.

In order to study the formation of 14 and 16, a number of kinetic and mechanistic

experiments have been performed.

These experiments include kinetic measurements

followed by 1H NMR spectroscopy and deuterium labeling studies analyzed by mass



This research is ongoing, and when appropriate, experiments will be

suggested that should illuminate facets of our proposed mechanism that are still ambiguous.

Synthesis. Characterization and Structure of the Metallacycle Complex

VanderLende reported the formation of a metallacycle complex from the thermolysis

of 13 in toluene at 70 C for two days.84 The assignment of structure 15 to this data was

based upon two geminal proton doublets at

49 and

-1.15 ppm in the 1H NMR.

However, these two doublets were not coupled to one another, as evidenced by selectively

irradiating one doublet and observing no perturbation in the resonance of the other doublet.

A COSY experiment unequivocally identified two additional doublets which were obscured

by the neopentane by-product and unidentified decomposition product resonances (Figure


The doublet at 8 2.49 is coupled to a doublet at 8 0.76; the doublet at 8 -1.15 is

coupled to a doublet at 8 0.86.

This revelation and other observations to be described later

suggested that complex 16 is formed upon thermolysis.

The 1H NMR spectrum of 16 (Figure 3.3) consists of the aforementioned sets of

geminal doublets.

The doublets at 2.49 and 0.76 ppm are attributed to the a-methylene

protons of the neopentyl ligand, and the doublets at 0.86 and

a-methylene carbon of the silacycle.

-1.15 ppm are assigned to the

Two singlets at 0.54 and 0.20 ppm integrate to three

protons each and are assigned to the [3 and P' methyl groups on the ring silicon.


singlets at 1.02 and 0.38 ppm are assigned to the t-butyl group of the neopentyl ligand and

the free trimethylsilyl group, respectively.

The 13C NMR spectrum shows 183W coupling

to the neopentyl a-carbon (8 92.9,

the silacycle (8 37.1,


1Jcw = 44 Hz).

= 103 Hz) and a smaller coupling the a-carbon of

No temperature-dependent dynamic behavior is

observed in the proton spectrum over the temperature range of 20 OC to 110 OC. Thus, the


i i -


.. ., ...

. A -

- S.-
-U* ,



yellow-brown single crystals suitable for x-ray diffraction studies, and the solid-state

structure of 16 was obtained (Figure 3.4).

given in Table 3.1.

Selected bond lengths and angles of 16 are

The geometry about the tungsten center of 16 is very similar to that of

14, reported by VanderLende.79 The geometry is best described as a square pyramidal

structure with the neopentyl ligand in the axial position.

metallated TMS2pda ligand lie in the basal plane. The t

The imido nitrogen and the

ungsten-carbon bond distance of the

neopentyl group is 2.12(2) A and is normal for tungsten-carbon single bonds.

C3 angle of the neopentyl ligand is 128.8(12).

is 1.695(13) A, and the W-N-phenyl ring bond angle is

The W-Ca-

The tungsten-imido nitrogen bond distance


phenylenediamido nitrogen-tungsten bonds are similar to those of other TMS2pda


The tungsten-a-carbon bond distance (2.18(2) A) for the silacycle is normal

for tungsten alkyl bonds.

Table 3.1:

Selected Bond Lengths (A) and Angles (0) for compound 16.









72 (f1i\





(V O

a --


oa o
m n

a e
mT ^- 0

N O'


cu .
cv) 0
C)) -

-> \\Y \lf or\/CJ ^V^ .

C) I0 4z CB

z a



The tungstasilacycle of 16 is related to other compounds reported in the literature.

Four-membered metallasilacycles have been formed by y-abstraction processes.106


has reported y-abstraction in bis(cyclopentyldienyl)bis(trimethylsilylmethyl)thorium

complexes to give a M-C-Si-C ring.33,107,108


se, similar chemistry is observed by

Diversi for the rhodium and thorium analogues.32109 Perhaps more analogous to the

formation of 16, Andersen reported the y-C-H activation of bis(trimethylsilyl)amido

complexes zirconium and thorium to give M-N-Si-C rings.110o111

C-H activation by singly bonded nitrogen or carbon atoms. Activ
the more basic alkylidene ligand have also been reported. Examp

These processes involve

ration of C-H bonds by

les are the E-activation of

the phenoxide ligand in the tantalum complexes of Rothwell (Figure 1.15)112,113 and the

orthometallation of a phenyl ring by a tungsten alkylidene reported by Schrock.41


scenarios seem possible for the formation of 16 from 13 or an unobserved alkylidene

species, and experiments described later will probe the origin of 16.

Kinetics of the Formation of the Metallacvcle Complex

To begin our investigation of the formation of the silacycle complex, 16, the

kinetics of the thermal decomposition of (TMS2pda)W(NPh)(CH2C(CH3)3)2 (13) was

followed by 1H NMR spectroscopy.

The rate of the decomposition of 13 to give 16 was

measured by observing the decreasing peak intensities of the t-butyl resonance of the

neopentyl ligands (0.99 ppm) and the trimethylsilyl resonance (0.48 ppm) of complex 13

(Figure 3.5a)

The decomposition of 13 was found to obey first-order kinetics for up to

five half-lives in dg-toluene over a temperature range of 80 OC to 110 OC.

The natural

logarithm of the two peak intensities were plotted versus time for at least 3 half-lives

(Figure 3.5b).

The rates, kNp and kTMS, were taken from the slope, and kNp and kTMS

1 I_ I I _

0C, and an average kinetic rate was determined for each temperature.

The kinetic data is

catalogued in Appendix B.

An Eyring plot was constructed by plotting ln(k/T) versus 1/T (Figure 3.6). The
enthalpy of activation (AHt) was calculated from the slope with the slope equal to -AHt/R,

and the entropy of activation (ASt) was calculated from the intercept, being equal to ASt/R

+ 23.76.

For the thermal decomposition of 13 to give 16, the activation parameters were

determined to be +25.6 0.3 kcalmol-1 for AH- and

-5.4 + 0.8 calmol-1K-1 for ASt

Since ASt is determined by extrapolation of a line over a large range, small errors can lead

to large variations in entropy terms, and conclusions from these values are limited.

However, at this point it will be pointed out that the observed negative entropy agrees with

other values determined for cyclometalation reactions, and suggests that in the transition

state the rotations of single bonds are restricted and the symmetries of the molecule


One thermolysis experiment was performed in the presence of 7.5 equivalents of

neopentane at 90.0 OC and the rate was observed to be 1.1

x 10-4 sec-1

This is

significantly slower than the rate without added neopentane (1.80 x 104 sec-1).


these conditions must be repeated to insure that such a difference is real.

These kinetic studies have given information that should help in the synthesis of

16 on a preparative scale.

The 1H NMR spectra of the kinetic runs indicate that there is a

marked increase in decomposition products at 110 C after five half-lives compared to the

reaction at 80 OC after three half-lives. An increase in decomposition products occurs as 16

remains in solution at elevated temperatures.

Therefore, in order to isolate the metallacycle

on a preparative scale, it is suggested that 13 should be thermolysed in toluene at 80 "C for

ten hours or 5 half lives.

Immediate evaporation of the toluene, extraction with pentane,

I __ I





O Np






t sec



S- 5.5

x 4.5

0. 4



4000 6000



Ila san


o o




o C



cw w


Nr 6
+ 8
n .
CU u
T? =



Kinetics of the Formation of the Alkvlidene Complex

The thermolysis of (TMS2pda)W(NPh)(CH2C(CH3)3)2 (13) in the presence of

excess PMe3 gives the tungsten alkylidene complex

(TMS2pda)W(NPh)(CHC(CH3)3)(P(CH3)3) (14) with coordinated phosphine.

VanderLende reported that the isolation of a Lewis base-alkylidene adduct was only

possible with trimethylphosphine.84

With PEt3, a mixture of the metallacycle and the

alkylidene-phosphine adduct were observed in the 1H NMR spectrum. Attempts to prepare

analogues of 14 with other Lewis bases (PMePh2, PCy3, PPh3 and quinuclidine) were not

successful, and only decomposition products or the silacycle 16 was observed.

The kinetics of the formation of 14 were followed over the same temperature range

used for the formation of the silacycle complex.

The alkylidene complex did demonstrate

fluxional behavior over this temperature range. At the temperatures studied, the

trimethylsilyl groups interchange faster than the NMR time scale.

resonance is observed for the time-averaged trimethylsilyl groups.

Thus, a coalesced

Variable temperature

NMR experiments showed the coalescence temperature of the TMS groups to be 39 OC.

The rate of exchange at this temperature is calculated to be 44.6 sec-1

and AG* is

determined to be 15.9 kcalmol-1.114


- PMe3


I 3





:. ::::..... .. -



Figure 3.7.

Mechanism for the equilibration of the TMS groups of 14.

The mechanism of the equilibration of the TMS groups is proposed to involve the


equilibrates the TMS groups as shown in Figure 3.7

Once the equilibrium in Figure 3.7

becomes rapid on the NMR time scale, an average resonance is observed for the TMS


No phosphine-free adduct is ever observed in the 1H NMR spectrum.

The kinetics of the formation of 14 were followed in the same manner as the

thermolysis of 13. The rate data for the thermolysis of 13 in the presence of PMe3 is

catalogued in Appendix B. Activation parameters were determined by plotting In(k/T)
versus 1/T, and AHt and ASt were found to be +28.1 0.3 kcalmol-1 and +0.8 0.8


, respectively (Figure 3.8).

The results indicate that upon addition of PMe3 the

rate of reaction is depressed slightly. A plot of the observed rate constants versus the

number of PMe3 equivalents added is included in Appendix B.

Preliminary results indicate

that this rate depression is greater at lower temperatures (80 OC).

At 80

C, the depression

of the rate upon adding 20 equivalents PMe3 is about 30 per cent (0.644 to 0.452 x 104


It is noted that such a small change in rate is not normally attributed to major

changes in the mechanism.

The source of this rate depression may be associated with an

increase in solvent polarity upon the addition of PMe3. For example, the addition of 20

equivalents (0.097 mL) of PMe3 to

solution of the highly polar phosphine.

mg of 13 in 0.6 mL toluene results in a 15 % v/v

The possibility of this change in rate being due to

changing solvent polarity needs to be investigated, and observing the reaction in the

presence of deuterated CD2Cl2 and CD3CN is suggested.

A Closer Look at the Role of PhosDhine

Trimethylphosphine has not been observed to interact strongly with any of the

tungsten bisalkyl complexes with TMS2pda as the chelating ligand.

This seems surprising

since (TMS2pda)W(NPh)(Cl)2 (12) and a related compound (TMS2dav)W(NPh)(Cl)2
, 4 W fl 1rrt. 1" 1 4 a n nL1 -rn n -r *. I a v n w t -.. . .






aE o.

so --



O 0

o 0

/ >t


attributed to tungsten bisalkyl phosphine complexes. However, the time scale of NMR

experiments range from 10-1 to 10-5 sec"1

and therefore the molecules involved in chemical

events that occur faster than 10-5 sec-1 are observed as a time-averaged species.

Ultraviolet-visible (UV-VIS) spectroscopy has a 10-14 sec-1 time scale limit, and it was
proposed that a short-lived bisalkyl phosphine adduct might be detectable by its UV-VIS


UV-VIS spectroscopy was useful in comparing complexes 12 and 17 and their

phosphine adducts.

The results of these studies and the observation of 13 in the presence

of PMe3 are described below.

The complexes (TMS2pda)W(NPh)(C1)2 (12) and (TMS2dav)W(NPh)(C1)2 (17)
are colored dark yellow to brown in toluene, and upon addition of a coordinating PMe3,
there is a marked color change to give intensely colored solutions of purple

(TMS2pda)W(NPh)(C1)2(PMe3) (18) and blue (TMS2dav)W(NPh)(C1)2(PMe3) (19).
These color changes were quantified by observing the visible spectrum of the compounds,
and the results are given in the Table 3.2 below.

Table 3.2.

UV-VIS Absorbance Data.

Compound Xmax, nm E, M-cm-1 fold angle, o
(TMS2pda)W(NPh)(C1)2, 12 447 4740 58
(TMS2pda)W(NPh)(C1)2(PMe3), 18 531 3012 20
(TMS2dav)W(NPh)(C1)2, 17 486 6759
(TMS2dav)W(NPh)(C1)2(PMe3), 19 575 4337
(TMS2pda)W(NPh)(CH2C(CH3)3)2, 13 393, est. 525 1344, 208
13 + PMe3 367, est. 525 1056, 112

The source of this spectral shift is related to the change in the fold angle of the

phenyl ring of the TMS2pda ligand (Figure 3.9).115-117

The fold angle is defined as the

180" complement of the dihedral angle between the plane of the phenyl ring and the plane


of the n electrons of the phenyl ring with the dz2 orbital of the five-coordinate metal center.

This folding demonstrates the high electrophilicity of the metal center In the presence of

PMe3, this interaction is interrupted, and a spectral shift is observed.

One problem with determining if there is a spectral shift from adding PMe3 to

(TMS2pda)W(NPh)(CH2C(CH3)3)2 (13) is that pure samples of the bisalkyl complexes
are difficult to obtain, and the color of the powders obtained are light yellow-brown at best.

Alternatively, washing of 13 with a small portion of acetonitrile affords a good yield of an

olive green powder and the 1H NMR spectrum of the material shows the powder to be

cleaner than the brown powders obtained by concentration of pentane solutions. A toluene

solution of the olive green powder of 13 is yellow in color and results in a broad maximum

in the visible spectrum centered around 525 nm and a UV absorbance is seen at 393 nm.

Addition of 1500 equivalents of PMe3 to 13 results in a blue-shift of the UV absorbance,

but a shift of the visible absorbance is not discernible in part due to the broad maximum

observed. Addition of 20 equivalents of PMe3 does not result in a blue shift, but the

intensity of all absorbances is reduced.

The spectral shift observed for the UV absorbance

may be attributable to the change in solvent polarity upon addition of the polar PMe3, but

source of the bleaching is not evident at this time.


N Me....- *
.N rtCI




Figure 3.9. Diagram depicting phenyl ring interaction with the metal upon
increased fold an le.

this temperature range.

The chemical shift of the free PMe3 (8 -61.8 at -60 OC) did not

change upon addition of the metal complex. The only effect on the spectrum observed was

significant broadening of the resonance at -61.8 ppm. The width at half the maximum

intensity (wi/) for free PMe3 is 17.6 Hz. The wi/2 values for 1.1 and 2.5 equivalents of

PMe3 are 41.1 Hz and 32.7 Hz, respectively. The trend observed for line broadening upon

addition of PMe3 suggests some interaction of the phosphine with the metal center, but the

absence of a difference in chemical shift means that any equilibrium between free and

coordinated phosphine lies decidedly towards free phosphine and unbound 13.

With the absence of a marked interaction between

(TMS2pda)W(NPh)(CH2C(CH3)3)2 (13)and PMe3 and only a slight decrease in reaction
rates observed with added PMe3 for the decomposition of 13, trimethylphosphine is not

deemed to be involved in the rate determining step of the formation of

(TMS2pda)W(NPh)(CHC(CH3)3)(P(CH3)3) (14).

Interconversion of the Metallacycle and the Alkylidene Complexes

An interesting result is the observation that the alkylidene complex,

(TMS2pda)W(NPh)(CHC(CH3)3)(P(CH3)3) (14), and metallacycle complex,

[(Me3SiN)C6H4(N'SiMe2CH2)]W(NPh)(CH2C(CH3)3) (16) can interconverted.
Removal of PMe3 from 14 gives 16 and unidentified decomposition products (Figure

3.10). This reaction has been observed in 1H NMR experiments with Cu(I)C1, MeOTf,

and Mel. The cuprous ion acts by irreversibly binding PMe3 to give an insoluble

CuCl(PMe3)x complex.

Cuprous salts are commonly used as "phosphine sponges" to

remove phosphines from metal complexes.48 The electrophilic methyl groups of MeOTf

and Mel act by forming the phosphonium salts, PMe4+OTf- or PMe4+I-.

The mechanism


The y-C-H activation by an alkylidene ligand proposed is precedented in the literature by

the examples mentioned earlier in this chapter.

These methods have not proven effective in

the preparation of 16 on a large scale.



..-NSi--N /
/ ",.---


Figure 3.10.

Proposed mechanism of PMe3 abstraction of 14 to give 16.

The silacycle complex 16 can also be converted to the alkylidene complex 14.
Addition of excess PMe3 to a dark yellow dg-toluene solution of 16 quantitatively gives an

orange solution of 14 over two hours at ambient temperatures (Figure 3.11).


the fact that removal of the PMe3 from 14 gives 16, the mechanism of the reverse reaction
is hypothesized to consist of an equilibrium between 16 and the four-coordinate alkylidene

complex 20.

This equilibrium decidedly favors 16 in solution, but the PMe3 present in

solution effectively traps 20 as the phosphine-bound alkylidene complex 14.

As a result,

the equilibrium of 16 and 20 is driven towards the formation of the alkylidene, and the
alkylidene complex 14 is formed quantitatively since the binding of PMe3 to 20 is
essentially irreversible at ambient temperatures.

PhN -s'
-Si--N SN

N..'I'"-W= NPt
N P-
Si- Pc



Figure 3.11.

Proposed mechanism of PMe3 addition to 16 to give 14.

Deuteration studies described below suggest that 16 is in equilibrium with 20,

however the existence of this equilibrium has not been rigorously demonstrated.

interconversions of 14 and 16 are amenable to kinetic study, and the following

experiments are planned.

The homogeneous reaction of 14 and MeOTf or Mel will be

studied by varying the concentration of the electrophiles over a temperature range.


study should define the molecularity and rate law for the abstraction of phosphine and

rearrangement to give the silacycle. Likewise, the kinetics of reacting 16 with varying

concentrations of PMe3 over a temperature range will also demonstrate the dependence of

the reaction on PMe3. Rigorous study of this process is contingent upon the preparation of

pure 16.

Another experiment to demonstrate a dynamic equilibrium between 16 and 20 is

spin saturation transfer studies with 1H NMR spectroscopy.

methyls groups were on the order of 1.2 seconds. Spin satu

T1 measurements of the silyl

ration transfer experiments at

20 and 110 OC did not demonstrate a dynamic equilibrium between 16 and 20.

Deuterium-Labeling Studies


experiments were performed by employing (TMS2pda)W(NPh)(CD2C(CH3)3)2 (13-d4).

Complex 13-d4 was prepared by alkylating (TMS2pda)W(NPh)(C1)2 (12) with two
equivalents of a,a'-dideuterioneopentyl magnesium chloride, Mg(CD2C(CH3)3)C1.18'119

The results were followed by mass spectrometry of the neopentane by-products and 2H

NMR spectroscopy.

Chemical Ionization Mass Spectrometry Study

Analysis of deuterated neopentane by-products (NpH-dn) by mass spectrometry

has been reported as a method to determine whether a-abstraction or y-abstraction

pathways are operative in the formation of alkylidene and metallacycle complexes. 1
seminal account of this method was by Schrock in his investigation of the role of a-

abstraction in forming Ta(CHC(CH3)3)(CH2C(CH3)3)3.38

Marks, using the same

method, has demonstrated that y-abstraction is the major pathway for the decomposition of

Cp2Th(CH2C(CH3)3)2 to give the metallacycle Cp2Th(CH2C(CH3)2CH2).108
These labeling studies are especially important to this research because in the

previous section the interconversion of the alkylidene and silacycle complexes was noted.
Therefore the origin of either species might arise from an initial a- or y-abstraction pathway

followed by an interconversion mechanism. Evidence for the two pathways is presented


For an initial a-abstraction pathway involving 13-4, the neopentane by-product

produced would have the formula C(CH3)3(CD3), or NpH-d3 (Figure 3.12).

only pathway that would result in NpH-d3

This is the

from the decomposition of 13-d4.

For an initial y-abstraction pathway involving 13-4, the neopentane by-product

produced would have the formula C(CH3)3(CHD2), or NpH-d2 (Figure 3.13).


I 1


of (TMS2pda)W(NPh)(CH2C(CH3)2CH2) (15), it will be assumed that any NpH-d2

formed results from the activation of a TMS group to give 16.



\- Si

UIII"'-"'W C




Figure 3.12.


Formation of NpH-d3 via the a-abstraction pathway.

The thermolysis of 13-d4 was performed at 80 oC and 100 OC with and without

added PMe3.

Also, deuterated and non-deuterated solvents were used in order to determine

whether C-H(D) activation of the solvent by 13-d4 is significant.

After thermolysis, the

volatiles of the reaction (NpH-d,, PMe3, solvent) were distilled from the reaction mixture
in order to eliminate possible secondary reactions catalyzed by any metal species present.

\ -

A./ ^

\ /

I .-NPh




Figure 3.13.


Formation of NpH-d2 via the y-abstraction pathway.

The volatiles were separated and analyzed by GC-MS techniques.

ioniztion (CI) by methane was used to generate charged analytes.

generates tert-butyl cations as the major ion observed.


For neopentane, CI

Thus, NpH-d3 generates ions with

m/z values of 60 and 57 in a 3:1 ratio, and NpH-d2 generates ions with m/z values of 59



neopentane generated. For this study, the t-butyl cation envelop was used to determine

whether a- or y-abstraction is operative.

Other studies have utilized electron impact

techniques (EI) to generate t-butyl cations. At the outset of this study, CI was favored over

El because CI is a gentler ionization technique. Despite the fact that CI is a "softer"

ionization technique, protonated neopentane readily eliminates methane to give the observed

t-butyl cations. Fortunately, there is no scrambling of protons and deuterons in the t-butyl

cations generated.

Table 3.3 summarizes the results of the CIMS data.

The % NpH-dn values given

are calculated from the relative m/z peak height of NpH-dn corrected for 13C abundance

divided by the sum of the relative m/z peak heights of NpH-d2 and NpH-d3 given in

Appendix B.

In Table B, the peak heights at m/z values of 58 and 61 result from the 13(


Table 3.3.

Relative Amounts of NpH-dn Produced in Thermolysis Reactions.

Entry temp, solvent PMe3, equiv. % NpH-d2 % NpH-d3
1 80 C, C6D6 0 33 67
2 80 oC, C7H8 0 36 64
3 80 C, C7H8 5 16 (2) 84 (2)
4 100 C, C7D8 0 40 60
5 100 C, C7D8 20 8 (1) 92 (1)
6 110 C, C7H8 0 45 55
7 25 C, C6D6 H2 93 7

Table 3.3 gives the CIMS data to date, and reactions will be repeated as necessary

to increase the confidence in these results.

Samples of NpH-d2 and NpH-d3 prepared by

the reaction of Mg(CD2C(CH3)3)Cl with CH30OH or CH30OD, respectively.

These samples

4 4t Sfl *n a a .

Comparing entries 2 to 3 and 4 to 5 does demonstrate the marked increase in the

relative ratio of NpH-d3 to NpH-d2 upon the addition of PMe3 to the thermolysis reaction.

The degree of the increase appears to increase upon increasing the temperature and PMe3

concentration, but this difference has not been shown to be significant due to the lack of

data. Also, performing the reaction in a non-deuterated solvent shows a trend of increased



This trend suggests that C-H activation of the solvent by 13-d4 may

occur to a minor extent, but this trend too must be confirmed with further analysis. Entry 7

is for the hydrogenolysis of 13-d4.

This reaction demonstrates that the tungsten alkyl

bond can participate in o-bond metathesis, and the more nucleophilic alkylidene ligand may

not be necessary for C-H bond activation.

With the limited data gathered, the a-abstraction

pathway is favored over the y-abstraction pathway, and addition of PMe3 increases the

relative rate of a-abstraction versus y-abstraction.

Deuterium NMR Spectroscopy

The thermolysis reactions of 13-d4 were also analyzed by 2H NMR spectroscopy.

The most striking result from the spectra is observed scrambling of deuterons into the TMS

groups of the alkylidene complex 14 and the metallacycle complex 16.

In fact, it was this

observation that eventually led to a reassignment of the metallacycle structure as 16 and not


The a,a'-deuterons of complex 13-d4 give rise to a single broad signal at 8

(Figure 3.14a).

After thermolysis of a do-toluene solution of 13-d4 at 110 OC for 90

minutes, several new signals are present (Figure 3.14b). Most prominent is a large singlet

at 8 0.95 assigned to the NpH-dn formed.

at 8 0.70, 0.53 and 0.36.

Three broad, overlapping signals are observed

These signals are approximately in a 1:3:1 ratio and assigned as

- a a.

3 - w * ** w


scale x2

U 5 U S

scale x2

. U








doublett', which overlaps with the NpH resonance, is also observed at 8 1.01.

assignment is not clear at present, but the t-butyl groups of the neopentyl ligands of 13 and
16 give 1H NMR signals at 8 1.00 and 1.02, respectively.

The 2H NMR spectrum of the thermolysis reaction of 13-d4 with excess PMe3

shows similar features (Figure 3.14c).

A large singlet for NpH-dn is observed, and a

single broad resonance at 8 0.41 is assigned to the TMS groups of the alkylidene complex,

14. No deuterium signal is observed that would correspond to an a-deuteron on the

alkylidene ligand.

The same unassigned doublet at 8 1.01 is observed as in the thermolysis

reaction without PMe3, but the doublet is of lesser intensity for the PMe3 case. The
alkylidene t-butyl group gives a 1H NMR signal at 6 0.98.

The observation of complete scrambling of deuterons to all silicon methyl positions

suggests that an equilibrium between the metallacycle complex 16 and the unbound

alkylidene complex 20 exists. In the case of added PMe3, this equilibrium exchange may

also be faster than PMe3 coordination to 20 to give 14. However, the additional

experiments described in the section on the interconversion of 14 and 16 are needed to

confirm these statements.

The doublet observed at 8 1.01 in both thermolysis reactions

may suggest a mechanism to exchange the a-methylene positions of the neopentyl ligand

with the t-butyl group of that same ligand.

Kinetic Isotope Effects in the Formation of the Metallaccle and the Alkvlidene Comlexes

The complex 13-d4 also allows investigation of the mechanism by determining if a

kinetic isotope effect (KIE) is observed. Primary kinetic isotope effects have been reported

in the literature for a-abstraction and y-abstraction pathways.

The relative rates of reaction

of the protonated versus deuterated analogues (kH/kD) is typically on the order of 6 when


From the kinetic data tabulated in Appendix B, the value for kHWkDwas determined

to be

+ 0.9 for the thermolysis of 13 in the presence of 20 equivalents of PMe3.

value of kH/kD for the thermolysis of 13 in the absence of PMe3 is 3.9 0.4.

Based on results from the CIMS studies, the KIEs measured reflect a mixture of

two competing pathways, a- and y- abstraction.

For 13-d4, a-abstraction involves the

breaking of a C-D bond, but 'y-abstraction involves the breaking of a C-H bond.

Therefore, only the a-abstraction pathway will incur a primary KIE. No primary KIE is

expected for the y-abstraction pathway, and a secondary KIE should not be appreciable

since no changes in hydridization occur.


The KIEs measured are


The NpH-dn product ratio determined by CIMS relates the relative rates of kDa and ky as



Solving for kDa in equation (2) and substitution into equation (1) allows for the

kHaoky ratio to be determined. For thermolysis in the presence of PMe3, kHa/ky is 6414.

In the absence of PMe3, kHo/ky is 9

2 for a preliminary KIE of 3.9 0.4.

Rearrangement and substitution of (1) with (2) give the equation

kna ( NpH-d3 k\y NpH-d2
tk"J NpHt kr NpHuJiJ

and solving for kHj/kDa, the actual KIE for a-abstraction can be extracted from the data.

In the presence of PMe3, kHa/kDa is determined to be 5.61.0.

In the absence of PMe3,

kHa/kDa is determined to be 5.8 0.5.


Proposed Mechanistic Scheme

Considering the experiments described above, a mechanistic scheme can be

proposed to explain our observations.

Further investigation will be necessary to increase

support for the proposed mechanisms and to expand their details.

The key points of the following scheme are as follows.

Competing a- and y-

abstraction pathways from the bisalkyl complex 13 compose the rate determining steps.

Trimethylphosphine is not implicated in the rate determining steps, but the ligand does

serve to trap the alkylidene complex 20. Also, a process capable of interconverting the

alkylidene and metallasilacycle complexes is incorporated.


The scheme is given in Figure

, and the supporting observations are summarized.



W _



- NpH

+ NpH


1 ^NPh



- PMe3

+ PMe3


\ /






thermolysis reactions involving 13-d4.

The kj/kD of 5.2 0.9 determined for thermolysis

in the presence of phosphine is in agreement with other values reported in the literature

where a-abstraction mechanisms are operative.

The CIMS analyses of the volatiles of the

reactions showed NpH-d3 to be the major isotopomer of neopentane.

besides a-abstraction would result in the formation of NpH-d3. The

No other pathway

reversible a-

abstraction process proposed provides a pathway by which the a-deuterons of 13-d4.can

interchange with the y-carbons of the neopentyl ligand.

The possibility of such an

interchange is suggested by 21 NMR spectra which show resonances at 5 1.01.

Evidence for the competitive y-abstraction includes the production of NpH-d2

during the thermolysis reactions, and it is noted that in the absence of PMe3, the relative

ratio NpH-d3 and -d2 decreases markedly. As well, initial KIE results for the thermolysis

reaction in the absence of PMe3 give a kH/kD of 3.9 + 0.4, and this lower observed value

agrees with the change in NpH-dn ratio.

The equilibrium postulated that interconverts the metallacycle 16 and the unbound

alkylidene complex 20 is necessary for the scheme due to several observations.


PMe3 is present or not, both a- and y-abstractions occur, and interconversion of 16 and

20 must occur in order to achieve the yields of 14 or 16 obtained.

Thus, in the absence of

PMe3, the alkylidene complex 20 formed by a-abstraction must convert to the metallacycle

complex 16. An equilibrium mechanism is also the most straight forward way by which
the a-deuterons of 13-d4 can interchange with the silicon methyl substituents of the pda


The ability to interconvert 14 and 16 by either the abstraction of the bound PMe3

of 14 or the addition of PMe3 to 16 indicates the relationship of the metallacycle and

alkylidene complexes.

This interconversion of 14 and 16 is amenable to further kinetic

study, and the results will add detail to this aspect of the scheme.

relative ratio of deuterated neopentane produced and the observed rates of reaction.

It has

been suggested that these slight changes for the reaction may be due to subtle solvent

polarity effects, and the fact that these changes are slight does not support major changes in

the mechanistic scheme upon the addition of phosphine.


In choosing an ancillary chelating ligand to stabilized alkylidene, a requirement that

is often taken for granted is that the ligand does not react with the alkylidene ligand. The

situation with the TMS2pda tungsten complexes in this chapter was surprising, but not

totally unexpected considering literature precedence. However, the chemistry presented

has lead to a unique opportunity to begin study of the C-H activation potential of alkylidene


The TMS2pda ligand was initially chosen because of its facile synthesis compared

to other N,N'-disubstitued pda ligands.

With the identity of the metallacycle established

and some understanding of the interconversion between the metallacycle and alkylidene

complexes, altering the substituents of pda ligands will possibly lead to new chemistry

favoring either formation of metallacycles or alkylidene complexes.

Without the studies

presented in this chapter, extremely little was known about the nature of the complexes

being formed and used for polymerizations.

Now more information is known, and new

ideas are presenting themselves.