CHEMISTRY of BIS(ALKYL) NITROSYL and RELATED COMPLEXES of MOLYBDENUM and TUNGSTEN by NEILH
CHEMISTRY OF BIS(ALKYL) NITROSYL AND RELATED COMPLEXES OF MOLYBDENUM AND TUNGSTEN By NEILH. DRYDEN B.Sc, The University of New Brunswick, 1985
A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
in
THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF CHEMISTRY
We accept this thesis as conforming
to the required standard
THE UNIVERSITY OF BRITISH COLUMBIA
August 1990
© Neil Howard Dryden, 1990 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.
Department of Chemistry
The University of British Columbia Vancouver, Canada
Date September 7, 1990
DE-6 (2/88) ii
Abstract
5 The reactions of (r7 -C5R5)M(NO)(CO)2 (R = H, Me; M = Mo, W) with PC15
5 result in the formation of the corresponding (TJ -C5R5)M(NO)(C1)2 products in high isolated yields (85-95%). These products have been fully characterized by conventional analytical and spectroscopic techniques including an X-ray crystallographic study of
5 5 [(rj -C5Me5)Mo(NO)(Cl)2l2. Alkylation of the (TJ -C5R5)W(NO)(C1)2 compounds with
Grignard reagents results in the formation of the corresponding complexes,
5 (r, -C5R5)W(NO)(R')2 (R = H, Me; R' = CH2CMe3, CH2CMe2Ph: R = Me;
R' = />-C6H4Me, Ph). An electrochemical study of (r^-C^) W(NO)(CH2CMe3)2,
5 5 (r, -C5H5)W(NO)(CH2CMe2Ph)2, and (T? -C5Me5)W(NO)(p-C6H4Me)2 shows 2 successive, chemically reversible, one electron reductions in THF for each complex.
5 The complexes (ij -C5R5)W(N"0)(R')2 (R = H; R' = CH2CMe3, CH2CMe2Ph:
R = Me; R' = CH2CMe2Ph,/?-C6H4Me) react with 1 atm of CO resulting in the formation
5 of the singly inserted products (r? -C5R5)W(NO){C(0)R'}(R') in good yields, presumably via initial CO coordination. These acyl complexes have been completely characterized by conventional techniques. The IR and NMR spectroscopic properties of these complexes are indicative of an f?2-acyl coordination mode for all of the acylalkyl and acylaryl products.
5 Under 6 atm of CO the (r, -C5H5)W(NO)(R')2 (R' = CH2CMe3, CH2CMe2Ph) complexes
5 are doubly carbonylated to form the corresponding (r? -C5H5)W(NO){C(0)R'}2 products.
5 Furthermore, (T7 -C5Me5)W(NO){C(0)/>-C6H4Me}(p-C6H4Me) reacts to form
5 (r? -C5Me5)W(NO)(CO){C(0)/?-C6H4Me}(p-C6H4Me) under 6 atm of CO, but there is no evidence for the formation of a bis(acyl) product. A preliminary X-ray crystallographic
5 2 investigation of (T7 -C5H5)W(NO){C(0)CH2CMe2Ph}2 reveals the presence of one 77 -acyl ligand and one n ^acyl ligand. The *H and ^C^H} NMR spectra of the
5 v7 -C5H5)W(NO){C(0)R'}2 compounds exhibit signals for only one type of acyl ligand, indicating that the complexes are stereochemically non-rigid in solution through a process which averages the signals for the and rp- acyl ligands. iii
5 The bis(benzyl) complexes, (r, -C5R5)M(NO)(CH2Ar)2 (R = H, Me; M = Mo, W;
5 2 Ar = Ph,/?-Tol), react with HC1 to form (r? -C5R5)M(NO)(r; -CH2Ar)(Cl) complexes.
These complexes have been reacted with Grignard reagents, R'MgCl, to form the
5 2 corresponding (r7 -C5R5)M(NO)(r? -CH2Ar)(R') (R' = CH2SiMe3, CH2CMe3, p-Td\) products. The benzyl ligand in all of these complexes is coordinated in an r?2-fashion, similar to that seen in the solid-state molecular structures of the bis(benzyl) precursors.
The coordination mode of the benzyl ligands has been confirmed by solid-state molecular
5 structure determinations of the representative examples (rj -C5Me5)Mo(NO)(CH2Ph)(Cl)
5 X and (r? -C5Me5)Mo(NO)(CH2Ph)(CH2SiMe3). The H and ^C^H} NMR spectroscopic data for all of the complexes are discussed, with focus on the diagnostic characteristics for the r/2-benzyl ligand. A possible qualitative interpretation for the symmetric r?2-bonding mode in these complexes is presented.
5 2 The (rj -C5Me5)M(NO)(r> -CH2Ph)(Cl) (M = Mo, W) complexes react with AgBF4 in CH3CN to form electrophilic complexes of the type
5 2 [(r; -C5Me5)M(NO)(r; -CH2Ph)(NCCH3)]BF4. The solid-state molecular structures of these complexes are discussed and contrasted with those found for the neutral r;2-benzyl
complexes. The reaction of racemic Ag02CCH(Et)(Ph) with the
5 2 (jj -QR5)M(NO)0? -CH2Ph)(Cl) (R = H, Me; M = Mo, W: R = H; M = Mo) compounds forms the corresponding diastereomeric
5 1 (»? -C5R5)M(NO)(r; -CH2Ph)(02CCH(Et)(Ph)) products. The r? ^coordination of the benzyl ligand and bidentate coordination of the carboxylate ligand are established by the
IR, 1H, and ^C^H} NMR spectroscopic data for these products. A mixture of two diasteromeric products is demonstrated by the 1H and 13C{1H} NMR spectroscopic data and initial attempts to separate these diastereomers by crystallization are described. iv
Table of Contents
Abstract ; . ii
List of Figures vii
List of Tables x List of Schemes xii List of Abbreviations xiii Acknowledgements xvi
CHAPTER 1 General Introduction 1
References and Notes 7 CHAPTER 2 Synthetic and Electrochemical Studies of 16-Electron Group 6 Nitrosyl Dichloro and Bis(alkyl) Complexes. 8
Introduction 9
Experimental Section 12
Electrochemical Measurements 13
Results and Discussion 23
Preparation of the Cp'M(NO)(Cl)2 Complexes 23
Mass Spectroscopic and Solid-State X-Ray Crystallographic
Investigation of the Molecular Structures of [Cp'M(NO)(Cl)2]n
(n = 1, 2) 26
Alkylations of Cp'W(NO)(Cl)2 With Grignard Reagents;
Improved Synthesis of CpW(NO)(R)2 (R= Npt, Neo) and
Preparation of Cp*W(NO)(R')2 (R = Neo, Npt, Ph, p-Tol) 31
Reduction Behaviour of CpW(NO)(Neo)2, CpW(NO)(Npt)2, and
Cp*W(NO)(p-Tol)2 37 V
Implications from the Electronic Structure of CpMo(NO)(Me)2 for the Electrochemical and Electronic Properties of
Cp'M(NO)(X)2 (X = alkyl, halide) Complexes 42
Epilogue 44
References and Notes 45 CHAPTER 3 Insertions of CO into Unsaturated 16-Electron Tungsten Bis(alkyl) and Bis(aryl) Complexes: Formation of Stable Acylaryl, Acylalkyl, and Bis(acyl) Species 48 Introduction 49 Experimental Section 51 Results and Discussion 65
Low-Pressure Carbonylation of Cp*W(NO)Q?-Tol)2 65
Reactivity of Cp*W(NO){C(0)/?-Tol} (p-To\) with H2 and CO 69
Low-Pressure Carbonylation of Cp'W(NO)(R)2 (Cp' = Cp, R = Npt, Neo; Cp' = Cp*, R = Neo) 71 Acylalkyl Complexes 78 High-Pressure CO Reactions of CpW(NO){C(0)R}(R) 80 Bis(acyl) Complexes 86
Epilogue . 88
References and Notes 89 CHAPTER 4 Preparation of Chiral Molybdenum and Tungsten Complexes:
Physical, Structural, and Spectroscopic Investigations of the Novel
7?2-Benzyl Ligand 92
Introduction 93
Experimental Section . 96
Results and Discussion 108
Preparation and Properties of Cp*M(NO)0?-Xyl)2 108
Preparation and Physical Properties of Cp'M(NO)(CH2Ar)(Cl). ... 109 vi
Preparation and Properties of the Complexes
Cp'M(NO)(CH2Ar)(R) (R = Tins, Npt,/>-Tol) 112 Molecular Structures of the Complexes Cp*Mo(NO)(Bz)(X)
(X = Cl,Tms) 114
NMR Spectroscopic Properties of the Cp'M(NO)(CH2Ar)(Cl) Complexes 124
NMR Spectroscopic Properties of the Cp'M(NO)(CH2Ar)(R) Complexes. 133 Epilogue 136 References and Notes 137 CHAPTER 5 Reactions of Cp'M(NO)(Bz)(Cl) With Silver (I) Salts: Formation of Benzyl Carboxylate and Benzyl Nitrile Complexes 140 Introduction 141 Experimental Section 143 Results and Discussion 150
Reactions of AgBF4 with Cp*M(NO)(Bz)(Cl) in CH3CN 150 Solid-State Molecular Structures of the
[Cp*M(NO)(Bz)(NCCH3)]BF4 Complexes 151
NMR Spectroscopic Properties of the Nitrile Complexes 158
Preparation, Physical, and Spectroscopic Properties of the
Cp'M(NO)(Bz)(02CR*) Complexes (R* = CH(Et)(Ph)) 159
Summary and Future Work 167
References and Notes 169 Appendix 172 vii
List of Figures
Figure 2.1 The SNOOPI plot of the solid-state molecular structure of
[Cp*Mo(NO)(Cl)2]2 with the hydrogen atoms omitted for clarity 27
J Figure 22 The 80 MHz H NMR spectrum of Cp*W(NO)(Npt)2 in C6D6 35
Figure 2.3 Ambient temperature cyclic voltammogram of CpW(NO)(Neo)2 in
THF (0.1 N [n-Bu4N]PF6, Pt working electrode) at a scan rate of
0.56 Vs1 39
Figure 3.1 Flow system for circulation of solutions for in situ IR monitoring of
reactions. (A) 0.2 mm NaCl flow cell (Wilmad,
Cat. # 105A10-5). (B) Teflon diaphragm pump (Cole-Parmer,
Cat. # 07090-62) 52
Figure 3.2 Spectra from in situ IR monitoring of the reaction of
CpW(NO)(Neo)2 with CO in CH2C12 with the flow system shown in
Figure 3.1 55
Figure 3.3 The 80 MHz JH NMR spectrum of Cp*W(NO){C(0)p-Tol}(p-Tol) in
C6D6. Inset shows irradiation of the doublet at 8.19 ppm causing the
decoupling of the doublet at 7.32 ppm 67
Figure 3.4 The 300 MHz JH NMR spectrum of CpW(NO){C(0)Neo}(Neo) in
C6D6. Inset is an expansion of the region from 6 2.6-2.8 75
Figure 3.5 The 300 MHz *H NMR spectrum of CpW(NO){BC(0)Npt}(Npt) in
C6D6. Inset is an expansion of the region from s 2.2-2.35 76
X Figure 3.6 The 300 MHz H NMR spectrum of CpW(NO){C(0)Neo}2 in
CDC13 83
Figure 3.7 The variable temperature 80 MHz *H NMR spectra of
CpW(NO){C(0)Neo}2 in CDC13 from -50 to 20 °C 84
Figure 3.8 The 75 MHz ^C^H} NMR spectrum of CpW(NO){C(0)Neo}2 in
CDCI3. Inset is an expansion of the region from 6 125-129 85 viii
l Figure 4.1 (a) The 200 MHz H NMR spectrum of Cp*W(NO)(p-Xyl)2 in C6D6. (b) The 50 MHz "C^H} APT NMR spectrum of
Cp*W(NO)(p-Xyl)2 in C6D6. The carbon signals for Cquat and CH2
groups are positive and the carbon signals for CH and CH3 groups are negative) . 110
Figure 4.2 The ORTEP plot of the molecular structure of Cp*Mo(NO)(Bz)(Cl). Hydrogen atoms have been omitted for clarity 115 Figure 4.3 The ORTEP plot of the molecular structure of Cp*Mo(NO)(Bz)(Tms). Hydrogen atoms have been omitted for clarity 117
l Figure 4.4 (a) The 80 MHz U NMR spectrum of CpW(NO)(Bz)2 in CD2C12 at - 100 °C.
l (b) The 80 MHz B NMR spectrum of CpW(NO)(Bz)(Cl) in C6D6 at 25 °C 126
Figure 4.5 (a) The 300 MHz *H NMR spectrum of Cp*W(NO)(Bz)(Cl) in C6D6. (b) The 50 MHz "C^H} NMR spectrum of Cp*W(NO)(Bz)(Cl) in
CDC13. 128
l Figure 4.6 (a) The 300 MHz H NMR spectrum of Cp*Mo(NO)(Bz)(Cl) in CDC13.
(b) The 75 MHz 13C{lH) NMR spectrum of Cp*W(NO)(Bz)(Cl) in
CDCI3 129
Figure 4.7 Variable temperature 80 MHz JH NMR spectra of Cp*Mo(NO)(Bz)(Cl)
in CDCI3 130
Figure 4.8 (a) The 300 MHz *H NMR spectrum of Cp*Mo(NO)(Bz)(Tms) in
C6D6.
(b) The 75 MHz 13C{!H} NMR spectrum of Cp*W(NO)(Bz)(Cl) in
C6D6. Inset is an expansion of the aromatic region from 135 to 127 ppm 135 ix
Figure 5.1 The ORTEP plot of the solid-state molecular structure of
[Cp*W(NO)(Bz)(NCCH3)]BF4. The hydrogen atoms on the Cp*
ligand have been omitted for clarity 153
Figure 52 The ORTEP plot of the solid-state molecular structure of
[Cp*Mo(NO)(Bz)(NCCH3)]BF4. The hydrogen atoms on the Cp*
ligand have been omitted for clarity 154
Figure 5.3 The 300 MHz *H NMR spectrum of [Cp*Mo(NO)(Bz)(NCCH3)]BF4
inCD3N02 • 160
Figure 5.4 The 75 MHzB C{1H} NMR spectrum of
[Cp*Mo(NO)(Bz)(NCCH3)]BF4 in CD3N02 161
Figure 5.5 The 200 MHz1 H NMR spectrum of
CpMo(NO)(Bz)(02CCH(Et)(Ph)) in CDC13 164
Figure 5.6 The 200 MHz lW NMR spectrum of
Cp*Mo(NO)(Bz)(02CCH(Et)(Ph)) in CDC13 166 X
List of Tables
Table 2.1 Selected Bond Distances (A) and Bond Angles (°) for the Molecular
Structure of [Cp*Mo(NO)(Cl)2]2 at 200 K 28 Table 2.2 Low-Resolution Mass Spectral Data for the Dihalo Complexes 30
Table 2.3 Data for the Reductions of the Cp'W(NO)(R)2 Complexes 40
Table 3.1 Infrared and Elemental Analysis Data for the Acylaryl, Acylalkyl, and
Bis(acyl) Complexes 60
Table 3.2 Mass Spectral, 1H, and 13C{1H} NMR Data for the Acylaryl,
Acylalkyl, and Bis(acyl) Complexes 61
Table 4.1 Infrared and Elemental Analysis Data for the New Bis(benzyl),
Benzylchloro, and Benzylalkyl Complexes 101
Table 4.2 Mass Spectral and JH and BC{1H} NMR Data for the Bis(benzyl)
and Benzylchloro Complexes 102
Table 4.3 Mass Spectral and JH and BC{1H} NMR Data for the Chiral
Benzylalkyl Complexes 105
Table 4.4 Selected Bond Distances (A) and Bond Angles (°) and Torsion
Angles (°) for the Molecular Structure of Cp*Mo(NO)(Bz)(Cl) 116
Table 4.5 Selected Bond Distances (A) and Bond Angles (°) and Torsion
Angles (°) for the Molecular Structure of Cp*Mo(NO)(Bz)(Tms) 118
Table 4.6 Comparison of Bond Lengths and Angles for the Benzyl Radical (Bz*)
and the r?2-Benzyl Ligands 123
Table 5.1 Infrared and Elemental Analysis Data for the Benzyl Acetonitrile and
Benzyl Carboxylate Complexes 146
Table 5.2 Mass Spectral and *H and BC{1H} NMR Data for the Benzyl
Acetonitrile and Benzyl Carboxylate Complexes 147
Table 5.3 Selected Bond Distances (A) and Bond Angles (°) and Torsion
Angles (°) for the Molecular Structure of
[Cp*W(NO)(Bz)(NCCH3)]BF4 155 xi
Table 5.4 Selected Bond Distances (A) and Bond Angles (°) and Torsion Angles (°) for the Molecular Structure of
[Cp*Mo(NO)(Bz)(NCCH3)]BF4 156 xii
List of Schemes
Scheme 2.1 10
Scheme 22 42 xiii
List of Abbreviations
A - thermolysis APT - attached proton test (in NMR spectrum) Ar - aryl, phenyl or p-tolyl arom - aromatic; ortho, meta or para
Bz - CH2C6H5, benzyl
13C - carbon-13 ^C^H} - proton-decoupled carbon-13 calcd - calculated
CgD6 - benzene-rf6
CDCI3 - chloroform-rf
CD2C12 - dichloromethane-J2
CD3N02 - nitromethane-d3 cm"1 - wavenumbers
5 Cp - T, -C5H5 Cp* - ^-CsMej Cp' - Cp or Cp* 8 - chemical shift in ppm referenced to Me^i at 6 0 d - doublet (in the NMR spectrum)
DME - MeOCH2CH2OMe, 1,2-dimethoxyethane
dippe - (Me2CH)2PCH2CH2P(CHMe2)2, l,2-bis(diisopropylphospino)ethane
dppe - (Ph)2PCH2CH2P(Ph)2, l,2-bis(diphenylphospino)ethane
Et - CH2CH3, ethyl
Et20 - (CH3CH2)20, ether, diethyl ether
EI - electron impact eV - electron volts hu - photolysis *H - proton
HOMO - highest occupied molecular orbital IR - infrared iR - (current)'(resistance) / - coupling constant (in the NMR spectrum) LRMS - low-resolution mass spectrum LUMO - lowest unoccupied molecular orbital m - multiplet (in the NMR spectrum) M - Mo or W m/z - mass-to-charge ratio in the mass spectrum
Me - CH3, methyl mmol - millimole MO - molecular orbital
v - stretching mode (in IR spectrum)
Neo - CH2CMe2Ph, neophyl NHE - normal hydrogen electrode
NMR - nuclear magnetic resonance
Npt - CH2CMe3, neopentyl
P+ - molecular ion (in the mass spectrum)
p-Tol - C6H4Me, para-tolyl
p-Xy\ - CH2C6H4Me, para-xylyl
Ph - C6H5, phenyl
q - quartet (in the NMR spectrum)
quat - quaternary carbon atom s - singlet (in the NMR spectrum) SCE - saturated calomel electrode t - triplet (in the NMR spectrum) XV
TBAH - tetrabutylammonium hexafluorophosphate, [(C4H9)4N]PF6
Tetraglyme - tetraethylene glycol dimethyl ether, Me(OCH2CH2)4OMe
THF - C4HgO, tetrahydrofuran
Tms - CH2SiMe3, trimethylsilylmethyl V - volts xvi
Acknowledgements
First and foremost, I would like to sincerely thank Peter Legzdins for his friendship and support throughout the course of this work. Without a doubt, he has made this an enjoyable and enlightening experience with his generous guidance and encouragement over •the past 5 years.
I would also like to thank the members of the Legzdins group, past and present, for the good working environment they have provided. In particular, special thanks goes to the people I worked most closely with (Nancy, George, Ev, and Teen), the last of the 319/325 crowd for their friendship in and out of the lab. Also I am indebted to all of my proofreaders (especially Peter, George, John, and Drs. F. Aubke and E. Piers) for their suggestions for improvements to this thesis. Thanks to Drs. V. Yee and J. Trotter of UBC and Drs. F.W.B. Einstein, R.H. Jones, and RJ. Batchelor of SFU for their collaboration on the X-ray crystallography work. Also thanks to everyone in the department who assisted in this work and in particular Peter Borda, Steve Rak, Marietta Austria, and the Mechanics and Electronics shop guys for helpful discussions. As well, I would like to acknowledge the people of Canada who supported my work through a NSERC postgraduate scholarship.
Finally, I sincerely thank my mother and father for their unwavering support and
patience through the years which have allowed me to achieve this goal. 1
CHAPTER 1
General Introduction 2
The field of transition-metal organometallic chemistry is one of the younger subdisciplines of modern chemistry. Although transition metal complexes containing carbon-bound ligands have been in existence for more than 160 years,1 the nature of the bonding present in these complexes was poorly understood until recent times. The beginning of the growth of transition-metal organometallic chemistry into the burgeoning field we know today can be traced to the synthesis of ferrocene,2'3 (C^lrLj^Fe, and the subsequent JT -bonding description for this complex proposed in 1952.4 It was this bonding description, along with those proposed to explain other metal ligand interactions, which initiated the interest in this general class of complexes that continues to this day. As well, the advent of better techniques since 1952 for studying organometallic complexes has also promoted the growth in this field; e.g. FT-NMR spectroscopy, solid state X-ray crystallographic analysis, and molecular orbital calculations all made more accessible by the introduction of inexpensive computer technology.
The compounds synthesized in this Thesis are all organometallic nitrosyl complexes, which are distinguished by the presence of a nitrosyl ligand (NO) among the ligands bound to the central transition metal.5 Some basic concepts necessary to describe the bonding of these groups to the metal centre are sketched out below.
The cyclopentadienyl group (Cp), along with its various substituted analogues, is a very common ligand in organometallic chemistry. Generally it is bound in a symmetric fashion through all five carbon atoms to a metal centre, i.e.,
Other coordination geometries with less than five carbons bonded to the metal exist but these are less common. In order to distinguish between these differing coordination
n numbers, the cyclopentadienyl ligand is more completely described in the form r? -C5H5
(n= 1-5). The prefix r)n denotes how many carbon atoms (n) of the cyclopentadienyl ligand are coordinated to the metal. This notation is used throughout organometallic chemistry, for any ligand which can exhibit varying coordination numbers.
The nitrosyl ligand can also adopt different bonding modes for monometallic complexes. The two linkage modes seen for monometallic complexes are:
a) linear nitrosyl b) bent nitrosyl and these will be considered in turn. The first case to consider is the linear nitrosyl. It is by far the most commonly occurring geometry for complexes of the earlier transition metals, including those in this
Thesis. In this linkage the nitrosyl is considered to be a formal 3-electron donor to the metal centre. The interactions that give rise to the hnear nature are diagrammed below.
M
M N=0
7r donation
MdTt—NO JI*
From this picture it is obvious why the NO ligand is linear, since the optimal overlap for both of these interactions occurs with a linear geometry along the M-N = 0 axis.
The degree of Md 7r-+NOp 7r * electron donation is a function of the amount of electron density on the metal centre. Since electron donation into the NOp *•* orbital reduces the
NO bond order, changes in this electron donation are reflected in the NO bond strength and can be quantified by the nitrosyl stretching frequency (I^NO) observed in the infrared
(IR) spectrum of the complex. This stretching frequency serves as a sensitive probe of changes in the electron-richness of a metal centre (generally, with more electron-rich
complexes exhibiting lower i/NO values), as well as serving as an easily monitored characteristic to follow the progress of reactions of nitrosyl complexes via IR spectroscopy.
The second case to consider is the bent nitrosyl ligand. In this case the bonding is through the singly occupied NO JT* orbital to a a orbital on the metal centre. In this fashion the NO acts as a 1-electron donor ligand.
M
A common concept encountered in describing organometallic complexes is the number of electrons in the valence shell of the metal centre. This is often referred to as the 18-electron rule, although it is more a guideline than a rule. It arises from the observation that complexes where the sum of the metal's valence electrons and electrons donated from the ligands equals 18 are generally most stable. Species of this type are called, appropriately 18-electron complexes, and their stability can be qualitatively understood by the simple fact that filling the 9 valence orbitals of a transition metal with 18 electrons (from the ligands and the metal) results in a closed shell. The electron counting formalism used in this Thesis considers all ligands and metals to be neutral (except where there is an overall charge on the complex).6 An example of how the electron count of a complex is determined is outlined below.
Consider the known complex CpMo(NO)(»?3-allyl)(I):
5 Cp (r? -C5H5) - 5 electron donor (a neutral cyclopentadienyl radical)
Mo - 6 electrons in its valence shell
NO - 3 electron donor (linear)
173 -allyl - 3 electron donor (neutral allyl radical, 1 each in 3 *• orbitals) I 1 electron donor (neutral iodide radical)
Total 18 electrons i.e., a closed valence shell. 5
These formalisms do not predict a stable or unstable configuration, but simply serve as a useful guide to more probable stable arrangements of ligands and metals.
The class of complexes which are the subject of this Thesis have a general formula
Cp'M(NO)(L)(L') M = Mo, W
5 Cp' = r, -C5R5, R = H or CH3
L, L' = halide, alkyl, aryl, »j2-benzyl or acyl
Chapter 2 deals with the preparation of some 16-electron tungsten bis(alkyl) and bis(aryl) complexes as well as the development of a convenient synthesis of the requisite
precursor complexes, Cp'M(NO)(Cl)2. An electrochemical study of the reduction behaviour of some of the bis(alkyl) and bis(aryl) complexes is also presented. The results of the electrochemical study are interpreted in light of the frontier MO's of these complexes predicted by Fenske-Hall MO calculations on the model complex
CpMo(NO)(CH3)2.
Chapter 3 contains an investigation of the reactivity of the tungsten bis(alkyl) and bis(aryl) complexes with carbon monoxide (CO). The preparations of stable acylalkyl,
Cp'W(NO)(C(0)R)(R), andbis(acyl), CpW(NO)(C(0)R)2 products complexes are described, as well as spectroscopic investigations on the bonding in these complexes.
Chapter 4 deals with the preparation of a series of chiral complexes,
2 CpM(NO)(Bz)(X) (Bz = CH2Ar; X = alkyl, chloride), which all contain novel >j-benzyl ligands which act as formal 3-electron donors. The characteristic spectroscopic properties of these complexes and the solid-state molecular structures for representative examples of these complexes are discussed. A rationale of the nature of the rj2-benzyl interaction is also presented.
Chapter 5 deals with the chloride abstraction reactions of the Cp'M(NO)(Bz)(Cl)
complexes with silver(I) reagents to form both [Cp*M(NO)(Bz)(NCCH3)]BF4 and
Cp'M(NO)(Bz)(02CCH(Et)(Ph)) products. The nature of the benzyl linkages in 6
the [Cp*M(NO)(Bz)(NCCH3)]BF4 complexes is discussed with regard to their solid-state molecular structures and spectroscopic properties. The spectroscopic properties of the
Cp'M(NO)(Bz)(02CCH(Et)(Ph)) products are discussed in light of the diastereomeric nature of these compounds, as well as establishing the presence of r? 1-benzyl ligands and bidentate carboxylate ligands in all of these species. The potential utility of this reactivity towards the resolution and isolation of optically pure complexes is also discussed. References and Notes.
The first known transition-metal organometallic complex is K+ [(C^H^PtCy:
Zeise, W.C. Ann. Phys. 1827, 9, 932.
Kealy, TJ.; Pauson, P.L. Nature (London) 1951,168, 1039.
Miller, S.A.; Tebboth, J.A.; Tremaine, J.F. /. Chem. Soc. 1952, 632.
Wilkinson, G.; Rosenblum, M; Whiting, M.C; Woodward, R.B. /. Am. Chem. Soc.
1952, 74,2125. An entertaining and personal account of the work leading to this publication can be found in: Wilkinson, G. /. Organomet. Chem. 1975,100, 273.
A recent review of organometallic nitrosyl chemistry has appeared: Legzdins,. P.;
Richter-Addo, G.B. Chem. Rev. 1988,88, 989. Earlier review articles dealing with organometallic nitrosyl chemistry are also referenced there.
An alternative counting scheme to the one outlined here where the metal centres are assigned formal oxidation states is presented in: Collman, J.P.; Hegedus, L.S.;
Norton, J.R.; Finke, R.G. Principles and Applications of Organotransition Metal
Chemistry, University Science Books: Mill Valley, CA, 1987; Chapter 2. The same electron-counts are obtained for a complex regardless of the method used to reach them. CHAPTER 2
Synthetic and Electrochemical Studies of 16-Electron Group 6 Nitrosyl Dichloro and Bis(alkyl) Complexes 9
Introduction
Previous work by Dr. Luis Sanchez of these laboratories resulted in the discovery1 of a family of 16-electron bis(alkyl) complexes of the general formula:
CpM(NO)R2
M=W, R= CH2SiMe3 (Tms), CH2CMe3 (Npt), CH2CMe2Ph (Neo); M = Mo,R=Tms
A surprising property of these formally unsaturated, 16 valence-electron complexes was their thermal and kinetic stability. For example, these complexes could be handled in air
as solids for short periods of time, although in solution they do react with 02 over longer periods of time. A Fenske-Hall MO study was performed on the model complexes
CpM(NO)(Me)2 (M = Mo, Ru), the results of which showed that the Ru HOMO (i.e. Mo LUMO) is metal-centred and non-bonding in nature. Therefore filling of this orbital in the 18-electron Ru model complex as compared to the Mo model complex should not affect the bonding framework of either species and no additional stability should be present for
the 18-electron CpRu(NO)(Me)2 over the 16-electron CpMo(NO)(Me)2. The reactivity of these bis(alkyl) complexes was investigated subsequently using
CpW(NO)(Tms)2 as a representative example of this group. The use of reagents with
Lewis base properties exploited the coordinatively unsaturated nature of the metal centre.2 A summary of the results of these studies is presented in Scheme 2.1, which includes the unprecedented insertion of elemental sulfur into a metal-carbon bond.3 Since only bulky alkyl ligands had been employed in the preparation of the
CpM(NO)(R)2 complexes, I set out to extend this synthesis to include related complexes containing other types of hydrocarbyl ligands. It was envisaged that the chemistry observed
for the general family of Cp'M(NO)(R)2 complexes could be dependent on the nature of the hydrocarbyl group. 10
A 0 0 A R o R R v N AIMe3 R ER i/8 Ea E=S|1/8S8
PMej 2/8 Sa • W. R-/\R A
PMe3 R N R CNR
2 NO
W W / • \ R N V KR'^ Q 0—N»S p RE N ER K 0 ^R R = CH2SiMe3, R'=t-Bu, E=S,Se Scheme 2.1
For instance, in the related family of complexes Cp*Ru(NO)(R)2, Bergman and coworkers have noted differences in the comparative reactivity of the alkyl versus aryl substituted complexes,4'5 i.e.,
Cp Ru(NO)Me2 + 2 PMe3 Cp Ru(CN)(PMe3)2 (2.1)
Cp Ru(NO)Ph2 + 2 PMe3 Cp Ru(NO)(PMe3)2 + Ph-Ph (2.2)
Thus, the synthesis of the aryl substituted complexes was attempted. The potential
to create a sterically less hindered metal centre in the Cp'M(NO)(aryl)2 complexes (as compared to the bis(alkyl) analogues) to enhance reactivity of the bis(aryl) complexes with larger substrates was also considered desirable. The preparations of the title complexes
are presented in this Chapter including the successful isolation of the new Cp*W(NO)(R)2
complexes (R = Neo, Npt, Ph, p-To\ (p-QH4Me)). Electrochemical studies of some of these complexes were undertaken to investigate the reduction behaviour of these species. 11
In earlier studies, Cp'M(NO)(I)2 starting materials were used as precursors in the synthesis of various group 6 alkyl and diene complexes.2,6 However, an electrochemical examination of the reduction behaviour of the dihalomolybdenum complexes,
Cp'Mo(NO)(X)2 (X = CI, Br, I), has shown that the chloro species are expected to be superior synthetic precursors for the preparation of the Mo bis(alkyl) complexes.7 It was reasonable to assume that the same conclusion could hold true for the tungsten congeners.
Thus, the Cp'W(NO)(Cl)2 precursor complexes were used in the preparations of the bis(alkyl) and bis(aryl) compounds mentioned in the preceding paragraph. Since the tungsten dichloro complexes were unknown at that time, I proceeded to develop a
convenient synthetic route to all four of the Cp'M(NO)(Cl)2 complexes. Their successful
preparations, along with the solid-state molecular structure of [Cp*Mo(NO)(Cl)2]2, are presented in this Chapter. 12
Experimental Section
All reactions and subsequent manipulations involving organometallic reagents were performed under anhydrous conditions in an atmosphere of dinitrogen unless specified otherwise. Conventional dry-box and vacuum line Schlenk techniques8 were utilized unless otherwise noted. The organometallic complexes CpM(NO)(CO)2
9 10 (M = Mo, W) and CpM*(NO)(CO)2 (M = Mo, W) were prepared by established procedures. All phosphorus chloride reagents were purchased from commercial suppliers
(PCI5, Aldrich; PCI3, BDH) and used without further purification. The alkyl halide reagents (Aldrich) neophyl chloride (2-phenyl-2-methyl-l-chloropropane), neopentyl chloride (2,2-dimethyl-l-chloropropane), chlorobenzene, and p-chlorotoluene were purified by distillation from P2O5 under reduced pressure. Solvents were freshly distilled • from appropriate drying agents under a dinitrogen atmosphere and purged for 10 min with dinitrogen to remove dissolved dioxygen prior to their use: Tetrahydrofuran (THF) and ether (Et20) were distilled from sodium/benzophenone; hexanes and pentane were distilled from sodium/ benzophenone/tetraglyme; dichloromethane (CH2CI2) was distilled from P2O5; toluene was distilled from sodium.
All IR spectra were recorded on a Nicolet 5DX FT-IR instrument, internally calibrated with a He/Ne laser, as solutions in NaCl cells or as Nujol mulls sandwiched between NaCl plates unless otherwise noted. All *H and ^C NMR spectra were obtained on Varian Associates XL-300 or Bruker WP-80 or AM-200E spectrometers, and the chemical shifts of the observed resonances are reported in parts per million downfield from
Me4Si referenced to the residual proton signal (C6D6, s 7.15; CDCI3, S 7.25) or natural
abundance carbon signal (C6D6, 6 128.00; CDC13, 6 77.00) of the solvent used. The
BC NMR with NOE enhancement and 1H coupling were obtained as for 13C{1H} spectra except with {1H} on between acquisitions (~1.6 s) and {^H} off during acquisition of the
FID (~0.8 s). The BC APT (Attached Proton Test) NMR spectra were recorded using standard routines included in the operating software for the Varian XL-300 and 13
Bruker AM-200E NMR instruments. In some cases long relaxation times for the Cp protons, indicated by low integrations for these signals, required the addition of a 10-20 s relaxation delay between successive acquisitions of the *H NMR FID signals to allow the proper integrations to be obtained. Mrs. M.T. Austria, Ms. L.K. Darge, and Dr. S.O. Chan assisted in obtaining the NMR data. Low-resolution mass spectra (EI, 70 eV) were recorded on a Kratos MS50 spectrometer using the direct-insertion method by Dr. G.K. Eigendorf, and Mr. M.A, Lapawa. All elemental analyses were performed by Mr. P. Borda of this Department.
Electrochemical Measurements. The detailed methodology employed for cyclic voltammetry (CV) studies in these laboratories has been outlined elsewhere.7'11 Cyclic voltammograms were recorded on a Hewlett-Packard Model 7090A Measurement Plotting System in buffered recording mode. The THF used for these studies was first treated as described above, then stirred with activated alumina (Woelm neutral, Super I grade) for 15 min before being transferred via cannula into the electrochemical cell. The potentials, E°, are defined as the average of the cathodic and anodic peak
potentials, (£pc + £pa)/2, and are reported versus the aqueous saturated calomel
electrode (SCE). Solutions were prepared at 0.1 M in the [n-Bu4N]PF6 (TBAH) support
electrolyte and ~6 x 10"4 M in the organometallic complex to be studied, and the solutions
were maintained under an atmosphere of N2. The 3-electrode cell consisted of a Pt bead working electrode (~ 1 mm diameter), a Pt wire auxiliary electrode, and a SCE in a separate holder connected to the cell compartment via a Luggin probe separated by a frit. Compensation for iR drop in potential measurements was not employed during this study. Ferrocene (Fc) was used as an internal reference in these studies, with the redox couple, Fc/Fc , occurring at E° = 0.54 V versus SCE in THF over the range of scan rates (y) used
(0.05 to 1.5 V s"1). The anodic and cathodic peak separation (AE) for this couple increases
with increasing scan rate (80-170 mV between 0.05-1.5 V s"1) but since the Fc/Fc+ couple is known to be highly reversible, other redox couples exhibiting similar peak separations to the internal standard were also considered to be electrochemically reversible. The ratio of 14
cathodic peak current to anodic peak current, /ptC//p>a, for the oxidation of Fc was unity over all scan rates used, as expected for a chemically reversible process. The linearity of a
1 2 plot of/pa vs. v / was checked for all reduction processes to establish the presence of diffusion control.
Treatment of CpW(NO)(CO)2 with PC13. An orange solution of CpW(NO)(CO)2
(1.67 g, 5.0 mmol) in Et20 (150 mL) was prepared in a 300-mL reaction flask fitted with a
pressure-equalized dropping funnel. The funnel was then charged with Et20 (50 mL) and
PCI3 ( 0.44 mL, 5.0 mmol). The solution of PCI3 was slowly added to the reaction mixture.
No reaction had occurred after 3 h as judged by the IR spectrum of the reaction mixture, which showed no change in position of the VQQ and ^NO Dands from those of the original
CpW(NO)(CO)2 complex.
Reaction of the Cp'M(NO)(CO)2 complexes with PCI5. All of these reactions were
performed in a similar manner, the reaction of Cp*W(NO)(CO)2 with PCI5 being described
as a representative experiment for the formation of all the Cp'M(NO)(Cl)2 complexes.
CAUTION: Large amounts of CO gas are evolved during the reaction. Care must be taken to avoid the buildup of excess pressure during the initial stages of the reaction and to vent the CO to a fumehood.
An orange solution of Cp*W(NO)(CO)2 (18.0 g, 44.5 mmol) in Et20 (120 mL) was
prepared and cooled for 15 min in a saturated CaCl2(aq)/Dry Ice bath (-20 °C). A sample of PCI5 (8.80 g, 42.0 mmol) was weighed into a glass vial, sealed with an airtight polyethylene stopper and removed from the glovebox. The solid PCI5 was quickly
transferred from the vial into the cooled, rapidly stirred solution of Cp*W(NO)(CO)2
under a constant countercurrent of N2. The flask was kept open to the bubbler on the N2 manifold and the cooling bath was removed. As the reaction warmed to room temperature, the solution changed colour from orange to green and vigorous gas evolution occurred along with the deposition of a brown solid. When all observable gas evolution had ended, the reaction solution was warmed to ~35 °C with a warm water bath, and the reaction vessel periodically evacuated to remove CO. When no more CO was evacuated 15 from the reaction mixture, as judged by the lack of pump noise during evacuation, the reaction was considered to be complete. The brown solid obtained was isolated from the pale green supernatant solution by filtration on a medium porosity glass frit, and this solid was washed with Et20 (3 x 40 mL) to remove traces of the PCI3 byproduct and excess
Cp*W(NO)(CO)2. The solid was then dried under vacuum (10"2 torr) for 12 h to obtain
Cp*W(NO)(Cl)2 (14.65 g, 83% yield based on PC15) as a brown powder.
Both the Cp and Cp* tungsten complexes were isolated in 80-85% yields as brown powders while the Cp and Cp* molybdenum complexes were isolated in 90-95% yields as orange powders.
Anal. Calcd for C5H5C12N0W: C, 17.16; H, 1.44; N, 4.00. Found: C, 17.01; .
1653 1 H, 1.50; N, 3.80. IR (CH2C12) ^No cm". *H NMR (CDCI3) 8 6.35 (s).
^C^H} (CDCI3) 6 107.62. Low-resolution mass spectrum: see Table 2.2.
Anal. Calcd for C5H5Cl2MoNO: C, 22.93; H, 1.92; N, 5.34. Found: C, 22.96;
1 H, 1.87; N, 5.26. IR (CH2C12) i/NO 1690 cm". *H NMR (CDCI3) 6 6.29(s).
^C^H} (CDCI3) not observed. Low-resolution mass spectrum: see Table 2.2.
Anal. Calcd for C10H15C12NOW: C, 28.60; H, 3.60; N, 3.33. Found: C, 29.00;
1 X H, 3.68; N, 3.06. IR (CH2C12) i>NO 1624 cm' . H NMR (CDCI3) 6 2.11 (s).
^CI^H} (CDCI3) 6 118.36,10.55. Low-resolution mass spectrum: see Table 2.2.
Anal. Calcd for CioH^C^MoNO: C, 36.17; H, 4.55; N, 4.22. Found: C, 36.17;
1 X H, 4.60; N, 4.30. IR (CH2C12) fNo 1659 cm". H NMR (CDC13) 6 1.95(s).
13C{1H} (CDCI3) 5 120.51,10.60. Low-resolution mass spectrum: see Table 2.2.
The Cp*Mo(NO)(Cl)2 complex could be crystallized from CH2C12 by Sohxlet extraction of the orange powder over a period of 8 h with a 95% recovery of solid. The extraction solution was allowed to cool slowly to room temperature which resulted in the deposition of large well-formed red blocks of the product. A single crystal suitable for
X-ray crystallographic analysis was selected from the product obtained in this manner.
An attempt to crystallize Cp*W(NO)(Cl)2 from a saturated CH2C12 solution by cooling to -25 °C resulted in the loss of the organometallic complex. This was 16 noted by the conversion of the initial green solution to a deep blue solution, along with the formation of a blue, oily residue on the sides of the flask. No attempt to purify this solution was made.
General Preparation of R2Mg(dioxane)2. A procedure for the preparation of the dialkylmagnesium reagents is outlined below, using a modification of a published procedure for the preparation of unsolvated dialkylmagnesium complexes.12 This method
was generally applicable, the reagents (p-Tol)2Mg(dioxane)2, (Neo)2Mg(dioxane)2 and
(Npt)2Mg(dioxane)2 being successfully prepared by this route.
13 14 A solution of the Grignard reagent in Et20 (NptMgCl, NeoMgCl ) or THF
(p-TolMgCl15) at 0.4 to 1.0 N concentration was prepared, and the solutions were treated with 1,4-dioxane (~4 equivalents). This resulted in the formation of a thick white slurry of
insoluble MgCl2(dioxane) in the solution. The supernatant solution containing
R2Mg(dioxane)2 was isolated by centrifuging portions of the slurry and cannulating the clear supernatant solutions out of the centrifuge tubes into a flask. The solvent was removed from the combined supernatant solutions under reduced pressure to give a white solid or a tarry residue. These products were triturated with hexanes, the hexanes were then carefully removed by cannulation, and the remaining solid dried in vacuo to obtain the organomagnesium products as free-flowing white powders. The solids were standardized by hydrolysis of weighed solid samples and titration of the resulting solutions with 0.100 N HC1 using phenolphthalein as an indicator. This allowed the composition of the powders to be determined for their use in further synthetic studies. In each case, these powders gave an amount of base approximately equal to that calculated for a complex with the
formulation R2Mg(dioxane)2.
Preparation of CpW(NO)(Npt)2. A solution of CpW(NO)(Cl)2 (5.00 g,
14.3 mmol) in THF (60 mL) was cooled in a saturated CaCl2(aq)/Dry Ice bath (-35 °C) for
15 min. A sample of (Npt)2Mg(dioxane)2 (4.50 g, 150 mg/mmol Npt", 30 mmol) dissolved in THF (30 mL) was rapidly cannulated into the cooled solution of the dichloride. The colour of the solution changed immediately from an emerald green to red-purple. The 17 reaction was stirred for 1 h, then the cooling bath was removed, and the reaction mixture stirred for an additional 1 h. The solvent was removed under reduced pressure, leaving a red residue which was then redissolved in Et20 (100 mL), treated with 2 mL of deoxygenated water, and stirred for 15 min. The resulting solution was filtered through a column of Florisil (80 x 30 mm) supported on a medium porosity glass frit and the column was washed with Et20 to completely elute a single red band. The red filtrate was reduced in vacuo to a viscous red oil. This product was freed of volatile oils by addition of a liquid nitrogen cooled sublimation probe to the flask and sublimation of the oils onto the probe under dynamic vacuum. The resulting red solid residue was dissolved in pentane (200 mL) and placed in a freezer (-25 °C) to induce crystallization. Overnight, the pentane solution deposited dark red needles and prisms of the product, which were collected by cannulating the mother liquor into another flask and washing the red solid with pentane (20 mL). The solid was then dried in vacuo to give CpW(NO)(Npt)2 (3.30 g) as a dark red crystalline product. The pentane wash solution was added to the mother liquor and the combined solutions saturated by removal of solvent under reduced pressure. Cooling of this saturated solution (-25 °C) afforded an additional
1.06 g of product as dark red needles. The CpW(NO)(Npt)2 product (72% overall yield) was characterized by comparison of its distinctive spectroscopic properties with those
2 1 1 previously reported for this complex : IR (Nujol) v^o 1560 cm". H NMR (C6D6)
2 6 5.04 (s, 5 H, Cp), 3.57 (d, 2 H, /HH= H-9 Hz, CtfaHb), 1.27 (s, 18 H, C(Cff3)3),
2 -1.55 (d, 2 H, /HH= H.9 Hz, CH^b).
Preparation of CpW(NO)(Neo)2. A solution of CpW(NO)Cl2 (3.50 g, 10.0 mmol) in THF (60 mL) was cooled in a saturated CaCl2(aq)/Dry Ice bath (-35 °C) for 15 min. Then an aliquot of NeoMgCl in Et20 (vide supra) (50 mL, 0.4 N, 20 mmol) was rapidly syringed into the cooled solution of the dichloride, causing an immediate colour change from emerald green to red-purple. The reaction mixture was stirred for 1 h, then the cooling bath was removed and the mixture stirred for an additional 30 min. The solvent
was removed in vacuo to give a red residue which was then redissolved in Et20 (100 mL). 18
The Et20 solution was treated with distilled, deaerated water (2 mL) and stirred for 10 min, during which time an off-white precipitate formed. The reaction mixture was then filtered through a column of Florisil (80 x 30 mm) supported on a medium porosity glass
frit. The column was washed with Et20 to completely elute a single red band; then the
column was washed with hexanes (100 mL). The volume of the combined Et20/hexanes solution was reduced in vacuo until crystals were observed to be forming on the sides of the flask. This solution was cooled in the freezer (-25 °C) to complete the crystallization. The solid which deposited from this solution was collected by cannulating away the supernate and washing the remaining red-purple crystals with pentane (20 mL). The crystals were
dried in vacuo to afford the CpW(NO)(Neo)2 product (3.21 g, 59% yield), which was characterized by comparison of its spectroscopic properties with those previously
2 1 J reported: IR (Nujol) i/NO 1561 cm". H NMR (C6D6) s 7.0-7.5 (m, 10 H, Q//5),
2 4.71 (s, 5 H, Cp), 3.56 (d, 2 H, JHH= 11.0 Hz, Ci/aHb), 1.68 (s, 6 H, C(CH3)2),
2 1.63 (s, 6 H, C(CH3)2), -0.97 (d, 2 H, JHH= H.O Hz, CHJ1h).
Preparation of Cp*W(NO)(Neo)2. A solution of Cp*W(NO)(Cl)2 (0.84 g, 2.0 mmol) in DME (30 mL) was cooled in an acetone/Dry Ice bath to -40 °C. A solution of
(Neo)2Mg(dioxane)2 (0.91 g, 223 mg/mmol Neo", 4.0 mmol) in DME (10 mL) was then cannulated into the cooled solution of the dichloride complex, causing an immediate colour change from green to purple. After 20 min, the cooling bath was removed and the mixture dried of solvent in vacuo to leave a red residue. This red residue was dissolved in benzene and transferred to a Florisil column (80 x 20 mm), and a single red-orange band was collected using benzene as eluant. The benzene was removed under reduced pressure to give an orange oil, which was dissolved in hexanes. The volume of the hexanes solution was reduced in vacuo to saturate it, and was then allowed to crystallize in the freezer (-25 °C). Cannulation of the supernatant solution from the product and drying the
remaining solid in vacuo afforded Cp*W(NO)(Neo)2 as red-orange crystalline plates (0.50 g, 40% yield). 19
Anal. Calcd for C^H^NOW: C, 58.54; H, 6.72; N, 2.27. Found: C, 58.37;
1 H, 6.90; N, 2.39. IR (Nujol) i/NO 1551 cm". Low-resolution mass spectrum (probe
+ 3 temperature 120 °C) m/z 615 [P] . *H NMR (C6D6) 6 7.49 (d, 4 H, /HH = 7.5 Hz, o-
ArH), 7.22 (t, 4 H, 3/HH = 7.5 Hz, m-ArH), 7.05 (t, 2 H, 3/HH = 7.5 Hz,/>-ArH), 2.95 (d, 2
2 H, /HH = 11-5 Hz, CtfaHb), 1.72 (s, 6 H, C(CH3)2), 1.65 (s, 6 H, C(Cr73)2), 1.37 (s, 15 H,
2 C5(Ctf3)5), -0.42 (d, 2 H, /HH = U-5 Hz, CH^). "C^H} NMR (C6D6) 6 153.19,
126.35,125.46 (C6H5), 109.73 (C5(CH3)5), 98.93 (CH2), 45.84 (C(CH3)2), 33.77, 32.36
(C(CH3)2), 9.38 (C5(CH3)5).
Preparation of Cp*W(NO)(Npt)2. A turquoise solution of Cp*W(NO)(Cl)2 (4.2 g, 10 mmol) in THF (100 mL) was cooled to 0 °C in an ice water bath. The solution was treated with an aliquot of NptMgCl in Et20 (25 mL, 0.8 N, 20 mmol). The colour changed immediately to a dark purple. The mixture was stirred for 2 h, the cooling bath was removed, and the reaction was treated with 1,4-dioxane (10 mL) which caused the immediate formation of a tan precipitate. The reaction solution was stirred for 10 min, then the solvent was removed in vacuo, and the residue was left to dry overnight under dynamic vacuum.
The dried residue was then suspended in 150 mL of 2:1 hexanes/Et20 and the
resulting red solution/yellow suspension was hydrolyzed with deaerated H20 (2.5 mL) and stirred for 10 min. The red solution was filtered through a column of Florisil (100 x 30 mm) supported on a frit, and a single dark red-brown band was washed off the column
with 2:1 hexanes/Et20 and dried of solvent in vacuo leaving a dark oily residue. The
residue was then dissolved in 2:1 hexanes/Et20 (15 mL) and applied to a silica gel column
(Merck Silica 60, 80-200 mesh, 300 x 15 mm) made up in 2:1 hexanes/Et20. The column was maintained at 10 °C by circulation of tap water through a cooling jacket on the column.
The column was eluted with 2:1 hexanes/Et20 and three different fractions were collected. A small yellow band eluted first and was discarded, followed by a second red band which was collected (-60 mL) until a brownish colouration .began to appear, and then a final green fraction (-200 mL) was then collected. 20
The second red fraction was dried of solvent under reduced pressure, dissolved in pentane, and reduced in volume in vacuo to give a saturated solution (~20 mL) which was
put in the freezer (-25 °C) to induce crystallization. Large red crystals of Cp*W(NO)(Npt)2 (0.365 g) were collected by cannulation of the supernatant solution and washing the remaining solid with pentane (10 mL). The supernatant solution was warmed to room temperature and reduced in volume until it was again saturated. An additional crop of product was collected from the second crystallization (0.100 g), resulting in a 10% overall
yield of Cp*W(NO)(Npt)2.
Anal. Calcd for C2oH37NOW: C, 48.89; H, 7.59; N, 2.85. Found: C, 48.60;
H, 7.63; N, 2.89. Low-resolution mass spectrum (probe temperature 80 °C) m/z 491 [P]+ .
1 X 2 IR (Nujol) i>NO 1553 cm". H NMR (C6D6) 5 2.73 (d, 2 H, JHH= 12.0 Hz, CtfaHb),
2 I. 51 (s, 15 H, C5(CH3)5) 1.34 (s, 18 H, C(CJf3)3), -1.42 (d, 2 H, JHH= 12.0 Hz, CH^fh).
^H) NMR (C6D6) s 109.57 (C5(CH3)5), 95.25 (CH2), 38.97 (Cquat), 34.47 (C(CH3)3),
9.83 (C5(CH3)5). The third fraction was dried of solvent to a green glassy solid. However attempts to purify this solid were not successful.
Preparation of Cp*W(NO)(Ph)2. A suspension of Cp*W(NO)(Cl)2 (2.00 g,
4.76 mmol) in THF (30 mL) was cooled in a saturated CaCl2(aq)/Dry Ice bath (-20 °C). An aliquot of PhMgCl in THF (5.6 mL, 1.75 N, 9.8 mmol) was added to the stirred solution via syringe, causing the green solution to immediately turn dark blue. The reaction mixture was stirred for 30 min and then dried in vacuo to leave a dark blue residue. The residue
was taken up in Et20 (15 mL) and added to the top of a refrigerated (acetone/Dry Ice, -20 °C) Florisil column (100 x 30 mm) supported on frit. The column was washed with
CH2C12 (100 mL) to remove a single red band which gave a blue solution in the receiving flask. Hexanes (100 mL) were added to the blue filtrate, and the solution volume was reduced in vacuo until the supernatant solution appeared dark olive-green in colour. The supernate was quickly removed by cannulation, and the remaining solid washed with
hexanes (3 x 30 mL) to obtain Cp*W(NO)(Ph)2 as a blue solid (0.92 g, 38%). 21
Anal. Calcd for C22H25NOW: C, 52.50; H, 5.01; N, 2.78. Found: C, 52.30; .
H, 5.00; N, 2.66. IR spectrum (THF) = 1588 cm"1. Low-resolution mass spectrum
+ + X (probe temperature 120 °C) 503 [P] , 473 [P-NO] . H NMR (CDC13, -30 °C) 8 7.7-7.81
B 1 (m, 4 H, AiH), 7.16-7.28 (m, 6 H, MH), 1.85 (s, 15 H, 05(0/3)5). C{ H} NMR (CDC13,
-30 °C) 5 136.29,129.96,128.73 (C^J, 10.80 (C5(CH3)5).
Attempts to prepare the CpW(NO)(Ph)2, CpMo(NO)(Ph)2, and Cp*Mo(NO)(Ph)2 analogues via similar methodology were unsuccessful. The desired products appeared to be formed in the reaction mixtures, with distinctive purple to blue solutions being formed
from mixtures of the Cp'M(NO)(Cl)2 complexes and PhMgCl in THF at -20 °C The IR
1 spectra of the THF reaction solutions showed strong bands (CpW(NO)(Ph)2,1601 cm' ;
1 Cp*Mo(NO)(Ph)2, 1613 cm") in the region anticipated for the i/^o of bis(phenyl)
complexes. However, attempts to isolate the Cp'M(NO)(Ph)2 products from these reaction solutions were unsuccessful, resulting in the decomposition of the purple- or blue-coloured solutions to pale yellow-green solutions.
Preparation of Cp*W(NO)(p-ToI)2. A solution of Cp*W(NO)Cl2 (3.0 g, 7.1 mmol) was prepared in DME (50 mL) and cooled in an acetone/Dry Ice bath (-40 °C). A solution
of (p-Tol)2Mg(dioxane)2 (2.5 g, 175 mg/mmol p-To\', 14.3 mmol) was prepared in DME (10 mL) and added to the cooled solution via a cannula tube. The solution colour immediately turned from green to deep blue, and the reaction was stirred for 2 h, during which time the cooling bath warmed up to -10 °C The solvent was then removed under reduced pressure at 10 °C and the remaining purple-blue residue was sequentially extracted
with hexanes (2 x 30 mL), 1:1 hexanes/Et20 (2 x 50 mL), and Et20 (3 x 30 mL). The combined extracts were placed in a freezer (-25 °C). The blue crystals deposited from this solution were isolated by removing the supernate via cannulation. The solid was washed
with hexanes (4 x 20 mL) and dried in vacuo to obtain Cp*W(NO)0?-Tol)2 as blue needles (2.40 g, 63% yield).
Anal. Calcd for C24H29NOW: C, 54.25; H, 5.50; N, 2.63. Found: C, 53.98;
H, 5.52; N, 2.62. IR (Nujol mull) U^Q = 1576 (s) cm"1. Low-resolution mass spectrum 22
+ 3 (probe temperature 100°C) 531 [P] . *H NMR (CDC13) S 7.75 (d, 4 H, /HH = 7.6 Hz,
3 Ar//), 7.09 (d, 4 H, /HH = 7.6 Hz, Ar//), 2.24 (s, 6 H, ArC//3), 1.91 (s, 15 H, C5(C//3)5).
^H} NMR (CDC13) 6 140.80 (Cipso), 135.79,128.10 (C^), 129.42 (Cpara),
112.71 (C5(CH3)5), 21.93 (ArCH3), 10.43 (C5(CH3)5).
The blue, solid Cp*W(NO)(p-Tol)2 changes to a white solid when left in the air overnight. The crude solid appears to consist mainly of Cp*W(0)2(p-Tol), indicated by the
= 1 IR spectrum (Nujol mull, VM=O 941, 898 cm") and low-resolution mass spectrum
+ (probe temperature 120°C, highest m/z 442 [Cp*W(0)2(p-Tol)] ) of this white solid. 23
Results and Discussion
Preparation of the Cp'M(NO)(Cl)2 Complexes. The established route to the dihalonitrosyl complexes is via reaction of the dicarbonyl precursors with the appropriate elemental halogen.
Cp'M(NO)(CO)2 + X2 • Cp'M(NO)X2 + 2 CO (2.3) (Cp'= Cp; M= W; X= I)16 (Cp'= Cp*;M= W;X= I)17 (Cp'= Cp*; M= Mo; X= I)18 (Cp'= Cp; M= Mo; X= I, CI, Br)19
This methodology has been successful for the production of the complexes indicated in
reactions 2.3. A need for the Cp'W(NO)(Cl)2 compounds as starting materials arose (vide infra), which were unknown before this study, so their efficient preparation has been pursued. However, when the route in equation 2.3 is extended to the reaction of the tungsten dicarbonyl complexes with chlorine, the result is less than ideal. Addition of one
equivalent of chlorine to a solution of CpW(NO)(CO)2 in CH2C12, followed by addition of hexanes, results in the precipitation of a brown powder. The IR spectrum of the brown
solid shows that it contains CpW(NO)(Cl)2 as the major product (when compared to the IR
spectrum of an authentic sample of CpW(NO)(Cl)2) but the brown solid is contaminated by additional nitrosyl-containing byproducts. Therefore this route is undesirable for the preparation of dichloro starting materials since additional purification steps would be necessary. Alternative chlorinating reagents were investigated before attempting to improve the reaction of chlorine with the dicarbonylnitrosyl complexes because gaseous chlorine is also very inconvenient to handle.
The successful preparation of the dichloronitrosyl complexes results from the 24 reaction of PCI5 with the dicarbonylnitrosyl precursors, i.e.,
Et20
Cp'M(NO)(CO)2 + PCI5 • Cp'M(NO)(Cl)2 + PC13 (2.4)
-2 CO This is an excellent synthetic route to the dichloro complexes, characterized by very high yields (85-95 %) and consistently good results. The reaction scale has been varied from 1 to 50 mmol with equal success, a factor which is crucial in the preparation of these species as they are primary starting materials in this research group, supplanting the diiodo analogues used previously. The products of equations 2.4 are moderately air-stable powders, orange in the case of the molybdenum complexes and brown in the case of the tungsten complexes. They are slightly soluble in common polar organic solvents giving emerald-green to turquoise-coloured solutions, with the tungsten complexes being noticeably more soluble than the molybdenum complexes. This solubility difference is the suspected reason for the lower yields of the tungsten complexes, vis-a-vis the molybdenum complexes, since more of the tungsten products are dissolved and lost during the filtration and washing of the products. The convenience and efficiency of the preparations of the dichloronitrosyl complexes via equations 2.4 can be attributed to a number of salient features of the reaction. Firstly, PCI5 is a solid which allows careful control of the stoichiometry by simply
weighing the reagent. Secondly, both the PCI5 and the Cp'M(NO)(Cl)2 products are very
insoluble in Et20, so there is never a large concentration of either present in solution. This presumably minimizes the possibility for further reaction of the products with PCI5, as may
be the case with Cl2 (vide infra). Thirdly, the phosphorus trichloride byproduct is unreactive with the product complexes under the reaction conditions used and it is very
soluble in Et20 so its removal is easily accomplished. Since it has been established that
20 CpMo(NO)(Cl)2 readily forms 1:1 adducts with electron-rich trialkyl phosphines, the lack of such adducts with PCI3 must be due to the poor donor properties of PCI3. 25
The observation of other products in the reaction of CpW(NO)(CO)2 with CI2, when compared to the successful preparation of the diiodide complexes as in equation 2.3, can be correlated with the reactivity of the halogen used. In the optimized halogenations of CpMo(NO)(CO)219 in equation 2.3, other nitrosyl-containing products are sometimes noted in the preparations using bromine and chlorine, but no byproducts are observed in the preparation using iodine. The authors attributed the problem to the purity of the CpMo(NO)(CO)2 starting material, but the problem could just as easily have resulted from the reaction of the CpMo(NO)(X)2 products with the halogen reagent. As well, the trend
in yields of the CpMo(NO)(X)2 products, I > Br > CI, are opposite to the trend in reactivity of the halogens used. The reactivity of the halogens can be put on a relative scale by considering their oxidation potentials in acetonitrile,21
E1/4atPt
X"/X2 (Vvs.NHE)
cr/ci2 1.15 '
Br-/Br2 0.70
r/I2 0.20 which show that I2 is a much weaker oxidizing reagent and therefore less likely to oxidize the products. A cyclic voltammetric study of the Cp'Mo(NO)(X)2 complexes has shown the presence of chemically irreversible oxidation processes.22 The use of phosphorus trihalides as alternatives to halogens in the synthesis of halide derivatives of early transition-metal organometallics has been reported.23 This route to form CpW(NO)(Cl)2 has also been attempted during this study, but PCI3 fails to react with the CpW(NO)(CO)2 complex. Apparently PCI5 is a more powerful chlorinating agent than PCI3, but less reactive than elemental chlorine in these syntheses. Phosphorus pentachloride has rarely been utilized as a chlorinating agent for organometallics. A survey of Chemical Abstracts under the heading of PCI5 from 1962 to 1988 reveals only one 26 citation where this reactivity had been exploited,24 namely CH2CI2
Cp*M(CO)3Me + PCI5 » Cp*M(Cl)4 (2.5)
M = Mo, W
The IR and NMR spectroscopic properties observed for the Cp'M(NO)(Cl)2 species are similar to those previously reported for the iodo analogues. A characteristic shift in the V^Q of approximately 30 cm"1 to lower energy is noted in going from the Cp to the Cp complexes, due to the greater electron-donating capacity of the Cp ligand. As
1 well, a 35 cm" difference between Mo and W complexes is noted, the lower i/NO values for the W species being a common observation for group 6 nitrosyl complexes.26 The 1H and 13C{1H} NMR chemical shifts of the Cp and Cp* ligands are as expected for these complexes.
Mass Spectroscopic and Solid-State X-Ray Crystallographic Investigation of the
Molecular Structures of [Cp'M(NO)(Cl)2]n (n= 12). Although reports of the
27 Cp'Mo(NO)(X)2 compounds have been in the literature since 1967, the recently reported
17 solid-state molecular structure of Cp*W(NO)(I)2 (vide infra) is the first to be determined for this general family of dihalonitrosyls. The X-ray crystallographic structural analysis of
28 Cp*Mo(NO)(Cl)2 has subsequently been carried out as part of these studies. A
SNOOPI29 plot of the solid-state molecular structure is presented in Figure 2.1 and important structural parameters are collected in Table 2.1. The most important structural feature is that the molecule is a symmetric dimer, possessing a centre of inversion. The general metrical parameters of the Cp*-Mo and Mo-NO groups are not exceptional and compare with those of other Cp*Mo(NO) containing complexes that have been structurally characterized.30 The Mo-Cl-Mo' bridging units are slightly asymmetric as indicated by the
Mo-Clbndgiug bond lengths (2.559(2), 2.495(2) A), however, both of these bonds are
significantly longer than the Mo-Clterminai bond length (2.406(2) A). 27
Figure 2.1. The SNOOPI plot of the solid-state molecular structure of [Cp*Mo(NO)(Cl)2]2 with the hydrogen atoms omitted for clarity. 28
Table 2.1. Selected bond distances (A) and bond angles (°) for the molecular structure of
[Cp*Mo(NO)(Cl)2]2 at 200 K.
Mo(l) - CF 2.035 A Mo(l) - N(l) - O(l) 169.7(4) ° Mo(l) - N(l) 1.781(5) Cl(l) - Mo(l) - Cl(l)* 76.00(6) N(l)- O(l) 1.174(7) Cl(l) - Mo(l) - Cl(2) 79.14(7) Mo(l) -Cl(l) 2.559(2) Mo(l) - Cl(l) - Mo(l)' 104.00(6) Mo(l) - Cl(l)' 2.495(2) N(l)- Mo(l) - Cl(l) 126.3(2) Mo(l) -Cl(2) 2.406(2) N(l)- Mo(l) - Cl(2) 86.1(1)
CP = centre of gravity of the C^Mes ring. ' denotes the symmetry related atom at -x, -y, -z. 29
The slight asymmetry may be a manifestation of the bridging chlorine being covalently bonded to one molybdenum atom while acting, in a formal sense, as a Lewis base to the other molybdenum atom. In any event, the markedly longer Mo-CUjridging bond's are indicative of weak interactions to form the dimer. This feature will become important in the discussion of the mass spectra of the Cp'M(NO)(Cl)2 complexes (vide infra). The
dimeric nature of [Cp*Mo(NO)(Cl)2]2 in the solid state is in contrast with the other structurally-characterized member of the family, namely Cp*W(NO)(I)2, which is a
16-electron monomer in the solid state.17
The family of complexes Cp'M(NO)(X)2 has generally been formulated as consisting of halide bridged dimers as in A below
A B
in order to satisfy the 18-electron rule. However, the only physical evidence proffered for
+ this formulation has been the mass spectroscopic observation of [CpM(NO)(X)]2
dimetallic ions for these complexes.20'31 These dimeric ion species are believed to be
formed by the loss of 2 halogen atoms in the mass spectrometer from a parent dimer as in
A. However, prior to this work no structural studies had been reported to confirm these
predictions. The solid-state molecular structure of Cp*W(NO)(I)2 reveals the complex to
be a 16-electron monomer as in B, and a molecular weight determination in CH2Cl2
indicates that Cp*W(NO)(I)2 in solution is also in the monomeric form B.17 This fact is of
interest when the mass spectrum exhibited by Cp*W(NO)(I)2 is compared to the mass
spectra of the dichloro complexes (Table 2.2). Essentially, all of these compounds exhibit
qualitatively similar mass spectra, even though Cp*W(NO)(I)2 is definitely monomeric and
+ [Cp*Mo(NO)(Cl)2]2 is dimeric. The observation of [Cp*W(NO)(I)]2 mass peaks would 30
Table 22. Low-Resolution Mass Spectral Data for the Dihalo Complexes.
Compound m/z" Intensity'' Assignment4"
+ CpMo(NO)(Cl)2 453 0.1 [2P-2C1]
423 0.1 [2P-2C1-NO]"1"
263 26.0 [P]+
233 100.0 [P-NO]+
+ CpW(NO)(Cl)2 351 273 [P]
321 100.0 [P-NO]+
+ Cp*Mo(NO)(Cl)2 563 0.1 [2P-2Cl-NO]
333 27.1 [P] +
303 80.1 [P-NO]+
267 100.0 [P-NO-Cl]+
+ Cp*W(NO)(Cl)2 738 0.3 [2P-2Cl-NO]
421 12 [P]+
+ 405 23 [P-CH4]
391 35 [P-NO]+
+ Cp*W(NO)(I)2 952 0.8 [2P-2I]
922 0.7 [2P-2I-NO]+
603 . 100.0 [P] +
573 75.9 [P-NO]+ a All spectra recorded at a probe temperature of 120 °C and m/z are reported for the most abundant mass of the isotopic cluster. b Relative percentage of the ion current of the most abundant mass of the isotopic cluster.
c P relates to the respective monomeric Cp'M(NO)(X)2 species. 31 seem to imply that ion-molecule reactions of monomeric Cp*W(NO)(I)2 and derived species in the mass spectrometer are possible sources of these dimeric mass ions.Thus, the presence of these dimeric peaks in no way confirms the degree of association of the complexes in the solid state. Similar dimeric ion-molecule products have been observed in
32 the mass spectra of the monomeric CpMo(NO)(SPh)2. The long Mo-Clt,ridging bonds in
[Cp*Mo(NO)(Cl)2]2 would seem to indicate that the associations of all of the dihalo complexes in the solid state, if present, are weak. Therefore, under the conditions of the mass spectroscopic analysis these bonds could be easily broken to form Cp'M(NO)(Cl)2 as the major species in the gas phase, which would explain the qualitative resemblance of the
mass spectra of the dichloro complexes to the mass spectrum of Cp*W(NO)(I)2. The largest features of the mass spectra of the former are consistent with this, being due to the
[Cp'M(NO)(Cl)2]+ species and fragments derived from them. At present, it would seem that the only qualitative indication of the degree of association of the dihalonitrosyls is their colour. The monomeric Cp W(NO)(I)2 is a
17 distinctive green colour while the dimeric [Cp*Mo(NO)(Cl)2]2 forms red crystals or orange powders. All of the dichloro complexes are orange to brown powders, indicating
that they are also probably dimers in the solid state similar to [Cp Mo(NO)(Cl)2]2.
However, since all the dichloro complexes are green to turquoise in solution, it is believed
that they are monomers in solution similar to Cp*W(NO)(I)2. Unfortunately, none of the
Cp'M(NO)(Cl)2 complexes are sufficiently soluble to allow solution molecular weight determinations in non-coordinating solvents, but their reactivity and electrochemical properties (vide infra) are most consistent with their formulation as monomers (possibly solvated) in solution. Therefore, when used as reagents in equations they will be represented as monomers.
Alkylations of Cp*W(NO)(Cl)2 With Grignard Reagents; Improved Synthesis of
CpW(NO)(R)2 (R= Npt, Neo) and Preparation of Cp*W(NO)(R')2 (R= Neo, Npt, Ph, p-To\). The dichlorides have been prepared with the expectation that they would be superior starting materials for the preparation of Cp'M(NO)(R)2 complexes. This 32 hypothesis is due to an earlier investigation of the reduction behaviour of the
7 Cp'Mo(NO)(X)2 (X= CI, Br, I) species. The cyclic voltammograms of the Cp'Mo(NO)(X)2 complexes in CH2CI2 all show a reversible reduction at easily accessible potentials (0.0 to -0.3 V vs. SCE). This previous investigation also showed that at slower
scan rates, follow up reactions of the reduced species to form [Cp'Mo(NO)(X)]2 dimers occur. Chemical reduction can also be performed using CP2C0 to afford the
[Cp2Co]+[Cp'Mo(NO)(X)2j" radical anion salts whose relative stability is found to be
CI > Br > > I. Since a 1:1 solution of Cp*Mo(NO)(Br)2 and TmsMgCl in DMF gives an
+ ESR signal identical to that observed for a DMF solution of [Cp2Co] [Cp*Mo(NO)(Br)2]"" it is reasoned that the metathesis reactions to form the bis(alkyl) complexes could very well proceed through radical anion intermediates.33 Therefore, it is hypothesized that the preparation of bis(alkyl) complexes should be designed to maximize the stability of these . proposed radical anion intermediates. This implies the use of the dichloronitrosyl starting materials, coordinating solvents to help stabilize the intermediates, and low temperatures to avoid possible decomposition pathways. Application of these ideas led to the successful
34 synthesis of Cp*Mo(NO)(Tms)2 (equation 2.6), i.e.,
Cp*Mo(NO)(Cl)2 + 2 TmsMgCl • Cp*Mo(NO)(Tms)2 + 2 MgCl2 (2.6)
Cp*Mo(NO)(I)2 + 2 TmsMgCl • [Cp*Mo(NO)(I)]2 (2.7) and also helped to explain the formation of [Cp*Mo(NO)(I)]2 as the major product (equation 2.7) when the diiodide precursor is reacted with the trimethylsilylmethyl Grignard reagent. The utility of the dichloro precursors is also realized in the synthesis of diene-containing complexes via metathesis with magnesium-diene reagents. Attempts to form a butadiene complex from CpMo(NO)(I)2 in this fashion result in the eventual formation of the desired product, but the 6 d reaction time (equation 2.8)6 33 is inconvenient for synthetic use.
THF
CpMo(NO)(I)2 + (C4rLj)Mg(THF)2 CpMoCNOX^-C^ (2.8)
6d
The reactions of the CpMo(NO)(Cl)2 and Cp*Mo(NO)(Cl)2 complexes with the same magnesium-diene reagent (equations 2.9) result in formation of the desired diene complexes in 40% isolated yields, but within hours instead of days.35
THF
4 CpMo(NO)(Cl)2 + (C4H6)Mg(THF)2 • CpMo(NC% -C4H6) (2.9)
3-6h
The success of reactions 2.6 and 2.9 thus establishes the dichloro complexes as superior starting materials in reactions involving metathesis of the halide ligands.
Another advantage to the use of the dichlorides is the ease of MgCl2 removal.
This is partially due to the lower solubility of the MgCl2 product as compared to the partially iodonated salts, e.g. MglCl, formed from alkylation of the diiodide complexes. As
well, the MgCl2 does not seem to form strong isonitrosyl linkages as often occurs during the preparations of bis(alkyl) complexes from diiodide starting materials, e.g.,2
Et20
CpW(NO)(I)2 + 2 TmsMgCl • [CpW(NO)(Tms)2]2MgI2-Et20 (2.10)
Use of the CpW(NO)Cl2 starting materials leads to improved yields (60-70%) of
CpW(NO)(Neo)2 and CpW(NO)(Npt)2 from those reported for their syntheses via the diiodide precursors (35-55%).2
The new complexes Cp*W(NO)(Npt)2 and Cp*W(NO)(Neo)2 are also prepared
via this route (10-40 % yield). Since small amounts of impurities in Cp W(NO)(Npt)2 or
Cp*W(NO)(Neo)2 result in oily products, they are crystallized from aliphatic solvents in 34 order to obtain tractable solids. As anticipated, these new Cp*-containing compounds are more soluble than their Cp analogues. Thus, the low yields of the Cp* compounds are most likely due to this increased solubility, making crystallization of the complexes difficult. As well, the steric bulk of the Npt and Neo alkyls coupled with the increased steric bulk of the
Cp* ligand may make alkylation of Cp*W(NO)(Cl)2 with these Grignard reagents more difficult.
The ^-H NMR, ^C^H} NMR and mass spectroscopic parameters for the Cp* bis(alkyl) complexes are comparable to those of their perhydro analogues and correspond to their formulation as 16-electron monomeric species. Most notable is the very distinctive coupling between the two protons of the methylene groups of the alkyl ligands in the
Cp*W(NO)(R)2 complexes. This coupling is due to the diastereotopic disposition of these two protons, i.e.,
1 which results in the inequivalence of the Ha and protons. The H NMR spectrum of
Cp*W(NO)(Npt)2 presented in Figure 2.2 displays this phenomenon, which is also observed for the other Cp'M(NO)(CH2R)2 complexes that have been synthesized.2
The i/jsjo bands in the IR spectra of the Cp* alkyl complexes are 10 cm"1 lower in energy than their Cp analogues, a feature attributable to the presence of the more electron- donating Cp* ligand. These V^Q bands are also very much lower than those seen for the corresponding dichloro complexes, a reflection of the larger electron donation from an alkyl ligand than an electronegative chloro ligand.
Since the use of the dichloro compounds results in the improved syntheses of the bis(alkyl) complexes, they were next used to prepare the Cp'M(NO)(aryl)2 species (equation 2.11), whose synthesis had been unsuccessful starting from the diiodide precursors.36 35
C(CH3h C5(C/f3)5
C#aHb I JVA_ 0 PPM
J Figure 22 The 80MHz H NMR Spectrum of Cp*W(NO)(Npt)2 in C6D6. 36
THF or DME
Cp*W(NO)(Cl)2 + ArMgCl (or Ar2Mg) • Cp*W(NO)(Ar)2 + MgCl2 (2.11)
-20 °C
The bis(aryl) complexes are obtainable as analytically pure solids in 30-60% yield. The initially blue reaction mixtures may, on occasion, spontaneously decompose in a matter of minutes, leaving a green-yellow solution. The nature of the decomposition is unknown. The pure bis(aryl) complexes can be weighed and transferred in air as solids but overnight
exposure of solid samples of Cp*W(NO)(p-Tol)2 to air results in bleaching of the blue
colour and formation of a white solid containing the Cp*W(0)2(p-Tol) dioxo complex, in a manner analogous to that observed previously for the bis(alkyl) complexes.34
02
Cp*W(NO)(p-Tol)2 • Cp*W(0)2(>Tol) (2.12)
Overall, the bis(aryl) complexes would appear to be less thermally stable than their
respective bis(alkyl) analogues, decomposing at -20°C under N2 over a period of weeks.
The bis(aryl) complexes are soluble in most common organic solvents, but to a noticeably lesser extent than their bis(alkyl) analogues. The /7-tolyl complex is somewhat soluble in hexanes but the phenyl complex is only sparingly so, the solubility difference being due to the methyl substituents on the former complex.
The spectroscopic properties of the bis(aryl) complexes are consistent with a 16-
electron configuration at the metal as seen for the bis(alkyl) complexes. The vNO values for the bis(aryl) complexes are in the range anticipated for a tungsten bis(alkyl) although
1 they are -25 cm" higher in energy than the corresponding Cp*W(NO)(R)2 (R= Npt, Neo)
a-alkyl complexes. The vNO values observed for the bis(aryl) complexes correlate with the
expected larger electron-withdrawing effect of an aryl group compared to an alkyl group, leading to a reduction of the W-NO jr-backbonding by increased competition for the
tungsten electron density by the two aryl groups. The 1H and 13C{1H} NMR spectra of 37 these bis(aryl) complexes are not unusual and are consistent with their formulation indicated above. The 1H NMR spectrum of the p-to\y\ group is distinctive in that it exhibits an AA'BB' coupling pattern due to the para-disubstituted aromatic ring.
Attempts to prepare other members of this class of complexes by reaction 2.11 indicate that the desired bis(phenyl) products are formed in solution. This conclusion is
m e reached by the observation of a single I>NO *h ^ spectra of the reaction solutions at values consistent with the formation of bis(aryl) complexes. As well, the reaction solutions are a very distinctive blue to blue-purple colour which is associated with this class of bis(aryl) and bis(alkyl) complexes. However, all attempts to isolate the products from these solutions have been unsuccessful, the decomposition of the products being indicated by the loss of the characteristic blue colour of the reaction mixtures.
In light of the stable CpW(NO)(alkyl)2 and Cp*W(NO)(alkyl)2 complexes that
have been prepared, the inability to isolate a CpW(NO)(Ar)2 complex given the isolation of its Cp* analogues is unexpected. It may be that a planar aromatic ligand is less sterically demanding than a bulky alkyl ligand, and consequently the aryl group does not shield the unsaturated metal centre as well as the alkyls do. The additional steric shielding of the Cp* ligand versus the Cp ligand may provide the small additional factor necessary to render the
Cp*W(NO)(aryl)2 product complexes stable enough to isolate. This feature may also be
the reason for the inability to isolate the less sterically hindered Cp'M(NO)(Me)2 complexes, although solutions of these species can be generated in situ at -60 °C and
34 reacted with H202 to form the corresponding Cp'M(0)2(Me) products. In a more general sense, it is not too surprising that the Cp* complexes are more stable than the Cp analogues, since this trend in thermal stability has been observed in other systems and is a major impetus for the synthesis of the Cp* analogues of Cp complexes.37
Reduction Behaviour of CpW(NO)(Neo)2, CpW(NO)(Npt)2, and
Cp*W(NO)(p-Tol)2. Electrochemical investigations of the Cp'W(NO)(Tms)2 complexes showed that they are reduced reversibly.22 Attempts to isolate a radical anion species via 38 chemical reduction, e.g.,
Et20
6 CpW(NO)(Tms)2 + CpFe^-QMeg) *> [CpFe(r, -C6Me6)][CpW(NO)(Tms)2] (2.13)
were not successful due to the intractable nature of the product formed.
The electrochemical behaviour of the CpW(NO)(Npt)2, CpW(NO)(Neo)2 and
Cp*W(NO)(p-Tol)2 complexes in THF has been investigated with regards to their reduction behaviour. All three of the complexes investigated show similar qualitative
features, with the cyclic voltammogram of CpW(NO)(Neo)2 (Figure 2.3) being representative of the cyclic voltammograms seen for reductions of the complexes. These cyclic voltammograms are consistent with the occurrence of two reversible, one-electron reductions as depicted in equation 2.14.
+e" +e"
2 Cp'W(NO)(R)2 ^—^ [Cp'W(NO)(R)2]- [Cp'W(NO)(R)2] - (2.14)
-e" -e"
Cp'= Cp,R= Neo, Npt
Cp'= Cp*,R=/?-Tol
The electrochemical data for all three compounds are collected in Table 2.3. The linearity of plots of for all of these reduction waves show that the reductions are diffusion controlled. Comparison of the AE values for these reductions with the
+ internal Fc/Fc reference indicate that the first reductions of CpW(NO)(Npt)2,
CpW(NO)(Neo)2 and Cp*W(NO)(p-Tol)2 are electrochemically reversible. The *p>aAp)C values for the first reduction of all three complexes are unity over the range of scan speeds used, showing that the reductions are also chemically reversible. The first reduction potentials for the bis(alkyl) compounds are similar, the slight change in alkyl between Neo
and Npt causing small changes in£° j, i.e., CpW(NO)(Npt)2 (£° i = -1.46 V) and
0 CpW(NO)(Neo)2(iE '1=-1.41 V). 39
Figure 2.3. Ambient temperature cyclic voltammogram of CpW(NO)(Neo)2 in THF
1 (0.1 N [/i-Bu4N]PF6, Pt working electrode) at a scan rate of 0.56 V s". 40
Table 23 Data for the Reductions of the Cp'W(NO)(R)2 Complexes."
C d Complex Scan rate B AE E--2 *p,a/'p,c .Vs'1) (V) (V) (mV)
CpW(NO)(Neo)2 0.30 -1.41 90(91) 1
1.20 -1.41 130(131) . 1 0.58 -2.63 240 ~1
CpW(NO)(Npt)2 0.40 -1.47 130(135) 1
0.80 -1.47 150(155) 1 0.58 -2.72 220 ~1
Cp*W(NO)(p-Tol)2 0.20 -1.20 80(80) 1
0.80 -1.20 130(130) 1 0.80 -2.45 140 ~1
In THF containing 0.1 M [n-Bu4N]PF6, at Pt-bead working electrode. Potentials
are measured vs. SCE.
Defined as the average of the cathodic and anodic peak potentials.
Defined as the separation of the cathodic and anodic peak potentials. Values of
AE given in brackets are for the Fc/Fc+ couple under the same conditions.
Ratio of anodic peak current to cathodic peak current. 41
5 However, the Cp*W(NO)(p-Tol)2 complex is easier to reduce by -250 mV (F '^ -1.20 V)
than these related bis(alkyl) species. This result is consistent with Cp*W(NO)(p-Tol)2
containing a less electron-rich metal centre than the CpW(NO)(alkyl)2 complexes. This
interpretation also correlates with the higher vNO value displayed by Cp*W(NO)(p-Tol)2
(1576 cm"1) compared with the bis(alkyl) complexes (~ 1560 cm"1). This difference in the electron density on the respective tungsten centres can be related to the relative electron
donating capability of an alkyl ligand versus an aryl ligand. From pKa measurements of
substituted benzoic acids, it can be shown that a CH3 group is a better electron donor than
38 is a C6H5 group. The relative electron deficiency of Cp*W(NO)(p-Tol)2 relative to the bis(alkyl) species is even more pronounced when it is noted that this species contains an electron-donating Cp* ligand while the bis(alkyl) complexes contain a Cp ligand. In the
related Cp'W(NO)(Tms)2 compounds, the Cp*-containing complex is more difficult to reduce by —130 mV than is the Cp analogue, due to the extra electron density donated by the Cp ligand.'
For all three complexes a second reduction is observed between -2.4 and -2.7 V vs.
SCE. Again, Cp*W(NO)(/7-Tol)2 is easier to reduce by -250 mV than are the bis(alkyl) complexes. The second reduction (equation 2.14) of each of these complexes also appears
to be chemically reversible, with *p)aAp)C values estimated to be unity. Unfortunately, accurate determination of this ratio is not possible, since the reductions occur very close to the solvent limit for the THF/TBAH electrolyte system. The second reduction for both of the bis(alkyl) complexes exhibit an increased AE separation, which is indicative of electrochemical quasireversibility39 in the electron transfer step. The AE for the second reduction of the bis(aryl) complex indicates an electrochemically reversible reduction process is occurring. This difference in the electrochemical reversibility for the second reduction is indicative of a slower electron transfer step for the bis(alkyl) complexes as compared to the bis(aryl) complex, but a detailed examination of the processes involved is beyond the scope of this investigation. This second reduction is not seen in the reduction
7 behaviour of the related Cp'W(NO)(Tms)2 complexes since the reduction potentials 42
apparently would lie outside the range available with the CH2C12/TBAH electrolyte system used in that study.
Implications from the Electronic Structure of CpMo(NO)(Me)2 for the
Electrochemical and Electronic Properties of Cp'M(NO)(X)2 (X= alkyl, halide)
Complexes. Fenske-Hall MO calculations have been performed on the model complexes,
CpM(NO)(Me)2 (M=Mo, Ru), to try to understand the electronic nature of the
16-electron bis(alkyl) species.1 The major finding of the study is that the LUMO for the
Mo species (HOMO of the Ru species) is metal-centred and non-bonding. Therefore, filling of this orbital should not substantially stabilize the existing bonding framework.
Conversely, the filling of this orbital should not destabilize the bonding framework to any substantial extent either.
The reduction behaviour of the Cp'W(NO)(R)2 species discussed in the previous
section correlates very well with MO picture of the CpMo(NO)(Me)2 model complex from the Fenske-Hall MO calculations (Scheme 2.2).
5P
Scheme 22 43
The observation of two one-electron reductions, corresponding to filling of the
LUMO, results in stable reduced species as indicated by the iptJiP)C values being unity for these processes. This is in accord with the non-bonding nature of the LUMO predicted from the calculations. Furthermore, the MO picture in Scheme 2.2 is useful in giving a framework for rationalizing the qualitative colour differences seen for various bis(alkyl)nitrosyl complexes. As can be seen in Scheme 2.2, the HOMO-LUMO energy gap for the 16-electron Mo bis(alkyl) is smaller than the same gap for the 18-electron Ru bis(alkyl) complex. This is reflected in a low energy HOMO-LUMO transition for all of
the Cp'Mo(NO)(R)2 and Cp'W(NO)(R)2 complexes prepared to date, a transition which is responsible for their distinctive violet or blue colours. In contrast, the 18-electron
4 Cp*Ru(NO)(Me)2 complex is red-orange in colour, consistent with a higher energy HOMO-LUMO transition absorbing higher energy wavelengths. This trend in colours of the 16- versus 18-electron complexes containing the Cp'M(NO) core seems to be quite
2 general. The 18-electron complexes CpW(NO)(Tms)2(PMe3) and Cp'M(NO)(Bz)2
(M = Mo, W)40 are yellow and orange respectively, as would be predicted from this argument.
The MO picture of CpMo(NO)(Me)2 appears to be a reasonably good model of
the CpMo(NO)(X)2 species as well. Since these dihalo complexes are also easily reduced reversibly (E°' = 0.0 to -0.3 V) to form radical anion species, this implies that they also contain a low-lying metal-centred LUMO as predicted for the bis(alkyl) complexes. As
discussed earlier, the orange or red colours of the Cp'M(NO)(Cl)2 species as solids are associated with 18-electron metal centres, i.e. halide-bridged dimers, while the turquoise to green colours of solutions of these complexes are taken as characteristic of monomeric 16-electron metal centres. These colours of the bis(chloro) compounds correlate with 16-electron (blue-green, small HOMO-LUMO gap) and 18-electron (red-orange, large HOMO-LUMO gap) metal centres by arguments similar to those outlined above for the bis(alkyl) complexes, lending additional support to the view that the MO picture of the bis(alkyl) species is qualitatively valid for the bis(chloro) complexes as well. 44
Epilogue.
Efficient and convenient syntheses of the Cp'Mo(NO)(Cl)2 complexes and the previously unreported Cp'W(NO)(Cl)2 complexes have been developed. The superiority of these species for the preparation of various complexes of group 6 nitrosyl complexes via
metathetical reactions, predicted from the reduction behaviour of the Cp'Mo(NO)(Cl)2 complexes,7 has been confirmed and utilized in the preparation of the new complexes
Cp*W(NO)(Neo)2, Cp*W(NO)(Npt)2, Cp*W(NO)(Ph)2, and Cp*W(NO)(p-Tol)2. The observation of two chemically reversible reductions during cyclic voltammetry
studies of CpW(NO)(Neo)2, CpW(NO)(Npt)2 or Cp*W(NO)(p-Tol)2 would indicate that
the successful isolation of stable complexes of the type [Cp'W(NO)(R)2]'" and
2 [Cp'W(NO)(R)2] " should be possible. Chemical reduction of the bis(alkyl) and bis(aryl) complexes has not been attempted as yet, but the outlook for success is excellent. The lack
of success observed for the isolation of [CpW(NO)(Tms)2]'" by reduction with
6 CpFe(r7 -C6Me6) may indicate that an alternative reducing agent is necessary to give tractable products. The E° values can serve as a guide to the choice of appropriate
41 reducing agents, sodium napthalenide (E1/2 = -2.50 V vs. SCE) being one possibility.
These reduced species would be very interesting as they should be nucleophilic and according to their MO description the charge should be concentrated at the metal centre.
This may be a route into 17-electron radical neutral complexes via addition of electrophiles such as Mel to the radical anions. The structures of the mono- and dianions would also be interesting since the MO picture above would indicate that the neutral bis(alkyl) compounds and their reduced forms should be essentially isostructural. 45
References and Notes
Legzdins, P.; Rettig, S.J.; Sanchez, L.; Bursten, B.E.; Gatter, M.G. /. Am. Chem.
Soc. 1985,107,1411.
Legzdins, P.; Rettig, S.J.; Sanchez, L. Organometallics 1988, 7, 2394.
Legzdins, P.; Sanchez, L. /. Am. Chem. Soc. 1985,107, 5525.
Seidler, M.D.; Bergman, R.G. /. Am. Chem. Soc. 1984,106, 6110.
Chang, J.; Bergman, R.G. /. Am. Chem. Soc. 1987,109, 4298.
Christensen, N.J.; Hunter, A.D.; Legzdins, P. Organometallics 1989,8, 930.
Herring, F.G.; Legzdins, P.; Richter-Addo, G.B. Organometallics 1989, 8,1485.
Shriver, D.F.; Drezdzon, M.A, The Manipulation of Air-Sensitive Compounds, 2nd
ed.; Wiley-Interscience: New York, NY, 1986.
Hoyano, J.K.; Legzdins, P.; Malito, J.T. Inorg. Synth. 1978,18, 126.
The Cp*Mo(NO)(CO)2 complex was prepared analogously to the published procedure for the W congener: Dryden, N.H.; Legzdins, P.; Einstein, F.W.B.;
Jones, R.H. Can. J. Chem. 1988, 66, 2100.
Legzdins, P.; Wassink, B. Organometallics 1984,3,1811.
Anderson, R.A.; Wilkinson, G. /. Chem. Soc, Dalton Trans. 1977, 809.
Schrock, R.R.; Fellmann, J.D. /. Am. Chem. Soc. 1978,100, 3359. An ethereal solution of NeoMgCl was prepared by slow addition of neophyl
chloride (18 mL, 111 mmol) to a suspension of Mg turnings (4.0 g, 163 mmol,
activated by addition of 0.5 mL of 1,2-dibromoethane) in Et20 (250 mL) over 1 h.
The reaction mixture was stirred for 2 h after the addition was complete and
filtered through Celite to remove excess Mg. Aliquots (1.0 mL) of the Grignard
reagent were hydrolyzed and found to be 0.4 N by titration with 0.100 N HC1 using
phenolphthalein as indicator.
Ramsden, H.E.; Balint, A.E.; Whitford, W.R.; Walburn, J.J.; Cserr, R. /. Org.
Chem. 1957,22,1202. 46
Hunter, A.D.; Legzdins, P.; Martin, J.T.; Sanchez, L. Organomet. Synth. 1986,3,
58.
Dryden, N.H.; Legzdins, P.; Einstein, F.W.B.; Jones, R.H. Can. J. Chem. 1988, 66,
2100.
Nurse, CR. Ph.D. Dissertation, The University of British Columbia, 1983.
Seddon, D.; Kita, W.G.; Bray, J.; McCleverty, J.A. Inorg. Synth. 1976,16, 24.
McCleverty, J.A.; Seddon, D. /. Chem. Soc, Dalton Trans. 1972, 2526.
Technique of Electroorganic Synthesis; Weinberg, N.L., Ed. in Methods and
Techniques in Organic Chemistry; Weissberger, A., Series Ed.; John Wiley and
Sons: New York, NY, 1985; Vol. 5, Part 2, p 3.
Richter-Addo, G.B. Ph.D. Dissertation, The University of British Columbia, 1988.
Moran, M. Transition Met. Chem. 1981, 6, 42.
Murray, R.C; Blum, L.; Liu, A.H.; Schrock, R.R. Organometallics 1985,4, 953.
Robbins, J.L.; Edelstein, N.; Spencer, B.; Smart, J.C. /. Am. Chem. Soc. 1982,104,
1882.
Hunter, A.D.; Legzdins, P. Organometallics 1986, J, 1001.
King, R.B. Inorg. Chem. 1967, 6, 30.
The structure was determined by Dr. RJ. Batchelor of Simon Fraser University on an Enraf-Nonius CAD4F diffractometer, in the laboratory of Professor
F.W.B. Einstein. X-ray diffraction data for [Cp*Mo(NO)(Cl)2]2:
Temperature: 200 K; Monoclinic; Space group P2j/c; a = 8.917(2) A; b = 8.449(2) A; c = 18.118(4) A; 0 = 110.17(2)°; V= 1281.2 A3; n (Mo Ka) = 13.96 cm-1; Scan range = 5° < 29 < 52°; No. reflections = 1966 with
F0> 5a (F0); R = 0.021; Rw = 0.025.
Davies, E.K. SNOOPI plot program. Chemical Crystallography Laboratory,
University of Oxford, Oxford, England, 1984.
Malito, J.T.; Shakir, R.; Atwood, J.L. /. Chem. Soc, Dalton Trans. 1980, 1253.
King, R.B. Org. Mass Spectrom. 1969,2,401. 47
Ashby, M.T.; Enemark, J.H. /. Am. Chem. Soc. 1986,108, 730. It is interesting to note that a similar halide dependence in metathesis reactions of
Cp2MX2 (M= Mo, W; X=Cl,Br,I) with Grignard reagents and lithium alkyls has been noted, with the isolated yields of products varying as I > Br > Cl. It may be that electron transfer steps are also occurring in these reactions although the
electrochemical behaviour of the Cp2MX2 (M=Mo,W) complexes has not been reported: Diversi, P.; Ingrosso, G.; Lucherini, A.; Porzio, W.; Zocchi, M. /. Chem.
Soc, Dalton Trans. 1983, 967.
Legzdins, P.; Phillips, E.C.; Sanchez, L. Organometallics 1989,8, 940. Christensen, NJ. Ph.D. Dissertation, The University of British Columbia, 1989. Phillips, E.C.; Sanchez, L., unpublished observations.
Gambarotta, S.; Floriani, C; Chiesi-Villa, A.; Guastini, C. Inorg. Chem. 1984,23, 1739 and references therein. This electron donating ability of substituent is quantified in the Hammett
parameter, a, derived from such pKa measurements, with more electron donating
(more negative a) substituents causing a greater reduction in the pKa of the acid.
For example, see: Lowry, T.H.; Richardson, K.S. Mechanism and Theory in
Organic Chemistry, 3rd ed.; Harper and Row: New York, NY; 1987, pp 143-159.
Kissinger, P.T; Heineman, W.R. in Laboratory Techniques in Electroanalytical
Chemistry; Kissinger, P.T; Heineman, W.R., Eds.; Marcel Dekker: New York, NY, 1984; Chapter 3, pp 90-93.
The Cp'M(NO)(Bz)2 complexes are 18-electron species by virtue of a 3-electron donation from one of the benzyl ligands: Phillips, E.C. Ph.D. Dissertation, The University of British Columbia, 1989. Reference 21, p 697. 48
CHAPTER 3
Insertions of CO into Unsaturated 16-Electron Tungsten Bis(alkyl) and Bis(aryl) complexes: Formation of Stable Acylaryl, Acylalkyl, and Bis (acyl) Species. 49
. Introduction
The insertion of CO into metal-alkyl bonds is one of the most extensively studied reactions in organometallic chemistry.1 The reason for this is the involvement of CO insertion steps in industrially important catalytic processes, e.g. hydroformylation of alkenes,2 as well as in the carbonylation of organic substrates by transition-metal complexes in synthetic organic chemistry.3 Two of the most widely investigated insertion reactions are those of the iron and manganese complexes outlined below,
L CpFe(CO)(L')R • CpFe(L')(L){C(0)R} (3.1)
L
Mn(CO)5R • Mn(CO)4(L){C(0)R} (3.2)
L, JJ = PR3, CO and studies on these two general reactions comprise a large part of what is presently understood about the detailed mechanisms of CO insertion.1 Some of the most important conclusions emanating from this work are:
(a) It is necessary for the CO and alkyl ligands to be arranged in a cis-fashion for the insertion to occur.
(b) When L = CO in equations 3.1 and 3.2 above, the CO that is inserted is always a coordinated CO and not the external CO. The incoming CO simply acts as a two-electron ligand.
(c) The insertions can be of two types, a formal alkyl migration onto the coordinated CO ligand or a formal insertion of the CO into the metal-alkyl bond. Both of these processes have been identified, the prevailing mechanism being strongly influenced by the solvents used. 50
(d) When the a-carbon of the alkyl group, R, is chiral there is no racemization
of this carbon after the insertion of CO.
These two systems (equations. 3.1 and 3.2) are typical of the majority of systems that have been studied with regard to CO insertion reactions, characterized by the insertion of a bound CO under the influence of an external ligand. Much less common are systems where an external CO molecule is the source of the inserted CO, and no incorporation of a trapping ligand is necessary for the insertion to occur. Possibly, this observation is due to a paucity of suitable coordinatively unsaturated metal-alkyl precursor complexes to allow the formation of a cis alkyl carbonyl complex.
The discovery of a class of stable 16-electron bis(alkyl) complexes,
CpM(NO)(R)2,4 prompted an investigation of their reactivity with small Lewis bases including CO. As outlined in the Introduction to Chapter 2, the coordinatively unsaturated
CpW(NO)(Tms)2 complex was observed to react with many Lewis bases, and these reactions led to the formation of stable products. However, the carbonylation of
CpW(NO)(Tms)2 did not lead to isolable products, although reaction with a bulky CO analogue, f-BuNC, did result in an iminoacyl complex via insertion of the isonitrile reagent into the W-alkyl bond.
Since a series of bis(alkyl) and bis(aryl) complexes closely related to
CpW(NO)(Tms)2 had now been prepared, a study of the reactivity of these alkyl complexes with CO was undertaken. It was hoped that for one of these bis(alkyl) species a stable product could be isolated after carbonylation in order to shed some light on the more
complex reaction of CpW(NO)(Tms)2 with CO. This Chapter discusses the reactivity of these bis(alkyl) and bis(aryl) complexes with CO to give stable insertion products. As well, the pathway of these transformations is discussed in light of intermediates observed spectroscopically during these reactions. 51
Experimental Section
The general experimental procedures employed in this study have been outlined in
detail in the Experimental Section of Chapter 2 of this Thesis. The CO and H2 gases (Matheson CP grade) were used as received without further purification. A flow system for in situ IR monitoring of reactions was constructed, as diagrammed in Figure 3.1, using a Teflon diaphragm pump (Cole-Parmer 07090-62) and a NaCl IR flow cell (Wilmad 105A10-5, 0.2 mm pathlength).
Reaction of Cp*W(NO)(>Tol)2 with CO. A blue solution of Cp*W(NO)(/?-Tol)2
(1.40 g, 2.63 mmol) in Et20 (50 mL) was prepared in a flask equipped with a calibrated gas reaction bulb filled with CO (475 mL @ STP, 20 mmol). The solution was cooled to -38 °C
with a saturated CaCl2(aq)/Dry Ice bath, the CO was then admitted into the reaction flask, and the mixture was stirred for 1 h. Over this 1 h period, the reaction solution changed from a deep blue to a pale orange along with the formation of a pale green-yellow precipitate. The orange supernatant solution was removed by cannulation, and an IR spectrum of the solid was recorded as a Nujol mull. This spectrum exhibited strong bands at 2014,1603, and 1585 cm"1 in the region 1400-2200 cm"1. As the mull warmed up, the
2014 cm"1 band diminished in intensity relative to the other two bands, and the colour changed from yellow-green to orange. The cooling bath was removed from the green- yellow solid, and after 3 h at room temperature, the solid had changed from light-green to
orange. This solid was dried of residual solvent in vacuo, dissolved in 1:1 CH2Cl2/hexanes (100 mL), and the volume of the solution was reduced until precipitate formation was noted. The saturated orange solution was placed in a -25 °C freezer to complete the crystallization of the product. A red microcrystalline solid formed (0.30 g) and was isolated by cannulating away the mother liquor and washing the solid with hexanes (50 mL). The mother liquor was again reduced to saturation and placed in the freezer. This yielded an additional 0.36 g of Cp*W(NO){C(0)/>-Tol}(>Tol) for an overall yield of 47%. The characterization data for the product are located in Tables 3.1 and 3.2. 52
A
Return
To reaction flask
Figure 3.1. Flow system for circulation of solutions for in situ IR monitoring of reactions. (A) 0.2 mm NaCl flow cell (Wilmad, Cat. # 105A10-5). (B) Teflon diaphragm pump
(Cole Parmer, Cat. # 07090-62). 53
Treatment of Cp*W(NO){C(0)/?-Tol}(p-Tol) with H2. In a glovebox, a red
solution of Cp*W(NO){C(0)/>-Tol}(p-Tol) (0.358 g) in CH2C12 (20 mL) was prepared in the Pyrex liner of a 300-mL Parr stainless steel pressure reaction vessel. The reactor was
assembled, removed from the box, and pressurized with H2 (450 psig). The reaction was
stirred for 5 h, then the excess H2 was vented, the reactor taken into a glovebox, and the solution transferred to a Schlenk tube. The resulting red reaction solution was reduced in vacuo to a red-orange solid. A portion of this red-orange solid was completely dissolved in CDCI3, and a 1H NMR spectrum of the resulting solution was recorded. This spectrum contained signals due exclusively to the starting material with no other proton-containing species being evident.
Treatment of Cp*W(NO){C(0)/?-Tol}(>Tol) with CO. A solution of
Cp*W(NO){C(0)/?-Tol}(p-Tol) (0.15 g) in CH2C12 (20 mL) was prepared in a Schlenk tube and cannulated into a 500-mL Fisher-Porter glass pressure vessel equipped with a pressure gauge and a magnetic stir bar. The vessel was pressurized with CO (75 psig) and stirred for 14 h. The pressure in the vessel was vented to a fumehood, and the solvent was removed under reduced pressure, leaving an orange-red solid. A IR spectrum of the red- orange solid as a Nujol mull displayed a new, strong band at 1970 as well a new band at
1620 cm"1 (m). Additional bands at 1605(m), 1553(s), 1540(m, sh), and 1535(m, sh) cm"1 were noted in the IR spectrum of the new product formed.
An attempt to crystallize the new product from 1:1 toluene/hexanes resulted in the deposition of a red solid material. The solid isolated from the crystallization was shown to be Cp*W(NO){C(0)p-Tol}(p-Tol) by comparison of its IR and ^ NMR spectra with those of an authentic sample, indicating that the product formed in the carbonylation had reverted to the acylaryl starting complex during the crystallization.
Low-Pressure Reaction of CpW(NO)(Npt)2 with CO. A red solution of
CpW(NO)(Npt)2 (0.30 g, 0.71 mmol) in CH2C12 (15 mL) was prepared in a Schlenk tube.
An IR spectrum of this solution displayed a i/^o absorption at 1570 cm'1. The tube was fitted with a gas reaction bulb filled with CO (475 mL @ STP, 20 mmol), and the stopcock 54 separating the gas bulb from the solution was opened to admit the CO. The mixture was stirred for 2.5 h during which time the red colour of the solution disappeared and was replaced by a yellow colour. An IR spectrum of the yellow solution contained two new bands at 1582 and 1557 cm"1. The yellow solution was reduced in vacuo to a yellow powder, redissolved in CH^Cl^, and transferred to the top of a Florisil column (120 x 10 mm). A single yellow band was collected by elution of the column with a 6:1
CH2Cl2/ether mixture. The solvent was removed from this yellow solution to give 0.24 g (80%) of CpW(NO){C(0)Npt}(Npt) as a yellow powder. The characterization data for this complex are located in Tables 3.1 and 3.2.
Low-Pressure Reaction of CpW(NO)(Neo)2 with CO. A red solution of
CpW(NO)(Neo)2 (0.500 g, 0.9 mmol) in toluene (30 mL) was prepared in a 100-mL 3-necked flask fitted with a gas reaction bulb filled with CO (475 mL @ STP, 20 mmol). The reaction was started by opening the stopcock on the gas bulb, and the mixture was stirred for 5 h. During this period the starting red colour completely discharged, leaving a yellow solution. The reaction mixture was then pumped free of solvent in vacuo, and the resulting yellow solid was dried overnight under dynamic vacuum. This yellow solid was washed with pentane (1 x 10 mL) and dried under vacuum for 2 h to give 0.480 g of CpW(NO){C(0)Neo}(Neo) (91% yield). The characterization data for this complex are collected in Tables 3.1 and 3.2.
In Situ IR Monitoring of the Reaction of CpW(NO)(Neo)2 with CO. A red
solution of CpW(NO)(Neo)2 (0.55 g, 1.0 mmol) in CH2C12 (100 mL) was prepared in a 250-mL 3-necked flask fitted with a gas reaction bulb filled with CO (475 mL @ STP, 20 mmol) and an IR flow cell system (Figure 3.1). The flow cell was placed in the sample compartment of the spectrometer, and the solution was circulated through the cell. After collecting an initial spectrum, the reaction was started by allowing the CO into the reaction vessel. The solution IR spectrum of the reaction was obtained periodically (Figure 3.2), and after 45 min the reaction was deemed complete when the IR spectrum recorded did not differ from the preceding spectrum. 55
o o o n • o o
WAVENUMBERS CCM-1>
Figure 32. Spectra from in situ IR monitoring of the reaction of CpW(NO)(Neo)2 with
CO in CH2CI2 with the flow system shown in Figure 3.1. 56
The solution in the flow apparatus was pumped completely into the flask and the flow system was removed from the reaction vessel. The yellow solution left in the flask was
reduced in vacuo to a yellow residue which was applied as a CH2Cl2 solution to a Florisil
column (80 x 15 mm) made up in CH2C12. A single yellow band was collected using
CH2C12 (2 x 30 mL) and then 2:1 CH2Cl2/ether (1 x 30 mL) as eluants. The solvents were removed from the eluates under reduced pressure to obtain 0.355 g (65%) of
CpW(NO){C(0)Neo}(Neo) as a yellow powder. The product was identified by comparison of its characteristic IR and 1H NMR spectroscopic properties with those of the previously prepared complex (vide supra).
In Situ Monitoring of the Reaction of CpW(NO)(Neo)2 with CO by NMR
Spectroscopy. A solution of CpW(NO)(Neo)2 in QDg was added to a sealable NMR tube, freeze-pump-thaw degassed two times, and frozen in a 0° ice-water bath. Then CO
(-0.3 atm) was admitted to the glass apparatus and condensed into the NMR tube with a
liquid N2 bath for 15 min, and the tube was flame-sealed closed. This amount of CO was calculated to give an internal pressure of approximately 1.5 atm of CO upon sealing the
NMR tube. After careful warming of the NMR tube to melt the QDg solution, an immediate reaction occurred with the solution changing rapidly (5 s) from red to yellow. A
41 NMR spectrum of this solution displayed peaks due to the singly inserted
CpW(NO){C(0)Neo}(Neo) species with no other proton signals being observed.
Preparation of CpW(NO){13C(0)Neo}(Neo) and CpW(NO){13C(0)Npt}(Npt).
Both of these complexes were prepared simultaneously to utilize a single bulb of labelled
^CO in their preparation. Solutions of CpW(NO)(Neo)2 (1.09 g, 2.0 mmol) and
CpW(NO)(Npt)2 (0.84 g, 2.0 mmol) were prepared in CH2C12 (45 mL each) in Schlenk tubes. These two tubes were then connected via a Y-adapter to a bulb of 13CO
(Sigma-Aldrich, 99 atom%, 100 mL, 4 mmol), and the reactions were initiated by breaking the glass break-seal on the gas bulb. The two solutions were stirred for 18 h to utilize the maximum amount of ^CO. After the 18 h reaction period both solutions, although still slightly red, had noticeably lightened in colour. The stopcock on the gas bulb was closed, 57 and the bulb was disconnected. The reaction mixtures were then worked up separately as follows:
The red solution containing CpW(NO){13C(0)Npt}(Npt) was freed of solvent
under reduced pressure to leave a red residue. This red residue was applied as a CH2C12
solution to the top of a Florisil column (90 x 30 mm) made up in 5:1 hexanes/CH2Cl2. The
column was sequentially eluted with 5:1 hexanes/CH2Cl2 (70 + 50 mL),
20:4:1 hexanes/CH2Cl2/ether (2 x 50 mL), 10:4:1 hexanes/CH2Cl2/ether (3 x 50 mL) and
1:1:1 hexanes/CH2Cl2/ether (2 x 50 mL). An orange band eluted first followed closely by a yellow band. The two separate fractions were each reduced in volume in vacuo until precipitation was noted, and the saturated solutions were then placed in a freezer (-25 °C) to crystallize. The solution resulting from the second band deposited a light yellow solid. This solid (430 mg) was collected by removing the mother liquor via cannulation and drying the solid under reduced pressure. The and ^CI^H} NMR spectra of the sample showed only resonances attributable to the singly labelled product
CpW(NO){13C(0)Npt}(Npt) by comparison with the similar spectra of the unlabeled analogue. No product was isolated from the solution of the first band since it contained a significant amount of the bis(alkyl) starting material. All of the spectroscopic data for
CpW(NO){13C(0)Npt}(Npt) are collected in Tables 3.1 and 3.2.
The solution containing CpW(NO){13C(0)Neo}(Neo) was pumped free of solvent
in vacuo and was applied as a CH2C12 solution to a Florisil column (140 x 25 mm) made up
in 3:1 hexanes/CH2Cl2. The column was eluted with 3:2 hexanes/CH2Cl2 (3 x 50 mL),
15:10:1 hexanes/CH2Cl2/ether (3 x 40 mL) and 2:2:1 hexanes/CH2Cl2/ether (4 x 40 mL) to give a single red-orange band which was collected in two fractions. The first fraction was taken to an orange solid in vacuo which was redissolved in
2:3 CH2Cl2/hexanes (25 mL) and put in the freezer to crystallize. The second fraction was reduced in volume until a precipitate began to form, and this solution was placed in a freezer to complete the precipitation. The first fraction deposited orange crystals which were collected by removing the supernatant solution via cannula, and they 58 were washed with hexanes (10 mL) to give 0.082 g of a solid orange product. The second fraction deposited an orange powder which was isolated in a similar manner to give 0.320 g of orange solid. The *H NMR spectra of both products showed that they were the desired
CpW(NO){13C(0)Neo}(Neo) compounds, slightly contaminated with -5% of the starting bis(alkyl). However, this was sufficiently pure to allow spectroscopic characterization of the product. Hence, the products were not purified further. The spectroscopic data for
CpW(NO){BC(0)Neo}(Neo) are collected in Tables 3.1 and 3.2.
Low-Pressure Reaction of CpW*(NO)(Neo)2 with CO. A Schlenk tube containing
a dark orange-red solution of Cp*W(NO)(Neo)2 (0.259 g, 0.42 mmol) in toluene (20 mL) was fitted with a gas reaction bulb filled with CO (475 mL @ STP, 20 mmol). The CO in the bulb was admitted to the reaction tube, and the mixture was stirred for 6 h. The volume of the reaction solution was reduced in vacuo to about 3 mL, and 30 mL of hexanes were added. The resulting cloudy solution was then left in a freezer (-10 °C) to complete the deposition of product. The yellow solid that formed was collected by filter cannulation and washed with hexanes (2 x 15 mL). This solid was dried to give 0.170 g of Cp*W(NO){C(0)Neo}(Neo) (63% yield) whose characterization data are collected in Tables 3.1 and 3.2.
High-Pressure Reaction of CpW(NO)(Neo)2 with CO. A solution of
CpW(NO)(Neo)2 (0.50 g, 0.92 mmol) in toluene (30 mL) was prepared in the glovebox in a Pyrex liner for a 300-mL Parr pressure reactor. The container was placed in the reactor, and the apparatus was assembled and removed from the box. The vessel was then pressurized to 450 psig and reacted unstirred for 24 h. The excess CO pressure was vented to a fumehood, and the reactor evacuated of non-condensable gases on a vacuum line. The apparatus was taken into the glovebox and disassembled to reveal a mixture consisting of long yellow fibers suspended in a light orange solution. The mixture was transferred into a flask and removed from the glovebox. The suspended matter was isolated by cannulating the supernatant solution into another flask and washing the remaining solid with 3:1 ether/hexanes (40 mL). This gave 0.28 g of a yellow solid product. 59
The orange supernatant solution was reduced to a yellow-brown powder in vacuo. The powder was dissolved in toluene (10 mL), and this solution was layered with hexanes
(50 mL). The mixture was placed in a freezer at -25 °C, and the layers were allowed to diffuse together to induce deposition of product. An additional 0.06 g of yellow product was isolated from this solution, giving an overall isolated yield of 68% of
CpW(NO){C(0)Neo}2. The characterization data for the complex are located in
Tables 3.1 and 3.2.
High-Pressure Reaction of CpW(NO)(Npt)2 with CO. A solution of
CpW(NO)(Npt)2 (0.56 g, 1.33 mmol) in toluene (30 mL) was prepared in the glovebox in a
Pyrex liner for a 300-mL Parr stainless steel pressure reactor. The liner was placed in the reactor, and the apparatus was assembled and removed from the box. The vessel was then pressurized to 80 psig with CO and was stirred for 14 h. The reactor was disassembled in a glovebox and the reaction solution was transferred to a Schlenk tube. The Schlenk tube was removed from the glovebox and the solvent was removed from the reaction in vacuo to afford a dark residue. The residue was dissolved in toluene and the solution was applied to a silica column (Merck Silica 60, 80 x 20 mm) supported on a frit. The column was washed sequentially with 4:1 toluene/ether (100 mL) and ether (20 mL) to elute a single orange band and the combined filtrates were reduced in vacuo to an orange-yellow residue.
Hexanes (40 mL) was added to the residue, resulting in a yellow solid suspended in an orange solution, and the flask was placed in a -25 °C freezer overnight to induce the deposition of more of the yellow solid. The yellow solid was isolated by removal of the supernatant solution by filter cannulation and was washed with pentane (20 mL). The yellow solid was dried in vacuo to obtain 0.24 g of CpW(NO){C(0)Npt}2 (37% yield) as an analytically pure yellow powder. The characterization data for this complex are collected in Tables 3.1 and 3.2. 60
Table 3.1. Infrared and Elemental Analysis Data for the Acylaryl, Acylalkyl, and Bis (acyl)
Complexes.
Compound IR (CH2Cl2) Analysis Found(Calcd) -1 % cm "NO "CO C H N
Cp*W(NO){C(0)/>-Tol}(p-Tol) 1562 1526 53.72 (53.68) 5.17 (5.22) 2.42 (2.50)
CpW(NO){C(0)Neo}(Neo)a 1582 1553 54.54 (54.46) 5.47 (5.45) 2.43 (2.44)
CpW(NO){C(0)Npt}(Npt)b 1582 1557 42.45 (42.77) 6.28 (6.06) 2.93 (3.12)
Cp*W(NO){C(0)Neo}(Neo) 1566 1547 57.69 (57.86) 6.60 (6.42) 2.23 (2.18)
CpW(NO){C(0)Neo}2 1591 1633,1552 53.77 (53.92) 5.19 (5.20) 2.30 (2.33)
CpW(NO){C(0)Npt}2 1588 1632,1553 42.57(42.78) 5.68 (5.70) 2.92 (2.93)
i:> 1
Data for CpW(NO){ C(0)Neo}(Neo) 1576 (s, z^NO) , 1520 (w, J>i3CO) cm".
13 1 Data for CpW(NO){ C(0)Npt}(Npt) 1575 (s, i/NO). 1526 (m, V13co ) cm*. 61 Table 32. Mass Spectral, 1H, and "C^H} NMR Data for the Acylaryl, Acylalkyl, and Bis(acyl) Complexes
Low-resolution JH NMR "CI1!"} NMR Mass spectra0 (C6D6) (C6D6)
m/zb 6 S
Cp*W(NO) {C(0)/>-Tol} (p-Tol) 8.19 (d, 2H, C(0)Artf, 279.16 (OO)
+ 531, [P-CO] 3/HH = 8.0 Hz) 168.22 (Cjpso) 501, [P-CO,NO]+ 7.85 (d, 2H, V/AiH, 147.08 (Cpara) 3/HH = 8.0 HZ) 13958 (Carom) 732 (d, 2H, C(0)Arff, 134.13 (Cpara) 3/HH = 8.0Hz) 13238 (Carom) 6.83 (d, 2H, WArtf, 130.10 (Carom) VHH = 8.0 Hz) 129.58 (Cjpso)
2.33 (s, 3H, C(0)ArCtf3) 129.13 (Carom)
1.92 (s, 3H, WArCfla) 109.56 (C5(CH3)5)
1.66 (s, 15H, C5(C/Y3)5) 21.75 (Ar CH3)
21.57 (Ar'CH3)
10.11 (C5(CH3)5)
CpW(NO) {C(0)Neo} (Neo) 7.66 (d, 2H, o-AiH, 292.23 (C=0)
+ 545, [P-CO] 3/HH = 7.5 Hz) 155.08 (Cjpso) 515, [P-CO,NO]+ 7.27 (d, 2H, o-AiH, 148.04 (Cjpso) 3/HH = 7.5 Hz) 128.70 (Carom) 7.25 (t, 2H, m-AiH, 128.11 (Carom) 3/HH = 7.5 Hz) 126.62 (Cpara) 7.15 (t, 2H, m-AiH, 126.43 (Carom) 3/HH = 7.5 Hz) 126.30 (Carom) 7.09 (t, \U,p-AiH, 125.13 (Cpara) 3/HH = 7.5 Hz) 99.34 (C5H5)
7.00 (t, m,p-AiH, 57.86 (C(0)CH2)
3/HH = 73 Hz) 44.48 (WCH2, Uwc = 86 Hz)
c 4.53 (s, 5H, CsH5) 44.18 (Cquat)
3.21 (d, 1H, C(0)Cf-AHB, 38.52 (Cquat)
2 /HH = 15.0 Hz) 35.23 (CH3)
2.71 (d,lH,WC//A'HB', 31.79 (CH3)
2 /HH = 12.0 Hz) 30.40 (CH3)
2.69 (d, 1H, C(0)CHA#B, 27.56 (CH3) 2/HH = 15.0 Hz) 62
2.65 (d, 1H,WCHA'#BS 2/HH = 12.0 Hz)
1.89 (s,3H, C(Cff3)A)
1.80 (s,3H,C(C/73)B)
1.30 (s, 3H, C(C/Y3)A')
1.13 (s, 3H, C(C/73)BO
C (C=0) CpW(NO) {C(O)Npt) (Npt) 5.01 (s,5H,C5^5) 291.95 449, [P]+ 2.86 (d, 1H, C(0)0/AHB, 99.53 (C5H5)
+ 2 434,[P-CH3] /HH = 135 Hz) 56.82 (C(0)CH2)
+ 421, [P-CO] 2.55 (d, 1H, C(0)CHA#B, 4335 (WCH2) 2/HH = 135 Hz) 37.77 (Cquat)
2.31 (d,lH,WC-/AHB, 34.66 (C(CH3)3) 2/HH = 132 Hz) 33.10 (Cquat)
2.25 (d, 1H, WCHA#B, 29.85 (C(CH3)3) 2/HH = 13.2 Hz)
1.53 (s, 9H, C(C/73)3)
0.97 (s, 9H, C(C/73)3)
CpW(NO) {13C(0)Neo} (Neo) 7.68 (d, 2H, o-AiH, 292.21 (C^.Vwc = 73 Hz)
+ 574, [P] VHH = 7.8 Hz) 155.05 (Cipso) 545, [P-13CO]+ 7.30,7.27 (m, 4H, o,m-AiH, 148.06 (Cjpso) 3/HH = 7.8 Hz) 128.73 (Carom) 7.18 (t, 2H, m-AiH, 128.12 (Carom) 3/HH = 7.8 Hz) 126.60 (Cpara) 7.10 (t, lH,p-AiH, 126.44 (Carom) 3/HH = 7.8 Hz) 126.30 (Carom) 7.02 (t, Ui,p-AiH, 125.12 (Cpara) 3/HH = 7.8 Hz) 99.34 (C5H5)
C 4.55 (s,5H,Cs^5) 57.82 (C(0)CH2, Vcc = 28 Hz) 3.24 (dd, 1H, C(0)0/AHB, 44.43 (WCH^/wc = 92 Hz) 2/HH = 14.9 Hz, 44.16 (Cquat) 2/CH = 7.1 Hz) 38.51 (Cquat)
2.73 (dd/ 1H, C(0)CHAi/B, 35.19 (CH3)
2 /HH = 14.9 Hz, 31.79 (CH3)
2 /CH = 4.8 Hz) 30.90 (CH3)
d 2.69 (m, 2H, WCHA>HB<) 27.58 (CH3)
1.89 (s, 3H, C(Ci/3)A)
1.80 (s, 3H, C(C/J3)B)
1.31 (s, 3H, C(CrY3)A')
1.15 (s, 3H, C(Cf/3)B') 63
c CpW(NO) {^CCOJNpt} (Npt) 4.97 (s,5H,CsH5) 291.93 (C=0,Vwc = 72 Hz) 450, [P]+ 2.83 (dd, IH, ^C^CHAHB, 99.53 (C5H5)
+ 435,[P-CH3] VHH = B.5 Hz, 56.83 (C(0)CH2, Vcc = 24 Hz)
2/cH = 7.1Hz) 43.35 (WCH2)
13 + B 421, [P- CO] 2.50 (dd, IH, C(0)CHA^& 37.77 (Cquat)
2 /HH = 13.5 Hz, 34.66 (C(CH3)3) 2/CH = 4.4HZ) 33.09 (Cquat)
29.85 (C(CH3)3) 2.32 (dd,lH,WCffAHB, 2/HH = 113 Hz, 3/CH = 13HZ)
2.26 (dd, 1H, WCHA#B, 2/HH = 113 Hz, 3/CH = 2.4HZ)
1.55 (s,9H,C(Cff3)3)
0.97 (s, 9H, C(C/73)3)
e CpW(NO){C(0)Neo}2 7.43 (d, 4H, 0-A1H, 279.88 (C=0) 545, [P-CO]+ VHH = 8.1 Hz) 148.67 (Cipso) 515, [P-CO,NO]+ 7.31 (t, 4H, m-AiH, 128.32 (Carom) VHH = 7.8 Hz) 126.01 (Cpara) 7.17 (t, 2H, p-ArH, 125.90 (Carom) VHH = 7.5 Hz) 100.51 (C5H5)
5.04 (s, 5H, CsHsf 63.31 (C(0)CH2)
3.49 (d, 2H, C(0)CtfAHB, 38.42 (Cquat)
2 /HH = 15.9 Hz) 30.49 ((CH3)A)
339 (d, 2H, C(0)CHA#B, 28.50 ((CH3)B) 2/HH = 15.9 Hz)
1.45 (s,3H, C(Ctf3)A)
1.44 (s, 3H, C(0/3)B)
C 27937 (C=0) CpW(NO){C(0)Npt}2 5.04 (s,5H,C5ff5)
+ 545, [P-CO] 3.49 (d, 2H, C(0)CffAHB> 100.45 (C5H5)
+ 515, [P-CO,NO] VHH = 15.9 Hz) 62.92 (C(0)CH2) 3.39 (d, 2H, C(0)CHA#B, 32.51 (Cquat)
VHH = 15.9 Hz) 30.11 (C(CH3)3)
1.45 (s, 3H, C(Ctf3)A)
1.44 (s, 3H, C(Cff3)B) 64
Cp W(NO){C(0)Neo}(Neo) 7.78 (d, 2H, o-AxH, 293.55 (C=0) 615, [P-CO]+ 3/HH = 7.2 Hz) 157.42 (Qpso) 584, [P-CO-NO-H]+ 7.29 (t, 2H, m-AiH, 148.27 (Cjpso) 3/HH = 7.2 Hz) 128.65 (Carom) 7.22 (d, 2H, o-AiH, 128.05 (Carom) 3/HH = 12 Hz) 126.46 (Carom) 7.16 (t, 2H, m-AxH, 126.20 (Carom) 3/HH = 7.2 Hz) 125.91 (Carom) 7.09 (t, m,p-AxH, 124.74 (Carom)
3 /HH = 7.2 Hz) 108.22 (C5(CH3)5)
7.02 (t, m,p-AiH, 54.30 (CH2)
3/HH = 7.2 Hz) 53.75 (CH2)
3.34 (d, 1H, C(0)C#AHB, 44.52 (Cquat) 2/HH = 13.2 Hz) 39.40 (Cquat)
2.97 (d, 1H, C(0)CHAHB, 32.49 C(CH3)
2 /HH = 13.2 Hz) 30.93 C(CH3)
2.20 (d, 1H, WO/A-HBS 28.87 C(CH3)
2 /HH = 12.6 Hz) 28.41 C(CH3)
1.82 (s, 3H, C(0/3)A) 9.60 (C5(CH3)5)
1.68 (s, 3H, C(CH3)B)
1.63 (d, 1H, WCHA'//BS
2 /HH = 12.6Hz)
1.41 (s, 15H, C5(Ctf3)5)
1.40 (s, 3H, C(CH3)A')
1.30 (s, 3H, C(Ctf3)B-)
Probe temperatures 80-150 °C (see Results and Discussion). Assignments are for the most abundant mass peak of the isotopic cluster. A 10-20 s relaxation delay necessary for proper integrated intensity (see Chapter 2, Experimental Section). The two overlapping signals are only partially resolved. Spectra recorded in CDCI3. 65
Results and Discussion
Low-Pressure Carbonylation of Cp*W(NO)(p-Tol)2. All of the complexes studied react quickly with CO. However, the reaction of the bis(aryl) with CO is sufficiently different to merit consideration separately from the other carbonylations. The reaction of the bis(aryl) complex with CO proceeds quickly under 1 atm CO at -20 °C (equation 3.3). This results in the clean formation of a green-yellow precipitate
o Et2O,-20 C
Cp*W(NO)(>Tol)2 • "Cp*W(NO)(CO)(p-Tol)2" ' (3.3)
blue CO, 1 atm. green-yellow which is formulated as the 18-electron, terminal carbonyl adduct of the starting 16-electron bis(aryl) complex. This formulation is based on the presence of a terminal CO ligand indicated by a distinctive IR absorption at 2014 cm"1 in the Nujol mull IR spectrum of the green product.
As the green product from reaction 3.3 is warmed to room temperature, the CO inserts into one of the tungsten-aryl bonds to form the acylaryl complex, causing a colour change from green-yellow to red-orange, i.e.,
25 °C
"Cp*W(NO)(CO)(p-Tol)2" * Cp*W(NO){C(0)/>-Tol}(p-Tol) (3.4)
green-yellow 3h red-orange
The red-orange Cp*W(NO){C(0)/?-Tol}(p-Tol) can be handled in air with no noticeable decomposition and is stable for at least 6 months when stored under dinitrogen at 10 °C.
The complex is less soluble in aliphatic solvents than is the Cp*W(NO)(p-Tol)2 precursor complex, but it is freely soluble in aromatic and polar organic solvents.
The IR spectrum of the orange complex (Table 3.1) is consistent with its
formulation as Cp*W(NO){C(0)^-Tol}(p-Tol), two bands at 1562 0/NO) and 66
1526 cm"1 (VQQ) being tentatively assigned as indicated. This acylaryl complex can be viewed either as a 16-electron species containing an r?1-acyl ligand (as shown in A below) or as an 18-electron complex containing an r/2-acyl ligand (as shown in B below).
W
A B
1 Since its vco band lies below 1560 cm" Cp*W(NO){C(0)^-Tol}(p-Tol) would seem to have an r?2-acyl ligand as in B. This region of the IR spectrum is the range expected for the
VQQ of an 772-acyl (1453-1625 cm"1)5 and is well outside the range of VQQ values generally observed for r^-acyls (1620-1680 cm"1).6 However, it should be noted that the rj1-acyl in
CpFe(dppe){C(0)Ph} has a VQQ of 1510 cm"1,7 so the position of a VQQ IR band is not a foolproof criterion for assignment of an »?2-acyl ligand. For low pressures of CO (~ 1 atm) the reaction appears to follow a two-step process. The incoming CO molecule first coordinates to the metal centre forming the 18- electron terminal carbonyl intermediate species as shown in equation 3.3, and then the coordinated CO subsequently inserts into one of the metal-aryl bonds. The coordination of CO before insertion into a metal-alkyl link is generally assumed to be occurring for carbonylations of unsaturated alkyl complexes by external CO, however only in a few instances have these coordinated CO intermediates been observed directly, e.g.,8
CO,-60°C -20 °c
RhCl2R(PPh3)2 • RhCl2R(CO)(PPh3)2 • RhCl2{C(0)R}(PPh3)2 (3.5)
VQQ = 2060 cm-1
The XH NMR and 13C{1H} NMR spectroscopic properties of Cp*W(NO){C(0)/>-Tol}(p-Tol) (Table 3.2) are consistent with its formulation as an
inserted product as shown in equation 3.4. The JH NMR spectrum of the acylaryl complex 67
II I
J
hk hk AAA KA
—I : 1 1 1 J _L ' 1 I I 1 1 I I L
8 7 6 5 4 3 2 pPM 1
Figure 3.3. The 80 MHz XH NMR spectrum of Cp*W(NO){C(0)^-Tol}(p-Tol) in (.).
Inset shows irradiation of the doublet at 8.19 ppm causing the decoupling of the doublet at
7.32 ppm. 68
(Figure 3.3) shows two inequivalent /Kolyl environments as expected for a singly-inserted product. The set of aryl proton resonances shifted downfield, relative to their position in the starting bis(aryl) complex, are assigned to the newly formed acyl ligand (Table 3.2). This assignment is based on the expected downfield shift caused by attachment of the alkyl group to an electronegative carbonyl, analogous to the shifts seen in the NMR spectra of carbonyl-containing organic molecules as well as metal-acyl complexes.6 The 13C{1H} NMR spectrum of this complex is also consistent with the proposed acylaryl structure, again showing two sets of /7-tolyl carbon resonances. The resonance for the acyl carbon occurs at 279 ppm, and this chemical shift is more consistent with an n2 rather than 771 type of acyl coordination by comparison with the commonly observed chemical-shift ranges for
2 5 1 6 both r/ -acyl (248-392 ppm) and r? -acyl (195-268 ppm) carbons. This adds additional support to the formulation of Cp*W(NO){C(0)p-Tol}(p-Tol) as an 18-electron complex. A molecular structure determination would be desirable both as conclusive evidence for this formulation as well as providing additional information on the degree of tungsten- oxygen interaction. Unfortunately, all attempts to grow suitable crystals of Cp*W(NO){C(0)j?-Tol}(p-Tol) for an X-ray structural determination have so far proven fruitless. Therefore, while it is not absolutely conclusive that the acylaryl species is an
?72-acyl complex, this formulation is the most reasonable given the available data. The reaction to form the acylaryl complex (equation 3.4) is apparently not reversible since decarbonylation does not appear to occur under ambient conditions. Also, there is no evidence for the acylaryl complex reverting to the bis(aryl) species when exposed to dynamic vacuum for extended periods of time. Decarbonylation has been observed for other acyl complexes, this being an especially good route for introducing alkyl ligands into anionic metal complexes via acid halides, e.g.,
-X" -co
LJJM" + XC(0)R • LNM{C(0)R} • I^MR (3.6)
R = Ar, CnF2n+2 69 when the corresponding alkyl halides are not sufficiently reactive to be used directly.9
Decarbonylation of Cp*W(NO){C(0)/?-Tol}(p-Tol) apparently occurs in the mass spectrometer since the mass spectrum of the acylaryl complex shows a highest mass isotope cluster attributable to the bis(aryl) starting material, corresponding to the loss of CO from the parent complex. Under the instrumental conditions used, this apparent decarbonylation process is very efficient as no signal for the parent complex ion is seen.
Reactivity of Cp*W(NO){C(0)/?-Tol}(p-Tol) with H2 and CO. The reactivity of
Cp W(NO){C(0)/7-Tol}(p-Tol) provides some insight into the nature of the proposed
772-acyl linkage. If it is indeed present, the tungsten-oxygen bond would not be expected to be strong. This is shown by the ability of CO to form an adduct with the acylaryl complex, i.e.,
CO, 75 psig, 14 h
Cp*W(NO){C(0)/>-ToI}(p-Tol)
toluene, N2,10 d
The formulation of the product of equation 3.7 as the CO adduct indicated above is due to the presence of a terminal CO stretch at 1970 cm"1 in the IR spectrum of the red solid isolated from this reaction. Furthermore, the r^-acyl of this proposed 18-electron product is indicated by a CO stretch in the IR spectrum at 1620 cm"1, in the range expected for an n 1-acyl.6
In contrast to its reaction with CO, Cp*W(NO){C(0)^-Tol}(p-Tol) does not
appear to react with H2, even under a H2 pressure of 450 psig. The lack of reactivity of the
acylaryl complex with H2 may be due to the inability of H2 to undergo oxidative addition to
the metal centre. This could be a result of H2 addition to the tungsten being unfavorable
with this ligand set, or the inability of H2 to displace the metal-oxygen bond and reach the metal centre. Dihydrogen has been shown to react with the bis(alkyl) species with ease, 70 i.e.,
H2,70 psig
CpW(NO)(Tms)2 • CpW(NO)(Tms)(H)(Phosphine) (3.8)
Phosphine except when the phosphine is PMe3, in which case the 18-electron adduct
CpW(NO)(Tms)2(PMe3) is formed, and this adduct is inert towards reaction with dihydrogen under these conditions.10 A similar situation may be present in
Cp*W(NO){C(0)/>-Tol}(p-Tol) with the oxygen of an r/2-acyl acting as a Lewis base, thus
forcing the metal into an 18-electron configuration which renders it inert to H2. The
differences in reactivity of Cp*W(NO){C(0)/?-Tol}(p-Tol) with H2 and with CO could be a manifestation of the better Lewis base properties of carbon monoxide versus dihydrogen, with the better base (CO) being able to displace an rj2-acyl interaction.
It is not surprising to note in a recent review of »?2-acyl complexes,5 that the majority of the reported r?2-acyl complexes have been of the actinide and early transition metals, paralleling the oxophilic nature of these metals. The metal-oxygen interaction is strong enough in these early transition-metal and actinide complexes to make the acyl ligand react more like an oxycarbene (as in C below) than an r?2-acyl interaction (B, vide supra).
C
This carbene-like form, C, is proposed since the distinctive reactivity shown by the acyl complexes of these oxophilic metals, e.g.,
CO
Cp2ThX{C(0)R} • [Cp2ThX{-0 = C(R)C(6) = /}]2 (3.9) 71 is best explained by viewing the acyl ligand in this manner. The Cp*W(NO){C(0)/>-Tol}(p-Tol) complex shows none of this type of reactivity with CO (vide supra), an observation which is consistent with the view that the proposed tungsten- oxygen interaction is weak and the acyl ligand is best viewed as an r?2-acyl as in B above.
Low-Pressure Carbonylation of Cp*W(NO)R2 (Cp' = Cp, R = Npt, Neo; Cp' = Cp*, R = Neo). Carbonylation of these bis(alkyl) complexes occurs under mild conditions, i.e.,
CO, 1 atm
Cp'W(NO)(R)2 • Cp'W(NO){C(0)R}R (3.10)
red 25 "C yellow to give good yields of the acylalkyl products. Reactions 3.10 are very clean, as is the NMR monitoring of the reaction to form CpW(NO){C(0)Neo}(Neo) (Experimental Section) which shows 100% conversion with no other proton-containing products present. This indicates that the isolated yields of products reflect losses in workup and not formation of mixtures of products. Monitoring of the same reaction by IR spectroscopy demonstrates the clean conversion of the bis(alkyl) to the acylalkyl (Figure 3.2), as indicated by the observation of two isosbestic points in these spectra. The acylalkyl complexes are all yellow solids which show reduced solubility in comparison to the corresponding bis(alkyl) starting complexes. They are, however, still somewhat soluble in aliphatic hydrocarbon solvents and very soluble in aromatic and polar organic solvents. The acylalkyl complexes all appear to have similar stability to the acylaryl complex discussed earlier, as they can be handled in air without noticeable decomposition,
and they can be stored unchanged under N2 for extended periods of time. The acylalkyl complexes all have a very distinctive cloying, fruity smell, a feature which is not shared with the acylaryl complex which has no noticeable odor.
The IR spectra of the acylalkyl complexes (Table 3.1) contain two strong bands in the region between 1550 and 1580 cm"1 consistent with their formulation as inserted 72 acylalkyl complexes. These two bands due to the vibrations of the nitrosyl and acyl ligands are of similar energy so they cannot be unambiguously assigned to V^Q or VQQ absorptions. Fortunately, the availability of the BCO labeled CpW(NO){13C(0)R}(R) species allows a more specific assignment of these bands. Using a simple two-atom harmonic oscillator model for the C-O stretch allows the calculation of an approximate
11 B ratio for VIIQQ/V^QQ of 1.0227. The CpW(NO){ C(0)Npt}(Npt) complex has two
1 bands at 1575 (vNO) and 1526 cm" (VUQQ) which are easily assigned since the V^Q should remain essentially the same as for the unlabeled complex. Using the frequency ratio above
1 1 along with the 1^3 co °f 1526 cm", allows the prediction of a vco °f 1560 cm" for
CpW(NO){C(0)Npt}(Npt). This value is remarkably close to that of 1557 cm"1 seen for one band of the unlabeled complex; thus this 1557 cm"1 band is assigned to the VQQ of the unlabeled CpW(NO){C(0)Npt}(Npt) complex. The utility of a simple two-atom harmonic oscillator model to predict an isotope shift for an acyl ligand is surprising given the complexity of the molecule, but this model has been used successfully in labelling studies of
2 12 1 other r/ -acyl complexes. Similarly, a VUQQ of 1520 cm" for
13 1 CpW(NO){ C(0)Neo}(Neo) leads to a calculated vco of 1554 cm" for the unlabeled analogue; thus the observed 1553 cm"1 band for CpW(NO){C(0)Neo}(Neo) is assigned to the i^co °f the acyl group. By analogy to the location of the VQQ for these two
CpW(NO){C(0)R}(R) complexes, the 1547 cm"1 band in the IR spectrum of
Cp*W(NO){C(0)Neo}(Neo) is assigned to the VQQ °f me acyl ligand. Since the VQQ for each of the acylalkyl complexes falls well below the range of VQQ values generally observed for r^-acyls (1620-1680 cm"1)6 all of the complexes are formulated as 18-electron species,
2 2 Cp'W(NO){r? -C(0)R}R, analogous to Cp*W(NO){r? -C(0)^-Tol}(p-Tol).
The transformations outlined in equation 3.10 are expected to proceed in a manner similar to the carbonylation of the bis(aryl) complex as shown in equations 3.3 and
3.4, though there is no direct evidence for an intermediate CO adduct in the carbonylations of the bis(alkyl) species. The assumption of an intermediate CO adduct is not unreasonable since the closely related CpW(NO)(Tms)2 complex appears to form CO 73 adducts readily at low temperature4, i.e.,
CO, -78 °C PMe3
CpW(NO)(Tms)2 • "CpW(NO)(Tms)2(CO)" •CpW(NO)(Tms)2(PMe3) (3.11)
purple yellow yellow
which can be displaced by the better Lewis base PMe3 as shown in equation 3.11. As noted earlier, the acylalkyl products are quite stable, a property which is in sharp contrast to the stability of the initial products observed in the reaction of
4 CpW(NO)(Tms)2 with CO, i.e.
CO,-20°C 25 °C
CpW(NO)(Tms)2 > yellow solution » decomposition (3.12)
VEO = 1635,1575 cm"1 ; E = C, N ??
The initial yellow solution formed in reaction 3.12 was presumed to contain acylalkyl complexes as judged by its IR spectrum, but no stable products could be isolated from this solution. Given the stability of the analogous CpW(NO){C(0)R}R complexes prepared in this study, the unsuccessful attempts to isolate any products from reaction 3.12 must stem from reactivity of the trimethylsilylmethyl alkyl groups on initially formed acyl products.
The trimethylsilylmethyl group has been shown to be reactive in ry2-acyl complexes, e.g.5'13
H20
Mo{C(0)Tms}(S2CNMe2)(CO)(PMe3)2 • Mo{C(0)CH3}(S2CNMe2)(CO)(PMe3)2 (3.13)
Cp2ThX{C(0)Tms} • Cp2ThX[OC(SiMe3) = CH2] (3.14)
However, since no product was isolated from reaction 3.12 the nature of the reactivity extant in this system remains unknown.
The 4l NMR spectroscopic properties of the Cp'W(NO){C(0)R}R products (Table 3.2) provide conclusive evidence of their formulation as singly-inserted products. 74
As expected, there are two different alkyl environments for these complexes corresponding to the alkyl and acyl ligands, with the set of alkyl signals shifted to lower field assigned to the acyl ligand due to the deshielding effect of the electron-withdrawing carbonyl group.
This can be seen in the 1H NMR spectrum of the CpW(NO){C(0)Neo}(Neo) complex (Figure 3.4). Both alkyl moieties in the acylalkyl complexes are diastereotopic, a fact which manifests itself in geminal AB coupling between the two protons of each methylene group. This diastereotopism is also mirrored in the observation of two methyl signals for each neophyl group in both of the Cp'W(NO){C(0)Neo}(Neo) complexes. The XH NMR data for the BCO labeled CpW(NO){13C(0)R}R compounds (Table 3.2) confirm the relative assignments of the methylene protons of the acyl and alkyl ligands, due to spin-spin coupling from the acyl carbon to these methylene protons. The acyl methylene protons
14 exhibit the expected larger spin-spin coupling to the acyl carbon (2JQH ~ 5-7 Hz) with the methylene protons of the alkyl group showing smaller coupling to the acyl carbon
(3/CH ~ 1 Hz). This distinctive coupling pattern can be seen in the XH NMR spectrum of
CpW(NO){13C(0)Npt}(Npt) (Figure 3.5).
Although the qualitative assignments of the spectra are straightforward, there is an unexpected trend in the degree of the chemical shift difference between the acyl methylene
protons (Ha and Hb) and the alkyl methylene protons (Hc and FLj).
'd c a
The observed chemical shift difference (AS) between Ha and H5 is larger than AS for Hc and FLj. This is not anticipated given the large AS seen for the diastereotopic methylene protons in the parent bis(alkyl) complexes (Chapter 2). The insertion of the CO group into one metal-alkyl bond increases the distance between the two bulky alkyl groups, thus changing the steric environment at the metal. This new environment could result in a new Figure 3.4. The 300 MHz XH NMR spectrum of CpW(NO){C(0)Neo}(Neo) in QDg. Inset is an expansion of the region from 5 2.6-2.8. 76
«
;' I l "l I | l I I l | l l I I' | I I I I | I I I I | I I I I | I I I I* | I I I I |' I T I 5 4 3 2 s 1 PPM
Figure 3.5. The 300 MHz *H NMR spectrum of CpW(NO){BC(0)Npt}(Npt) in QDg.
Inset is an expansion of the region 6 2.2-2.35. 77
preferred rotational orientation which has both alkyl methylene protons (Hc and HA in more similar chemical environments than in the corresponding bis(alkyl) precursors, resulting in the reduced AS observed. The observed diastereotopism of the acyl methylene protons can be ascribed mainly to chemical shift anisotropy due to the carbonyl group. This effect has been studied in organic carbonyl complexes, and it has been qualitatively shown that an a-proton in the plane of a carbonyl group is shielded relative to one out of the plane.15 The large chemical shift differences observed for the acyl methylene protons in the acylalkyl species would indicate that these complexes adopt a configuration which has one proton very much more in the plane of the carbonyl than the other.
The 13C{1H} NMR spectra of these complexes are also consistent with the proposed acylalkyl structures, the most important feature of these spectra being the location of the resonance for the carbon of the r?2-acyl group at -292 ppm for each of the acylalkyl complexes. This chemical shift is more consistent with an 772- rather than r^-type of acyl coordination when the generally observed ranges for both r?2-acyl (248-392 ppm)5 and r;--acyl (195-268 ppm)6 ligands are considered. This adds additional support to the formulation of the acylalkyl complexes as 18-electron species, but conclusive evidence for
T72-acyl coordination will require determination of the solid-state molecular structure. A structural determination of CpW(NO){C(0)Neo}(Neo) confirm this coordination mode, i.e.,
W '"NO
O
with the geometry indicated above, but there is substantial disorder in the structure so accurate bond distances cannot be determined.16 78
There is no tendency for the CO of the acylalkyl complexes to deinsert under high vacuum. However, the mass spectra of these species are variable depending upon the conditions used to obtain the mass spectrum. When the probe temperature is kept to 80 °C, the parent ion for the acylalkyl complexes (Cp' = Cp; R = Neo, Npt) is observed in the mass spectrum, but at temperatures over 100 °C the highest m/z feature observed for
R = Neo is the [P-CO]+ mass peak, due to a decarbonylation process similar to the one observed for Cp*W(NO){C(0)/>-Tol}(p-Tol) in its low resolution mass spectrum. The parent ion for CpW(NO){C(0)Npt}(Npt) can still be observed at 150 °C, but at a greatly reduced intensity relative to the fragmentation peaks when compared with the 80 °C spectrum. The mass spectrum of Cp*W(NO){C(0)Neo}(Neo) obtained at 150 °C contains no parent ion peak, the highest m/z peak observed under these conditions being
attributable to Cp*W(NO)(Neo)2. Acylalkyl Complexes. The successful isolation of these complexes affords a new family of stable acylalkyl organometallic complexes. The isolation of stable acylalkyl complexes is an uncommon occurrence in transition-metal organometallic chemistry, an observation which has been noted in papers dealing with complexes of this type, notably
17 18 CpRe(CO)2{C(0)Me}(Me) and [Re(CO)4{C(0)Ph}(Me)]\ The paucity of isolated acylalkyl complexes is due to the strong tendency for complexes of this type to reductively eliminate the acyl and alkyl ligands as ketones. Formation of ketones is a very common mode of reactivity for organometallic bis(alkyl) complexes with CO, i.e.,18,19
CO
,, LIJMR2 • LnMR(COR)" » I^M + R2CO (3.15) presumably through intermediate acylalkyl complexes as indicated in equation 3.15. The fate of the I^M fragment in reaction 3.15 is usually the formation of metal carbonyl complexes by trapping of the unsaturated LQM fragment with the excess CO present in the reaction mixture, e.g.,19 79
Fe(Npt)2(dippe) + 4 CO • Fe(CO)3(dippe) + (Npt)2CO (3.16)
or by the formation of isolable ??2-ketone complexes, e.g.,20
2 Cp*TaMe4 + CO • Cp*TaMe2(r? -Me2CO) (3.17)
The acylaryl and acylalkyl complexes isolated in this study are formulated as 18-electron species containing an r/2-acyl ligand, a description which is supported by their spectroscopic parameters (vide supra). This configuration may also account for some of their stability since r/2-coordination is seen in one of the other known classes of stable acylalkyl
5 complexes, namely Cp'2M{C(0)R}R (M = Zr, Hf). These zirconocene acylalkyl complexes are very close analogues to the acylalkyl complexes isolated in this study. The analogy stems from the observation that the C^M (M = Group 4 metal) and CpM'(NO) (M' = Group 6 metal) moieties are 14-electron fragments and are formally isolobal. This similarity is manifested by the formation of similar complexes of both metals, for example r?4-diene complexes (M = Zr;21 M' = Mo22) and 16-electron bis(alkyl) complexes (M = Zr,
Hf;23 M' = Mo, W4). These Group 4 bis(alkyl) and bis(aryl) complexes when reacted with
CO form the acylalkyl complexes mentioned above,24 i.e.,
CO
Cp2MR2 • Cp2M{C(0)R}R (3.18) R = alkyl, aryl, benzyl M = Zr, Hf
However, there are some differences between the complexes formed in this study and the Group 4 metallocene complexes formed in reactions 3.10. A major difference is the observation that the carbonylations of the Zr and Hf bis(alkyl) precursors (R = alkyl, benzyl) in reactions 3.18 are reversible. This is seen in the XH NMR spectra of the 80
Cp2M{C(0)R}R complexes, where toluene solutions revert in vacuo to the bis(alkyl) starting materials.24 This behaviour is not seen for the tungsten acylalkyl complexes, a fact which is demonstrated by their 1H and ^C^H} NMR spectra which show no evidence of the starting bis(alkyl) complexes being present even though the solutions are sealed in vacuo. The Zr and Hf acylaryl complexes are more stable than the corresponding acylalkyl species, with no reverse reaction to form the bis(aryl) precursors being noted. This added stability is attributed to an increased metal-acyl interaction due to the influence of the aromatic group.24
High-Pressure CO Reactions of CpW(NO){C(0)R}R. The reactions of the acylalkyl complexes with higher pressures of CO do not lead to the expected reductive elimination of ketone and formation of the well-known dicarbonylnitrosyl products as depicted in equation 3.19.
CO
CpW(NO){C(0)R}R —H—• CpW(NO)(CO)2 + R2CO (3.19) R = Npt, Neo
Instead, carbonylation of the CpW(NO){C(0)R}R complexes results in a more interesting transformation,
CO, 70-450 psig
CpW(NO){C(0)R}R - • CpW(NO){C(0)R}2 (3.20)
R = Npt, Neo affording very novel bis(acyl) complexes in good yields. The most convenient route to the bis(acyl) products is from direct high-pressure carbonylation of the bis(alkyl) precursor complexes, without the need for the isolation of the intermediate acylalkyl complexes.
The observed reactivity of Cp*W(NO){C(0)^-Tol}(p-Tol) with CO (equation 3.7) contrasts with the CO insertion seen for the CpW(NO){C(0)R}R complexes in equation
3.20. No insertion of CO into the second metal-aryl bond is seen after formation of the 81
Cp*W(NO){C(0)^-Tol}(p-Tol)(CO) adduct, so there must be some factor mitigating against CO insertion in this complex. The most probable factor could be the anticipated stronger metal-aryl bond relative to a metal-alkyl bond,25 thereby rendering the second insertion energetically unfavorable in the aryl case.
These bis(acyl) complexes are stable in solution and the solid state, with no evidence for decarbonylation processes at ambient conditions. However, the mass spectra of these complexes are again variable as seen for the precursor acylalkyl species, exhibiting very facile loss of both of the incorporated CO molecules at higher temperatures, with the parent ion being observed (R = Npt) if the probe temperature is kept around 80 °C.
Both bis(acyl) complexes have qualitatively similar IR spectra whose most notable feature consists of three strong bands in the range 1635-1550 cm"1 (Table 3.1). The highest band for each complex is near 1630 cm"1 and is most certainly due to a simple v 1-acyl ligand. The additional two bands can then be assigned to the nitrosyl and jj2-acyl ligands by analogy to the IR spectra of the acylalkyl complexes (vide supra). These assignments lead to the description of these complexes as 18-electron
CpW(NO)(r?1-acyl)(r/2-acyl) species. The location of the r^-acyl band near 1630 cm"1 for both complexes can be taken as a representative value for the VQQ of a complex containing a Cp'W(NO)(r71-acyl) moiety. This hypothesis lends additional support to the formulation of the precursor CpW(NO){C(0)R}R complexes as r?2-acyls since they contain no bands above 1600 cm"1. This expected value for the VQQ of an r?1-acyl ligand is also consistent with the formulation of the product of reaction 3.7, since the IR spectrum of this product contains an absorption at 1630 cm"1 corresponding to the rj 1-acyl ligand which this species must contain if it is indeed the terminal CO adduct Cp*W(NO){C(0)/?-Tol}(>Tol)(CO).
An X-ray crystallographic investigation of the molecular structure of
CpW(NO){C(0)Neo}2 indicates that in the solid state it possesses the predicted r?2,^1 bis(acyl) coordination mode around the metal centre.26 82
NO
However there is a great deal of disorder in the alkyl residues, a problem which precludes refinement of the data below an R-value of 13%. Although this level of refinement allows the gross connectivity of the ligands around the metal to be determined with some certainty, it is not sufficient to allow accurate metrical data to be derived. These bis(acyl) compounds show very intriguing NMR spectra (Table 3.2), the room temperature 1H NMR spectra being suggestive of the presence of symmetric molecules with identical acyl ligands (Figure 3.6). This is contrary to the apparent rj2, rj1 configuration of the two acyl ligands indicated by the IR spectra of these complexes
and the solid-state structure of CpW(NO){C(0)Neo}2. Therefore, it would appear that the two acyl ligands are fluxional, with the acyl oxygen atoms quickly exchanging from r?1- to f?2-coordination in concert, i.e.,
Neo Neo resulting in NMR spectra that are a time average of the two possible environments.
Unfortunately, variable temperature 41 NMR spectra of CpW(NO){C(0)Neo}2 between -50 and 25 °C (Figure 3.7) are inconclusive since a static configuration cannot be frozen out. The changes in the general appearance of the NMR spectra also cannot rule out this fluxional process.
The ^C^H} NMR spectra of the complexes also show a symmetric coordination mode for the acyl ligands, as demonstrated by the 13C{1H} NMR spectrum of
CpW(NO){C(0)Neo}2 (Figure 3.8). The signals for the acyl carbons in the bis(acyl) X Figure 3.6. The 300 MHz H NMR spectrum of CpW(NO){C(0)Neo}2 in CDC13. 84
8 4 5 0 PPM
X Figure 3.7. The variable temperature 80 MHz H NMR spectra of CpW(NO){C(0)Neo}2
in CDC13 from -50 to 20 °C. '631-SZI 9 UIOJJ UOI§3J aqj jo uoisiredxa ire SI J9SUT oo Z £ (•) DCD.in ^{03N(0)D>(0N)Md3 jo uirupads HWN {Hx}3£1 FM S£ 9qx 86 species are shifted ~ 10 ppm upfield relative to the acyl carbons in the acylalkyl complexes.
This shift is consistent with a fluxional rj1,^2 bis(acyl) complex, with the observed chemical shift of the acyl carbon being roughly the average of the chemical shift for a tungsten r)2- acyl (-292 ppm, vide supra) and a tungsten rj ^acyl (240-250 ppm).27 If the symmetry observed in the NMR spectra is due to an alternative formulation, namely a symmetric 16- electron bis(rj1 -acyl) complex, then the acyl resonance would be expected to be closer to that of an rj1-acyl. Thus an r?1,^2 bis(acyl) complex best fits the data obtained. Bis(acyl) complexes. The bis(acyl) complexes prepared here are very rare examples of isolable bis(acyl) compounds. In most cases the carbonylation of bis(alkyl) compounds, seemingly the most straightforward route to complexes of this type, does not result in the formation of stable bis(acyl) species but rather to the formation of ketones via unstable acylalkyl intermediates (vide supra, equation 3.9).In cases where the acylalkyl complexes are stable, reactivity patterns suggestive of the formation of bis(acyl) complexes are noted. This reactivity can be divided into two major groups. The first category of reactivity involves the carbonylation of late transition-metal bis(alkyl) complexes, which generally results in the formation of ketones via acylalkyl intermediates. However in some cases low yields of diacyl products, RC(0)C(0)R, have been observed as well.28'29
co
LnM(R)2 » "LnM{C(0)R}R + LnM{C(0)R}2" <*• LnM + RC(0)R + RC(0)C(0)R
(3.20) M = Ni, Pd
These organic diacyl products are presumably a result of the formation of short lived bis(acyl) intermediates which reductively eliminate the diacyl product, although these intermediate bis(acyl) complexes have not been isolated. Another possible mechanism for the formation of diacyl products is via CO insertion into a metal-acyl bond, but this is generally believed to be energetically unfavorable. This conclusion is reasonable since 87 there is only one known example of CO insertion into a metal-acyl bond reported for a Mn complex under specific conditions.30
A second class of reactivity is seen in carbonylations of bis(alkyl) metallocenes,
CO
Cp*2M(R)2 - "Cp*2M{C(0)R}2" • Cp*2M[0(R)C=C(R)0] (3.22)
M = Th, U;31 Zr32 where the bis(acyl) species in equation 3.22 have not been isolated due to their rapid conversion to the enediolate complexes indicated above. Nevertheless, bis(r?2-acyl) intermediates have been observed by NMR spectroscopy at low temperature for the Th- and U-containing cases.33 The coupling of the »?2-acyl ligands in reaction 3.22 is attributed to the carbene-like character observed for r?2-acyl ligands in these oxophilic metals. This coupling is not seen for the tungsten bis(acyl) complexes prepared in this study, possibly stemming from a weaker r/2-acyl interaction due to the lower oxophilicity of tungsten.
Theoretical studies on the mechanism of the coupling reaction in equation 3.22 have shown that a 20-electron bis(r?2-acyl) species is a likely intermediate in the formation of the enediolate complexes of Th and U. Such 20-electron species are unlikely for the tungsten complexes, and this is another possible reason for not observing a reaction similar to equation 3.22 for the tungsten bis(acyl) complexes.
Stable bis(acyl) complexes of Re and Mn have been successfully synthesized and isolated by nucleophilic attack on an acylcarbonyl complex,18
+ M(CO)5{C(0)R} + R'Li » Li [M(CO)4{C(0)R}(C{0}R')]" (3.23)
M = Mn, Re
Nevertheless, the synthesis of bis(acyl) complexes by CO insertion appears to be a rarely observed reaction. One instance where this has been reported34 is in the reaction of a very electron-deficient homoleptic titanium complex 88
CO
Ti(Bz)4 —• Ti(Bz)4(CO)2" Ti(Bz)2(C{0}Bz)2 (3.24)
1 1640 1635 cm 1 vco = 1952,1867 cm- "CO = >
In a system more closely related to the bis(acyl) compounds in this study, the metallocyclic
Hf bis(alkyl) complex below has been shown to insert CO into both metal-alkyl bonds35
O CO CPYCO * °M }^~JJ (325)
Cp" = Me3SiC5H4 to form a stable bis(acyl) complex. It is possible that the rigidity imparted by the chelating ring inhibits the formation of an enediolate ligand as in reaction 3.22.
Epilogue
These studies have shown that the products obtained from the carbonylation of
Cp'W(NO)R2 complexes are very dependent upon the nature of the alkyl groups. For the case where R = Tms, the carbonylation products are not stable. However the use of other alkyl or aryl groups, results in the successful synthesis of novel, stable acylalkyl, acylaryl and bis(acyl) complexes. Differing reactivity with CO is also seen between the bis(alkyl) and bis(aryl) complexes, showing that there are very subtle differences in the energetics of these
processes that can affect the outcome of the reactions. The formally isolobal Cp2MR2
(M = Group 4) complexes exhibit reactivity toward CO that is somewhat different to that observed in this study, an indication that the isolobal analogy is probably not that applicable for these acyl complexes. The ability to form the acylalkyl and acylaryl complexes cleanly should allow for the investigation of the kinetics of these carbonylation reactions. It will be of interest in the future to investigate the characteristic reactivity of these new acylalkyl, acylaryl, and bis(acyl) complexes in order to compare it with that observed for related complexes and intermediates. 89
References and Notes
Alexander, JJ. In Chemistry of the Metal-Carbon Bond; Patai, S.; Hartley, F.R.,
Eds.; John Wiley and Sons: New York, NY, 1985; Vol 2, Chapter 5.
Collman, J.P.; Hegedus, L.S.; Norton, J.R.; Finke, R.G. Principles and Applications of Organotransition Metal Chemistry, University Science Books: Mill Valley, CA,
1987; Chapter 12.
Reference 2, Chapter 16.
Legzdins, P.; Rettig, S.J.; Sanchez, L. Organometallics, 1988, 7, 2394.
Durfee, L.D.; Rothwell, I.P. Chem. Rev. 1988,88, 1059.
Kegley, S.E.; Pinhas, A.R. Problems and Solutions in Organometallic Chemistry,
University Science Books: Mill Valley, CA, 1986; Chapter 1.
Felkin, H.; Meunier, B.; Pascard, C; Prange, T. /. Organomet. Chem. 1911,135,
361.
Baird, M.C.; Mague, J.T.; Osborn, J.A,; Wilkinson, G. /. Chem. Soc. (A) 1961,
1347.
Reference 2, Chapter 3, Section 3.5.
Martin, J.T. Ph.D. Dissertation, The University of British Columbia, 1987.
Herzberg, G. Spectra of Diatomic Molecules, 2nd ed.; D. Van Nostrand: New York,
NY, 1963; p 141.
Curtis, M.D.; Shiu, K.-B.; Butler, W.M. / Am. Chem. Soc. 1986,108,1550.
Carmona, E.; Sanchez, L.; Marin, J.M.; Poveda, M.L.; Atwood, J.L.; Priester, R.D.;
Rogers R.D. /. Am. Chem. Soc. 1984,106,3214.
2 The normal range of /CH observed for a-keto protons in organic molecules is 6-
7.5 Hz: Levy, G.C; Lichter, R.L.; Nelson, G.L. Carbon-13 Nuclear Magnetic
Resonance Spectroscopy, 2nd ed.; Wiley-Interscience: New York, NY, 1980; p 158.
Karabatsos, GJ.; Sonnichse, G.C; Hsi, N.; Fenoglio, DJ. /. Am. Chem. Soc. 1967,
89, 5067. 90
The structural determination was hampered by disorder due to the presence of two molecules in the unit cell. Isotropic refinement of the structure allows the determination of the connectivity of the molecule but accurate bond distances and angles could not be obtained. Einstein, F.W.B., Simon Fraser University, personal communication, June 1990.
Goldberg, KL; Bergman, R.G. /. Am. Chem. Soc. 1989, 111, 1285.
Casey, CP.; Scheck, D.M. /. Am. Chem. Soc. 1980,102,2123, and references cited therein.
Hermes, A.R.; Girolami, G.S. Organometallics 1988, 7, 394, and references cited therein.
Wood, CD.; Schrock, R.R. /. Am. Chem. Soc. 1979,101, 5421.
Erker, G.; Kruger, C; Muller, G. Adv. Organomet. Chem. 1985,24,1.
Christensen, N.J.; Hunter, A.D.; Legzdins, P. Organometallics 1989, 8, 930.
Cardin, D.J.; Lappert, M.F.; Raston, CL. Chemistry of Organo-Zirconium and
-Hafnium Compounds; John Wiley and Sons: New York, NY, 1986; Chapter 6.
Fachinetti, G.; Fochi, G.; Floriani, C. /. Chem. Soc, Dalton Trans. 1977,1946.
Stoutland, P.O.; Bergman, R.G.; Nolan, S.P.; Hoff, CD. Polyhedron 1988, 7, 1429 and references therein.
(a) Yee, V.C, University of British Columbia, personal communication, June
1989. Some selected bond distances (A) are: W - N, 1.97(4); W - Cl, 1.95(5); W - C2, 2.33(5); W - Ol, 2.31(5); W - 02, 2.84.
(b) A related acyl complex, CpMo(NO)(I){C(0)/?-Tol} has been shown to have an r/2-acyl ligand with a similar coordination geometry 91
with the acyl oxygen bonding trans to the NO ligand: Bonneson, P.V.; Yau, P.K.L.; Hersh, W.H. Organometallics 1987, 6, 1587.
Mann, B.E.; Taylor, B.F. 13C NMR Data for Organometallic Compounds;
Academic Press: New York, NY, 1981; p 147.
Akermark, B.; Ljungqvist, A. /. Organometal. Chem. 1978,149, 97. Ozawa, F.; Sugimoto, T.; Yuasa, Y.; Santra, M.;Yamamoto, T.; Yamamoto, A.
Organometallics 1984,3, 683. Sheridan, J.B.; Johnson, J.R.; Handwerker, B.M.; Geoffroy, G.L.; Rheingold, A.L. Organometallics 1988, 7, 2404. Moloy, K.G.; Fagan, P.J.; Manriquez, J.M.; Marks, T.J. /. Am. Chem. Soc. 1986,
108, 56. Manriquez, J.M.; McAlister, D.R.; Sanner, R.D.; Bercaw, J.E. /. Am. Chem. Soc.
1978,100, 2716.
Tatsumi, K.; Nakamura, A.; Hofmann, P.; Hoffmann, R.; Moloy, K.G.; Marks, T.J.
/. Am. Chem. Soc. 1986,108, 4467.
Roder, A.; Thiele, K.-H.; Palyi, G.; Mark6, L. /. Organometal Chem. 1980,199, C31.
Lappert, M.F.; Martin, T.R.; Atwood, J.L.; Hunter, W.E. Chem. Commun. 1980, 476. 92
CHAPTER 4
Preparation of Chiral Molybdenum and Tungsten Complexes: Physical, Structural, and
Spectroscopic Investigations of the Novel r?2-Benzyl Ligand.1 93
Introduction
The family of bis(alkyl) complexes, Cp'M(NO)(R)2 (M = Mo, W), has exhibited
2 3 3 4 many novel transformations, most notably when treated with C*2, Sg, Seg, and CO
(Chapter 2, Introduction). These reactions involve the transformation of at least one of the metal-alkyl bonds in these compounds. Therefore, it became of interest to prepare mixed bis(alkyl) species, Cp'M(NO)(R)(R'), to allow a comparison of reactivity of two different alkyl groups on the same Cp'M(NO) fragment. Prior to this work, there had been unsuccessful attempts by other members of this research group to prepare mixed alkyl complexes of this type. Although they were unsuccessful, it is useful to consider the observations collected during these attempts in order to analyze the possible sources of failure. An attempt to form a mixed bis(alkyl) species via stepwise alkylation employed a
reaction having a 1:1 stoichiometry of TmsMgCl and CpW(NO)(I)2. Surprisingly, this reaction did not result in the simple metathesis of one halide for one alkyl to form the expected CpW(NO)(Tms)(I) product. Instead, the known bis(alkyl) species was formed:5
CpW(NO)(I)2 + TmsMgCl » Vi CpW(NO)(I)2 + Vi CpW(NO)(Tms)2 (4.1)
Subsequent studies on the electrochemical properties of the dihalonitrosyl precursor complexes suggested that the reaction might possibly proceed through a radical anion intermediate.6 If this radical anion reacts with a second equivalent of Grignard reagent faster than the diiodonitrosyl starting material, it would explain the observed selectivity in alkylation giving the bis(alkyl) product. 94
As well, attempts to form these species via the reaction of a mixture of Grignard reagents, e.g.,
CpW(NO)(I)2 + TmsMgCl + BzMgCl —+• Vi CpW(NO)(Bz)2 + Vi CpW(NO)(Tms)2 (4.2) were also unsuccessful, the known bis(alkyl) species being the only products isolated.7 This lack of success can be attributed to various possible complications. The first one could be that the desired CpW(NO)(Bz)(Tms) compound is inherently unstable with respect to disproportionation to form the symmetrically substituted products, i.e.,
2 Cp'M(NO)(Bz)(Tms) • Cp'M(NO)(Bz)2 + Cp'M(NO)(Tms)2 (4.3) thereby explaining its lack of formation from the mixture of alkylating agents. This type of alkyl exchange has been seen in a mixed bis(alkyl) zirconocene species,8 i.e.,
hi/
4 2 Cp2Zr(Bz)(CH = CHPh) • Cp2Zr(Bz)2 + Cp2Zr(r? -PhCH = CHCH = CHPh) (4.4)
A second possible problem associated with the mixture of Grignards may be due to the relative rates of reaction of the two alkylating agents. If one Grignard reagent is more reactive than the other, then the selective nature of the metathesis may be a manifestation of this kinetic effect.
Only one other mixed alkyl species was known prior to the beginning of this work,
9 namely Cp*W(NO)(Tms)(CH2CPh3). It was formed serendipitously by the reaction of
+ Ph3C PF6" with Cp*W(NO)(Tms)2, i.e,
+ Cp*W(NO)(Tms)2 + Ph3C PF6- • Cp*W(NO)(Tms)(CH2CPh3) + PF5 + Me3SiF (4.5) but this synthesis proved to be of minimal synthetic utility due to difficulties in isolation, poor yields, and a lack of generality since this was the only case for which the reaction was successful. 95
This Chapter outlines the successful synthesis of the general class of compounds
having the composition CpM(NO)(r?2-CH2Ar)(Cl) and their subsequent alkylation to form
the previously elusive mixed alkyl species CpM(NO)(r/2-CH2Ar)(R). The novel spectroscopic and structural properties of this whole range of complexes, especially with regard to the nature of the benzyl-metal interaction, are discussed in detail. 96
Experimental Section
All manipulations were performed under anaerobic conditions under an atmosphere of dinitrogen using the general experimental procedures outlined in Chapter 2. Reagent chemicals were obtained from commercial sources (anhydrous HC1, Matheson; TmsMgCl in Et20, Aldrich), and were used as received. The organometallic starting
10 materials Cp*M(NO)Cl2 (M = Mo or W) ; and Cp'M(NO)(Bz)2 (M = Mo, W; Cp'= Cp,
Cp*)11 were prepared by known routes.
Preparation of Cp*W(NO)(/>Xyl)2. A suspension of Cp*W(NO)(Cl)2 (6.0 g,
14.3 mmol) in Et20 (100 mL) was cooled to 0°C in an ice water bath. An aliquot of
12 />-XylMgCl in Et20 (33 mL, 0.86 N, 28.6 mmol) was added to the light green solution/brown suspension via a syringe. An immediate reaction ensued, as evidenced by the formation of a light orange flocculant suspension. After being stirred for 1 h, the
reaction mixture was treated with CH2C12 (50 mL) to dissolve the suspended solids. Deaerated water (3 mL) was added to the reaction mixture, which caused the formation of a white gel on the sides of the flask and a darker red-orange solution. After a further
15 min of being stirred, the solvents were removed under reduced pressure to give a red-
orange solid. The residue was extracted with CH2C12 (30 mL) and the extract was transferred to the top of an alumina column (activity III, 200 x 30 mm) supported on a glass
frit. The column was washed with CH2C12 to give a single orange band which was collected and reduced in vacuo to about 100 mL in volume. The resulting solution was diluted with an equal volume of hexanes and reduced in volume until precipitation had been initiated. The saturated solution was then placed in a freezer (-25 °C) to complete the crystallization of product.
An orange microcrystalline solid formed which was collected by cannulating away the mother liquour and washing the solid with pentane (20 mL). This solid was then dried
in vacuo to give 2.38 g of Cp*W(NO)(/7-Xyl)2. The mother liquour was reduced in volume again until incipient precipitation, and another crop of product was isolated (2.48 g). A 97
final crop of 0.80 g of orange microcrystals was obtained in a third crystallization (72%
overall yield).
In a similar manner the Cp*Mo(NO)(p-Xyl)2 complex was obtained in 76% overall yield starting from 19 mmol of Cp*Mo(NO)(Cl)2. The analytical and spectroscopic data for
both of the Cp*M(NO)(p-Xyl)2 complexes are collected in Tables 4.1 and 4.2.
Preparation of Cp*W(NO)(Bz)(Cl). An orange solution of Cp*W(NO)(Bz)2
(6.70 g, 12.6 mmol) in CH2C12 (100 mL) was prepared in a 300-mL 3-necked flask. The reaction flask was partially evacuated and then filled with dry HC1 gas. The subsequent reaction was monitored by periodically transferring by cannulation a sample of the reaction solution to an IR cell and recording its IR spectrum. The reaction appeared to proceed in a straightforward manner with the diminution of the starting nitrosyl absorption band at
1548 cm"1 occurring concomitantly with the appearance of a new band at 1582 cm"1. When the spectrum obtained showed no qualitative difference from the one previously recorded, the reaction was deemed to be complete. There was no evidence for the appearance of any
= 1 spectral features due to Cp*W(NO)(Cl)2 O/NO 1624 cm") in the IR spectra obtained. The solution was then treated with hexanes (100 mL), and the volume was reduced in vacuo. When an orange precipitate began to form, the reaction solution was put into a freezer (-25 °C) to induce the crystallization of the product. This solution deposited a orange solid which was isolated by cannulation of the supernate into another flask. This orange solid was washed with hexanes (20 mL), and the wash solution was added to the supernatant solution. The solid was dried in vacuo for 2 h under dynamic vacuum to obtain
4.10 g (69% yield) of Cp*W(NO)(Bz)(Cl) as a red-orange crystalline solid. The combined solutions were reduced in volume to induce precipitation and returned to the freezer. An additional amount of product (1.08 g) was isolated from the second crystallization (86% overall yield).
The other congeners were prepared similarly to give Cp*Mo(NO)(Bz)(Cl) (66% yield), CpMo(NO)(Bz)(Cl) (87% yield), CpW(NO)(Bz)(Cl) (60% yield), and
Cp*Mo(NO)(p-Xyl)(Cl) (85% yield) as microcrystalline solids. A crystal of 98
Cp*Mo(NO)(Bz)(Cl) suitable for X-ray crystallographic analysis was chosen from product prepared in this manner. The analytical and spectroscopic properties of these compounds are presented in Tables 4.1 and 4.2. There was no evidence for the formation of
Cp'M(NO)(Cl)2 compounds in these preparations, as no nitrosyl absorptions for these dichloro complexes were observed in the IR spectra of the reaction solutions.
Preparation of Cp*W(NO)(p-Xyl)(CI). A sample of Cp*W(NO)(p-Xyl)2 (2.38 g,
4.26 mmol) was dissolved in toluene (80 mL). A gas reaction bulb containing HC1 (475 mL
@ STP, -20 mmol) was then attached to the reaction flask and opened to the reaction solution. The reaction mixture was stirred for 18 h during which time the solution became darker red in colour. The solution was reduced in vacuo to obtain an orange-yellow solid
which was dissolved in 1:1 hexanes/CH2Cl2. The volume of this solution was reduced in vacuo until solid had formed and the supernatant liquid was a pale orange. The
Cp W(NO)(p-Xyl)(Cl) product was collected by cannulating away the supernatant solution and dried under vacuum to an orange feathery solid (1.86 g, 3.8 mmol, 89% yield).
Characterization data for this compound are collected in Tables 4.1 and 4.2.
Preparation of Cp'M(NO)(Bz)(Tms) Complexes. The four congeners were prepared similarly, the specific procedure for Cp*Mo(NO)(Bz)(Tms) being given as a representative example.
A suspension of Cp*Mo(NO)(Bz)(Cl) (1.45 g, 3.75 mmol) in Et20 (100 mL) was prepared and an IR spectrum of the supernatant solution was recorded (v\so = 1626 cm"1). The mixture was treated with an aliquot of TmsMgCl in Et20 (5.0 mL, 0.75 N,
3.75 mmol) and an instantaneous reaction occurred as evidenced by the disappearance of the suspension and the formation of a clear red solution. An IR spectrum of the red
1 solution contained a single V^Q band at 1597 cm". The MgX2 salts were precipitated by the addition of a small quantity of water (-0.1 mL), and the reaction mixture was taken to
dryness. The residue was extracted with 2:5 CH2Cl2/hexanes (2 x 70 mL), and the extracts were filtered through a short column of Celite (50 x 20 mm) supported on a frit. The combined filtrates were reduced in volume in vacuo to afford a saturated solution which 99 was placed in a freezer (-25 °C) to induce crystallization. The solution deposited large red crystals (e.g. 2x2x3 mm) of Cp*Mo(NO)(Bz)(Tms) (0.730 g) which were isolated by removal of the supernatant solution via cannula and washing the crystals with a small amount of hexanes (10 mL). The supernatant solution was again concentrated, and an additional two crops of Cp*Mo(NO)(Bz)(Tms) as large red crystals were collected (0.615 g,
82% overall yield). A crystal suitable for X-ray crystallographic analysis was chosen from the Cp*Mo(NO)(Bz)(Tms) product isolated in this manner.
The analogous CpW(NO)(Bz)(Tms) (65%), Cp*W(NO)(Bz)(Tms) (86%), and
CpMo(NO)(Bz)(Tms) (31%) complexes were isolated similarly as large red crystals in the indicated yields. The characterization data for all these complexes are collected in Tables
4.1 and 4.3.
Preparation of Cp*M(NO)(p-Xyl)(Npt). The complexes Cp*Mo(NO)(p-Xyl)(Npt) and Cp W(NO)(p-Xyl)(Npt) were prepared in a similar manner. The procedure for the tungsten complex is given as a representative example.
A sample of (Npt)2Mg(dioxane)210 (0.54 g, 170 mg/mmol Npt", 3.2 mmol), was
weighed into a Schlenk tube in a glovebox. The solid was dissolved in Et20 (20 mL), and this solution was then cannulated into an orange suspension of Cp*W(NO)(/?-Xyl)(Cl)
(2.4 g, 2.5 mmol) in Et20 (60 mL). The reaction mixture was stirred for 2 h during which time the suspended organometallic reactant was consumed. The final orange reaction
mixture was treated with H20 (0.4 mL), stirred for 10 min, and then reduced in vacuo to a solid residue. The residue was taken up in 1:1 CH^C^/hexanes (30 mL) and filtered through a column of alumina (activity UJ, 20 x 70 mm) supported on a frit. An orange filtrate was collected by washing the column with a 1:1 CH^C^/hexanes mixture (60 mL).
This solution was concentrated under reduced pressure until a precipitate began to form and was placed in a freezer (-25 °C) to induce the crystallization of product.
An orange microcrystalline solid was collected by removal of the supernatant solution by cannulation and washing the solid with hexanes (20 mL). The solid was dried in vacuo for 2 h to obtain 0.45 g of Cp*W(NO)(p-Xyl)(Npt) (34% yield). The supernatant 100 and washing solutions were combined, reduced in volume and returned to the freezer to induce the deposition of an additional 0.14 g of product (0.59 g total, 45% overall yield)
The preparation of the Mo congener on a similar scale gave Cp*Mo(NO)(p-Xyl)(Npt) as a feathery orange solid in 54% yield. The characterization data for both complexes can be found in Tables 4.1 and 4.3.
Preparation of Cp*Mo(NO)(p-Xyl)(p-Tol). Solid samples of Cp*Mo(NO)(Bz)(Cl)
10 (0.40 g, 1.0 mmol) and (p-Tol)2Mg(dioxane)2 (0.26 g, 175 mg/mmol p-ToY, 0.75 mmol) were weighed into a Schlenk tube in a dry box. The tube was removed and cooled to 10 °C
for 10 minutes. Then Et20 (45 mL) was rapidly syringed into the tube, and the mixture was stirred overnight at ambient temperature. The final orange reaction solution was then treated with water (0.1 mL) and stirred for 5 min to destroy the excess Grignard reagent. The reaction mixture was dried in vacuo to yield a sticky orange residue, which was then
dissolved in 2:1 hexanes/CH2Cl2 (60 mL) and filtered through an alumina column (activity
III, 20 x 70 mm). The column was then washed with 1:1 hexanes/CH2Cl2 to complete the removal of a single orange band. The combined filtrates were taken to dryness in vacuo to obtain an orange glass which was dissolved in hexanes (20 mL). Cooling of this solution to -78 °C resulted in the precipitation of an orange solid. The cooling bath was removed and the solvent was then slowly removed under reduced pressure as the solution warmed up to obtain Cp*Mo(NO)(/>Xyl)(/>Tol) as an orange microcrystalline solid (0.365 g, 80% yield). The physical and spectroscopic properties of this complex can be found in Tables 4.1 and 4.3. 101 Table 4.1. Infrared and Elemental Analysis Data for the New Bis(benzyl), Benzylchloro and Benzylalkyl Complexes
Compound IR (J^NO) Analysis Found(Calcd)
cm"1 % NujolMull Solution C H N
CpW(NO)(Bz)(Cl)c 1597 1597* 35.29 (35.54) 2.95 (2.98) 3.41 (3.45)
Cp*W(NO)(Bz)(Cl)d 1580 1582° 42.92 (42.92) 4.63 (4.66) 2.88 (2.94)
CpMo(NO)(Bz)(Cl) 1620 1627° 45.11 (4538) 3.96 (3.81) 4.39(4.41)
Cp*Mo(NO)(Bz)(Cl) 1609 1611° 52.89 (52.66) 5.78 (5.72) 3.66 (3.61)
E Cp*W(NO)(p-Xyl)2 1561,1541 1550° 55.74 (55.82) 5.93 (5.94) 2.51 (2.50)
Cp*W(NO)(p-Xyl)(Cl) 1568 1578° 43.95 (44.15) 4.92 (4.94) 2.82 (2.86)
Cp*Mo(NO)(p-Xyl)2 1580,1562* 1575° 66.06 (66.23) 6.96 (7.05) 3.04 (2.97)
Cp*Mo(NO)(p-Xyl)(Cl) 1612 1607° 53.68 (53.81) 5.90 (6.02) 3.60 (3.48)
CpW(NO)(Bz)(Tms) 1554 1593FE 41.76 (42.02) 5.07 (5.07) 3.09(3.06)
Cp*W(NO)(Bz)(Tms) 1549 1576& 48.00 (47.82) 6.29 (6.31) 2.70(2.65)
CpMo(NO)(Bz)(Tms) 1578 1613FE 51.88 (52.02) 6.35 (6.28) 3.83 (3.79)
Cp*Mo(NO)(Bz)(Tms) 1570 1597* 57.15 (57.39) 7.60 (7.57) 3.22 (3.19)
Cp*W(NO)(/P-Xyl)(Npt) 1541 1578* 52.38 (52.58) 6.84 (6.72) 2.71 (2.66)
Cp*Mo(NO)(p-Xyl)(Npt) 1599,1560* 1595^ 62.95 (63.14) 8.22 (8.06) 3.30 (3.20)
Cp*Mo(NO)(p-Xyl)(p-Tol) 1590 1603B 65.64 (65.64) 6.86 (6.83) 2.94 (3.06)
CH2C12
Et20 Cl; Found - 8.86; Calcd - 8.74. Cl; Found - 737; Calcd - 7.45. Two peaks of similar intensity are observed, presumably corresponding to the mull spectrum and a Nujol solution spectrum due to some solubility of the complex in Nujol. 102 Table 42. Mass Spectral and *H and "Ci1!!} NMR Data for the Bis(benzyl) and Benzylchloro Complexes
Low-resolution *H NMR "C^HINMR mass spectrum0 (C6D6) (CDCI3)
c Complex m/zb S S [VCH, Hz]
CpW(NO)(Bz)(Cl) 405, [P]+ 7.57 (t,lH,/>-Ar/f, 137.62 (Cpara, d, [164]) 375,[P-NO]+ VHH = 75 Hz) 135.35 (Carom, d, [163]) 6.88 (d, 2H, 0-A1H, 129.68 (Carom, d, [164]) VHH = 7.5 Hz) 109.62 (Cjpso)
6.73 (t, 2H, m-ArH, 101.14 (C5H5) d, [180])
3/HH = 7.5 Hz) 42.89 (CH2, t, [149])
D 5.00 (S)5H,C5H5)
3.15 (d, 1H, CHAHB, 2/HH = 6.0 Hz)
2.70 (d, 1H, CHA#B, 2/HH = 6.0 Hz)
Cp*W(NO)(Bz)(Cl) 475, [P]+ 7.55 (t, lH,p-AiH, 137.83 (Carom, d, [161]) 445, [P-NO]+ 3/HH = 7.5 Hz) 133.62 (Cpara, d, [161]) 6.88 (d, 2H, 0-A1H, 129.33 (Carom, d, [161]) 3/HH = 7.7 Hz) 112.63 (Cjpso)
6.76 (t, 2H, m-AiH, 110.08 (C5(CH3)5)
3/HH = 7.5 Hz) 48.20 (CH2, t, [147],
3.20 (d, 1H, C//AHB, Vwc = 53 Hz)
2/HH = 6.0 Hz) 10.49 (C5(CH3)5, q, [128])
2.14 (d, 1H, CHA#B, 2/HH = 6.0 Hz)
1.58 (s, 15H, C5(Cr73)5)
CpMo(NO)(Bz)(Cl) 319, [P]+ 7.54 (t, m,p-AxH, 136.70 (Cortho/ 289, [P-NO]+ 3/HH = 7.5 Hz) 135.45 (Carom, d, [161]) 6.83 (t, 2H, m-AiH, 129.40 (Carom, d, [161]) 3/HH = 7.5 Hz) 111.27 (Cjpso) 6.70 (br s, 2H, o-AxH) 102.45 (C5H5, d, [179])
d 5.03 (s, 5H, C5/75) 47.51 (CH2> t, [149])
3.25 (d, 1H, C/f AHB, 2/HH = 5.0 Hz) 103
2.73 (d, IH, CHA/^B, 2/HH = 5.0 Hz)
Cp*Mo(NO)(Bz)(CI) 389, [P]+ 7.51 (t, lH,p-ArH, 137.18 (Cortho, br d, [152]) 359,[P-NO] + D3.84 (Cpara, d, [153]) 3/HH = 6.1 Hz) 6.88 (t, 2H, m-AxH, 12930 (Cmeta, d, [160]) 114.78 (Cjpso) 3/HH = 6.1 Hz)
6.68 (br s, 2H, o-AxH) 111.29 (C5(CH3)5)
3.29 (d, IH, CHAHB, 51.78 (CH2, dd, [147,151])
10.61 (C5(CH3)s, q, [127]) 2/HH = 4.6 Hz)
2.07 (d, IH, CHA#B, 2/HH = 4.6 Hz)
1.41 (s, 15H, C5(Ctf3)5)
Cp*W(NO)(p-Xyl)(Cl)e
+ 489, [P] 6.84 (d, 2H, Ari/AHB) 145.89 (Cpara) 459, [P-NO] + 138.09 (Cortho) 3/HH = 7.5 HZ) 6.68 (d,2H,ArHA#B, 13039 (Cmeta) 109.47 (Cjpso) 3/HH = 7.5 Hz) 3.21 109.20 (d, IH, CHAHB, (C5(CH3)5) 47.83 (CH ) 2/HH = 7.3 Hz) 2
2.20 (s, 3H,ArCtf3) 22.26 (ArCH3)
2.18 (d, IH, CHA^B, 10.26 (C5(CH3)5) 2/HH = 7.3 Hz)
1.64 (s, 15H, C5(Cff3)5)
Cp*Mo(NO)(p-Xyl)(Cl)e 403, [P] + 6.80 (d, 2H, AT/ZAHB, 146.14 (Cpara) 373, [P-NO] + 137.80 (Cortho) 3/HH = 8.1 Hz) 6.66 130.16 (br s, 2H, ArHA#B) (Cmeta) 3.30 (d, IH, CHAHB, 111.11 (Cjpso) 111.00 (C (CH ) ) 2/HH = 5.0 Hz) 5 3 5
2.20 (s, 3H, ArO/3) 51.70 (CH2)
2.10 (d, IH, CHAi/B, 22.18 (ArCH3) 10.38 (C (CH ) ) 2/HH = 5.0 Hz) 5 3 5
1.55 (s, 15H, C5(0/3)5) 104
Cp*W(NO) + 529,[P-NO] VHH = 6.9 Hz) 132.49 (Carom) 6.79 (d, 4H, ATHAHB, i30.32 (Cjpso) VHH = 6.9 Hz) 129.18 (Carom) 2.31 (d, 2H, CKAHB, 107.06 (C5(CH3)5) VHH = 9.0 Hz) 42.46 (CH2) 2.02 (s,6H,ArCff3) 2130 (ArCH3) 1.54 (s, 15H, C5(Cff3)5) 9.96 (C5(CH3)5) 0.74 (d,2H,CHAffB, VHH = 9.0 Hz) € Cp*Mo(NO)(p-Xyl)2 473, [P] + 6.81 (m,8H,Ar#) 137.20 (Cpara) 443, [P-NO] + 2.38 (d, 2H, CHAHB, 131.55 (Cjpso) VHH = 7.2 Hz) 131.44 (Carom) 2.09 (s,6H,ArCff3) 128.81 (Carom) 1.51 (s, 15H, C5(Cff3)5) 108.00 (C5(CH3)5) 0.57 (d, 2H, CHA#B, 43.93 (CH2) VHH = 7.2 Hz) 21.40 (ArCH3) 10.08 (C5(CH3)5) a Probe temperatures 120 °C. ^ The m/z values are for the highest intensity peak of the calculated isotopic cluster. c Values of VCH were determined using a gated decoupled pulse sequence. ^ Relaxation delays (10-20 s) required to observe full integrations of Cp proton signals. e "C^H} NMR spectrum recorded in Q5D6. f The low solubility of CpMo(NO)(Bz)(Cl) in CDCI3 and the very broadened nature of this signal make its assignment tentative at best. 105 Table 4.3. Mass Spectral and XH and 13C{1H} NMR Data for the Chiral Benzylalkyl Complexes Low-resolution !H NMR 13C{1H}NMR 0 mass spectrum (C6D6) (CDC13) Complex m/zb 6 6 IA/CH, Hz]c CpW(NO)(Bz)(Tms) 457, [P]+ 7.58 (t, m,p-ArH, 135.27 (Cortho, d, [161]) 442,[P-CH]+ 132.00 (Cpara, d, [161]) 3 3/HH = 75 Hz) 427,[P-NO]+ 6.65 (d, 2H, o-AiH, 129.44 (Cmeta, d, [161]) 111.51 (Cjpso) 3/HH = 7.7 Hz) 6.51 (t, 2H, m-AiH, 99.46 (C5H5, d, [179]) 38.91 (ArCH , t, [147], 3/HH = 75 Hz) 2 5.01 (s, 5H, CsHsf Vwc = 46 Hz) 3.09 (d,lH,ArC/7AHB, 13.84 (SiCH2, t, [112], 2/HH = 5.8 Hz) Vwc = 77 Hz) 2.60 (d,lH,ArCHAtfB, 2.59 (S^CHsH q, [118]) 2/HH = 5.8 Hz) 0.22 (s, 9H, Si(Ctf3)3) -0.19 (d, 1H, SiCtfAHB, 2/HH = 12.9 Hz) -3.68 (d, 1H, SiCHAtfB, 2/HH = 12.9 Hz) Cp*W(NO)(Bz)(Tms) 527, [P] + 7.50 (t, lH,p-AiH, 134.65 (Carom) 512, [P-CH]+ 130.41 (Cpara) 3 3/HH = 7.5 Hz) 497, [P-NO]+ 6.62 (t, 2H, m-AiH, 129.52 (Carom) 115.54 (Cjpso) 3/HH = 7.5 Hz) 6.54 (d, 2H, o-AiH, 108.27 (C5(CH3)5) 47.05 (ArCH ) 3/HH = 7.5 Hz) 2 3.25 (d,lH,Ara/AHB) 20.02 (CH^i) 10.51 (C (CH ) ) 2/HH = 5.0 Hz) 5 3 5 2.10 (d, 1H, ArCHAtfB) 3.05 (Si(CH3)3) 2/HH = 5.0 Hz) 1.55 (s, 15H, C5(0/3)5) 0.32 (s,9H,Si(CJ/3)3) -0.73 (d, 1H, SiCifAHB, 2/HH = 13.5 Hz) -4.06 (d, 1H, SiCH^B, 2/HH = 13.5 Hz) 106 CpMo(NO)(B2)(Tms) + 371, [P] 7.39 (t, m,p-AiH, 134.10 (Cortho) + 3 356,[P-CH3] /HH = 7.7 Hz) B1.87 (Cpara) + 341, [P-NO] 6.62 (t,2H,m-ArJJ, 128.56 (Cmeta) 3/HH = 7.7 Hz) 11237 (Cjpso) 6.26 (br s, 2H, o-AiH) 99.89 (C5H5) d 5.00 (s,5H,CsHs) 39.16 (ArCH2) 3.11 (d, lH,ArCffAHB, 15.93 (SiCH2) VHH = 43 Hz) 2.40 (Si(CH3)3) 2.60 (d, IH, ATCHA^B, VHH = 43 Hz) 0.26 (s,9H, Si(Cff3)3) 0.14 (d, IH, SICHAHB, VHH = 12.8 Hz) -4:15 (d, IH, SiCHA#B, VHH = 12.8 Hz) Cp*Mo(NO)(Bz)(Tms)e + 441, [P] 7.36 (t, m,p-AiH, 133.84 (Cortho) + 426,[P-CH3] 3/HH = 7.5 Hz) 130.54 (Cpara) + 411,[P-NO] 6.66 (t, 2H, m-AiH, 128.63 (Cmeta) 3/HH = 73 Hz) 115.72 (Cjpso) 6.20 (d, 2H, o-Arif, 108.39 (C5(CH3)5) 3/HH = 7.5 Hz) 46.09 (ArCH2) 3.23 (d,lH,ArCi/AHB) 19.52 (SiCH2) VHH = 4.5 Hz) 10.60 (C5(CH3)5) 2.00 (d,lH,ArCHA/7B) 3.77 (Si(CH3)3) VHH = 4.5 Hz) 1.53 (s, 15H, C5(Cff3)5) 0.30 (s,9H, Si(Ctf3)3) -0.64 (d, IH, SiCHAHB, VHH = 13.1 Hz) -4.49 (d, IH, SiOWfa, VHH = 13.1Hz) Cp*W(NO)(p-Xyl)(Npt)e + 525, [P] 6.63 (s,4H,Ar#) 140.44 (Cpara) + 493,[P-NO-2H] 3.24 (d, IH, ArCHAHB) 133.76 (Carom) VHH = 7.0 Hz) 130.59 (Carom) 2.17 (cUH.ArCH^e, 118.32 (Cjpso) VHH = 7.0 Hz) 107.72 (C5(CH3)5) 2.03 (s, 3H, ArCff3) 6530 (CCH2) 1.58 (s, 15H, C5(Ci/3)5) 48.51 (ArCH2) 1.35 (d, IH, CCffAHB, 37.57 (Cquat) VHH = 13.7 Hz) 34.81 (C(CH3)3) 107 1.20 (s, 9H, C(C#3)3) 21.88 (ArCH3) -2.41 (d, IH, CCHA^B, 10.29 (C5(CH3)5) 2/HH = 13.7 Hz) Cp*Mo(NO) (p-Xyl) (Npt)c + 439, [P] 659 (d, 2H,ArHAHB, 142.10 (Cpara) 407,[P-NO-2H] + 3 133.76 /HH = 7.6 Hz) (Cortho) 6.18 (d,2H,ArHAtfB, 129.85 (Cmeta) 114.07 3/HH = 7.6 Hz) (Cjpso) 333 (d, IH, ATCHAHB, 108.12 (C5(CH3)5) 57.66 (CCH ) 2/HH = 4.8 Hz) 2 2.13 (s, 3H, ArCr/3) 45.70 (ArCH2) ^ 2.00 (d, IH, ATCHAHB, 37.06 (Cquat) 34.64 (C(CH ) ) 2/HH = 4.8 Hz) 3 3 138 (s, 15H, C5(Ci/3)5) 21.78 (ArCH3) 1.17 (s,9H, C(Cff3)3) 10.55 (C5(CH3)5) 0.94 (d, IH, CCHAHB, 2/HH = 12.5 Hz) -2.88 (d, IH, CCHA#B, 2/HH = 12.5 Hz) Cp*Mo(NO)(p-Xyi)(p-Tol)e • 459, [P] + 6.88 (d, 2H, AiH, 173.14 (Cjpso, p-Tol) 429, [P-NO] + 142.77 (Cpara, p-Xyl) 3/HH = 7.5 Hz) 6.2-6.7 (m, 6H, AiH) 138.70 (Cortho'P-Xy1)) 3.42 (d, IH, ATC/TAHB, 134.69 (Cmet3i,p-Xy\) 132.94 (Cpara, i»-Tol) 2/HH = 4.5 Hz) 2.23 (s,3H,ArCff3) 129.58 (Carom, P-Tol) 127.19 2.12 (d, 1H, ArCHA#B, (Carom, />-Tol) 112.12 (Cjpso, ^-Xyl) 2/HH = 4.5 Hz) 1.88 (s,3H,Ara/3) 109.02 (C5(CH3)5) 1.54 (s, 15H, C5(Cff3)5) 47.65 (ArCH2) 21.71 (ArCH3) 21.39 (ArCH3) 10.53 (C5(CH3)5) a Probe temperatures 120 °C. * The m/z values are for the highest intensity peak of the calculated isotopic cluster. c Values of VCH were determined using a gated decoupled pulse sequence. ^ Relaxation delays (10-20 s) required to observe full integrations of Cp proton signals. e "Cf^H} NMR spectrum recorded in CoT>6. 108 Results and Discussion Preparation and Properties of Cp*M(NO)(/>Xyl)2. These complexes are prepared in the same manner as the related bis(benzyl) complexes,11 i.e., 1) Et2o Cp*M(NO)(Cl)2 + 2 /?-XylMgCl • Cp*M(NO)(p-Xyl)2 + MgCl2 (4.6) 2) H20 The liberation of the final bis(p-xylyl) products from intermediate MgCl2 adducts in reactions 4.6 is accomplished with water, as has been used in preparations of other bis(alkyl) complexes.3 The bis(/>xylyl) complexes are notably more soluble than their bis(benzyl) analogues due to the methylation of the aromatic rings. As solids, they can be handled in air for purposes of transfer and weighing without noticeable decomposition. When stored under dinitrogen, they appear to be indefinitely thermally stable, as are the analogous bis(benzyl) complexes. The IR and mass spectroscopic data (Tables 4.1 and 4.2) are consistent with their formulations and are comparable to those exhibited by the bis(benzyl) analogues. Prior to the inception of this work, the series of bis(benzyl) complexes, Cp'M(NO)(Bz)2, had been structurally characterized and found to be a unique group of the larger class of compounds, Cp'M(NO)(R)2. Their uniqueness stems from the fact that they all contain one benzyl ligand coordinated in an r?2 fashion as a 3-electron donor, i.e., M A to give 18-electron complexes. These original bis(benzyl) complexes, Cp'M(NO)(Bz)2, are also fluxional in solution via a process that changes the benzyl group which is coordinated in an ^-fashion to the metal centre. 109 This process causes the *!! and ^C^H} NMR signals of both the r/1- and r?2-benzyl ligands to be averaged, resulting in one set of benzyl resonances. Both of the newly J 13 1 prepared Cp*M(NO)(p-Xyl)2 species also exhibit signals in their H and C{ H} NMR spectra for one type of /?-xylyl ligand (Figure 4.1). This apparent symmetry is seen most dramatically in the *H NMR spectra of these complexes, a single distinctive AB doublet pattern diagnostic of a para disubstituted benzene ring being evident. This pattern is useful in following subsequent chemistry since it simplifies an otherwise crowded aromatic region in the 1H NMR spectra. The methylene proton resonances of the bis(/>xylyl) complexes occur at values that closely correspond to the values observed for the methylene protons of the bis(benzyl) complexes. Preparation and Physical Properties of Cp'M(NO)(CH2Ar)(Gl). It was hoped that the ability of a benzyl ligand to provide extra electron density to the metal centre, via the novel r?2-coordination mode, would allow the isolation of a stable complex, namely 2 Cp'M(NO)(r? -CH2Ar)(Cl). This hypothesis was strengthened by the knowledge that similar species among the known CpM(NO)(r;3-allyl)(I) (M = W,13 Mo14) complexes had been prepared previously. The apparently straightforward route to these complexes, namely selective alkylation with one equivalent of a Grignard reagent, was not attempted due to the poor results from similar reactions noted in the Introduction.15 Therefore, the reaction of the bis(benzyl) complexes with HQ was attempted to try to form the benzyl chloro species. 110 (a) JU A i—i—I i i—i—i—I—i—i—i—i—i—r i—i—i—j—i—i—i—i—|—i—i—i—i—|—r—i—i—i—|—r 7.0 6.0 5.0 4.0 3.0 2.0 1.0 6 0.0 PPM (b) r I r —i • r— 1 140 120 100 BO 60 40 20 6 0 PPM J Figure 4.1. (a) The 200 MHz H NMR spectrum of Cp*W(NO)(p-Xyl)2 in QDgf.). n l (b) The 50 MHz C{U} APT NMR spectrum of Cp*W(NO)(/>Xyl)2 in Cfi>6(*). The carbon signals for Cquat and CH2 groups are positive and the carbon signals for the CH and CH3 groups are negative). Ill These reactions are successful, the complexes being prepared most conveniently with an excess of HC1 (equation 4.7). xsHCl Cp'M(NO)(CH2Ar)2 • Cp'M(NO)(CH2Ar)(Cl) + ArCH3 (4.7) The progress of reaction 4.7 is easily followed by IR monitoring of the solutions since the product benzylchloro complexes exhibit VNO absorptions (Table 4.1) approximately 30 cm*1 higher in energy than the bis(benzyl) starting materials. This shift in v^o 1S m accord with a reduced electron density on the metal centre available for the M d ?r-»NO n backbonding, the result of replacing a benzyl group with a more electronegative chlorine atom. In the Cp'M(NO)(CH2Ar)(Cl) complexes, the v^o hands of the Cp*-containing species are ~ 15 cm"1 lower than those for the Cp-containing ones, due to the greater electron-donating ability of the Cp* ligand. As well, the vNO bands of the tungsten complexes are ~30 cm"1 lower in energy than those for the molybdenum complexes. Similar trends in the J^NO °f me starting bis(benzyl) complexes and of other classes of compounds containing Cp'M(NO) moieties prepared in this research group have been noted.11'16 As noted in the Experimental Section, there is no deleterious effect in using excess HC1 to effect conversions 4.7. Selective hydrochlorination of an alkyl ligand has been used to prepare other mixed alkylhalonitrosyl species, e.g., acetone Cp*Ru(NO)(Ph)2 + HCl(aq) » Cp*Ru(NO)(Ph)(Cl) + PhH (4.8) but in this reaction the stoichiometry of the HC1 source must be carefully controlled to 17 avoid formation of Cp*Ru(NO)(Cl)2. The Cp'M(NO)(CH2Ar)(Cl) complexes are orange to red-orange solids which are soluble in polar organic as well as aromatic solvents, but to a lesser degree than their corresponding bis(benzyl) precursors. The benzylchloro compounds are markedly less 112 soluble in aliphatic solvents, with only the (^-containing members showing any appreciable solubility. They are air stable as solids, enabling their handling without an inert atmosphere for the purposes of weighing and transfer. When stored under N2 at 10 °C in the dark, they remain unchanged over a period of at least 6 months. The mass spectra (Table 4.2) of all of these complexes are simple, each displaying a highest mass isotopic cluster of signals corresponding to a parent ion of composition + [Cp'M(NO)(CH2Ar)(Cl)] . The expected monomelic nature of these complexes is consistent with this result. Daughter peaks due to loss of NO are noted for all of these complexes, an observation which is common for organometallic nitrosyl complexes.18 Preparation and Properties of the Complexes Cp'M(NO)(CH2Ar)(R) (R = Tms, Npt,j?-ToI). The reactions of the benzyl chloro species with the trimethylsilylmethyl Grignard reagent, e.g., 1) Et20 Cp'M(NO)(Bz)(Cl) + TmsMgCl + Cp'M(NO)(Bz)(Tms) + MgCl2 (4.9) 2) H20 proceed quickly and cleanly, as indicated by the IR spectra of the final reaction solutions which reveal a single strong absorbance due to the V^Q of the product (Table 4.1). In no case is there any indication for the formation of a MgCl2 adduct involving the nitrosyl ligand of the bis(benzyl) complexes. It appears that the formation of these adducts may be due to the synthetic route used to prepare the bis(alkyl) species, since they are seen in most of the bis(alkyl) (Chapter 2) and bis(benzyl) (vide supra) preparations via alkylation of the dihalonitrosyl starting materials. The complexes formed via reactions 4.9 are thermally stable as solids and can be handled in air without any noticeable decomposition. The Cp'M(NO)(Bz)(Tms) compounds all exhibit a single strong nitrosyl absorption in their IR spectra between 1576- 1 1613 cm* (Table 4.1), these values corresponding closely with the yNO parameters seen in the bis(benzyl) analogues. This is not surprising since the electron-donating ability of simple a alkyls would not be expected to vary widely, leaving the metal with essentially the 113 same electron density in related Cp'M(NO)(Bz)2 and Cp'M(NO)(Bz)(R) complexes. In their low-resolution mass spectra (Table 4.3), all of the complexes exhibit a highest mass isotopic cluster of signals assignable to the parent ion peaks of the respective Cp'M(NO)(Bz)(Tms) complexes. In the preparations of the 16-electron Cp'M(NO)(R)2 complexes, it has been observed that the trimethylsilylmethyl group generally results in higher yields of products and produces the most stable complexes wherever it has been used. The Cp'M(NO)(Bz)(Npt) derivatives have thus been synthesized to compare the relative ease of formation and ultimate stability of these neopentyl analogues of the Cp'M(NO)(Bz)(Tms) complexes. Reactions to form the neopentyl substituted products, i.e., 1) Et20 Cp*M(NO)(p-Xyl)(Cl) + 0.65(Npt)2Mg • Cp*M(NO)(p-Xyl)(Npt) + MgCl2 (4.10) 2) H20 result in isolation of the desired complexes in good yields as orange microcrystalline solids. The Cp M(NO)(/7-Xyl)(Npt) products are similar in their physical properties and in their spectroscopic parameters to the benzyl trimethylsilylmethyl complexes (vide infra). The only notable difference noted for the neopentyl derivatives, from the trimethylsilylmethyl analogues, is in the reaction conditions used for their preparation. Initial attempts at forming these non-trimethylsilylmethyl substituted complexes were disappointing, giving intractable oils. Utilization of longer reaction times, slight excesses of alkylating reagents, and chromatography on alumina give much better results. Therefore, the future preparation of other benzyl alkyl complexes is probably best accomplished using these general procedures. The aryl-substituted complex, Cp*Mo(NO)(p-Xyl)(p-Tol), is also preparable via a route similar to equation 4.10 and its stability mirrors that of the benzyl alkyl complexes. During the attempted preparation of the 16-electron bis(aryl) complexes, Cp'M(NO)(Ar)2, 114 (Chapter 2) it was observed that the bis(aryl) complexes are much less stable than their bis(alkyl) analogues, with only the Cp*W(NO)(Ar)2 complexes being isolable. This instability is not exhibited by Cp*Mo(NO)(p-Xyl)(p-Tol), a manifestation of the influence of the r?2-benzyl linkage which greatly stabilizes this general family of mixed alkyl complexes. The low-resolution mass spectra of the Cp*M(NO)(p-Xyl)(R) complexes (Table 4.3) exhibit highest mass isotopic clusters attributable to [Cp*M(NO)(p-Xyl)(R)]+ parent ions, consistent with the formulation of these compounds as 18-electron monomers. Molecular Structures of the Complexes Cp*Mo(NO)(Bz)(X) (X = Cl, Tms). Given the novel rj2-benzyl link found in the parent bis(benzyl) complexes, it was of interest to determine the nature of the metal-benzyl linkages in these new classes of benzyl- containing complexes. The single-crystal X-ray crystallographic analysis of a prototypal member, namely Cp*Mo(NO)(Bz)(Cl), has been performed by Vivian Yee of this Department,19 and it confirmed the anticipated r/2-coordination mode for the benzyl ligand. An ORTEP diagram of this complex is given in Figure 4.2 along with pertinent structural parameters in Table 4.4. The specific details are discussed below, but for now it is important to note that the M-CH2-Ph unit is isostructural with that found in the other r?2- benzyl structures previously determined.11 * The Cp Mo(NO)(Bz)(Tms) complex has also been analyzed by single-crystal X-ray crystallography,19 thereby allowing a comparison of the relative effect on the r?2-benzyl ligand of a trimethylsilylmethyl group versus a chlorine atom. The ORTEP plot of this alkyl complex is given in Figure 4.3 and its structural parameters are presented in Table 4.5. The Cp Mo(NO)(Bz)(Tms) complex contains an r/ -benzyl ligand identical to that extant in the chloro analogue. This indicates that the alkyl and chloro groups do not significantly change the nature of the interaction between the metal and the benzyl ligand in spite of the different steric and electronic effects provided by these groups. The intramolecular dimensions involving the Mo-Cp* portions of both molecules are normal and comparable to those found in related compounds.20 Figure 42. The ORTEP plot of the molecular structure of Cp*Mo(NO)(Bz)(Cl). Hydrogen atoms have been omitted for clarity. 116 Table 4.4 Selected bond distances (A) and bond angles (°) and torsion angles (°) for the molecular structure of Cp*Mo(NO)(Bz)(Cl). Mo-CP" 2.0404(12)A Mo- N-O 170.3(2)° Mo-N 1.778(2) Mo- Cll- C12 81.10(15) N-O 1.198(3) C12 -Cll- Hlla 118(2) Mo - Cll 2.193(3) C12 -Cll- Hllb 117.8(15) Mo - C12 2.439(2) Cll -Mo- C12 36.25(9) C11-C12 1.460(4) C13 -C12- Cll - Mo -91.0(2) C12 - C13 1.404(3) C17 -C12- Cll - Mo 84.5(2) C12 - G17 1.392(3) C13-C14 1.376(4) C14 - C15 1.375(4) C15 - C16 1.367(4) C16-C17 1.383(4) a 5 CP - centre of the r/ -C5Me5 ring. Figure 4.3. The ORTEP plot of the molecular structure of Cp*Mo(NO)(Bz)(Ti Hydrogen atoms have been omitted for clarity. 118 Table 4.5 Selected bond distances (A) and bond angles (°) and torsion angles (°) for the molecular structure of Cp*Mo(NO)(Bz)(Tms). 0 Mo-CP 2.058(1) A Mo- N-O 169.7(2)° Mo-N 1.760(2) Mo- C15 - C16 82.97(12) N-O 1.217(2) C16 - C15 - H15a 119(2) Mo - cn 2.218(2) C16 -C15-H15b 115(2) Mo - C16 2.473(2) C15 - Mo - C16 35.60(8) Mo - C15 2.188(2) C17 - C16 - C15 - Mo 82.3(2) C15 - C16 1.450(3) C21 - C16 - C15 - Mo -91.2(2) C16 - C17 1.397(3) C16 - C21 1.417(3) C17 - C18 1.375(4) C18-C19 1.372(4) C19 - C20 1.371(5) C20-C21 1.375(4) a 5 CP - centre of the rj -C5Me5 ring. 119 The M-N-O bond angles for both compounds are essentially linear, being 170.3(2)° for the chloro complex and 169.7(2)° for the alkyl complex. This feature is consistent with the NO ligand acting as a 3-electron donor in both of these compounds. In both of these complexes the Mo atom lies under the CH2-Ph bond axis, in a direction perpendicular to the plane of the benzyl ligand, Mo This orientation is best quantified by the dihedral angles Corth0-Qpso-Ci-Mo (0) and o C'ortho-Qpso-CrMo (tf). For the chloro species these are 84.5(2)°(0) and 91.0(2) (<£), respectively, and for the trimethylsilylmethyl analogue 82.3(2)°(0) and 91.2(2)0(tf). For comparison, in CpMo(r?3-p-Xyl)(CO)2,21 the dihedral angles are radically different, being 60.9°(e) and 111.59°(0), and reflect the asymmetric interaction of the benzyl ligand with the metal. The hapticity of the benzyl ligand is also characterized in a comparison of the metal-carbon bond lengths for both of these coordination modes. In both of the f?2-benzyl complexes, the molybdenum atom is within bonding distance of the methylene carbon (Mo-Q = 2.193(3), 2.188(2)) and the ipso carbon (Mo-Qpso = 2.439(2), 2.473(2)). However the molybdenum atoms are at least ~3.0 A away from either of the ortho carbons 2 of the r/-benzyl ligands, a separation which would appear to preclude a Mo-Cortno bond since a Mo-C single bond is estimated to be -2.39 A.22 In the ??3-benzyl complex the Mo-Cortho bond length (2.480(6) A), as well as the Mo-Cx (2.269(7) A) and Mo-Qpso (2.364(5) A) bond lengths, all agree reasonably well with this estimated Mo-C bond length. The aromatic ring systems are also significantly different in the rj2- and r/3-cases. In the classic r/3-case (vide supra), jr-bond localization is evident in the aromatic ring in short C3-C4 and C5-Q bonds and long C4-C5 and Q-C7 bonds. Thus the structure is best 120 represented as M C7 C6 C1 C2 C5 M = CpMo(CO)2 C3 C4 where the aromatic ring has a diene-like structure. However, in the r;2-benzyl ligands for both of the structures determined here, there is no evidence for bond localization. Indeed, the four C-C bond lengths in the C3-C4-C5-C6-C7 groupings are the same, suggesting delocalized multiple bonding over these carbons. The C2-C3 and C2-C7 bond lengths are noticeably longer, showing a reduced bond order between these carbon atoms. Therefore, the overall interaction is best represented in valence bond terms as M C7 C6 / \ / \\ M = Cp*Mo(NO)(X) C1 • C2 ) C5 \ / X = Tms, CI C3 C4 indicating that there has been a net donation of two electrons from the aromatic n system to the metal centre. This results in an overall three-electron donation, with the benzylic methylene radical supplying the other electron. The rj2-benzyl containing compounds are formulated as 18-electron species since they exhibit diamagnetic properties, e.g. unbroadened NMR spectra, so a three-electron donation from the r?2-benzyl ligand is required for these complexes. It should be noted however, that there is still a degree of aromatic character to the ring since the lengthened C2-C3 and C2-C7 bonds are still reasonably close to the value of 1.395 A expected for an aromatic C-C bond.23 In the structure of Cp*Mo(NO)(Bz)(Tms), the methylene group of the trimethylsilylmethyl ligand is located directly over the aromatic ring of the benzyl group. A 121 closer examination of the CH.2-Si group shows that it is oriented so as to place one proton over the plane of the benzyl aromatic ring, e.g., This proton lies almost over the exact centre of the six carbons, 2.50 A above the plane of the ring. This feature will be important in the discussion of the NMR spectroscopic properties of the Cp'M(NO)(CH2Ar)(R) (R = Npt, Tms) complexes presented later on in this Chapter. A theoretical paper has appeared recently dealing, in part, with the rotational barrier of the methylene group in and the optimized geometry of the benzyl radical.24 The rotational barrier is due to a partial double bond resulting from a bonding interaction between the methylene pn- orbital and the aromatic * system. The major form of this interaction can be represented by a simple orbital diagram, i.e. CH2 p?r orbital ring pff MO where the singly occupied p* orbital and the filled HOMO of the aromatic ring interact. It is tempting to speculate that the interaction between the r/2-benzyl ligands and molybdenum, as seen in the crystal structures of Cp*Mo(NO)(Bz)(Cl) and 122 Cp*Mo(NO)(Bz)(Tms), arises from a three-electron donation from the MO's in the above diagram to the metal centre. Thus, the r/2-benzyl-Mo bonding may be viewed as a benzyl radical stabilized by * -coordination to a metal centre. Consistent with this view, in both of the r/2-benzyl complexes there is some multiple bond character between the methylene and ipso carbons. This is reflected in the QpSO-Qmethylene bond distances (1.460(3) A and 23 1.450(3) A) being significantly shorter than a CH3-Caromatic single bond (1.53 ± 0.01 A). It is interesting to compare the optimized geometry of a free benzyl radical24 with the geometry determined for the bound r/2-benzyl ligands. The bond distances found for the »?2-benzyl groups and those calculated for the benzyl radical are listed in Table 4.6. The general trends predicted for the aromatic ring bond distances are similar to those found in the structures, the Qpso-Cortho bonds being longer than the other ring C-C bonds. There is a large difference in the predicted methylene to ring bond distance for the radical and that found in r/2-benzyl ligands. If the two frontier orbitals of the benzyl radical are indeed donating electrons to the metal, then a reduced methylene-ring interaction would be expected and a lengthened bond should result. Therefore, the interpretation of a metal- stabilized„benzyl radical is at least consistent with the observed structures. Apparently similar symmetric metal-benzyl interactions have been observed for some Group 4 25-28 and actinide 29,30 benzyl complexes, these metal-benzyl interactions being variously described as r;4,30 r;n,29and r?2.28 The f?2 description invoked for [Cp2Zr(r/2-Bz)]+ 28 did not discuss the nature of the bonding, but assumed it to be the same as that found in the above-mentioned r/4 and r/n configurations. In all of these cases, the interaction is rationalized on the basis of the predicted Huckel MO 7r-electron density for the benzyl anion. These calculations predict that the net negative charge of the benzyl anion lies primarily on the methylene carbon (4/7) with the rest being on the ortho carbons (1/7 each) and the para carbon (1/7).31 Necessarily, these calculations also predict no charge on the ipso and meta carbons. The assignment of the r?4 and riD configurations, then, seems to stem from the apparent necessity to have M-Corth0 interactions for benzyl-* metal electron donation to occur. However, this interpretation implicitly assumes 123 Table 4.6 Comparison of Bond Lengths and Angles for the Benzyl Radical (Bz*) and the f?2-Benzyl Ligands. 6 a b c d (°) (A) (A) (A) (A) Cp*Mo(NO)(Bz)(Cl) 118 1.46 1.40 1.37 1.38 Cp*Mo(NO)(Bz)(Tms) 117 1.45 1.41 1.37 1.37 QH5CH2 121 1.40 1.43 1.39 1.40 c 124 that the total electron density donated to the metal is distributed throughout the benzyl anion in the same manner as the net negative charge. This would seem to be an oversimplification of the real situation. In all of these early transition-metal and actinide- containing species the molecular structures exhibit M-Cortii0 distances marginally small enough to consider a bonding interaction. The qualitative metal-benzyl bonding description proposed above, namely a metal-stabilized benzyl radical, eliminates the need to invoke metal-Cortho bonding interactions and gives some indication as to how it can be viewed as a three-electron interaction. A reexamination of these structures may show they can be considered as examples of the same r?2-bonding mode observed for the Cp'M(NO)(CH2Ar)(X) complexes discussed in this Chapter. Given the very similar ligand 2 3 21 sets of the related Cp*Mo(NO)(r?-Bz)(X) and CpMo(CO)2(r> -/>-Xyl) complexes, the 3 factors which favor an rj2- or r?-benzyl ligand, at least in these cases, would appear to be electronic in nature and possibly a theoretical study of these complexes could reveal the source of this effect. As a final point, the methylene groups of the r?2-benzyl ligands in Cp Mo(NO)(Bz)(Tms) and Cp Mo(NO)(Bz)(Cl) are approaching a geometry indicative of sp2 hybridization. This fact is best demonstrated by the structurally determined C-C-H bond angles, which are approaching the 120° value expected for an sp2-hybridized carbon atom. This hybridization is reflected in the characteristic NMR spectroscopic values found for the methylene groups discussed in the next sections. NMR Spectroscopic Properties of the Cp'M(NO)(CH2Ar)(Cl) complexes. An examination of the 1H NMR spectroscopic data (Table 4.2) reveals that all the new benzylchloro complexes possess a set of methylene resonances with similar chemical shifts and 2/HH coupling constants. This similarity indicates that qualitatively all of the Cp'M(NO)(CH2Ar)(Cl) complexes are isostructural in solution. The proton chemical shifts of the »?2-benzyl methylene signals are located at a lower field than seen for methylene protons in the rj1 alkyls studied previously. This observation is consistent with 125 the methylene group being more alkenyl in character in the 172 bonding mode of the benzyl ligand. X The low-temperature H NMR spectrum of CpW(NO)(Bz)2 and the room- temperature XH NMR spectrum of CpW(NO)(r/2-Bz)(Cl) are presented in Figure 4.4. In the low-temperature *H NMR spectrum (Figure 4.4(a)), signals for a CpW(NO)(r72-Bz)(r?1-Bz) species are evident, including two sets of AB doublets corresponding to the methylene protons of each type of benzyl ligand present. The low field methylene proton signals (5 3.2,2.7; 2/HH = 5 Hz) are assigned to the methylene protons of the r?2-benzyl ligand of the CpW(NO)(r;2-Bz)(r71-Bz) species, due to their similarity to those for CpW(NO)(Bz)(Cl). The pair of high field AB doublets (s 1.4, -0.7; 2 /HH = 10 Hz) are then assigned to the ^-benzyl methylene protons of CpW(NO)(r?2-Bz)(r;1-Bz). The original assignments of the methylene proton signals in the 1 11 low-temperature H NMR spectra of the Cp'M(NO)(Bz)2 complexes were reversed. However, unambiguous examples of r/2-benzyl ligands in the Cp'M(NO)(Bz)(Cl) compounds allows the proper assignments to be made by similar comparisons. A closer examination of the observed 2/HH coupling constants for the methylene protons provides additional structural information about the nature of the r/2-benzyl group 2 in all of the Cp'M(NO)(CH2Ar)(Cl) complexes. In all cases this geminal /HH coupling constant (4 - 7 Hz) is smaller than observed previously for the diastereotopic methylene protons of the 16-electron bis(alkyl) complexes CpM(NO)(R)2 (M = Mo, R = Tms; M = W, R = Tms, Neo, Npt) or for the 77 1-benzyl ligands in the low-temperature 1H NMR 2 spectra of the Cp'M(NO)(Bz)2 compounds (8 -12 Hz). This corresponds to the r/-benzyl methylene fragment having a hybridization predominantly sp2 (cf. geminal H-H coupling 32 3 constants of -3 to +2 Hz for C=CH2 groups ) rather than sp (cf. geminal H-H coupling 33 constants of 12-15 Hz for R'RCH2 groups ) in character. The similarity in the *H NMR parameters of the methylene groups indicates that all of the Cp'M(NO)(CH2Ar)(Cl) complexes possess structurally similar r/2-benzyl linkages. 126 (a) PPM Figure 4.4. (a) The 80 MHz *H NMR spectrum of CpW(NO)(Bz)2 in CD2C12 (•) at -100 °C. X (b) The 80 MHz H NMR spectrum of CpW(NO)(Bz)(Cl) in C6D6(.) at 25 °C. 127 Since the Cp'M(NO)(CH2Ar)(Cl) complexes contain only one benzyl group, the assignments of the 1H and ^C^H} NMR signals for the »j2-benzyl ligands in the aromatic region are straightforward. At ambient temperatures, the tungsten complexes exhibit XH NMR signals for the aromatic ring protons indicative of a symmetrically disposed benzyl ligand (Figure 4.5(a)). This apparent symmetry is also reflected in the observation of only four carbon resonances in the ^C^H} NMR spectrum for the aromatic ring (Figure 4.5(b)) of these species. This is surprising since in spite of the symmetric r/2-benzyl tungsten interactions, these tungsten centres are asymmetric and this should result in two different chemical shifts each of the ortho and meta aromatic proton signals. It is possible that the apparent symmetry reflects coincidental overlap of the signals due to insufficient differences in the relative chemical environments of these protons. There is a marked difference in the appearance of the signals due to the ortho protons of the tungsten complexes and those of the molybdenum complexes. Although all of the Cp'M(NO)(Bz)(Cl) complexes display 1H NMR spectra with two triplets of relative intensity 1:2 for the meta and para protons in an apparently symmetric benzyl environment, the molybdenum-containing complexes display a broad singlet for the ortho proton signals (Figure 4.6(a)). This broadening effect is also seen in the ^C^H} spectra where the signals due to the ortho carbon atoms of the Mo complexes have noticeably wider lineshapes (Figure 4.6(b)). A similar comparison exists between the Mo and W congeners of the Cp*M(NO)(p-Xyl)(Cl) complexes. These observations are indicative of the molybdenum complexes being stereochemically nonrigid. A variable temperature 1H NMR study of the Cp*Mo(NO)(Bz)(Cl) complex (Figure 4.7) provides proof for this fluxional process. As can be seen in these spectra, as the sample is cooled from room temperature the aromatic signals first broaden to featureless humps and then begin to sharpen. The signal for the para proton (s 7.51) does not change significantly. However, the broad signal for the ortho protons in the room temperature spectrum (5 6.75) splits into two signals (6 6.18, -7.3 ) at -50 °C. These two 128 (a) —i 1 1 1 1 1 1 1 1 1 1 1 1 1— 140 130 120 1 10 100 90 80 70 60 50 40 30 20 c 10 PPM 0 Figure 4.5. (a) The 300 MHz *H NMR spectrum of Cp* W(NO)(Bz)(Cl) in QD^.). (b) The 50 MHz ^C^H} NMR spectrum of Cp*W(NO)(Bz)(Cl) in CDC13(.). 129 (a) | ' i ' i I I 6 111111111111111111111111111111111111 II II 11 II 11' 5 * 3 2 c 1 PPM (b) 1 1 1 l 1 l T T T | I |-T-T-| IT I '6|) ' "l"' ' " '4[(' ' ' 'I" " tdo' I J^T 1 T I | I 1 'J"! I T I T1"1' T ' J '» ' '" '"I 1 1 1 J Figure 4.6. (a) The 300 MHz H NMR spectrum of Cp*Mo(NO)(Bz)(Cl) in CDC13 (.). (b) The 75 MHz ^C^H} NMR spectrum of Cp*Mo(NO)(Bz)(Cl) in CDC13 (•)• 130 Figure 4.7. Variable temperature 80 MHz *H NMR spectra of Cp*Mo(NO)(Bz)(Cl) in CDC13(.). 131 signals are displaced from the original position of the ortho resonance in equal but opposite amounts. The same observation is seen for the meta proton signals (25 °C, s 7.17; -50 °C, S 7.07, -7.25). The low-field signals for the ortho protons and meta protons overlap and are difficult to distinguish, but the higher-field signals for both types of protons are unobscured. The observation of broadening for only the ortho proton signals of the Cp>Mo(NO)(CH2Ax)(Cl) complexes at room temperature can be explained from these data. In the XH NMR spectrum of Cp*Mo(NO)(Bz)(Cl) obtained at -50 °C, the difference in chemical shifts (AS) for the two types of ortho protons observed is large (> 1 ppm) in comparison with the difference in the two meta proton environments (0.17 ppm). Since the temperature of coalescence, Tc, is proportional to AS the ortho proton signals should coalesce at a much higher temperature than do the meta proton signals. It is probable that a similar fluxional process is present in the tungsten-containing • complexes but at faster rates. This process would explain the apparent symmetric benzyl environments in these complexes and faster rates could be the reason that the tungsten complexes display no appreciable broadening of signals due to the ortho groups in the benzyl ligands. There is no evidence for any change in the signals due to the methylene protons of Cp*Mo(NO)(Bz)(Cl) in the spectra recorded between -50 and 25 °C, a fact which necessitates a fluxional process which does not significantly change the nature of the r?2- benzyl linkage. A fluxional process that fits the variable temperature 1H NMR data is a motion which has the ring tilting between two limiting orientations. M This tilting would bring either side of the ring, in turn, closer to the metal centre. Since the metal centre is the source of asymmetry in the molecule, the groups closest to it should exhibit the largest differences in chemical environment. The conclusion that the metal 132 induces the most asymmetry is supported by the low temperature 1H NMR spectrum of Cp*Mo(NO)(Bz)(Cl) which reveals larger chemical shift differences between the two ortho protons than the two meta protons. The changing orientation of the benzyl ligand could result in an averaging of the environment for the aromatic ring and thus the seemingly symmetric environment. In this case, then, as the solution is cooled, the benzyl group adopts the most energetically favorable conformation leading to a static asymmetric benzyl ligand as is evident in the low temperature spectrum. No significant change in the methylene proton signals in the variable temperature study would be expected since the r/2- benzyl linkage would remain intact throughout this process. An alternative fluxional process which would explain the symmetric aromatic environment is a hapticity change followed by a rotation of the phenyl group, i.e, M H H C?V _ M V V2 V 3 34 as proposed for the fluxional behaviour of CpM(CO)2(r/ -benzyl) (M = Mo, W). However, as noted above, the necessity of a breakage of the r/2-benzyl linkage in this process makes it highly unlikely for these CpM(NO)(Bz)(Cl) complexes. A comparison of the NMR spectroscopic parameters for the r?2-benzyl ligand of the Cp'M(NO)(Bz)(Cl) complexes with those for the r/2-benzyl of the low-temperature 2 1 Cp'M(NO)(r/ -Bz)(r/ -Bz) species also seems to preclude any breakage of the Mo-Qpso bond (vide infra), adding additional support to the "tilting" process explained above. B 1 Firstly, from the low-temperature C{ H} NMR spectra of the Cp'M(NO)(Bz)2 1 complexes, the expected chemical shift for Qpso in an r/ -benzyl is above 145 ppm whereas that for an r/2-benzyl is in the range 110-120 ppm.11 Additionally, in the solid state CP 13 1 2 MAS C{ H} NMR spectrum of CpMo(NO)(Bz)2, the Qpso chemical shift for the r?- 11 benzyl is 113.8 ppm. These observations establish the expected chemical shift for Qpso of 133 2 a non-fluxional r/-benzyl metal interaction. Therefore, the chemical shifts for Qpso in the Cp'M(NO)(CH2Ar)(Cl) complexes of -110-113 ppm support the conclusion of a continuously intact r?2-benzyl. This conclusion is also supported by the proton-proton and carbon-proton coupling constants for the methylene groups (2/HH = 4-7 Hz; ^cu - 147- 151 Hz) in the Cp'M(NO)(Bz)(Cl) compounds. These coupling constants are comparable with those for the methylene group of the r;2-benzyl group in the low-temperature NMR 2 spectra of the Cp'M(NO)(Bz)2 complexes ( /HH = 4-6 Hz, VCH = 148-152 Hz). Any breakage of the M-Qpso bond in the fluxionality of the Cp'M(NO)(Bz)(Cl) complexes should result in these values reflecting some r/1-benzyl contribution. This would cause an increase in 2/HH ana" a decrease in 1JQH relative to the values found for the r;2-benzyl ligand in Cp'M(NO)(r/2-Bz)(r;1-Bz). Since these changes are not seen, it therefore seems unlikely that the fluxional processes involve breaking of the r/2-benzyl linkage. NMR Spectroscopic Properties of the Cp'M(NO)(CH2Ar)(R) complexes. Examination of the 1H and13 C{1H} NMR spectra of the alkyl benzyl complexes reveals that their benzyl ligands are very similar to those found in the corresponding benzyl chloro precursor complexes. The proton-proton coupling constants (2/HH = 4-7 Hz) for the methylene group and the chemical shifts for QpSO (111-115 ppm) are again diagnostic for an i72-benzyl. This similarity may be expected, given the close correlation between the solid-state molecular structures of Cp*Mo(NO)(Bz)(Cl) and Cp*Mo(NO)(Bz)(Tms) (vide supra), thereby again indicating that the ancillary chloro or alkyl ligands have little effect on the metal-benzyl interaction. The *H and13 C{1H} NMR spectra of the complexes again show symmetric environments for the benzyl ligands and broadening of the ortho signals for the CpMo(NO)(Bz)(Tms) complex. In the *H NMR spectra of Cp Mo(NO)(Bz)(Tms) (Figure 4.8(a)) there is no broadening of the ortho proton signals as seen for the Cp'Mo(NO)(CH2Ar)(Cl) and CpMo(NO)(Bz)(Tms) species. However, in the ^C^H} NMR spectra of Cp*Mo(NO)(Bz)(Tms) (Figure 4.8(b)) the signals for the ortho aromatic-ring carbons do exhibit a slight noticeable broadening. A similar situation exists in the JH andB C{1H} NMR spectra of Cp*Mo(NO)(/--Xyl)(Npt). TheJ H NMR 134 spectra of these Cp*Mo-containing complexes show no broadening of the ortho proton signals simply because the rate of fluxionality in these complexes has increased to the point where Tc has become lower than ambient temperature for the ortho proton signals. Since the chemical shift differences (in Hz) are much larger for ^C NMR signals than for JH B NMR signals, the Tc values are proportionately higher (vide supra) for the C NMR spectra and some broadening is seen. This shows that the Tc values for the fluxionality of these complexes are near room temperature and can vary between metal centres. This is again believed to be the reason for the apparently symmetric benzyl ligands in the tungsten complexes, with Tc for these complexes again being sufficiently below ambient temperature so no noticeable broadening of signals is observed. There are intriguing JH NMR chemical shifts for the methylene proton signals of the Tms and Npt groups in the benzyl alkyl complexes, a feature which is demonstrated by X Figure 4.8(a). Since these CH2 protons are diastereotopic, they exhibit two different H NMR signals and are coupled to one another. However, the observed chemical shifts for the higher field methylene proton in the Cp'M(NO)(Bz)(Tms) complexes (S -3.7 to -4.5) and the Cp*M(p-Xyl)(Npt) complexes (s -2.5 to -2.9) are quite unexpected, given the chemical shifts for similar protons in the previously prepared bis(alkyl) complexes.2'3,10 The solid-state molecular structure of Cp Mo(NO)(Bz)(Tms) provides a clue to the source of this distinctive chemical shift. The significant shielding of one of the methylene protons is a result of this proton lying directly over the ring of the benzyl group (vide supra), therefore experiencing a net shielding effect due to an aromatic ring current. The 3.3 ppm upfield shift for in Cp*Mo(NO)(Bz)(Tms) (5 -4.49) from the location of the highest-field proton signal in Cp*Mo(NO)(Tms)2 (6 -1.17) agrees very well with the theoretically calculated shielding of 3 ppm for a proton situated in such a position over a benzene ring.35 This good correlation indicates that the structure that exists in solution for Cp*Mo(NO)(Bz)(Tms) is very similar to the solid-state molecular structure. Furthermore, the observation of similar ring current effects for the methylene protons in all of the Cp'M(NO)(CH2Ar)(R) (R = Npt, Tms) complexes is evidence that all these complexes 135 (a) Figure 4.8. (a) The 300 MHz *H NMR spectrum of Cp*Mo(NO)(Bz)(Tms) in C6D6 (.). (b) The 75 MHz ^C^H} NMR spectrum of Cp*Mo(NO)(Bz)(Tms) in C6D6 (.). Inset is an expansion of the aromatic region from 135 to 127 ppm. 136 possess a similar molecular geometry in solution, corresponding to the solid-state geometry of Cp*Mo(NO)(Bz)(Tms). The magnitude of this ring-current effect implies that the proximity of the methylene proton and the benzyl ring does not significantly change during the fluxionality of the benzyl ligand in solution. Again, this is evidence for there being no M-Cjpso bond breakage occurring to substantially reduce this ring-current effect. Epilogue. The most interesting result to emanate from this study is the actual isolation of the desired mixed bis(alkyl) complexes. They do not appear to be extremely unstable in any manner nor do they show any propensity to undergo disproportionation to symmetric bis(alkyl) complexes. The previous attempts to synthesize CpW(NO)(Bz)(Tms) must have failed due to the route chosen to prepare them, not because of the ultimate instability of the products. In the other bis(alkyl) complexes prepared in this research group there has been a marked reduction in the stability of the complexes when the metal is changed from tungsten to molybdenum. However, these benzyl-containing complexes are uniformly stable regardless of the metal centre involved. This last observation is in line with the view that the extra electron density donated from the benzyl ligand forming 18-electron species is the preeminent factor in determining the reactivity of these complexes. These new complexes allow variation in alkyl groups in a systematic manner, a feature which could be used to probe relative alkyl reactivities. Given the existence of these complexes, it is hoped that a wider ranging investigation of the remarkable transformations observed for the CpW(NO)(Tms)2 complex can now be accomplished for these complexes. The Cp'M(NO)(CH2Ar)(X) species also contain a chiral metal centre, since the four different ligands are arrayed in a pseudotetrahedral arrangement. Therefore, as well as being possible starting materials for reactivity studies, detailed mechanistic investigations of any transformations observed could be carried out on their optically active forms. 137 References and Notes Portions of this Chapter have been communicated: Dryden, N.H.; Legzdins, P.; Phillips, E.C.; Trotter, J.; Yee, V.C. Organometallics 1990, 9, 882. Legzdins, P.; Phillips, E.C.; Sanchez L. Organometallics 1989,8, 940. Legzdins, P.; Rettig, S.J.; Sanchez, L. Organometallics, 1988, 7, 2394. (a) Chapter 3, this Thesis, (b) Presented in part at 72nd Canadian Chemical Conference, Victoria, B.C, June 1989, Abstract 396. Martin, J.T, unpublished observations. Herring, F.G.; Legzdins, P.; Richter-Addo, G.B. Organometallics 1989,8,1485. Martin, L.R, unpublished observations. Czish, P.; Erker, G.; Korth, H-G.; Sustmann, R. Organometallics 1984,3, 945. Brunet, N. M.Sc. Dissertation, The University of British Columbia, 1988. Chapter 2, this Thesis. Phillips, E.C. Ph.D. Dissertation, The University of British Columbia, 1989. A solution of />-XylMgCl was prepared by slow addition of a-chloro-p-xylene (20 mL, 148 mmol) to a suspension of Mg turnings (5.0 g, 200 mmol) in Et20 (200 mL). After 2 h the supernatant solution was filtered through Celite and was determined to be 0.86 N by standardization with 0.100 M HCl/phenolphthalein. Greenhough, TJ.; Legzdins, P.; Martin, D.T.; Trotter, J. Inorg. Chem. 1979,18, 3268. (a) McCleverty, J.A.; Murray, A.J. Transition Met. Chem. (Weinheim) 1979, 4, 273. (b) Faller, J.W.; Whitmore, B.C. Organometallics 1986,5,752. Faller, J.W.; Shvo, Y.; Chao, K.; Murray, H.H. /. Organomet. Chem. 1982,226, 251. 138 A similar observation has been made in unsuccessful attempts to selectively methylate CpMo(NO)(I)2 with MeLi or MeMgX: Alegre, B.; de Jesus, E.; Vasquez de Miguel, A.; Royo, P.; Lanfredi, A.M.M.; Tiripicchio, A. /. Chem. Soc, Dalton Trans. 1988, 819. These authors also report that alkylation with Me3Al in toluene under heterogeneous conditions affords [CpMo(NO)(Me)(I)]2. Hunter, A.D.; Legzdins, P. Organometallics 1986,5; 1001. Chang, J.; Seidler, M.D.; Bergman, R.G. /. Am. Chem. Soc. 1989, 111, 3258. Charalambous, J. in Mass Spectrometry of Metal Compounds; J. Charalambous, Ed.; Butterworths: London, Eng.; 1975, p 55. Yee, V.C Ph.D. Dissertation, The University of British Columbia, 1990. Malito, J.T; Shakir, R.; Atwood, J.L. /. Chem. Soc, Dalton Trans. 1980,1253 and references therein. Cotton, F.A.; LaPrade, M.D. /. Am. Chem. Soc. 1968, 90, 5418. (a) Curtis, M.D.; Shiu, K.-B.; Butler, W.M. /. Am. Chem. Soc. 1986,108,1550. This estimated bond length is the sum of the covalent radii for Mo of 1.62 A (1/2 22b Mo-Mo single bond length in Cp2Mo2(CO)6)) and for C of 0.77 A. (b) Adams, R.D.; Collins, D.M.; Cotton, F.A. Inorg. Chem. 1974,13, 1086. CRC Handbook of Chemistry and Physics, 60th ed.; Weast, R. Ed.; CRC Press: Boca Raton, FL, 1979; p F-216. Dorigo, A.E.; Li, Y.; Houk, KN. /. Am. Chem. Soc. 1989, 111, 6942 Latesky, S.L; McMullen, A.K.; Niccolai, G.P.; Rothwell, LP.; Huffman, J.C. Organometallics 1985,4, 902. Davis, G.R.; Jarvis, J.A.J.; Kilbourn, B.T.; Piols, AJ.P. /. Chem Soc, Chem. Commun. 1971, 677. Davis, G.R.; Jarvis, J.A.J.; Kilbourn, B.T. /. Chem Soc, Chem. Commun. 1971, 1511. Jordan, R.F.; LaPointe, R.E.; Bajgur, C.S.; Echols, S.F.; Willett R. J. Am. Chem. Soc. 1987,109,4111. 139 Mintz, E.A.; Moloy, K.G.; Marks, T.J.; Day, V.W. /. Am. Chem. Soc. 1982,104, 4692. Edwards, P.G.; Andersen, R.A.; Zalkin, A, Organometallics 1984,3, 293. This result is simply the relative carbon pit orbital contributions to the HOMO of the benzyl ic system, e.g.: Lowe, J.P. Quantum Chemistry; Academic Press: New York, NY, 1978; p 506. Becker, E.D. High Resolution NMR, Theory and Chemical Applications, 2nd ed.; Academic Press: New York, 1980; p 96. Reference 32, p 95. Cotton, F.A.; Marks, TJ. /. Am. Chem. Soc. 1969, 91,1339. Reference 32, p 74. 140 CHAPTER 5 Reactions of Cp'M(NO)(Bz)(Cl) with Silver(I) Salts; Formation of Benzyl Carboxylate and Benzyl Nitrile Complexes. 141 Introduction In the preceding Chapter, the reactivity of a series of Cp'M(NO)(CH2Ar)2 complexes with HC1 to form Cp'M(NO)(CH2Ar)(Cl) products was reported. The benzyl ligands acted as formal 3-electron donors to the metals via a novel r/2-bonding mode in these benzyl chloro products, forming thermally stable products. These benzyl chloro complexes were alkylated with Grignard reagents in order to prepare a series of stable mixed alkyl species, Cp'M(NO)(CH2Ar)(R). The benzyl chloro complexes were also considered to be potential precursors for cationic compounds of the type + [Cp*M(NO)(Bz)(NCCH3)] via halide abstraction with silver (I) salts. It was hoped that the r/2-benzyl linkage could act as a stabilizing influence in these cationic species. The successful syntheses of the [Cp*M(NO)(Bz)(NCCH3)]BF4 (M = Mo, W) complexes are presented in this Chapter along with the results of single crystal X-ray crystallographic structural determinations of both complexes. The bonding and spectroscopic characteristics of these cationic species are contrasted with those of the neutral complexes discussed in the preceding Chapter. Chloride abstraction from the Cp'M(NO)(Bz)(Cl) compounds with silver salts of coordinating anions was also investigated. The specific objective was to probe the possibility of synthesizing neutral diastereomeric complexes of the type Cp'M(NO)(Bz)(Y), i.e, Cp'M(NO)(Bz)(Cl) + AgY • Cp'M(NO)(Bz)(Y) + AgCl (5.1) via coordination of a chiral anion (Y"). These products could potentially be separated, to give optically active metal complexes. Therefore, a preliminary investigation using a racemic silver carboxylate salt as the chloride-abstracting reagent to evaluate the viability of this methodology has been performed. The synthesis of the desired products is 142 presented and initial attempts to separate the diastereomeric products via fractional crystallization is also discussed. The presence of rj ^benzyl and bidentate carboxylate ligands in these products is deduced from the distinctive IR and NMR spectroscopic properties of the complexes. 143 Experimental Section All manipulations were performed under anaerobic conditions in an atmosphere of dinitrogen using the general experimental procedures outlined in Chapter 2. Reagent chemicals were obtained from commercial sources and were used as received. The organometallic starting materials Cp*M(NO)(Bz)(Cl) (M = Mo, W) were prepared by the methodology outlined in Chapter 4 of this Thesis. Silver 2-phenylbutyrate 1 (Ag02CCH(Et)(Ph)) was prepared according to a known method. Preparations of [Cp*M(No)(Bz)(NCCH3)]BF4. The Mo and W complexes were prepared in a similar manner. The preparation of the W complex is given as a representative example. A solution of Cp*W(NO)(Bz)(Cl) (1.43 g, 3.0 mmol) in CH3CN (45 mL) was prepared in a Schlenk tube. An IR spectrum of the resulting orange solution was recorded 1 (AgCl cell, vNO = 1584 cm"), and then solid AgBF4 (0.58 g, 3.0 mmol) was added to the stirred reaction solution. The reaction mixture immediately lightened in colour, and a flocculent white precipitate formed. The reaction mixture was stirred for 8 h to ensure that the reaction had gone to completion and was then filtered through a Celite plug (50 x 20 mm) supported on a glass frit. An IR spectrum of the light orange filtrate was 1 recorded (AgCl cell, t/NO = 1642 cm"), and then the volume of the solution was reduced in vacuo to ~8 mL. The concentrated solution was treated with Et20 (30 mL), and the resulting solution was placed in a freezer (-20 °C) to induce the crystallization of the product. The solid that deposited from the solution was collected by removing the supernatant solution via cannula. The solid remaining was washed with Et20 (15 mL) to obtain [Cp*W(NO)(Bz)(NCCH3)]BF4 as a red-orange crystals (1.18 g, 69% yield). In a similar fashion [Cp*Mo(NO)(Bz)(NCCH3)]BF4 was prepared in 80% yield as an orange crystalline solid. Crystals of both complexes for X-ray crystallographic analysis were selected from samples prepared in the manner outlined above. The characterization data for both complexes are collected in Tables 5.1 and 5.2. 144 Preparation of Cp'M(NO)(Bz)(02CR*) (R* = -CH(Et)(Ph)). The complexes CpMo(NO)(Bz)(02CR*), Cp*Mo(NO)(Bz)(02CR*), and Cp*W(NO)(Bz)(02CR*) were prepared in a similar manner. The preparation of the Cp Mo(NO)(Bz)(02CR ) species is given as a representative example. A Schlenk tube was charged with Cp*Mo(NO)(Bz)(Cl) (0.412 g, 1.07 mmol) and silver 2-phenylbutyrate (0.288 g, 1.07 mmol). The Schlenk tube was then cooled in a 10 °C water bath, and CH2C12 (40 mL) was added to the tube via syringe. The cooling bath was removed, and the reaction mixture was stirred for 3 h. During the course of the reaction, the initial red solution became lighter in colour and a white solid precipitated. The reaction mixture was filtered through a Celite plug (40 x 20 mm) to remove the AgCl precipitate, and the light orange filtrate was reduced in vacuo to a glassy foam. This foamy solid was triturated with hexanes (15 mL) to obtain an orange powder. Toluene was added dropwise to the hexanes suspension of the orange powder until all the solid had dissolved. The saturated solution was placed in a refrigerator (5 °C) to induce crystallization of the products. The solid that crystallized from the toluene/hexanes solution was isolated by removing the mother liquour by cannulation and washing the remaining solid with hexanes (10 mL). This procedure gave Cp*Mo(NO)(Bz)(02CR*) as large red-orange crystals (0.227 g, 41% yield). The *£! NMR spectrum of the product showed the presence of two diastereomers in a ~ 1:1.5 ratio as determined by integration of the proton signals for the CH3 group of the carboxylate ligand. The mother liquor was placed in a freezer (-10 °C) to induce the crystallization of more product. In this manner an additional 0.211 g (38%) of product were obtained (1:1.4 ratio of diastereomers). On a similar reaction scale Cp*W(NO)(Bz)(02CR*) was isolated as red-orange blocks in two crystallizations (0.257 g, 39%, 1:1.4 diastereomeric ratio; 0.190 g, 29%, 1:1.5 diastereomeric ratio) and CpMo(NO)(Bz)(02CR*) was isolated as orange needles (0.140 g, 29%, 1:1 ratio of diastereomers; second crop crystallized at -25 °C: 0.205 g, 42%; 1: —1.1 145 ratio of diastereomers). The characterization data for all three complexes are collected in Tables 5.1 and 5.2. 146 Table 5.1 Infrared and Elemental Analysis Data for the Benzyl Acetonitrile and Benzyl Carboxylate Complexes. Compound IR (Nujol) Analysis Found(Calcd) cm-1 % "NO other C H N [Cp*W(NO)(Bz)(NCCH3)]BF4 1618 2280,2300° 40.16 (40.13) 4.42 (4.43) 4.86 (4.93) [Cp*Mo(NO)(Bz)(NCCH3)]BF4 1638 2282,2313° 4735 (4752) 5.25 (5.25) 5.69 (5.84) b CpMo(NO)(Bz)(02CR*) 1613 1450,1495* 5935 (5933) 5.21 (5.20) 3.16 (3.14) b 0 Cp*Mo(NO)(Bz)(02CR*) 1602 1448, mo 62.98 (62.91) 6.49 (6.45) 2.70 (2.72) b Cp*W(NO)(Bz)(02CR*) 1570 1454,1488^ 53.62(53.74) 5.55 (5.51) 233 (232) The two bands are v and a combination band (y QQ + & CH) °f the CH3CN ligand. b R* = CH(Et)(Ph) c These are v QQ bands of the bidentate carboxylate group4. Table 5.2. Mass Spectral and 1H and 13C{1H} NMR Data for the Benzyl Acetonitrile and Benzyl Carboxylate Complexes. Low-resolution mass spectrum0 *H NMR 13C{1H}NMRC b Complex m/z 6 S [VCH, Hz] d [Cp*W(NO) (Bz) (NCCH3)] BF4 8.20 (t, IH, AiH, 144.54 (Carom) 3/HH = 73 Hz) 137.71 (Carom) 7.71 (d, IH, AiH, 132.00 (Carom) 3 /HH = 73Hz) 131.42 (Carom) 7.55 (t, IH, AiH, 130.22 (Carom) VHH = 7.5Hz) 112.52 (C5(CH3)5) 7.41 (t, IH, AiH, 111.62 (Cjpso) 3 /HH = 7.5 HZ) 49.13 (PhCH2) 6.42 (d, IH, AiH, 10.56 (C5(CH3)5) VHH = 7.5 Hz) 5.96 (NCCH3) 3.47 (d, IH, CffAHB) VHH = 6.0 Hz) 2.65 (s, 3H, NCC#3) 2.50 (d, IH, CHAHB, 2/HH = 6.0 Hz) 2.10 (s, 15H, C5(Ctf3)5) d [Cp'Mo(NO) (Bz) (NCCH3) ] BF4 7.99 (t, IH, AiH, 144.05 (d , Carom, [166]) 3 /HH = 7.2 Hz) 137.17 (d, Carom, [163]) 7.65 (d, IH, AiH, 136.82 (NCCH3) 3 /HH = 7.5 HZ) 131.94 (d, Carom, [160]) 7.61 (t, IH, AiH, 131.19 (d, Carom, [164]) 3 /HH = 7.5 Hz) 130.39 (d, Carom, [165]) 7.49 (t, IH, AiH, 114.61 (C5(CH3)5) 3 /HH = 7.5 HZ) 114.02 (Cjpso) 6.38 (d, IH, AiH, 55.51 (t, PhCH2) [155]) 3 /HH = 73Hz) 10.68 (q, C5(CH3)5> [129]) 3.74 (d, IH, CHAHB) 5.23 (q, NCCH3, [139]) 2/HH = 5.0 Hz) 2.55 (d, IH, CHAHB, 2/HH = 5.0 Hz) 2.48 (s, 3H, NCCiJ3) 2.00 (s, 15H, C5(Ctt3)5) 148 e CpMo(NO)(Bz)(02CR*) 7.05-7.35 (m,9H,Ar#) 191.39 (C02) 447, [P]+ 6.85-6.95 (m, 1H, ArH) 149.92 (Cip^, Bz)) 417,[P-NO]+ 5.39,5.38(2 s, 4.5H* C5H5) 138.18 (Cipso,'1 Ph) 3.42,3.39-f (2 d, 1H, PhC/YaHb, VHH = 9.0 Hz) 128.71,127.95, 127.84, 3.30,3.29^ (2t,lH,C/f, 127.43,127.40,124.02 (Car0m) 3/HH = 75 Hz) 3.21,3.19-f (2 d, 1H, PhCHa/Yb, 105.54 (d, C5H5, [180]) 2/HH = 9.0 Hz) 56.38 (d,CH, [132]) 2.05' (2 m, 1H, CHCrYaHbMe) 50.71, 50.67 (t, PhCH2, [134]) 1.85' (2 m, 1H, CHCHatfbMe) 26.35 (t, CH2, [130]) 12.23 (q, CH3, [125]) 0.86,0.85-^ (2t,3H,C#3, 3/HH = 75 Hz) e Cp*Mo(NO) (Bz) (02CR*) 6.95-7.35 (m, 9H, AiH) 190.74,190.50 (C02) + 517, [P] 6.80-6.95 (m, 1H, AiH) 148.54,148.15 (Cipso, Bz) 3.14/3.08* (dd, 1H, CH, 138.87,138.64 (Cipso/1 Ph) 3/HH = 7.0, 8.5 Hz) 2.24/ 2.22* (2 d, 1H, PhCffaHb, 128.48,128.43, 128.22,128.18, 2/HH = 8.6 Hz) 128.10,127.31, 127.04,126.98, 2.00/ 1.95* (2 d, 1H, PhCHatfb, 126.92,123.44, 123.27 (Carom) 2/HH = 8.6 Hz) 1.50-1.95 (4 m, CHCZ/y/bMe) 112.93 (C5(CH3)5) 1.64/ 1.56 * (2 s, 15H, C5(C//3)5) 56.46 (d,CH,[126]) 0.70/0.59* (21, 3H, C/Y3, 53.17, 53.14 (t, PhCH2, [133]) 3/HH = 7.4 Hz) 26.38, 25,78 (t, CH2, [127]) 12.22,12.17 (q, CH3, [127]) 9.12, 8.94 (q, C5(CH3)5, [128]) e Cp'W(NO)(Bz)(02CR*) 7.00-7.35 (m, 9H, AsH) 190.78, 190.72 (C02) + 603, [P] 6.80-6.95 (m, 1H, AiH) 148.32,148.29 (CipSo, Bz) 3.09/ 3.02* (2 dd, 1H, CH, 137.99,137.61 (Cipso/' Ph) 3/HH = 7.3, 9.0 Hz) 2.12/2.10* (2 d, 1H, PhO/aHb, 128.63,128.57,128.47,128.21, 2/HH = 10.8 Hz) 128.12,12759, 127.34,126.82, 1.80/1.76* (2 d/ 1H, PhCHatfb, 126.78,123.01,122.99 (Carom) 2/HH = 10.8 Hz) 1.45-2.00 (4 m, CHCtfaflbMe) 11158 (C5(CH3)5) 1.78/1.68* (2 s, 15H, C5(Cff3)5) 56.90,56.81 (d,CH, [135]) 0.70/0.58* (2t,3H,CrY3, 50.66, 5057 (t, PhCH2, [129], 3/HH = 7.5 Hz) Vwc = 71 Hz) 25.97, 25.34 (t, CH2, [130]) 12.09,12.05 (q, CH3, [128]) 9.08, 8.91 (q, C5(CH3)5, [128]) 149 a Probe temperatures 150 °C. No attempts were made to obtain mass spectra of the cationic species. ^ The m/z values are for highest intensity peak of the calculated isotopic cluster. c Multiplicities of carbon resonances were determined by a gated decoupled pulse sequence. ^ Spectrum recorded in CD3NO2, with the chemical shifts relative to Si(CH3)4 referenced to the residual proton signal (CX^HNO^S 4.33) or carbon signal (CE^NO^S 62.8) of the solvent. e Spectrum recorded in CDCI3. Integrations for the proton signals are the totals for both diastereomers in the product isolated from the first crystallization. f Spectrum contains signals for an equal amount of both pairs of enantiomers. £ Intensity of Cp proton signals are lower than expected due to long Ti values for these proton resonances. ^ Assignment by analogy to the free acid: CjpSO = 6 138.38 in CDCI3. 1 Indistinguishable overlapping signals for both sets of enantiomers. J Minor pair of enantiomers. * Major pair of enantiomers. " These proton signals were located by resolving them from obscuring resonances of the Cp* rings by use of the HOM2DJ pulse program supplied with the Varian XL-300 NMR spectrometer for homonuclear 2-dimensional 5 J-resolved *H NMR spectroscopy. The standard parameters supplied for acquisition and processing were used. 150 Results and Discussion Reactions of AgBF4 with Cp*M(NO)(Bz)(Cl) in CH3CN. The newly prepared Cp*M(NO)(Bz)(Cl) complexes are close analogues to a family of nitrosyl allyl complexes, CpM(NO)(r?3-allyl)(X) (M = Mo6,7, W8; X = halide). One of the characteristic reactions of these allyl complexes has been with silver salts to form cationic complexes,9 e.g., CH3CN 3 3 CpW(NO)(r/ -allyl)(I) + AgBF4 • [CpW(NO)(r? -allyl)(NCCH3)]BF4 (5.2) -Agl The desired cationic benzyl complexes were synthesized in a similar manner from the Cp*M(NO)(Bz)(Cl) compounds, i.e, CH3CN Cp*M(NO)(Bz)(Cl) + AgBF4 • [Cp*M(NO)(Bz)(NCCH3)]BF4 (5.3) -AgCl Reactions 5.3 result in formation of the desired products in good yields as red-orange or orange crystalline solids. The progress of reactions 5.3 can be followed easily by IR spectroscopy, with the diminution of the I/^Q band for the starting material (W, 1584 cm"1; 1 Mo, 1610 cm") being accompanied by the growth of a new i/N0 band at higher energy (W, 1622 cm"1; Mo, 1647 cm"1) for the cationic product. + The [Cp*M(NO)(Bz)(NCCH3)] cations are electrophilic and are able to abstract CI" from a NaCl IR cell. This process can be followed by leaving an acetonitrile solution of the complex in a NaCl cell and monitoring the progress of the chloride abstraction by IR spectroscopy. This monitoring reveals the complete loss of the v^o band of the cationic complex and the return of the U^Q band for the starting benzyl chloro species in a matter of minutes. The use of AgCl plates in the solution cell allows IR monitoring of reactions 5.3 without halide abstraction from the cell material. The electrophilicity of the benzyl 151 3 nitrile complexes is similar to that displayed by the [CpW(NO)(r/ -allyl)(NCCH3)]BF4 complex as shown in equation 5.4.9 CH3CN 3 3 [CpW(NO)(r/ -allyl)(NCCH3)]BF4 > CpW(NO)(r/ -allyl)(X) + KBF4 (5.4) KX X = Cl, Br The [Cp M(NO)(Bz)(NCCH3)]BF4 complexes are soluble in polar organic solvents such as CH3CN and CH2Cl2 that are known to be good ionizing solvents, but they are much less soluble in solvents less able to support ionic complexes, e.g. Et20. This solubility difference is exploited by using Et20 to precipitate the products from the CH3CN reaction solutions. Solutions of these compounds in CH3CN are quite thermally stable at room temperature and a solution of [Cp*Mo(NO)(Bz)(NCCH3)]BF4 in CD3NC»2 stored at -20 °C shows no noticeable decomposition over a period of months. The solids show no noticeable decomposition after storage under N2 at 10 °C for months. The IR spectral data of the cationic complexes indicate a shift to higher energy for the I/]NJO bands of the products, as compared to the Cp*M(NO)(Bz)(Cl) starting materials (vide supra), consistent with the presence of an electron-poor metal centre in these compounds resulting in less Md-»N07r* back-bonding. Along with the diagnostic U^Q bands found for these complexes, the presence of an N-bound CH3CN ligand is indicated by two bands in the region 2280-2315 cm'1 (Table 5.1). These IR absorptions are characteristic of a bound CH3CN molecule, with the increase in i/CN from that of free 1 CH3CN (2230 cm") being a commonly observed phenomenon for coordinated nitrile ligands.10 Solid-State Molecular Structures of the [Cp*M(NO)(Bz)(NCCH3)]BF4 Complexes. A solid-state X-ray crystallographic study of these cationic products has been performed11 in order to determine the metal-benzyl linkages present and compare them with those of the neutral benzyl complexes. The ORTEP plots of the molecular structures 152 of [Cp*W(NO)(Bz)(NCCH3)]BF4 and [Cp*Mo(NO)(Bz)(NCCH3)]BF4 are presented in Figures 5.1 and 5.2, and selected structural parameters collected in Tables 5.3 and 5.4 respectively. The molecular geometries of the two cationic complexes are essentially the same. All the metal-ligand bond distances are essentially the same for both congeners, which is not too surprizing since the covalent radii of molybdenum and tungsten are the same. The Mo and W complexes also contain essentially isostructural M-Ci-Qpso moieties, with similar metal-Cmethyigne (Mo, 2.182(6) A; W, 2.175(8) A) and metal-Cipso (Mo, 2.389(5) A; W, 2.383(7) A) bond lengths. The bonding in these complexes is somewhat different from that seen in the neutral r;2-benzyl complexes described in the preceding Chapter. One difference to be noted between the neutral and cationic complexes is in the bond lengths within the * aromatic ring. In the neutral complexes, Cp Mo(NO)(Bz)(X) (X = Cl, Tms), the C-C bond lengths of the ring are indicative of delocalized 7r-electron density as shown below. M C7--=C6 M = Cp*Mo(NO)(Cl) / \ / \\ C3 - C4 = 1.38 A C1 C2 .} C5 C4 - C5 = 1.38 A \ C5 - C6 = 1.37 A C3 C4 C6 - C7 = 1.38 A Conversely, alternation of long and short C-C bonds around the phenyl ring in these cations suggests that the TT-electrons are more localized as depicted below. /\ f7 °\ M = [CP*m°(n°)(ncch3)]+ / \ / \\ C3 - C4 = 1.36 A C1 C2 ^ C5 C4 - C5 = 1.40 A C5 - C6 = 1.36 A \_/ C6 - C7 = 1.42 A C3 C4 The localization in the cationic species is reminiscent of that seen in the complex, CpMo(CO)2(f?3-p-Xyl),12 although the differences in C-C bond lengths are even more pronounced in this latter complex. 153 Figure 5.1. The ORTEP plot of the solid-state molecular structure of [Cp*W(NO)(Bz)(NCCH3)]BF4. The hydrogen atoms on the Cp* ligand have been omitted for clarity. 154 Figure 52. The ORTEP plot of the solid-state molecular structure of [Cp*Mo(NO)(Bz)(NCCH3)]BF4. The hydrogen atoms on the Cp* ligand have been omitted for clarity. 155 Table 5.3. Selected Bond Distances (A), Bond Angles (°) and Torsion Angles (°) for the Molecular Structure of [Cp*W(NO)(Bz)(NCCH3)]BF4 W-CP* 2.021(4) A W-Nl-O 173.1(6)° W-Nl 1.767(7) W - N2 - C18 176.7(6) W-N2 2.105(6) W-C11-C12 79.1(5) W-Cll 2.175(8) C12-Cll-Hlla 117(3) W-C12 2.383(7) C12 - Cll - Hllb 111(5) Nl-O 1.220(9) Cll-W-C12 37.2(3) N2 - C18 1.142(9) C13-C12-C11- W 91.6(7) C11-C12 1.468(12) C17-C12-C11 - W 79.1(7) C12 - C13 1.437(12) C12 - C17 1.398(12) C13 - C14 1.346(13) C14 - C15 1.409(13) C15 - C16 1.378(13) C16 - Cll 1.393(12) C18-C19 1.452(12) a 5 CP - centre of the r? -C5Me5 ring. 156 Table 5.4. Selected Bond Distances (A), Bond Angles (°) and Torsion Angles (°) for the Molecular Structure of [Cp*Mo(NO)(Bz)(NCCH3>]BF 4 Mo-CP0 2.020(3)A Mo- Nl-O 171.2(4)° Mo-Nl 1.768(5) Mo-• N2 - C18 175.6(5) Mo-N2 2.140(5) Mo- C11-C12 79.4(4) Mo - Cll 2.182(6) C12 -Cll-Hlla 115(4) Mo - C12 2.389(5) C12 -Cll-Hllb 123(5) Nl-O 1.204(6) Cll -Mo-C12 36.7(2) N2 - C18 1.134(7) C13 - C12 - Cll - Mo 93.5(5) C11-C12 1.453(9) C17 -C12-C11-MO 78.1(5) C12 - C13 1.424(8) C12 - C17 1.400(9) C13-C14 1.355(10) C14 - C15 1.404(10) C15 - C16 1.362(10) C16 - C17 1.416(10) C18-C19 1.464(9) a CP - centre of the rr-C5Me5 ring. 157 Another difference between the structures of the neutral benzyl complexes and the Mo cations is in the dihedral angles subtended by the Corth0-QpSO-Cmethyiene" groupings (e). Mo rou In the neutral r? -benzyl complexes the Mo-Corth0-Qpso g P hes in a plane nearly perpendicular to the plane of the benzyl ligand. In these cationic complexes the metal is slightly more tilted towards one side of the benzyl ligand. Quantitatively, the smaller dihedral angles for the cationic tungsten (79.1(7)°) and molybdenum (78.1(5)°) complexes compare to an angle 6 of 84.5(2)° for the neutral Cp Mo(NO)(Bz)(Cl) complex. However, e for all these Cp*M(NO) benzyl complexes is still very different from that in 3 CpMo(CO)2(r?-/>-Xyl) (6 = 60.9°). Even with the differences noted above, the benzyl ligands in these cations are still essentially rj2 in character. This conclusion is reached primarily because the smallest metal-Cortno separations for both of the cationic complexes (Mo-Cortno 2.83 A, W-Cortrio 2.86 A) are still much larger than that expected for a Mo-C single bond length (2.39 A).1- Therefore both of these separations seem to preclude a Mo-Cortii0 bonding interaction. 3 For comparison, the Mo-Cortho bond distance in CpMo(CO)2(»? -p-Xyl) is 2.43 A, which compares very well with the estimated Mo-C single bond length. The differences in benzyl- metal interactions observed in the solid-state structures of Cp*Mo(NO)(Bz)(X) (X = CI, Tms) and [Cp*M(NO)(Bz)(NCCH3)]BF4 (M = Mo, W) are best ascribed to differences in the electronic structures of the respective metal centres. Since a linear acetonitrile ligand should not be substantially more sterically demanding than a chloro ligand, steric effects do not seem to be sufficient justification for the differing structures observed. The minimal influence of steric effects is reinforced by the similarity observed in the structures of Cp*Mo(NO)(Bz)(Cl) and Cp*Mo(NO)(Bz)(Tms) as described in the preceding Chapter, since a Tms group is unquestionably more sterically demanding than a CI group. The 158 differences between the neutral and cationic r;2-benzyl complexes prepared in this study are minor when compared to the radically different r/3-benzyl ligand in Q>Mo(CO)2(r,VXyl). The observation of three distinct types of benzyl linkage is interesting, since the respective 5-coordinate, 15-electron fragments, CpMo(CO)2, Cp*Mo(NO)(X), + [Cp*Mo(NO)(NCCH3)] , would appear to be very similar. Contrasting coordination modes between these fragments have also been seen in the bonding of the allyl ligand in 3 3 related CpW(NO)(r/ -allyl)(I) and CpW(CO)2(r/ -allyl) complexes. From X-ray crystallographic and 1H NMR spectroscopic studies, CpW(NO)(allyl)(I) is formulated as in 8 3 A below with an asymmetric a, n coordinated allyl ligand. The CpW(CO)2(r7 -allyl) complex has a symmetric JT allyl ligand as in B below, as deduced from XH NMR studies of this complex.14 A B A comparison of the solid-state molecular structures of the corresponding CpMo(NO)(allyl)(I) and CpMo(CO)2(allyl) complexes shows the same structural differences (asymmetric vs. symmetric) in the metal-allyl bonding.15 Molecular orbital calculations need to be performed, in order to determine the source of the differences in the natures of the metal-benzyl linkages in these 5-coordinate fragments. NMR Spectroscopic Properties of the Nitrile Complexes. As would be expected from the similar molecular structures determined for the Mo and W cations, the 1H NMR and ^C^Ff} NMR spectra of both complexes are qualitatively the same. The asymmetric disposition of the metal with respect to the benzyl ligand seen in the solid-state molecular structures is also reflected in the signals observed for the benzyl ligand. The asymmetry is manifested in the 1H NMR spectra by the observation of five signals for the aromatic 159 protons of both of these complexes (Figure 5.3) and in the 13C{1H} NMR spectra by six signals for the aromatic carbon atoms of the benzyl group (Figure 5.4). Although slightly asymmetric, the benzyl ligands in the cations still have NMR signals that compare favorably with other r?2-benzyl complexes that have been prepared. The cationic nature of these complexes results in the 1H chemical shifts of the Cp* and methylene protons moving to lower field relative to those in the benzyl chloro starting materials. The 2/HH coupling constants for the methylene protons are in the range observed for the neutral r?2-benzyl complexes, and a VCH value of 155 Hz for the benzyl methylene carbon of [Cp*Mo(NO)(Bz)(NCCH3)]BF4 is not significantly different from that of the neutral benzyl complexes (148-152 Hz). A convincing piece of evidence for the similarity between neutral and cationic benzyl complexes in solution is provided by the chemical shift of the Qpso signal for [Cp*Mo(NO)(Bz)(NCCH3)]BF4 of 114.0 ppm and for [Cp*W(NO)(Bz)(NCCH3)]BF4 of 111.6 ppm, which compare to values of 114.8 ppm and 112.6 ppm for the respective neutral benzyl chloro complexes. The cationic benzyl complexes appear to be stereochemically rigid in solution, with no fluxionality of the benzyl ligand being evident, as seen in the neutral r?2-benzyl complexes. The differences in metal-benzyl bonding extant in the cationic and neutral benzyl complexes (reflected in their solid-state molecular structures) may make a fluxional process as seen in solution for the neutral benzyl complexes energetically unfavorable for the cationic complexes at room temperature. Indeed, the low-temperature *H NMR spectrum of Cp*Mo(NO)(Bz)(Cl) (Figure 4.8) is very similar in appearance to the *H NMR spectra of [Cp*Mo(NO)(Bz)(NCCH3)]BF4 (Figure 5.3). Preparation, Physical, and Spectroscopic Properties of the Cp'M(NO)(Bz)(02CR*) Complexes (R* = CH(Et)(Ph)). Given the success of the reactions of AgBF4 with the Cp*M(NO)(Bz)(Cl) complexes, investigation of similar reactions with silver carboxylate reagents has been performed. This reactivity has been pursued for the A LJ' I 1 1 i • . • ! | • i • • i • i i • j • • ' I p'pM' i—i—I—i—i—i—i—|—i—i—r—i—|—i—I—i—i r Figure 5.3. The 300 MHz *H NMR spectrum of [Cp*Mo(NO)(Bz)(NCCH3)]BF4 in CD3N02 (•)• o 1 1 ' I—i—I—i—l—l—i—I—i—i—i—I—I—i—r- I i I i I | I i—j—i—i—i—i—|—i—i—i—r~]—i—i—i—i—|—i—T—i—i—|—i—i—i—i—|—i—i—i—i—|—i—i—i—i—|—n I I—r"' ' ' J 140 120 1 00 80 60 40 20 PPM 0 Figure 5.4. The 75 MHz ^C^H} NMR spectrum of [Cp*Mo(NO)(Bz)(NCCH3)]BF4 in CD3N02. ON 162 purposes of evaluating it as a potential method of resolving the racemic Cp'M(NO)(Bz)(Cl) complexes. Silver carboxylates have been chosen since optically active carboxylic acids can be easily obtained, either commercially or by established resolution procedures,16 and silver carboxylates are easily prepared from these acids.lb The reactions of the silver carboxylates with the benzyl chloro complexes produce the desired neutral carboxylate complexes, i.e, CH2CI2 2 1 Cp'M(NO)(r, -Bz)(Cl) + Ag02CR* • Cp'M(NO)(r; -Bz)(02CR*) (5.5) -AgCl R* = CHEtPh The products of reactions 5.5 form thermally stable solutions, with no change in the XH NMR spectra of CDCI3 solutions of these complexes being seen after storage at room temperature for several days. As solids, the complexes can be transferred and weighed in air with no noticeable decomposition occurring. The slight solubility of the complexes in aliphatic solvents, as well as their free solubility in aromatic solvents, are consistent with their formulation as shown in equation 5.5 with covalently bonded carboxylate ligands. Also the highest mass m/z features in the low-resolution mass spectra of all three + complexes may be assigned to the monomeric [Cp'M(NO)(Bz)(02CR )] mass 10ns. Since the silver 2-phenylbutyrate used in these reactions is racemic, there are 2 diastereomeric products formed in reactions 5.5, consisting of the (RR,SS) pair of enantioihers and the (RS,SR) pair of enantiomers.17 Since these two sets of enantiomers should possess different physical properties, an initial attempt to separate these two if, different products through fractional crystallization has been performed. The Cp - containing Mo and W tungsten products appear to be somewhat separable by this route, with one crystallization causing some enrichment in one set of enantiomers. The crystallizations of CpMo(NO)(Bz)(02CR*) were not as successful in enriching the product in either of the diastereomers, with only a slight excess of one diastereomer being observed 163 in the second crop of product. These preliminary results indicate that the Cp* complexes may be easier to resolve in future.18 The bonding in these carboxylate complexes is different from that invoked for all the other compounds containing the "CpM(NO)(Bz)" moiety described in this Thesis. All of the spectroscopic data (vide infra) for the carboxylate complexes are consistent with their formulation as 18-electron Cp'M(NO)(r;1-benzyl)(02CR*) complexes, containing bidentate carboxylate ligands acting as formal 3-electron donors. The IR spectra of the complexes are indicative of bidentate carboxylate ligands, with two bands assigned to VQQ stretching modes in the range 1450-1500 cm"1. The separation of these symmetric and asymmetric stretching modes is about 50 cm"1, which has been shown to be characteristic of a bidentate carboxylate ligand.4 The 41 NMR and 13C{1H} NMR spectra of these complexes (Table 5.2) are in accord with their formulations as presented in equation 5.5. The diastereomeric nature of the products is exemplified in the 1H NMR spectrum of the CpMo(NO)(Bz)(C»2CR*) (Figure 5.5). In this spectrum the presence of two sets of enantiomers is indicated by the two sets of signals for the methylene protons of the benzyl ligand, as well as for the CH and CH3 protons on the carboxylate ligand. It has been observed that 2-phenyl substituted ligands, as in the carboxylate used here, usually show well resolved signals for different diastereomers in 1H NMR spectra from ring current effects of the aromatic ring.18b The CH2 protons of the 2-phenylbutyrate ligand are diastereotopic. This property is manifest in slightly different coupling constants between each of these protons and the CH proton (Table 5.2), causing the CH proton signal to appear as a doublet of doublets for the Cp*- containing complexes. Unfortunately, their diastereotopic nature also results in the CH2 proton signals being very complicated. Consequently, no attempt has been made to assign the signals due to these protons. For CpMo(NO)(Bz)(02CR*), the CH proton signals of the carboxylate ligand are separated enough to determine that 164 A A I —i— 4 PPM X Figure 5.5. The 200 MHz H NMR spectrum of CpMo(NO)(Bz)(02CCH(Et)(Ph)) in CDCI3 165 the ratio of the two diastereomers isolated from the crystallization is nearly equal. For the two Cp* complexes the CH3 proton signals are well separated (e.g. Figure 5.6), and integration of their intensities allows a direct determination of the relative amounts of each set of enantiomers present in the sample. The presence of two diastereomers is also seen in the ^C^H} NMR spectra of all the complexes. The obvious result is the presence of two signals for some of the carbon atoms, corresponding to each of the respective diastereomeric species. The presence of an r;1-benzyl ligand is also established by the NMR spectra of these complexes. The first indication of this is the chemical shift of Qpso for the benzyl ligand, which occurs above 145 ppm for all three complexes. Additionally, the proton coupled ^C NMR spectra exhibit VCH coupling constants of 130-134 Hz for the methylene carbon atoms of the benzyl groups. Finally, the 2/HH coupling constants are between 8-11 Hz for the diastereotopic methylene protons of the benzyl ligand. These parameters are all comparable to those seen for the r/1-benzyl ligands in the low-temperature NMR spectra of 2 the Cp'M(NO)(Bz)2 complexes (QPSO = 148-153 ppm ; VCH = 126-132 Hz ; /HH = 9-10 Hz).19 This result is consistent with the IR data which indicates a bidentate carboxylate ligand, therefore an rj ^benzyl ligand acting as a formal 1-electron donor is necessary for the Cp'M(NO)(Bz)(02CCR*) products to attain an 18-electron configuration. The presence of an r) ^benzyl ligand in the products of reactions 5.5 shows that the second carboxylate oxygen atom is apparently a strong enough Lewis base to displace an r?2-benzyl linkage present in the Cp'M(NO)(Bz)(Cl) precursor complexes. Figure 5.6. The 200 MHz *H NMR spectrum of Cp*Mo(NO)(Bz)(02CCH(Et)(Ph)) in CDC13 167 Summary and Future Work The results of the chemistry presented in this Chapter are important for indicating the direction of future studies. This work has shown that halide abstraction from the neutral benzyl chloro complexes results in the formation of electrophilic cationic compounds. The chemistry of the cationic species should be investigated towards establishing the degree of the electrophilic character of these complexes and exploitation of this property by examining these complexes as possible Lewis acid promoters for organic transformations. The properties of the [Cp*M(NO)(Bz)(NCCH3)]BF4 complexes make them attractive for application in this manner. They are easily prepared and can be stored as solids for extended periods of time. The conditions necessary to handle them are easily attainable with normal precautions routinely used in organic synthesis. Organometallic nitrosyl complexes are being used towards this end, some recent examples including the 20 report of [(PMe3)(CO)3(NO)W]SbF6 as a catalyst for Diels-Alder reactions and condensation of methyl propiolate with alkenes to form lactones mediated by 21 [Cp'M(NO)2]BF4 (Cp' = Cp, Cp*; M = Cr, Mo, W). The potential applications of the [Cp M(NO)(Bz)(NCCH3)]BF4 cations as Lewis acid promoters is especially exciting when the potential for the development of a chiral Lewis acid is considered. The initial results of the investigation of the formation and separation of the diastereomeric complexes, Cp'M(NO)(Bz)(02CR*), is a positive indication that resolution of Cp'M(NO)(Bz)(02CR*) complexes may be possible using optically active Ag02CR*. These complexes could be used as a route into optically active benzyl halide species, e.g., Cp'M(NO)(Bz)(02CR*) + HX Cp'M(NO)(Bz)(X) + R*C02H (5.5) or benzyl alkyl complexes, e.g. Cp'M(NO)(Bz)(Li02CR*) + LiR Cp'M(NO)(Bz)(R) + LiQ2CR* (5.6) 168 Chiral organometallic nitrosyl systems have also been used for the synthesis of optically active organic complexes. The configurationally stable chiral Lewis acid [CpRe(NO)(PPh3)]+X" has been prepared in optically pure form, and has been used as a chiral auxiliary for directing stereospecific nucleophilic attacks on carbonyl groups to form optically active alcohols and in the synthesis of optically active CHDTCC^Na.22 Resolved molybdenum allyl complexes have been shown to react with aldehydes to form alcohols of high optical purity.23 Obviously most of these future avenues of investigation hinge upon the successful resolution of a Cp'M(NO)(L)(L') system in an optically pure form. However, the enormous potential these systems possess for synthetic purposes as well as for reactivity studies would seem to justify a thorough survey of possible resolution schemes. 169 References and Notes (a) The silver salt was prepared by treatment of the 2-phenylbutyric acid with NaOH followed by AgN03. (b) Wilson, CV. Org. React. 1957, 9,332. Milne, J.B. Can. J. Chem. 1970,48, 75. Storhoff, B.N.; Huntley, C.L. Coord. Chem. Rev. 1977,23,1. Nakamoto, K. Infrared and Raman Spectra of Inorganic and Coordination Compounds, 4th ed.; John-Wiley and Sons: New York, NY, 1986; p 232. For example see: Kessler, H.; Gehrke, M.; Griesinger, C. Angew. Chem. Int. Ed. Engl. 1988,27, 490 and references therein. McCleverty, J.A.; Murray, AJ. Transition Met. Chem (Weinheim) 1979,4, 273. • Faller, J.W.; Shvo, Y. /. Am. Chem. Soc. 1980,102, 5398. Greenhough, T.J.; Legzdins, P.; Martin, D.T.; Trotter, J. Inorg. Chem. 1979,18, 3268. Martin, D.T. Ph.D. Dissertation, The University of British Columbia, 1984. Reference 4, p 280. Yee, V.C. Ph.D. Dissertation, The University of British Columbia, 1990. Cotton, F.A.; LaPrade, M.D. /. Am. Chem. Soc. 1968, 90, 5418. (a) Curtis, M.D.; Shiu, K.-B.; Butler, W.M. /. Am. Chem. Soc. 1986,108,1550 and references therein. Since W and Mo have similar covalent radii,1315 their M-C single bond lengths are also expected to be comparable. (b) This similarity in size is due to the "lanthanide contraction" resulting from inefficient shielding of the tungsten nucleus by its 4f orbitals: Sharpe, A.G. Inorganic Chemistry; Longman: New York, NY, 1981; Chapter 25. Faller, J.W.; Chen, CC; Mattina, M.J.; Jakubowski, A. /. Organomet. Chem. 1973, 52,361. 170 Faller, J.W; Chodosh, D.F.; Katahira, D. /. Organomet. Chem. 1980,187, 227. The resolution procedures for a series of acids of the general formula (R)(R')CHC02H have been reported: Levene, P.A.; Mikeska, L.A.; Passoth, K. /. Biol Chem. 1930,88, 27. 5 The priority sequence of the ligands would be r/ -Cp > 02CR* > NO > CH2Ph for assignment of R or S stereochemistry to the metal centre in a pseudotetrahedral geometry. For example, the complex below would be the (R,S) diasteriomer with the metal stereochemistry preceeding the carboxylate stereochemistry. The pseudoatom atomic weights of polyhapto ligands in organometallics have been assigned: Stanley, K.; Baird, M.C. /. Am. Chem. Soc. 1975, 97, 6598. (a) If optically active Ag02CR* is used, only two of the four diastereomers would be formed. Their separation may be easier by fractional crystallization due to the simplified product mixture but this cannot be predicted with confidence. (b) For example see: Brunner, H. Adv. Organomet. Chem. 1980,18, 151. Phillips, E.C. Ph.D. Dissertation, The University of British Columbia, 1989. Bonnesen, P.V.; Puckett, C.L.; Honeychuck, R.V.; Hersh, W.H. /. Am. Chem. Soc. 1989, 111, 6070. This article also includes an extensive compilation of references for Lewis acid catalysis of Diels-Alder reactions. Legzdins, P.; Richter-Addo, G.B.; Einstein, F.W.B.; Jones, R.H. Organometallics 1990, 9, 431. 171 (22) Fernandez, J.; Gladysz, J.A. Organometallics 1989,8,207 and references therein. (23) Faller, J.W.; Linebarrier, D.L. /. Am. Chem. Soc. 1989, 111, 1937. 172 Appendix 173 The IR spectrum of a 1:2 mixture of CpW(NO)(Cl)2 and PhMgCl as a solution in THF. 174 The IR spectrum of a 1:2 mixture of Cp*Mo(NO)(Cl)2 and PhMgCl as a solution in THF. CP«MOCNO> CC3_> 2 • PHMGCL. IN THF S-20C CO. 2MM THF> A hJ J 22S2.7 204B.B JB17. 8 lBlO. 2 1703.2 1SBS. 1 14SB. 1 1382.0 1275. O WAVENUMBERS CCM-1> 176 177 178 13 The IR spectrum of CpW(NO){ C(0)Npt}(Npt) as a solution in CH2C12. g CPW rr—^ o o o o to 8? P £. o 1 O + •4- 2400. O 2012. 5 1812. S 1818. 7 1425. O 1231.2 1097. 5 643. 78 030.00 WAVENUMBERS CCM-1> 180 The IR spectrum' of Cp*W(NO)(p-Xyl)2 as a Nujol mull. 181 The IR spectrum of Cp*W(NO)(Bz)(Cl) as a Nujol mull. 182 The 300 MHz XH NMR spectrum of Cp*W(NO)(p-Xyl)(Npt) in QDg. 3 3 185 186 The IR spectrum of Cp*W(NO)(Bz)(02CCH(Et)(Ph)) as a Nujol mull. 187 The 50 MHz ^C^H} APT NMR spectrum of Cp*Mo(NO)(Bz)(02CCH(Et)(Ph)) in CDCl3(«). The carbon signals for Cquat and CH2 groups are negative and the carbon signals for CH and CH3 groups are positive. r ° JL _o CM _o o to o to 0Z Q. O •CU 2 o o CO