r5.Y.q2

Negative Ion Rear{angements rn the Gas Phase

A thesis presented for the degree of Doctor of Philosophy

by

Peter Charles Hans EICHINGER B.Sc. (Hons.)

Deparhent of Organic Chemistry The University of Adelaide

Sept., t99L Contents

List of Fisures v1

List of Tables xiii

- Statement xvu

Acknowledeements XVlU

Abstract xx

1, Introduction. L - 1.1 Generation of Negative lons. 2 7.2 Formation of Negative lons. J

1,.2.1,. Primary Formation. 3

7.2.L.7 Resonance Capture (Electron Capture). 3

1,.2.1,.2 Dissociative Resonance Capture (Electron Dissociative Attachment). 4

L.2.1.3 lon-Pair Formation. 5

1.2.2 Secondary lons. 6 7.2.2.I Proton Transfer. 7

7.2.2.2 Charge Exchange. 7

1.2.2.3 Nucleophilic Addition. 8 I.2.2.4 Nucleophilic Displacement. 8 1.3. Collisional Activation 9 1.4 Mass Analysed Ion Kinetic Energy Spectrometry

( MrKES ). 10 1.5 Detection of Negative Ions. 11 1.6 Vacuum Generators ZAB 2HF Reversed Geometry

Mass Spectrometer. 74

1.7 Structural Identification of Ions. 76

1..71 Charge Reversal. 18

L.72 Neutralisation - Reionisation Mass Spectrometry 22 7.73 IonDissociation Characteristics. 25

1..74 Kinetic Energy Release. 28 1.8 Rules for Fragmentation of Anions. 32 1.8.1 Simple homolytic cleavage. 33

7.8.2 Formation of an ion- neutral complex. 33 1.8.3 Proton transfer preceding ion - neutral

complex formation. 34

7.8.4 Rearrangement, preceding fragmentation 34

1..9 Rearrangement reactions. 35 1.10 Intrinsic Reactivity 37

1.11.1 Solvent 38

1,.11.2 Volume of Activation ÂV# 39

1.11.3 pKa 40

7.17.4 Counter Ions 47

1.11.5 Ion Pairing 43

1,.11.6 Aggre gregation 50

L The Wittig Rearrangement. 2.1, Introduction. 64 2.1.1. The Condensed Phase Wittig Rearrangement. 64

2.L.2 The Gas Phase Wittig Rearrangement. 73 2.1.2.1Atkyl- and Aryl Benzyl Ethers. 74

2.1..2.2 Diallyl Ether. 79 111

2.7.2.3 Absence of the Wittig Rearrangement

from Allyl Alkyl Ethers. 83

2.2 Results and Discussion 85

2.2.1 Dibenzyl Ether. 85

2.2.2 Allyl Phenyl Ether. 87 2.2.3 Allyl Benzyl Ether. 97

2.2.4 A Reinvestigation of the Condensed Phase

Rearrangements of Deprotonated Allyl Benzyl Ether. 111 2.2.5 Diphenylmethyl Phenyl Ether. 772 2.2.6 Vinyl Alkyl Ether. 776

2.2.7 Benzyl Vinyl Ether. 130

2.2.8 Divinyl Ether. 131 2.2.9 Allyl Vinyl Ether. 133

2.3 Conclusions and Summary. 145

2.3.I The Scope of the Wittig Rearrangement in the Gas Phase. 145

2.3.2 Loss of Alkyl Radicals. 747

3. The Smiles Rearrangement.

3.1, Introduction. 159 3.2 Results and Discussion. 166

3.2.L Phenoxyalcohols. 1.66

3.2.2 Thiophenoxyalcohols. 180 3.2.3 Phenoxyalkanethiols. IU 3.3 Synthetic studies directed toward Identification

of the Intermediates in the Gas Phase Rearrangement. 186 lr The Benzilic Acid Reanangement and related rearrangements. 4.1, Introduction. 190 4.2 The Gas Phase Benzilic Acid Rearrangement. 791 1V

4.3 Rearrangements of Deprotonated

o - Hydroxylactate Esters. 797

4.4 Rearrangement Reactions of

Deprotonated Alkyl Pyruvates. 207

å The Acvloxvacetate / Acvl hvdroxvacetate rearransement and related reanansements. 5.1 Introduction. 278 5.2 The Gas Phase Rearrangement of Acyloxyacetates. 223 5.3.1 Fragmentations of the Alkoxycarbonyl

- a-Dicarbonyl Complex(L90). 231

5.3.2 Loss of Alkyl Formate. 23r 5.3.3 Fragmentations of the Alkoxycarbonyl

a-Dicarbonyl Complex (19L). 237

5.4 The Gas Phase Acyloxyacetonitrile / AcyL Hydroxyacetonitrile Rearrangement. 248

6. Future Directions. 6.7 Future Directions. 256 6.2 Dissociation of the deprotonated molecule to form a radical anion -radical or anion - neutral complex. 257 6.2.7 The Wittig rearrangement. 257 6.2.2 Other Rearrangements. 259 6.3 Internal nucleophilic addition of an ion

to a suitable r-System. 259 6.3.1 The Smiles Rearrangement. 259 6.3.2 Miscellaneous. 267 6.4 Nucleophilic addition of the base to yield an anionic species which may then undergo reatrangement. 262 v

6.4.7 Benzilic Acid rearrangement. 262

6.5 Solvolysis reactions. 264

6.6 Miscellaneous Reactions. 264

Exoerimental. 266

References. 296

339 vl

List of Fisures.

Figure Title of Figure Page Number Number

1.1 Morse curves -indicating the vertical transition on

addition of an electron to a diatomic system to form a

stable radical anion M-. ( resonance capfure, process a) and an unstable radical anion ( dissociative resonance

capture, process b ). 6 r.2 Schematic representation of the electron multiplier

commonly used to detect positive ions. 72

1.3 Schematic representation of the electron multiplier showing modifications designed to enhance the

sensivity of the detector to negative ions. T4 '1..4 Simplified schematic representation of the VG ZAB

2HF Mass Specbometer. 15

1.5a Conventional Positive Ion Mass Spectrum of Flavanone at 70 eY. 20

1.5b The Charge Reversal Mass Spectrum of the Flavanone

Molecular lon. 20

1..6 Schematic representation of a Neutralisation Reionisation experiment. 23 1.7 Simplified schematic representation of the Gas Collision Cell and the Effect of Applying an Electric

Potential to the Collision Cell. 26 vtl

1.8 The spectrum of deprotonated benzyl ethyl ether obtained by floating the gas collision cell to +1,000V shown offset from the normal spectrum. 28 r.9 The Effects of Orientation of an Ion leading to Kinetic

Energy Release. 29

1.L0a Peak Shape resulting from small Kinetic Energy 31

Release.

1.10b Peak Shape resulting from large Kinetic Energy Release. 31

2.1 The Collisional Activation Mass Spectrum of Deprotonated PhCH2OEt recorded on a VG ZAB zF{F

Mass Spectrometer. 75

2.2 The Collisional Activation Mass Spectrum of Deprotonated Ph(EI)CHOH recorded on a VG ZAB zF{F

Mass Spectrometer. 75

2.3 The Collisional Activation Mass Spectrum of Deprotonated Diallyl Ether recorded on a VG ZAB z}{F

Mass Spectrometer. 80

2.4 The Collisional Activation Mass Spectrum of Deprotonated 5 - Hexenal recorded on a VG ZAB zH.F

Mass Spectrometer. 80

2.5 Comparative Reaction Coordinates for allylic migration

in the presence and absence of a Lithium Cation. 83

2.6 The Collisional Activation Mass Spectrum of

PhOC-(H)CH=CH2. 89

2.7 The Collisional Activation Mass Spectrum of Ph(cH2=cH)cHo- 89

2.8 The Collisional Activation Mass Spectrum of Deprotonated g-Allyl Phenol. 96 vul

2.9 The Collisional Activation Mass Spectrum of

Deprotonated Allyl Benzyl Ether. 100 2.70 The Collisional Activation Mass Spectrum of the Expected Wittig rearrangement ion

Ph(CH2=CHCH2)CHO- (82). 101

2.r1. The Collisional Activation Mass Spectrum of the Expected Wittig rearrangement ion

PhCH2(CHz=CH)CHO- (83). 101

2.12 The Collisional Activation Mass Spectrum of

PhCD2OC-(H)CH=CHz (8a) recorded on a VG ZAB 2FIF

Mass Spectrometer. 104 2.r3 The Collisional Activation Mass Spectrum of PhCD-OCH2CH=CH2 (85) recorded on a VG ZAB z}{F

Mass Spectrometer. 104

2.74 The Collisional Activation Mass Spectrum of CH2=Ç-

OCH2CH3 prepared by an S¡2(Si) reaction. 123

2.15 The Collisional Activation Mass Spectrum of Deprotonated Ethyl Vinyl Ether. 723

2.1.6 The Collisional Activation Mass Spectrum of the Expected Wittig Rearrangement Ion, the Butan-2-one Enolate lon. r24 2.77 Partial Spectrum of the "Collision - Induced" vs "IJnimolecular" Decompositions of Deprotonated Ethyl Vinyl Ether recorded on a VG Z.AB 2HF Mass Spectrometer. 126

2.78 The Collisional Activation Mass Spectrum of Deprotonated n-Butyl Vinyl Ether. L28 2.r9 The Collisional Activation Mass Spectrum of CH2=Çg--OBur. r29 IX

2.20 The Collisional Activation Mass Spectrum of the Flexan

- 2 - one Enolate Ion. 129

2.21 Collisional Activation Mass Spectrum of Deprotonated

Allyl Vinyl Ether. 137

2.22 Collisional Activation Mass Spectrum of Deprotonated

Penta-1.,4-Pentadien-3-o1. 137

2.23 Charge Reversal Mass Spectrum of Deprotonated Allyl

Vinyl Ether. 138

2.24 Charge Reversal Mass Spectrum of Deprotonated Penta-

1,4-dien-3-ol. 138

2.25 The Collisional Activation Mass Spectrum of

Deprotonated Benzyl Diethyl Acetal [PhC-(OEt)z]. r57

3.1 Collisional Activation Mass Spectrum of PhOCH2CH218O- recorded on the VG ZAB 2HF Mass

Spectrometer. 769

3.2 Collisional Activation Mass Spectrum of PhOCH2CH2CH218O- recorded on the VG ZAB 2HF

Mass Spectrometer. 769

3.3 Collisional Activation Mass Spectrum of PhO(CHz)¿18O- recorded on the VG ZAB 2HF Mass

Spectrometer. 170

3.4 PhO- peak (m/z = 93) from PhO(CH2)zO-. 171. 3.5 PhO- composite peak (m/z = 93) from PhO(CH2)sO-. 177 3.6 Partial Spectrum of the "Collision Induced" vs "IJnimolecular" Decompositions of PhO(CH2)3189- io.,

recorded on the VG ZAB 2HF Mass Spectrometer. 172

3.7 Collisional Activation Mass Spectrum of the Phenoxide

Ion, recorded on a VG ZAB 2F{F instrument. 178 X

3.8 Partial Collisional Activation Mass Spectrum of the

PhO- ion ( the rn/z = 64 - 66 region ). 179 3.9 Partial Collisional Activation Mass Spectrum of the

2,4,6 Dg -PhO- ion ( the m/z = 64 - 66 region ). 779

3.10 Partial Collisional Activation Mass Spectrum (the m/z

= 64 - 66 region) of (Ph-1-13C)O- ion (m/z = 94). 180

3.11 Partial Collisional Activation Mass Spectrum (the m/z = 64 - 66 region) of (Ph-1-t3C)O- ion (m/z = 94) trom 2-

(phenoxy - 1-13c)ethoxide ion. 180

3.72 Collisional Activation Mass Spectrum of PhSCH2CHzO- recorded on the VG ZAB 2HF Mass

Spectrometer. 181

3.13 PhS- peak (m/z = 109 ) from PhS(CH2)2O-. 183 3.14 Composite PhS- peak (m/z = 109 ) from PhS(CH2)3O-. 183 3.15 The Collisional Activation Mass Spectrum of

P h O ( C H z)zS- , recorded on the VG ZAB 2HF instrument. 185

4.1 Collisional Activation Mass Spectrum (CA MS/MS) of the [Benzil + - OH ] adduct recorded on the Kratos MS 50 TA instrument. 793

4.2 Collisional Activation Mass Spectrum (CA

MS/MS/MS) for t}lte m/z = 183 ion from the I benzil + -OH I adduct recorded on the Kratos MS 50 TA instrument. 194

4.3 The Collisional Activation Mass Spectrum (CA MS/MS) of Deprotonated MeCH(OH)13COzMe recorded on the VG ZAB 2HF instrument using an electric sector

scan 202 XI

4.4 The Collisional Activation Mass Spectrum (CA MS/MS) of H2C-CO13COzCH3 recorded on the VG ZAB

2HF instrument, using an electric sector scan. 21.0

4.5 The Collisional Activation Mass Spectrum (CA MS/MS) of H2C-COCOzCDzCDg recorded on the VG

ZAB zF{F instrument, using an electric sector scan 21.4

4.6 The Collisional Activation Mass Spectrum (CA MS/MS) of H2C-COCOzC(CH3)3 recorded on the VG

ZAB zI{F instrument, using an electric sector scan. 2t6

5.1 Collisional Activation Mass Spectrum of CH3OCON(O-)CH2CH=CH2 recorded on the VG ZAB

2HF Mass Spectrometer. 220

5.2 Collisional Activation Mass Spectrum of CH31aç92C-(H)COzCHg recorded on the VG ZAB 2HF Mass Spectrometer. 225

5.3 Collisional Activation Mass Spectrum of lBuCOzC-(H¡taga2Et recorded on the VG ZAB z''F Mass Spectrometer. 228

5.4 Collisional Activation Mass Spectrum of -CHzCOCHO

recorded on the VG ZAB 2HF Mass Spectrometer. 232

5.5 Collisional Activation Mass Spectrum of Deprotonated CHgCOzCHzCO2tBu recorded on the VG ZAB 2HF Mass

Spectrometer. 246

5.6 Collisional Activation Mass Spectrum of Deprotonated CH3CO2CH2CN recorded on the VG ZAB 2HF Mass

Spectrometer. 25t xtl

6.r The Collisional Activation Mass Spectrum of Deprotonated Benzyl ethyl thioether. 259

6.2 The Collisional Activation Mass Spectrum of Deprotonated Phenylpropane thiol. 259

6.3 The Collisional Activation Mass Spectrum of Deprotonated 2-Methoxy Phenoxypropanol. 26L xll1

List of Tables

Table Title of Table Page Number Number

1.1 Effect of Solvation on the Gas Phase reaction. 39

7.2 The Affect of Viscosity on the proportion of Geminate

Pair Recombinations. 40

1.3 Comparitive Acidities of Selected Compounds in

solution and the Gas Phase. 4I

1..4 The Affect of Counter ion on the ratio of products formed. 47

1.5 The Effect of Counter ion on the Course of the

Decomposition of the Alkoxide (29). 49

7.6 The Effect of Solvent on the Product formed in Scheme

1.8. 55

7.7 The Relative Rates of Nucleophilic Substitution of Benzyl Tosylate in Water and in Acetonitrile in the

presence of 18{rown-6. 58

1.8 The Effect of L5 - Crown - 5 on the Course of the

Reaction Between Ethyt Magnesium and Pyridine. 61

7.9 The Affect of Increasing the Ionic Character at the

Oxygen atom on the Adjacent C-H bond. 62

2.7 Products obtained on rearrangement of Dibenzyl Ether

and Methyl Lithium (in excess). 72

2.2 The Collisional Activation and Charge Reversal Mass

Spectra of the C3H5O- Ions. 78 XlV

2.3 Collisional Activation Mass Spectra of Ph(PhCHfCHO-

and Isotopically Labelled Analogues. 86

2.4 The Collisional Activation (CA) Mass Spectra of

P - hOC (H) CH =CH 2, Ph(CH2 =QH ) C HO-, 2- (CH2=Çll- CH2)C6H4O- and related Deuterium Labelled

Derivatives. 90

2.5 Charge Reversal (CR) Mass Spectra of

PhOC-(H)CH=CH2 and Ph(CH=CHz)CHO- Ions. 9T

2.6 The Collisional Activation Mass Spectra of Benzyl Allyl

Ether and Isomeric C19H11O- Ions. 107

2.7 Charge Reversal (CR) Mass Spectra of the C1gH9- ion from Deprotonated Allyl Benzyl Ether and Isomeric

CrOHq- Ions. 108

2.8 Collisional Activation Mass Spectra of Ph2C-OPh,

Ph3CO-, and isotopically labelled analogues. 113

2.9 Collisional Activation Mass Specfra of Deprotonated Vinyl Ethers and of the Wittig Rearrangement Products. r19-r20

2.1.0 Collisional Activation Mass Spectra of C3H5O- Ions. 126

2.71. Collisional Activation Mass Spectra of Deprotonated

Benzyl Vinyl Ether and Benzyl Vinyl Alcohol. 131

2.72 Collisional Activation Mass Spectra of Deprotonated Divinyl Ether and the expected Wittig Rearrangement Product. 132

2.r3 Collisional Activation Mass Spectra of Deprotonated Allyl Vinyl Ether and Isomers. 740 2.r4 The Occurrence of the Wittig rearrangement and Elimination reactions from Deprotonated Ethers. 146 XV

3.1 The Affect of Solvent and Base on the Intramolecular

Nucleophilic A¡omatic Substitution of (3.6). 762 3.2 The Collisional Activation Mass Spectra of Deprotonated Phenoxy alcohols, Thiophenoxy alcohols

and Phenoxy alkanethiols. 767

4.7 The Charge Reversal (CR) Mass Spectra of the m/z = 183 lons. 795 4.2 Collisional Activation (CA) Mass Spectra of

Deprotonated cr-Hydroxycarboxylate Esters. 199

4.3 Collisional Activation (CA) and Charge Reversal (CR) Mass Spectral Data for the Product Ions from

Deprotonated a-Hydroxyacetate Esters. 205

4.4 Collisional Activation Mass Spectra of Deprotonated

Alkyl Pyruvates (CH3COCOzR). 209

4.5 The Collisional Activation (CA) and Charge Reversal

(CR) Mass Spectra of Product Ions in the Mass Spectra of Deprotonated Alkyl Pyruvates (CHgCOCO2R). 2t2

5.1 The Collisional Activation Mass Spectra of

Deprotonated Acyloxyacetate Esters I R1CO2C- (R2)CO2R3 I and Deprotonated Acyl Hydroxyacetate

Esters I R1CO2CR2(OH)CO2R3 - H+ ]. 224 5.2 Collisional Activation Mass Spectra (CA MS/MS/MS)

of selected Product Ions in the Mass Spectra of

Deprotonated Acyloxy Acetate Esters

(RtCO2C-(H)CO2R3). 232 xvl

5.3 Collisional Activation Mass Spectra of Deprotonated

Acyloxyacetonitrile ( RCO2CH2CN - H+ ) Ions and Labelled Analogues. 250

5.4 Collisional Activation and Charge Reversal Mass

Spectra of Selected Product Ions in the CA MS/MS of Deprotonated Acetoxyacetonitrile (CH3CO2CH2CN). 253 xvll

Statement

This thesis contains no material which has been submitted for the award of any degree or diploma in this of any other university. To the best of my knowledge this thesis contains no material previously published or written by any other person, except where due reference has been made in the text.

A NAME P c il. [tcurñc,Eg COURSE: i 3 - il i:,ü ijgl I give consent to this copy of my thesis, when deposited in the Univ Ml],:fi'#l#'ffi',-' available for loan and photocopying.

) 2 SIGNED DATE XVIII

Acknowledeements

I would like to express my deep gratilude to my supervisor, Prof.

John H. Bowie for his stimulating direction of this work. I have been afforded a rare opportunity to explore aspects of chemistry which have intrigued me. The only demands placed upon me have been to rigorously study the systems chosen and ensure that questions are asked in a manner such that results will be meaningful.

The very nature of this project has meant that the assistance of many people has been necessary to allow rapid progress to be made. The assistance has taken many forms. Intellectually, the questions and comments of Dr. George E. Gream and Dr. Ch¡is. J. Easton have been greatly appreciated. They have proved to be enlightening/ providing valuable perspectives on problems that may have otherwise been missed.

The able technical assistance of Lee Paltridge and Tom. Blumenthal cannot be underestimated. The expertise that they have acquired has ensu¡ed that the ZAB is almost always working at its optimum, in spite of the operators using the instrument. Dr. Roger N. Hayes deserves special mention. He has performed all of the experiments requiring the use of the Kratos mass spectrometer. Without these results it would not be possible to proceed in the manner that has been outlined in the thesis.

Dr. Guy Krippner and his wife Lorely, Dr. Richard Warren and Dr. George Skouroumounis have been valuable friends. They have an infectious zest for life and have made the time in the lab. both memorable and enjoyable. xlx

Fellow colleagues in the Bowie group, Kev. Downard and Greg Adams have been helpful in their comments and deserve thanks.

Finally, Mum and Dad, Magoo and Sharon Watkins (friend, fiance and now wife) have given me a great deal of support. Mum has endured quite a lot while typing sections of the initial draft of this thesis. Sharon, perhaps more than anyone else is acutely aware of the traumas involved in my persuit of mass spectrometry and writing this thesis. XX

Abstract

In this thesis the investigation of rearrangements known to proceed in solution were examined in the gas phase. Where possible the results obtained in solution are indicated in the text and in this way the behaviou¡ of the compounds in condensed phase and the gas phase may be directly compared.

In solution there are many factors influencing the course of a reaction. The term "intrinsic" reactivity is often applied to reactions occurring in the gas phase. In the introduction, this concept is dealt with to demonstrate the influence of the solvent and counter ions which are seldom considered when dealing with anionic chemistry. In this way, the unique insight that the study of gas phase ion chemistry provides is higNighted.

The systems chosen for study have been carefully selected. In solution there are significant differences in the behaviour of the systems.

The Wittig rearrangement was chosen because the mechanism is believed to involve a dissociated intermediate, the Smiles rearrangement involves nucleophilic attack at the ipso position of the aromatic ring, and the Benzilic Acid rearrangement was studied because in solution chelation of a metal ion is important to effect nucleophilic addition of the base. xxl

"Commit thy works unto the Lord and thy thoughts shall be established."

Proverbs 16:3

"Entia non sunt multiplicanda praeter necessitetem"

Occam's Razor Cha Í 1

Introduction.

In the past twenty years, the study of negative ion chemistry in the gas phase has emerged as a powerful method for investigating the behaviour of anions and radical anions in the absence of solvent or counter ion.

Gas phase aciditiesl, nudeophilic displacement reactions2, atomatic substitutions3, eliminations3 and reductions4 as well as steteoelectronic effectsS have been studied. This new perspective has served to highlight the differences between the gas phase and solution, yielding at times quite unexpected results.

The use of negative ion mass spectrometry either as an analytical tool6 or to study gas phase ion chemistry7 has trailed positive ion mass spectrometry. There are several reasons that may be advanced to account for this. These are:

L) perceived difficulties in generating negative ions8. 2) the lack of decompositions of the negative ions8.

3) the low sensitivity of negative ion detectionS.

The solutions to these problems are addressed here Chapter 1 2

L.L Generation of Negative Ions:

The advent of the chemical ionisation source has permitted the study of an extended range of compounds. Limitations still exist concerning the types of ions which may be studied; however, such deficiencies often appear to be overstatedT. Analytical chemists have

shown that it is possible to detect certain drugsS at levels generally below those obtained by positive ion mass spectrometry. It has recently been

noted that "negative ion chemical ionisation (NICI) has been found to be

at least as sensitive as positive ion chemical ionisation (PICI) and may be

up to 100 times more sensitive in favourable cases"9. For example, the limit of detection of clonazepaml0 ( 1 ) by negative ion mass spectrometry is about 4 pg. whereas 100 pg. represents the limit achievable by positive ion mass spectrometry.

H o

Cl

1

Numerous other examples exist where negative ion mass spectrometry is the method of choice'l'1,72. Most molecules contain acidic hydrogens and are readily deprotonated and the anions produced are readily detected. However ions where the cognate radicals have negative electron affinities such as simple alkyl anions (except eg. methyl) cannot be observed directly. Chapter 1 J

1.2 Formation of Negative Ions.

1.2.L Primarv Formation.

Direct electron impact results in the formation of positive and negative ions13,14 in the source of a mass spectrometer. The process is -affected by the energy of the electrons, the pressure and temperature of the source1S, and the neutral precursor molecule.

Electron capture by u molecule may result in the formation of a short lived negative ion state. The term "resonance" is applied to these states and only form over a narrow range of energies. Primary negative ions can be formed by either resonance or non - resonance processes

(where the electron removes some of the excess energy).

1.2.1.L Resonance Capture (Electron Attachmentl

A molecule having a positive electron affinity (ER¡a , nãy capture a low energy electron ( 0 - 10 eV ). Electron attachment to a neutral molecule ( AB ) leads to the formation of a radical anion (equation 1.L).

AB+e -) AB-' equation 1.1

At low pressures (< tO-¿ Torr), the lifetimes of the ions may vary from 'J,0'72 to 10-1a secondsl6. Ions having lifetimes > L0-6 seconds can be studied using conventional mass spectrometers. At high pressures ( > 1

a This is a fundamental properqr of a molecule defined as M+e -+ M-' -ÂHo = EA (M) Chapter 1 4

Torr ) the formation of certain long lived molecular ions becomes favou¡able. These high pressures can be generated by the use of an inert buffer gas. Excess electron attachment energy is dissipated through loosely bound van der Waals complexes pre-existing in the same gas. This process is indicated in the exampIeTT (equations 1..2a, 1.2b and 1,.2c), where the asterisk indicates the species bearing excess energy and G represents the inert buffer gas atom or molecule.

Oz.G+e-+*O2-'.G equation 1.2a *O2-'.G -) 02-'.ç* equation 1.2b O2-'.G* l Oz-' +G* equation 1.2c

These molecules may decay by loss of an electron (autodetachment) and / or dissociative resonance capture.

In general, long lived molecular radical anions are formed for molecules that are aromatic, highly conjugated systems, or highly substituted (eg. halogenated or nitro) systemslS.

1.2.1.2 Dissociative Resonance Capture. (Electron Dissociative Attachmentì

The capture of an electron may result in a molecule undergoing a vertical transition to form a radical anion lying on the repulsive region of it's potential energy surface; the resulting unstable radical anion fragmenting extremely rapidly to yield an anion.

Dissociative resonance capture generally increases in rate with increasing temperafure since the probability of dissociation increases with Chapter 1 5 increasing internal energy. At low source pressures, collisional stabilization does not occur and excess energy remains in the molecule. Pre-existing loosely bound Van der Waal's complexes are not present. This process is favoured by higher energies than for simple resonance capture, over an electron energy range of 0 - 15 eV.

The products may be formed in electronically excited states. In the context of this work, the fragmentation of HzO-* is describedlg. Dissociation of HzO-* leads to th¡ee anionic products, H-, O-* and HO-. Three electronic resonances at 6.5, 8.6 and 11.8 eV were studied. Translational energy measurements of H- produced during dissociation of HzO-* f¡om the 6.5 eV and 8.6 eV resonances indicated that most of the energy is released as translational energy. By contrast OH- was found to be formed in a rotationally excited state . Dissociation of HzO-* from the 11.8 eV resonance yielded OH- in an electronically excited (Az L+) state.

The resonance processes are represented schematically in Figure 1.1.

L.2.1,.3 lon - Pair Formation.

This process ocqrrs over a wide range of electron energies. It is generally observed for energies of greater than l-0 eV. In the chemical ionisation source high energy electrons can be "thermalised" in this manner

(equation L.3).

e -(fast) + AB + A+ + B- + e -(slow) equation L.3 Chapter 7 6 M

M (Stable)

oO I ç() _t ill M '(Unsta ble) a

AEDir.o.i",ion 2 34 I 0

b

0

Atomic SeParation Figure 1.L Morse curves indicating the vertical transition on addition of an electron to a diatomic system to form an unstable radical anion (dissociative resonance capture, process a ) and a stable radical anion M-' ( resonance capture, process b).

1..2.2 Secondarv lons.

The chemical ionisation (CI) source typically operates at a very high pressure ( 0.2 - 2 Torr ) compared to the pressure found in the flight tube (< fO-z Torr). At a pressure of L Torr the mean free path of an ion is 0.2 x l-0-3 mm. Under these conditions the electrons and primary ions formed undergo multiple collisions, becoming "thermalised" (le. excess internal energy of the ions is transferred to the buffer gas). The primary ions may then react with a molecule. Generally, the proportion of sample to reagent gas is kept very low to ensure that electron capture by the sample is avoided2o. Chapter 1 7

'j,.2.2.1 Proton Transf er.

Deprotonation is arguably the most important process for generating anions in the CI source. This reaction is dependent on the relative gas phase acidities2l of the species present.

A-+BH+AH+B- equation L.4

The proton transfer ( equation L.4 ) occurs when BH is a stronger acid than AH. The usual bases employed in the gas phase are NH2- Q2\ or HO- (23) as the conjugate acids of these ions are very weak. For many organic compounds the deprotonation reaction is exothermic.

'j..2.2.2 Charee Exchanse.

When the electron affinity of a neutral species B is greater than that of A- then an electron transfer may occur24,25 (equation 1..5).

A-+B + A+B- equation L.5

This is a fundamental property of an ion of great interest to physical chemists. The superoxide radical anion ( Oz-') is frequently present in the chemical ionisation source due to the inadvertent admittance of air with the sample or reagent gas. It has a low electron affinity and will tranfer an electron to compounds of higher electron affinity (eg26. equation 1.6). Chapter 1 8

CI

+ (1.6) +O 2 + 02 Noz LA Noz

1.2.2.3 Nucleophilic Addition.

This process is quite important when weak bases such as C1-, O-'or O2-', which do not readily deprotonate molecules are used27. The resulting addition compounds are classified as "adducts" if covalent bonding occurs2S or "solvated ions" if covalent bonding is absent29.

L2.2.4 Nucleophilic Displacement.

This type of reaction occurs when deprotonation is unfavourable3o.

For example,(l) fluoride ion has been used to displace silyl groups to yield highly reactive anions3l (eg. equation L.7) which are inaccessible by other routes, and (ii) the ion O-' will displace hydrogen atoms from aromatic compounds32 (equation 1.8).

F- + (CH3)sSiCOCH3 + CH3CO- + (CH3)aSiF equation 1.7 ArH+O-'+A¡O-+H' equation 1.8

There are reactions which may also cause the anionic species to decompose in the chemical ionisation source especially when the source pressure is excessively high. These are collisional detachment33 and Chapter 1 9 associative detachment34 reactions. Little data on these reactions appears to have been reported. Radicals are generated, however these species react far more slowly with neutral molecules and do not present a major complicating factor3S.

1.3 Collisional Activation.

Ions formed by chemical ionisation (CI) are generally formed with low internal energies36. These ions do not fragment readily and consequently only a minimal amount of structural information can be obtained. Decomposition can be induced when an ion accelerated to high translational energy collides with a surface or a neutral molecule or atom. In the ZAB, this is achieved by focussing the ion beam into a small 'collision' cell containing a gas ( typically He or Ar ) at a moderate pressure. On collision, a proportion of the translational energy is transferred to the ion often resulting in electronic excitation. Radiationless transitions occur extremely rapidly to yield the ion in an electronic ground state, with the energy redistributed over the entire ion.

The quasi equilibrium theory3T (QET) rationalises these observations. The amount of translational energy convertedb into internal energy is given by equation 1-.9. h= Ro^.t i Consta",t â.= i,-{¿.l^aø}rå^ d¿llancq ¿tl = -{rqnrtahoï"l c'"e1X h l2trJ \rtz h^ L Yr^a5S ÂE..,"* ø \ml- II equation 1.9

b It is possible that this relationship does not hold for large molecules such as peptides.(ref. 39) Chapter 1 10

During an 8 kV collision an ion of m/z = 100 might typically gain 1 - 3 eV. The excess energy imparted into the ion may be dissipated by framentation. The rate of decomposition is directly related to the amount of excess energy (le. the greater the excess energy the greater the rate of decomposition). Only reactions with a rate constant of > 105 - 106 sec.-1 are observed as the time of flight through the spectrometer (m/z = 100; accel. voltage 8kV) is 24 msec. This technigue4O is commonly referred to as Collisional Activation (CA) or Collisional Induced Dissociation (CID).

It has been stated that CID gives a great range of dissociation channels of high abundance arising from simple cleavage rather than rearrangement4l and as a consequence the technique has considerable structural application. In this thesis it will be demonstrated that rearrangements are quite common and that the structures of the rearranged ions can be easily determined.

L.4 Mass Analysed Ion Kinetic Energy Spectrometry (MIKESì

An ion M- transmitted through the magnetic sector and down the flight tube may fragment into many daughter ions m7-, m2-, ñ3-... nìn- on collisional activation. The original translational energy of the parent ion M- is shared between the fragment ion and the neutral species formed. These ions may be detected when an electric sector is placed after the collision cell. The MIKE technique42 therefore produces peaks that may be broadened due to energy release (termed kinetic energy release) arising from the dynamics of fragmentation. In an analytical sense the low mass resolution is not desirable, but this effect is very useful in examining mechanistic problems and will be discussed later. Chapter 1 11

There are numerous linked - scanning techniques43 that are available to study decomposition reactions in the gas - phase depending on the type of information required. They will not be discussed here as they have not been used for the work described in this thesis.

1.5 Detection of Negative Ions.

In principle, the detection of negative ions is a relatively straight forward process44. In a magnetic sector mass spectrometer for example, this is achieved by reversing the polarity of the accelerating voltage and the magnetic current4S.

In practice however, it is generally found that the sensitivity of detectors is very low for negative ions when compared with positive ions. To understand this it is essential to describe the operation of the most versatile and commonly used detector, the electron multiplier. The electron multiplier effects amplification by converting the kinetic energy of the ion beam, imparted into the surface of a copper-beryllium dynode (a type of electrode) into the emission of secondary electrons46. The secondary electrons are accelerated onto an appropriately positioned second dynode, which in turn emits an even greater number of secondary electrons. The acceleration of electrons from one dynode to another is achieved by connecting each stage to a successively higher potential. A typical electron multiplier may be comprised of up to 20 stages, allowing amplification in excess of 106. A conceptual diagram of the dynode array is given in Figure 1.2. The positive gradient from plate A to plate B is typically from -3kV up to ground potential. Chøpter'l- 6. 72 do ca-)

Conversion of Positive Ions into Electrons.

Amplified Signal

Figure 1.2 Schematic representation of the electron multiplier commonly used to detect positive ions.

This arrangement is ideally suited to the detection of positive ions as they are attracted toward the first negatively biased dynode.

By contrast, negative ions are repelled by the negatively biased dynode array resulting in decreased incident kinetic energy. This results in a far weaker signal being generated. Furtherrnore, the ion beam is defocussed; many ions are deflected away from the dynode array to be discharged on the walls of the flight tube. This problem is particularly severe for the quadrupole mass spectrometer as the accelerating voltage applied to ions leaving the source is very small ( 3-15 V ).

The problem of low sensitivity has been tackled in two ways. The initial approach4T was to "float" the electron multiplier at +6 kV. In this way, the incident negative ions will be attracted to the dynode array and Chøpter 7 13 defocussing of the ion beam could be avoided. Unfortunately this arrangement proved unsuccessful due to the unavoidable leakage of small amounts of current#.

An alternative approach was devetoped by Stafford48. A specialised positively charged electrode held at a positive voltage (+L to +3 kV) and isolated from a normal dynode array is placed near the entrance to the electron multiplier. The negative ions are focussed onto this special electrode (called a conversion dynode) and upon collision, form secondary positive ions by fragmentation, charge stripping and sputteringc. These secondary positive ions are detected by the electron multiplier as described before. A conceptual diagram49 of this electron multiplier designed for detection of negative ions is given in Figure 1.3. A similar approach has been described by Markey et al.

The detector is likely to exhibit mass and energyS0 discrimination as observed for positive ions.

Very recently, changes to the positioning of the detector have allowed an enhancement in the signal to noise ratio. The detector is aligned "off axis" thereby preventing fast neutrals which may have considerable kinetic energy from striking the dynodes. It is thought that this will contribute to an enhanced life for the electron multiplier.

# Overload and saturation effects result when the ouþut current of the dynode array exceeds l0-8 amperes. c The major source of positive ions is thought to be the layer of contamination present on the elechode surface (ref. 51). Chapter 1 ''Off-Axis" Conversion Dynode \4 Negatrve Ions are convened to Posirive Ions. (A)

NeuÚals Fast 4=t4

I \ I \on Bearn I

\ \

ì .9

?,

(B) Conversion of Positive Ions into Electrons. (c) Amplifìed Signal

Figure L.3 Schematic representation of the electron multiplier showing modifications designed to enhance the sensivity of the detector to negative ions.

1.6 Vacuum Generators ZAB 2HF Reversed Geometry Mass Spectrometer.

The problems alluded to above have been addressed and many coûunercially available instruments are capable of detecting negative ions

-with high sensitivity. None the less there are a number of features of this instrument53 that deserve comment as it is an essential tool in this work. Firstly it must be noted that this instrument ( Figure 1.4 ) is of "reversed geometry". This means that unlike conventional ( 'Nier - ]ohnston' ) two sector instruments, the electric sector is placed after the magnetic sector. Chapter L 15

iln r rilr I ilt ll Second Field Free Region Gas Collision Cell Magnettc Electric Sector Sector

Alignment and Alignment and Focus Plates Focus Plates First Field Free Region Gas Collision Cell Ï

EI-CI Source Electron Multiplier Ion Repeller

Figure L.4 Simplified schematic representation of the VG ZAB 2HF Mass Spectrometer.

Unlike earlier instruments designed to select ions on the basis of their energy, the magnetic sector can be used to select an ion (negative or positive) of predetermined m/z vaIue. Operating both sectors as mass analysers in this configuration is termed tandem mass spectromehy (MS / MS). This modification is valuable in the study the dynamics and mechanism of fragmentation. Furthermore it is possible to examine gas phase reactions using the techiques of charge reversal and kinetic energy

release. These techniques will be described later.

It must be noted that the inverse arrangement of the electric and magnetic sectors creates one minor problem. This is the presence of artefact peaksSa that may on occasion be observed. They arise when metastable transitions occur in the field free region to yield an ion with the same momentum as the ion of interest. These peaks are often very Chapter 1 76 sharp when compared to the peaks normally observed in a MIKE spectrum.

In principle, a quadrupole mass spectrometer can be used for the examination of negative ion reactions. Unfortunately, problems arise because the neutral sample and reagent gas diffuse out of the exit slit along with thermal electrons extracted by the positive accelerating lens used under negative ion chemical ionisation conditions. The positive acceleration lens does not permit positive ions to exit into the flight tube. However, positive ions may form in the mixture between the source and the quadrupole. Unlike the magnet used in the ZAB, the quadrupole does not discriminate between positive and negative ions and consequently positive ions formed outside of the source are recorded as "pseudonegative" ions55. The appearance of the CHs- ion in negative ion chemical ionisation spectra, when using CFIa as a moderating gas is typical of this effect56.

The energy imparted into an ion during collision with a neutral gas is low due to the low accelerating voltage ( 30- 100 eV). It is generally not possible to use the technique of kinetic energy release using this type of instrument.

1.7 Structural Identification of lons.

A full description of a molecular ion is essential to uniquely identify the species5T. It is essential that the widest possible range of techniques available are used when applicable to uniquely identify the ions. There are relatively few techniques available. Using an instrument Chapter 7 77 such as the VG ZAB zF{F it is possible to use the techniques of collisional activation, charge-reversal and kinetic energy release. The structure of daughter ions may be compared with ions obtained from independently synthesized neutral precursors and the fragmentations compared. The technique of neutralisation - reionisation is available but this is only applicable for a limited number of select dissociations.

Characterisation of ion dissociations in the gas phase constitute major challenges. The methods which have been devised for gas phase product ion identification can be readily employed on the VG ZAB 2HF are as under :-

(1) Collisional Activation

(2) Charge Reversal

(3) Metastable Ion Dissociation Characteristics

The use of facilities made available to us, namely a Kratos MS 50 TA hybrid triple analyser (EBE) instrument I by Dr.R.N. Hayes, Midwest Centre for Mass Spectrometry, University of Nebraska, Lincoln, Nebraska, U.S.A. ] allowed us, in addition, to perform the following experiments:-

(4) Neutralisation-Reionisation (s) MS/MS/MS58

For many polyatomic ions the number of isomeric ion st¡ucfures formed on fragmentation are considerable. A recent review on negative ion dissociations has shown that many such reaction processes in the gas phase are adequately rationalised using the same "arrow - pushing" formalism used for solution chemistry (this is expanded upon later). Chapter 7 18

Reactions typically observed in solution (eg.base catalysed elimination) may also occur in the gas phase. Where a number of isomeric product anions may be formed, thermochemical data including gas phase acidity, electron affinity and hydride affinity can be applied to determine the probability of such processes. Finally the structure of an ion may be assigned by comparing the various spectra of that ion with those of various isomeric ions formed by unequivocal synthetic procedures. Collisional Activation has already been dealt with in an earlier section

(Section 1.3).

7.7 j[, Charse Reversal.d

In 7975, Bowie et a1.59 pioneered a technique for converting non- decomposing negative ions into decomposing positive ions.

The technique had its genesis in the work of Beynon and Cooks eú a1.60 They had shown that when positive ions (M+) underwent collision with a suitable target gas (N) at high pressure, charge stripping ( equation

1.10 ) and charge exchange (equation 1.11) reactions occu¡red.

N+M++M2++N+e- equation 1.1.0

N+M++M-+N2+ equation 1.11

Bowie was able to demonstrate that under similar conditions negative ions would undergo a charge stripping reaction to form decomposing positive ions ( equation 1.72 ). Interestingly the formation of

d In the literature this reaction is also referred to as a "charge permutation" reaction. Chapter 1 19 a doubly charged negative ion [which might arise on charge exchange (equation 1.13)l was not observed.

N + M- + M+ + N+ +2e- equation 1.12

N+M-+M2-+N+ equation 1..L3

The spectra obtained were poorly resolvedS9. This is undoubtedly due to the combination of scattering (arising from the high pressure used) and the effect of kinetic energy release ( discussed later ). The relatively abundances of many fragment ions differed significantly from conventional electron impact ( EI ) spectra and this is taken to indicate a difference in internal energy of ions formed by the two processes. Consequently the spectra obtained are referred as charge reversal ( CR MS/MS ) spectrae . The differences are highlighted in the spectrum of flavanone ( Figure 1.5a and 1.5b ).

Collision of an ion of high translational energy with a target gas molecule (N) is a rapid process6l (t = 19-15 sec ). During this time it is argued that nuclear motion is slow and can be ignored relative to molecular vibrations (t = 10-13sec.) and electrons must undergo vertical ionisation (a Franck-Condon process). Rearrangements do not occur during the ionisation process, although they may occur before the translational energy is distributed throughout the isoelectronic vibro - rotational energy levels of the molecule. Ions arising from rearrangement processes are not very common in charge reversal spectra. Ions produced from slow skeletal rearrangements are generally found to be weak, whereas fast rearrangements such as those involving ortho substifuents62

e Originally these spettra were referred to as +E spectra. Chapter L 20

224 100 t20

a

E €(l 50 104 r47 -ó 92 o .! (.) ú, & 78 2A'l 5l 131 r95 39 ll5 139 165 178 0 40 60 80 100 na 140 160 180 2û 2n Figure 1.5a Conventional Positive Ion Mass Spectrum of Flavanone at 70 eV. 104

r00 r20 l3l

78 89 gz ñ 65 oC) F ! ll ,d 50 o I t) r78 5l 224 139 r5Z 165 20? 39 rgg 195 223

0

40 60 80 lm t20 140 160 180 2n 220 Figure 1.5b The Charge Reversal Mass Spectrum of the Flavanone Molecular lon. Chapter 1 2l are more pronounced. Lifetimes of S I0-7 secs have been inferred fo¡ decomposing ions. These data support internal energy being high, ie. simple cleavage dominates ovet rearrangments.

In most instances, it has found that the charge reversal process is not particularly sensitive to the collision gas pressureS9 (le. the absolute intensity of the spectrum increases with an increase in pressure but relative peak heights are not markedly affected). The two processes, outlined in equations 1.14 and 1.15, therefore, seem unlikely.

AB-+A-*+B+A+. equation 1.L4

AB-+A*+B-rA+ equation 1..15

In a recent study of this process Beynon63 examined a system in which the collision gas pressure had a profound effect on the relative abundances of ions. By deflecting the negative ions away from the collision cell, the fast metastable neutrals presumably formed by a single electron detachment were readily reionised. The spectra obtained were strikingly different from the directly ionised negative ion. Under these conditions the amount of energy converted to internal energy was determined to be in the range 10 - 25 eV. The amount transferred was dependent on target gas (N2 vs He) and substrate.

The minimum energy ÂE necessary for the reaction to convert AB- into AB+, assuming both are in their ground state is given by the sum of the electron affinity ( EA ) and ionisation energy ( IE ) of the neutral molecule64 ( equation 1.16).

ÂE = 1g + EA (AB) equation L.16 Chapter 1 22

In polyatomic ions loss of excess internal energy proceeds through fragmentation via ( equation 7.77 )

[An*1*+A++B equation 1.17

and not the alternative process ( equation1.18 )

[AB-],+ -å A-* -) A+ equation 1.18

Variation of collision gas pressure is found to have almost negligible effect on the spectra.

Rearrangement of ions after charge reversal has been reported. It has been asserted that the charge reversal spectra of cyclo-CH2-CH=CH- or CH2=Ç=CH- results65 in the formation of a mixture of the cyclo-

CH2=Ç1¡=CH+ and CH2=C=CH+ ions. The rearrangement occurs within 8 ps (microseconds) of charge inversion.

A list of examples of charge reversal spectra of negative ions66 has been reported.

1.7.2 Neutralisation - Reionisation Mass Spectrometry.

This technique cannot generally be readily applied to the V.acuum Generators ZAB 2HF mass spectrometer in this laboratory, but an experiment will be described (see Chapter 3) using the MS 50 TA instrument. Chøpter 1 23

The neutralisation - reionisation technique6T lr.as proven very

useful as an aid in the identification of neutral species, including radicals and carbenes, derived from very small polyatomic ions. Unfortunately,

large polyatomic ions usually decompose by many competitive processes leading to a large number of neutral species. Structural identification of the individual neutral species under these circumstances is very difficult.

The experiment is, in principle, relatively simple and is illustrated

in the diagram below ( Figure 1.6 ).

Gas Collision Cell Gas Collision Cell

Ions Field Free Region 02, He Gas

Fast Beam s=__ Neutrals fr ---- IonBeam#

-lon

Mass Selection Analyser Magnetic Sector Electric Sector

Ion Deflection Plaæ

Figure 1.6 Schematic representation of a Neutralisation - Reionisation exPeriment.

The negative ion (ABX-) (or positive ion) of interest is selected by use of the magnetic sector. This transmitted ion ABX- is allowed to enter

the first of two gas collision cells. A neutral species ABX may be formed by

collisional detachment (equation 1.1.9) with the collision gas N. Chapter 1 24

N + ABX-+ N +ABX+¿- equation 1.19 or alternatively it may fragment to yield a neutral and a charged ion

(equation 7.20).

N+ABX-+N+AB-+X equation 1.20

The collision gas is typically argon or xenon. The negatively charged ions; namely the parent ion ABX-, fragment (daughter) ion AB- and any positive ions formed by charge reversal are deflected toward the walls of the flight tube by the use of a suitably charged electrode placed immediately after the first gas collision cell.

The beam of metastable neutrals are allowed to pass into a second gas collision cell. By contrast with the first gas collision cell mercury or oxygen are used as collision gases. The use of oxygen appears to most favoured for the generation of positive ions, as electron transfer (equation

1.2L) occurs as well as collisional detachment (equation 1.22).

Oz*+ABX+ABX+* +Oz- equation 1.21

02* * ABX + ABX+* + C,2+ e- equation 1..22

Mclafferty has used mercury extensively and has found this produces negative ions68 in high abundance.

The excess energy imparted into the ionised neutrals is dissipated by fragmentation; the products of which are analysed using the electric sector. Chapter 1 25

For polyatomic molecules in particular the results obtained must be carefully scrutinized, because under some instances rearrangement of the neutral species may occur prior to reionisation. Burgers and Villeneuve69 have shown that the benzyl anion forms the benzyl radical only under unimolecular electron detachment whereas the cycloheptatrienyl radical is formed on collisional detachment. Clearly considerable caution must be exercised when dealing with systems which undergo very facile radical rearrangements and cyclisations.

'1,.7.3 Ion Dissociation Characteristics.

Ionisation of a molecule leads to the formation of ions with a wide range of internal energies. Many ions will fragment within the source, resulting in poor yields of molecular ions. However with soft ionisation techniques such as Negative Ion Chemical Ionisation (NICI), the deprotonated molecular ion can be extracted from the sou¡ce in high yield. The extracted ions are called metastable if fragmentations of the ions occur after the source exit slit but prior to the collector. If the fragmentations are not initiated by a collision with a second body but are due to excess internal energy then the process is a unimolecular decomposition. Metastable ions having less energy may decompose readily on collision with a second body ( such as an inert gas ): this process then is a collision- induced decomposition.

The relative proportions of ions fragmenting by unimolecular vs

collision induced decompositions are characteristic of a particular ion57, as this is dependent on the thermochemistry of the decomposition channels. Clearly for an exothermic rearrangement we would expect the rearranging Chapter 7 26 ion and authentic product ion to show very similar ion dissociation characteristics.

Measurement of the relative proportions of ions fragmenting by unimolecular vs collision induced decompositions is readily accomplished in the ZAB 2HF by floating the collision gas cells electrically above (or below) ground potentialT0. This process is illustrated below

(Figure 1.7).

V

I L m;+'(u)t mj Ion B

Gas Collision Cell Field Free Region

eV eV (mz/mr) eV + (m2-m2lm1) eV' Kinetic (mzlmr) eV ,----- (mzlmr) eV Energy e(V-V') - (eV-V') m2lm¡ (c) (m2lm¡) eV - eV'

Position

Figure 1.7 Simplified schematic representation of the Gas Collision Cell and the Effect of Applying an Electric Potential to the Collision Cell.

A non decomposing ion, m1-accelerated through a voltage V is further accelerated toward a collision gas cell held at a positive voltage V'. The ion will have greater translational energy prior to a collision induced

decomposition; the fragmentation leading to product ion m2- o f translational energy lmze (V-V')]/m1. After the collision cell the daughter Chapter 1 27 ion is no longer under the influence of the electrically charged collision cell and the product ion will have an energy given by equation 1.23.

E = m2eVlm1-eV' (m1-m2/m1) equation 1.23

An ion which decomposes unimolecularly to yield ion m2- (ie. outside of the collision gas cell) will not have the translational energy altered. It will have translational energy given by equation !.24.

E = mz/m1 eV' equation 1..24

Thus an electric sector scan will result in a MIKE spectrum in which the ions formed by unimolecular dissociation are found in the usual position but the collision induced fragmentation is shifted by an amount dependent on the voltage applied to the cell. An example from these laboratories is illustrated in Figure 1.8.

In principle, the proportion of ions decomposing unimolecularly or via collisionally induced fragmentation should not vary from one instrument to another. In the ZAB zI{F the collision cell is placed at the focal point of the ion beam in the 2nd fietd free region. Consequently the dimensions of the collision cell are minimised, thereby ensuring that a region of high gas pressure can be produced without excessive leakage of collision gas into the flight tube. When leakage occurs a significant proportion of the ions will undergo collision induced decompositions outside of the cell. For this reason to be rigorously correct the terms "shifted" and "unshifted" should be applied when describing the affect of placing a charge on the collision cell. This is because gas leaking out of the collision cell may cause collision - induced decompositions which are Chapter 1 28 included loosely with the fragmentations that are designated unimolecular. In an instrument such as the vG ZAB 2HF the leakage is minor. None the less even though the terms "unimolecular" and

"collision - induced" are used loosely, it is this author's opinion that these terms more readily convey the intended meaning than the other terms.

135

106(c) 106(u) 57(c)

106 51

57(u) 1't (c)

105(c) 105(u) 17(u) ,,

Figure 1.8 The spectrum of deprotonated benzyl ethyl ether obtained by floating the gas collision cell to +1,000V shown offset from the normal sPectrum.

L.7.4 Kinetic Enerev Release.

In solution it is not possible to gain any information regarding the potential energy profile f or decompositions because of rapid interconversion of- the internal energy of a system amongst the various rotational, vibrational and translational modes. A Maxwell-Boltzmann energy distribution occurs. The absence of solvent molecules colliding Chapter 1 29 with the decomposing ions allows us to gain some information about the potential energy profile of the final step.

In the mass spectrometer, the pressure in the flight tube is < I0-7

Torr, with the exception of the chemical ionisation source and the gas collision cell. The mean free path of an ion at these pressures is calculated to be approx. 200 cm ( ie. an ion will travel on average 200 cm before a collision). Therefore excess energy of the transition state cannot be partitioned to other molecules by collision. As a consequence, the excess energy can only be lost during fragmentation processes as either rotational or translational energy. Any differences in rotational oi translational energy will cause broadening of the observed signal. This process is termed kinetic energy release71. This phenomenon is most readily understood by use of the diagram ( Figure 1.9 ).

-Sll.l2 - "Àñ- \ DE Gas Collision Cell Electric Sector

"8" ions A "8" ions ions

mlz

Figure 1.9 The Effects of Orientation of an Ion leading to Kinetic Energy

Release. Chapter 7 30

Ions admitted into the 2nd field free region will contain a range of energies; for the most part considerably below the energy required for fragmentation. Those anions having excess energy will have fragmented in the 1st field free region and hence do not reached this position. As the particles are collisionally activated they gain between 1-3 eV ( 115-345 kJ *o1-1 ) of excess internal energy7z. There is also a small loss of translational energy of the anion73, referred to as the "disproportionation factor"; this will not be discussed further here.

The loss of a neutral molecule from a decomposing ion is markedly affected by the thermochemistry of the fragmentation. The process is often complex, however to illustrate this point a simple representation of the fragmentation process is provided. The "white" sphere represents the incipient neutral species and the "black" sphere the incipient daughter ion immediately prior to decomposition.

In the first instance shown (A), the daughter ion would gain extra momentum. By contrast in (B) fragmentation leads to formation of an ion retarded with respect to the motion of the ion beam. In (E) the products move orthogonally to the ion beam which has no effect on the measured momentum.

When the neutral and the fragment ion drift apart (le. only a small amount of energy is released) a metastable peak which is narrow and gaussian will form. This commonly occurs for endothermic processes. If however the fragment ion and neutral repel then there is a discrete kinetic energy release giving the particles equal and opposite momenta. As the fragmentations are generally extremely rapid a large proportion of the Chapter 1 31 collisionally activated ions will decompose in the collision cell. to form what has been described as an exploding sphere. The size of this "sphere" is determined by the amount of energy released. The physical constraints of the collision cell can become important in determining the shape of the observed peak. As the magnitude of the kinetic energy release is increased the shape changes from a narrow gaussian shaped peak to a broad gaussian peak , then to a flat topped peak at intermediate energies and finally to a dish shaped peak when the kinetic energy release is large. The slits on the cell prevent ions which have moved significantly orthogonally (in the X direction) from continuing along the ion beam. The amount of energy that can be released is dependent on the energy profile for the decomposition reaction. These processes are shown in Figures 1.10a and b.

trarsition state

'_-"-products

E

E reacting configuration

reåctmg configuration products

Figure 1.L0a. Peak Shape Figure 1.L0b. Peak Shape resulting from sÆ+ll Kinetic resulting from tars€ Kinetic farXe. Sr+.4 tl Energy Release. Energy Release. Chapter 1. 32

The energy release will depend upon the amount of excess energy in the transition state and the relative energy of the transition state and the products. The fraction of the reverse activation energy released as kinetic energy for reactions involving a cyclic fransition state increases as the size of the transition state decreasesT4. It is generally considered that decompositions of isomeric ions will involve transition state structu¡es of different energies and consequently different kinetic energy releases will be observed. If two reacting configurations are available, composite peaks may be observed. This observation is of great importance in the investigation of gas - phase rearrangements. If a negative ion undergoes complete rearrangement followed by conversion, then comparison of the peak widths for the rearranging negative ion and the independently synthesised product ion should be the same.

L.8 Rules for Fragmentation of Anions.

Recently a reviewT5 of the fragmentation reactions of even electron anions has appeared. It was shown that the fragmentations could be explained in terms of a number of 'rules'. For convenience, fragmentations from delocalised ions (eg. enolate ions) are represented as proceeding from the negative charge localised on one atomf . In summary the decompositions were classified, as follows:-

f Fragmenüations remote from and uninfluenced by the charge cmtre have been reported for systems where simple heterolytic cleavage involving the charge site is energetically unfavourable. Chapter 1 JJ

1.8.L. Simple homolytic cleavage.

The most common homolytic cleavages involve the losses of H'

(eg.76 equation 1.25) or Me' (eg.77 equation 1.26). These losses are expected to be most favourable when a stabilised radical anion is formed.

FJ Et \o CH-CO"- c-co"- + I{' (1.25) /' /o B Et -l-

PhcH-o-cH¡ PhCHO-' + 'CH3 G.26) --+ l-.8.2. Formation of an ion - neutral complex.

Ion - neutral complexes are formed when a heterolytic cleavage of a parent ion occurs to yield two species which are bound by ion - (induced) dipole interactions and hydrogen bonding. The binding energy may be quite large ( up to 80 kJ mol-1 ). These complexes may decompose by dissociation ( eg.78 equation L.27) or may undergo proton transfer prior to dissociation ( eg.equation 1.28). oo :r, ll ll HrCJC¡C-cH3.\? o il HrC:C:O + C- cH¡ (7.27) o HrC:C:O il _c-cH3 )l HC:C-O- + CH3CHO (1.28) Chapter 1 34

When (l) an acidic site is present in the negative ion of similar

acidity to the initialty deprotonated ion (eg.79 equation 7.29) or (li) a simple fragmentation does not occur readily from one site in an anion (eg.76 equation 1.30), then proton transfer may occur. In certain instances proton transfer may be effected by collisional activation.

o o

H C OMe

It o o o ",õl.,Ao* H2C:C:O + (1.2e)

Et /'CH-CO.- Et

I E._V 7t. \-I tõ". Et'L\",'llu'/C-CO2H*ll C:C:OI OHI* f:Ç:Q + HzO (1.30)

1-.8.4. Rearraneement, precedins fraementation.

Rearrangements may proceed either via the intermediacy of ion -

neutral complexes (eg.80 Beckmann rearrangement, equation 1.31,) or by an associative mechanism (eg.8t Claisen ester enolate rearrangement,

equation 7.32). Chapter 1 35

H"C H2C C:NO / /C:NOH [{cnr=c=NcHr¡ óH ] H¡C HsC

CHr=Ç=¡ç¡1, + HzO (1.31)

- Ph\ CH.CO 2 (1.32)

I

'Remote' fragmentationsS2 appear to proceed in systems where simple heterolytic cleavage do not involve the charge site cannot occur. They are quite pronounced in long chained saturated carboxylate salts

@g.æ equation 1.33), and sulphonates .

H \tc" n + 4cH;¡$ + H2

2 + 4c^;îco; (1.33)

LS Rearransement Reactions.

Anionic and base-catalysed rearrangements are generally less well

known than carbonium ion rearrangements and are often thought to be few in number. The migration of an electron rich alkyl group, for example to an electron rich group seems highly unlikely and yet such Chapter L 36 rearrangements do occur. In solution, the Wittig rearrangement is a well known example.

There are, in fact, many examples of rearrangement reactions occurring in solution. These include the acyloin (cr-ketol)8a, anionic ortho

Fries85, anionic oxy-Cope86, Baker-Venkataram art87 , benzilic acid88, Favorskii89,9, Fritsch-Buttenberg-Weichell91, Grovenstein-Zimmermang2, Hofmann93, Lossen94, Neber95, Payne96, Ramberg-Bäcklund97, Truces-

Smiles98, Wittig rearrangements99 and many carbanionic cyclisationelOO. This list is by no means exhaustive but is designed to illustrate the wide range of rearrangement reactions known in solution.

In most cases the scope of synthetic applications of rearrangement reactions and the mechanisms of such processes operating in solution have yet to be fully examined. From the perspective of the work described

in this thesis it is useful to know something about the mechanism of a reaction in solution in order to compare that with the results of cognate

gas phase studies.

In solution, mechanistic studies are often complicated by a number

of factors, not least of which is the solvent. Not only can the reagents react by a number of different reaction paths, but the products obtained may undergo further reaction. Most importantly the solvent and the counter- ion present modify the reactivity of a base or nucleophile in solution. Rearrangements in solution will be influenced by solvent and counter- ion. In this thesis some rearrangements known to proceed in solution are Chapter 1 J/ examined in the gas phase. In this way it is hoped that knowledge about the "intrinsic " reactivity8 of rearranging systems may be examined.

In the following section the effect of solvent and counter-ions will be dealt with in detail. In particular, comparisons will be made to indicate where divergent behaviour has already been observed between solution and gas phase reactions.

1.Ll- Intrinsic Reactivitv.

In the gas-phase, reactions of anions with neutral molecules or - radicals can be probed in the absence of the affects of the counter-ion and solvent. In this way the "intrinsic " reactivity of the ion may be examined.

The simple fact that the anionic species has been produced with thermal energies does not imply that the reactions (and in the context of this work, the rearrangements) of this ion will be the same as in solution. This is because the anions are influenced myriad of complex by ^

8 There appears to be no definition of the term "intrinsrc" reactivity. Kayser (ref. 101) has defined "intrinsic" reactivity in terms of a theoretically defined quantity. In these terms it is the theoretically calculated coefficient of the frontier orbital of the reactants and reagent. In the opinion of this author this is not satisfactory , especially when experimentally determinable quantities are measured. The "infiinsic" reactivity of an anion is solely its reactivity in the absence of solvent and counter ion. Steric effects may affect the reactivity of an anion, however to take account of this affect requires the experimentalist to calculate the affect of this group on the reactivity of the system. Thus the result is not the simplest experimentally determined quantig and is subject to the vagaries of molecular orbital calculations (in particular the level of the calculations and any approximations made therein). Chapter L 38 interactions between the solvent, reactants and reagents in solution. Furthermore in classical chemistry there is a constant transfer of energy between an ion and its environment. In aacuo, in the gas phase this type of energy transfer does not occur.

1.L0.1 Solvent.

Heterolyic cleavages involve the formation of charged intermediates, which are usually solvated to a differing extent to the intermedi¿¡sslO2. This is best illustrated by considering the solvolysis of ú - butyl chloride (equation 1.34). In water the observed enthalpy of formation of the solvated intermedi¿¡ss103 is g5 kJ mol_l, whereas in the gas phase the = 630 k] mol-l. ^Hof

H.C-\ H.C,l equation L.34 H3C";C-Cl t\C+ CI Hsc H¡C' HsC -{> - This ¡eaction is quite favourable in solution but not favou¡ed in the gas phase. However, the presence of charged intermediates on the reaction co-ordinate does not necessarily imply that increasing the polarity of the solvent will lead to an increase in reaction rate. For example, the

[3,3] sigmatropic rearrangement of N-phenyl-O-acetylhydroxylaminel04 (2, equation1.35) proceeds rapidly to completion in toluene. Under the same conditions, but using tetrahydrofuran as solvent, only 5% of the starting material has been converted to product. Chapter 1 39

But B u I I N\ 1. KHMDS H o rì (1.3s) -78oC, thr. CO2Me H3 o 2. CH2N2 / TÍß 2

The affect of solvation has been studied in the gas phase105. The rate of reaction of methyl bromide with hydroxide ion is one such example (equation 1.36). Solvating the hydroxide ion with successive molecules of water resulted in a significant rate retardation. (l*) CH3Br + (HO-/n á cH3OH + Br- +,^ [J],O (1.36) ñ

Table 1..1. Effect of Solvation on the Gas Phase reaction.

n k M-1 s-l

0 6.0+0.12x1011

1 3.8+1.5x1011

2 I.2+ 0.6 x LOe

3 < 1.2 x 108

oo 1.4 x 104

L.L0.2 Volume of Activation ÂV#

The solvent may also influence a reaction in a manner seldom considered. The solvent may have a particular overall structurel06 by virtue of ion - dipole, dipole induced interactions, etc. In this way the motion of molecules and especially ions may be greatly influenced. For Chapter L 40 example, the rate of diffusion of a K+ ion from solutionlOT into 18-crown-6 's 10-12_sec. The viscosity of the solvent may at times affect the ability of the reactive intermediates to diffuse into the solution. This is true for uncharged species such as radicalsl0S where increasing viscosity increases the proportion of geminate pair recombinations.

Table L.2 The Affect of ViscositylOl on the Proportion of Geminate Pair Recombinations.

Solvent Viscosity Radical Precursors (centistokes) Di-t-buty1 di-t-butyl Di-!-butyl peroxyoxalate hyponitrite peroxide isooctane 0.562 0.086 0.098 0.12

20% Nujol 0.865 0.13 0.16 0.19

70% Nujol 5.85 0.31 0.36 0.48

L00% Nujol 92 0.il 0.68 0.76

(cH3) 3co2cocoo 2c(cH3) 3 (CH3)3C-O-N=N-O-C(CH ùz Di-ú-butyt peroxyoxalate di-ú-butyl hyponitrite

(cH3)3c- o-o-c(cH ùz di-f-Butyl peroxide

L.L0.3 oKa.

Relative acidity scales have been determined for a wide range of -compounds in methanol and dimethyl sulphoxide (DMSO)110. It is immediately apparent that solvationlll 6ur a dramatic affect on the measured acidity of the compound. This effect is dramatically shown in

Table 1.3. Chapter 7 47

Table 1,.3 Comparitive Acidities of Selected Compounds in solution and the Gas Phase.

Compound pKa ÅlI0¿6lda

Hzo DMSO Gas - Phase

PhCOzH 4.25 11.1 1.41.8

CFbC02H 4.75 12.3 1457

(CFIeCO)zCHz 8.9 13.3 7439

CHgNOz 10.0 77.2 1492

PhOH 10.0 18.0 1.462

CH2(CN)2 11.0 11.0 1407

CHsCONHz 15.1 25.5 1430

CH3OH 15.5 29.0 1595

Hzo 15.75 32 7637

HNs 4.7 7.9 t440

a kI mol-1

L.L0.4 Counter Ions.

The nature of counter ions in solution is usually deemed to be of little consequence in affecting the course of a reaction. If one considers a simple experimentally measured physical quantity such as pKa, then the change from Li+ to K+ is almost negiligiblell0. This simple subsitution often does result in dramatic changes in the rates and direction of reaction and these are invariably the result of ion-pairing between an anionic reagent or chelation of the cation to the substrate. Chapter 1 42

The metal ion may lead to preorganisation of the reacting species. This has been demonstrated in a variety of ways. An investigation into the aldol reaction of (3) by Denmarkll2 has shown that a closed transition state is formed in the presence of a counter ion, leading to the formation of the syn product (a). The addition of Krytofix@ í2.2.21 (a crown ether) effectively eliminates the possibility of co-o¡dination of the metal ion to the reactants. The aldol reaction still proceeds, however an open transition state is formed leading to the formation of the anti product [( 5 ), Scheme 1.1 ]. Diastereofacial selectivity due to the presence of the counter ion has been reported in other systems.

R L R 1 L I H I M\I o' L H R4 o R3 R3

Closed Transition State Open Transition State

O M* base , TTIF o o -7goc + CHO H3 H. OH HO 3 4 ,syn 5 ,øtti

Scheme 1.1

A fu¡ther example was reported by Crombie st al.l73 Treatment of the pyrone (6) with sodium methoxide (NaOMe) (equation 1..37) afforded the substituted phenol (7) whereas the use of magnesium methoxide (Mg(OMe)2), (equation 1.38) gave (8) and (9). The effect is by no mearìs subtle and has been attributed to the role of chelation. Chnpter 1 43

CH J COMe zCHs CH J RO CH= C a NaOMe R o _> ( 1.37) I MeOH OH H Rl o COR 6 7 Mg(OMe )2 MeOH

CH3 cH3o cocH3 zCHs

+ ( 1.38) o ocH3 co2cH3

I 9

L.L0.5 Ion Pairine.

This is an affect which is usually easily measured. It has shown that for a variety of organic salts in dimethyl sulpho¡¡¿s114 the association order is K+ < Na+ < Li+. This is demonstrated by comparing the association

constantlls of CsOMê ( Kassoc. = 200 ) with LiOMe ( K"rro.. = 108 ) in DMSO. It is dependent of the ability on the counter ion to chelate to the nucleophile and follows the order:

RCO2- > RCONR'- > RCOC-R'2

A relationship between ion-pairing and basicity unfortunately does not exist. It is well documented that K+ pairs strongly with the methoxide

ion whereas the equally basic thiazolide (L0) and phenyl acetylene ions (11.) do not. Chapter 1 44

S N. \J Ph-C=C - 10 11

Ion pairing is observed for many reactions. The lone pair of electrons of the anion is localised in ion-pair formation and is thus unavailable to react at an electrophilic site. This has been recently highlighted by Cacciapaglia and ¡4"tr¿o11ni116. The rate of cyclisation of ethyl (5-bromopentyl) acetoacetate ( L2 ) ( equation 1.39 ) decreased by a factor of 105 when the counter ion was changed from Mg++ to K+.

zBt Br M* corEt (1.39) COrEt --+ t2 COrEt

grr¿ne117 has demonstrated that in solution, the Claisen rearrangement ( a [3,3] sigmatropic rearrangement ) of the alkoxide ( 13 , equation 1,.40), proceeds - 1016 - !017 ümes faster than the corresponding rearrangement of I,5 hexadiene (1,4, equation 1.41). A resonance contributor involving disruption to the cr C-C bond may also be drawn for the alkoxide ( 1.3 ). o' (1.40)

13

4/g ._

=rJ' L3 Chapter 1 45

(1.41)

L4

The [1,3] sigmatropic rearrangement of o-phenyl selenyl ¡s¡enssl18 (equation 1.42) is also affected by ion pair formation. The lithium enolate has been found to be very stable, whereas the sodium or potassium enolate ions undergo facile rearrangement. [1,3] and [1,5] Sigmatropic shiftsl19 are also found to proceed more rapidly as the degree of ionisation is increased.

o SePh

NaH or KH 3 3 (7.42) THF/HMPA --+

The stability of propargylic anions to rearrangement and 3-siloxy allyl metallated species to the Brook rearrange-sn¡120 have also been reported to be similarly affected by the choice of counter ion.

A reversible rearrangement involving migration of an ethoxy carbonyl group from N to C is influenced by ion pairingtzt. The thermodynamically more stable anion is thought to be the (15) and this predominates where ion-pairing is unimportant, but the amide ion (L6) forms a strong ion-pair with Li+ driving the equilibrium toward the left (equation 1.43).

Na* Ph2C(CO rEt)NPh .-- Ph.CNPh.l (1.43) Li* corEt 1 th \ - b 5 Chapter 1. 46

The vinyl amide ion (17) is alkylated on when the counter - ion is potassium. Flowever, when the Ag+ - amide ion-pair is prepared alkylation occurs exclusively on carbor122 ( Scheme 1.2 ).

H o

M* cH3I cH3I t7

o o

H I + CH¡ å M* = As*) M rcl Scheme 1.2

Ion pairing may affect the stability of the reactive intermediates. When the reactive intermediates are of similar energy it is possible that the formation of one product may be favoured by this process. This effect has been noted by Staley and Erdmanl23 who found that 6-methyl-6- phenylcyclohexa-1,,3-dienyl anion (L8), yields many products ( Scheme 1.3, Table 1.3 ); the proportions of (19), (20) and (21) are most significantly affected by the choice of counter-ion. The hydride transfer reaction from

(22) to the starting material (18) is fastest in the presence of the solvated Li+ ion. By contrast, the cyclopentadienes (20) and (2L) are formed in the highest yield in the presence of the Cs+ ion. Chapter 1 47

H Ph CHs CH¡ CH¡

22 t9 / lvb

18 \ CH¡ CH¡ CH: Ph

CH3 + - 20 2l

Scheme 1.3

Table 1.4 The Affect of Counter ion on the ratio of products formed.

Product Counter Ion

Li Na K Cs

L9 24 6 6 3

20+2L 5 6 t4 30

A change in mechanism has also been observed when benzyl thioallyl s¡þs¡e124 are deprotonated. Two rearrangements are possible from the deprotonated ion (23); the 17,21 and/or Í2,31 thio-Wittig rearrangements to yield (2a) and the Sommelet rearrangement yielding (25) shown in Scheme 1.4. In the presence of the Li+ ion the thio - Wittig rearrangement is favoured whereas the Sommelet rearrangement is favoured where free ions are present. A similar result is obtained for Chapter 1 48 dibenzylthioetherl2S; the Sommelet rearrangement being favoured from "free" ions whereas the contact ion pair leads to the formation of the Stevens type rearrangement product.

S. I CH-CH 2-CH=CH2 SCH2-CH=CH 2

23 24

S

CH=CHz Scheme 1.4 25

Another example is shown in Scheme 1..5. In this case, skeletal rearrangement of deprotonated dihydro [10] annr¡len-11--onet26 (26) in the presence of the "soft" potassium ion lead to the dihydronaphthalene (27) whereas the tricyclic alkene (28) is formed when Li+ is the counter ion.

o (Me) 3SiO,

OC KDA / THF I -20 (cH¡) 3sicl

26 27

LDA / TItr / -20 OC (cH3)3sicl

)g

) )

Scheme 1..5 28 Chapter 1 49

ç¡a^727 has shown that homolytic and heterotytic cleavage oÍ 7,7,2 - triphenyl - 2 - methylbutanoxides (29, Scheme 1.6) occurs competitively with the nature of the counter ion exerting a determinative role on the outcome. Anionic cleavage was favoured when the counter ion was K+, whereas Li+ favoured radical formatiorr

lvb O- M+ NrÊ o M* tt Homolysis I I Et-C-C-Ph Et-C. C- Ph tt I Ph Ph Ph Ph 29

Heterolysis

lvß O til N,þtt lvß Et-C.M++Ph-C-Ph + I Et-C-C-Er Ph tt Ph Ph

o OH lvß il I I Ph-c-Ph + Ph-CH-Ph Et-CH* I Ph

Scheme L.6

Table 1.5 The Effect of Counter ion on the

Course of the Decomposition of the Alkoxide (29).

Counter lon kh"t"ro

khorn,

K 2.2

Na t.6

Li 0.05 Chapter 1 50

On occasions, the size of the counter ion is found to be particularly important128. For example, deprotonation of pyridone (30) occurs very readily with lithium diisopropyl amide, leading to rapid rearrangement

(equation 1.44).

Ph Ph

LiNiPrt (1.M) o THF, -76OC I I CH Ph Ph Ph -cH, 30

Changing the base to potassium dimsylate proved unsatisfactory as no reaction was observed even under forcing conditions. It was argued that steric congestion arising during the formation of the intermediate (31) was severe and thus a larger cation, cannot be accommodated.

Ph

o'M*

Ph 3

31 1.L0.6 Assresation.

The role of aggregation in solution is poorly understood. It has been only recently shown that simple reactions such as the aldol condensation may proceed via a complex series of aggregated species.

- Chapter 1 51

Consider the reaction of lithium diisopropylamide in tetrahy-drofu¡an with a methyl ketone. Prior to deprotonation, the base is present as a mixture of dimeric and monomeric forms (32) and (33), of which the monomeric form (33) is considered to be most ¡s¿q¡iys129.

Pr',,.. Prt,.. N-Li-(THF) Pri/ Prl n

(THF) n

32 33

The addition of two equivalents of the ketone leads to displacement of THF ligands. Deprotonated by the amide ligand ensues to produce a tetrameric "ladder" - type u¿¿u6130 (34). Addition of another equivalent of the ketone leads to closure of the adduct to afford the "cubic" adduct (35,

Scheme 1.7).

Subsequent chelation of a ketone leads to co-ordination of the carbonyl oxygen to a lithium cation. This enhances the susceptibility of the carbonyl group to nucleophilic attack by the proximal enolate ion. Re- organisation of the ligands about the "cubic" adduct (36) results in displacement of the weakly co-ordination carbonyl oxygen for the alkoxide ion to yield (37). Crystal structures for the aggregates along this reaction pathway have been found and are used to account for the products from enantioselective reactions and to explain C- vs O- alkylation prod.rs¡s130. Chapter L 52

THF o I {t Li R-C-CH, + LiNiPr2 ). / o- Li \ N Y I \ 34

{ a\ R Li-o l-..- l-..

{ 35

,/-,. l.-

- ( 36 37

Scheme 1.7

It has been argued that radical cleavages become favourable as the degree of aggregation is increased131. Kochi has demonstrated that alkyl radicalsl32 can be formed readily from tetrameric (LiaMea¡ methyl lithium. Lithium dialkylamides were studied using cyclic voltammetry and found Chapter 1. 53 to be better electron donors in aggregated specis5133. Electron transfer reactions are sometimes observed during the reaction of Grignard reagents to account for the formation of pinacol products. Hirota and Weiss¡¡¿n134 have demonstrated that the paramagnetic species are dimeric.

Complex interactions frequently arise between the anion, counter ion and solvent, especially when dealing with carbanionic species. This is dramatically highlighted by an X-ray investigation of the organometallic reagent benzyl lithium. Structural changes occur to the site of lithium binding when the donor ligands are changed. Thus with crystallisation of benzyl lithium in diethyl s¡hs¡135 yields (38), whereas addition of an excess of tetramethylenediamine (TMEDA¡136 ¿ffs¡ds (39). Addition of 1 equivalent of tetramethylenediamine to a solution of benzyl lithium in tetrahydrofuranl3T leads to the formation of the benzyl lithium (40).

OEt 2 .ß) ü ..H Li H

oEr 2

3E 39 40

Theoretical calculations were in accord with the obserrru¡¡e¡s137.

Significantly, differences in the reactivity of these complexes was observed during the Haller - Bauer cleavage of optically active ¡.¡onss138.

The natu¡e of the solvent can dramatically affect the stability of an anion and its propensity to rearrangement. For example, the 7,2 Chapter 1 54 rearrangement of the Grignard reagentl39 formed in equation 1.45 occurs very readily in diethyl ether (a poor donor ligand) but proceeds only sluggishly in tetrahydrofuran.

Ph Ph Ph Ph \^/ \/ HrC=CH.CH2-C-CH2- MgCl J-.+ Ph-CH2-C-CH2'MgCl (1.4s)

Equally dramatic is the affect of solvent on the reaction of lithiated acetylenes and allenes. Brandsma et al.140¡¿ys shown that in diethyl ether the o- lithio trimethylsilyl cumulene ether (41) undergoes a reversible Brook rearrangement to yield the alkoxide (a2) (equation 1.46) whereas (a1) is quite stable in tetrahydrofuran. This again reflects the "donicity" of diethyl ether vs tetrahydrofuran.

R 1 ri R\ TMS la- 11- /1- C .-_> (1.46) R2'/ OTMS R2/ ou 4t 42 -

A number of solvent affects are observed during alkylation of propynyl lithium (a3) with propylens e¡¡¿s141. In tetrahydrofuran, alkylation occurs at both of the sites of charge localisation to afford (aa) and (45). The addition of one equivalent of hexamethylphosphorous triamide changes the stability of the delocalised anion, resulting in increased steric bulk at the terminal position. As a consequence, attack is favoured at the site adjacent to the alkyl group with (45) formed as the major product. A further increase in the amount of hexamethylphosphorous triamide is thought to favour formation of a "free" ion which may then undergo hydride transfer to produce (a6). The results are summarised in Scheme

1.8 and Table 1.6. Chapter 1 55

OH I P-C=C-(CIJ)zCH'CH3 o 44+ /\ Pr-C=C-CH2Li CHs Pr oH 43 H2c: I cH2cH-cH-cH3 45+ OH I PrCH=C=CH- CH2 CH-CH-CH3

46

Scheme 1.8

Table 1-.6 The Effect of Solvent on the Product formed in

Scheme 1.8.

Solvent Product

44 45 46

THF 45 55

HMPA (0.5-1.0 equiv.) T2 87 1

HMPA (5.0 equiv.) 5 19 7L

A different result was obtained during studies of L,or proton shifts in

3- ,4- and 5- lithio trimethyl si1¿n"s142. In contrast to the results described above, the lithiated species were stable in diethyl ether. In tetrahydrofuran

the 3-lithio compound underwent a [L,5]H shift (equation 1.47), while the 4- lithio species underwent ring closure (equation 1.48) and the 5- Iithio species underwent both ring closure (equation 1.49) and a [1,7]H shift Chapter 1 56

(equation 1.50). The relative proportion of products obtained from the 5- lithio s-pecies are dependent on the polarity of the solvent.

)¡ Li (1,.47)

Li -- Li+ (1.48) /

- + MeLi

\si- Li* / Si (1.4e)

+ MeLi

(1.s0)

The alkylation of carboxylic acid salts with methyl lithium only proceeds when the polarity of the solvent is increased from diethyt ether to 80% tetrahydrofuran in diethyl .1¡s¡143. The reaction is postulated to proceed via a dissociative mechanism. It is thought that solvation of the intermediates, in particular allyl lithium is most important (Scheme 1.9). Chapter 1 57

H¡9 H¡9 l¿cHz MeLi H,C:CH.L H,C:CH ;¿CHS tcorH Etzo tc-cH, il' o MeLi 807o THF-EI ,O

o II Scheme 1.9 HsC CH 2C- CH J

The solvent and counter ion are often crucial in determining the course of the reaction. It is possible that the role of the counter ion can be more positively identified by the use of powerful chelating agents. Several years ago, crown ethers and cryptands ( were shown to complex very strongly with alkali metal g¿¡isns144. 18-Crown-6 has been found to complex particularly effectively with K+, 12-Crown-4 with Li+ and 15- Crown-S with Mg**. Furthermore, it was shown that many simple inorganic salts could be dissolved in non polar solvents such as benzene, dichloromethane or acetonit¡ile. Potassium permanganatel45, for example could be obtained in a concentration of 0.06 M in benzene by adding 18-Crown-6. Electrical conductivity measurements show that "free" ions are present. In solvents of low polarity, solvation of ions is likely to be very weak. It is thus now possible to study the reactivity of anions in a larger range of solvents than were previously accessible.

¡io¡1¿146 investigated the relative rate of substitution of benzyl tosylate by various nucleophiles in acetonitrile. The potassium salts of the nucleophiles were not very soluble in acetonitrile, but the addition of 18- crown-6 allowed a high concentration of the inorganic salts to dissolved Chapter 1 58 into the solvent. The results were compared with the data obtained independently for reactions in water

Table 1..7 The Relative Rates of Nucleophilic Substitution of Benzyl Tosylate in Water and in Acetonitrile in the presence of 1.8-crown-6.

Relative Rates (Acetonitrile) (Water)

N3- 10.0 20

OAc- 9.6 1.0

CN- 2.4 250

F- 7.4 0.2

CT 1.3 2

Br- 1.3 1,6

I- 1.0 100

SCN- 0.3 125

There are two obvious features evident from this table. Firstly the ¡elative reactivity of the nucleophiles varies considerably more markedly in water than in acetonitrile and secondly the order of reactivity is not the same. The halides exemplify this point. In water nucleophilicity is in the order I- >> Br- > Cl- > F-, whereas in acetonitrile the order F- > Cl- , Br- ) I-. This result is ascribed to solvation of ions in water being F- > Cl- > Br- > I-. Nucleophilic substitution had been studied in the gas phase and had yielded the same relative rates as Liotta's work. The rates observed in Chapter 1 59 the gas phase are estimated to be approximately 1011 times faster than in rolrr¡ien147.

Decarboxylation of benzisoxazol.e - 3 - carboxylates is enhanced by a factor of 105 when the solvent is changed from methanol ¡o p¡4591a8.

The rate of decomposition of alkali metal carboxylate salts is enhanced L3 - 100 fold by the addition of 18-crown-6149. Whilst it is the free carboxylate ion that decomposes, it has been noted that the rate enhancement correlated best with the stability of the intermediate anion formed.

A remarkable change in selectivity has been noted for the reduction of 2-cyclohexenone with LiAIFI¿ or LiBru in ether when l2.Z.Llcryptate is u¿¿s¿150. In the absence of the complexing agent the sole product obtained is cydohexenol obtained by a 1,,2 addition of hydride ion. In the presence of the complexing agent, L,4 addition occurs to afford cyclohexanone as the major product.

The Cannizarro reaction has been shown to require a Lewis ¿si¿151 for hydride ion transfer to occur (equation 1.51).

oo HOO ll ll oH- til Ph-C-CH _ Ph-CH.C-OH (1.s1) 30"c ca**>Tl**>sr**>Ba*

Addition of a suitable crown reduces the yield of product dramatically as the carbonyl groups are not polarised by the Lewis acid (a7) thereby disfavouring nucleophilic attack at the carbonyl group. Chapter 1 60

orM8.o 'n Pd'" nn/ 47

Addition of ketone (a8) to KH in THF leads to reduction1s2, whereas deprotonation occu¡s when cryptand Í2.2.21is present (Scheme 1.10).

o OH il KH, THF I C- Ph CH.Ph 48 KH, THF 18-crown-6 , cH3I

o il C-Ph Scheme 1..10

From the foregoing discussion it is clear that the factors influencing reactions in solution are very complex indeed. While it may be argued that the gas - phase is a non polar medium, reactions performed in a non polar medium such as hexane will not necessarily yield the same results as the gas phase. In solution, the interactions observed are not readily separated. The use of crown ethers and cryptands provides a valuable aid in identifying the role of the counter ion. None the less, this method is not without its limitations. For example, the reaction of pyridinslS3 \^/i1þ diethyl magnesium has been studied in diethyl ether and tetrahydrofuran

(THF) in the presence and absence of 15-crown-5 (equation 1.52). In these experiments 2- and 4- ethyl pyridine were obtained. Ch"apter 7 67

Et

1). Et2Mg - THF + (r.s2) 2). H2O 3).tol EI

The rate of reaction was increased by a factor of >105 in ether but the ratio of the products had changed dramatically. It was proposed that the reactive species was no longer Et2Mg , but Et3Mg- and Et5Mgz-. The higher aggregates were considered to be more basic and underwent reaction via an electron transfer process.

Table 1.8 The Effect of 15 - Crown - 5 on the Course of the Reaction Between Ethyl Magnesium and Pyridine.

Reaction ProductYield(%) Conditions 2-Ethyl pyridine 4-Ethyt pyridine

THF,40oC ,24hf . 0.6 < 0.01 THF ,400C ,24hf ., 6 76 L5-crown-5 THF,40oC ,72hr. 13 37 L5-crown-5

It has already been shown that in many reactions of alkyl lithiumls4 or alkyl Grignardlss reagents that the reactive species is the monomeric

species. As interesting as the result may be, it does not give a view of the fundamental reactivity of the alkyl anion.

The effect of crown ethers and cryptands on the reactivity of nucleophiles is not clear. ço¡.1156 contends that nucleophilicity is Chapter 1 62 enhanced to a greater extent than basicity. Flowever, PerraudlsT has stated that nucleophilicity is no lonþer manifested but that basicity is enhanced..

There are sound reasons to believe that reactivity of systems may change in the gas phase. Theoretical and experiment results indicate that in methoxide salts the C-H bond strength is affected by the increase in ionic character of the adjacent oxygen1S8. The C-H bond in sodium and potassium methoxide is very polar and the bond is much weaker than in methanol. This affect can presumeably be attributed to hyperconjugation (equation1.53). In the gas phase this o C-H bond is predicted to be even weaker than for the ion pairs.

H

<-> Hr.ç:e (1.s3) H

Table L.9 The Affect of Increasing the Ionic Character at the Oxygen atom on the Adjacent C-H bond.

Bond Strength Change'in

( observed ) Bond Strength (kJ mol-t¡ (kJ mol-t¡

H{H2OH 379 0

H{HzONa 337 42

H{H2OK 331 -48

H-CH2O- 311 -68

The rate of nucleophilic substitution has been found to between 1010 - 1020 times faster in the gas phase than in solution1sg. The Chapter 1. 63 acceleration of the rates of many other reactions will not necessarily be the same. is conceivable reactions being _ It that this may lead to some favoured in the gas phase relative to solution. Indeed, the alkylation of the cyclohexanone enolate ion by methyl iodide has recently been investiga¡sd160. In solution, alkylation occurs on carbon (equation 1.54) whereas in the gas phase reaction occurs on oxygen (equation 1.55).

o

o cH3I (1.s4)

OCH 3

cH3I (1.ss)

A similar divergence in gas phase behaviour from the results obtained in solution have been noted when n-propyl bromide was treated with methoxide ion. Nucleophilic substitution occurs in solution (equation 1.56) but elimination takes place in the gas phase161 (equation

1,.57). v ./\.r'OW (1.s6) I\rfeo - +MeOH + Br (1.s7) ^ An understanding of these effects will be of interest to the mass spectroscopist using this tool for structural elucidation as well as those interested in mechanistic problems. ha t 2

ifti

2J1, Introduction. 2.1.1- The Condensed Phase Wittig Rearrangement.

Ethers which can be readily deprotonated on a carbon o to the ether oxygen atom can undergo rearrangementl62 to yield an alcohol (equation 2.7). This is known as the Wittig rearrangement and has been observed in solution for a wide range of benzylic and allylic ethers. Eliminations and cr-cleavages163 often accompany this rearrangement and this is especially so when simple ethers are studied.

pht-i ,nõro* - * PhcH2oR - [u.[ I --* phcH-oä- Li (2.1) 49

The earliest example of this rearrangement was recorded for the sodium metal reduction of benzhydryl phenyl ether16a. Migration of the phenyl group gave diphenylmethanol in poor yield. It was unclear whether the reactive species was a radical anion formed by electron capture or an a-carbanion formed by deprotonation. This ambiguity was removed by Wittig and ¡5¡r¡1¿nn165 who demonstrated that the rearrangement was initiated by deprotonation at the benzylic position. Treatment of benzyl methyl ether with phenyl lithium produced 1- phenylethanol in very good yield ( equation 2.2 ). The thermodynamic driving force for the migration of the alkyl or aryl group is the formation Chapter 2 65 of the alkoxide ion. Kinetic studies have shown that the rate of reaction is first order in both ether ut',¿ 6"r.166.

PÐLi PhCH2ocH3 - PhçH-oH (2.2) CH¡

In spite of the early advances, the mechanism for the transformation is by no means well established. Initially it was thought that a concerted internal nucleophilic subsititution (SNi) mechanismlíT could be used to describe this process ( equation 2.3 ).

=-/ I/-'\'C- /\ -Q - (2.3) c-o¿ Rr,..)c-o -O-+ R2/

If the rearrangement is concerted then the rules arising from the theory of the Conservation of Orbital Symmetryl6S are applicable to reactions in solution. Thus a concerted process in which a suprafacial [ 1s, 2s I shift occurs is highly unlikely as it is symmetry forbidden. The allowed lla,2s I and f 1^s,2a I shifts with inversion at the sp2 carbon atom or antarafacial migration respectively are not geometrically feasible. The Wittig rearrangement could not be expected to proceed by a concerted reaction if these rules apply.

The same conclusion is drawn if one considers the number of rc electrons in the transition s¡¡¡s169. Hückel transition states leading to suprafacial alkyl migration is favoured when (ah+2)rc electrons are involved in the rearrangement. Such a transition state has aromatic character. By contrast, a Möbius transition state requires 4nr electrons and involves antarafacial alkyl migration. This Wittig rearrangement Chapter 2 66 involves 4rc electrons however due to geometric constraints the concerted reaction would not be feasible.

Dewar is reported to have claimed that MINDO calcul¿¡i6ns170 indicate that the Wittig rearrangement may be concerted. This result would clearly be at variance with the simplified rules advanced earlier. Whilst no reference to any published article dealing specifically with the rearrangement has appeared, calculations on the isoelectronic Stevens' rearrangement indicates that rearrangement proceeds via an antiaromatic cyclopropenyl type anion intermediate. Dewar stated that on the basis of the Bell-Evans-Polanyi principlelTt, the activation energy would be low for this very exothermic reaction and the transition state will be closer to the reactants in structure.

Woodward and Hoffmann suggested that predictions for anionic rearrangements based on the Conservation of Orbital Symmetry may require modification to account for the effects of metal ions. This is particularly so when a metal ion is strongly bound to the carbanion centre at the outset of the reaction. The p orbital of the metal ion would need to be considered. Wittig had shown that the rearrangement of phenyl benzhydryl ether for various alkali metal salts was Li+ > Na+ > K+ , indicating that the least dissociated salt rearranged f¿s1ss1172. This result suggests that the cation is strongly coordinated to the deprotonated ether as is described in the equilibrium (equation 2.4). Incorporating a metal ion into the transition state structure as described by Woodward and Hoffmann would yield an intermediate (50) isoelectronic with the intermediate obtained in the Stevens' rearrangement ( 51 , equation 2.5). Chapter 2 67

M M (Ph) + (ph) (2.4) 2C-O-Ph 2C-o-Ph+

(M=Li*,Na*,K*) 50

3 +/ oon - +/ l Ph-c-cHz-\,, --+ Ph-C-C-ì!'.., -o+ Ph-C-CH-N(Me) 2 Q5) H\ ilb 51

P\ PhCR=O cR-o \C-lvÊ .l HttC" Et tc' ùÊ Nrb H Et Et H

R_ H 52 R- Ph 53 Ph. P\ CR-OH CR-OH I I Httc Et Nde "{'* Retention Inversion of of configuration configuration

R=H 54 R=Ph 55

SchötlkopflTz examined the rearrangement of opticatly active tertiary alkyl benzyl ethers ( 52 and 53 ) in strong base. If a concerted process was involved then the product would be optically active. Flowever, under the conditions described above considerable diminition in optical activity was found. Rearrangement of ether ( 52 ) yielded

alcohol ( 54 ) in which 80% racemisation had occurred, whereas ( 53 ) rearranged to alcohol ( 55 ) with orúy 20% racemisation. The degree of Chapter 2 68 retention of configuration was reduced as either the polarity of the solvent or the temperature were increased174.

Stevens also found evidence indicating that a concerted mechanism did not occur and was further able to show that the course of allylic rearrangements was indeed different from the Stevens' rearrangement. It had been shown that migrations of allyl groups in the Stevens' rearrangement occur with almost complete inversion17S, whereas almost complete retention of configuration was observed for the Wittig rearfangernsn¡176. Clearly the mechanisms for these rearrangement are different.

To account for these results two dissociative mechanisms have been proposed177. These are outlined in Scheme 2.L. The mechanisms involve the intermediacy of either a ketyl radical anion - radical pair ( 56 ) or an anion solvated ( 57 ). The tightly bound intermediates may then collapse in the solvent cage to yield the Wittig product or escape to react intermolecularly.

a PhCH-O + R 56 R\ PhCHOR + ll' + PhDc-o- 'R PhCH=O + H 57

Scheme 2.1 Chapter 2 69

Rearrangement via a dissociative mechanism would lead to an increas_e in the transition state volume of activation As the pressure ^V#. applied to the reaction mixture was increased the yield of rearranged productlTS decreased (90% conversion at 0.L MPa; 20"/o conversion at 1000 MPa). Most importantly an optically active ether was studied and the degree of retention of configuration remained the same irrespective of the

Pressure applied to the system indicating that only one mechanism lead to the formation of products.

Lansbury and PattisonTTg studied the rearrangement of alkyl benzyl ethers and found that the migratory aptitude was t Bu >> i Pr > Et > Me t. Tertiary alkyl benzyl ethers were also investigated and the migratory aptitude compared with the rates of thermolytic cleavage of peresters. In both studies the migatory aptitude was deemed to be consistent with bond dissociation energy; the rate of migration decreasing as the bond dissociation energy increased. Lansbury and Pattison placed an important caveat on this work. In their studies the ethers were treated with butyl lithium in diethyl ether. Subtle changes to the counter ion or solvent may result in a change of mechanism.

+ Rächardt (ref. 180) has shown that this order is not found for homolytic cleavage of the C-O bond of methyl alþl ethers. The bond dissociation energy of dimethyt ether is found to be approx.376 kJ mol-l (90 kcal. mol-l). However, the difference in bond dissociation energy between alþl groups Me and I Bu is only 8.5 kJ mol-l (2 kcal. mol-l). Deprotonation at the a-carbon results in the formation of a negative charge adjacent to the oxygen atom. This is likely to significantly reduce the (- I) inductive effect of the oxygen atom considered responsible for the anomalous values for the bond dissociation energy. Chapter 2 70

Garst and Smit¡181 s¡¿mined the rearrangement of benzyl hex-S- enyl ether ( 58 ).

IOH Ph-CH (2.6) Ph \ (cH2)4cH-cHz 58

OH I Ph-CH Q.n

The rearrangement was considered to occur both intra- and inter- molecularly, with the intramolecular process predominating. Rearrangement within the solvent cage ( equation 2.6 ) ís considered to be much faster than hex-S-enyl radical cyclisation ( k = 105 sec-1 ) and consequently it is argued that cyclised product ( equation 2.7 ) results from the minor intermolecular rearrangement process. The authors noted that the intramolecular process may proceed via an alternative mechanism.

The free radical mechanism does not appear entirely satisfactory. Pinacol products which are observed for many reactions in which ketyl radical ut ions13L ,734 are formed I eg. Grignard single electron transfer (SET) mechanism I have not been reported.

Furthermore, attempts to use chemically induced dynamic nuclear polarisation ( CIDNP ) experiments to demonstrate the intermediacy of radical intermediates have yielded unexpected results. Chemically Chapter 2 77

Induced Dynamic Nuclear Polarisation182,183 is a phenomenon caused by the coupling of radical pairs in a solvent cage. It is measured using NMR and is manifested by the enhancement of absorptions or emissions in the spectra. Felkin and Frajerman noted that during a study of butyl alkyl s¡¡s¡s184 an elimination product and not the Wittig rearrangement product exhibited a strong CIDNP effect. This may be due to spin lattice relaxation resulting from the presence of the metal ion. A similar result was obtained for the rearrangement of allylic s¡þs¡s185. This interesting result will be discussed later in this chapter.

The results of earlier *o.¡186 indicates that under appropriate conditions evidence for anionic mechanism can be found. Dibenzyl ether was treated with a vast excess of methyt lithiumlsT yielding tl diphenylethanol ( 59 ) and L-phenylethanol ( 60 ) ( equation 2.8 ). As the polarity of the solvent increased intermolecular rearrangement products were observed ( Table 2.1 ).

MeLi PhcH2ocH2Ph PhCHCH"Ph + PhCHCH. (2.8) -U¡-> lo at OH OH

59 60 Chapter 2 72

Table 2.L Products obtained on Rearrangement of Dibenzyl Ether

with Methyl Lithium ( in excess ).

Solvent Ratio se (%) 60 (%)

Ether 100 0

Ether - THF L:1 97 3

TI-IF - Et3N L:L 93 9

THF 82 18 THF - ( MeOCH2CH2OMe ) 1:1 75 25

RautenstauchlSS has investigated the rearrangement of benzyl vinyl ethers. For example, (Z) - benzyl propenyl ether ( 61 ) required quite vigorous conditions to effect rearrangement. O.ly one product ( 62 ) was obtained resulting from the migration of the propenyl moiety. If a radical

had intervened then both the (E) and (Z) alkenes l( 62 ) and ( 63 ) respectively ] would be expected due to the very rapid rate of radical inversion. Vinyl anions have been shown to be configurationally s¡¿61s189 under similar conditions. These results were found to be consistent with an anionic dissociation recombination mechanism

(equation 2.9 ).

OH n-Bu Li I Ph-cH"-o¡ rr vrr2 "\J Ph-CH (2.e) TMEDA, -27oC , oln 61 62 63

Migrations involving the phenyl group may proceed by an associative mechanism ( equation 2.10 ). Nucleophilic addition to the Chapter 2 /J

aromatic ring would afford an unstable spiro intermediate ( 64 ). Evidence for this intermediate has yet to be found.

ï

Ph-CH-O-Ph Ph-çH-o' (2.10) Ph

64

2.L.2 The Gas Phase Wittie Reananeement.

Unlike solution chemistry, molecular motions in the gas phase are not constrained by the solvent sphere or influenced by the counter ion. Consequently the restrictions of the Woodward-Hoffmann rules have been called into question. It has been argued19O that the treatment of Woodward and Hoffmann may be only applicable where molecular vibrations are not severe (ie. the movements of the atomic nuclei do not deviate significantly from the positions in the ground state). In the gas phase a molecule may exist in a vibrationally excited electronic ground state and the motion of the nuclei may be sufficient to destroy the symmetry of the transition state. At best, these rules are only useful indetermining incipient reaction paths. On this basis extreme caution must be used when using the Woodward - Hoffmann rules to draw even the simplest of analogies with solution chemistry.

Evidence for the Wittig rearrangement proceeding in the gas phase has recently been found. Chapter 2 74

2.1.2.1 Alkyl- and Aryl- Benzyl Ethers.

Alkyl- and aryl- benzyl s¡þs¡s191 were examined along with the corresPonding alkyl- and aryl- benzyl alcohols. The spectra were found to be very similar. This is most readily illustrated if the collisional activation mass spectra of deprotonated ethyl benzyl ether ( Figure 2.1 ) and ethyl benzyl alcohol ( Figure 2.2 ) are considered. The peak heights were very similar indicating that the same proportion of ions was decomposing via a given reaction channel. Furthermore, the peak widths of the daughter ions were the same in both spectra. Thus it may be inferred that the ion decomposition characteristics were the same. The fragmentations of benzyl ethyl ether are summarised in Scheme 2.2. There is only one fragmentation which proceeds from the deprotonated ether and that is the loss of the "Et'" radical (equation 2.7I). Rearrangement to the alkoxide ion ( 65 ) precedes all other fragmentations. Several decomposition channels are available. These are loss of a hydrogen atom (H') unexpectedly from the aromatic ring ( equation 2.12 ), loss of molecular hydrogen ( Hz ) presumably by the intermediacy of a hydride ion - ketone complex ( equation 2.'l,,3 ), and deprotonation of the phenyl ring ( ortho hydrogen ) by the incipient ethyl anion to effect elimination of ethane ( equation 2.1.4 ). The remaining products are accounted for by decomposition of a phenyl anion - propionaldehyde complex ( 66 ) which yields phenyl anion by dissociation ( equation 2.1,5 ) or the propionaldehyde enolate ion on proton transfer ( equation 2.L6 ). Chapter 2 75

l-tÍ t-

106 57

77

76 105

Figure 2.L The Collisional Activation Mass Spectrum of Deprotonated PhCH2OEt recorded on a VG ZAB 2F{F Mass Spectrometer. t3s

5'7

77

76 105

Figure 2.2 The Collisional Activation Mass Spectrum of Deprotonated

Ph(EÐCHOH recorded on a VG ZAB 2H'F Mass Spectrometer. Chapter 2 76

o- a a

---O+ Ctrat] + 'cH2cHî (2.11)

? þ \ o-

65

(Ph-H)CHO + H (2.t2)

o o

H + Hz (2.13) o-

J cHo 65 + czHa Q.14)

+ CeHs (2.1s) .r'H CoHs H ) o + ceHo Q.16) 66 H

Scheme 2.2 Chapter 2 77

The mechanism by which this rearrangement proceeds is not clear. The appearance of the benzaldehyde radical anion I electron affinity (E.4.) = 70 kJ mol-l (0.72 eV)] does not constitute proof that the rearrangement proceeds via the intermediacy of this anion. Independent work to show that a benzaldehyde radical anion will react with an ethyl radical has yet to be reported. It is possible that this is a side reaction which does not lead to the rearrangement product.

The fragmentation pathways were established using deuterium labelling, in conjunction with the comparison of collisional activation (CA) and charge reversal (CR) mass spectra of the product ions and of independently synthesised alkoxide ions. Perhaps the most compelling evidence for the occurrence of the rearrangement has been the identification of the C3H5O- ion found in the spectra of PhCH-OEI and Ph(EI)CHO-. The loss of benzene for example could yield the two ions

indicated in Scheme 2.3.

Ph H o + CoHs Q.l7)

o' -> Ph tÅrl + CH3CH=CH-O' + CoHo (2.18)

Scheme 2.3 Chapter 2 78

The CA MS/MS and CR MS/MS data (TabIe 2.2) of the product ion are consistent with the formation of the propionaldehyde ion (equation

2.1,9).

Table 2.2 The Collisional Activation and Charge Reversal

Mass Spectra of the C3H5O- Ions.

Initial Ion CAMS/MS (m/z) CR MS/MS (n¿)

56 55 41 27 5655il5343424740

PhCH'OEt 100 70 72 3 671 722782

PhcH(Et)O- 100 68 72 4 5 722 2 2I9 2

cFbcHcHo- 100 72 L0 2 6 7L2 2 2 5 7 2

CRÀ,fSllv[S (rnlz)

39 38 37 29 28 27 26 25 15 74 73 72

PhCH-OEt 15 551004559551067342

PhcH(Et)O- r5 66100436256116t442

cFbcHcHo- 12 55100496t531051137

The formation of this ion is rationalised as proceeding directly from the Wittig rearrangement alkoxide by u heterolytic cleavage to yield a transient ion molecule complex (equation 2.L8). Deprotonation of propionaldehydele2 [ (CHeCII2CHO)= 1553 kJ mol-l ] by the ^Hoacid phenyl u.tien193 [ ÂHoacid (C6H6)= 1,667 kI mol-l ] within the ion molecule complex would undoubtedly occur readily. There is no obvious rational for the formation of this ion directly from the ether.

O^ly one ion is formed exclusively from the deprotonated ether. This involves the loss of " Et' " (equation 2.1L) which should lead to the Chapter 2 79 formation of the benzaldehyde radical anion194 [electron affinity = 70 kJ mol-1 (0.72 eV)]. This is an unusual process as the loss of "R'" might be reasonably expected to be consonant with the order of bond dissociation energies. For example, the moiety "! B.r'" should be lost more readily than other alkyl radicals. However this does not appear to be so. It is therefore conceivable that this loss is not a simple loss of a radical. It has been demonstrated in positive ion chemistry that fragmentations leading to formation of the ethyl radical may be complex. This issue is discussed later in section 2.3.2.

The reactions of deprotonated benzyl alkyl ethers in the gas phase are different from reactions in solution. For example, in solution, elimination from deprotonated benzyl ethyl ether (equation 2.1,9) competes with the Wittig rearrange¡nsn¡179. No elimination is detected from the deprotonated ethers in the gas phase.

( o- + }J2C=CH2 Q.L9\ -->

2.7.2.2 Diallvl Ether.

Deprotonated diallyl ether195 (Figure 2.3) was also studied, and a Wittig rearrangement occurs here also. In this instance oxy-Cope rearrangement of the alkoxide ion ( 67 ) was thought to occu¡ subsequently to form the S-hexenal anion [( 68 ), Figure 2.4), ( see Scheme 2.4). Chapter 2 80

H.C=Cil-CH: 4l n

95 43 (-c4l-{ó)

(-cHi) 82

(-H20) 19 69 55 93 57 65 53 6I

(xt) l¡2) (xl) Figure 2.3 The Collisional Activation Mass Spectrum of Deprotonated

Diallyl Ether recorded on a VG ZAB zI{F Mass Spectrometer.

4l(-ctH¡o) vt

4l (-C¡Hc) 95

(-cHi) 82

cH2o) 't9

55 69 65 93 51 53 67

(xl) (x 2)

Figure 2.4 The Collisional Activation Mass Spectrum of Deprotonated 5 -

Hexenal recorded on a VG ZAB z}{F Mass Spectrometer. Chapter 2 81

o- o il CH

--+

67 68

Scheme 2.4

Nakai196 has reported that in solution, deprotonated diallyt ether undergoes a [2,3] sigmatropic rearrangement exclusively. Flowever in the gas phase the results support both [1,2] and [2,3] rearrangements. These results were inferrred by a detailed examination of the decompositions of the deprotonated labelled diallyl ether ( 69 ) shown in Scheme 2.5.

If the [2,3] rearrangement had occurred exclusively the ions of m/z = 43 and m/z = 45 would have been observed. This was not the case; ions of m/z = 43 and m/z = 44 formed via the [L,2] rearrangement were also observed. The ratio of the diagnostic ions m/z = 44 to 45 was 10 : 7, indicating that both processes had occurred. A deuterium isotope effect operates against the formation of m/z = 45. It is quite possible that the [1,2] and [2,3] rearrangements are equally facile; since a 2H/ 1H ratio of L.45 would account for the differing relative abundances of the ions. Chapter 2 82

o\CD CD 2 ^ ) 69 TI,2] 12þ) o'

-CD 2 )

/\ /\ o o o o il I il -CD CD 2 CD 2 DCH t ) ) CD CD 2 I CHDCDO ( núz 45 CH2CH=CDr( m/z =43) = ) + CH2{HCDO + 2

CH2CDO ( r¡t/z=4) CD2CH=CH2 ( m/z = 45 ) + CH2=QflCH=CDz + CH2=CHCDO

Scheme 2.5

Theoretical calculations for a model slrster¡l97 indicate that the [2,3] rearrangement in solution is influenced by a lithium counter ion ( Figure

2.5 ). In the absence of the counter ion the [2,3] sigmatropic rearrangement proceeds without any barriers. It is tempting to postulate that the absence of the l\,2) rcarrangement pathway may arise because the counter ions have a determinative effect on the course of the reaction. Chapter 2 83

I ô + 63kJmol o- L

l\ ou.ot'' o-.

¡\ 0 kJ mol'l o- Li

AE

t l0 U mol-l

-21ó kJ mol -224 kJ mol'l fr Lio)

Reaction Coordinate Figure 2.5 Comparative Reaction Coordinates for allytic migration in the presence and absence of a Lithium Cation.

2.L.2.3 Absence of The Wittig Reanangement From Allyl Alkyl Ethers.

In principle, since there is evidence for the Wittig rearrangement proceeding in solution for allyl n-butyl ether198 (equation2.20), it may be exPected that this reaction would occur in the gas phase. This is not the s¿sg199, as loss of a hydrogen atom (equation 2.2I) and an elimination reaction (equation 2.22) are observed. This result is quite unexpected and has yet to be satisfactorily explained. Chapter 2 84

OH PrLi (2.20) 4r.o-"oo pentane, THF

4.' + H' (2.21) a

- 4=. + + CHz=C¡1-ç¡1, (2.22) ç

In this chapter a series of deprotonated ethers will be investigated. These include the following : dibenzyl-, diphenylmethyl phenyl-, phenyl allyl-, benzyl allyl-, benzyl vinyl-, divinyl-, and allyl vinyl ethers. They have been chosen so that a systematic investigation into the role of

electronic and steric factors affecting the course of the reactions in the gas phase may be made. Where ever possible, the condensed phase rearrangement (or their analogues) is noted. In this way it is hoped that the behaviour of the species in solution and the gas phase can be easily compared.

The general conclusions concerning the gas phase and solution

phase reactions will be recorded at the end of this chapter. Finally, the loss of "R'" from the deprotonated ethers is described. The results obtained in

this work are analysed within the framework of "R'" losses observed from previous studies. A rationale for this specific fragmentation is given. Chapter 2 85

2.2 Results and Discussion.

2.2.1 Dibenzyl Ether.

The rearrangement of dibenzyl ether proceeds readily in solutiorLl6T' but unexpectedly the deprotonated anion was too unstable to be observed in the gas phase. Instead only PhCH2- was observed. This may have been formed from the initially deprotonated ether by ^ simple heterolytic cleavage (equation 2.23) or possibly by decomposition of the rapidly formed Wittig rearrangement product (equation 2.24).

/1 Ph PhCHO + PhCH2' (2.23)

-+

-i+nnôJJn PhCHO + PhCH2 (2.24)

By contrast with the deprotonated ether, the alkoxide Ph(PhCH2)CHO- was sufficiently stable to be readily investigated. The relative abundances for the fragmentation of this ion and its labelled analogues are listed in Table 2.3 and summarised in Scheme 2.6. Chapter 2 86

Table 2.3 Collisional Activation Mass Spectra of Ph(PhCH2)CHO- and Isotopicalþ Labelled Analo gues.

Initial Ion Loss Ha HD QF{e GHsD CzF{e C7H5D2 CzHDz PhCHO Ph(PhCH2)CHO- 46 9 76 100 Ph(PhCHz)13CHO- M 8 15 Ph(PhCD2)CHO- 2T 3 12 100 Ph(C5D5CD2)CHO- 20 4 13 100

Loss Phl3CHO PhCH2CHO PhCU'1S6¡'O PhCD2CHO QD5CQCHO Ph(PhCH2)CHO- 0.5 Ph(PhCH2)13CHO- 100 0.6 Ph(PhCD2)CHO- 0.1 Ph(C5D5CD2)CHO- 0.1

+ Hz (2.25) f(,,;;":1 "]*,,:)":o 70

Ph- + PhCH2CHO (2.26) / o- Ph PhcH2cHo 71 \ PhCH-CHO + CoHo Q.2n

PhCH2' + PhCHO (2.28) 'H2cPh] / [(rr,cro) CHO 72 + PhcH3 e.29)

Scheme 2.6 Chapter 2 87

simple heterolytic cleavage of deprotonated 1,2 diphenyl ethanol would yjeld three ion-molecule complexes. The hydride ion-

deoxybenzion complex (70) would simply collapse by loss of H2 to yield the benzoin enolate ion (equation 2.25). The phenyl anion phenylacetaldehyde complex (71) may dissociate (equation 2.26) or undergo proton transfer (equation 2.27). In a similar way abenzyl anion - benzaldehyde complex (72) may either dissociate (equation 2.28) or the benzyl anion lÂHoacid (CeHsCHj)2OO = T594 + L2 kI mot-l ) may deprotonate benzaldehyde (equation 2.29). lÂHoacid (CeHsCHO)201 ='J,6I2 t24kJ mol-1 l.

The loss of PhCH2' does not occu¡ from PhC-(H)OCH2Ph and yet in terms of the relative radical stability this fragmentation should be extremely favourable. This observation tends to support the earlier suggestion (see int¡oduction section 2.I.2.'J., p.79) that the loss of an alkyl radical is not a simple homolytic cleavage [c/ equation 2.1L].

222 Allvl Phenvl Ether.

Simple deprotonated alkyl allyl ethers fragment predominantly by an ct,P - elimination (equation 2.30). This decomposition channel seems unlikely for allyl phenyl ether since deprotonation of the phenyl ring would be required (equation 2.31).

Z\r.. 4/O + H2C=CH-CH3 (2.30)

-> Chapter 2 88

_) + 4.,o (2.31) H

Although it is conceivable that the Wittig rearrangement may occur in this system an anionic Claisen rearrangemsn¡203 is also feasible (equation 2.32).

o- (2.32)

Comparison of the spectra of PhOC-(H)CH=CH2IQ3), Figure 2.61and expected Wittig rearrangement alkoxide ion Ph(CHz=CH)CHO- lQ4), Figure 2.7] shows that the Wittig rearrangement accounts for most of the decompositions observed. Peak width measurements are recorded in

Table 2.4 of ions from the Wittig product and the ether and are found to be identical (within the limits of experimental error). The peak heights indicate similar proportions of products are formed in each instance. Furthermore, the charge reversal spectra (Table 2.5) are very similar and this indicates that most of the deprotonated allyl phenyl ether rearranging in the source has formed (74). Chapter 2 89

Bl iìl (-Hz)

I 15 (-H20) CcHi 77

o HrC=C.Ci H 55

105

r03 106

CH2=CK n Pho- 93 (x 100) Jt

(x l0)

Figure 2.6 The Collisional Activation Mass Spectrum of PhOC-(H)CH=CH2.

ceHj

ltl

CH-o I ICH Hzc

H2C.C' .i: 55 (-H20) ll5 l

(-H2C=O) 76 l03 r05 (

n 4l

(x l0)

Figure 2.7 The Collisional Activation Mass Spectrum of Ph(CH2=ÇH)CHO- (-) \ I N)

Table 2.4 The Collisional Activation (CA) Mass Spectra of PhOC-(H)CH=CH2, Ph(CH=CH2)CHO-,2-(CH2=ç¡1-CH2)C5HaO- and related deuterium labelled derivatives.

Initial Ion Loss H2 Hp HZO HOD C2H3' CO CHZO CHDO CSFI4 C3H4O C3H3DO C6,H5 QH5D CoHZDA QHDS PhCHO PhCDO cóDscHo PhOC-(H)CH=CH2a Fig.2.3 c5D5oC-(H)CH=CHz 6 100 7 I 2 2 3 279 3 2 6 0.1 Ph(cHz=CH)cHo-b 1001044 27 10 01 QHs(CHz=CH)CHo- 83100 9 9 4 5 3 s2 3 1 13 0.2 Ph(cH2=cH)cDo- 10093284704 5 38 79 0.3 Ph(CH2=ç¡r)cHo- 100671885 5 % 18 q0s a. Peak widths at half height lm/z (+ 0.2 Volts ) loss I : 131(56.5)Hz 115(51.1)H2O, 105 and 106 (not resolved), 103(44.8)CH2O, 77(39.1)CsHaO, 55(35.ÐC6H6,. When a potential of +1,000 Volts is applied to the collision cell the following collision - induced : unimolecular ratios are obtained : m/z (c:u) 131(65:45), 115(10:90), 106(100:0), 105(80:20), 103(5050), 93(955), T7(70:30),.$(a0:60). b. Peak Widths at half height Im/z (+ 0.2 Volts ) loss ] : 131(56.5)H2 115(51.5)HzO, 105(53.9)CzH+ 103(45.2)CH20,77(39.4)CgH¿O, 55(3a.5)C5H5. When a potential of +1,000 Volts is applied to the collision cell the following collision - induced : unimolecular ratios are obtained : m/z (c:u)

737(7 0:30), 1 15(10:90), 105(70:30), 103(50:50), 77 (80:20), 55(30:70).

\o O Chapter 2 9I

Table 2.5 Charge Reversal ( CR ) Mass Spectra of PhOC-(H)CH=CH2 and Ph(CH2=çH)CHO-.

Initial Ion m/z

732 131 115 105 103 9t 89 77 75

PhOCH--CH=CHz 27219522714 18 100 20

Ph(CHz=CH)CHO- 21118542513 18 100 21,

Initial Ion m/z

65 63 55 51 50 39 37 29 27 PhOCH--CH=CHz 72317666552615621 10281866532415720

Most of these fragmentations (Scheme 2.7) are analogous to decompositions of alkyl benzyl ethers (see ethyl benzyl ether described earlier in Scheme 2.2). Chapter 2 92

'/ o H2Ç:Ç-Ç tPh+ H2 (2.33)

o H-

+ H2 Q.34)

75

o' CHO

"u=",1 + HC=CHz (2.35)

74 76 o + CoHs (2.36) H

H o- HrC:C:C. + CoHo Q.37) H

Scheme 2.7

The loss of H2 is slightly different from the process described for deprotonated ethyl benzyl ether because the hydride ion neutral complex (75) fragments either by ring deprotonation (equation 2.33) or by removal of a vinylic hydrogen (equation 2.34). Deuterium labelling indicates that deprotonation of the phenyl ring is favoured but as a deuterium isotope effect is likely to be observed for both losses (equation 2.33 and 2.34) a quantitative determination of absolute yields from each path is not possible. Chapter 2 93

The formation of the ethene anion (equation 2.35) is observed in this systern. Presumeably the ethene anion being a weaker base than an alkyl anion does not deprotonate benzaldehyde readily in the ion complex

(76). The loss of acrolein (equation 2.36) and benzene (equation 2.37) are consistent with the formation of a phenyl anion - acrolein complex.

The losses of CO and CH2O were not observed for benzyl alkyl ethers but have been observed for simple aryl benzyl ethers and arise from the Wittig rearrangement product. By analogy with the decomposition path described for the diphenylmethoxide ion (formed on rearrangement of deprotonated benzyl phenyl ether), a plausible rationale is given in Scheme 2.8. The vinyl anion - benzaldehyde complex (76) may form by simple heterolytic cleavage. The vinyl anion does not appear to act as a base : instead it may act as a nucleophile. Ipso attack on the aromatic ring would yield a transient "Meisenheimer" type o adducÛ03 Q7). Decomposition of the o adduct may then proceed via the styrene- deprotonated formaldehyde ion complex by hydride ion transfer (equation

2.39) or proton transfer (equation 2.40). It seems likely that these reactions reflect the greater stability and weaker acidity of the vinyl ¿nisn204 when compared to the alkyl anions. Chapter 2 94

o il L('^, 76

CHO tn + C=o (2.ss) HC:o p( t\. It Ph H2 + CHr=g (2.40) 77

Scheme 2.8

The loss of water was unexpected2OS as it had not been observed from the Wittig alkoxide ions. This intense ion is observed from both the deprotonated ether and the Wittig ion. A plausible rationale for decomposition (Scheme 2.9), consistent with the deuterium labelling

(Table 2.4) involves an initial specific 1,3 hydrogen transfer206'207, followed by the formation of an ion complex (78). The hydroxide ion deprotonates the phenyl ring most readily (equation 2.aÐ ; however an allenic proton may also be removed (equation2.42) but this process is minor.

P\ P\ cH.o CH.OH / H2C:CH -> H2C:C - H C:C:CHz

+ HzO (2.4r) *] l(",':':"í;') C:CHz 7E + HzO (2.42)

Scheme 2.9 Chapter 2 95

In the spectrum of (73) there are two peaks not observed in the spectrurn-of the Wittig product ion (74). The first, corresponding to the formation of PhO-, can be readily formed by an cr- cleavage of the deprotonated ether. The other peak is formed by the loss of CzHg.. Flowever, Claisen rearrangement (Scheme 2.10) should yield deprotonated q-allyl phenol, and this species would be expected to lose C2H3' to form the q-quinone methide (79). The spectrum of deprotonated q-allyl phenol is shown in Figure 2.8 ; as expected it fragments by loss of CzH3'. Thus it is likely that m/z = 106 in Figure 2.3 is formed by a Claisen process.

H

o' + HC=CHz

79 Scheme 2.10 Chapter 2 96 (-CzH¡') lll 106

Pho- 93

Figure 2.8 The Collisional Activation Mass Spectrum of Deprotonated q- Allyl Phenol.

The Claisen process may proceed by a specific proton transfer from a phenyl ortho hydrogen to the allylic position, followed by an SN2' displacement reaction. Specific proton transfers of this type have already been observed during the fragmentation of deprotonated uttiro1"208

(equation 2.43).

@Hr ocH a .+--.+ (2.43)

An alternative rationale is indicated in Scheme 2.1.1. In solution allyl phenol ¡earranges via a cyclic intermediate analogous to ion (80). In the analogous anionic process, the cyclic intermediate may collapse via the hydride ion complex (81) losing of CH2=ÇH. to form g-quinone methide radical anion (79, equation 2.45). This seems a less likely proposition since Chapter 2 97 the hydride ion complex (81,) may also undergo competitive elimination of H2 (eqqrtion 2.44).

+ H2 e.M)

'l o' H < 80 8L

o 'HC=CHz + e.45)

79 Scheme 2.LL

The Wittig rearrangement product ion rearrangement has a pronounced unimolecular component ( footnote to Table 2.4), whereas the products arising via the Claisen rearrangement are produced only by collisional activation ( footnote to Table 2.4) indicating that this process is disfavoured energetically.

22.3 Allvl Benzvl Ether.

In solution, the rearrangement of allyl benzyl ether seems at variance with the usually accepted ketyl radical anion - radical mechanism because migration of an allyl or benzyl moiety does not occur readily. Instead, using KNH2 / liq, NH3209 or KOtBu / dimethyl sulphoxide Chapter 2 98

(DVÍSO¡t88a 1g¿¿t to isomerisation of the double bond to yietd benzyl propenyþther (equation 2.46). The vinyl group migrates under more

vigorous conditions (equation 2.47).

KNFI2 / liq.NH3 or KOIBU / DMSO PhCH2-O-CH2CH=CH2 PhcH2-o-cH=cH-cH s Q.46)

KNFI2 / liq.NFI3 ^ OH I pnzcr\^ (2.47)

In the gas phase, deprotonation of allyl benzyl ether is expected to yield two delocalised . Intermolecular protonation, which results in isomerisation of these anions in solution will not occur in the gas phase because of the absence of collisions with other species. Consequently, it is possible that the Wittig rearrangement may be observed in the gas phase. Atlylic migration to the benzylic carbanion should yield Wittig product ion (82, equation 2.48) whereas benzylic migration to the allylic carbanion should yield Wittig product ion (83, equation 2.49), shown in Scheme 2.12. Chapter 2 99

o

Allyl Benzyl migration migration

o- (2.48) o - Q-4e)

82 83

Scheme 2.12

The spectrum of deprotonated allyl benzyL ether is shown in Figure

2.9. It is most complex, and is dearly not just a combination of the spectra of the two Wittig product ions Ph(CHz=CHCH2)CHO- (Figure 2.10 ) and PhCH2(CHz=CH)CHO- ( Figure 2.I1) Chapter 2 100

129 (-H2O)

69

CLl.=CH-CHz 4l PhcH2 Ph 91 1'7 105

Figure 2.9 The Collisional Activation Mass Spectrum of Deprotonated Allyl Benzyl Ether.

The spectrum of allyl benzyl ether is dominated by the loss of water.

This loss is quite minor in the spectra (Figures 2.10 and 2.11) of the Wittig product ions, indicating that the loss of water probably precedes lhis rearrangement. A new rearrangement, of a type not observed in the spectrum of allyl phenyl ether ( see Section 2.2.2 ) is in operation. The other fragmentations are minor but appear to be consistent with the decompositions of the two possible Wittig product ions.

This is substantiated by peak width measurements, eg. m/z = 105 [ Figure 2.9,T1¡2= 42.6 V; Figure2.t0,Tt /z= 42.7 V ),m/z = 91 [Figwe2.9,

T t / z = 26.6 V ; Figure 2.17, T 1 ¡ 2 = 26.4 Y l, m/z = 77 lFigure 2.9, T 1 ¡ 2 = 39.I

Y ; Figure 2.I0,T1¡2=39.0 Vl and m/z = a1 [ Figure2.9,Tt/2=21,.7Y ; Figure 2.70,T1/Z = 2!.8Y l. However the relative abundances of these peaks do not correlate as well as those of previous examples. Chøpter 2 101

CH'=CH-CHz l-17 4l

(PhcHo - H)- 105 (-HzO) 69 j'l 9l rl'7 t29 146 (x l0) 106 145

Figure 2.10 The Collisional Activation Mass Spectrum of the Expected Wittig rearrangement ion Ph(CHz=CHCHz)CHO- (82).

PhcH2 t47 9l

129 (-H:O)

ll9 55

43 tt7 4l 69 77

(x 25) (x 25) t46 I15

Figure 2.L1 The Collisional Activation Mass Spectrum of the Expected Wittig rearrangement ion PhCH2(CH2=CH)CHO- (83). Chapter 2 102

Treatment of PhCD2OCH2CH:CHz with DO- resulted in the formati,o_4_of two ions, a benzylic carbanion (85) (Scheme 2.13) arising from dedeuteration ie. (M-O*¡- and an allylic carbanion (84) arising from deprotonation (M-H*¡-. The ratio of (85) to (8a) is 5 : 1.x This result is as expected in view of the gas-phase acidities of unsubstituted propene20o (ÂHoacid = 16221 11 kI mol-1) and toluene2OO (ÂHo¿si¿ = 7594 + L0 kJ mol- 1). Deprotonation at the benzylic position is clearly favoured. The alkoxyl subsitution is likely to decrease the acidity of each position but only but a small amount. This evidence provides a reasonable explanation for the comparatively minor abundance of ions resulting from deprotonation at the allylic site. The collisional activation mass spectra of the ions (84) and

(85) are reproduced in Figures2.72 and 2.13 respectively.

The spectra of the two initially formed carbanions provides evidence about differences in the reactivity of the ions. The Wittig rearrangement of PhCDzOC-(H)CH=CHz yields many ions which are formed following proton and deuterium transfer. The complexities of the hydrogen and deuterium transfers are summarised in the Scheme 2.1.3. The allylic carbanion (8a) undergoes deuterium transfer from the benzylic position to yield ion (85). Alternatively, proton transfer from the phenyl ring yields ion (87), which may undergo benzylic deuterium transfer to yield (88). Fragmentation of (86) yields PhCD2- ions (m/z = 93) and PhCD-

CHO. There is no evidence for the formation of the PhCHD- ion (m/z = 92) indicating that the hydrogen/deuterium transfer is not reversible I le. (84) + (85) l. It is concluded that the allyl carbanion (8a) is unstable with respect to the benzyl carbanions (85) in the gas phase. Both (8a) and (88)

x It should be noted that an isotope effect disfavouring dedeuteration undoubtedly occurs , and the likely proportion of benzylic carbanion formed in the unlabelled allyl benzyl ether will be higher. Chapter 2 103 may undergo the Wittig rearrangement : cf. equations 2.50 and 2.51, and 2.52 - 2.54_respectively.

By contrast, if we consider the fragment ions arising from the Wittig rearrangement of PhCD-OCH2CH=CHz only one ion (PhCHD- ,m/z = 92) has arisen following a proton transfer. The relative intensity is very small, indicative that proton transfer is a very minor process for PhC- (D)OCH2CH=CHz.

PhcD2 + 4lo (2.s0) / H-O' \ PhCD-CHO + DHC=CHz (2.51) 86

* PhCD2-O-CHCH=CH 2 (C 6H)-CD 2-O-CH2CH=CHz 84 87

(C 6H4D)-CD' O- CH 2CIJ=CIJ2

(2.52)

(2.s3)

(2.54)

PhCD2-O-CHCH=CH2 *:,- PhCD-O-CHDCH=CHz (2.5s) 84 85

Scheme 2.13 Chapter 2 104

(-Hzo) l.¡e 130

(-HDO) l3l

CoHs PhCD; '7'l 93 'n 4t c6H4D 70 78

147 J|L t46

Figure 2.12 The Collisional Activation Mass Spectrum of PhCD2OC-(H)CH=CHz (8a) recorded on a VG ZAB zÍIF Mass Spectrometer.

70 (-CoHo) 148

(-HDO) 129

c6H4'cDo (-HzO) Ph 106 '17 130 CHr=ç¡1-ç¡1, 4l

PhCHD. 92

t46 (-Hz)

Figure 2.13 The Collisional Activation Mass Spectrum of PhC-(D)OCH2CH=CH2 (85) recorded on a VG ZAB 2HF Mass Spectrometer. Chapter 2 105

Migration of the allylic group during the Wittig rearrangement of

(85) may-proceed by a 17,21process to yield alkoxide ion (90a) or a [2,3]

Process to yield (90b). The fragmentations of the deprotonated benzylic ether (85) are summarised in Scheme 2.14. The alkoxide ions (90a) and (90b) decompose either via the intermediacy of an allyl ion - benzaldehyde complex (equation 2.56), or via a phenyl ion - but-4-enal complex (equations 2.57, 2.58 and 2.59). Unfortunately it is not possible to determine which process(es) is/are occurring.

PhCD-O-CHDCH=CHz 85

þ o' PhCD-CDH-CH=CHz 90a --+ o' and / or I PhcD-cH2-crFCHD CH2CH=CHD + PhCDO (2.s6) 90b cH2{H-CHD-CDO Ph' + and/or (2.57) DHC{H-CH 2-CDO and / or l"( DHC=CH-CH 2-CDO ) c6HsD + CHr=ç¡1-6H-CDO (2.s8)

DHC{H-CH-CDO CoHo + and/or (2.se) cH2{H-CD-CDO

Scheme 2.14 Chapter 2 106

The major fragmentation from both the allylic anion (84) and the benzyl anion (85) is loss of water. The allylic anion (84) readily undergoes proton transfer to yield benzylic anion (85). Ion (84) is the major species and it seems likely that loss of water proceeds via this ion. There are two possible explanations. Cyclisation of either the benzylic or allylic ether anion had occurred, perhaps in a manner analogous to the loss of water from allyl phenyl ether, or a new rearrangement process had occurred.

Since the loss of H2O is not originating via a Wittig rearrangement, a number of other possible rearrangement ions were investigated. The spectra ( Table 2.6 ) of all of the ions formed by deprotonation are different from those of allyl benzyl ether. This result indicates that none of the deprotonated ions are formed from allyl benzyl ether. In the sou¡ce it is quite likely that more than one deprotonated ion may be present, whereas the rearrangement may lead to one ion specifically. As the identity of the ion which loses water could not be obtained this way another approach was required. This was possible using the VG ZAB 2}{F spectrometer because the loss of water had arisen predominantly by a unimolecular decomposition. Consequently the structure of the product ion (C1gH9-) is likely to be essentially the same in the source and collision cell. Therefore the collisional activation and charge reversal mass spectra of source formed CtoHg- ions were determined and they were compared with the corresponding spectra of C1gH9- isomers of known structure (Table 2.7). The CA MS/MS proved to be of very limited diagnostic value, since the loss of H'and H2 were the only fragmentations. Forfunately, the charge reversal mass spectra (Table 2.7) did allow a determination of the structure to be made; the data from the table being wholly consistent with that of a deprotonated dihydronaphthalene. Chapter 2 r07

Table 2.6 Collisional Activation Mass Spectra of Benzyl Allyl Ether and isomeric C16H110- ions.

Loss

Initial lon H. H2 Cru HzO CzFL CH2O C3Ffu C¡F|¡O C¿FkO eFk CzHe CsFrs phCHO

(PhCHzOCHzCH=CH2 4 - H+)- 3 7 100 1 2 2 19 4

å. Ph(CHz=CsCHz)CHo- 11 18 15 24 1 4 100

d' PhCH2(CH=CH)CHO-c 18 4 2 1 100 02

PhCH2CH2CH-CHO 100

H 16 52 14 100 831 12

cH3

11896359 100 25

o-

9 35 12 76 36 100 2L

o

13 100 T9

ß. Peak width at half height: lm/z (10.2 Volts)lossl : 145(37.9)H2, 729(23.7)H2O, 105(42.6)CsiH6, 9t(26.6)C3HaO, 77(39 'L)CaH6O, 69(33.6)C6H 6, 7Q7.7)PhCHO. When a potential of +1,000 Volts is applied to the collision cell the following collision - induced to trnimolecula¡ ratios are observed Im/z (c:u) ]: 1a5(20:80), 129 (10:90), 105(90:10), 91 (50: 50), T (7 0:30), 69 (30:7 0), 41 (1 0:90). b. Peak width at half height: lm/z (fr.2 Volts)lossl : 145(37.7)Hz, 729(2t.5)H2O, I0S(42.7)C3H6,77(39.0)ClHoO, ó9(33.8)CøHe, 41(21.8)PhCHO. When a potential of +1,000 Volts is applied to tJre collision cell the following collision - induced to unimolecular ratios are observed I m/z (c:u) l: 145(5:95), 129 (30:70), 105(85:15), 77(80:20), 69(65:35), a1(10:90). c. Peak width at half heighh Ím/z (!{.2 Volts)lossl : 145(37.8)H2, I29(20.1)II2O, 119(45.5)CzH+, 9l(26.4)C3HaO, 55(20.3)PhMe. When a potential of +1,000 Volts is applied to the collision cell the following collision - induced to unimolecu-lar ratios are observed I m/z (c:u) l: 145(10:90), L29 (3070), 119(90:10), 91(45:55), 55(10:90). d. The collisional activation mass spectrum of PhCH2(CHz=CH)CDO- is as follows I loss (relative abundance %) ] : H'(6),D'/rlrz(12), HD(2), Hzo(1.s), HoD(0.s), czFI¿(1), caIl3Do(100), CzFIs(0.3). Chapter 2 108

Table 2.7 Charge Reversal ( CR ) Mass Spectra of the C16H9- ion from deprotonated AJlyl Benzyl Ether and isomeric

CroFIs- ions I relative abundance (%) ].

Neutral m/z Precursor

r29 r28 727 126 115 1r4 113 103 99 89 87 n 74 63 51 50 43 41 39 29 27

a PhCH2OCH2CFI=CH2 54 100 33 2 17 8 35 4 20 11 37 25 y 40 36 3 6 20 2 6

L 48 100 35 3 L9 8 35 52r t42629 40 32 33 7 6 18 2 4

r 46 100 37 4 20 9 27 5 18 13 2t 25 38 25 27 7 4 16 1 s

b

36 100 39 25 10 15 9 14 2 7 10 t2 18 27 74 16 16 7

o 42 100 51 18 t4 13 I 16 2lt 13 18 243432 33 16 3 CH:C:c]r,

ø PhCH2CH=CdH2 20 100 46 L6 I 8 6 t6 2 17 7 27 18 26 25 24 11 29 24 14 72

¿. The collisional activation mass spectrum shows only loss of H' and H2.

å. The collisional activation mass spectrum is as follows I m/z (telaltve abundance %), loss I : 128(100) lH.', 727(I0)Hz,

1la(0.05)Me', 1 13(0.1)CFI1. Chapter 2 109

This result may appear at variance with earlier results which where the loss of water from 1-hydroxytetralin (Table 2.6) was shown to be a minor Process. The species formed in the source as almost certainly the alkoxide ion as the loss of H2 common to other alkoxides was very pronounced. The immediate precursor to the deprotonated dihydronaphthalene would be a benzylic deprotonated 1-hydroxytetralin. 1-Methoxytetralin is therefore a better analogy: the CA MS shows that fragmentation occurs almost exclusiaely by a pronounced loss of methanol with only very minor losses of H' and H2.

A plausible mechanism for this process, which does not involve any of the deprotonated ions described in the Table 2.5 but accounts for the losses of Hz O and HDO arising from deprotonation of PhCD2OCH2CH=CH2 is given in Scheme 2.15. To account for the loss of H2O and HDO it is postulated that hydrogen and deuterium transfers from the g-phenyl anion and the benzylic ion precede the rearrangement. Chapter 2 110

CD o o

\ / \

D H

o o H H H D

OH OH

D H

H D

I HO D HO H

H D

I D H

HO' l- H D

D H

+ HzO HOD +

Scheme 2.15 Chapter 2 111

of Deprotonated Allyl Benzyl Ether.

The results reported in the gas phase (above) show little analogy to the condensed phase reactions described earlier. In particular, no hydrocarbon products were reported. There are examples of hydrocarbons being formed in the course of the Wittig rearrangement and this prompted us to reinvestigate this reaction using more forcing conditions.

Allyl benzyl ether was treated with lithium di-isopropylamide in a tetrahydrofuran - hexamethylphosphoramide mixture. The th¡ee major products isolated were purified by gas chromatography. The products, identified by ir. ,1H nmr and comparison of g.l.c. retention times using authentically prepared compounds, were 1-phenytbut-3-en-1-oI (91), 1,2- dihydronaphthalene (92) and 1-s-tolylprop-2-en-L-ol (93) in the approximate ratio 4:3 :1 indicated in Scheme 2.16.

PhCH2OCH2-CH=CH 2

Ph\ 3 /cH.oH H,C:CH-CH, / CH=CHz HO

9l 92 93

Scheme 2.16 Chapter 2 772

The resr¡lts of this reaction are unexpected. Onty one of the possible Wittig products, ie. (91), is observed in this condensed phase reaction. Flowever in the gas phase, there is evidence for the formation of both of the Wittig rearrangement products.

The alcohol (93) is a minor product in solution but no evidence to support its formation is to be found in the gas phase spectrum of the deprotonated alcohol. An alkyl group migration of this type has been observed 6"¡6¡s210.

The most important finding is that one of the major products in the condensed phase experiment was 7,2 - dihydronaphthalene$. This is the product of the corresponding gas phase reaction.

2.2.5 Diphenylmethyl Phenyl Ether.

The Wittig rearrangement of this ether has been observed in sels¡¡sn211. Phenyl group migration was observed from benzyr phenyl ether in the gas phase. However in this system phenyl migration results in formation of an ion which is already sterically congested. Thus, in this system the effect of steric factors may be exarnined.

The collisional activation mass spectra of PhzCH-OPh, the Wittig rearrangement alkoxide ion Ph3CO- and various isotopically labelled compounds are recorded in Table 2.8.

S If 1,+ - dihydronaphthalene had been formed it would isomerise to 1.,2 - dihydronaphalene under the reaction conditions. ô

a l'\)

Table ZE Collisional Activation (CA) Mass Spectra of Ph2C-OPh, PhgCO-, and Isotopically Lâbelled Analogues.

Initial Ion Lßs H. H) HD c6Hr C¡tlqD: C¿HsD C¿H2D¿ C¿DsH CzHsDO CzH¿O C5I3CH5O C7H5DO C7tfuQO CTH2DaO C7H2D5O CpHl¡ C¡2H5D5 C12ËlaD5 CBHI0 Ph2CO Ph213CO Ph(C5D5)CO (C5H2Q)2CO

PhO-CPh2¡ 100 72 55 5 2 372

PfuCo+ 100 24 57 5 2 13

PÌç BCO- 100 23 55 4 2 72

Ph2(C5D5)CO-c 7æ 48d 48d t5 20 74 20 1.3 2.7 73 3 1 1 3 I

n0 3q 39d 8 23 30 1 J 2 10

a Fo¡med by dçrotøration of PhOCHPhz. b. Fo¡¡ned by dçrotøration of Pt¡eCOH. c. The collisional activation ma6s spectÌa of the deuterium - labelled de¡ivatives showed composite peaks for the læses of CøHe, C7H5O, CrzHro, (C6HS)ZCO and labelled analogues. The ratios were determined using the linked - scan (E/B) technique for ions decomposing in the fi¡st collision cell. d. D' and H2 = 2 a.m.u.

P P OJ Chapter 2 774

The spectra of the deprotonated ether (9a) and of the possible Wittig rearrangement ion (95) are almost identical. Thus the fragmentation data are again consistent with rearrangement of preceding fragmentation. The decompositions of deprotonated diphenytmethyl phenyl ether (94) are indicated in Scheme 2.17.

The alkoxide ion (95) fragments by loss of hydrogen (H2) possibly to yield the 9 - phenyl fluorenyl alkoxide ion (96) (equation 2.60) or via a phenyl anion - benzophenone complex (97). The complex (92) may decompose by dissociation to yield the phenyl anion (equation 2.6L); the phenyl anion may deprotonate benzophenone (equation 2.62) or effects a nucleophilic aromatic substitution (S¡Ar) reaction (equation 2.63). This type of reaction has been reported in the gas phass203. Chapter 2 115

Ph

+ H2 (2.60)

Ph2c-o-Ph 96 94 Ph + Ph2C=O (2.67) þ Ph3c-o' en' (n o I rc=o)] il 95 97 h + CoHo Q.62)

o ll + Ph-Ph (2.63) Ph/C

o- Ph2C-O-Ph + Ph (2.64) -l(- Ph

Scheme 2.17

As with benzyl phenyl ether, loss of a phenyl radical is absent (equation 2.64). The loss of H' is a major fragmentation, but the available data does not permit a determination of whether the phenyl hydrogen is lost from a specific position or randomly from the ring, Aryl scrambling

has been noted previously for negative ior6191,206,272, however numerous

examples also exist indicating specific aryl hydrogen 1oss213-215. Chapter 2 176

2.2.6 Vinvl Alkvl Ethers-

In solution, there have been reports of wittig rearrangements involving migration to a vinylic position188, although none involve an alkyl group.

This rearrangement would lead to the formation of enolate ions and whilst thermodynamically such a process would be considered energetically favoured any condensed phase sfudy would prove to be very difficutt due to the tendency of these ions to undergo numerous secondary reactions such as aldol condensations (equation 2.65). Furthermore, in many instances deprotonation occurs at the p - position. These ions fragment by elimination producing the alkoxide ion (equation 2.66).

Aldol condensation + (2.6s) àot* + products

-ãotR + RO- + HC=CH (2.66)

In the gas phase it is possible to study the reaction of the ions without the accompanying side reactions. Furthermore, it is of particular interest to establish if the Wittig rearrangement occurs in this system as the behaviour of benzyl- and allyl- alkyl ethers (described in section 2.7.2.I and 2.I.2.3 respectively) have been shown to be quite divergent.

Deprotonated benzyl alkyl ethers have been shown to undergo facile rearrangement (equation 2.67 ) prior to fragmentation. By contrast there is no evidence of any rearrangement of deprotonated allyl alkyl ethers. Chapter 2 177

Fragmentation of these ions has been shown to proceed almost exclusively by elimination (equation 2.68 ). o' I phCH_O_R phCHR -o+ (2.67) CH¡ + H2C=CH-CH 3 e.68) H

The investigation of deprotonated vinyl alkyl ether is expected to be considerably more complex than those described previously. Unlike the studies of allyl- and benzyl atkyl ethers (where the ions studied were produced by deprotonation at only one site), it was unclear whether deprotonation of vinyl alkyl ethers would occur at the vinylic position cr to the oxygen atom (ø-carbanion, 98) and/or at the terminal vinylic methylene group ( p-carbanion, 99 ).

ào'* lotA

98 99 Furthermore, the possibility of a 7,2 hydrogen migration between the c¡- and p- vinylic positions must also be considered. L,2 Hydrogen migrations are generally considered symmetry forbidden. However theoretical calculations (MINDO) indicate that for simple vinylic systems2l6 the barrier ( 130 to conversion may not be particularly high = k] mol-l ).

In spite of the potential difficulties it was anticipated that the characteristic fragmentations of the ketone enolate product isr$217 '278 Chapter 2 118 resulting from Wittig rearrangement should allow the presence of these rearranged ions to be readily discerned.

It was envisaged that specific formation of the cr-carbanion would all<-rw (t) an unambiguous assignment of fragmentations of the deprotonated vinyl alkyl ether and (li) allow an accurate assessment of the proportion of the o-carbanion rearranging to the methyl ketone enolate ion prior to fragmentation. The spectra of the o-carbanions were obtained by Srrl 2(Si) desilylation reaction of the appropriate o - trimethylsilylvinylalkylethers, using NHz- as a nucleophile ( equation

2.6e ).

N,b 3 Si

+NHz- ào'* + Me 3SiNH z e.69) 98

The data obtained for the collisional activation mass spectra of the vinyl alkyl ethers ( deprotonated by NH2- ), the a-carbanions of vinyl alkyl ethers, and enolate ions (the expected products from the wittig rearrangement) are listed in Table 2.9.

When vinyl alkyl ethers were treated with NHz- in the source fragmentations characteristic of both deprotonated species were observed. When R = Me, decomposition of the a- deprotonated ion is expected yield Me- (equation 2.70) and HC=CO- (equation 2.7t) ions whereas fragmentation of the p deprotonated ion shor¡ld form Ro- (equation 2.72) and HC{- ions (equation 2.73). Table 2.9 Collisional Activation Mass Spectra of Deprotonated Vinyl Ethers and of the Wittig Rearrangement Products ô Loss Initial Ion \s H' H2,D HD CHg' CH¿ CDgH HzO CzHz CzH+ C2H2D2 C2F{6 C2HaD2 C2H3D3 CH2CO f\ì NJ (cH2=ç¡19Me)-H+ 100 924 6 2 CH2=Ç-9¡4" 100 96 8 I MeCOCH2- 100 76b 4 (CH2=ç1¡9Et¡-t¡+ 1.5 6 41, 100 25 CH2=ç-98¡ L05 100 16 (CH2=ç1¡OCD2Me)-H+ 15 20 2 14 37 100c 1.3 (CH2=ç¡1OCH2CD3)-H+ L58 6 4 7L 100c 66 (MeCOEI)-H+ 15 100 18 51 6 10 80d (CH2=6¡¡9Pr)-H+ 186 6 2 100 42e (MeCOPr)-H+ 48% u 1, 1008 (CH2=Q1¡giPr)-H+ 8 6 46 (MeCOiPr)-H+ 100 72 n 3L 6 24 (CH2=6¡198u)-H+ T6 9 2 100 CH2=Q-93,t t4 31 (MeCOBu)-H+ 92 100 79 4 23 & (CH2=6ggtBu)-H+ 6 6 4 f% (MeCOtBu)-H+ 26 100 ?3 38 4 (CH2=ç¡{OCH2tBu)-H+ 805 12 2 100 24 CH2=ç-OCH2tBu 85 10 25 28k 22 (MeCOCH2tBu)-H+ 100 15 6 3 (cH2=6¡lgPh)-H+ 100 38 n (MeCOPh)-H+ 100 1 2

F F¡ a. Width of HC=CO- peak at half height = 70 t 2Y b. Widh of HC=CO- peak at half height = 72 + 2Y c. Composite peak centred at \o m/z = 43. Central Gaussian peak superimposed on dish-shaped peak , width of composite peak at half height = 775 + 3 Y d. Width of HC=CO-peakathatf height =47t.2Y e. fi/z =57, (MeCOCHZ)- widthof peakathalf height =l07t2Y /. composite peakcentred on m/z = 43, central C'aussian peak superimposed on dish-shaped peaþ width of composite peak at half height = 166.5 t 3 V g. m/z = 57, (MeCOCH2)- width of peak at hatr h"ight = 10t t 2 V Ototnotes continued on nÐd paç with Table 2.9) Table 2.9 Collisional Activation Mass Spectra of Deprotonated Vinyl Ethers and of the Wittig Rearrangement Products. continued) r') Initial Ion Loss Formation C\ CsFk CgHa C¿He C¿Hto CsHro CsHtz QH¿ QHe lpn- CzHs- CzH- HO- Me- \ NJ (CH2=(¡¡gMe)-H+ 61. 5 CH2=Ç-9¡4" 5 MeCOCH2- (CH2=6¡-19Et)-H+ 53 CH2=ç-gB1 (CH2 =6gOCD2Me)-H+ 21. (CH2=Q¡1OCH2CD3)-H+ 36 (MeCOEt)-H+ 1 1 1 (CH2=ÇggPr)-F¡+ 56r 92 1.4 (MeCOPr)-H+ 18 (CH2=61¡9iPr)-H+ 100h 2 4 (MeCOiPr)-H+ 12 81 (CH2=6¡¡gBu)-H+ 23 4L 18 6 CH2=Ç-93' 39 100 30 (MeCOBu)-H+ 67i 218 (cH2=6¡19tBu)-H+ 100h 1 (MeCOtBu)-H+ 21 348 (CH2=6¡1OCH2tBu)-H+ UJ 50 M CH2=f-9çH2tBu 100 67 62 (MeCOCH2tBu)-H+ 26 4 (CH2=6¡19Ph)-¡¡+ 1 1 1 0.5 (MeCOPh)-H+ 33 u

case, the dish shaped comporient is ttæ onty peak detected (route A, Sctæme B (Scheme 2) is either not FJ h. n/z = 43 , in this 1. l..J operational, or altematively is minor in comparison with route A. The reason for this observation is not apparent, but it is interesting O that this scenario only occurs when R = iPr or tBu. i. m/z = 57 , (MeCOCH2)- width of peak at half height = lM + 2 Y i. m/z = -CH=CHOCH2tBu, 57 , (MeCOCHZ)= width of peak at half height =7't t 2Y k. the loss of CZHZindicates the presence of some formed from a minor amount of CH(SMe)3=CHOCH2IBu. Chapter 2 727 4o'w (2.70) L00 Itde-+ H2C=C=O

Me (H2C=C=O) þ L03 o' CH¿ + HC=C-O- (2.77)

102

N,IcO' + HC= CH (2.72) 4o'w lvfeO' (HC=CH) 104 L01 MoOH + HC {' (2.73)

Scheme 2.L8

The spectra of the a-vinyl alkyl ether anion (produced by desilylation) confirm that the ions RO- and HC=C- do not arise via this ion. Thus it is reasonable to assume that these ions arise from the p- carbanion. Furthermore, the absence of these ions in this spectrum indicates that the interconversion cr- and p- anions of vinyl alkyl ether (equation 2.74) does not occur under the conditions of collisional activation.

ào'* 4o'w (2.74)

100 L01

The complexity of the fragmentation processes increases dramatically as the length of the alkyl chain is increased. For the sake of Chapter 2 r22 clarity examples will be selected to highlight the processes involved rather than examining one particular system.

The decomposition of methyl vinyl ether may occur eithe¡ via the Wittig rearrangement product (102) or form directly via an ion complex

[(L03), Scheme 2.18].

Measurement of the peak widths at half height for the product ion HC=CO- (m/z = 4L ) in the spectra of CH2=Ç-OCH3 and the acetone enolate ion are identical, within the limits of experimental error (Tt/z= 70 + 2 eÐ. The Wittig rearrangement (Scheme 2.1S) can only be advanced tentatively as the evidence is, at this point, considered incondusive.

The spectra of deprotonated ethyl vinyl ether (Figure 2.14), the a' deprotonated ethyl vinyl ether (Figure 2.15) and the butan-2-one enolate ion (Figurc 2.16) are different from one another, but one of the ions formed is common to all spectra. The peak widths of the HC=CO- ion (a characteristic loss of methyl ketone enolate ions) were measured for the species examined and were found to be identical, within the limits of experimental error (Tt/Z = 47 * 2 eV). However, it is apparent that if indeed a Wittig rearrangement has occurred then it must be a minor

Pfocess. Chapter 2 \23

(-CHr=6¡1t; 43

4l

Figure 2.14 The Collisional Activation Mass Spectrum of CH2=Ç-

OCH2CH3 prepared by an S¡.,¡2(Si) reaction (equation 2.69). (-cHr{Ht 7t 43

HCtC- 25 EtO

45

HC:C-o' 4l

10

69

Figure 2.15 The Collisional Activation Mass Spectrum of Deprotonated Ethyl Vinyl Ether. f-crc-o - Chapter 2 4l 124

70

55

69

56

43 68 53 l5 t7 2'l

Figure 2.16 The Collisional Activation Mass Spectrum of the Expected Wittig Rearrangement lon, the Butan-2-one Enolate lon.

The peak width at half height of the composite peak (Figure 2.1,4) is 177.5 + 3V. When a potential of + 1,000 volts is applied to the collision cell about 50 - 60% of all of the peaks are shifted indicating that this proportion of decompositions is occurring in the collision cell. For m/z = 43 the collision induced component (ie. produced by decomposition in the collision cell) shows a major gaussian component, whereas the unimolecular component shows a dish shape peak. Thus the kinetically more favoured process occllrs from the u- anion.

The elimination of ethene dominates the spectrum of the deprotonated ethyl vinyl ethers. The ion at m/z = 43 is formed as a composite peak. The a-deprotonated ion ( L05 ) fragments to yield the dish-shaped peak exclusively ; the central Gaussian component being presumably arising from the p-{eprotonated ion ( 1-06 ). Deuterium Chapter 2 I25 labelling has shown that the loss of ethene involves transfer of a B-proton from the alkyl side chain to the vinyl anion (Table 2.9). The appropriate fragmentations are surnmarised in Scheme 2.19. \rq L,,t a o L05 il H2C-C. + [J2C=CII2 H ,1

) L06 Scheme 2.19

Determination of the proportion of ions decomposing unimolecularly vs collision induced provides an insight into the kinetics of the fragmentation (see Legend to Figure 2.16). The process occurring via pathway (b) is mainly collision induced (esú. c:u L0:1) whereas the process occurring via pathway (a) is mainly unimolecular (c:u 1:10). Thus fragmentation from the cl-deprotonated ion I pathway (a) ] appears to be kinetically more favourable than the fragmentation from the P - deprotonated ion I pathway (b) ]. Similar fragmentations are observed for all of the higher alkyl ethers ( R > Et ) , except for neopentyl vinyl ether which has no p-hydrogen. Chapter 2 726

4l(u )

4 4

25(c) 25(u) 4l(c) +1,000 V

0v

Figure 2.77 Partial Spectrum of the "Collision - Induced" vs "IJnimolecular" Decompositions of Deprotonated Ethyl Vinyl Ether recorded on a VG ZAB 2H.F Mass.spechometer. The spectrum obtained on floating the collision cell to +1,000Volts is offset for ease of comparison.

Table 2.10 Collisional Activation Mass Spectra of C3H5O- Ions.

Precursor Ion cAlcn m/z (rel. abundance) Spectrum Type

CH2-COCFI3 CA MS/MS4 56(100), 55(12), 4L(29) 55(3), 54(1), 53(3), 43(36), 42(700),41(31), CR MS/MS4

40(12), 3 9(30), 38( 1 4), 37 (72), 29 (3 6), 28(73), 27(34), 26(24), 25(7), t5(74), L4(22), 13(7), 12(2)

CH2=Ç-OCng, CAMS/MS/MSb 56(100), 55(13), 49(31) 43(39), 42(100), 41(28), 40(10), CRMS/lvtS/MSb 55(2), 39(30), 38(15), 37(7r), 29(33\, 28(9), 27(37), 26(27), 25(Ð, 1s(11), 14(21), 13(5) a. This data is taken f¡om reference 77, recorded on a VG 7AB ZIJF Mass Specfrometer. b. This was recorded on the Kratos MS 50 TA Mass Spectrometer. Chapter 2 727

Compelling evidence for the operation of the Wittig rearrangement is to be found in the spectra of CH2=Ç--OR ( R = Pr , Bu and neopentyl ). In addition to the formation of HC=C-O-, ã competing reaction forms an ion of moderate intensity at m/z = 57.

The spectra of deprotonated n-butyl vinyl ether (Figure 2.18), u- deprotonated n-butyl vinyl ether (Figure 2.19) and the hexan-2-one enolate ion (Figurc 2.20) are typical of these ions. The various spectra of the deprotonated ethers are similar to those described above, except for the formation of m/z = 57.

The peak width of the ions at m/z = 57 are (103.5 + 2 Volts) identical within the limits of experimental error for the related ions. When a potential of + 1,000Volts is applied to the cell only 70% of this peak is shifted indicating that this process is largely unimolecular. The identity of this ion was shown by CA MS/MS/MS and CR MS/MS/MS data to be acetone enolate ion (Table 2.9 and 2.10). The acetone enolate ion cannot be formed directly from either the cr,- or P- deprotonated n-butyl vinyl ether; it is formed via the Wittig rearrangement shown in Scheme 2.20 and illustrated in Figures 2.I8,2.I9 and 2.20. Chapter 2 I28 -o

7 )

L07

o

+ H3C-CH=CH 2

Scheme 2.20

This mechanism is also consistent with the observation that the branched vinyl ethers CH2=Q1-JO!Pr and CH2=Ç¡1OtBu do not fragment in this manner, since they are unable to undergo the standard enolate fragmentation shown in Scheme 2.20. Bu.O- 7l q,

(-C¡H¡)

43

I urö'c'clr,

57

4l 97 (-Hr) 7l (-cFI.) lc!c- 83 25 69

Figure 2.18 The Collisional Activation Mass Spectrum of Deprotonated n- Butyl Vinyl Ether. Chøpter 2 r29

99 (-C¡He)

43

o il Hrö'c'cu,

51 GCrI4) 1t 83

69 'lt 8t

Figure 2.L9 The Collisional Activation Mass Spectnrm of CH2=CH--OBun.

99

(-cru) n (-Hr) 83 o( Hrö'c'or, 57 69

7l ltcrc-o' 4l

81

Figure 2.20 The Collisional Activation Mass Spectrum of the Hexan - 2 - one Enolate lon. Chapter 2 130

Deprotonated phenyl vinyl ether (Table 2.9) does not undergo the

Wittig rearrangement. Instead the formation of PhO- ( presumably a fragmentation of the B- vinyl carbanion) proceeds almost exclusively (equation 2.75).

lorn PhO' + HC= CH (2.75)

22.7 Benzvl Vinvl Ether.

BenzyI propenyl ether provides the only example in which vinyl migrationlSSb to a benzylic position has been observed in solution

(equation 2.76 )

nBu Li Ph-cH"-of rr \,fr2 "\J (2.76) TMEDA, -270C , oln 108 1,09

The conditions required to effect this Wittig rearrangement are more vigorogs than used for benzyl alkyl ethers indicating that in this reactiory the migratory aptitude of the vinyl group is less than that of an alkyl $oup.

It was thus of interest to ascertain whether vinyl group migration was sufficiently facile to be observed in the gas phase,. The gas-phase behaviour of the prototypical example benzyl vinyl ether will be discussed Chapter 2 131 here. The collisional activation mass spectra of deprotonated benzyl vinyl ether and the-expected Wittig rearrangement product are recorded in Table

2.77.

Table 2.L1 Collisional Activation Mass Spectra of Deprotonated BenzyI Vinyl Ether and Benzyl Vinyl Alcohol.

Initial Ion Loss Formation

H' H2 CH+ CzHq CH2O PhCH2- Ph- C3H5O- CzFIsO-

PhCH-OCH=CH2 100 9 3 3 2 7 1 0.5

PhCH(CH=CHz)O- 100 10 3 3 1.6 5

The spectra are very similar and indicate that decompositions proceed through the same ion. The fragmentations can be readily explained using the scheme proposed for benzyl alkyl ethers (Scheme 2.2), although unlike benzyl alkyl ethers the loss of "R'". (CH2=çg') is not observed. Fragmentations of the unrearranged ether ( for example elimination ) have not been detected. The data supports the contention that the Wittig rearrangement occurs here also.

2.2.8 Divinvl Ether.

The data for the spectra of deprotonated divinyl ether, o- deprotonated divinyl ether and the but-3-en-2-one enolate ion are recorded in Table 2.1.2.

- Chapter 2 \32

Table 2.12 Collisional Activation Mass Spectra of Deprotonated Divinyl Ether and thé expected Wittig Rearrangement Product.

Initial Ion Loss Formation

H' H2 CH¿ F{zO CzHz HCzO- CzHg- CzH-

(CH2=(¡¡6CH=CHz)-H+ 3 0.5 0.2 0.1 100 4 0.5 3.5

CH2=Q-OCH=CHZ 0.5 100 28 .2

CH2=Ç1¡çOC1¡t- 6 412 L 100 T2 1.0

lvÍe ?Si 'NHz + A"^ + âOA + Me3SiNH2 Q.77)

The a-divinyl ether was prepared by an S¡¡2(Si) nucleophilic displacement on silicon using amide ion (equatton 2.77). The data indicate that the major process is loss of C2H2, and that it does not originate from the expected Wittig rearrangement product. Elimination of C2H2, could in principle proceed from both the o- and the F- carbanion. Decomposition of the ø-deprotonated anion may occur by a concerted process as indicated in (equation2.78) or alternatively it may undergo a proton transfer to yield the p - deprotonated anion in a stepwise mechanism as indicated in

(equation 2.79). , a o- HC=CH + \ (2.78) ìJ. L HC=CH2 Chapter 2 133

o' HC=CH + \ (2.7e) ìj2 HC=CH2 -

The formation of HC=CO- could conceivably arise either directly from the deprotonated ether (equation 2.80) or from the Wittig

rearrangement ion (equation 2.8L).

-Hc=ctr] 'HC=CHz [(trc=c=o) H2C=C=O + (2.80) / - âo o \ -- H2C=C=O +'HC=CHz (2.81)

Peak width measurements however , indicate that the loss of HC=CO- is different for the deprotonated ether (Ttn= 83.2!2.0 eV ) and the ketone enolate ion ( Ttn = .116.3 + 2.0 eV ). Thus the Wittig rearrangement is not a significant feature of this system.

22.9 Allvl Vinvl Ether.

In solutiory treatment of allyl vinyl ether ( 110 ) with butyl lithium

/ TMEDA in ether188a results in the formation of 1.,4-pentadien-3-ol ( 111 )

(equation 2.82). The migration of the vinyl group is not particularly facile ; a result that has already been noted for the Wittig rearrangement of benzyl

propenyl ether. Rautenstrauch e¡ a¡.788 have shown that migration of a vinyl group to an allylic centre proceeds with retention of configuration (equation 2.83), indicating that the intermediacy of a vinyl radical intermediate is highly improbable. Chapter 2 134

nBuLi TMEDA / Ether 4o/\-/ cH-oH (2.82) -27"C , oln J

1.L0 111.

nBul.i TMEDA/ Ettrer (2.83) oC -27 , oln OH

The dramatic divergence in the fragmentation behaviour of deprotonated allyl alkyl ethers and benzyl atkyl ethers has already been noted. Elimination was not observed during the decomposition of deprotonated allyl phenyl ether but this process does occur for allyl alkyl ethers. This reaction may occur in this case (equation 2.84), in competition with the Wittig rearrangement (equation 2.85). In addition, a rearrangement analogous to that observed for allyl phenyl ether could also occur (equation 2.86). F") \oaz + HC=CH Q.U) - Chapter 2 135

o - (2.8s)

LL3

7L2 (2.86)

7t4

The CA MS/MS spectrum for deprotonated allyl vinyl ether (Figure

2.18) is very complex. The elimination reaction (equation 2.84) must be a

very minor process as the HC=C- (m/z = 25 ) and CH2=CHCHzO- (m/z = 57 ) ions are very weak. If the Wittig rearrangement is operational then the only expected differences between the spectra of (112) and (LL3) should

be the loss191 of C2H3' (equation 2.87), and possibly the formation of C2H3-

(equation 2.88) and 16rt21e of CzFI¿ (equation 2.89).

a 4/o\/ + + HC=CHz Q.8n tt2

o' H2C: + H2C=CH2 (2.88) +l{^') / >" \ 'HC=cHz 1,L3 4lo + e.Bs)

The acrolein radical anion has a very low electron affinity2zo but if

formed it may be observed as it has a lifetime at thermal energies of 38+3 psec. This ion is not observed in the spectra of allyl alkyl ethers. The Chapter 2 136 second pathway, leading to the formation of the ketone - anion complex is favoured in.-solution over the ketyl radical anion - radical complex, perhaps reflecting the relative stability of the "free" vinyl anion.

The CA MS/MS spectra of (L12) and of the expected Wittig product ion (L13) illustrated in Figures 2.21, and 2.22 respectively, are significantly different. Surprisingly, the losses of CH2=CH. and C2H4 and the formation of C2H3- are observed in both spectra. The peak widths, particularly those formed by the loss of CzH+ and the formation of C2H3- are different in the two spectra. The experimental evidence thus precludes the possibility that (112) is fragmenting mainly via the Wittig rearrangement ion ( 1L3 ). This result is further confirmed by comparison with the charge reversal spectra of the ions (L12) and (113) shown in

Figures 2.23 and 2.24 respectively. Chapter 2 737

8l 83

55 (-CHr=ç¡1t¡

53

65 (-H2o)

56

21 29 43 4l 57 67 n 25

,r ìt

Figure 2.21. Collisional Activation Mass Spectnrn of Deprotonated Allyl Vinyl Ether. 55 ({n2-ç¡¡¡ 8l t3

27 53

56 ó5 GHzO) 4t

t7 25 29

(x 3) Figure 2.22 Collisional Activation Mass Spectnrm of Deprotonated Penta- 1,4-Pentadien-3-ol. Chaptø 2 138

2l

55

26

39

28

29

4l 50 53 49 63 65 62 66 gr82 l4

Figure 2.23 Charge Reversal Mass Spectrum of Deprotonated Allyl Vinyl Ether.

39

27

29

53

26 38

50 55 28 4t ll st82 l4 66 r3 1.5 43 6365

Figure 2.24 Charge Reversal Mass Spectrum of Deprotonated Penta-1.,4- dien-3-ol. Chapter 2 739

It is possible that a more plausible rearrangement pathway involves an anionic Claisen rearrangement. Sigmatropic rearrangements are often dramatically accelerated by the presence of n electron donors. Indeed, evidence for this type of anionic rearrangement has already been found during the investigation of phenyl allyl ether (see section2.2.1).

The spectra of (1-12) and the Claisen rearrangement product, [deprotonated 4-pentenal ] are identical (Table 2.73). Peak width measurements are in agreement within the limits of experimental error and the relative proportions of products from collision induced to unimolecular decompositions were the same.

The thermal Claisen rearrangement of allyl vinyl ether is well known and since it was remotely conceivable that rearrangement had occurred prior to deprotonation this possibitity was investigated by heating a sample of 4-pentenal at 150oC for 30 mins. These conditions are more vigorous than the sample would experience in the inlet (temp. 100oC) or briefly in the source (temp. cø. l.50oC). No evidence for the presence of the rearrangement product was found either by g.I.c. or i.r.. It is thus believed that the rearrangement observed is an anionic tearrangement occurring from the deprotonated ether (112)

The decompositions of this compound have been rationalised in the following way. The reaction leading to this product may involve either classical Claisen rearrangement to yield higlly reactive pent-4-enal ion (115) directly or alte¡natively via the intermediacy of a cyclised ion such as (114) ( Scheme 2.21). Chapter 2 140

Tablc 2.13 colrisional Activation Mass spectra of Deprotonated AJlyl vinyr Ether and Isomcrs.

Initial Ion Data m/za (Loss) Type 82 81 65 56 55 53 43 41 29 27 t7

41 100 13 10 30 22 6 5 4 CH2=Ct1ç¡1-OC"a"t 5 0.5 Tt/zb 50 43 60 50 47 48 54 36 43 c:uc 3:7 3:.7 1:0 3:7 1:1 7:3 3:2 2:3 /:5 100 5 Sd 50 9 4 0.5 10 0.1 .H Tt/z 53 43 45 34 68 26 c:u 3:/ 3:7 19:1 2:3 1:1 2:3 t9:1 o/od 40 100 '1 13 4 31 20 7 6 4 5 0.5 Tt/z 50 M 67 49 40 50 56 36 42 c:u 2:3 2:3 1:- 2:3 1:1 /:3 7:3 2:3 7:3 o/oe 28 100 t2 3Í 30 20 5 6 5 6 0.4 Tuz 48 u.5 4{t.5 39 49 55 37 & c:u 2:3 3:7 3:7 1:1 3:2 3:2 2:3 3:2 26 100 13 5d u 26 3 4 3 4 0.3 Tvz ß u.5 \) 50 4t 50 57 36.5 38.5 c:u 2:3 3:7 3:7 2:3 3:2 1:1 2:3 3:2 24 100 29 7d 23 28 2 2 5 4 0.5 Tt/z 48.5 43.5 50 39 36 40 c:u 1:4 4:5 2:3 5:4 7:3 7:3 2:3 2:7 a. with respect to the b' base peat ( 100% ). ts ) is the average of at c. least five scans measu¡ed to an accuracy of t 2.0 V. to the collision cel]. d' The cA spectrum of cH2=cH-cH2-cD-{HO is as follows m/z (ross) rel. I abundance ] : B3(H.)28 , 83(Hz)100 , 81(HD)18 66(Hzoßa , 6s(HoD)8 ' , 57(czH:')8 , 56(c2Ha)ae , 55(c2H3D)s , sa(cHzo)s s3(cDHo)2 27(C3H3DO)1 , , 2e(caH5D)1 , ¿' obtained by deprotonation of 3,4 - d-ihydro-2H-pyran ; the spectra oÍ 3,4 -dihydro-2H-pyran and 3,6 - dihydro,2H- pyran are identical. /. Poorly resolved. Chapter 2 '1.47

-..+

1L8 t19 I o OH

( +

172 - 1,t4 720 - 721

?

ü tL6 ü 115

ü 177

Scheme 2.21

The ion (114) is expected be a very strong base I eg. compare ÂGo¿s¡d

(CH3OCHs) =1704 kJ mol-1 ¡221 ¿n¿ to have a very low electron affinity [eg. E.A. ('CH2OCH3) = 0 eV]221, but it may have a transient existence, since delocalisation of the charge could occur to afford homoallylic anion (122).

+

lt4 - 122

This type of delocalisation has been postulated to account for acidity of the aromatic anion (123) in the gas - phase2zz, and has been observed in Chnpter 2 742 solution for the norbornenyl anion (L24) - nortricyclyl anion (125) slstem223 an-d the pyramidane anion224 $26).

123

--> 124 !2s

._ -+

126

Intramolecular 1,3 hydrogen transfer of an allylic proton of (11a) could occur leading to the more stable deprotonated tetrahydropyran (1L8). Rearrangement of the tetrahydropyran (1L4) to the cydopentanol alkoxide ion (L20) may proceed either by anion rearrangement (LL5 + 11,4) or by a radical anion - radical intermediate (equation 2.90). Homolytic cleavage of the vinylic CÐ bond would be presumably quite favou¡able as the radical anion should be quite stable I c/. electron affinity CH3CHO" = L66 kJ mot-1 (1.2 ev¡ltr+.

o' .n (2.e0) Chapter 2 143

It is also possible that in the gas-phase the rearrangement may occur by a 1.,4 addition (Michael "type" addition) of a vinyl anion to acrolein within the ion - neutral complex rather than 1,2 addition (wittig rearrangement). It should be noted that in solution the regiospecificity of reduction of a,p unsaturated ketones is strongly influenced by the metal ion. 1,2 Reduction occurs in the presence of the metal ion whereas only

L,4 reduction occurs in the absence of the metal ¡on150.

If the species (114) is formed then proton transfers to yield ions (LL6) and (117) seems feasible (Scheme 2.22). Fragmentation of ion (116) would lead to the formation of CzHs- ( equation 2.9r ) and loss of C2Ha ( equation 2.92 ) and the decomposition of (1L7) would result in the formation of CHO- (equation 2.93 ) and loss of CH2O ( equation 2.94 ). An alternate decomposition channel would result from nucleophilic addition of the vinyl anion to the carbonyl group to yield the cyclopentenol alkofde ion

(120) (Scheme 2.21). A proton transfer yielding ion (121.) might also occur. This ion (121) may decompose by losing H2 (equation 2.95) or H2O

(equation 2.96, Scheme 2.22). Chapter 2 744

o 'HC=CHz lY + e.97) H Q: H H Lt6 ln / o' H2C=C:C. + H2C=CH2 e.92) H 'HC=o 4/ + (2.s3) o *[ (+z) '""=o] H \17 C¿Hs + H2C=O (2.e4) o'

OH + H2 e.e')

\21 + HzO (2.96)

Scheme 2.23

The labelled deprotonated pentenal enolate ion CH2=CH-CHzCD--

CHO shows losses of C2H4, C2I{1D, CHzO and CHDO (footnote to Table

2.72). This result is quite surprising because it implies that the anions (L16) and (117) are interconverting. Direct interconversion is highty unlikely because the process is symmetry - forbidden. There have ?ä*ff"1.tr19,8 Sr .tt_. 1,2 hydrogen migrations recorded in solutionSg. The(degenerate 1,ä

hydride shift for the ethyl anion is calculated to be 202 kf mol-1, while the 1,2 hydride shift converting Me3Si- into -CH2MezSiH is calculated to be 376 kI mol-l and is not obserus¿Z?S. A possible mechanism for the reversible interconversion of (1L6) and (L17) is indicated below. Chapter 2 145

ü U ü 116 117

Finally, deprotonated 2 - cyclopropylethanal was also examined (Table 2.77). In solution, Rautensttu,r.6226 had shown that the deprotonated tetrahydropyran (118) rearranges very slowly ( > 200hr. ) to afford the 2-cyclopropylethanal anion (119). The spectrum of this ion is recorded in Table 2.11. and it is very similar to that of (112).

2.3 Conclusions and Summary.

2.3.L The Scope of the Wittig Rearangement in the Gas Phase.

The scope of this rearrangement has been studied in some detail and the results summarised in Table 2.14. It is interesting to consider the reasons for the difference in the gas phase behaviour of deprotonated benzyl-, allyl- and vinyl alkyl ethers. No elimination is observed from the deprotonated benzyl alkyl ethers whereas elimination occurred exclusively for deprotonated allyl alkyl ethers. Surprisingly, the Wittig rearrangement is observed from the deprotonated vinyl alkyl ether. The gas phase acidity of the unsubstituted moieties toluene200 (^Ho¿sid = 1594 kJ mol-l), propene2OO ( 1647 (^Hoacid = kJ mol-l), and ethene2O4 = ^Hoacid 1700 kJ mol-1) give an indication of the relative order of acidities the benzyl-, allyl- and vinyl- alkyl ethers respectively. Elimination therefore, is expected to be most favoured from the deprotonated vinyl alkyl ether Chapter 2 746 and be less favourable for the deprotonated allyl alkyl ether. What could account for the unexpected behaviour of these systems?

A possible explanation may be found if the electron affinities of deprotonated ethers is considered. These are not known, however the electron affinities of unsubstituted benzyt-z0O, v11y1-200 and vinyl-20a anions are known. They are 87,53 and 74kI mol-1 respectively. on this basis, the likelihood of rearrangement proceeding will be in the order of anion stabilities ie. benzyl > vinyl > atlyl. This is consistent with the tabulated results. Additional work will be necessary to determine whether this postulate is co¡rect.

Table 2.14 The Occurrence of the Wittig rearrangement and Elimination reactions from Deprotonated Ethers.

Compound Wittig rearrangement Elimination observed observed

Benzyl Alkyl Ether Yes No

Diallyl Ether Yes No

Atlyl Alkyl Ether No Yes Dibenzyl Ether No No Allyl Phenyl Ether Yes (major) No

Allyl Benzyl Ether Yes (minor) No

Diphenylmethyl Yes No Phenyl Ether

Vinyl Alkyl Ether Yes (minor) Yes (major)

Benzyl Vinyl Ether Yes No

Divinyl Ether No Yes

Allyl Vinyl Ether No Yes (minor) Chapter 2 747

2.3.2 Loss of Alkyl Radicals.

The loss of a radical "R'" from deprotonated ethers has been mentioned throughout this chapter. It is observed from the benzyl alkyl ethers but not other benzylic ethers. The loss of benzyl, aIIyI, phenyl and vinyl radicals are not observed. Delocalisation of the unpaired electron is generally considered to lead to stabilisation of the benzyl and allyl radicals and leads us to suspect that the "radical" losses observed are not the resr:lt of simple homolytic fragmentations. A major peak in the spectrum of deprotonated benzyl methyl s1þs¡191 is the loss of methyl radical. However, the loss of alkyl radicals is not in the order of radical stabilities. In particular, the relative losses of radicals R = Me, Et, Pr, iso Pt, tert Bu is U,39, 12,13,3L% respectively. In solutiory it is interesting to note that the elimination of alkenes competes with the Wittig reartangementlT9, and that there is evidence to suggest that the elimination reaction184,185 involves a radical process leading to the loss of the B - hydrogen atom. In the following section the reported losses of an alkyl group are examined. In this way it is hoped that evidence may be obtained which may shed light on this anomalous observation.

The loss of an alkyl radical "R'" has been observed in the spectra of a number of systems. Where it has been possible to examine this decomposition closely, the loss of the radical "R'" has been shown to proceed via a complex mechanism. The decomposition has been shown not to be the simple homolytic cleavage of an alkyl moiety that one expects.

Consider the loss of "Et'" from the 2-ethyl butanoate anion227 (127). By using a combination of C-13 and deuterium labelting227a i¡ has been Chapter 2 1.48 shown that the radical loss is a stepwise process involving a specific loss/transfer-of H' from the C-4 position prior to carbon - carbon bond cleavage. This result indicates that the incipient radical is stabilised in some way. It is possible that interaction of the semi occupied p orbital with the anti-bonding æ* orbital of the carbonyl group228 would result in ring closure in a 5 - endo - trigonal maruter229. The radical anion is stabilized by delocalisation of the carbon centred radical with non - bonding fully occupied p orbitals. The two processes involving (l) initial hydrogen atom transfer (Scheme 2.24) and (il) hydrogen atom loss (Scheme 2.25) will be discussed in detail.

o- o o'

Et Et Et a a -+ 127 L28

?

OH

HO Et H2C=CH2 + ,!

130 - 729

a o a + H

Scheme 2.24 Chapter 2 749

o-

Et

127

o' # o' o a

a Et Et H a H'

o' Et + Ij,2C=CIJ2 + H' a

Scheme 2.25

The two processes would eventually lead to the same product. Transfer of a hydrogen atom would lead to diradical anion (128) which may cydise to the unstable anion (129). Either of these species would readily lose ethene and the carboxylate anion (130) would form. Hydrogen atom loss would need to occur very readily if this process occurred in order that it is corrsistent with the fact that loss of ethene is not observed. 9'¡1ui¡230 ¡¿g shown that the loss of H' does not proceed readily from the acetic acid enolate ion (equatron 2.97),, and furthermore it occurs less readily than the loss of water (equation 2.98). The hydrogen transfer process therefore seems unlikely. Chapter 2 150

o a + H' (2.97) OH H2 o

o HC=C-O- + HzO (2.98)

Evidence for the alternative process involving hydrogen atom loss

(Scheme 2.25) can be fotmd in the reported deuterium labelled carboxylate ions, specifically Et2CDCO2-, (MeCDz)zCHCO2- and (CDsCHz)2CHCO2-. These ions show losses of H' : D' of 45 : 55, 100 : 0, M : 76 rcspectively. This indicates that the formation of the cr - centred radical anion EI2C'CO2- occurs most readily but is accompanied by the formation of the 1 - centred radical anion 'CH2CH2(Et)CHCOz- or as proposed here, the cyclised lactone radical anion.

Cyclisations similar to that proposed in Scheme 2.25 have been observed in the condensed phase23l,232 and an exampls231 is shown here (equation 2.98). The ring opening reaction233 has been observed with neutral and radical intermediates.

o- o

o + o (2.e8\

-

It should also be noted that whilst radical reactions are sensitive to polar and steric effects, the kinetically formed product usually predominates over the thermodynamically more stable 6-membered ring produg¡228. Direct radical cleavage of this intermediate would be presumably very facile. Coulson s¡ 4¡.234 have studied the gas-phase Chapter 2 151 acidities of alkoxides and proposed that long chained alkyl group may in fact interact with the alkoxide ion by anisotropic polarisation and adopt a more constrained conformation (equation 2.99).

o (2.ee) ü

Extending that observation to the 2-ethyl butanoate anion leads to the conclusion that the anionic centre causes anisotropic polarisation of the alkyl group. Incipient radical formation becomes favourable because of the pre-organisation of the transition state caused by the polarisation.

A similar process may occur during the fragmentation of deprotonated aldoximines23s ( 131 ). The H/D isotope effect occurring during the loss of "Et'" is very large (= 3.0) indicating that the loss of H' is perhaps even more advanced during that decomposition than in the carboxylate anion ( 1,27 ). This is readily understood if the stability of the radical being formed is considered. Cyclisation onto the unsaturated slster¡23sa would yield the resonance stabilised radical ('J-,32 ) , where the major resonance contributor results in a positive charge residing on nitrogen ( Scheme 2.26 ). Chapter 2 I52

o- o- - I P I cH3cH2cH2. O a +. - -\ / N \ -)-ñ- / "] l.aaCH"CH"CHI H) L3L r32

o I N a + CHr=ç¡7, + H

Scheme 2.26

The loss of "R'" from dialkylated benzylic anions (133) has been demonstrated to involve the loss of H' and an alkene u1ss236. Cyclisation as before leads to a transiently stable radical anion ( Scheme 2.27 ). Cyclisation of a y centred radical has recently been reported to proceed in uncharged intermediates in an analogous manner. Chapter 2 153

R

I 1.33 R

I

R R

H H a I H R R 1

R

+ H2C=CH-RI + H' a

Scheme 2.27

It is possible that the loss of "R.,, from deprotonated N- monosubstituted amides (134) also proceeds via a similar mechanis m237 (equation 2.100). unfortunatery this cannot be proven because the deprotonated N,N disubstituted amide (13s) undergoes elimination# (equation ,,p.,, 2.101) in preference to the loss of (equation 2.102).

# The reason for the divergent behavior can be explained if the relative acidities of the anions formed a¡e considered. An elimination reaction competes with the loss of ,R..,,from (134) ÂHo¿ç¡¿ ( [ HcoNHMe ) = 1509 kI mol-l ] whereas exclusive elimination occurs from (135) (CIfuCONMe2) l^Hoacid = 1569 kf mol-l ]. The relative order for the loss of "R'"increases in the order fBu < Et = lBu = gPr < oBu << iPr. This observation is certainly not consistent with the loss being a simple homolytic cleavage. Chapter 2 154

o CH 3 CH. I t" .,-c>. + H2C=CHz + H (2.100) N^ N H o/c\N a t34 ö

or (2.101) -{ \J H2 o il + H2C=CH2 e.rcz)

While it could be argued that these are "remote" fragmentatioru (le.

fragmentations uninfluenced by the charge centre in any *uy) it must be stressed that this is not the case. The loss of H' occurs from a specific site

so that a transiently stable radical anion is formed. The formation of a five

membered ring is often associated with these processes and this reflects the optimum arrangement of atoms that allow interaction of the developing

semi-occupied p orbital with the anti-bonding rc* orbital adjacent to the negatively charged centre.

The loss o¡ 't¡o" from deprotonated benzyl ethers similarly does not aPPear to be a straightforward process. On the basis of radical stabilities, the loss of ¡-butyl should be extremely pronounced: experimentally that is not so in this system. Thus, the fragmentation of deprotonated benzyl alkyl ethers may also be more complex involving the loss of H' followed by an alkene. In this case a six membered ring would be formed. Whilst Chapter 2 155 this mode of cyclisation is not common in solution this process has been observed for cyclisu¡ion238 onto an aromatic ring (eg. equation 2.103). Attack at the ipso carbon (phenyl C-1 position) is not favoured as this would prevent delocalisation of the negative charge. The possible intermediates are (136) and (137) of which (136) appears more likely. A rationale for the decomposition of the alkyl ethers from (136) is summarised in Scheme 2.2g.

a

o o

(2.103)

H

o

H

L36 137

Qo + CH2-CH2 H

136 Scheme 2.28 Chapter 2 156

If the substitution of the atkyl group is increased it is likely that the stability of the intermediate will alter due to ste¡ic effects. The obvious way to probe the mechanism here is to examine the relative losses of isotopically labelled ethyl ethers from deprotonated benzyl diethylacetal (138). Deuterium labelling on the p position of the alkyl group should have little effect on the loss of "R'" if a simpte cleavage is involved. If the p- H' loss occurs then the loss of "Et'" ( equation2.t04 ) should be far more pronounced than of "CD3CH2'" ( equation 2.105 ).

o' Ph-C.-OCH + H2C=CD2 + D' (2J,04) OCH2CH3 /// 2CH3 Phö( ocH2cD3 o' Ph-C-OCH + lH2C=ClH2 + H' (2.105) 1,38 2CD,

Unfortunately, the loss of an alkyl radical was not observed when the unlabelled benzyl diethylacetal was examined (Figure 2.25). Instead, loss of ethene predominates (equation 2.106).

oEr o- I Ph-CH-OEr + H2C=CIJ2 (2.106) H2

- Chapter 2 t57

l]9

l5l (-CHt=6¡1t¡

120

104

133 t77 45 76 43

Figure 2.25 The Collisional Activation Mass Spectrum of Deprotonated Bet:øyl Diethyl Acetal [PhC-(OEt)2]. Cha ter a

Smiles Rearrangement.

3.L lntroduction.

Since the discovery of the Smiles rearrangement239 in 1930 , this rearrangement has become one of the most thoroughly examined and exploited of all anionic rearrange*sn¡s24O. The rearrangement is formally

described in equation 3.1..

X Ph-x-(cHz)o-Y - : (CHr)o: Ph-Y-(CHz)o-X' (3.1)

Nucleophilic addition to the electron rich aromatic n system is generally unfavourable; however this impediment is removed when higttly polarisable electron withdrawing groups ( such as NO2, SO2, NR2

and halogens ) are present in the ortho and / or pjuê positions.

For the reaction to proceed in the direction indicated (to the right) X should be a good leaving group and Y a powerful nucleophile. Often vigorous conditions are required to allow these rearrangements to proceed.

Addition of the nucleophile to the ipso (C-L) position is favoured by

polarisation at C-1 caused by the inductive effect of X, although this is not essential to form the spirocyclic o complex. The closely related Chapter 3 159

Grovenstein-Zimms¡¡n¿n89,92 rea¡rangement proceeds via a spirocyclic complex eveft though X =Y = CHz.

o - Complexes, whether bridged or not, are usually referred to as

Meisenheimer complexes241. They are often very stable and in a number of favourable instances X-ray crystallography has been used for structural verification. They are more commonly identified, or their presence inferred, by their UV specfra in solution.

The length of the bridging group which may lead to a stable o- complex has not been studied in great detail. Typically spirocyclic rings containing 3, 4 or 5 atoms appear to be very common. Three examples

(1l}¡z+2, $ZS¡z+z and (L40)244 are given below :

N/ P1

cD3 S o R H3 cD3 Noz

138 139 L40

The stability of the Meisenheimer complex is affected by the presence of counter-ions and the solvent. The equilibrium concentr¿1¡6n245 of the 1.,1 dialkoxy adduct (141) for example, is dramatically affected by chelation of an alkali metal ion ; increasing in the order K+ < Na+ < Li+. Chøpter 3 160

H3CO OCH ¿ 3 M. Noz : M=Li*, Na*, K

Noz r41

Mutai and Kobayashi246 have reported the Smiles rearrangement of a series of (4-nitrophenoxy) N alkylanilines (142) and studied their rearrangement in the presence of base ( equation 2.2 ).

P\ (cHtnoH t oNHPh N

EtrN n=2-5 (3.2) CH3CN

Noz -+ Noz 1l2

As expected, the relative rate of reaction is in the order 2 > 3 > 4 > 5.

Curiously, this order changes when the solution is irradiated: thus the relative rate of reaction is now in the order 4 > 5 >> 3 > 2.

A limited systematic study of the effects of solvent and base on intramolecular rearrangements of 1-(P-N-Me-aminoethylthio) -2,4,6- trinitrobenzene (L43) was reported by a Russian group247. They have found that using a weak base (pyridine) in DMF, initial nucleophilic substitution at the ipso position (le. the Smiles rearrangement) is not favoured. However in the presence of a strong base (KO!-Bu), the Smiles Chapter 3 767

rearrangement occurs completely in preference to ortho attack (Scheme 3.1). The pr-oportion of ortho vs ipso product formed under various

conditions is indicated in Table 3.1. These results were rationalised on the basis of the nucleophilicity of the anion of (1a3).

n /-a NHCHl S HS CH 3 ozN Noz ozN Noz o2 Noz

Noz Noz NO 2 t43

CH3 I ozN

¡ Noz CH¡ Noz 144 145

Scheme 3.1 Chapter 3 762

Table 3.1 The Affect of Solvent and Base on the

Intramolecular Nucleophilic Aromatic Sub stitution of (1a3).

Base Solvent Ratio '1.44:145

t-BuOK DMF 100: 0

t-BuOH 86: 14 c6Fk 25:75

NEt3 DMF 67: 33 t-BuOH 30:70

C6I{6 -: -a rT DMF 0:100 t-BuOH 0:100 CeFk 0:L00

ø This was not reported.

o-Adducts may also arise by attack at a carbon atom which is not bearing a leaving group. In many instances this is readily explicable. For example, treatment of 3,S-dinitrobenzonitrilez4S (146) with methoxide yields two products ( Scheme 3.2 ). Nucleophilic addition to the aromatic ring occurs solely at C-2 and this is rationalised as being due to the powerful electron withdrawing effect of the three nitro groups on the aromatic ring. Chapter 3 163

CN H

ON{e

o2 2

CN /

a t46 MeO -cl*" lvfÐ -"7NH \ H

o a o2 2

Scheme 3.2

The Von Richter reaction249 is an example where attachment at C-2 occurs readily ( Scheme 3.3 ). This complex reaction has been extensively studied and proceeds as a result of the high nucleophilicity of the nitrile anion and the powerful electron withdrawing effect of the nitro group which activates the ortho oosition.

o co 2H ...H I Noz + \o' Hzo CN ffi_ steps - Scheme 3.3 Chapter 3 764

"Vicarious" nucleophilic aromatic substitution reactions2sO form a distinct class_-of substitution reaction. FIowever, attack at C-2 is favoured in these instances by the presence of a suitable leaving group on the

benzylic carbon (equation 3.3 ).

o 3 'CN + + (3.3)

Meta bridged o-adducts have seldom been reported, requiring the aromatic system to be highly activated by electron withdrawing groups. In

the presence of strong bases, substituted acetoacetanilides react with L, 3,5 - trinitrobenzene2sl ( equation 3.4 ) affording the meta bridged o-adduct (147). In a similar way the reaction of diethyl sodio-3-oxopentadioate

reacts with l-substituted 3,5 dinitro - 4 - pyridones2îz ( equation 3.5 ) to

yield the meta bridged o-adduct ( 148 ).

2

o o E0z 2 2 ,,\,Å NHC6H4X (3.4) I H I Noz o="\ o NH r47 I c6H4x

o *r'* ozN NO" o ' EÐzc-^-rcorEt ErO 2 o (3.s)

R Noz- EtO 2C 1.48 Chapter 3 165

It is apparent that the degree of activation of the aromatic æ system, the presence of counter-ions and the solvent are important in

determining the rate and course of reaction in solution. In the gas phase the absence of solvation and ion-pair formation will undoubtedly affect the rate of reaction. Indeed, the work of Sams and Simmons253 has shown this very clearly in solution. The addition of 18-Crown-6 to a solution containinr 7,2-, 1,3- or 1,4- dichlorobenzene, and sodium methoxide (NaOCH3) in acetonitrile allowed nucleophilic aromatic substitution to proceed under conditions which would otherwise fail to effect this type of reaction (equation 3.6). Significantly, there was no evidence for the formation of a benzyne intermediate indicating that the basicity has not been enhanced to a greater extent than the nucleophilicity. Unfortunately it is not always possible to predict the effect that the absence of solvent induced polarization will have on the reaction in the gas phase.

ON{e

NaOMe (3.6) cr CH3CN cl 18-Crown-6

under similar conditions it has been possible to observe Meisenheimer complexes by nmr2s+.

Most recently, Nibbering25s has produced evidence in support of o - complex (149). Whilst this structure is compatible with results obtained in analogous solution phase reactions, the structure proposed is by no means unequivocally established. Chapter 3 166

749

3.2 The Gas Phase Smiles Rearrangement.

In this chapter, the gas-phase rearrangements described in equation

3.L for the systems X = Y = O ; X= S, Y = 0 and X = O ,Y = S are reported. The data for the collisional activation mass spectra are recorded in Table

3.2.

3.2.1 Phenoxvalcohols.

The spectra obtained for all deprotonated phenoxyalcohols Pho(CH2)noH (n = 2 ,3 ,4 ) are totally dominated by the formation of the phenoxide ion (PhO-). Other decomposition channels are observed for the higher phenoxyalcohols but these are very minor. The relative proportion of ions decomposing unimolecularly is indicative of the energetics of the process(es) forming the phenoxide ion. The relative proportion of ions fragmenting outside of the gas collision cell is high.

Hence the conclusion may be drawn that this is a very facile process as collisional activation is not required to effect decomposition. Chapter 3 767

Table 3.2 The Collisional Activation Mass Spectra of Deprotonated Phenoxy alcohols,Thiophenoxy alcohols and Phenoxy alkanethiols.

Parent Ion Loss Formation

H' H2 CH2O phOH Pho- PhS- Peak Peak width ,rr.b width ,r,.b a a

PhO(CH2)2O- 15 100 40.5 10:3

PhO(CHz)gO- t2 100 c l0',2

PhO(CH2)aO- 76 100 395 10:6

PhS(CH2)2O- 8 26 100 23.t 10:1 57 d l0:4

PhS(CH2)3O- I 100 47.7 l0:2 %

PhS(CH2)aO- 10 4 100 10:5

PhS(CHz)sO- 8 lm 47.5 10:4.5

PhS(CH2)5O- t2 100 4.9 10:4

PhOCH2S- 100 88 t6

PhO(CH2)2S- 100 15 52 26 9 10:3

PhO(CH2)3S- 100 I 25 15

PhO(CHz)¿S- 72 5 100

PhO(CHz)sS- 6615 100

ø The measurement of the peak width at half height (Volts) is accurate to + 0.5 v. b The relative ratio of peaks obtained on applying a potential of +1,000 Volts to the gas collision cell. c Composite peak (Figure 3.5). d Composite peak (Figure 3.14).

These data however do not permit any conclusions to be drawn on whether the product PhO- ion is formed by a simple internal nucleophilic Chapter 3 168

( SNi ) displacement reaction, ( equation 3.7 ) or whether the phenoxyalkoxide ion fragmentation had been preceded by equilibration of

the open chain alkoxide ions via a Smiles rearrangement ( equation 3.8 ), or whether the decomposition occurs directly from the spirocyclic Meisenheimer intermediate [(Ls0), equation 3.9]. These reactions are shown in Scheme 3.4.

'o o Pho- I Pho' + (3.7) (CHz)" L, (CHÐ"

l8o Ph-o-(cHto-l!o (CHr)o Ph-18O-(CH2),-O ' - o/ - 150

sNi Slri

Pho' (3.8) Pho' + Phl8o' (3.9) Phl8o' (3.8)

Scheme 3.4

Both processes will result in a degenerate rearrangement in which both oxygen atoms become equivalent. In order that this reatrangement could be examined without altering the nature of the leaving group or the nucleophile, the terminal oxygen atom was replaced by isotopically labelled ( O-18 ) oxygen. The results of the isotopic labelting experiments

are shown in Figure 3."J.,3.2 and 3.3. Chapter 3 169

ph'8o .. qs Il9 Ph'"o -" 93

Figure 3.1 Collisional Activation Mass Spectrum of PhOCH2CH2189- recorded on the VG ZAB 2HF Mass Spectrometer.

Ph lóo | 5-3 93

Ph¡80 95

l2r t23

Figure 3.2 Collisional Activation Mass Spectrum of PhOClIzCH2Ctt2lþ- recorded on the VG ZAB 2HF Mass Spectrometer. Chapter 3 170

I rrJ Ph'oo 93

Figure 3.3 Collisional Activation Mass Spectrum of PhO(CH2)a189- recorded on the VG ZAB 2FIF Mass Spectrometer.

For phenoxyethanol ( n = 2 ), the relative abundances of the labelled and unlabelled product ions are very similar. On the basis of the measured relative peak heights for Ph16O- (m/z = 93 ) and Ph18O- (m/z = 95), the ratio of L00 : 95 ( + O.S ) respectively is obtained. Assuming that all of the 2-phenoxyethoxide ions have undergone the Smiles rearrangement then the 016 / 018 isotope effect is calculated to be 1.05 + 0.03.

By marked contrast, the abundance ratio of Ph16O- : Ph18O- from 3 phenoxypropanol (n=3) is L00 :76. The spectrum of PhO(CH2)¿18O- shows only Ph16O-. Clearly the Smiles rearrangement proceeds to a much lesser extent with the addition of a methylene ( CH2 ) group ( ie. n= 3 ) and does not proceed at all when n = 4. Chapter 3 771

The ion dissociation characteristics of PhO(CH2)rtag- are noticeably different from PhO(CH2)218O-. The peak profiles of the PhO- ion ( m/z = 93 ) are illustrated in Figures 3.4 and 3.5 respectively.

Figure 3.4 PhO- peak (m/z = 93) Figure 3.5 PhO- composite peak from PhO(CH2)zO-. Peak width (m/z = 93) from PhO(CH2)sO-. at half height = 40.5 t 0.5 Volts. Peak width of major component at half height ca 35 Volts: minor component ca I05 Volts.

The peak arising from decomposition of PhO(CH2)2189- is narrow and gaussian ( width at half height 40 + 0.5V ). However the peak from PhO(CH2)g18O- appears to be composite; a major component ( width at half height L05 = V ) superimposed on a narrow peak (width at half height = 35 V ) The minor process has a major collision induced component indicating that this is energetically a more unfavourably process (see the broad displaced Ph16O- peak in Figure 3.6). Two mechanisms must be Chapter 3 172 occurring to account for the formation of PhO- from PhO(CH2)gO-, whereas single mechanisms occur for the homologous Pho(CH2)no- ions (whenn=2and4).

Ph róo

Ph ''o

Ph¡60- Ph'60

+ 2,000 V

0v

Figure 3.6 Partial Spectrum of the "Collision Induced" vs "IJnimolecular" Decompositions of PhO(CH2)3189- ion recorded on the VG ZAB 2HF Mass Spectrometer. The collision induced spectrum, recorded with the collision cell floated at +2,000 Volts is offset above to the left, for ease of comparison.

If the acyclic 3-phenoxypropoxide ion was interconverting via the spirocyclic Meisenheimer complex and then fragmenting the Ph18O-

(m/z = 95 ) peak should also be composite. It is a narrow peak indicating that the acydic Ph18O(CHz)nO- ion is presumably not formed otherwise a composite peak would be expected from this ion also. It seems reasonable to propose that the Ph16O- (m/z = 93 ) and Ph18O- (m/z = 95 ) ions formed by the more facile process are derived from a common precursor, namely Chapter 3 773 the Meisenheimer complex and that the acyclic species fragments by an energetically_Less favourable process prior to formation of the spirocyclic intermediate.

An alternative mechanism for the fragmentation of the PhO(CH2¡rtaO- ion can be envisaged ( equation 3.10 ) involving a sequential loss of formaldehyde and ethene.

18 (3.10) -.-+ Pho + H2C=CH2 + CH2=O|8 Pho

Flowever, it is unlikely that this can account for the presence of the minor Process. If fragmentations are proceeding from the acyclic species then both Ph16O- (m/z = 93 ) and Ph18O- (m/z = 95 ) should be composite and this is not so. The PhO- ion [ ÂHo¿si¿ (PhOIÐ= 1465 18 kI mol-111e2 ig a weak base and would not deprotonate CH2=Q ( ÂHoacid = L6471 3 kJ mol-l )256. Evidence of this fragmentation, if it did occur, could not be readily observed. In view of these observations, the following equations tentatively summarise the processes observed.

Pho(cH ùzo' PhO' + (3.11) -+ - -o Pho PhO + rî (3.12) - 'o PhO + (3.13)

-> Chapter 3 174

The propensity of the acyclic ions to form the spirocyclic Meisenheimer intermediates is directly attributable to the loss of entropy as the size of the spirocyclic ring being formed increases. The intermediacy of the spirocyclic Meisenheimer complex needs to be demonstrated if the occurrence of the Smiles rearrangement is to be substantiated. It is conceivable even though considered highly improbable that nucleophilic attack may occur at either the ortho- or meta- position of the aromatic ring. Conceptual alternative pathways are given in Schemes 3.5 and 3.6 respectively.

a -ol H 018 a - H - rï "l H

018 018 - - lG)"1 Scheme 3.5 Chapter 3 175

<_ 018

- H

H a 018

<- "l + ol8 ol8

Scheme 3.6

The structure of the neutral product formed has tentativety been assigned as ethylene oxide. This structure would logically result from a simple internal nucleophilic displacement ( Sxi ) reaction , however cleavage of the phenoúde alkyl CÐ bond may involve a far more complex reaction, perhaps involving hydride transfer to form acetaldehyde. As it aPpears that deprotonated 2-phenoxyethanol fragments by only one mechanism to yield one product it was hoped that a neutralisation - reionisation experiment would allow the neutral product to be identified. The experiment was especially simple to perform because the fragmentation was largely unimolecular (le. did not require collisional activation) and it was therefore possible to use the ZAB to attempt identification of this neutral. A similar experiment was performed on the Kratos MS 50.T4 instrument. The accelerating potential of the ZAB was set at -7 kV. The gas collision cell (in the 2nd field free region) was floated at '7.2 V thereby deflecting the negative ion beam but allowing the neutral beam to pass through the collision cell. The neutrals were ionised using

He or 02 to yield positive ions which were analysed used an electric sector scan. Unfortunately a spectrum very similar to ionised benzene was Chapter 3 776 obtained, obscuring the spectrum of any neutnl C2Hao species. The mechanism of the formation of benzene is not clear. If it is formed by a heterolytic fragmentation then the negative ion produced must be very unstable having undergone electron detachment to yield its radical. The same result was obtained when the neutralisation - reionization experiment was performed on the KRATOS MS 50 TA instrument. An attempt was also made to identify the neutral product obtained from the postulated S¡¡i reaction for Pho(CH2)+o-. once again the spectrum obtained is consistent with that of ionized benzene. Thus the identity of the neutral species whilst consistent with the results obtained can only be taken as tentative.

The processes shown in Schemes 3.5 and 3.6 seem unlikely: ortho attachment requires nucleophilic attack at a site of high electron density and meta addition involves the formation of a highly strained unactivated site. In solution meta bridged adducts have been reported to be stable because the charge is delocalised onto an electron withdrawing group. This can not occur in this example.

Several approaches to the resolution of this problem were attempted. Initially the approaches were centred on the independent synthesis of a spirocyclic intermediate (L51) or (1S2). n n o.. o o,

L51 152 Chapter 3 777

Deprotonation of either compound (or a mixture) would have allowed a direct comparison of the behaviour of the spirocyclic Meisenheimer complex with the species formed from the acyclic PhO(CH2)2O- ion. Several approaches to the resolution of this problem

were attempted. Unfortunately this approach proved unsuccessful. The reasons for the failure of the synthesis will be discussed later in this chapter.

An alternative and ultimately successful strategy was to examine the decompositions of the phenoxide ion formed independently and during the fragmentation of the 2-phenoxyethoxide ion. It was hoped that a specific fragmentation could be observed which would allow the site of attachment of the oxygen atoms to be identified.

Fortunately, the collisional activation mass spectrum of the phenoxide ion ( Figure 3.7 ) reveals two major decomposition pathways. Chapter 3 778

9l 65 (_co)

&

75(-H2o)

4l 68

39

Figure 3.7 Collisional Activation Mass Spectrum of the Phenoxide Ion , recorded on a VG ZAB zF{F irstrument.

These were losses of m/z = 28 and m/z =29,ie CO and'CHO respectively. This assignment was confirmed by examining the spectra of

tl;re 2,4,6-trideuteriophenoxide ion which showed losses of CO and 'CDO and 1-13C phenoxide ion which lost 13CO and '13CHO. These fragmentations are reproduced as the partial spectra in Figure 3.8 and 3.9 respectively. In the partial spectrum of the d3-PhO- ion no evidence for the loss of 'CHO could be found. This indicates that the loss of 'CDO is very specific; the hydrogen (deuterium) presumably arising from the ortho position. Chapter 3 6s (-co) 179 68 (-co)

(-'cHo)

64 (-'cDo) 66

Figure 3.8 Partial Collisional Figure 3.9 Partial Collisional Activation Mass Spectrum of the Activation Mass Spectrum of the PhO- ion ( tlire m/z = 64 - 66 2,4,6 Dg -PhO- ion ( the m/z = 64 - region ). 66 region ).

The d3-PhOCH2CH2O- ion was examined by consecutive MS/MS. However, in this instance the decomposition of the fragment d3-Pho- ion produced losses of 'CHO and 'CDO ( Figure 3.10). This result indicates that either the Smiles rearrangement had not occurred or more likely that scrambling of the label had occurred on formation of the spirocyclic

Meisenheimer complex and/or during its fragmentation. Scrambling has been reported in negative ion chemistry79l,206,212 but it is rare. The CA MS/MS/MS spectrum of the daughter (1-13C phenyl)O- ion formed on decomposition of the 1r-t3Cphenyl)ocH2CH2o- ion shows exclusioe loss of 13CO ( Figure 3.11 ). There is no evidence for the loss of 12CO (a shoulder or peak at m/z = 66 would be expected). Clearly the formation of Chapter 3 180

the spirocyclic Meisenheimer complex is tine only intermediate compatible with this result. 65 (-'3co) 6s (- co)

& &

Figure 3.L0 Partial Collisional Figure 3.11 Partiaf Collisional Activation Mass Spectrum (the Activation Mass Spectrum (the m/z = 64 - 66 region) of (Ph- m/z = 64 - 66 region) of (Ph-l- O- ion (m/z = 94). 13C)o- ion (m/z = 94) from 2- (phenoxy - 1-13c)ethoxide ion.

3.2.2 Thiophenoxyalcohols.

As with the phenoxyalcohols, the spectra of the thiophenoxy alkoxides are very simple. By contrast the possible extent of the Smiles rearrangement is more easily ascertained as the collapse of the spirocyclic

Meisenheimer complex would yield two ions , Pho- (m/z = 93 ) and PhS- (m/z = 109 ). The ipso carbon atom is attached to a less electronegative Chapter 3 181 atom ( s vs o ) in the starting material ; however, as before the nucleophile is the alkoxide ion.

Examination of the data in Table 3.1 and Figure 3.12, shows that the

Smiles rearrangement is occurring very readily for n = 2 and 3 (phO- is the base peak) and to a very minor extent for n = 4. The homologous alkoxides PhS(CH2)nO- ( n = 5, 6 ) fragment to yield only the phS- ion. The nucleophilic displacement ( SNi ) reaction dominates for n > 4.

Pho'

PhS

Figure 3.12 Collisional Activation Mass Spectrum of PhSCHzCHzo- recorded on the VG ZAB 2HF Mass Spectrometer.

o PhS- + l_\ (3.14) Phs(cH2)2o' S - PhO- + l_\ (3.1s) L53 Chapter 3 182

It is interesting to note that the relative proportion of ions decomposing- unimolecularly to yield the phenoxide anion is even higher than for the corresponding phenoxyalkodde ions. The thiophenoxide ion is formed far less readily, based on the proportion of ions formed within the collision cell ( le. requiring collisional activation ). This suggests that sulphur is a better leaving group than oxygen.

The phenoxyalkoxide ion PhX(CH2)3Y- ( X =1 @, Y - 1&O ), as noted above, yields Phx- ions ( X = 169 ) via two decomposition channels. As the nature of the X group has changed, so also may the energetics of the decomposition of the thiophenoxyalkoddes.

Consequently, the peak shapes of thiophenoxide ions formed on fragmentation of Phs(cHz)no- ions ( n = 2, 3 ) were scrutinized; the expanded peaks are reproduced in Figure 3.13 and 3.14. When n = 2 a narrow Gaussian peak ( width at half height = 46 ! 0.5V ) was observed for the fragmentation leading to the PhS- ion. Thus it is probable that the Smiles process occurs exclusively (equation 3.14 and 3.15). Flowever, when n = 3, a composite peak was observed. In this case it was not possible to determine the width of the narrow peak arising via fragmentation of the spirocyclic Meisenheimer complex (L53), however the width of the broad peak is determined to be 96 + 0.5V. This component is likely to arise from an S¡¡i process ( equation 3.16 ) and is of comparable abundance to the ion formed on rearrangement.

o e' + (3.16) Ph S + ü Chapter 3 183

Figure 3.13 PhS- peak ( m/z = Figure 3.14 Composite PhS- peak

109 ) from PhS(CH2)2O-. (m/z = 109 ) from PhS(CH2)3O-.

It is conceivable that a subtle change in the structure may lead to a change in the course of the rearrangement. Therefore it is necessary to confirm that the spirocyclic Meisenheimer complex (153) is indeed involved in the rearrangement. Thus the collision induced spectra of the appropriate product ion from (1-13C phenyl)SCHzCHzO- ion were recorded. If the heteroatoms do become attached to the same carbon atom then (1-13C phenyl)O- and (1-13C phenyl)S- ions witl be formed exclusively. The MS/MS/MS experiment on the m/z = 94 ( C-13 labetted PhO-)showsaloss of m/z =29(13CO)andm/z =28 (12CO)inthe ratio of 95 : 5 . Clearly the Smiles rearrangement has occurred and the spirocyclic Meisenheimer complex is thus dominant. The fragmentation of the PhS- ion does not proceed in a manner directly analogous to the

PhO- ion (ie no losses of 'CHS or CS) are observed. However the loss of C¿H¿ noted for this ion is most useful2s7. If the sulphur atom has Chapter 3 784 migrated from C-1 ( labelled with 13C ) during the rearrangement then the loss of 12c313cH 4may also be observed. The cA MS/MS/MS experiment on (1-13C phenyl)S- shows a broad peak for loss of CaHa. The low sensitivity, coupled with the width of the peak do not allow a small loss 12c313cH4 (certainly < r0%) to be excluded. Thus the evidence in both cases is indicative of reaction through a Meisenheimer complex.

3.2.3 Phenoxyalkanethiols.

The collisional activation mass spectra of Pho(CHz)r,S- ions (Table 3.1) provide a constrast to the spectra of the corresponding phenoxyalkoxide ions. In these instances oxygen is attached to the aromatic ring and consequently the ipso carbons (C-1) would be subject to essentially the same inductive electron withdrawing effects. However the nucleophile approaching the n aromatic system is now S- rather than O-.

O.ly the 2-phenoxyethanethiolate PhOCH2CHzS- ion (Figure 3.15) can have undergone the Smiles rearrangement: in this case PhS- is observed but it is a very minor process. The PhS- ion is absent from the other spectra fie. of PhO(CH2)nS- when n = 3 ,4 ,S l. This is presumably because s- is a poorer nucleophile than o-. This is reflected in the gas phase acidity of ethanol (^Hoacid = 1580 t 10 kI mol-1)200 versus ethanethiol (ÂHo¿qi¿ = L487 + 12 kI mot-1)200. Chapter 3 185 l._i3

CH2CHS

59

Pho

93

PhS. 109

Figure 3.15 The collisional Activation Mass spectrum of pho(c$z)zs-, recorded on the VG ZAB 2HF instrument.

The characteristic fragmentations of Pho(CHz)ns- for n = r ,2 or 3

are the formation of Pho- [eg. see equations 3.17 and 3.18, Scheme 3.2 ]

and the loss of PhOH I eg. equation 3.19 ].

- phlo\,/S pho' + H2C=S Q.ln -.- oPh S Pho' + /) (CH)1n-r) (3.18) ( / ( (?")" -{> Pho- p,"",,,",)] S \ PhoH + 4rr^z)o S' (3.19)

Scheme 3.7 Chapter 3 186

3.3 Synthetic Studies directed toward identification of the 'Smiles' Intennediate.

It was necessary to determine the site of attachment of the nucleophile to the aromatic ring, in order to establish whether a Meisenheimer intermediate formed during the course of this rearrangement. To do this there were two synthetic approaches that appeared viable. These were the synthesis of a suitable molecule which would yield the anionic intermediate postulated for this rearrangement by a deprotonation reaction, or synthesis of 1-13C labelled phenoxyethanol and 1-13C labelled thiophenoxyethanol. The initial approach was directed toward formation of (L5L) and (152).

one strategy is outlined in.scheme 3.8. A short synthesis of 4- acetoxycyclohex-2-enone (L54) had been reported2S8. In principle, the desired compounds (151) and (152) may be formed by converting (154) to the acetal (155) followed by deacetylation.

o on o n o -#

ococH3 ococH3 154 L55 L52

Scheme 3.8

The ketone (15a) was formed in slightty poorer yield than reported in the literature. However, all attempts to form the acetal (155) failed. Reflux with ethylene glycol in a variety of solvents including Chapter 3 787 dichloromethane, chloroform, toluene and 1,2 dichloroethane and using methanesulphonic acid, p-toluenesulphonic acid or conc. sulphuric acid failed. Attempts to effect transacetalation using ethylene glycol acetone acetal also proved fruitless. In all of these reactions much of the starting material was converted to phenol. \

A milder route to the desired compounds was considered and is summarised in Scheme 3.9. This synthetic approach seemed more promising as the removal of the iron tricarbonyl moiety from (L57) could be effected using trimethylamine N-oxide at room temperature. Flowever, similar attempts at formation of the acetal described above failed to yield

(1s6).

o ft f-\ o o -r(* 3 3 L56 157 152

Scheme 3.9

The successful approach involved the preparation of 1-13C labelled phenoxyethanol (Scheme 3.10) and L-13C labelled thiophenoxyethanol (Scheme 3.11). These syntheses were slightty modified from the reported methods to ensure optimum yields of each product. The other modifications were minor and are recorded in the experimental section. Chapter 3 188

KC*N,18-Crown-6 ,t * CN CH3CN, reflux , 16 hr conc. HCI o 6hI.

BaC*O 3 * * Hotc- co2H 3050 - 3100C r, ., r,

"Mitsubishi" dehydrogenation catatyst 3900 - 4000c

OH OH o activated charcoal * l_\ 400-450C ,4g hr

Scheme 3.10 Chapter 3 189

CH3C*O2Na

cH3c*o2cH2cH3

BrMg r oC Ether , 0 OH 3 CH¡ 12 (several crystals) 140oC ,20 mins.

"Mitsubishi" dehydrogen ation catalyst 0c 3900- 400 * 2H CH¡ KtvInO4

HzO, reflux NN: I00Vo, IISO¿

NH2.HCl

1)NaNO2,HCl KSC(S)OCH2CH3 2)Znl conc. HCI SH O activated charcoal / \ gooc ,24hr. OH

Scheme 3.11.

Note Added In Proof: Theoretical Calculations (AM1 MO) have recently been reported2s9 which are consistent with the experimental results described herein. Cha ter 4.

The Benzilic A id And R lated Rqarrangements.

4.L Introduction.

In solution, ø-dicarbonyl compounds have been found to rearrange readily. The best known of this class of rearrangement is the benzilic acid rearrangementSS. Benzil (L58), on treatment with potassium hydroúde ion affords the potassium salt of benzilic u.i¿260 [(160), equation 4.L].

o o o gH OH _> OH 2 (4.1) Ph --O+ Ph L58 L59 160

The rate of rearrangemenÉh is dramatically affected by the presence of metal ions. The metal ions present are thought to chelate to the carbonyl group inducing further polarisation of the carbonyl group. In this way nucleophilic addition to the carbonyl group is enhanced to yield a transient complex (161).

Á M' X=O,NH,S Rl =H,alkyl :ñ:: R = R2 = H, alkyl, aryl,COrEt, CO2H 1.6L Chapter 4 191

Migration of R2 proceeds to yield the rearrangement product. Recently evidence has been found which suggests that at least in certain circumstances single electron transferSSb may play a determinative role in course of the reaction.

42 The Gas Phase Benzilic Acid Rearangement.

This gas phase study was initiated to investigate whether this class of base catalysed reaction would proceed in the absence of metal ions.

Preliminary attemps were made to observe the existence of an adduct from biacetyl (1,62) using HzN- and Ho- ions. It was immediately apparent that deprotonation had occurred ; no ions attributable to the formation of the adduct could be found.

162

An alternative strategy to investigate the benzilic acid rearrangement involved using a substrate which is not readily deprotonated. Benzil (15S) was introduced into the ion source of the ZAB in the presence of HO- at high pressure ( ca. 10-2 to 10-1 Torr ). Nucleophilic addition reactions have been found to occur more readily at higher pressure s267 . An adduct was observed but due to the low abundance of this ion it could not be identified in this instrument.

This result indicated that nucleophilic addition was less favourable than deprotonation. Attempts to increase the source pressure from 1.0-2 to Chapter 4 792

10-1 Torr in the hope of generating solvated hydroxide ions (HoH...-oH) which coul4 hydrogen bond to the carbonyl group (L63), thereby

enhancing the induced charge on the carbonyl carbon, or simply to remove excess energy on desolvatiorL failed.

o o $ il ..oH c-c + HOH....OH (4.2) Ph Ph Ph

1.59

When the estimated source pressure was raised to S x 10-1 Torr using the KRATOS MS 50 TA instrument an adduct was observed that could be identified by its collisional activation mass spectra. This spectrum (Figure a.1) is identical with the spectrum of the authentic deprotonated benzilic acid.

The major fragmentation involves decarboxylation to yield an ion of mass m/z = I83. The loss of CO2 cannot form directly from the simple adduct (159). Within the limits of experimental error the peak widths at half height of m/z = 183 ions formed from the authentic benzilic acid and

the adduct formed from Ho- addition were identical ( 33.1 + 0.5 volts ). This is consistent with rearrangement having taken place. The relative adundances of other ions were also the same. If rearrangement has indeed taken place then decarboxylation (equation 4.3) would yield deprotonated diphenylmethanol (164).

HO o 4 Ph,"lC-C\ + COz (4.3) Ph o- -.+ 160 764 Chapter 4 193

The strrrctu¡e of m/z = 183 was probed by comparing the collisional activation mass spectra (Figure 4.2) and the charge reversal mass spectra with the authentically prepared diphenylmethanol anion, and authentic decarboxylated benzilic acid (Tabte 4.1).

221 183 (-co,r

22ß çH)

,09 (-H2O) Phco2 tzl 153

Figure 4.1 Collisional Activation Mass Spectrum (CA MS/MS) of the [Benzil + - OH ] adduct recorded on the Kratos MS 50 TA instrument. This spectrum is identical to that of Ph2C(OH)COz-. Experimental conditions : ionizing energy 70eY , ion source temperature L50oC , accelerating voltage 8 kV. The reactant ion was -OH formed by ionising HzO. The measured source pressure was 4 x L0-5 Torr. Benzil was introduced using a direct probe (without heating) increasing the source pressure by 1 x 10-5 Torr to produce a total source pressure of approx. 5 x 10-5 Torr. Helium was introduced into the collision cell at a pressure of 2 x 10-6 Torr causing a reduction of 30% in the main beam. Chapter 4 794

tt' I 182 (-H')

105

Ph 7'l

153

ix 3)

Figure 4.2 collisional Activation Mass Spectrum (cA MS/MS/MS) for the m/z = 183 ion from the þenzil + -OHl adduct recorded on the Kratos

MS 50 T{ instrument. This spectrum is indistinguishable from the m/z = L83 ion formed from decarboxylation of Ph2C(OH)COz-. Chapter 4 195

Table 4.1 The Charge Reversal (cR) Mass spectra of the fti/z = 183 Ions

Precursor lon Spectrum a Relative Abundanc/

782 181 165 151 139 727 115

[(PhCoCOPh+-oH) - COz] MS/MS/ÀíS 2033563411977

Ph2C(OH)CC,2--CO2 NIS/MS/MS 24 38 67 35 11 10 11 Ph2CHOH - H+ MS/MS 15224fi309910

Precu¡sor Ion Spectrum a Relative Abundance b

105 89 n 63 51 39

[(PhCoCOPh+-oH) - Coz] MS/MS/lv[S 74 7 100 9 ?5 6

Ph2C(OH)COz=COz MS/MS/lv[S 76 6 100 9 22 6

PhzCHOH - tt+ MS/MS 81 10 100 12 24 5

a. The MS/MS/MS spectra were recorded on the Kratos MS 50 TA instrument. The MS/MS spectrum was recorded on the VG zAB 2HF instrument. b. The majority of peaks are broad: the peak maximum onty is recorded, relative to the base peak (100%)

The fragmentations of the ion (164) have been describedl9l; in brief they are summarised in Scheme 4.1 Chnpter 4 196

OH I

Pn'cipt, 164

o I + H' (4.4) Phrci Ph

o- CHO + Ph- (4.s) I + Phlci" Ph H CHO-H*) + CoHo @.6) 1.65

+ C t+¿/O

(4.7)

Scheme 4.1

The spectra of all m/z = 183 ions are remarkably similar (minor differences being due to the different instruments used for recording the spectra.). The ions formed on collisional activation are identical indicating that the benzilic rearrangement has occurred. The product arising from nucleophilic addition is formally the product of a r,2 rearrangement.

Attempts to observe this rearrangement were made on the substrates methyl glyoxal (L66), ter'!-buglglyoxal (L67) and phenyl gtyoxal (168). Chøpter 4 797

oo oo o o \L lt \il $ il ,C-C.. c C C-C H¡C H "rc. * H,C"'\, H Ph H CH, 166 L67 L68

A range of bases was chosen from HzN- (ÁHoacid of the conjugate acid = 1690 + 3 kJ mol-1¡263, e¡f-(1,696 t 0.4)264, MeO- (1593 + 9)200, EtO- (1580 t 10)200,i-PrO- (ISTZ t 10)200,!-BuO- (1568 t9)200,n-BuS- (1431 t 72)265. In no instance could an adduct be detected. Methyl glyoxat I cf. ÂHoacid(acetone)792 - Ls49 + 8 kJ mol-11 was however very readily deprotonated by all alkoxides and amide ion. If the fragmentations of the rearranged ions proceeded very readily in the source it is remotely conceivably that they will not be observed. For simplicity the rearrangement product derived from the addition of methoxide ion to methyl glyoxal (166) was examined. This ion, deprotonated methyl lactate was readily fo¡med and formed a strong ion in the source. The fragmentations of this complex are recorded in Table 4.2.

4.3 Reanangements of Deprotonated Alkyl Lactate Esters.

The decompositions of the deprotonated methyl lactate proved to be quite intriguing, especially the loss of the fragment of mass = 28 a.m.u. The neutral fragment tentatively assumed to be carbon monoxide (Co) could not have arisen from a simple bond cleavage process. Earlier work had shown that the base peak in the CA MS/MS spectrum of g|ycins266 involved a loss of CO (equation 4.8). Chapter 4 198

NH2CH2CO2'

NHCH2CO zH [1'*=."1 E-OH HNCH2OH + CO (4.8) -

A systematic study into the fragmentations of ø-hydroxycarboxylate esters was therefore undertaken. These data from the collisional activation

mass spectra are recorded in table 4.2.

Deprotonation of the o-hydroxycarboxylate esters with amide ion (NH2-) will yield two deprotonated species; the alkoxide (L69) and the enolate ion

(170).

R-!-CozMe R-CH-C:O2Me OH o- 769 170

The relative proportions of each ion carrnot be readily determined due to the lack of independently determined acidity values. A rough

qualitative estimate of the enolate acidity can be made if the change of a Me substituent for a Meo substituent on aceto¡¡¡i¡s267 of ÂHo¿çid = 13 kI mol-l is applied to methyl acetate. The acidity of these protons will

decrease-from ÅHoacid = L557 kJ mol-l b 1544 kJ mol-l. The acidity of a simple u1.o¡o1200 is ÂHo¿qid = 1580 kJ mole-l, however internal hydrogen bonding may alter this value significantly. Unfortunately no data is

available to allow this to be determined. Flowever, data (Table 4.2) for the methyl O-silyl phenyl lactate (1,71) and the methyl a-hydroxy phenylacetate (172) allow differentiation to be made between the enolate

and the alkoxide ions. Chapter 4 799

Table 4.2 Collisional Activation (cA) Mass Spectra of Deprotonated a - Hydroxycarboxylate Esters.

Parent lon Lossa Me' CO 13co MeoH HCo2Me Hl3cozMe

MeCH(OH)CO2 Me-H+ 15 100 4 20

MeCH(OH)1369t¡4"-1¡+ 13 100 3 12

EtCH(OH)CO2Me-H+ 4 100 2 39 iPrCH(OH)CO2Me-H+ 4 100 2 54

IBuCH(OH)CO2Me'H+ 2 100 1

PhCH(OH)CO2Me-H+ 10 100 52 8

PhCH(O-)CO2Meb 100 4c 3

Formationa MeOCO- Meo13cû Meû

MeCH(OH)CO2 Me-H+ 7 2

MeCH(OH)1369t¡4"-1¡+ 4 2

EtCH(OH)CO2Me-H+ 3 2

!PrCH(OH)CO2Me-H+ 2 1

IBUCH(OH)COzMe'H+ 1 1

PhCH(OH)CO2Me-H+ 0.5

PhCH(O-)CO2Msb 0.3

a Peak height listed as relative abundance with respect to the base peak (100%). b Prepared by an S¡2 (Si) reaction as shown in equation 4.9. c This is a minor and unexpected loss of methanol. Chnpter 4 200

(Me) 3SiO oHo a// Ph_CH Ph_CH.C.-oM"

171 172

PhCH(OSMe 3)CO2Me + NH i phCH(O')CO2Me + Me ,SiNHz (a.9) - Desilylation of (T7l) by an s¡2(si) reaction using amide ion (equation 4.9) leads specificalty an alkoxide ion (120) whereas deprotonation of (172) is expected to yield the alkoxide ion (170) and the enolate ion (169). In Table 4.2 the pronounced differences in the behaviour of these ions is summarised. In particular, the spectrum of (169) showed losses of Me'and MeoH and the formation of Meo- presumably arising from the enolals 1sn218. The fragmentations arising

from decompositions of the alkoxide ion are indicated in scheme 4.2.

*-1-.o2Me R-C-CO2' + Me' (4.10) -+ OH 169 OH

R Ç:Ç:Q + MeO' (4.11) HO

HO R ì* ] / C:C:O + MeOH (4.I2) 173 o

Scheme 4.2

Homolytic cleavage of the ester-alkyl C-O bond occurs to give the stablised radical anion (equation 4.10). The other fragmentations occur Chapter 4 207 through the ion-neutral complex (173). Dissociation of the complex (123) (equation 4.11) accounts for the formation of MeO-, while proton transfer prior to dissociation (equation 4.72) results in the loss of methanol. Stringer t¡ a1.278 had shown that these were major fragmentations of o- deprotonated esters. A minor contribution to the loss of methanol presumably occurs from the alkoxide ion by a cyclisation reaction

(equation 4.13). The alternative process (equation 4.14), involving a 7,2 proton 5¡i¡¡225 forming the enolate ion (169) seems unlikely because of the high activation energy of either a concerted process or a stepwise migration of a hydride ion to the nucleophilic oxygen.

o o Ph I \ Ç:e:e + MeOH (4.13) ^..] / 1{.,*"\") o

o oH o 9't___ !r" I /î Ph-CH-C-ON,, - Ph-ç-a_O* (4.t4)

The loss of CO (the base peak in all spectra) was shown by C-13 labelling ( Figure 4.3 ; Table 4.2) to proceed f¡om the ester carbonyl group. This loss (as with most others) is largely collision induced indicating that there is a substantial barrier to decomposition. A possible mechanism, given in scheme 4.3 proceeds via the intermediacy of an ion-complex (174). The possibility of any alternative intermediate leading to the products will be discussed later. This ion - neutral complex could fragment by a number of decomposition channels, namely simple dissocation (equation 4.15), deprotonation of the aldehyde by the alkoxycarbonyl anion (equation 4.1,6) or by a decarbonylation process

(equation 4.17) to yield an adduct (175). Chapter 4 202

75 (-rrco)

104

cH2cHo gg (-Me ) 43 McO GMcOrÐ 60 31 72

Figure 4.3 The Collisional Activation Mass Spectrum (CA MS/MS) of

Deprotonated MeCH(OH)l3ggtMe recorded on the VG ZAB ZIHF instrume¡rt using an electric sector scan. Applying a potential of +1,000 Volts to the collision cell gave the following results : I m/z (collision induced : unimolecular) 89 (50:50) ,75 (70:30) ,72 (70:30) , 60 (90:10) , 43 (80:20),31 (80:20) l Chaptn 4 203 o' I R-CH-CO2Me

,170

MeOCO' + RCHO (4.1s) o. r\ R + _ C-ON4e (R-H)CHO + HCO2Me (4.76) H

174 ' RCHO + MeO + Co $.77) 175

Scheme 4.3 The reaction described in equation 4.17 is analogous to the fragmentation of the hydroxycarbonyl ion. o'Hair t¡ a7.268 studied the decomposition of oxalic acid (Scheme 4.4). The anion decomposed via a I (COz) HOCO- I ion complex. The HOCO- ion fragments further within the complex by an a-cleavage yielding -OH and CO. Subsequent nucleophilic addition of HO- to CO2 produces the bicarbonate anion.

o o I I HO-C-C-O

o o I (co) Ho-c-o CO + HO-C-O

Scheme 4.4

The structure of the ionic species (175) from equation 4.17 could conceivably be either a hydrogen bonded adduct (L76) or the tetrahedral nucleophilic addition producf (177). When R = !-Bu and Ph there is no acidic cr-hydrogen to the carbonyl group. Hydrogen bonded adducts for these species seem unlikely. The adduct formed is likely to be the Chaptu 4 204 tetrahedral ion (177). The complex decompositions of the adducts support this (Table 4.3). The authentic ion PhCH(oCHs)o- was synthesised by an s¡2 reaction in the ion source of vG zAB 2HF (equation 4.18): its spectrum is listed in Table 4.3.

o tl o- N{eO HCR,C R-CH H Olvfe L76 tTl

- PhCH(OMe) 2 + MeO' -+ PhCH(OMe)O + MeOMe (4.18)

The CA mass spectra are the same, within the limits of experimental error . A hydrogen bonded adduct may form when R = Me, Et and i-Pr. The loss of CO from deprotonated methyl lactate is expected to yield the acetaldehyde enolate ion irrespective of whether the hydrogen bonded adduct or the nucleophilic adduct product was formed. The ion (177, R = CHg) was prepared independently by an S¡2 nucleophilic displacement with methoxide ion (equation 4.L9). The CA MS/MS and CR MS/MS spectra of this ion were very similar to, but not identical to those of the ion obtained by Co loss from deprotonated methyl lactate.

- MeCH(OMe) + MeO' MeCH(OMe)O + MeOMe (4.19) 2 --+ Chapter 4 205

Table 4.3 Collisional Activation (CA) and Charge Reversal (CR) Mass Spectral Data for the Product Ions from Deprotonated a-Hydroxyacetate Esters.

Precu¡sor [.oss Product Ion SpectrurrÉ rn/z m/z type tn/z (rel.abundance) MeCH(OH)CO2 Me-H+ co 75 cA 43(100) 1ß cR 59(3) , 44(33), 43(100) , 42(68) , 33(15) , 32(30), 31(89) ,29(55) , 27(29), 15(56) ,74(9) ,73(2)

MeCH(OMe)2 MeCH(OMe)O-ü c$d 43(100) 75 q¡¡d 59(2) , 44(38) , 43(100) , 42(72) 33(12) , 32(28) , 3r(92) , 2e(4e) 27(20),15(10)

EtCH(OH)CO2Me -H+ co 89 cA s7(100) 717 cR 58(42) , 57(81) , 56(22) , 4L(22) , 39(52) ,33(22) ,37(46) ,29(100) , 27(42), 15(14) , 74(6) fBUCH(OH)CO2Me CO TL7 cA 101(5), 87(100), 85(6),31(45) 745 cR 101(4) , f36(6), 85(4) ,77(8) , 69(71) , 58(10) ,57(t00), 55(15) , 43(16) , 4t(72) ,39(42) ,39(42) ,37(t7) , 2e(43),27(73)

PhCH(OH)CO2Me co 737 cA 121(5) , 107(100) , L05(76) ,7M(72) ,77(16),37(74) 165 cR 121(10) , 105(100) , 90(6) ,77(82) , æ(6) ,57(22), 50(18) ,39(6)

PhCH(OMe)2 PhCH(OMe)O-c cAd 121(11), 107(100), 705(74), 104 (9) ,77(t8\,31(18) JK cRd 121(1.4), 105(100) ,90(11) ,77(86) , t3-1 63(16) ,51(20) ,50(15) ,39(5)

a All spectra (MS/MS/MS) were recorded using a Kratos MS S0 TA mass spectrometer. b This ion was prepared by an S¡2 reaction in the source of the ZAB spectrometer (shown in equation 4.1.9). c This ion was prepared by an S¡2 reaction in the source of the ZAB spectrometer (shown in equation 4.18). d MS/MS spectrum using VG ZAB 2HF Mass Spectrometer. Chapter 4 206

It is possible that the minor differences observed are due to differences in instrumental parameters used to obtain MS/MS/MS and MS/MS spectra, with the two instruments, however one cannot exclude

the possibility that the presence of the hydrogen bonded adduct (tZ6) is the cause of the differences in the spectra. Given the magnitude of the differences in experimental data it must be concluded that the hydrogen

bonded adduct (176) must be, at most, a minor product.

The product of the decarbonylation reaction, namely the addition product (177), formally arises from a L,2 rearrangement. No condensed phase analogy can be found in this system, and this is in complete : constrast to the benzilic acid rearrangement already described.

Finally, although the initial aldehyde-alkoxyl carbonyr anion complex (774) (Scheme 4.3) appears to adequately account for the formation of the products, decompositions proceeding via the intermediacy of a radical-radical anion complex needs to be considered.

An alternative radical mechanism is given in Scheme 4.S.

o- o o' o' o I tr- I R-CH-C-OCH3 + R- CH a R-CHaa - C-OCH¡ -

o' I R-CH-OCH¡ + CO

Scheme 4.5 Chapter 4 207

Flowever, this mechanism appears unlikely as the spectra of the a- hydroxy carboxylate esters failed to show loss of the radical 'COR1. The benzaldehyde ketyl radical anion has already been observed. in the Wittig

rearrangement. It has a modest electron affinityl94 [ z0 kJ mol-1 (0.72 ev)] and the acetaldehyde ketyl radical anion194 is reported to be stightly more

stable having an electron affinity of 116 kI mol-1 (1.20 ev). Neither of the kefyl radical anions have been observed.

4.4 Reanangement Reactions of Deprotonated Pyruvates.

Louw and Kooymans269 have reported the thermal loss of carbon monoxide from alkyl pyruvates (an a-keto carboxylate ester). This decomposition was suggested to involved homolytic cleavage (equation 4.20) to yield an acyl (R-'C=o) acyl radical / alkoxyacyl radical (Ro-'C=o) pair, which then lost carbon monoxide (Co) from the alkoxyacyl moiety prior to recombination as shown by C-13 labelling.

oo oo o [il iltl tl 1 R-c-e-oRr R- c. a C- OR RC-ORI + CO (4.20) -+ ->

o o o o o I il I H2C:C. .c-oR1 + H2C-C-OR 1 + CO (4.21) oR1 Chapter 4 208

The intriguing possibility that such a 1,2 ndical rearrangement may show an analogous gas - phase anionic rearrangement (equation 4.zr) warranted consideration. The electron affinity of ketenez7} has yet to be experimentally determined but it must be very low (< 0.5 eV) as two mass spectrometric studies have shown that only fragment ions are formed. The collisional activation mass spectra (CA MS/MS) of deprotonated alkyl pyruvates are recorded in Table 4.4.

Consider the fragmentations of the deprotonated 13C labelled methyl pyruvate (illustrated in Figur e 4.4). Loss of 13Co is observed, but the loss is not as Pronounced as it was for deprotonated methyl lactate (see Figure 4.4). Interestingly, the loss of 13co is predominately unimolecularly. All the fragmentations shown in Figure 4.4 can be adequately rationalised using heterocyclic cleavages. The fragmentations of deprotonated methyl pyruvate are summarised in Schem e 4.6. The intermediacy of ion-molecule complex, (178) is used to explain the decomposition processes. Chapter 4 209

Table 4.4 Collisional Activation Mass Spectra of Deprotonated Alkyl Pyruvates (CH3COCOzR)

Parent Loss a

Compound H. CO 13CT) CHzCO (R-H) (R-D) CH3COCOzCHe 25 45 58 CFfuCa136O2CFI3 25 36 56 CHgCOCOzEt 30 38b 64 38b CH3COCO2CH2CD3 26 11 60 28 CH3COCOzCzDs 20 8 18 CHgCOCOziPT 85 7 100 c 100 c CHsCOCOztBu 3 4 85 CFbCOCozC(CDg)s 2 5 70

Formation a

RO- HC2O- C3H5O- C3D5û CH3COCO2CH3 T4 100 CH3Co136qcFIg 13 100 CH3COCO2Et 72 100 CH3C@O2CH2CD3 78 100 CH3COCOzCzDs 75 100 CH3C@O2!Pr 87 47 CH3COCO2IBu 100 10 T2 CH3COCO2C(CD3)3 100 I 8

a. Peak leights listed as relative abundance with respect to the base peak (100%) b. Composite peaþ comprised of a sharp gaussian peak superimposed on a dish - shaped peak c. Composite peak, comprised of a sharp gaussian peak superimposed on a dish - shaped peak (An approximate ratio of gaussian : dish - shaped peak is L00 :45 ) Chapter 4 270

t-tc= c-o 4l t02

60 (-H2C=C=O)

73 (-t3co)

McO 3l

Figure 4.4 The collisional Activation Mass spectrum (cA MS/MS) of Hzc-col3CozcHg recorded on the vG zAB 2HF instrtunent , using an electric sector scan. The relative ratio of ions undergoing unimolecular vs collision induced decompositions was obtained by applying a potential of

+L,000 Volts to the collision cell. Ím/z (collision induced : unimolecular) l 73 (20:80) ,60 (70:30) ,41 (50:50) ,31 (90:10) Chapter 4 2I7

o o II tl

cH2c-cocH J o ll NlkO -C + H2C=C=O (4.22)

o HC=C-O- + MeOCHO (4.23) (Hrc:c:o) il C-ORV (4.24) 178 N{eO +CO+H2C{=O

-otvfel [Hrc=c=o + +co (4.25) t79

Scheme 4.6

Decomposition of (L78) involves either dissociation (equation 4.22), deprotonation (equation 4.23) or c-cleavage of the methoxycarbonyl CÐ bond followed by dissociation (equation 4.24). In addition the methoxycarbonyl anion may act as an methoxide donor leading to the formation of either a hydrogen bonded adduct (180) or deprotonated methyl acetate formed by nucleophilic addition (t 81) (cl, equation 4.2S). The structure of the product ion (m/z = Z3) was investigated using both the vG zAB ZIJ'F and the KRATOS MS 50 TA instruments. The CA- and

CR mass spectra of the (m/z = 73) product ions are recorded in Table 4.5.

o II I I,feO'...H2C=C=O H2C-C\ ON{e L80 lE1 Chapter 4 212

Table 4.5 The Collisional Activation (CA) and Charge Reversal (CR) Mass Spectra of Product Ions in the Mass Spectra of Deprotonated Alkyl Pyruvates .(H2C-COCO2R) a

Precursor Ion Loss Spectrum (m/z) (m/z) TyPe m/z (abundance) -CH2CO2Me CH2CO CAMS/lvfS/MS 31 (100) (101) (se) CRMS/MS/MS 44(76), 30(22), 29(700), 28(67), 15(30) -CH2COCO2Me co4 CA MS/MS 58(7),41(100) ,31(2) (101) (73) CRlvfS/lvfs 59(7), 57(2), 45(8), 44(72), 42(700), 37(6), 29(29), 27(73), 75(78), 74(7), 13(3)

cob CAMs/lvls/lvfsc 41(100) (101) (73) CRÀ/tSlMS/N4Sc 42(60), 41 (100), 31(45), 29(30\, 27 (11) , t5(17),14(10)

-CH2CO2Me d CAMS/tvfS/MS 58(7),41(100) ,37(4) (n) CRN4S/lvrS/lv[S 59(8) , 57(2) ,45(10) , M(10) ,42(100) ,31(6) ,29(30) ,27(15) , 75(17), 14(8) ,13(3) -CHzCOCOztBu EtCOCHO CAMS/MS/MSC 56(100) ,4t(æ) (143) þn cRMs/lv[s/lvfsc 55(10) , 43(65) , 42(100) , 47(25) , 39(60) ,29(40) ,27(40) , 26(75) , 1s(1s)

CHzCOCFIs e CA lvls/lvfs s6(100), 55(12), 4t(29) (sz) 55(3) CRMS/lvfS ,54(1) ,53(3) ,42(100) ,4r(3r) , 40(12) , 39(30) , 38(14) , 37(72) , 29(36) , 28(13) ,27(U) ,26(24) ,25(7) , 15(14) ,14(22\ ,13(7\ ,72(2\ a. Decomposition data for the ion formed in the ion source. b. Decomposition data for the ion formed in the gas collision cell c. The spectra of the product ion are weak and represent an average of 100 scans. Peaks of low intensity ( < 1.0% of the base peak ) are lost in the background noise. d. Authentic -CH2COzMe formed by deprotonation of methyl acetate. e. taken from reference 191 Chapter 4 273

The (m/z = 73) ion formed in the ion source of the ZAB was compared with the mass spectra data for authentic deprotonated methyl acetate ion. Fragmentations characteristic of the enolate ion are loss of Me' to yield m/z = 58 ion, loss of MeO- (m/z = 3L) and formation of deprotonated ketene ion (m/z = 41) and these were observed in both spectra. The CR mass spectra were identical within the limits of experimental error. Decomposition in the ion source thus yields solely the 1,2 rearrangement product (181).

The product ion (m/z = 73 ) formed on collisional activation was also studied. The CA mass spectra of the product ion showed only the ions m/z = 4J.. Other ions may be lost in the noise or as the result of differences in the instrumental parameters; however, the charge reversal spectrum is not identical with that of -CH2CO2Me. The ions aL m/z = 4L

(HCO2+) and m/z = 3I , (CH3O+) are not formed from (181) (see Table a.5) ; almost certainly they arise from hydrogen bonded ion complex (L80). Thus we propose that a mixture of (LBO) and (181) are formed on collisional activation.

The data in Table 4.2 show that the ester alkyl group has a profound influence on the course of the fragmentation of the deprotonated pyruvates. In particular, (l) the loss of carbon monoxide and the formation of HC=CO- decrease rapidly in the order MeO- > EtO- > i-PrO- > t-BuO- whilst (il) the elimination of the alkene from the ester alkyl group increases in the order Et < i-Pr < t-Bu. The former reactions presumably reflect the stability of the alkoxy carbonyl anion which is expected to decline in the order of the basicity of the alkoxide (equation 4.24). The basicity of the alkoxycarbonyl anion and the alkoxide ion will diminish in Chapter 4 214 the same order and the formation of HC=C-O- is expected to decline correspondingly (see equation 4.23)

The elimination of an alkene has been observed previously for unsaturated ethersl99, enola¡ss218 and amides237. It is best illustrated for pyruvates in Figure 4.5 ( labelled ethyl pyruvate ) where the losses of Co and labelled ethene are readily observed.

88 (-QD¡)

120 HCrC-O 4l

92 (-co)

c2D5o o tr c 50 c2D50-

78

(x t0)

Figure 4.5 The Collisional Activation Mass Spectrum (CA MS/MS) of

C-(H)2COCO2CD2CD3 recorded on the VG ZAB 2Í{J' instrument, using an electric sector scan. The peak width of the m/z = 88 ion measured at half height is 15L + 2 Volts.

The loss of alkene results in the formation of a dish shaped peak, indicating that the process is accompanied by a substantial kinetic energy Chapter 4 275

..1"¿ts218 (le. the process has a large reverse activation energy). Unlike the earlier relrorts of internal elimination reactions2l8,237 which are thought to proceed via a 6 - membered transition state it seems likely that this

elimination proceeds through a seven centred state (equation 4.27). A less likely alternative involves formation of a t¡ansient spirocyclic ion (f8z) which may effect elimination via a 6-membered transition state. (equation

4.28).

o

+ IJ2C=CIJ2 ?t @.27) -

'- + IJ2C=CH2 e.28) H2

The probability of elimination will be related to the number of available B - hydrogen atoms and on this basis alone the observed order of elimination described above is readily comprehended. Chapter 4 276

8'l / ..n,\ \ "'t't'.",/

Bu'O 73

o I o rc!c-o- Hrð-c-cÍr, T 4t Bu'o-C- 51 l0t 98 ils (-CO)

Figure 4.6 The Collisional Activation Mass Spectrum (cA MS/MS) of

H2c-CoCo2c(CH3)3 recorded on the vG zAB 2IlF instrument, using an electric sector scan.

The spectrum of the t butyl ester (Figure 4.6) differs from the other pyruvate esters as an ion of nt/z = 57 is also observed. In the spectrum (Table 4.4) of the d9-deuterium labelled ester an ion of m/z = 62 is present, suggesting that an ion of empirical formula C3D5O- has been formed. The use of the KRATOS MS 50 TA instrument identifies m/z =57 to be the acetone enolate ion from its MS/MS/MS data (Table a.5). The formation of such an io_n demands methyl migrationz7l ¡o the tert - butyl group prior to fragmentation: a possible mechanism is shown in equa+ton 4.29.

oo o oo o [il il ilil il Et-C-C- + H3CCCH 3 EI-C-C-H + H2CCCH3 ø.29)

HsC CH¡ Chapter 4 277

In conclusiory unlike the thermal rearrangement of pyruvates (cr- keto carboxylate esters) there is no evidence that the gas phase rearrangement of pyruvates involves homolysis to yield an acyl radical anion / alkoxyacyl radical intermediate. Chapter 5.

eA ate eta an earra

5.L Introduction.

In Chapter 2 (section 2.10 ) the fragmentation of the altyl vinyl ether system was described. The possibility of the formation of a homoallylic anion (114) was advanced. It is dearly of considerable interest to establish whether or not this type of rearrangement can occur in the gas phase.

-{> \14 - 122 Rearrangements involving nucleophilic addition to a double bond, especially when a bridge of one atom separates the nucleophile from the double bond have long been observed in solution chemistry (equation 5.1).

Baldwin229 has classified the initial step as a 3-exo trigonal ring closure.

+ w Z *'\"2? { tYl (s.1) -> - - The 4-butenyl anion is the simplest example of this rearrange¡n¿¡¡¡Z72. It readily undergoes a degenerate rearrangement via a cyclopropylcarbinyl anion (equation 5.2). Chapter 5 279

:^- + =a..,/ (s.2) -+ - There is a growing body of evidence to suggest that homoenolization of ketones occurs via a similar processSg. A similar rearrangement has been reported for the phosphate system273 sþs\l/n below (equation 5.3).

(s.3)

--- Many other systems are thought to rearrange in this manner. An alternative mechanism has recently been advanced in which the intermediacy of a radical anion - radical pair has been postulated2T4, similar to the Wittig rearrangement (equation 5.4). No evidence for this mechanism was provided.

o o S o -s il I Ph,.. Ph... I Ph u_ P\ -+ I a c\. R åqå, Ph/ s-c\'åt P( R Ph

(s.4)

A particular example of this type of rearrangement has been reported in the gas phase. Deprotonated N-allyl N-hydroxy carbamv¡e:275 decomposes to yield the methyl carbonate anion ( MeOCO2- ) as the base peak of the spectrum (Figure 5.1). Chapter 5 220

o l ì() I I 5 C cHrO o

89 G'CH2CH=CH2)

59

.L

6t

3l 98 CMcoH¡

(x l0)

Figure 5.1 Collisional Activation Mass Spectrum of CH3OCON(O-) CH2CH=CH2 recorded on the VG ZAB zIfF Mass Spectrometer.

A relatively minor, but none the less significant loss of methanol is

consistent with the intermediacy of the cyclised ion (183) which may be

formed reversibly. The processes are rationalised in Scheme 5.L Chapter 5 227

o

MeO Å o-

o

^"çN'', L83

o lv{eO' Ä ÌlfeO o'N\^

o oo -O2CN=CH-CH=CHz MeOH + +. a lvfeO

Scheme 5.1

The acyloin rearrangementS4, also known as the cr-ketol rearrangement/ is a base catalysed reaction initiated by u deprotonation step. It is proposed that it may proceed via the intermediacy of a deprotonated hydroxy oxirane (Scheme 5.2) Chapter 5 222

o o o

*Å I o OH

o o o o R1 R1 RÅ Å R R R1 I o o o

Scheme 5.2

The rearrangement is often thermally reversible in the condensed phase and the position of the equilibrium is affected by subtle changes in reaction çen¿i¡iens84a. An example of the reversibility of the reaction is shown in equation 5.5.

o (s.5) OH occH tt 3 o cocH3 -

In this chapter, two gas phase systems will be examined which are closely related to the acyloin rearrangement. Lee , Chan un¿ 1ç*on276 have reported that under basic conditions in solution, acyloxyacetates rearrange to acyl hydroxyacetates ( Scheme 5.3 ). The intermediacy of a hydroxy oxirane anion (184) was postulated. Chapter 5 223

o 'o OH oRl oRl R R' o o o o - L84

Scheme 5.3

Both the precursors and the products are stable to heat. Indeed acetyl hydroxyacetate I CHsCozCH(oH)CozCHs ] can be distilled with ns thermal rearrangement having taken place. This system appears eminently suited to gas phase study since it should be possible to unambiguously demonstrate that any rearrangement that occurs is due to the anion and not the neutral precursor.

5.2 The Gas Phase Reanangement of Acyloxyacetates.

The collisional activation mass spectra (cA MS/MS) of deprotonated acyloxy acetates (i.85), together with the appropriate isotopically labelled compounds and selected isomeric acyl hydroxy acetates are recorded in Table 5.1.

o R2 oR3 R L85 o

Examination of the spectrum of methyl acetyloxyacetate (Lgs ;

R1=R3=CH3;R2=H) (see Table 5.1 and Figure s.2 for C-13 labelled r.85) reveals that this compound fragments by a variety of decomposition channels. Chapter 5 224

Table 5.1 The Collisional Activation Mass Spectra of Deprotoruted AryloxyAcetate Esters I RICO2C-(R2)CO2R3 ] and Deprotonâted Acyl Hydroxyacetate Esters IRICOR2(OH)CO2R3 - H+ ¡.

o Neutral lnccc Formation c à P¡ecursor n' H?/D CO Cñz R3OH R3oD HCOTR3 DcorR3 I McCIC2CHæO2M¿ 42(40.s) 1.5 100(34) s4{37) 2 2 McCC2ltH2CøM¿ Fig. 5.2. 1

3 CD3CO2CH2CO2Mcd r00 't9 77 a 4 McCOCH(OH)COZMc 1 43(39 s) 100(34) 46(3ó.5) 2 5 McCICH(OSiMq)C02Mc c 44 2 100 47 ) 6 MGCO2CH2CO2ET 6 58 2. 100 52 I 7 MæO2CH2CO2C2D5 4 52 4 94 100 I 8 McCO2CH2CO2jPr/ l8 46 4 62 100 2 0.5 2 9 McCO2CH2CO2¡Bu 11 23 188 100 89 a 3 t2 l0 E¡CO2CH2CO2E! 7 7l t4(3't) 9l(32.5) r00(34) l(32) I u E¡OOCH(OH)CøEr l8 E5 10(37) 8ó(32.5) 100(34) l(32.5) I 12 ¡rco2cHæøEt 15 8 69 ll 3l r00 6 l3 iPcøcH(oH)CøE¡ 178 100 6 26 u 0.5 7 t4 BUCOzCH2CO2ET 5 l8 3t 100 7 l5 iBuCû2CHæO2E. 8 50 6 87 100 9 l6 ¡B'¡CI}2CH2CO2Er 6 48 1 45 100 t4 t7 lBuCøCH2CøEr 210 100 u(37) 30(29.s) 4e(36) l l(33.5) It ¡Buco2cH2l3cøE¡ Fig. 5.3 l9 ÉuCìOCH(OttrcûZE¡ r00 32(37) 32(30) 24(36t 4 13(34) . 2 20 CSHr ¡COZCHZCÞZFr 244 22 39 100(38.5) 1 3s(38.s) l 6(38) 2l PbCO2CH2CO2ET 68(76.5) 52(5r) 6(37) r00(38.s) 2 36(39) 2 7 (3 8.5 22 PùCoCH(OH)CøEr 42(75.s) 44(50) 5(37.s) 100(34.5) ) 23 McCÞZCHMc)CØMc 6 l l(s0) 7l(38) 100(34) J 24 McCOC(lv{cxO$C02Mc 8 r0(5 l) 68(37.5) 100 25 McCQ2CH(Et)Cû2Mc 5 E4 88 26 McCO2CH(¡R)CO2Mc 4 2.5 100 76 /t 27 MæO2CH(Ph)CqMc 8 100 67 28 tBuCÛ2CH0r,lc)C02Me I t22 t2 100 2 I a . Compound number b. Neut¡al preo¡rsors were deprotonated by amide ion (NH2-) c. relative abundance (peak width at haff height, measu¡ed for an average of at least five scans to an accuracy of Í 1.0 Volts) d. CDaCOzCH-CO2Me was the only ion observed in the sou¡ce on deprotonation of the neutral with amide ion (NH2-) ¿. Ion formed by desilylation reaction to yield specifically MeCOCH(O-)CO2Me; see text equation 5.11 /. There was a minor loss of CHz=CH{H3 rel.abund.(9) g. Composite peak presumably the broad component being a loss of CH2{(CH3) The I¡. following losses were also observed:- loss (rel.abund.), CH2{=O (29) , Me. (a) Chapter 5 225

t32

100 CMeoH)

12(HCOzlr'.c¡ 104 cco)

cHrl3co, Hc-rrc_o- 60 76 GC2O2) Meo_ cHr¡rco 31 4244 l0)

(x l0)

Figure 5.2 Collisional Activation Mass Spectrum of

CH31aç92C-(H)COzCHa recorded on the VG ZAB 2HF Mass Spectrometer.

The site of initial deprotonation in the ion source was determined using the deuterium labelled compound CD3CO2CH2CO2CHs. OnIy an anion formed by loss of H+ was observed indicating that deprotonation occurred yiel

HCO2CH3). This result is not unexpected. While the gas phase acidities of the sites in this molecule are not known, an estimate of the relative acidities can be deduced. The acetyl position would be similar in acidity to methyl acetate2O0 ( = 1557 + 17 kI mol-l). The acidity of the ^Hoacid substituted acyl position can be inferred by comparing the acidity of acetone with CH3CO2 substituted acetone [ ÂHor.¡¿ (CIfuCOCH3)20O = 1550 Chapter 5 226 t 8 kI mol-1; ÁHo¿.i¿(CH3CO2CH2COCHT¡7e2 = L465 + 10 kJ mol-l l. Thus the difference in acidity of the two positions would be very Large; ie., of the order of 85 kJ mol-l.

Most of the decompositions cannot arise directly from the unrearranged ester enolate ion (L85). Three of the ions observed can be accounted for by direct decomposition. Another ion may arise by proton transfer. These decompositions are shown in Scheme 5.4.

MeCO,CHCOTMe + MeCO2 + !CHCO2Me (5.6)

r MeCO ' lvkO + 2CH=C=O þ.n [Ctut.co 2cH{{) Meo ] L86 MeCO 2CrC0 + MeOH (5.8)

MeCO 2CHCO2Me I cH2co2cH2corMe \87

+ 'OCH2COTM" - [1CH'=C=6¡ ] -+ HCIC-O + HOCH2CO2Me (5.9)

Scheme 5.4

The ester enolate ion may fragment directly, by an cr-cleavage to yield the acetate ion (equation 5.6). A transient methoxide ion - acetyloxy substituted ketene complex (186) may form, fragmenting by loss of methoxide ion via dissociation of the complex (equation 5.7) or deprotonation of the ketene by methoxide ion (equation 5.8). Finally proton transfer via a 5 - membered transition state may occur to form enolate ion (L87). This ion may decompose via an ion-ketene complex Chapter 5 227 which after proton transfer would account for the formation of the deprotonated-ketene (HC{-O-) as in equation 5.9. The processes described in equations 5.6 , 5.7 and 5.9 are minor and it will be shown later that the loss of methanol does not arise in the marner indicated in equation 5.8. It is interesting to note that the loss of the ester alkyl group by homolysis, a reaction typical of ester enolate ions does not occur in this instance (see equation 5.10).

+ MeCO 2CHCO2Me MeCO .ClFrCO2' + Me' (5.10)

A number of processes including the losses of carbon monoxide (CO) , alkyl formate (HCO2¡a) and dicarbon dioxide (also called ethylene dione i O=C=C=O) cannot be rationalised as proceeding from the unrearranged ion CH3CO2C-(H)CO2CH3.

The relative abundances of the fragment ions are very dramaticatly influenced by the degree of substitution of the acyl alkyl group (R1). [see Figures 5.2 and 5.3, and Table 5.1, (compound numbers 6,'J.0,I2,I7,2I)].

For example, the loss of carbon monoxide increases in the order Me

< Et = i-Pr - Ph < t-Bu. The loss of CzOzincreases dramatically in the order Me < g1 = i-Pr << Ph < !-Bu whereas the loss of methanol decreases dramatically in the order Me > Et > !-Pr = g-Bu > ph. By contrast, changes to the degree of subsifution of the ester alkyl group appear to have little effect on the relative abundances of any of the ions. (Table 5.1., compound numbers 1,6,8,9). Chapter 5 228

(-'tco) I 59 188

l31i-r2crr69r;

tBu CO 85

142 (-E¡OH) (-Hrrco2Er) ¡BuCO2 l0l ll3

EtO Erorllc' 45 '74

(x l0)

Figure 5.3 Collisional Activation Mass Spectrum of IBuCO2C-(H)13CO2Et recorded on the VG ZAB 2HF Mass Spectrometer.

The collisional activation mass spectra (CA MS/MS) of the deprotonated acyloxyacetate ions and of the possible rearrangement ions (the acyl hydroxy acetate ions) are very similar (Table 5.1). The widths at half height of the major peaks in the spectra are the same (within the limits of experimental error) for the isomeric pairs, although relative abundances are not always the same (Table 5.1). This result is consistent with the isomeric ions decomposing via a common intermediate to yietd the major product ions.

In the source formation of the ester enolate ion (1-88) may accompany the formation of the alkoxide ion (189), when R2 = H. Whilst the affect of hydrogen bonding is not readily determined, an estimate of the relative acidities of the ions may be made. Thus the acidity of the ester enolate ion would be similar to the CH3COzCH¿COCH319z ( ÂHo¿qi¿ = Chapter 5 229

1466 + 10 kJ mol-l ). Furthermore, the replacement of the hydrogen atom by a hydroxy-group may have little influence on the acidity of the position. For example26z, the acidity of MeOCHzCN (^Hoacid = 1556 t 15 kJ -o1-1) is very similar to C!fuCN (^Hoacid + = 1561 11kJ mol-1 ). Thus ÁHoacid ( t88 ; RTCOCH-(OH)CO2Rg ) 1,466 kJ mol-1 , whereas ( 189 ; = ^Hoacid RICoCH(oH)Co2Rs ) = 1450 kJ mol-1 ie. typical for a simple alkoxide ion.

o o o o

R oR3 R oR3 I OH"\ o R2

188 189

The spectra of (l) CH3CO2C-(CH3)CO2CH3 and the isomeric CH3COCH(CH3)(O-)CO2CH3 were compared as were the spectra of (li) CH3CO2CH-CO2CH3, I CHaCOCHOHCO2CH3 - H+ ] and authentic

CH3COCH(O-)COzCHe (formed by nucleophitic displacement of a trimethylsilyl group by amide ion, equation 5.11). oo oo

OCH3+NHr'-H¡ OCH3 + (CH,).SiNH2 (5.11) os(cH3)3 o_

The spectra obtained were very similar for the isomeric species and it will therefore be argued that the major fragmentations proceed from the alkoxide ion. (189) When R2 = H it is possible that the ion (1S8) undergoes a facile interconversion into (189). A I,2 hydride transfer by a concerted

Process is reported to symmetry forbidden; the barrier to interconversion being of high energy89,277 . A possible mechanism by which interconversion occurs is given in equation 5.12. Chapter 5 230

o o oo

R oR3 R 3 þ.72) I OH"\ [('t¿4" o H ll OH

Rl oR3 o-

The rearranged ion (189) may decompose to afford two ion- molecule complexes, namely an alkoxycarbonyl anion cr-dicarbonyl complex (1-90) and an acyl anion a-ketoester complex [(i.91), Scheme 5.5]. oo

R oR3 -o R2 189

o oo 3 il[ ¡_OR R,_L R2-c-c-oR 16,_u_E_ì L90 19L

Scheme 5.5 Chapter 5 23I

D Complex (L90).

The decompositions of the alkoxycarbonyl anion cr-dicarbonyl

complex (190) are surnmarised in Scheme 5.6. -

+ HCorR' 1s.rs¡ o il / 192 c-oR3 l{-'-rE-ù \ oo 190 Rr_C-C_R2 +R3o + Co (5.14)

193

Scheme 5.6 The reactions of the ion-complex (190) are expected to show some

similarities to the fragmentations of the alkyl hydroxyacetates ( see Chapter 4 ). Thus the alkoxycarbonyl anion deprotonates the a-dicarbonyl compound (equation 5.L3 ) or alternatively reacts as an alkoxide ion donor , eliminating carbon monoxide to form either a hydrogen bonded adduct or products arising from nucleophilic addition to a carbonyt group.

(equation 5.14 )

5.3.2 Loss of Alkyl Formate.

The loss of alkyl formate forms the base peak in most of the spectra. Identification of the product ion (192) has been possible for CH3CO2C-

(H)CO2CH3 ( Table 5.2 , m/z = 71. ,Figvre 5.4 ) and PhCO2C-(H)COzCzHs (Table 5.2; m/z = 133). Chapter 5 232 10 CH) oo ilil H2 CC CH

H2CCHO 43

HCE C-O 4l

o 69 (-H) il HC 45 29

Figure 5.4 Collisional Activation Mass Spectrum of H2C-COCHO recorded on the VG ZAB 2FIF Mass Spectrometer.

Table 5.2 Collisional Activation Mass Spectra (CA MS/MS/MS) of selected Product Ions in the Mass Spectra of Deprotonated Acyloxy Acetate Esters (R1CO2CH2CO2R3).

Product Spectrum Precursor Ion l¡ss Ion TyPe CA Specrum

(m/z \ (m/z ) m/z ( rclaÊve abundance )

MeCO2C-(H)CO2Me co 103 MS/MS/lvfs 7 5 (700),7 t(32),45 (72),43 (6) 7 L(L00),57 (86\,55 (37 ),43 (53),47 (24), MeOH 99 MS/MS/N4S 2e(12)

HCO2Me 714 MS/MS/NT^S 70(100),45(8),43(88),41 (37),29 (2) -CH2COCHO 77 MS/MS 70 ( 1 00),45 ( 9),43 (90),47 (3 6),29 (2)

PhCO2C-(H)CO2Et HCO2Et 133 lvß/MS/NrS 7ûs(rffi),77(28)

PhCOCHO - H+ 133 MS/MS 105(7ffi),77(32) t BuCO2C-(H)CO2Et CzØ 131 MS/MS/lv[S 1.01.(1û,87 (s3 ),8s(12),4s ( 1 00) Chapter 5 233

a. CR MS/MS/MS data for tn/z = 7I are m/z ( rel. abund. , loss ) : 56(14,Me' ), 54(76,HO'), 53(I2,HzO), 42(100,CHO' ), 29(83,CH2CO), 27(76,CO2),26(17,HCOz'),I5(7,C2O2),74(7,C2HO2') This has been accomplished by comparison of the MS/MS/MS spectral data , obtained on the Kratos MS 50 TA instrument , for the products ion m/z = 77 and m/z = 133 with the CA MS/MS spectra of the authentic methyl glyoxal anion (-CH2COCHO) and the phenyl glyoxal anion (PhCOCO-) respectively. The variation in the spectral data are small and probably reflect differences in instrumental parameters.

The use of a C-13 label has shown conclusively that the loss of carbon monoxide proceeds only from the alkoxycarbonyl anion (see Figures 5.2 and 5.3). The loss of carbon monoxide increases only slightly with increased substitution of R1 and is a very minor process for R2 = a1kyl, declining with increased substitution at p2 (Table 5.1, compound number t,23,25,26,27). It is quite conceivable that steric factors would adversely affect deprotonation by the alkoxycarbonyl anion, (c/ equation s.13).

The structure of the product ion (193) formed on loss of CO is not well defined. For the parent compound CH3CO2CH2CO2CH3 the possible products are the hydrogen bonded adduct (194) and the deprotonated hemiacetals [ (195) and (196) ].arising from nucleophilic addition to the carbonyl groups.

oo- o o I cH,å-f ...H H3C C-CH t ocH3 cH3o

L94 195 1'96 Chapter 5 234

The independent synthesis of suitable precursors to these ions has not been possible using our facilities. Predictions can be made regarding the probable fragmentations of these ions, however the results cannot be completely unambiguous. The cA MS/MS/MS data for the decarbonylated ion ( Table 5.2 , m/z = 103 ) indicates that carbon monoxide (Co) and methanol (CH3OH) losses occur readily. The formation of

CH3CO- and HCO2- ions also occu¡s.

Possible fragmentations of the hydrogen bonded adduct (194) are suggested in Scheme 5.7 : the major fragmentation is expected to be the loss of methanol (equation 5.15).

o I oo lvfeO .... cH3c- MeOH + 'CH2C-CH (s.15) 194 -

H2 HOMe + HOMe o o

oo H2C{. + MoOH + HCOi (5.16) t

- MeO + MeOH + HC r CH + HCOi þ.I7) o 1l

Scheme 5.7 Chapter 5 235

The formate anion may be formed by an a-cleavage analogous to

the decomposition of deprotonated methyl glyoxal ( equation 5.16) cr,- Cleavages are not common but do occur I see phenyl allyl ether (section 2.2.2) , and N-altyl N-hydroxy carbamate (scheme 5.1) l. An alternative mechanism could involve proton transfer from CH3oH to the cr-vinyl anion followed by a p-elimination induced by methoxide ion (equation

5.L7) | B elimination is observed during the decompositions of vinyl alkyl ethers (section 2.2.6)1. A rational explanation for the loss of carbon monoxide and the formation of CH3CO- cannot be made for the decompositions of the hydrogen bonded ion (194).

oo- oo taI 'CH2C-CHilll CH?C-C-\ .., f{ + MeOH( CH2COCHO)I -+ MeOH + (5.18) ocH3 195

o o o o il I I I cH3c HC-OCH cH3c + HC-OCH (s.1e) ) 5 \97 - o il CH¡C-H + OCH3 + CO (s.20) o o_ il ( il cH3c- c-ocH 3 ") o il 198 cH3c- H + +CO (5.21) 199

Scheme 5.8 Chapter 5 236

Postulated Éragmentations of the deprotonated hemiacetal (195) are summarised in Scheme 5.8. Loss of methanol may occtu readily (equation 5.18 ). An acyl anion-methyl formate complex (197) may be formed and dissociation of the complex could account for the formation of the acyl anion (equation 5.19). The acyl anion ( ÂHoacid CHsCI{O = 1633 t 8 kJ mol-1) is of simitar basicity to the formyl anion ( ÂHoacid IICozCH g = 163z t 19 kI mol-1 ). Therefore under the conditions of collisional activation, proton transfer may occur to produce a new ion - complex (L98). a- Cleavage could lead to the formation of the methoxide anion ( equation

5.20 ). This is, however, expected to be a very minor process as the Riveros reaction ( equation 5.2L) is favoured over simple dissociation of the ions because of the thermodynamics involved in forming the "adduct ion " (199). The adduct ions formed could have the deprotonated hemiacetal structure (200) or be simple hydrogen bonded ions (20L). The formation of the formate anion is not readily accounted for from this ion.

o' I CH"-CH,\ (cH3oH)'H2CCHO ocH3 200 201

Suggested fragmentations of hemiketal ion (196) are summarised in Scheme 5.9. The following fragmentations are possible; ie. the formation of Meo- (equation 5.22) and loss of CH3oH ( equation 5.23 ) , and loss of co ( equation 5.26 ). The formation of HCoz- ( equation 5.27 ) could only proceed by intermediates which appear to be highly unstable. Furthermore the loss of CO, presumably forming via the ion complex

(203) should also be accompanied by the loss of CH2o ( equations.2s ) -CHO [ ion is a very strong 6"ss256 ( ÂHoacid (CH2O) = 7647 t 3 kJ mol-1 ) and could readily deprotonate methyl ¿çs¡¿1s200 ( ÂHo¿sld (CI!¡COzCH3) = Chapter 5 237

7557 t 17 kI mol-l )1. This ion is not observed. The process outlined in equation 5.2*, involving initial formation of a methyl glyoxalate - methyl

anion complex (202) is likely to be a minor process. Thus it seems improbable that (196) is the product ion produced by loss of Co. perhaps the ion species (L94) and (1.95) are both formed; however the relative proportion of each is not known.

ooilil o o\ lvleO' + CH3C-CH (5.22) ililI N4eO' CH 3c-cHJ o o il lt MeOH + 'CH2C-CH (5.23)

'cH¡]* N'reo' + c2o2+ cHa (s.24) 'o o \ il It^--E-[l Hsc ...C-CH 202 t o cH3o ll o H2CC-OCH3 + CH2O (5.25) 196 ,) CH I(',.8-*' ,o' 203 H3C-CH + Co (s.26) ocH3

/o-C,H "'t'7\'t + H3C-C; \ + H3CCOMe + HCO 2- þ.27) MeO O- Olvfe

Scheme 5.9

5.3.3 Fragmentations of the Alkoxvcarbonvl anion a-Dicarbonyl Complex (L91)

- Chapter 5 238

Acetoxyacetate anions may decompose via the ion complex (19L)

The proposed decomposition pathways are shown in scheme 5.10.

oo o oo [il il ilil R1 H-C-C-OR R C . + H-C-C-OR 3 (s.28)

1.9L

oo o ilil il 3 'OR3 C-C-OR R C- H + + CrO2 |(-'-t-l - þ.29) 204 - 205 Scheme 5.10

The acyl anion - alkyl glyoxalate ester complex (19L) may dissociate (equation 5.28). Alternatively, transfer of the aldehydic proton to the acyl

anion leads to the formation of a new complex (204) which may then lose CzOz ( equation 5.29 ). The possible formation of neutral dicarbon dioxide (CzOz)is interesting since all attempts to generate this compot¡n¿278 ¡¿ys

, to date, been unsuccessful. Theoretical calcula¡i6n"279 predict that CzOz

should be unstable. Therefore this species, once formed, may decompose

in the gas phase to two molecules of CO.

The positions from which the carbon atoms are derived can be inferred from C-13 labelling. It is clear that the terminal acyl carbonyl group is not involved in the loss of CzOz ( Figure 5.2 ), but the ester carbonyl group is lost ( Figure 5.3 ). The loss of CzOzcannot be rationalised on the basis of the sequential loss of two carbon monoxide molecules. Had Chapter 5 239

this process occurred then the loss of CO should have been observed from

the C-13 labelled ion (I-BuCO2CH-13çO2EI; Figure 5.3 ).

The process of CzOzloss is only observed when R2=fl but becomes more favourable as the degree of substitution at R1 increases. [ie !-Bu > Ph > i-Pr = Et > Me ( Table 5.L ; compound numbers IZ,2!,12,I0, 6 ) l

A plausible explanation for this order is found if one considers the presence of an acidic hydrogen on the aryl/ alkyl glyoxal. The deprotonated glyoxalate ester is tikely to be a very strong base; the gas- phase acidity probably very similar to the alkoxycarbonyl anion ( ÂHoacid

IICOzCHg = 1638 + 19 kJ -o1-1¡280. The acidity of the protons q to the carbonyl group would be expected to be similar to those of ketones (ÂHoacid acetone2O0 = 1549 t s kJ mot-1). clearly the alkoxycarbonyl anion can readily deprotonate the glyoxal. Steric and statistical ( ie the number of acidic hydrogens decreases from Me > Et > lPr ) factors could presumably affect the deprotonation to a minor extent. When R1 = ph or tcr.!-butyl the protons present are considerably less acidic and as a consequence the relative abundance of the fragment ion formed by loss of CzOz increases dramatically. The aryl-!{ and CIIO protons on benzaldehyde are of almost equal acidity ( ÂHoacid = 1626 t 26 kl mol-1 )201. However this process is less favourable than removal of the cr-proton of a carbonyl group

(equation 5.30 ).

CHO Ç:Q ooilil c-c-oR3 and l(er-'*) /or

ooilll + H-C-C-OR3 (s.30) Chapter 5 240

The hydrogen atoms on trimethylacetaldehyde (R1 = t-Bu ) are only weakly acidic. Exact values are not known, however Nibbering has shown that this compound may be deprotonated using amide ion to form a homoenolate ion ( equation 5.31 ).

H o H¡C, o H3 (s.31) H¡C H3 - It is therefore unlikely that the alkoxycarbonyl anion will deprotonate trimethylacetaldehyde. The affect of the substituent R1 on the extent of C2O2loss is thus the result of a competition between the processes indicated in equations 5.28 and S.29.

The structu¡e of the adduct (205) was probed for t-Buco2CHzcozEt (see Table 5.2 , m/z =131 ). Sensitivity problems using the Kratos MS 50 TA instrument did not permit other systems to be examined. The fragmentations of the ion are best accounted for by assuming an S¡r¡2 addition product has formed. The postulated fragmentations are indicated in Scheme 5.11. Chapter 5 241

o

ocH2cH )"]

H"C o- HsCa' zCHt H¡C H H¡C o 205 H3 + Hz + H2C=CH2 þ.32) H¡C o

I H3 o 'oEt H3 + (5.33) H¡C

206

HsC. o- + HOEI (5.34)

H.C - Hr¿).ç-cH2o + CHTCHO HsC (s.3s)

Scheme 5.1.1

Fragmentation of the s¡2 adduct (20s) may occur via two ion complexes. Formation of a hydride ion solvated ethyl trimethylacetate complex would result in the loss of "C2FI6" possibly by an elimination reaction (equation 5.32). Alternatively, an ethoxide ion trimethylacetaldehyde complex (206) could form, fragmenting to yietd

ethoxide ion (equation 5.33), deprotonating the weakly acidic B-methyl grouPs (equation 5.34) or undergoing hydride ion transfer to afford the

neopentanoxide ion (equation 5.35). Chapter 5 242

when R1 = CHe, Et and i-pr , as well as the formation of the s¡2

addition product, the hydrogen bonded adduct will presumably also be

formed. The driving force for the loss of CzOzmay be the formation of the stable adduct (205) as is observed for the Riveros reaction. When R2 = alkyl, the ion complex (204) cannot form.

Most of the major decompositions have arisen from the deprotonated acyl hydroxyacetate ion (189). In view of this the mode of formation of the carboxylate ion ( RtCo2- ), the alkoxide ion R3o- and the loss of R3OH must be reconsidered.

The loss of the carboxylate ion RICOz- is observed from both ions. The peak width at half height was measu¡ed and is the same for both ions suggesting that uPon collisional activation interconversion of (L89) to (207) may occur ( equation 5.36 ). This is a very minor peak and unfortunately the structure of the neutral cannot be probed. Increasing the steric bulk at R2 might be expected to the favour the reversion to (202) however there is no evidence for the toss R1Co2- increasing as R2 is increased.

o o R2

oR3 (s.36) R1 oR3 R o R2 ---l> o L89 207

The loss of R3oH is observed for the acyloxyacetates and the isomeric acyl hydroxyacetates (Table 5.1.). The relative abundances of the ions produced by this major loss are similar , and for a selected pair of isomeric ions the peak widths at half height are the same (within the Chapter 5 243

Iimits of experimental error). This indicates the intermediacy of a common decomposition channel. The anions (199) and (207) may be interconverting on collisional activation, possibly decomposing via (207) as described earlier (equation 5.8). Indeed an examination of the proportion of ions having undergone unimolecular vs collision induced decomposition shows that an appreciable proportion of the rearranged ions decomPose without having undergone collisional activation. An alternative mechanism, involving a transient hydroxy oxirane intermediate (cf. chapter 4) is advanced in equation 5.37. oo

CH 3 ocH3 cHs ocH3 o H

189

o I + HOCH3 þ.37) o 208

The possible structure of this process has been investigated for the [cHeCozc-(Fr)co2cH3 - MeoH] ion. The CA MS/MS/MS data for this ion is recorded in ( Table 5.2 , m/z = 99 ). The decompositions of the ion (20g) are rationalised in Scheme 5.12 and are consistent with the fragmentations recorded in Table 5.2. Deprotonated ketene ( HC{-O- ) may be formed by proton transfer within a transiently formed ketene - ion complex prior to dissociation of the complex (equation 5.38). Alternatively, ion (208) may undergo an internal nucleophilic displacement to yietd the cyclopropanone carboxylate ion (209). Decarboxylation is quite feasible (equation 5.39). Proton transfer may also proceed yietding the carboxylic Chapter 5 244

o I H2C{=O

o I CCH"- Io ".="-"] 208 o [þ*i,)

o I + HC=C-O- (s.38) 2

o o + COz (s.3e)

CHCO 2 209

^ o '"r'o c co2H 4cora + co (s.40) 2H 21.0 ^ o il (Hc:c-co2H) -cH 2H

HC=C-CO| + CH2O (s.41)

Scheme 5.12 Chapter 5 245

acid enolate ion (21-0). The highly strained ring may open and either

undergoes an c-cleavage to lose co ( equation 5.40 ) or lose cH2o as indicated in equation 5.4i..

The probable fragmentations of the ion described in equation 5.8 (ie. CH3co2C:Co-) are sununarised in Scheme 5.13. Not atl of the fragments observed in the spectra can be readily accounted for by decompositions of this ion.

o o il I - cH3c-o-c=c-o CH3C-O - + C{=O (5.42) 211,

o I o CH3C-O-ç-C:O O{=C=O + CH3å - (s.43)

-+ o -il cH2c-o\ + (H2C=C=O) o / Ç:Ç:Q H H

HC=C-O' + o (s.M) H

Scheme 5.13

A simple heterolytic cleavage directly from (211) would account for the loss of C2O2. (equation 5.43) Alternatively an intramolecular proton transfer followed by a simple heterolytic cleavage would yield a transient ion - ketene complex. Deprotonation of ketene prior to dissociation would Chapter 5 246

presumably occur readily (equation 5.M). In fact the deprotonated acetoxy

ketene is expected to afford CH3COz- by a-cleavage (equation 5.42) which is not observed. On this basis, the loss of methanol is expected to arise via the reaction path outlined in equation 5.37. In the absence of the

formation of the ion by an independent route this assignment must be assumed to be tentative.

Finally, it should be noted that fragmentations typical of ester enolate ions I namely losses of R3. and (R3-H) ] are minor in these spectra. In the case of CH3Co2CH-CO2iBu ( Figure S.5 ) no loss of tBu' is observed and the loss of CH2=Ç(çH3)2 is a minor fragmentation.

lll (-tBuOH) oo 99 ilil H2CC-CH 7t

(-c2o2')

(-CH2=C(CHr) ô)

(-co)

145 cHrco2 ¡BuO' lr7 I],CCHO 73

HC- C.O (-Hz) 4J 56 r29 t7l 4t 55 ss l3t rÂ^ tx l0)

Figure 5.5 Collisional Activation Mass Spectrum of Deprotonated cH3co2cH2co2tBu recorded on the vG zAB 2HF Mass spectrometer.

The evidence Provided above clearly supports the contention that anionic heterolytic cleavages lead to the observed decompositions. None Chapter 5 247

the less, the intermediacy of radical intermediates cannot be rigorously

excltided. It could be argued that perhaps the loss of Czoz may arise by a homolytic (diradical) fragmentation since the stability of CzOz is calculated to depend upon its spin multipticity. A possible mechanism for its

formation is indicated in Scheme 5.14. oo

R oR3 -o H

I o ö o _C-C-C-ORlt I lt. a )"]

oo o oo ilil ----)t----> 1_c_I il[ .c-c-oR)l R H + .c-c-oR3 (s.45)

o- ooilil o- n1-å¡"'on + aaC-C Pt-ç'.-O*, * CzOz H

Scheme 5.14

This mechanism seems unlikely however since a transient ion - molecule complex (212) would be expected to form. Dissociation of the complex (equation 5.45) would yield the ketyl radical anion. These species are not observed. The electron affinities of the benzaldehydelga [70 kJ mol- 1(0.72 eV)l and acetaldehldgle+ [116 k] tt'to1-1(1.20 eV)l ketyl radical anions Chapter 5 248

are positive and indeed the benzaldehyde radical anion is observed in the collisional activation mass spectra of benzyr arkyl ethers

5.4 The Gas Phase Acvloxyacetonitrile / Acyl Hydroxyacetonitrile Rearrangement.

The acyloxyacetate/acylhydroxyacetate rearrangement is best described as Proceeding via anionic intermediates. There appears to be no conclusive evidence for the intermediacy of radical anions. In principle , the acyloxyacetonitrile anion may undergo a similar rearrangement, possibly via an hydroxyoxirane intermediate (equation 5.46).

o o' CN I oo-ilt RCOCHCN RC-CH-CN (s.46) RV.o H

However, in this case (unlike the earlier example), rearrangement via a acyl radical-cyanoformaldehyde radical anion complex formed by homolytic cleavage may be more favourable. (equation S.4Z)

o o oo- [- il ilt RC-OCHCN RC ocHCN RC-CH-CN (s.47) - o I il RC . - OCHCN

Dawson and Nibbering2Sl have shown that the deprotonated methoxyacetonitrile anion fragments readily by loss of a methyl radical. (equation 5.48) Chapter 5 249

. C.A. H3C-OCHCN > . OCHCN (s.48) 213

The stability of this radical anion (213) has been attributed to the capto-dativs sffsç1282.

-aa--a o-cH-c-\ +-+ o_cH-c=N .__+ O_CH=C:N 213

It is therefore desirable to consider the acyloxyacetonitrile system. The collisional activation mass spectra of several deprotonated. acyloxyacetonitriles are recorded in Table 5.3.

consider the prototypical example acetoxyacetonitrile

[(cHgcozcHzcN; R = CFk ) ( Figure 5.6 , Table 5.3 )]. Deproronation with amide ion (NH2-) would be expected to yield two ions, oiz. the ester enolate ion (214) and the a-acetonitrile ion (215).

o o -l¡ cH2cocH2cN cH3cocHCN 2L4 2\5

The gas phase acidity tables allow us to predict that both positions would have very similar ÁHo ¿çid values I eg.ÂHo acid (CFI3COzCH3¡zoo - 1557 + 17 kJ mol-l ; ÂHoacid (CHeOCIIzCw¡zez = 1556 + j.5 kJ mol-11. Thus one might expect almost equal proportions of both ions to form in the ion source. Determination of the sites of initial deprotonation in the source was made by measuring the relative abundances of the dedeuterated to deprotonated anions from trideuteroacetoxyacetonitrile ( CD3CO2CH2CN Chapter 5 250

R = CDg ). A ratio of -CD2CO2CH2CN : CD3COzC-(H)CN of 100 : 40 is obtained. A deuterium isotope effect is perhaps likely to disfavour formation of the ester enolate ion (2L4). This unexpected result possibly reflects the greater propensity of the ion (21s) to undergo source decompositions and will be discussed later in this section.

Table 5.3 Collisional Activation Mass Spectra of Deprotonated Acyloxyacetonitrile ( RCozCH2CN - H+ ) Ions and Labelled Analogues.

Initial lon L.oss

H. HCN DCN CH2o cHDo CH2Co CD2Co

CFIgCØCHzCN-H+¿ 1 1,6 82 100

CD2-CO2CH2CN 23 77 100

CDsCO2C-(H)CN 10 L2 63 r.00

ECO2CH2CN-H+ 29 1.00

PTCO2CH2CN-H+ 26 100

PhCH2CO2CH2CN-f¡+ 20 13 t8

lnitial Ion Lcs Formation

MeCHCO ETCHCO phCHCO HOCH2CN cN-

CFIgCO2CHzCN - H+ 22 -CD2CO2CH2CN u

CD3Co2C-(H)CN 20

EICO2CH2CN-H+ 22 1

PTCO2CH2CN-H+ 33 2

PhCH2CO2CH2CN-H+ 5 100 0.5 a. The ion -OCH'CN is observed ( rel.abund. 9o/"). Chapter 5 251

56 (-H2c=c=o) 98

68 (-CHzCo)

CN-

26

-OCH.CN ]t (-HCg

55 (-H') 97

Figure 5.6 Collisional Activation Mass Spectrum of Deprotonated cH3co2cH2CN recorded on the vG zAB 2HF Mass spectrometer.

The CA MS/MS spectra of both the deuterated acetoxyacetonitrile anions (Table 5.3) showed the same decomposition channels were available to each iory indicating that equilibration of the ions by H+ or D+ transfer (as aþpropriate) preceded fragmentation. The fragmentations of deprotonated acetoxyacetonitrile are indicated in scheme 5.i.5. Chapter 5 252

o CH2CO + OCH2CN (5.49) Hr¿< -ocHrcv] o [(c*rrco) I cH2cN 2.t4 [cHrco + cN' I + cH2o (s.sO) 2't6

o o o o ö I ? I H3C-C-OCHCN -o+ H3C-C-CH-CN .--{> H3C-C. HC-CN (5.53) 2t5 217

I o o I I t( H3C-C-CH ).,-

o il CN- + MeCCHO (5.51) CH2COCHO + HCN (5.52)

Scheme 5.15

Firstly, it is clear that not all of the major fragmentations occrlr from the ion (215). The ester enolate ion (21a) decomposes via an ion complex. One decomposition channel involves dissociation to lose ketene (equation 5.49). The deprotonated formaldehyde cyanohydrin is clearly a very weak

base* as the deprotonated ketene ion [(HC=C-O'),m/z = 4U is present in <7'/o relative abundance. The alternative mode of decomposition involves transfer of the cyanide ion (-CN) to ketene (equation s.sO). The structu¡e of the adduct formed (216) could be either the hydrogen bonded ion (218) or the product of S¡2 addition to the carbonyl group (219). cA

* The gas pbase acidity of ketene ÅHo¿si¿(CHz=C{) = 1528 + 11 kJ mol-l Chapter 5 253

MS/MS/MS and cR MS/MS/MS data for the product ion ( m/z = 6g ) are recorded in Table 5.4

o [Nc- lHrc-r-=o)] H 218 2t9

Table 5.4 Collisional Activation and Charge Reversal Mass Spectra of Selected Product Ions in the CA MS/MS of Deprotonated. Acetoxyacetonifrile ( CH3CO2CH2CN ).

Product Loss CA MS/MS/MS Ion [m/z , (loss) , relative abundance ] (m/z ) 7t HCN 70(H' )100, 5(C2H 2)I0, 43(CO)S2, 41(CHz0)36, 2e(CH2Co)z

68 CH2O 4 1 (HCN)S, 26(CF{2CO) 1 00

Product Loss CR MS/MS/MS Ion lm/z , (loss) , relative abundance ] (m/z ) 7t HCN 56(Me')78, 54(HO')16, 53(H2O)10, 42(CHO')100, 29 (CH2CO)89, 27 (CO)14, 26(IJCO2' )19, L5(C2O )9, 74(C2HO2')I2 . 68 CH2O 5 4 ( C H)78, 52(0)36, s 1(HO )62, 42(CN) 100, a0(H2CN /CO)84, 39(CHO')24, 38(CH2O)16,

2 9 (C2HN) 1 26 2, (CH2CO)20, 1 a (CzNO ) 12, 73 (C2HNO' ) 2, 12(C2H2NO)1

This product was compared with cA MS/MS and cR MS/MS of deprotonated acetyl cyanide and within the limits of experimental error was identical to the product formed in equation 5.50. Chapter 5 254

A 7,2 anionic rearrangement of ion (215) yietds íon (217). This ion may decompose via an ion complex which may dissociate to form -CN (equation 5.51) or lose HCN to yield deprotonated methyl gtyoxal (equation 5.52).

The structure of the ion formed by loss of HCN ( m/z = 7r ) has been confirmed by comparison of CA MS/MS/MS spectrum of the m/z = z1 ion (Table 5.4) with (the cA MS/MS specrrum of ) authenric -CH2coCHo. (Figure 5.6 , Table 5.4) The loss of HCN however appears to be counter to the relative acidities of the neutral species. Methyl glyoxal will be of comparable acidity to acetone ( ÂHoacid = L549 t S kI mol-l )200 The cyanide ion is a very weak base [ ÂHoacid (HCN)= L47O + 8 k] mol-l1200. A plausible explanation for this may be fotmd by examining the ion-induced dipole energy of the ion-molecule complex formed after the endothermic proton transfer283. Metþl glyoxal has a high dipole moment which may be even larger after a proton transfer. This could compensate for the endothermicity of = 70 kI mol-1 loss arising from basicity differences.

The compounds R = Et , Pr and PhcHz, yield spectra consistent with the decompositions already described. When R = PhCH2 deprotonation at the benzylic position is very favourable and the cleavage of the ester enolate ion is dominant. In contrast to the cleavages when R= alkyl , deprotonated ketene is produced ( equation s.il ), forming the base peak in the spectrum.

PhcHCO2CH2CN

PhCH=C=O ( 'OCH ,cNl --+ PhC=C-O' + HOCH2CN (s.s4) Chapter 5 255

When R = t-Bu , a deprotonated ion is not observed but when R = Ph , the CA MS/MS spectrum shows that an a-cleavage reaction occurs exclusively. (equation 5.55).

a a PhCO2CHCN PhCO2- + a (s.ss) a CHCN -

Finally the spectra of deprotonated acetoxyacetonitrile only shows only a weak radical anion (213). This may have arisen directly from deprotonated glycolonitrile by hydrogen atom loss on collisional activation (equation 5.56) or from the radic al / radical anion complex (equation 5.53). Loss of H' from deprotonated glycolonitrile is a facile process#. This fact, combined with the failure to observe the formation of the radical anion from benzoyloxyacetonitrile ( R = Ph ) indicates that the radical / ndical anion mechanism is unlikely.

C.A a o-cH 2cN O-CHCN + H (5.s6)

-

# The CA MS/I\4S of deproûonaæd glycolonitrile (HOCH2CN - H+) is a.s follows - I nr'z(loss) relative abundance I 55GI)100 ,54{Hz)14 ,26(CHzO)35 Cha er 6.

6.7 Future Dire ons.

Several types of ¡earrangements, observed in solution, have been examined in the gas phase and where it has been possible the reactivity of anions formed in solution have been compared with the "intrinsic,, reactivity of the ions in the gas phase.

In the course of this work it has become increasingly apparent that reactions in the gas phase do not always proceed in the same way that they do in solution. The simple answer would be to say that the results reflect the " intrinsic " reactivity of the anions. Is it possible to predict those reactions which are likely to proceed in the gas - phase on the basis of these results?

The range of anionic rearrangements studied allows some tentative predictions to be made. Reactions which are intermolecular will not be observed. Indeed a preliminary investigation rearrangement of deprotonated phenyl alkyl s¡þs¡s284 (equation 6.1) shown to occur intermolecularly in solution, did not occur in the gas phasea.

a The NICI MIKE sPecha lm/z (rel.abundance: loss/formation)l are : 2-MeC5H4OMe -H+ 120(72,-H' ), 119(10,-H2),106(100,-CH3'), g2(2,-"8t "), I7(76,-CHZO), 77(ph-,2): (expected rearr¿rngement ion) 2-EtC5FIaOH-H+, 120(94,-H'), 1,7g('1,4,-H2), 106(100,-cH3.), 93(5,C2 Ha), 92(6,-"Et'"), 77(pl] ,76) ; 2-MeCe H4oEI-H+ 133(7,-H2), 106(100,-"Et. "), 91.('t ,- MeCHO), 77(7,Ph-); 2-MeC5 H4Oípr-H+ 134(1,-Me. ), 133(1,_CH4 ), 107(3,CaHs), 106(100,_ "iPr' "), 1 - 9 (1,-MdOMe), 7 7 (7,Ph-),57( 1,MeCOCH2 ) . Chopter 6 257

OR o-

CHz 2R (6.1) -+

The categories studied, based on mechanisms postulated for solution chemistry and for which gas - phase studies have been conductedb involve:

anion - radical or anion - neutral complex.

6.2.L The Wittig rearrangement.

Alkyl benzyl ethers undergo rearrangement exclusively. Unlike solution studies elimination of an alkene was not observed. Ailyl alkyl ethers decomposed by elimination exclusively, whereas alkyl vinyl ethers undergo rearrangement to a limited extent but the dominant fragmentation route is via an elimination ¡eaction.

Further work may be directed toward the investigation of mixed acetals of benzaldehyde. It may be possible to examine the deprotonated ioru to determine the relative migratory aptitude of the alkyl groups. This may Provide evidence to support either the radical anion - radical complex or the anion - benzaldehyde complex.

b The mechanism for the transformation in solution may be significantly different from ttnt in the gas - phase. Unfortunately this information is Áetdom well defined in solution. Chapter 6 258

An alternative target could involve investigation of the thio - Wittig rearrangemsn¡l24,l2S. This rearrangement is influenced by the counter ion in solution. Unfortunately a preliminary investigation into this system, in particular of deprotonated benzyl ethyl thioether (Figure 6.1) and 1-phenyl thiopropanot (Figure 6.2), indicated that the wittig rearrangement did not occur. It is (remotely) possible that the deprotonated thioether may have undergone a sommelet type rearrangement (see p. 48)to afford deprotonated2 - ethyl benzylmercaptan.

An ion of this type is expected to decompose by loss of a methyl radical to yield the thioquinonemethide radical anion (cf. p.9S).

6.2.2 Other Rearrangements.

The Beckmann and Lossen rearrangements have been observed. If these rearrangements proceed in the same manner as in solution then the reactive intermediate is a nitrene - anion complex. An investigation to consider the effect of the basicity of the leaving group is necessary.

6.3 Internal nucleophilic addition of an ion to a suitable n system.

This rearrangement proceeds in the gas phase without activation of the aromatic ring by electron withdrawing groups present in the ring, as required in solution. It may useful to examine this rearrangement further to investigate the likelihood of attack at the ortho position28S when suitable leaving groups are present (Scheme 6.1). t5l Chapter 6 259

r22 ( "F,1"')

123 (-CHr=ç¡1r;

PhcHi 9l 149 (-Hz) 59 ól

(x l0) Figure 6.1 The Collisional Activation M¿ss Spectrtr.rr of Deprotoru&d Benzyl ethyl thioether. n2 \_,,,.t..) 15¡ I 150(-H )

123(-CHr=ç¡1t¡ PhCH; r9 (-Hz) 33 13 9l

(xl0) Figurê 6.2 The Collisional Activation Mass Spectrum of Deprotonated Phenylpropane thiol. Chapter 6 260

o- o(cHt3o - + Lr L

L +

L LH+ H

Scheme 6.1.

This type of cyclisation has been observed in solution, but it requires very forcing conditions. A pretiminary investigation indicates that this occurs (Figure 6.3). In addition substituent solvation-assisted resonance (SSAR) have been postulated to affect the o- values in solution. "¡¡..¡s286 These effects will be absent in the gas phase and a plot of the amount of rearran$ement vs o- would indicate if an anomalous value for o- exists in solution. It would also allow quantification of this affect to be made. Chapter 6 r83 267 t6o -

OMe 15l (-cH3oH)

123

OMe

125

93 108

Figure 6.3 The Collisional Activation Mass Spectrum of Deprotonated 2- Methoxy Phenoxypropanol.

6.3.2 Miscellaneous.

N-Allyl N-Hydroxycarbamate and acyloxy acetates also undergo learrangement. The results obtained are consistent with nucleophilic addition to the carbonyl group. Acyloxyacetonitriles may rearrange via radical intermediates. At present there appears to be no obvious way to investigate tlUs problem. Chapter 6 262

In the systems sfudied, deprotonation occurs more readily than the required nucleophilic addition in the gas phase. The rearrangement proceeds for benzil, a molecule that does not possess acidic protons but nucleophilic addition does not occur for several related systems. This result is interesting because chelation of the carbonyl gtoups to a metal ion promotes rearrangement in the condensed phase. The cannizarro reaction is strongly influenced by the nature of the counter ion but none the less it seems highly likely that this reaction may also proceed in the gas phase under favourable circumstances. Further work in this area is warranted to investigate conditions in which nucleophilic addition may dominate over deprotonation. systems in which the metal ion is not important are indicated in Scheme 6.2287 ¿^¿ 5.32s8. Chapter 6 263

o o o

HO OH

o o

OH

Scheme 6.2

Rl0 \ R2o' / C:N Rl o OSO2AT OSO2AT -+ -

Rl0 R20 tc:N + C:N /\ I R1 o oR2 Rlo OR

Scheme 6.3

The Háller - Bauer cleavage, a reaction of non- enolisable ketones

(equation 6.2) may occur in the gas phase. Nibbering20r noted during an investigation into benzaldehyde that amide ion would add to the carbonyl group. This would constitute an area of further study. Chapter 6 264

NHz

NH2 o' (6.2) o NHz

6.5 Solvoll¡sis reactions.

7-Acetoxynorbornadiene and 7-hydroxynorbornadiene react almost instantly in methanolic bicarbonate at 25oC to yield methyl tropyl s1þs¡289

(equation 6.3). In the gas phase deprotonation occurs in preference to rearrange¡¡1sn¡29O. It is likely that when solvation of the transition state is important these reactions will be less favourable in the gas phase. É

OAc

- H lvfeO (6.3) Olvfe

Reactions in which solvation may be very important, eg. pinacol and Favorskii rearrangements could be considered.

6.6 Miscellaneous reactions.

In the course of the study of deprotonated altyl alkyl ethers199 the acetone enolate ion was identified amongst the products (equation 6.4). Curiously, the Precursor of this ion was identified to be deprotonated cyclopropanol. In solution, ring opening yields the propionaldehyde (equation 6.5). The reasons for this divergent behaviour is unclear. It would be interesting to examine the affect of replacing the methine Chapter 6 265

hydrogen with methyl or phenyl to obtain further mechanistic information. -

o'

cH3cH2cHo (6.4)

o o lo 1--|t- (6.s)

Ê Collisional activation mass spectra and charge reversal mass

spectr¿292 were recorded on vacuum Generators LAB-2]HF mass

spectrom sls¡293 operating in the negative chemical ionisation ¡¡16¿s294. All slits were fully open to obtain maximum sensititivity and to minimise energy-resolution effects. The chemical ionisation slit was used in the ion source; ionising energy 70 eY (tungsten filament), ion source temperafure

L50oc, accelerating voltage 8 kV. Liquids were int¡oduced through the septum inlet at 100oC; solids through the direct probe at between 50oC and 100oc. Carbanions were generated by H* abstraction by Ho- (or H- or o- ) or H2N- or D+ abstraction as appropriate by Do- (or D- or o-). Reactant negative ions were generated from either Hzo or D2o by 70 ev electrons.

The indicated sou¡ce gauge pressure (of H2o or D2o) was typically 5x 10-a

Torr. The substrate pressure was typically 5x10-7 Torr. The estimated. total

Pressure within the source is L0-1 Torr. The pressu¡e of He in the second. collision cell was 2 x t0-7 Torr, measured by an ion gauge situated between the electric sector and the second collision cell. This produced a decrease in the main beam signal of ca.10% and thus corresponds to essentially single collision conditions. Consccu(ivc collision induccd dissociation spcctra ( MS/ MS/MS) wcrc mcasurcd with a Kratos MS 50-TA instrumcnt dcscribcd prcvi- ously.e In a typical cxpcrimcnt, a spccifìc fragmcnt ion M, formcd in thc MS/MS spcctrum of M¡ was transmittcd to thc thi¡d ficld frcc rcgion by sctting bo(h ESA-I and thc magnaic s¡ctor to (Mr/M¡) of thc normal sctting uscd to focus M¡. Mass analysls of fragmcnt ions rcsulting from collision induccd dccompcition in thc third ficld frcc rcgion was accom- plishc/ by scanning ESA-2. Ncutral substratcs wcrc dcprotonatcd by McO- (from McONOro) in a Kratos Mark lV chcmical ionization soure: ion so{.¡rcc tcmpcraturc 100 oC. clcct¡on crrcrgy 280 cV. c¡nission cuÍcnt 50O ¡Â ¿nd accclcrating toltagc -8 kV. lìquids wcrc introduccd through an all-glass hc¿tcd inla sys(crn at 100 "C. Thc indicatcd sourcc prcssurc o[ c¿ch substratc was 2 x lO-J Torr and of mcthyl nitritc t x l0{ Torr giving an cstimatcd sotircc prcssurc of ld Torr- Thc indicatcd pr&urc of Hc in thc collisioo cclls was 2 x t0{ Torr. producing a dccrc¿sc in thc main bcam signals of 30ft. (9) Burisky. D. J: Cooks. R. G-: Cher E- K-: Crcs. M- L- Aml. Chcm- 1982- Jl. 295. Græ.. M. L: Chcr E- K.¡ Lyon. P. A-: Crcw. F. W.: Equ. S.: Tudgc. H. Int. J. Mut Spcctrom. lon Phys. t982- 12.24]. (tO) Ridgc. D. P.: Bauch:mp, J. L. l. Am. Chcm. Sæ. 1974. 9ó. 1595 Experimental 267

General.

Melting points were measured using a Kofler hot-stage melting point apparafus under a Reichert microscope and are uncorrected. Infrared spectra were recorded on a Jasco IRA - 1 or Hitachi 270 - 30 spectrophotometer . Analytical mass spectra were recorded on a AEI MS -

3010 spectrometer. 1¡¡ runr spectra were recorded on a varian T-60. Unless otherwise stated, runr spectra were recorded as dilute solutions in deuterochloroform using tetramethylsilane as an internal standard. 13C runr spectra were recorded on a Bruker WP - 80 spectrometer. They were determined in deuterochloroform using tetramethylsilane as an internal reference. !j

Radial chromatography was carried out on a Chromatotron 79247 (Harrison Researcþ Palo Alto/TC Researclu Norwich), using Merck silica gel 60 PFzs+. Alternatively, purification was effected using flash chromatography295.

All solvents were purified by standard procedures. Light petroleum refers to the fraction of bp. 55-65oC. Unless otherr,r¡ise stated all organic solutions were dried with anhydrous magnesium sulphate.

Alkyl Lithium reagents in ether solution were accurately titrated296 using sec BuOH and Z,2'biqr:jnoline prior to use. Experimental 268

Isotopically Labelled Reagents.

Reagent Isotopic Purity Commercial Supplier

Dzo D > 98o/" Australian Institute of

Nuclear Science and Engineering

Hz18O t&o ß.zu YEDA Stable Isotopes CD3GIO D>98% Aldrich

LiAlDa D > 98"/" Aldrich

NaBDa D > 98"/" Aldrich CH313ggr*" 73c=g7o/o Cambridge Isotope Laboratories C

Bal3COs 13ç = 69.3% British Oxygen Corporation K13CN 13C=97Vo Cambridge Isotope Laboratories cD3oD D > 99.8% CEA (France)

CD3I D > 99.5% Aldrich C2D5OD D>99% CEA (France) (CD3)3CoD D>98% Aldrich

Chapter 2.

3, 4 - DihydropyÍan, dibenzyl ether and 1 - hydroxy -L,2, 3, 4 - tetrahydronaphthalene were commercial products. Experimental 269

Diphenylmethyl phen)¡l ether This was prepared by heating dry potassium phenolate with diphenylmethyl bromide in anhydrous tetrahydrofuran according to the literature ¡¡g¡¡e¿297. The yield on recrystallisation from ethanol was 48"/o, mp. 52.5-54oC, (lit. mp.2ez 52.5-53oC)

Allyl phenl¡l ether.

A mixture of phenol, allyl bromide and potassium carbonate were used to prepare the g¡hs¡298. It was obtained as a colourless tiquid bp. 33 - 5oÇ/20mm.Hg. (lit. Up.zra 85oC/19mm.Hg.) in86% yield.

I2EISI Ptrenyt aUyt eUrer. !

[2Hs] Phenyl allyl ether was made from C6D5OD (zlHo 98%; Stauffer Chem. Co.) by a reported procedure2es in86% yield (2Hs> 99%), bp.83 -soc / 20mm. Hg. o-Allyphenol.

Thermal rearranges¡gn¡299 of phenyl allyl ether was effected by reflux over 6 hr. The crude product was carefully washed according to the reported procedure and distilled to give a colourless liquid, bp. 102 - soc / 20mm. Hg. (lit. bp.2es 103 - loS.SoC / t9mm.Hg.), ytetdZ3%.

1 -Phenylprop-2-en-1 -ol.

Acrolein was added slowly to a cold (<-15oC) ethereal solution of phenyl magnesiumbromide30O. Y = 68o/o,bp.76 -77.soc / 0.65mm. Hg. (lit. bp.aot 72 - 3oC / 0.5mm. Hg.) Experimmtal 270

t -(tzEISl Ptrenyt )p r o p-Z -e

1-([2Hs]Phenyl)prop-2-en-1-ol was made from C6D5MgBr and acrolein as described above3ffi, yield 62"/o,bp.76 -77.5oC / 0.65mm. Hg. (2HS > g9%).

L -Phen$ [1 -2Hil pro p-2-en-ol. 1-Phenyl[1-2H1]prop-2-en-ol was made from phenyl-magnesium bromide and [1-2H1]acrolein by a standard method30O, yield 85/o, (2lH199%). t1- 2H1]Acrolein was prepared by Sarett oxidationSo2 6¡ 9, I0 - ethano[13- 2H2lanthracen-11-yl-methatro1303,followed by pyrolysis using a woods

Metal bath3Oa in an overall yield 60"/", (2H1= 99%).

1 -Phenvl [2-2Hil prop-2-en-ol C 1-Phenyl[2-2H1]prop-2-en-ol was made from phenyl-magnesium bromide and [2JH1]acrolein303,304 by the reported method30O, yield 58"/", (2Hr = ee%).

All)¡1 benzyl ether. This was by addition of benzyl bromide to sodium atlyloxide in dry dimethylformamide according to the reported procedure3Os. Y = 62"/",bp. 123 - 5oC / 40mm. Hg.(lit. bp.30s 204 - soc)

L -Phenylbut-3-en-L -ol.

This was prepared by addition of benzaldehyde to an ethereal solution of allylmagnesium bromide according to the reported method306. Reduced pressure distillation gave a colourless liquid, bp.74 - 6oC / 0.1mm. Hg. (tit. bp.306 71oc / 0.75mm.Hg.) ,yierd69"/". Experimental 277

1-Phenylbut-2-en-3-ol.

This was prepared by the addition of phenyl magnesium bromide to 4- chlorobut-2-en-1-ol according to the method of Delaby307. Distillation afforded the product as a colourless liquid, bp. ßa - 7oC / 18mm Hg. (lit. bp.307 725 - 6oC / I2mm. Hg.), yie\d46%.

4-Phenvlbutanal.

4-Phenylbutanol, prepared by addition of 2-phenylethyl magnesium bromide in diethyl ether to ethylene oxide308 (Y = 77%, lit. bp.ao+ 14goç/74mm. Hg.), was oxidised using ch¡omium trioxide / pyridine in - dichloromgfþ¿ns3O2. ! = 79%,bp.1j.5 - 7oC / t4mm. Hg. (lit. bp.30e 720 - zoc / 16mm. Hg.) Ë

3- (o-Tolyl)nropanal.

This was prepared in two steps. The addition of ethylene oxide to a cø. 1M solution of g-tolyl magnesium bromide in diethyl ether310 at < SoC followed by worked rp in the usual way using saturated ammonium ciloride and distillation provided the alcohol as a colourless liquid. Y = 78% bp. 1.34 - 5.5oC / 15mm. Hg. (lit. bp.31o 136oC / 15mm. Hg.) The alcohol was oxidised using chromium trioxide / pyridine in dichloromsfþans3O2. The aldehyde was obtained on distillation at reduced pressure as a colourless liquid. Y = 87"/",bp.48 - 50oC / 0.05 mm. Hg. (lit. bp.31r 49 - sooc / o.oz mm. Hg.)

1 -(o-tolyl)prop-2-en-1-ol.

This was prepared3Tz 6t the addition of a solution of vinyl magnesium bromide in tetrahydrofuran to o - tolualdehyde. Y = 96"/o,bp.73.5 - 75oC / 0.08mm. Hg. (lit. bp.37z 65 -73oC / 0.05mm.Hg.) Experimental 272

1 .4-Dihydronaphthalene.

Reduction oj naphthalene3l3 using sodium in ethanol afforded the desired compound. Y= 96"/",bp.94.5 -96oC / 18mm. Hg.( lit. bp.ste 96oC / 18mm. Hg.)

1 .2-Dihydronaphthalene.

This was prepared by u base catalysed isomerisa¡i6¡314 (5% sodium methofde at 140-50oC) of 7,4 - dihydronaphthalene. Y =92"/o,bp.86oC / 13mm. Hg. (lit. bp.31+ 86oC / 13mm. Hg.)

3-Methvlindene.

Indanone was treated with methyl magnesium iodide in diethyl ether. C The alcohol obtained was purified by distillation Y = 56"/o, bp. 120 -zoc / 23mm. Hg. (lit. bp.31s 12I - zoc / 25mm. HS. ) and dehydrated using 10% -aqueous sulphuric acid according to the reported metho¿315. Y = 88"/",bp. 9'J. -2oC / r6mm. Hg. (tit. bp.31s 20S - 6oC)

L-Benzvlallene.

This was prepared by the method of Brandsm¿316 ssing benzyl magnesiumbromide and propynyl methyl ether. Y -- 69"/",bp.72 - 74oC / 11mm. Hg. (lit. bp.377 74 -74.5 oC / 11mm. Hg.) -

1-Methoxy-1 .2.3.4-tetrahydronaphthalene.

An ethereal solution of L, 2, 3, 4 - tetrahydro - l. - naphthol was treated with an excess of diazomethane in the presence of the Lewis acid catalyst boron trifluo¡ide etherate according to the literature method319. Y = 98"/", bp. 113.5 - 114.5oC / 11,.5 mm. Hg. (tit. Up.ata f.4-I14.5oC / 1.1mm.Hg. ) Experimental 273 o-Allybenzyl alcohol. Benzyl alcohol in pentane was treated with 2.5 equivalents of n-butyl lithium in pentane at reflux for 6h¡. The product was monoalkylated using allyl bromide according to the reported metho¿319. The product was impure and was purified by preparative gas chromatography (pye unicam L04) using a 5% SE 30 on Ch¡omasorb A(40-60) 2.4 m x 7 mm column; column temperature L60oC, flow rate 60 cm3 min-l, retention time 6.45 mins. The compound is not particularly stable and its 1H n.m.r. signals, when freshy prepared, are 2.42 (1H, s), 3.46 (2I{, m),4.65 (2F{, m),5.13 (2H, br s),5.66 - 6.42 (1 H, m), and7.28 (4H,br s) (Found: M+', 148.0891. CroFIrzO requires M, 148.0888).

1-(o-Tolyl)allene. 1-(o-Tolyl)allene was prepared in 59% yield, by the addition of o-tolyl magnesium bromide to propynyl methyl ether in diethyl ether at -L5oC using the method reported for pheny¿11gns316. It is a liquid, bp. 82.5 - 84oC

/ 16mm. Hg. It is not particularly stable32O but when freshly prepared shows the following 1g n.m.r. spectrum: 2.33 (3H, s), 5.08 (2H, m),. 5.48 (1H, m), and 7.12 (4H, br s) (Found: M+' 130.0783 CroHro ; requires: M+' 130.0783).

The compounds Ph(PhCH2)CHOH, Ph(PhCHz)13CHOH, ph(phCDÐCHOH, and Ph(C6D5CD2)CHOH were available from a previous study321.

Similarly, Ph3COH, Ph313ç9" and Ph2(C6D5)COH were available from a previous study322. Experimental 274

U4,f¿Hal pnenvt )a-coFI. This was prepared by treating benzophenone with Í 2,4,6 - 2Hs l phenylmagnesium bromide in diethyl s¡¡s¡322. ! = 76o/", mp. 161 - ZoC, (lit. mp.322 t6O - 3oC)

Iæ,ÉZLIsl phenylb romidc. Aniline hydrochloride was converted into Í2,4,6-2Hsl aniline hydrocNoride by several exchanges (5 exchanges) with deuterium oxide according to the method of Best323. The diazonium salt was prepared and then treated with a potassium bromide solution to produce phenylbr6¡¡i61s.324 Y = 64o/o, bp. 155.5 - 156.5oC (lit. Upaz+. 156oC) [ 2Hs > 92% ; z¡¡Z> 7"/" l Ê

Base-catalvsed Rearrangement of Allyl Benzyl Ether.

To a solution of di-isopropylamine Q.azg.) in anhydrous tetrahydrofuran (50 cm3¡ maintained at -78oC was added a-butyl-lithium (1.40M in hexane; 23.t cm3 ) dropwise. After stirring at -78o C for 15 min, hexamethylphosphoramide (6.059.) was added, followed by attyl benzyt ether (a.60g) in anyhydrous tetrahydrofuran (20 cm3) The dark red solution was allowed to warm to 20oC and stirred for 2 hr. at that temperature. The mixture was then poured into aqueous ammonium chloride (saturated; 300 cm3), the organic layer separated and extracted with diethyl ether (5 x 40 cm3) the combined ethereal extracts were washed with aqueous hydrogen chloride (2M, 30 cm3) and water (3 x 30 cm3), and dried (Na2SOa). After removal of the solvent, the residue was distilled in vacuo to yield th¡ee discrete fractions. The first fraction, bp. range 76 - 94oC at 10mm. Hg., was l,2-dihydronaphthalene (L.17 g.), identified by comparison with authentic material3l4 using i.r., 1H n.m.r., and g.c. - m.s. The second fraction, bp. range 74 - 92oc / 2 mm. Hg., was 1-(o-totyl)prop -2- Experimental 275 en-1-ol (0.39 g) identified by comparison with an authentic specimen3l26, i.r., 1g n.m.r, and g.c.-m.s. The thi¡d fraction, bp. range 94 - gSoC / 2 mm. Hg., 1-phenylbut-3-en-ol G.a6 Ð was identical with authentic ¡¡¡¿¡s¡i¿1306, as above.

Vinvl Ethers. All vinyl ethers were prepared by heating a mixture of ethyl- or n-butyl vinyl ether and the appropriate alcohol in the presence of the catalyst -mercuric acetate according to the procedure of Watanabe and çe¡en325. The yields and boiling pt. ranges are recorded in the table below.

Experimental Data for Vinyl Ethers. È

Vinyl Ether Yield (%) Boiling Pt. oC /mmHg. Reference actual literature

Me 76 7t-2 11 326

Et 74 u-6 36-7 326

CD3CH2 71 34-6 36-7 326

CH3CD 69 u-6 36-7 326 n-Pr 8L 53-5 54-6 325 i-Pr 79 55.5-7 55{ 327 l-Bu 63 76-8 77-8 325

CH2C(CH3)3 87 89-9L 88-90 325

Ph 76 156-9 158{0 329

Altyl 59 67-8 66-7 328

CD2CH{H2 47 67-8 66-7 328 PhCH2 72 w7 /t+ 1034/25 330 Experimental 276

Divinyl Ether.

This was prepared by adding 2-chloroethylether to potassium tert-butoxide

according to the reported procedureSst. y=g{o/o, Bp. 3TgoC (lit. bp.33t 39oC).

Neopentyl Methyl Ketone.

This is prepared by a chromic acid oxidation of diisobutyleng332 over 10

days at room temp. in32o/" yield, bp.72 - 6oC (lit. Up.aaz I22 - 6oC)

The Trimethvlsilvl Vinyl Alkyl Ethers. General procedure.

!-Butyl-lithium (1.70 mol dm-3 in pentane, 10cm3) was added to a sti¡red solution of alkyl vinyl ether (20 mmot) in anhydrous tetrahydrofuran (60 c

.*3) at -78oc under nitrogen. The mixture was allowed to warm to -2soC and maintained at that temperature for 2 h¡. The mixtu¡e was then cooled to '78oC, trimethylsilyl ctrloride (2.359.) was added and the mixture was stirred for 3 hr. The mixture was allowed to warm to 20oC, filtered under anhydrous nitrogen gas and the filtrate distilled through a 30 cm. column of glass helices. Alt these compounds are readily hydrolysed and must be kept uner a N2 atmosphere. [N.B.: Particular care must be taken to avoid the admission of oxygen and water, especially during the silylation of methyl- and ethyl vinyl ethers as distillation to remove hexamethyldisiloxane (bp. 101 - 3oC) is very tedious.l. Experimental data for these compounds are recorded in the table below. Experimental 277

Experimental Data for cr-TrimethylsilvlVin)¡l Atkyl Ethers.

Compound Yield Boiling Pt 1H nmr [ôH(60MHz)] (oClmmHg)

CH2=ç15iMe3)OMe 91 103 -4 / 760 0.10 (9FI,s), 3.51(3H,s), 4.29(IHIJ=2.0Il2), 4.60(7HdJ=2.0Il2)

CH2=ç15¡¡4e3)OEt 84 737 -8 / 760 0.L0(9H,s), 1.14(3H,t,l=7 .}F{z), 3.57(2H,qJ=7.ÍIz),

4.74(1HdJ=1.8H.2), 4.42(L}JdJ=t.9FIz) CH2=ç15iMe3)OBun 76 97-9/73 0.08(9H,s), 1.05-1.90(7H,m),È

3.83(2f1,tJ=7.0}{2), 4.22(IHdJ=2.0tl2), 4.55(lHdJ=2.0F{2)

CH2=ç15iMq)OCH2But 77 62-4 / 48 0.14(9H,s), 0.96(9H,s\,

3.28 (2IJ,s), 4.24(LH d J =2.0Il2), 4.56(TH¿J=2.0IJ2\

All of the c- trimethylsilyl vinyl alkyl ethers were hydrolysed readily and are unstable in air. They were not submitted for elemental analysis. Accurate mass measurement (positive ion) was only possible for

CH2=ç1SiMe3)OMe (M+' = 130.0809; C6HlaSiO requires 1.30.08L0) as the molecular ion M+' and (M - Me')+ ions from other compounds were absent. Experimental 278

Pent-4-enal. This was prepared by " thermal Claisen rearrangement of allyl vinyl s¡þs¡328, effected by passing the aforementioned ether through a quartz tube heated to 255oC. Redistillation of the product afforded a colourless liquid Y=94"/", Block bp. 105oC (lit. Up.eza 103 - 104oC / 749m*.Hg.)

GåzH¿)bt¿-enat (2,2-2HùPent-4-enal was prepared by exchange of pent-4-enal with deuterium oxide by u standard procedure333. Y = 64"/o, bp. 103 - 4oC

ç2t12>ost"¡

ç

3. 6-Dihydropyran. This was prepared by the method of Collonge3S+, using 4-butenol, formaldehyde and hydrogen chloride.Y=56"/o, bp. 86-8oC (lit. Up.aa+ 86oC)

2 { I¿cloprop ylethanal. This was prepared by the base catalysed rearrangement of 3,6-dihydropyran according to the method of Rautens¡¿ush335 and purified by preparative gas chromatography (Pye Unicam 104) using a 5"/" SE 30 on Ch¡omasorb

A(40-60) 2.4m x 7 mm column; column temperature L25oC, flow rate 60 cm3 min-l, retention time 5.55 mins. and redistilled from a T-tube in a sublimation block (Block temp. 115 - 120oC ; lit. bp.336 112oç / 760mm. Hg.) 1H nmr õ(CHCI3), 0.1.6-0.55(5H,m), 2.17(2H, dd, I = 2LIz,l = 8Hz), 9.66(l}J,t,l =2ldz). Experimental 279

Cydopent-3-en-1-ol.

Addition of diborane to cyclopentadiene followed by oxidative work-up by the methodl of winstein ¿ú al.gsz gave the product as an unstable colourless liquid Y=30"/", bp. 77 - 8oC / 55mm. Hg. (lit. bp.337 67 - ïoC / 5Omm.Hg.)

Cyclopent-2-en-l--ol

Cyclopentadiene was reacted with hydrogen chloride (gas) to affo¡d 1- chloropentadiene (used without purification) and then treated with saturated sodium bicarbonate according to the method of Alder338. Distillation afforded a colou¡less liquid Y = 77"/o, bp. 56 - 8oC / 15mm. Hg. (lit. Up.aaa 52oC / 12mm. Hg.)

Penta-1.4-dien-3-ol. This was prepared by the addition of vinyl magnesium bromide in tetrahydrofuran to acrolein according to the reported method339 Y = 48"/o, bp.70 - LoC / tS0 mm Hg. (lit. Up.a+o 64 - 6oC / 110mm. Hg.)

Phenylacetaldehyde Diethylacetal. This was prepared by the addition of ethyt orthoformate to a refluxing solution of be¡rzylmagnesium bromide in ether according to the method of Blicke.34l Y. = 67To, bp 118 - 200C / 15 mm Hg. (lit. bp. 114 - 20cC / 15 mm. Hg.)

Chapter 3.

2-Phenoxyethanol was a commercial product Experimental 280

PhenoxyAlcohols. These compounds were prepared from anhydrous potassium phenoxide and the appropriate chloroalcohol. The results are tabulated below.

Experimental Data for Phenoxvalcohols.

Yield Boiling Pt. (oclmm. Hg.) Reference (%) observed literature

PhO(CH2)sOH 7T 82.s-4/t 170/60 u2 PhO(CH2)¿OH 69 88-e0/0.s 7e/20 u2

(mp.94-5oC) (mp.9a-5oC) 343

G

Thiophenoxymethanol

This was prepared344 ¡.o- thiophenol, formaldehyde solution and a catalytic amount of sodium methoxide. Y = 48"/", 62 - 4oC/ 0.09mm. Hg. (lit. Up.ea+ 60 - soc / 0.1mm. Hg.)

Thiophenoxyalcohols.

These were prepared according to the reported literafure procedures.

Experimental. Data for Thiophenoxyalcohols.

PhS(CH2)nOH Yield Boiling Pt. (oClmm. Hg.) or Reference

(%) Melting Pt. 1oç¡ observed literature PhS(CHf3OH 78 rs9 - 62/15 768/19 345 PhS(CHfaoH 68 mp.22-4 mp.24 346 PhS(CH2)5OH 69 mp.29.5 - 31 mp. 31.5 347 PhS(CHf60H 51 mp.41-43 mp. 43 u8 Experimental 287

Phenoxvalkane thiols. % These compounds were prepared from the appropriate phenoxyalkylbromide or chloride, by nucleophilic substitution using sodiumthiocyanate,- followed by acid hydrotysis according to the reported methods. The crude reaction mixtures of all of these products were initially purified using silica gel flash chromatography, eluting with 2"/" ethyl acetate in petrol ether (bp. 68- 72oC) grading up to 10% ethyl acetate in petrol ether.

Experimental Data for the Phenoxyalkanethiols.

g

PhO(CH2)"SH Yield (%) Boiling Pt. 1oç / mm. Hg.) Reference observed literature

PhOCH2SH 28 I07-4 / 12 68-72 / 2 349 PhO(CH2)zSH 34 7r3-s/e L12/7.s 350

PhO(CH2)gSH 78 130-3/rr ry/12 351

Pho(CH2)+SH 81 128-1.30/4 e6/0.s 352

Pho(CH2)sSH 67 137 -e / 4 rT-s / 3.8 353

The Labelled Compounds.

2-Phenoxyethan-18O-ol and Related Compounds.

A susperuion of phenoxyacetic acid (1.52 g.) in thionyl chloride (4.5 cm3) was stirred at 35oC for 5 hr. Excess thionyl chlorid.e was removed, in uacuo, the product purified by distillation [Y = 88"/o, bp. 140 - zoc / 22mm. Hg. (lit. bp.3s+ I42oC / 25 mm. Hg.)l and anhydrous tetrahydrofuran (1.5 sm3) and Hz18O fieda Research and Development Co.,20.8"/" 18O, 300 mg.¡ was added. The mixtu¡e was stirred at 20oC for 2 hr., the solvent remove to give a residue which was dried ln oacuo (0.02mm., 80oC, 3 hr.), Experimental 282 dissolved in anhydrous tetrahydrofuran (4 cm3), and added dropwise at OoC over a 30-min. period and under nitrogen to a stirred suspension of lithium al.r*i.,.r* hydride @7a mg.) in anhydrous tetrahydrofuran (2 cm3). The mixture was heated under reflux for 3 hr. ,then cooled to LOoC and aqueous sodium sulphate (saturated, 1 cm3) was added. The mixture was extracted into diethyl ether (3 x 10 .*3), dried (MgSo¿) and distilled to yield 2-phenoxyethan-18O-ol (1.019. ,73"/" yield, bp.IIT.S - tt9 oÇ / 20 mm. Hg. (t89 = I0%)

3-Phenoxyprop an-18O- ol.

3- Phenoxypropionic acid was prepared from 3 - iodopropionic acid and potassium phenoxide using the method of Bentley et s¡.355y = 52"/" bp. L05 E - 7oC/ 8mm. Hg. (lit. bp.355 235 - 45oC) [This compound was directly exchanged with deuterium oxide by refluxing a vigorously stirred mixture several times for 8hr. It has been shown that treatment of this compound with thionyl cNoride does not yield the acid chloride but instead affords chromanone.356] Reduction with lithium aluminium hydride, as described above, afforded the product. Y = 87"/o, Block bp. 80 - BS oC / 0.09mm. Hg. (lit. bp.uzI70oC/ 60mm. Hg.)

4-phenoxybut4nJ8o-ol.

4- Phenoxybutanoic acid was prepared by the method of Bentley st a1.355 y = 63o/o, mp. 64-5oC (lit. bp.355 64-6oC) It was converted into the acid chloride using thionyl chloride. Y = 79%o, bp. 6a - 7oC/ O.Lmm. Hg. (lit. bp.3s7 154 - 6oC / 20mm. Hg. ) Reduction with lithium aluminium hydride afforded the desired product. Y = 84"/", Btock bp. 85 - 90oC / 0.45mm. Hg. (lit. bp.Mz164oc / 20mm. Hg.) Experimental 283

2-(Phenoxy-1 -flC) ethanol. This compound was prepared by the following sequence of reactions which are modifications of a reported proce¿.r¡s3S8;

la) 1 5-(Dicyano-13C¿)pentane.

A mixture of L,S-dibromopentane (5.22 g), 18-crown-6 Q.a}g), and potassium cyanide-13C (Cambridge Isotopes, 99"/",13C, 3.0g) in anhydrous acetonitrile (30 cm3) was sti¡red at reflux for L6 h¡. on cooling to 20 oC, the mixture was filtered through silica gel (1-cm thickness in a 50-cm3 sintered glass funnel), the silica gel was washed with acetonitrile (2 x 20 cm3), and the filtrates were combined. Distillation gave 1,5- (dicyano-13C2) pentane as a colourless oil (2.879,99% yield), bp. L19 - 120oC / 1.0 mmHg. (lit. Up.att t L71 - zoc / 12 mm. Hg.)

(b) Pentane-l 5 (dicarboxylic acidJ3Cd

1.5- (Dicyano-13C2) pentane (2.8I g.), tetrahydrofu¡an (1 cm3), and aqueous hydrogen chloride (concentrated, 8.5 cm3) were heated under reflux Íor 6 hr. The solution was cooled to 20oC and the hydrolysis product was extracted into diethyl ether (5 x 8 cm3). washed with aqueous sodium chloride (saturated, 8 cm3), and dried (Mgsoa). Removal of the solvent gave pentane-_l,S (di-carboxylic acid-13C2) e.44 g.,94%" yield), mp. 103 - 105oC (tit. mp.3s8 tOZ - 104oC).

(c) Cyclohexanone-1 J39

This was prepared by modification of a reported procedure.3s8 A mixture of pentane-1,5(dicarboxylic acid-13C2) (3.aag.) and barium carbonate-l3C

(British Oxygen çs. 13Ç= 62.8"/o,225 mg) was heated at 305 - 31.0oC (Woods metal bath) in an apparatus which allowed the product to distill out under Experimental 284

a slow stream of nitrogen (5 cm3 min-1). The cyclohexanone-1-l3c $.73 g 82% yield) was used directly in the next step without purification.

(d) Phenol -r-13q

gg% Cyclohexanone -1-13c $.73 g) was converted into phenor -1-13C Q,.46 g, yield) by a reported pyrotysis reaction35S using a mixed base metal catalyst.35l bp. 77 - zoc / 15mm. Hg. (lit. bp.3a0 182oC) 13C Incorporation (determined by positive ion mass spectromehy) = 99"/".73C NMR (CDCþ),

155.4 (C-l,singlet), other peaks, 129.9,72'J,.2, rL3.7,less than 3"/" of C-1 peak height.

(e) 2- (Phenoxy -1-13C) ethanol. !

Phenol-L-t3C (O.+Z5g) and ethylene oxide (2.40cm3¡ were heated. at 40-45oC for 48 h¡ in a Carius tube. on cooling to ooc, the reaction vessel was opened, and the excess ethylene oxide was allowed to evaporate. Distillation using a T tube in a sublimation block (1i.0 - 120oC / 10mm.

Hg. ; lit. bp. 237oC) gave 2-(phenoxy-1-13c)ethanol (0.649,9I% yield¡, 13Ç = 98%.

2- (Thiophenoxy-1 -13C) ethanol.

This was prepared by a seven stage synthesis from Mel3CozNa (Cambridge

Isotope Laboratories. 13C = 99%)

Ethyl 1J3C acetate.

The 13C labelled sodium acetate treated with triethyl phosp¡¿¡s360 at 170oC and labelled ethyl acetate distilled out in 93% yield.bp.76 - 77oC (tit bp.a6o 770C). Experimental 285

1 -(Methyl-l3C) -cyclohexan-1 -ol.

13C Labelled ethyl acetate was treated with the bis Grignard reagent from 1,S-dibromopentane in diethyt s¡þs¡361 to afford 1-(methyl-13c)- cyclohexan-L-ol in87"/' yield. bp.53-4oC/7mm. Hg. (lit. bp.362l:ß-4oC / 7mm. Hg.)

1 -(Methyl-l3C ) -cyclohex-1 -ene. l-(Methyl -13C)-cyclohexan-1-ol (3.01 g.) was heated at 135oC in a distillation flask fitted with a vacuum jacketed Vigreux column. Addition of one crystal of iodine at 10-min intervals on eight occasions gave a distillate which was a mixture of water and product. Separation of the organic phase, followed by distillation gave 1-(methyl-13C)-cyclohex-1-ene I

(1.91 g. 75% yteld), bp. L04 - 106oC / 760 mmHg.(tit.bp.362110 - 111oC )

Methyl(benzene-1 -13Ç)

Dehydrogenation of 1-(methyl-tac¡-cyclohex-L-ene using a mixed base metal catalyst359 gave methyt(benzene-L-r3c) in 9t"/" yield. The crude product was used in the next step.

Benzoic acid 1J3C. oxidation363 qf the labelled toluene with potassium permanganate gave benzoic acid 1-13C in 90% yield, mpt i.21oc (tit. mp.36s 121, - 2oC)

Aniline 1-LLC hydrochloride.

A Schmidt rearrangement364 using sodium azide in 100% sulphuric acid on the labelled benzoic acid gave aniline 1-13C hydrochloride Y = 91"/o, mp. 196 - 8oc (lit. mp.364 t96-8oc) [NB: The use of > 100% sulphuric acid causes strlphonation of the product.] Experimmtal 286

Thio-phenol-113{l

Diazottzation- of the labelled aniline, followed by addition of potassium ethyl xanthate and heating the resulting mixtu¡e for 2 - 3 h¡. at 50 - 60oC, followed by acid hydrolsis according to the reported method364 resulted. in the one - pot formation of thio-phenol-113c. y - 7r"/o ,bp. t69 - 70oc, (Lít. bp.3a0169oC).

(Thi ophe noxy- 1 -13C ) e thano l. Thiophenol-L-13C, ethylene oxide and a catalytic amount of activated

charcoal were sealed into a Carius tube and heated to 45 - 50oC for 7 days. The Carius tube was cooled in ice - water opened and the ethylene oxide

allowed to evaporate in a fume cupboard using the reported procedure.365 c

Distillation afforded the product as a colourless tiquid. y = 9gyo,bp.l4goC / 14mm. Hg. 13ç = 99o/o (by positive ion mass spectrometry). 13ç NMR (cDCl3) ru.9 (C-1,s). other peaks r29.s, 128.7,1.26.t,60.1 and 36.2 less than 10% o1C1. peak height.

Chapter 4.

General methgd for the Preparation of Alkyl pyruvate Esters. Pyruvic acid, the appropriate alcohol and a catalytic amount of methane sulphonic acid were heated ín1,2 dichloroethane according to the method of Clinton and Lasko*t¡i.366 The yields and boiling pt. data are recorded. below. Experimental 287

Yield (%) Boiling Pt. 1oç7tt mHg.) Reference observed literature

74 t35-7 7U-7 /760 366

CH3COCQCH2CH3 78 s8-s9/20 s440/20 367

69 534/15 s0.5-51l13 368

74 55-8/12 5+7 /11, 366

Labelled Compounds.

The deuterium-labelled compounds ethyl [2,2,22Hg]-, ethyl [2Hs]-, and t- butyl [2Hg]-pyruvate were prepared using, respectively, f2,2,22H31 ethanol, È [2H5]-ethanol, and [2H10]-l-butyl alcohol, by the stand.ard procedur e.366

Incorporation; D3, D5, and D9 99%.

I22À2ÉIaLsrhaad

A solution of f2,2,2 2H3lacetyl chloride (5.009., 61.1mmole) in anhydrous 1,2 dibutoxyethane (25 cm3) was added dropwise to a suspension of lithium aluminium hydride (L.74g.,46.0 mmole) in 1,2 dibutoxyethane (50 cm3) at Ooc, under nitrogen. The mixfure was allowed. to warm and was stirred for 12h¡. 2 - Butoxyethanol (32 cm3, 3o.9g.,19L mmole) was added and the labelled ethanol obtained by distillation as the mixture was heated to 120oC. Y =2.749.,91"/o,bp.72oC

Lactic acid (2-hydroxypropanoic acid) and mandelic acid were commercially available products. Experimental 288

(t) 2-H)¡droxybutanoic acid.

(t) 2-Bromobutanoic acid was hydrolysed using the method of Bischoff.369 Y = 82"/"

(t) 2-Hydroxy-3-methylbutanoic acid.

This was prepared from (t) 2-bromo-3-methylbutanoic acid by hydrolysis using the method of Schmidt.37O Y = M/o, mp. 86 - 7oC (lit. mp.3z0 86oC), bp.70 - 1oC / 0.2mm. Hg. (|it. bp.370 70 - IoC / 0.2mm. Hg.)

2-Hvdroxv-3.3-Dimethylbutanoic acid. This was prepared from tert butyl glyoxal by ^ Cannizzaro reaction according to the method of Fuson et a1.377 y - 94"/", bp. 82.5 - B4oC / É 1.0mm. Hg.(lit. bp.371 86 - 7oC / 1.a mm. Hg.)

Preparation of the Methvl Esters of cr-Hydroxycarboxylic Acids. General Procedure.

The methyl esters were prepared by heating a solution of the ester tn 1,2 dichloroethane with methanol and methanesulphonic acid according to the method of Clinton and Laskowsy.366 The data is recorded below.

Experimental Data for the Methyl Hydroxyacetates.

Compound Yield Boiling Pt. 1oç / mm. Hg.) Reference (%) observed literature CH3CH(oH)CQCHe 79 r45 / 760 145 / 760 366 82 M-6 / 21 æ / 30 372 (CH3)2CHCH(oH)CQCHs u 73 -74.5 / 12 70 -L / 10 373 71 69 -71. / 13 69 -70 / 16 374 77 133-4.5 / 15 \M / 20 375 Experimental 289

This compou¡d was prepared from methyl hydroxy(phenyl)acetate using hexamethyldi'silazane/trimethylsilyl cNoride, according to the method of q¡.376 Langer s¡ The product was obtained as a colourless liquid. Yield = 94"/",8p. 99 - 101oC / 0.5mm. Hg. (lit. bp.377 98 oC / 0.4mm. Hg.).

Ethyl 2-trimethylsiloxypropionate.

This compound was prepared using the method of Langer ¿¡. s7.376 The product was obtained as a colourless liquid. Y = 86/o, bp. 60 - LoC / 12 mm. Hg. (lit. bp.37z 62oC / r2mm. Hg.)

Sodium Lactate - 1- l3C. Ê

This was prepared using the method of Sakami s¡ s¡.378, except that potassium cyanide (tgC) was used in place of sodium cyanide (13C). Thus, acetaldehyde was converted to lactonitrile-L-13C, hydrolysed using concentrated hydrochloric acid, and basified with sodium hydroxide. Y = 92%

Methyl [lJiCILacrate.

A mixtu¡e of methanesulphonic acid (1.5g.) and anhydrous methanol (5 cm3) was added to stirred suspension of sodium [1-13C] lactate (0.539.) in 1,2-dichloroethane (20 cm3¡. The reaction mixture was heated under reflux for 24 h¡. The acid was neutralised with sodium methoxide in methanol

(0.1 mol dm-3 ), and distillation yielded methyl [1-13C] lactate (0.389., 77"/"), b.p. 145oC / 760 mm. Hg. (lit. Up.aø 1.M - 45oC),(13C > 91%).

Methyl [L-lÍlC] Pyr,rrrut..

Methyl [1-13C] pyruvate was prepared by oxidation of methyt [1-13C] lactate (tgC gfY") with 0.167M potassium permanganate solution as described for Experimental 290

the preparation of butyl pyruvate.379 y = 74"/o, bp. 135 - 7oC (lit. Up.aø 135 - 7oc)

Chapter 5.

Preparation of Alkvl Acyloxyacetate esters.

The anhydrous sodium salt of the appropriate carboxylic acid was heated to reflux with the ethyl bromoacetate in acetone according to the method of Lee, Chan and Kwo^.276 The appropriate data are recorded in the following Table.

Preparation of Methyl 2-Acetoxycarboxylate esters.

These compounds were prepared by acetylation of the methyl 2 hydroxycarboxylates (preparation described in experimental for Chapter 4) by the method of Neises and Steglich.379 The appropriate data is recorded in the following Table.

Experimental Data for the Alkyl cr-acyloxycarboxylate esters.

Compound Yield Boiling Point 1oç/mm. Hg.) Reference (%) observed literature

CH3CO2CH2CO2Me u e6-7.5 / 80 170-2/760 380 86 L08-9 / 60 108/60 381 ECO2CH2CC2Et 81 118-719.5 / 60 Lr9/60 380 iPrC02CHzCOzEt 71, 723-4 / 60 L23/60 380 Et 85 94- 6 / 7.5 IM/øO 380 iBuCO2CHzCOzEt 69 1,07-3/13 737/60 380 Experimental 291

Compound Yield Boiling Point (oClmm. Hg.) Reference (%) observed literature sBuCO2CH2CO2Et 74 1.02-3.s / 73 736/60 380 tBuCO2CH2CQEt 73 88-9 / 20 724/60 380 nCsHTTCQCH2CQEI 78 721-3 / 17 755/60 380 PhC02CH2CQEt 97 I73 4 / 7.2 286.5-288.5 382 MeCO2CH(Me)C02Me 82 66-8.s / L2 64/10 383/384 MeC02CH(Et)CO2Me 87 73-5 / 72 70/10 384 MeCO2CH(iPr)C02Me 79 50-51.5/1 50/1 385 tBuCO2CH(Me)CO2Me 84 50 -2/ I.2 52-3/ r.5 386

2-Acetoxy-3 3-dimethylbutanoic acid methvl ester. Ë This was prepared by the acetylation of methyl 2-hydroxy-3,9- dimethylbutanoate using the method of Neises and Steglich.379 Y=87"/o, bp. range 107 - 9oC / 32mm. Hg. 1H nmr (õH,CDCI3): 0.90 (9H,s), 2.I4 (3H,s), 3.75 (3IJ.,s), 5.09 (1H,s), ir (film): 1728 cm-7. Accurate Mass M+.

188.1051 ; requires C9H160a: 188.1048

Experimental Data for the Methyl 2-Acetoxycarboxylate esters.

Compound Yield Boiling Point (oClmm.Hg.) Reference (%) observed literature

MeCOCH(OH)CO2Me 68 65-67 / 71, 68 / L3 387 MeCOCH(oH)C02Et 62 92-94 / 11, 98 -101 / 14 388 MeCOC(Me)(OH)CO2Me 69 70-1. / 7 77 -80 / 10 389 MeCOC@t)(OH)C02Me 48 68-70 / 5 80 -83 / 10 38e PhCOCH(OH)CO2Et 68 90 -99 / 0.0t 90 - 100 / 0.02 388 Experimental 292

Labelled Compounds.

Eth)¡l (acetoxf l J3C) acetate

This was prepared by the method of Lee ¿¡ s¡.276 using 1-13C sodium acetate and ethyl bromoacetate. Y = 74"/o, bp. 106.5 - 108oC / 60mm. Hg. (lit. bp. 108oC / 60mm. Hg.) (taç > 98%)

ft. 7. 2.2.2 -

This was prepared from sodium acetate and 1, 7,2,2,2 -z}i.s Ethyl bromoacetate [obtained from bromoacetyt bromide and l, I, 2,2, 2 - 2lHs

Ethanol - OD (Y = 94"/o, bp. 159oC)l according to the method of Lee ,¡ a¡.276 Y = 63"/", bp. 106.5-108oc / 60mrn. Hg. (lit. bp. 108oC / 60mm. Hg.) (2Hs > e8%) ç

Ethyl (2. 2 - Dimethylpropionvloxy)acetate-1-13e.

The ¡eation of the sodium salt of 2,2 - dimethylpropionic acid and ethyl L-

13C bromoacetate (obtained by treating t-13C acetic acid with PBr3 , then ethanol and distilled, bp. 157.5 - 158oC) by the method of Lee ¿¡ q1.276 afforded the desired product as a colourless liquid. Y = 57"/" (overall), bp. 122.5 - 4oC / 58mm. Hg. (lit. bp.124oC / 60mm. Hg.) (tsç > 98%)

General Methgd for the preoaration of Acyloxyacetonitriles. These compounds were prepared by the reaction anhydrous sodium carboxylate with chloroacetonitrile according to the method of Wagenkneckt ¿l 61.399 The relevant data is recorded in the following Table.. Experimental 293

Experimental Data for the Acyloxyacetonit¡iles.

Compound Yield Boiling Point (oClmm. Hg.) Reference (%) observed literature MeCO2CH2CN 86 82-4 / 30 179-80/755 400 EtCO2CH2CN 84 77 -79 / 22 188-9/759 407 nPTC02CH2CN 89 84-5.5 / 22 200/758 401. PhC02CH2CN 67 88 - 90 / 0.55 9r/O.e 402 PhCH2CO2CH2CN 74 72-74 / O.tZ 96-97 / 0.5 403

Labelled Compound

€2LZÉh ecetoxy) a cetonitrlU !

This was prepared by the method of Wagenkneckt ¿ú a1.399 using sodium 2IJg - acetate and chloroacetonitrile. Y = 84o/obp.82 - 4oC / 30mm. Hg. (lit. bp.ffi 179 - 80oC)

Chapter 6.

Alkyl Phenyl Ethers.

Alkylation of anhydrous potassium salt of 2 - cresol with the appropriate alkyl bromide in tetrahydrofuran was achieved by prolonged reflux (ca.

72hr.). Thus, the following ethers were obtained : methyl .¡¡.¡404 I = 72%,bp. (lit. bp. 770-2oC / 760 mm. Hg.); ethyl ethel0s Y =77"/o,bp. (lit. bp. 190oC); iso propyt ethel06 Y = 43/o,bp. (lit. bp. 192oC) Experimental 294

:2-Ethvlohenol. This was prepared by heating the aqueous diazonium salt of 2 - ethyl

aniline at 50oC for t hr, using the reported proceduue407. Y = 4B/o, bp. (lit. bp.a07 206.5 - 20ToC)

2-Propylphenol.

This was prepared using phenol, zinc and allyl bromide according to the reported methoda0s. Y = 60%bp. (lit. bp. 224.5 - 6.5oC)

Benzyl Ethyl Thioether.

The sodium salt of benzyl mercaptan was heated with iodoethane as

desc¡ibed for the preparation of benzyl alkyl ethers. Y = 7I"/o, bp. (lit. bp.aOr t 21,4-60C)

Phenvlpropane-L -thiol.

This was prepared by heating phenyl - 1 - bromopropane with sodium hydrogen sulphide as described for the preparation of phenyl-1- mercaptoe¿r¿ng4lO. Y = 69"/o,I37-9oC / 75 nn. Hg. (lit. bp.al1 1,Og - 4oC / mm. Hg. )

2-Methoxv Phenoxypropanol.

To a solution of 2 - methoxyphenol (1,.249, 10.0 mmole) in anhydrous tetrahydrofuran (L2 cm3), was added oil - free potassium hydride (418mg,

L0.2 mmole). The mixture was briefly warmed (s}oc/20 min.) to ensure complete reaction and 3 - chloropropanol (1.65, 17.45 mmole) added and heated to reflux for 96 hr. The reaction mixture was washed with ammonium ciloride (saturated, 2 x 7 cm3) and brine (2 x 7 cm3), then dried and the solvent removed in aacuo. The residue was purified using radial chromatography, eluting with a gradient of 5% + r2.5% ethyl Experimental 295

acetate in petrol ether, affording the desired product as a colourless oil. Y = 7H 82% nmr(CDCIs): 2.0aQH, þ I = 6.5]fl12),3.92(3H, s), 3.94(2H, t,l = 6.5), 4.73(2H, t,l = 6.5),6.78(4H, s); Accurate Mass M+' obtained 782.0939;

C 1 9H1aO3 requires 182.09 43

f References.

1. I.E. Bartmess and R.T. Mclver,]r. ,"The Gas phøse Acidity Scale" in "Gas - Phase Ion Chemistry", Vol. 2, Chapter I\, (Ed. M.T. Bowers),

Academic Press, New York, 1979.

ZaJ.H. Bowie, Mass Spectrometry Reaiews 1984,9,I - 37 ; 2b J.M. Riveros, s.M. José and K. Takashima, "Aduances in Physical organic Chemistry", YoL 2r, pr97 - 240, (Eds. v. Gold and D. Bethell), Academic press, orlando

(usA), 1985 F

3. N.M.M. Nibbering, "Aduonces in Physical Orgønic Chemistry", vol. 24,

(Edits. V. Gold and D. Bethell), Academic Press, Orlando (USA), 1988

4. M.E. Jones and G.B. Ellison, I. Am. Chem. Soc., !989, 1-1-1-, 1,645; V.M. Bierbaum, J. Filley, C.H. DePuy, M.F. Jarrold and M.T. Bowers, I. Am.

Chem. Soc., 1985, 107, 28'/-.8

5. H. van der_Wel, N.M.M. Nibbering and M.M. Kayser, Can. l. Chem., 1988,66,2597

6a. C.L. Fischer, S.D. Kahn, W.J. Hehre and M.C. Caserio, I. Am. Chem.

9oc.,1989, 1L1,7379 ; 6b. l.{. Tabet, C. Prevost, A. Bouillot, J. Bastard, D.K. Manh, Y. Tandeur, HeIo. Chim. Acta, L986, 69,806

7. K. Biemann irt"Mass Spectrometry - Organic Chemistry Applications",

McGraw-Hill, p160, New York (USA), 1962 References 297

8. K.R. Jennings ín "specialist Periodical Report, Mass spectrometry", yol.

4 (Ed. R.A.W- Johnstone), p.203, Chemical Society,London, (I9ZZ)

9. K.R. Jennings in "specialist Periodical Report, Mass spectrometry", Yol L4, p 217, Chemical Society,London, $9n)

10. W.A. Garland and B.H. Min, /. Chromntog.,I9T9,1Z2, ZZ9

11. D.F. Hunt and F.W. Crow, Anal. Chem.,7978, 50, I78I ; A.G. Netting and A.M. Duffield, Biomed. Mass Spectrom.,L985, 72, 668; G.K.- C. Low and A.M. Duffield, Biomed. Mass Spectrom. , L985, 1-2, 948; G.C. Ford, K.N. Cheng and D. Halliday, Biomed. Mass Spectrom. ,L985,12,432 C

12. W.A. Garland and M.P. Barbalas, in "Mass Spectrometry in Biomedical

Research" (Ed. S.K. Gaskell) , J. Wiley and Sons Ltd., Chichester, 1986, p379

13. M. von Ardenne, K. Steinfelder, R. Tümmler, in " Elektronenanlagetungs - Massenspektrographie Organischer

Substanzen", p10, Springer - Verlag, Berlin (Germany), L7TI

L a. H.S.W. Massey in "Negøtioe lons", 3rd. Edition, Cambridge University Press, Cambridge, 1.976 ; I4b. A.G. Harrison, in "Chemical Ionization Mass Spectrometry", CRC Press, Boca Raton (uSA),1983 ; L4c. J.G. Dillard, Chem. Reaiew,7973, 6,589

15. D.R Dakte¡nieks, I.W. Fraser, J.L. Garnett and I.K. Gregor, Orgønic Mass Spectrom.,7979,'1,4, 330; E. Houghton, P. Teale, M.C. Dumasia and ].K. Wellby, Biomed. Møss Spectrom.,'1982, 9, 459; R.C. Dougherty, M.l. Whitaker, L.M. Smith, D.L. Stalling and D.W. Kuehl, EHP Enpiron. References 298

Health Perspect.,1980,36, 103; R.c. Dougherty and K. piotrowska, l. Assoc. Off. AnaI. Chem., 1976, 59, 1023

16. L.G. Christophorou and S.R. Hunter in "sutarms of lons and Electrons in Gases", (Edits. W. Lindinger, T.D. Mark and F. Howorka), Springer -

Verlag, Vienna, 1984

17. Y. Kokaku, Y.Hatano, H. Shinamori and R.W. Fessende^, I.Chem.

Phys.,7979, 71, 4883; Y. Kokaku, M. Toriumi and Y. Hatano, ibid., LgB}, Zg, 6167 ; H. Shimamori and R.W. Fessenden, ibid., L98L, Z4,4Sg; H. Shimamori and H. Hotta, ibid.,7983,7g,I3lB

I

18. M.E. Rose and R.w. Johnstone in "Mass spectrometry for chemists and Biochemists", p. 34-6, Cambridge University Press, Cambridge, I9B2

19. L.G. Christophorou, D.L. McCorkle and A.A. Christodoulides in

"Electron Molecule Interactions and their Applications", Vol. L., p. 477,

(Ed. L.G. Christophorou), Academic Press, New York, L984

20. Ref. 18 , p t20 - 7

21. H. Budzikiewicz, Mass Spectrom. Rea., 1986,5,345; see 1,4b. p.27 ; C.H. DePuy and V.M. Bierbaum, "Ptoton Transfer Reøctions of Anions" in " Structure / Reactiaity and Thermochemistry of lons",(Edits. P. Ausloos and S.G. Lias) , D.Reidel Publishing Company, Dordrecht, Holland, t987, p. 293

22. I.B. Westmore and M.M. Alauddin, Mass Spectrom. Reo., L986,5, 381 References 299

23. J.A.D. Stockdale, R.N. Compton and P.W. Reinhardt, Phys. Rea.,796L, 8L,184

24. T.O. Tiernan in "Interactions Between lons and Molecules", (Ed. P Ausloos), Plenum Press, New York,1975

25. G. Panl and P. Kebarle, I.Am. Chem. 9oc.,7989,1-1-1-, 464; J.J. Grabowski, J.M. Van Doren, C.H. DePuy and V.M. Bierbaum, l.Chem. Phys.,1984,80, 575

26. R.C. Doughertf , AnaI. Chem., !98I,53, 625A

27. S.A. Sullivan, C.H. DePuy and R. Damrauer, l. Am. Chem. Soc., 198'1., L03,480

28. D.J. Hajdasz and R.R. Squires, I.Am.Chem.9oc.,L986, 108,3139

29. J.C. Kleingeld and N.M.M. Nibbering, lntl. l. Mass Spectrom. Ion Phys., 1983, 49,3I'J.

30. C.H. DePuy, R. Damrauer, J.H. Bowie and J.C. Sheldon, Acc. Chem. .

Res., L987,20,I27

31. R.C. Lum and J.J. Grabowski,l.Am. Chem. 9oc.,1988, 1L0, 8568

32. C.H. DePuy , V.M. Bierbaum, R. Damrauer and J.A. Sonderquist, l.Anr

Chem. Soc., L985,107,3385 , c¡v,cl' lzl¿c¿racrs L*td' 14.¿cei* . References 300

33. L.G. Christophorou, D.L. McCorkle and A.A. Christodoulides, in "Electron - Molecule Interactions and their Applications", Chapter 6, vol.

1, (Ed. L.G. Christophorou), Academic press, 1984

34. M.McFarland, D.L. Albritton, F.C. Fehsenfeld, E.E. Ferguson, A.L. Schmeltkopf, l.Chem.Phys.,1973, 59, 6629; E.E. Ferguson, Intl. l. Mass spec. Ion Phys.,7976, 19,53 - 70 ; F.C. Fehsenfeld "Interactions between Ions and Moleculesi', p. 387 (Ed. P. Ausloos), Plenum publ. Corp., N.y., 1975

35. G. Drabner, A. Poppe, H. Budzikiewicz, Intl. l. Mass Spectrom. Ion

Processes, 1990, 97, 7 !:

36. K.R. Jennings, lntl. f. Mass Spectrom. Ion Phys., 1968, 1,227 ; W.F. Haddon and F.w. Mclafferty, I. Am. chem. soc., \968, go,474s; M.E. Rose and R.W. Johnstone in "Mass Spectrometry fo, Chemists ønd Biochemists", p. L39, Cambridge university university Press, cambridge, 7982

37. I.H. Beynon and ].R. Gilbert in "Application of Transition State Theory to Unimolecular Reactions", Wiley, Chichestet, r9B4; C. Lifshitz, l. phys. Chem., 1983, 87, 23C4

38. H.S.W. Massey, Rep. Prog. Phys.,1949, L2,3279

39. R.G. Gilbert, M.M. Scheil and P.J. Derrick, Orgønic Møss Spectrom.,

1985, 20,430 References 301

40. I.M. wilson in "Mass spectrometry" (Ed. D.H. williams ),,vol. L, p. r- 30, Chem.Soc.,London,(797t); yol. 2,(t973) p. I-32; (1975) Vol.3 p.B6-II6 (ed.R.A.w. Johnson) ;(1977) vol. 4, r02-g1r; K. Levsen and H. Schwarz, Mass Spectrom. Rea.,1983, 2, T

41. R.G. cooks and G.L. Glish, chem. Eng. News, rgg"J., sg,40; F.w.

Mclafferty, Sciutce, 798I, 2lq, ZB0

42- I.H- Beynon and R.G. Cooks, Intl. l. Mass spectrom. Ion phys.,"J.976, tg, 707

43. A.P. Bruins, K.R. Jennings and s. Evans, rntt. l. Mass spectrom. Ion Phys., 7978, 26, 395 ; R.K. Boyd and J.H. Beynory org. Mass spectrom., rgzz,

12' 763; D.s. Millington and J.A. smith, ibid., p.264; R.p. Morgarç C.J.

Porter and J.H. Beynon, ibid., p. 73s ; cJ.porter, A.G. Brenton and J.H. Beynon, Int. l. Møss Spectrom. Ion phys., 1980, 96, 69 ; M.I. Lacey and C.G. Macdonald, Org. Mass Spectrom., 1929, lq, 465

44.R.C Dougherty Anal. Chem.,1926, qg,2042

45. R.c. Dougherty , J. Dattory and F.J. Biros, org. Mass spectrom., rg7z,6, It77

46. F.M. Harris, G.w. Trott, T.G. Morgan, A.G. Brentory E.E. Kingston and

I.H. Beynon, Mass Spectrom. Rea., 1984, g, Z0g; I.S. Allen, Reu. Sci. |nstrum,,1947,18,739; C. Ia Lau,"Mass discrimination caused by electron multiplier detectors." in "Topics in organic Mass spectrometry", (Ed. A.L.Burlingame), p. 93-1.20, Wiley, New york, 1970 ; N.H. Tolk, W. References 302

Heiland, J.C. Tully, C.W. White (Edits.) "Inelastic Ion-Surface Collisions"

Academic Press, New York, (1977)

47. A.L.c. smit, M.A.I. Rossetto and F.H. Fierd, Anal. chern., 19T6, 4g,2042

48. G-C. Stafford Finrrigan Corporation, communication at 25th Annual Conference on Mass Spectrometry and. Ailied Topics. Washington D.C.$9n) ; G.C. Staff ord,, Enairon. Health perspectives,19g0,36, gS.

49. G.R. waller in "Biochemical Applications of Mass spectrometry", First supplementary volume, Wiley-Interscience, New yotk, I9T2

50. s.P.o. Markey, A.J. Lewy, A.p. zavadíl,J.A. poppiti and A.w. Hoveling, Quantitation of biogenic amines and metabolites using negative ion chemical ionization, in Extended Abstracts of the 25th Annual Conference on Mass Spectrometry and Allied Topics, p.276-27g., American Society for

Mass Spectrom etry, 1977

51. J.L. Holmes and J.E. Szulejko, Org. Mass Spectrom., Lgg3,]:g,27g

52. T.R. Kemp in"Mass Spectrometry" yol. g, Chap.4. p I20-t

53. R.P. Morgan, J.H. Beynory R.H. Bateman and B.N. Green,

Intl. l. Mass Spectrom. Ion Phys.,I7ZB, Zg, IT'!.

54. T. Ast, M.H. Bozorgzadeh, J.L. wiebers, J.H. Beynon and A.G. Brenton,org, Mass spectrom., 1979,1.4,3r3; M.J. Lacy and C.G. Macdonald, org. Mass spectrom., \980,ls, 484; M.J. Lacy, c.G. Macdonald., K.F. Donchi and P.J. Derrick, Org. Mass Spectrom.,IïBI, L6, 3S1 References 303

55. A.P. Bruins, Anal. Chem..,I9BS,SZ,S72

56. A.Poppe, E.schroder, H. Budzikiewicz, org. Mass.spectrom., 1986,2r, sg

57. J.L. Holmes, Org. Mass Spectrom., I9BS, 20, 169

58. D.J. Burinsky, R.G. Cooks, E.K. Chess and M.L. Gross, AnaI. Chem., 1982,54,295

59. J.H. Bowie and T. Blumenthal, l. Am. Chem. Soc., I9TS, gZ, 2gS9 ; I.H. Bowie and T. Blumenthal, Aust. l. Chem.,19Z6, Zg, Ll^s

60. RG. Cooks, J.H. Beynon and T. Ast, I. Am.Chem. Soc. ,1922,94,1004 ; T. Ast, ].H. Beynon and R.G. Cooks,.I. Am. Chem. Soc., 1922,94,66I!,T.

Keough, J.H. Beynon and R.G. Cooks,l. Am. Chem. Soc., !973, gS,1695

61. J.E. Szulejko, I. Howe and J.H. Beynon, Intl. l. Mass Spectrom. Ion Phys.,198I,37,27

62.J.H. Bowie, P.Y. White, J.C.Wilson, F.C.V. Larsson, S.-O. Lawesson,I.Ø.

Madsen, C. Nolde and G. Schroll, Org. Mass Spectrom., LgT7, t2,19'J,

63. R.G. Kingston, A.G. Brenton, M. Guilhaus and J.H. Beynon, Org. Mass Spectrom., 1986, 2'1., 697 ; I.E.Szulejko, f.H. Bowie, I. Howe and ].H. Beynon, Intl. l. Møss Spectrom. Ion Phys.,1980, 94,99

64.I. Flowe, J.H. Bowie, J.E. Szulejko and J.H. Beynon, l. Chem. Soc.,Chem. Comm., 1979,984-5 References 304

65. P.c. Burgers, J.L. Holmes, A.A. Mommers and J.E. szulejko, org. Mass Sp ectrom., t983, 1,8, 596

66. M.M.Bursey, Mass Spectrom. Reo.,1990,9, SSs

67. J.K. Terlouw in"Adoances in Mass spectrometry",yol11A, p. 9g4, (Ed,. P. Longevaille), Heyden and sons, L9B9 ; I.K. Terlouw and H. schwarz, Angew.Chem.Intl.Ed.(Eng.),I9BZ, 26, B0S ; Angeut.Chem.,IggT,99, g29 ; C. Wesdemiotis and F.W. McL affefty, Chem.Rea., I9BZ, 97, gs

68. R. Feng, C. westdemiotis, F.w. McLafferty, l.Am.chem.soc.,Lggr,1.og, 6521 ç

69. S. Villeneuve, P.C. Burgers, Org. Mnss Spectrom.,1986,2]^,733

70- J.H- Beynon and R.G. Cooks, Intl. l. Mass spectrom. Ion phys.,, 19z6, ]rg, 707

71'. R-D. Bowen and D.H. williams, "Reanangements in Ground and Excited states",Yol. L.,Essay 2, pss, (ed. p. de Mayo), Acad.emic press, New York, 1980

72. M.S. Kim and F.W. McL affefty, I. Am. Chem. Soc., I9ZB, L00, 3279

73. R.G. cooks (Ed.) "Collisional spectroscop/", plenum, New york, !9TB

74. R.G. cooks, J.H. Beynon, R.M. caprioli and G.R. Lester, "Metastøble Ions", Elsevier, Amsterdam, 1973 ; R.G. Cooks and J.H. Beynon, "Møss spectrometry", (Ed. A. Maccoll), MTP Intl. Rev. of science, physical References 30s

chemistry, series 2,Yor.5, pr59, Butterworths, London. ; R.K. Boyd and J.H. Beynon, IntI. l. Mass Spectrom. Ion phys.,I97Z,2g, 163

75. J.H. Bowie, Mass Spectrom. Reo.,L990, g, g4g

76. M.B. Stringer, J.H. Bowie, P.C.H. Eichinger, C.J. Currie, /. Chem. Soc.,Perkin Trans. 2, 1987, 385

77. P.c.H. Eichinger, I.H. Bowie and T. Blumenthar, l.org.chem., 19g6, s't,, 5078

78. c.H. DePuy, v.M. Bie¡baum, R. Damrauer, J.A. sonderquist, I. Am. C Chem. soc.,1985,102,3385 ; s. Chowdhury and A.G. Harrison, org. Møss Spectrom., 7989, 24, 123

79. P-c.H. Eichinger and J.H. Bowie, org. Mass spectrom.,rggz, 22,'!.09

80. G.w. Adams, J.H. Bowie and R.N. Hayes, l. chem. soc.,perkin Trøns. 2, \989,2159

81. P.c.H. Eichinger, I.H. Bowie and R.N. Hayes, l.org.chem., r9gr, sz, s224

82. I. Adams, Mass Spectrom. Reo., 1990, g,'1,4I

83. N.l. Jensery K.B. Tomer, M.L. G¡oss, I. Am. Chem. Soc.,I98S,].07, L86g

84a. M.B. Rubin and S. Inbar, /. Org. Chem., 1988,53,33SS ;8ab !.y. Paukstelis and D.N. stevens, Tetrahedron Letters, \gz'J,,3s49 ;84c X. creary, P.A. Inocencio, T.L. Underiner and R. Kostromin, l. Org. Chem.,1985,50, Rrtrences 306

1932 ;84d D.G.H. crout and C.J.R. Hedgecock, I. Chem. soc., perkin Trans. 1, 7979, 7982 ;84e s. Grelin and R. Grelin, I. org. Chem., 7gzg, 44, g0g ; g4f M.]. Nye and w.P. Tang, l. Chem. soc. D., 197L, 1394 ; gag G. palmisano, B. Danieli, G. Lesma and M. Mauro, l. chem. soc., Chem. Commun., 19g6, 7564 F.M. Dean ;84h and L. E. Houghton, Tetrahedron Letters, 1969, 3sr9 ; 84i M. Rentzea and E. Hecker, Tetrøhedron Letters, 19g2, rZgs ; g4i M.p. Hartshorn, w.T. Robinson, K.H. sutton, J. vaughan and G.J. wright, Aust. I. Chem.,7983,96, glz

85. D. Hellwinkel, F. Laemmerzahl and G. Hofmarvt, chem. Ber., rggg, 7L6' 3375 E.J.Madaj, ; Jr., D.M. snyder and w.E. Truce, I. Am. chem. soc., 7986, 1'08, 3466; H.P. Abricht, H. schmidt and K. Isslieb, Z. chem., rggs, 2s, ç 470 ; D. Hellwinkel and M. supp, chem. Ber., L976,'r.0g,3749 ; w. Dannecker and M. Fariborz, Z. Naturforsch., Teil B, Lgr4, 2g, s7s ; B. Dhawan and D. Redmore, l. chem. Res. synop., \9gg,222 ; J. Heinicke, E. Nietzschmann and A. Tzschach, l. organomet. chem., r9g3, z4g, L; L.s.

Melvin, Tetrahedron Letters, rg8r, 22, 3zrs; J. Heinicke, I. Boehle and A. Tzschach, /. organomet. chem., 19g6, grz, 11 ; J. Heinicke, E. Nietzschmann and A. Tzschach, /. organomet. chem., 19g6, J1,0, ClT ; G. B. Hammond, T. Calogeropoulou and D. F. wiemer, Tetrøhedron Letters, 1986,27,4265; T. calogeropoulou, B. Hammond and D. F. wiemer, I.org. Chem.,1987,52, 4l8S

86. M.D. Rozeboom, J.P. Kiplinger and J.E. Bartmess, I. Am. chem. soc., 'l'06, 7984, 1025 ; c.s.s. Rao, G. Kumar, K. Rajagopalan and s. swaminathan, Tetrahedron, rgïz, gg,2lgs; RC. Gradwood and R.M. Lett, I. Org, Chem.,7982,47,2268; T. Miyashi, A. ]Hazato and T. Mukai, I. Am. chem. soc., 7982, lo4, 89r ; p.A. wend.er and s. McN. siebu¡th, Tetrahedron Letters, 1981., 22, 247],; R.w. Thies and J.R. pierce, I. org. References 307

chem., 7982, 47, 798 ; w.L. Brown and A.G. Fallis, Can. l. chem., 19g7, 6s, pierre 7828; L-4. Paquette, F. and C.E. Cottrell, I.Am.chem. soc., r9g7, 'l'09' 5737; s.R. wilson, R.N. Misra and G.M. Georgiadis, J. org. Chem., L980, 45,2460; R.L. Danheiser, c. Martinez-Davila and J.M. MorinIr., I. Org. Chem., 7980, 45, Ig40

87. G. warm pharm.,7gzs,30g, and Ch. Lachmann, Arch. 3g9 ; G. Deri, R. s. Kapil and s. P. popli, Ind. l. Chem., sect. B, Lgzg, rzB, zs; R. s. Gupta, T. R. seshadri, C. s. sharma and N. D. sharma, Ind. l. Chem., sect. B, 1.979, 178,37 ; M. Hauteville and J. Chopin, BuIl. Soc. Chim. (Fr.),1979,II9 ; M. Hautevill€, M. chadenson and J. chopin, BuII. soc. chim. (Fr,), 1979, rls ; K.D' Banerjii and D. Poddar,l. Ind. chem. soc., L97g,, ss, sM ; R.C. Reddy, Ê G.L.D. Krupadanam and G. srimannarayana, Ind. J. chem. sec. B, rgg7, 268'974; A.K.D. Mazumdar, G.c. saha, T.K. sinha and K.D. Baneqi, proc. NatI. Acad. sci., India sect. A,l9ïs,5s, 109 ; H.L. Graggard, K.N. wadodkar and B.]. Ghiya, Ind. l. chem. sect. B, rggs, zqB, 1244 ; s. saxena, I.K. Makrandi and S.K. Grover, Synthesis,IggS, 69T

88a. s. selman and J.F. Eastham, e. Rea. Chem. soc., !960,14,227; ggb c.G. Screttas, M. Micha-Screttas and C.T. Cazianis, Tetrahedron Letters,Iggg,24,

3287 ; 88c I. Rajyaguru and H.S. Rzepa, l. Chem, Soc., perkin Trans. 2, 19g2, 1'819 ;88d K. s. warren, o.K. Neville and E.c. Hendl"y,l. org. chem., \96g, 28' 2152; 88e D. Klenburg, Ann. Acad. sci. Fennicae ser. A II, 1962,]r].g, s2 ; p.p. 88f F. sanchez-viesca and Marroquin, Rea. I-atinoøm. euim., rgg4,7s, 60 88g H. Dahn, ; M. Ballenegger and H.p. schlunke, chimia., 1964, tg, s9 ;

88h L. D. Huu, F. Dayer, L. Duc, H. Rodecowal and H. Dahn, HeIa. Chim.

Ãcta.,7974,57,2215; 88i H. Dahn and s. Karoui, Hela. Chim. Acta., 1969, 52, 249]. ; 88j H. Gowal, A. spiess, M. Ballenegger, L. Duc, H. Moll, H.p. Schlunke and H. Dahn, Heht. Chim. Acta.,I9gS,6g, 2Ig2 References 308

89. D. H. Hunter, I. B. stothers and E. w. warnhoff, in "Rearrangements in Ground and Excited states", vol. r, p. 391, (Ed. p. de Mayo), Academic

Press New York, 1980

90. A.S. Kende,Org. React.,7960,1'l,, 26I; N. Schamp, N. de Kimpe and W. coppens, Tetrahedron,7975, i]-, z09r; p. Matzinger and C.H. Eugster,

Hela. chim. Actø., 7979, 62, z32s; A. Baretta and B. waegel, Tetrahedron Letters, 1976,753 ; R. verhe, N. schamp, L. De Buyck and R. vanloocke,

BUII. Soc. Chim. Belg., 7975, 84, 381

91.P Fritsc}:., Liebigs Ann. chem., !894, 2zg, 3Lg ; w.p. Buttenburg, Liebigs Ann. chem.,'J'894,279,327; H. wiechell, Liebigs Ann. chem., rg94, z7g, ç

337 ; R. Rebois and J. chauveau, c. R. Hebd. seønces Acad. sci. ser. c, 1976,

282,947 ; G. Koebricþ H. Trapp and I. Hornke, chem. Ber., 196T,100,96r ; W.M. Jones and R. Damico, I. Am. Chem. Soc., 1963, g5,2223; B.A. Shainyan and A.N. Mirskova, Zh. Org. Khim., 1983, 'tg, 'J.344 ; Z.A. Talaikite, I.A. Dembinskiene, V.I. vidugiriene, Z. Stumbreviciute and L.p.

Rasteikiene, Zh. Org. Khim.,I9B3,lg, 146'/-.

92. E. Grovenstein Jr. and A.A. Morton, Angew, Chem.,Ig7B, 90,3t7 ; E.

Grovenstein jr. and A.B. cottingham, I. Am. chem. soc., r9zz,99, rggr ; A. Dagan and M. Rabinovitz, l. Am. Chem. Soc., 1976,9g, g26g ; E.

Grovenstein Jr., Ado. organomet. chem., r97z,'r,6, 167; E. Groverutein Jr., K-W. Chiu and B. B. Patil, /. Am. Chem. Soc., 1980, 102,5848 ; E. Grovenstein )r., K.w. Black, s.c. Goel, R.L. Hughes, J.H. Northrop, D.L.

St¡eeter and D. Van Derveer, l. Org. Chem.,I9Bg,Sq,16TI; J.J. Eisch, C.A. Kovacs and P. Chobe, I.org.chem.,r9B9,s4,rzzs; E. GrovensteinJr. and P.C. Lu, I. Am. Chem. 9oc.,1982, 104, 668I ; A. Maercker and M. passlack, References 309

Chem. &er.,1983,1-L6,710 ; E. Grovenstein Jr. and G. Wentworth, I. Am. Chem. 9oc.,7963, 85, 3305

93. E.S. Wallis andJ.F. Lane, Org. React.,t946,g,267

94. V.L. Bauer and O. Exner, Angew. Chem. lntl. Ed. (EngI.),1974,'/.,9, 376, lAngew Chem. 1974,86, 419); H.L. Yale, Chem. Rea., !943,99,209

95. R.T. Conley and S. Ghosh, Mech. MoI. Migr., 797I, 4, 797 ; G.H. Corkins, L. Storace and E. Osgood, /. Org. Chem., L980, 45,3156 ; A. Delgado, ].M. Garcia, D. Mauleon, G. Minguillon, J.R Subraits, M. Feliz, F. Lopez and D. Velasco, Can. l. Chem.,1988, 66,517 ; A.V. Eremeev, I.P. E Piskunova, R.S.E1'kinson, Khim. Getercsikl Soedin., 1985, L202 ; R.F. Parcell and J.P. Sanchez, 1.Or9.Chan.,L987,46,5229; J.A. Hyatt, I.Org. Chem.,1981,,46, 3953; K.R. Henery-Logan and T.L. Fridinger, I. Am. Chem. Soc., 1967,89,5724 ; J.L.La Mattina and R.T. Suleske, Synthesis, 1980,329 ; W.B. Renfrow, J.F. Witte, R.A.Wolf and W.R. Bohl, I.Org. Chem., L968,33, 150

96. P.C. Bulman Page, C.M. Rayner, LO. Sutherland, l. Chem. Soc., Chetn. Comm., 1988,356 ; C.H. BehÍens, Y.S. Ko and K.B. Sharpless, /, Otg.

Chem., 1985, 50, 5687 ; I.C. Gill, P.M. Lockey, B.A. Marples and J.R. Traynor, l. Med. Che¡n., L986,29,7537

97.L. Paquette, Org. Reøct.,\977,25,7; Mech. Mol. Migr.,1968,7,I21 ; Acc. Chem. Res.,1968,1,209; F.G. Bordwell, Acc. Chnn. Res., 1970,3,281 ; C. Rappe, S. Patai in "The Chemistry of the Cørbon HøIogen Bond" Part 2, p.

1L05-L0, John Wiley and Sons, New York, 7973. References 310

98. V.N. Drozd, IntI. l. Sulphur Chem.,I9Tg, g, 443; D.M. Synder and W.E. Truce, I. Am.Chem. Soc., \979,].01,5432 ; y. Fukazawa, N. Kato and S. Ito, Tetrahedron Letters,7982, Zg, 432

99. K.I. H. Kruithof and G.w. Klumpp, Tetrahedron Letters, LgBz,23,3r0r ; B. cazes and s. Julia, Tetrahedron Letters, 1924,2077 ; J.A. Marshall, T.M.

]enson and B.S. Detloft, I. Org. Chem.,1986, SL, 43'J,6; A.G. Anderson Jr. and M.T. Wills, I.Org.Chun.,t967,g2,3214; R.R.Schmidt and B. Beitzke, Chem. Ber., 1983, 1'1.6,2].15 ; K. Mikami, K. Fujimoto and T. Nakai, Tetrahedron Letters, 7983, 24, 5'J,3 ; D.J. Peterson and J.F. Ward, l. Organomet. Chem.,1974, 66, 209 ; S. Julia, V. Ratovelomanan and H.

Chanh, C. R. Acad. Sci., Ser. C,1974, ZZB, 37I; R. H. Mitchell, T. Otsubo I and V. Boekelheide, Tetrahedron Letters, I9TS, 2I9 ; S. Apparao, A. Rahman, H. Ila and H. Junjappa, Tetrahedron Letters, 7982, 23,97I ; R. Gleiter, H. P. schaaff, u. Huber-Patz, H. Rodewald, w. Goetzmann and H. Irngartinger, l. Org. Chem., 1987, 52, 3979

100. G. Boche, Top. Current Chem.,1988,146,1

1,01. M.M. Kayser and O. Eisenstein, Can.l.Chem., l9ï'l,, Sg,24ST; M.M.

Kayser and G. Wipff, Can.l.Chem., 1982, 60, II92

102. E. Buncel and H. Wilson, Acc. Chetn. Res.,I9T9,12, 4I

103. B.K. carpenter in "Determination of organic Reaction Mechanism", p. 154, Wiley-Interscience, New York, 1.984

104. Y. Endo, S. Hizatake and K. Shudo, Tetrahedron Letters, L99I,92,2803 References 311

105. D.K. Bohme and G.I. Mackay, I. Am. Chem. Soc., I9B\, 10J,979 ; see also N.M.M. Nibbering in "Kinetics of Ion - Molecule Reactions",vol. 840, p. 165, (Edit. P. Ausloos), Plenum press, New york,1979

106. I.v. Aleksandrov and V.I. Gol'danskii," The Theory of Elementary

Processes in Condensed Media." in "sooiet Scientific Reaiews / Section B, Chemicøl Reuielus", Vol. 11, Chapt.1, (Edit. M.E. Vol'pin), Chur, Switz,

Norwood Academic Press, (1988)

107. P.B. Chock, Proc. Natl. Acad. Sci.(USA), 1922,69,1939; see also

H.Diebler, M. Eigeo, G. Ilgenfritz, G. Maas and R. winkler, pure AppL

Chem.,1969,20,93 !

108. R.A. Sheldon and J.K. Kochi ,l.Am. Chem. 5oc.,L970, 92, 4395; F.D. Greene, M.A. Berwick, J.C. Stowell ,l. Am. Chem. Soc., L970,92,867; K.- W. Lee, N. Horowitz,l. Ware and L.A. Singer, I. Am.Chem. Soc.,\977,99, 2622

109. H.Kiefer and T.G. Traylor, l. Am. Chem. Soc.,196Z , Bg, 6662

11.0. F.G. Bordwell, Acc. Chem. Res.,1988,21,456

111. R.W.Taft and F.G. Bordwell, Acc. Chem. Res.,1988, 21,469

112. S.E. Denmark and B.R. Henke, I. Am. Chem. Soc.,I989,l1-L,8032

113. L. Crombie, D.E. Games and M.H. Knight, l.Chem.Soc.,Chem.Comm., 1966,355 References 312

114. T.E. Hogen-Esch in "Ada. Phys. Org. Chem.", Vol.L5, p1.53-266 "Ion pairing Effects in Carbanion reaction." ;

115. J.H. Exner and E.C. Steiner, I. Am. Chem. )oc.,1974,96, I7B2

116. R. Cacciapaglia and L. Mandolini,/. Org. Chem.,I?BB ,59,2579

117. D.A. Evans and A.M. Golob, I. Am. Chem. Soc., L975,97, 4765

118. D. Liotta, M. saindane,R. Monahan III, D. Brothers and A. Fivush,

Syn. C0mm.,7986, 16, 1467

F

119. s.R. wilson, D.T. Mao, K.M. Jernberg, Tetrahedron Letters, rgzr,2ssg ; M. Bhupathy and T. Cohen, I. Am. Chem. 9oc., 1983,705,6978 ; M.T. Zoeckler and B.K. Carpenter, l. Am. Chem. Soc., L9BL, ]-}g, T661, ; T. Miyashi, A. Hazato and T. Mukai,l. Am. Chem. Soc., 1982, lO4,89I; T. Uyehara, K. Ohmori and Y. Kabasawa, Chem. Letts., 1994, IB79 ; I,s Sigmatropic Rearrangement: P.J. Battye and D.W. Jones, I. Chem. Soc.,Chem. Comm., 1984, 990

720. J. Enda and l.Kawajima, I. Am. Chem. Soc.,I9BS, !07,5495

121. J.G. Smith and J.M. Sheepy, l. Chem. Soc.,Chem. Comm., 1976,999

122. G.W. Cline and S.B. Hanna, I.Org.Chem.,1988, 53,3586

123. S.W. Staley and J.P. Erdman, I. Am. Chem. 1oc.,1970,92, 3832

724. J.F. Biellmann, J.B. Ducep and D. Shirlin, Tetrahedron, L980, ZG, 1249 References 313

125. J.F. Biellmann and J.L. Schmitt, Tetrahedron Letters,7973,4675

726. Y. Fujise; Y. Hasegawa and s. Ito, Tetrahedron Letters,1985, 26, gstl

127. D.I. Cram, A. Langemann, W. Lwowski and K.R. Kopecky, I. Am.

Chem. Soc., 1959, 8'].., 57 60

128. A.R. Katritzky, J. Arrowsmith, z. Bin Bahara, C. Jayarañ, T. siddiqui and S. Vassilatos,l. Chem. Soc.,Perkin Trans.'1., I9B0,28SI

129. A.S. Roth and D.B. Gollum, I. Am. Chem. Soc.,I9B9, !\\, 6772

Ê

130. P.G. Willard and M.J. Hintze, I.Am.Chem.50c.,1987,1,09,5539 and references cited therein.

131. L. Eberson, "Electron Transfer Reactions in organic Chemistry" , Chapter 8, Springer - Verlag , BerIin,1987

132. K.S. CherV F. Bertini, J.K. Kochi, I. Am. Chem. Soc., !973,95,1340

133. P. Renaud and M.A. Fox, /. Am. Chem. Soc.,1988, 110,5702

134. N. Hirota and S.I. Weissman, I. Am. Chem. 50c.,1964,9G,2532

135. M.A. Beno, H. Hope, M.M. Olmstead and P.P. Power, Organometallics, l.985, 4,2717

136. S.P. Patterson, I.L. Karle and G.D. Stucky, I.Am.Chem. 9oc.,1970,92, 1150 References 374

737. W. Zarges, M. Marsch, K.Harms and G. Boche, Chem.Ber.,Ig8g,'1,22, 2303

138. L.A. Paquette, J.P. Gitday and C.S. Ra, I. Am. Chem. Soc., I9BZ,'!,09, 6858 ; L.A. Paquette and J.P. Gilday, I. Org. Chem., 1988,53,4972 ; L.A. Paquette and C.S. Ra, /. Org. Chem.,I9BB, Sg,49TB

139. E. Grovenstein,Jr., and A.B. CottinghamJ. Am. Chem. ]oc.,1972,99,

1881

140. R.G. Visser, L. Brandsma, H.J.T. Bos, Tetrahedron Letters, l9BT,22,

2827 g

141. C. Huynh and G. LinstrumelLe, l. Chem. Soc,.Perkin Trans. 2, I9Bg, 1133

1,42. A. Maercker, M. Eckers and M. Passlack, I. Orgønomet. Chem., 1980, 186,L93

143. J.C. Dalton and B.G. Stokes, Tetrahedron Letters, t97S,3LT9

144. G.W. Gokel and H.D. Durst, Synthesis ,1976,168

145. D.I. Sam and H.E. Simmons, /. Am. Chem. Soc., LVTZ, 94, 4024

146. C.L. Liotta, E.E. Grisdale and H.P. Hopkins,Tetrahedron Letters, 1975,

4205 References 315

L47. L.B. Young, E. Lee-Ruff and D.K. Bohme, l. chem. soc.,Chern. Comm., 7973,35

148. D.s. Kemp and K.Paul, l. Am. chem. soc., 1970, 92,2ssg; see also D.H. Hunter, W. Lee and S.K. Kim, l. Chem. Soc.,Chem. Comm., 1974, ],0].g

149. D.H. Hunter, M. Hamity, V. Patel and R.A. patel, Can. l. Chun., IgTg, 78,I04

150. A. Loupy and J. Seyden-Penne, Tetrahedron Letters,I7ZB,2STI

151. T. okuyama, K. Kimura and r. Fueno, Bull. chem. soc. (lpn.), rggz, ç 55,2295

152. H. Handel, M.- A. Pasquini and J.-L. pierre, BuIl. soc. chim. (Fr.) II, 1980,351

153. P.R. Markies, o.s. Akkerman, F. Bickethaupt, w.I.I. smeets and A.L. Spek, I. Am. Chem. 9oc.,7988, t1..}, 4284

154. T.L. Brown, /. Orgønomet. Chem., 1966, S, L9]- ; M. Schlosser in "struktur und Reactioität Polarer organometalle", springer-verlag,

BerlirU L973

155. M. Andrac, F. Gaudemar, M. Gaudemar, B. Gross, L. Miginiac, p.

Miginiac and C. Prévost, Bull. soc. chim. (Fr.), 1963, 138s and references cited therein.

156. G.W. Gokel and H.D. Durst, Aldrichimica Acta ,1976,9 ,3 References 316

157. R. Peraud, H. Handel and J.L. pierre, BuII. soc. chim. (Fr.) Il, rgg0, zgg

158. M.D. Rozeboom, J.P. Kiplinger and J.E. Bartmess, I. Am.chem. soc., 1984,1.i06,1025

159. J.H. Bowie, Acc. Chem. Res.,l9B0,19,76

160. M.E- ]ones, s.R. Kass, ¡. Filley, R.M. Barkley and G.B. Erlison, l. Am Chem. Soc., 7985, 107, I09

161. M.E. Jones and G.B. Ellison I. Am. Chem. )oc.,1989, L]-]-,1645

ç

'J'62. G. wittig and L. Löhmann, lustus Liebigs Ann.chem., rg4z, ss}, 260

763. A. Maercker, Angew. Chem. IntI. Edit. (EngI.), I9g2,26, g72,l Angew.Chem., 1987 , 99, 1002)

1,64. P. Schorigen, Dtsch. Chem. Ges. 8er.,L924, 57,1634 ; ibid.,,Ig2S, Sg, 2024 ; ibid.,1926,59,2510

165. G. wittig -and L. Löhmann, lustus Liebigs Ann. chern.,7942, ss}, 260

166- w. schlenk and E. Bergmann, fustus Liebigs Ann. chem., r92g, 464, is

167. C.R. Hauser and S.W. Kantor, I. Am. Chem. Soc.,I9S1,, Zg,I43Z

168. R.B. woodward and R. Hoffmarrt, Angat¿. chem. Intt. Ed. (Engl.), 1969,9,797 References 3L7

769- E. Buncel, "carbanions: Mechanistic and lsotopic Aspects" in "Reaction Meehanisms in organic chemistry, Monograph 9", p.834 (Edits. C. Eaborn and N.B. Chapman), Elsevier, Amsterdall:r, \975.

170. M.J.S. Dewar and C.A. Ramsden,l. Chem. Soc., Perkin Trans.'1, 1974,

1333,1839. ; See also J.F.Garst and C.D.Smith,l.Am.Chem. 50c.,1973,95,

6870

77'J,. M.I.S. Dewar in "The Molecular Orbital Theory tf Organic Chemistry", McGraw Hill, New York, 1969

172. G. Wittig and E. Stahnecket, lustus Liebigs Ann.Chem.,l9S7, G}S, 69 Ê

I73. U. Schöllkopf , Angaa. Chem.,7960,72, 570

174. U. Schötlkopf and W. Fabian, lustus Liebigs Ann. Chem.,I96t, G42,I ; U. Schöllkopf and H. Schäfer,lustus Liebigs Ann. Chem.,L963,663,22; lJ.

Schöllkopf and D. Walter, lustus Liebigs Ann. Che¡n.,1962, GS4, 27

I75. B.I. Millard and T.S. Stevens, l. Chem. 50c.,1963,3397

176. I. Cast, T.S. Stevens and J. Holmes, l. Chem. Soc., L960, 3521 ; B.l. Ballard and T.S. Stevens, l. Chem. Soc., 7963, 3397 ; J.H. Brewster and M.W. Kline, I. Am. Chem. Soc.,\952,74,5L79

I77. U. Schöllkopf , Angew. Chem. IntI. Ed.( Engl.),1970,9,763

178. E. Hebert, Z. Welvart, M. Ghelfenstein and H. Swarc, Tetrahedron Letters,7983,24 ,7381 References 318

I79. P.T. Lansbury, V.A. Pattison, J.D. Sidler and J.B. Barbier, l. Am. Chem. Soc.,1966,88,-78

180. C. Rüchardt, Angew. Chem. IntI. Ed. (Engl.),1970 ,9, 830

181. J.F. Garst and C.D. Smith, I. Am. Chem. Soc.,I9T3,95, 68T0

182. R. Kaptein, l. Chem. Soc. D,1,97L,732 ; R. Kaptein, Adrs. Free Radical Chem.,1975, 5, 319

183. U. Schöllkopf , Ind. Chim. Be\ge,1971., i6, 1057

ç

184. H. Felkin and C. Frajerman, Tetrahedron Letters, 1977, 34Bs

185. R. Briere, M. Cherest, H. Fetkin and C. Frajerman, Tetrahedron

Letters, 7977, 3489

186. G. Wittig and H. Schlor, Soumen Kemistilehti,I}SB, g1-,2

L87. P.T. Lansbury and V.A. Pattison, I.Org.Chem.,1962,27,\923

188a. V. Rautenstrauch, G. Büchi and H. Wüest, I. Am.Chem. Soc., 1974, 86,2576; 1-88b V. Rautenstrauch, l. Chem, Soc.,Che¡n. Comm.,1970, 4

189. G.M. Whitesides, C.P. Casey and J.K. Krieger, I. Am. Chem. \oc.,7971, 93,1379

190. T.F. George and J. Ross, l. Chem. Phys.,197t,55, 3851 References 379

191. P.C.H. Eichinger, J.H. Bowie and T. Blumenthal, l. Org. Chem., 1986, 5'1.,5079

792.1.8. Cummings and P. Kebarle, Can. l. Chem.,IgTB, 56,I

193. M. Meot-ner, L.W. Sieck, l. Phys. Chem.,1986, gO, 6687

194. W.T. Nafl R.N. Compton and C.D. Cooper, I. Chem. phys.,'J.gT2, SZ, 1303 ; an earlier reference yields a somewhat lower value for the electron affinity of benzaldehyde of 0.42 + 0.010 ev w.E. wentworth and E. Chen, l. Phys. Chem.,1967,7L, 1929

ç

195.P.c.H. Eichinger and J.H. Bowie, l. chem. soc.,perkin Trans. 2, 1.987, 1.499

196. T. Nakai and K. Mikami, Chem. Rea.,1986, 86, 8Bs

797. Y.-D. Wu, K.N. Houk and ].4. Marshall, I. Org. Chan., \990, SS, t42L

198. H. Felkin and A. Tambuté, Tetrahedron Letters, L969,821

199. R.J. Waugh, R.N. Hayes, P.C.H. Eichinger, K. M. Downard and J.H. Bowie, I. Am. Chem. 5oc.,1990, 112,2537

200. J.E. Bartmess, J.A. Scott, R.T. Mclver,Jr., I. Am. Chem. Soc., L979,10"L, 6047

201. I.C. Kleingeld and N.M.M. Nibbeing, Tetrahedron,IgS4, q0, ZTB9 References 320

202. R.P. Lutz, Chem. Rea.,l9B4, gL,20S

203. s. Ingemann, N.M.M. Nibbering, s.A. sulivan and c.H. Depuy,l. Am.

Chem. Soc., 1982, 104, 6520

204. ÂHoacia (CHz=CHz) = 7700 kJ mofl j ÁHoacid (CzHo) = 1763 kJ mol-l C.H. DePuy, V.M. Bierbaum and R. Damrauer, l. Am. Chem. Soc., 1984, L06,4051

205. A.P. Bruins, AnaI. Chem.,1979,57,962.

206. M.B. Stringer, D. J. Underwood, J.H. Bowie, J.L. Holmes, A.A. ts Mommers, and ].E. Szulejko, Cøn. l. Chem., \986,6q,264.

207. For a review see J.H. Bowie, M.B. stringer, R.N. Hayes, M.J. Raftery,

G.J. Currie, and P.C.H. Eichinger, Spectros. IntI. l.,I9BS, 4, ZTZ.

208. P.C.H. Eichinger, J.H. Bowie and R.N. Hayes, Aust. l. Chem.,19ï9,42, 865

209. C.C. Price- and W.H.Synder, I. Am. Chem. Soc., L96!, Bg, t77J

2\0. u. schöllkopf, K. Fellenberger, and M. Rizk, lustus Liebigs Ann. Chem., 1970, 734, 106

21,1,. G. Wittig and E. Stahnicker, lustus Liebigs Ann. Chern., \957, 605, 69

212. I.H. Bowie and B. Nussey, l. Chem. Soc., Chem. Comm.,t9T0, !7; Org Mass Spectrom., 1970, 3, 933 References 327

213. I.H. Bowie, Aust. l. Chem.,1977,24,989

274. J.H. Bowie and S. Janposri, Org. Mass. Spectrom.,1976,l1-, IZ90

275 J.H. Bowie and M.B. Stringer, Org. Mass Spectrom.,I9BS,20,1.3B.

276. C.E. Dykstra, A.J. Arduengo and T. Fukunaga,I.Am.Chem.Soc., I9TB,

100, 6007 ; M.I.S. Dewar and M. Shanshal, l.Am.Chem.Soc., 1969, 91,, 3654

277. D.F. Hunt, A.B. Giordani and J. Shabanowitz, I. Org. Chem.,I9B2,47, 738

!

218. M.B. Stringer, J.H. Bowie and J.L. Holmes, I. Am. Chem. Soc., 1986, 108, 3888 ; M.B. Stringer, J.H. Bowie and G.J. Currie, l. Chem. Soc., perkin Trans. 2,1986, L82I

279- G. Klass, J.c. sheldon and J.H. Bowie, l. chem. soc., perkin Trans. 2, 1983, 1337

220. A. Hadjiantoniou, L.G. Christophorou and f.G. Carter, l. Chun. Soc.,

Faraday Trnns.,2,1973, 69, 1704; M.D. Rozenboom, J.P. Kiplinger and J.E. Bartmess, I. Am. chem. soc., 1984, L}G, 1025 (electron affinity 40 t 40 kI mol-1)

221,. C.H. DePuy, V.M. Bierbaum and R.J. Damrauer, l. Am. Chem. Soc., 7984,106,4051 References 322

222. (Gas Phase) R.E. Lee and R.R. squires, I. Am. chem. soc., 19g6, Los, 5078 ; (solution) s. winstein, M. ogliaruso, M. sakai and J.M. Nicholson, I. Am. Chem. 9oc.,7967, 89, 3656

223.J.K. Stille and K.N. Sannes, I.Am.Chem. Soc.,\972,94,g494; M.N. Paddon-Row, D.N. Butler and R.N. Warrener, /. Chem. Soc., Chem. Comm., !976, 747

224. R.M. Minyaev, V.I. Minkin, N.S. Zefirov, V.I. Natanzon and S.V

Kurbatov, Zh. Org. Khim., 7982,'/.,8, 3

225. J.c- sheldon, J.H. Bowie and P.c.H. Eichinger, l. chem. soc.,perkin c Trans. 2,7988,1263

226. V. Rautenstrauch, HeIa. Chim. Acta, 1972, SS, 594

227.M.8. Stringer, J.H. Bowie, P.C.H. Eichinger, G.I. Currie, l. Chem. Soc.,Perkin Trans. 2, 7987,385 ; a). When examining kinetic isotope effects, it is considered that the ratio of ion abundances (h/rù is approximately equal to the kinetic isotope effect (k1,/k2) for intramolecular fragmentations. P.]. Derrick, Mass Spectrom. Raninls, IgB3, 2, 2Bs

228. A.L.J. Beckwith in "Essøys on Free Radicøl chemistry" Chemical

Society Special Publication No. 24 Chapter 9 "Some Aspects of Free Rødical Rearrangement Reactions" p.239-69; A.L.I. Beckwith, CJ. Easton, A.K. Serelis, A. Lawrence, /. Chem. Soc.,Chem. Comm., I9BO,4B2.

229. J.E. Baldwin, l. Chem. Soc.,Chem. Comm., 1926,234 References 323

230. R.A.I. O'Hair, S. Gronert, C.H. DePuy and J.H. Bowie, I.Am.Chem. 9oc.,7989, 1LL, 3105

231. A.B. Ernst and W.E. Fristad, Tetrahedron Letters,I98S,26, gZ6I

232. W.E Fristad and J.R. Peterson, I. Org. Chem.,'J,985,50, L0; other examples involving the addition of a radical to an anionic species, see : J.F. Bunnett, Acc. Chem. Res.,7972, 5, 139; R.A. Rossi and J.F. Bunnett, I. Am. Chem. Soc., 1972, 94, 683; R.A. Rossi and R.A. Alonso, I. Org. Chem., 1989, 45, 1239; J.F. Wolfe, M.C. Sleevi and R.R. Goehring, I. Am. Chem. Soc., 1980,102,3646

F

233. R. Lalande, B. Maillard and M. Cazaux, Tetrahedron Letters, 1969,74s ; cleavage of ester radical anion, see M.D. sevilla, K.M. Morehouse and S.Swartz, l. Phys. Chem., I98!, gS, 923 ; P. Cremashi, G. Morosi,

M.Simonetta, Angew. Chetn. Intl. Ed. (Eng.),I98L,20, 673

234. c.A. Coulson, A. Maccoll, L.E. suttorç Trans. Faraday soc., r9s2, 49, 106

; see also RJ.W. Le Fevré, Ado. Phys. Org. Chem.,L96S, g, L

235. G.J. Currig, J.H. Bowie, R.A. Massy - Westropp and G.W. Adams, /. chem. soc.,Perkin Trans. 2, 1988,403 ; a). c/. addition of an alkyl radical to the nit¡o enolate ion, G.A. Russell and R.K. Khanna, I. Am.Chem. Soc., 1985, I07, 1450; G.A. Russell and R.K. Khanna, Tetrahedron,ISSS, q1,, 4\33 ; G.A. Russell and D.F. Dedolph, I, Org. Chun.,I98S, S0,2gZ8

236. G.W. Adams, J.H. Bowie and R.N. Hayes, l. Chem. Soc.,Perkin Trans 2,7ggg,2l5g References 324

237. M.I. Raftery and J.H. Bowie, IntI. l. Mass spectrom. Ion processes, r9gg, 85, 767

238. M. Newcombe and J.L. Esker, Tetrahedron Letters, 199L,32, j.035

239. L.A. Warren and S. Smiles, l. Chem. Soc. ,I9i0, 956 and I\ZZ.

240. For a review, see: w.E. Truce, E.M. Kreider, w.w. Brand, organic

Reactions; Iohn wiley and sons; New York, 1970; vol L8, p. gg.see arso D.M- schmidt, G.E. Bonvicinio , I. org. chem. 1984, 49, 'J.664 and references cited therein.

g

241. F. Terrier, Chern. Rea.,1982,82,77 ; G. Artamkina, M.p. Egorov and

I.P. Beletskaya, Chem. Rer¿., 1982, 82, 427

242. E. Grovenstein, Jr. and P.{.Lu,l. Am. Chem. Soc., !982,104, 66gI

243. N.A. suttle and A. williams, l. chem. soc.,perkin Trøns. 2, 1983, 2, 1369

244. G.W.H. Cheeseman and A.A. Hawi, l. Heterocycl. Chem.,1983, ZO, SgI

245. C.F. Bernasconi and K.A. Howard, I. Am. Chem. Soc.,I9B2,10q,7248

246. K. Mutai and K. Kobayashi, Kokngøku Toronlcni Koen yoshishi ,1979, 86-7 ; Chem. Abst. ,92:180332t

247. v.N. Knyazev, v.N. Drozd and V.M. Minov, Tetrahedron Letters, 7976,4925 References 325

248. T. Abe, BUII. Chem. Soc. (lpn.),1983, 56, 1206

249- B.K. Carpenter ín "Determination of Organic Reaction Mechanisms", chapter 9 , wiley - Interscience, New york, 1984 and references cited therein.

250. M. Makosza and T. Glinka, I. Org. Chem.,I9B3,48, g860

251. L.M. Grandos and N. Radha, I. Org.Chem.,1983,4g, ST0

252. E. Matsumura, M.Ariga and Y. Tohda, BulI. chem. soc. (lpn.),19g0, s3,

2891 E

253. D.J. Sam and H.E. Simmons, /. Am. Chem. Soc., !974, gZ, ZZS2

254. F. Terrier, G. Ah - Kow, M.-J. Pouet and M.p. simonnin, Tetrahedron Letters, 1976, 227

255. W.P.M. Maas, P.A. Van Veelery N.M.M. Nibbering Org. Mass 9pectr0m.,1989,24,il6; W.P.M. Maas, N.M.M. Nibbering

"Adaances in Mass Spectrometry", (Ed. p.Longevaille), Vol 11A, p.g5g-9,

Heyden and Sons, London, 1989.

256. K.K. Murray, T.M. Miller, D.G. Leopold, W.C. Lineberger, l. Chem. Plryt., L986, 84,2520

257.1.H. Bowie and M.B. Stringer, Org. Mass Spectrom.,1985, ZO,IgB Refermces 326

258. T. Toru, S. Kurozumi, T. Tanaka, S. Miura, M. Kobiashi and S.

Ishimoto, Synthesis, \974, 86Z

259. C.K. Kim, I. Lee and B-S. Lee,l.phys.Org.Chem.,1997, q,3Is

260 G. Scheuing, Dtsch.Chem.Ges.B er., 1923, ffi , 252

26L. J.A.G. Roach, J.A. Sphon, D.F. Hunt and F.W. Crow, l. Assoc. Off. Anal. Chem., L980, 63, 452; R.C. Dougherty and J.D. Wander, Biomed.

Mass Spectrom.,1980, 7, 401.

262. Yariation in the spectra obtained using different mass spectrometers Ê have been reported : B. Arbogast, W.L. Budde, M. Deinzer, R.C. Dougherty, I. Eichelberger, R.D. Foltz, C.C. Grimm, R.A. Hites, C. Sakashita, E. Stemmler, Org. Mass Spectrom.,L990,25,I97

263. P.A. Schulz, R.D. Mead, P.L. Jones and W.C. Lineberger, I. Chem. Phys.,L982,77,1153

264. G.I. MacKay, R.S. Hemsworth, D.K. Bohme, Can. l. Chem., 1976, 54, 1624

265. I.E. Bartmess "The Gas Phase Acidity Scale", The University of Tennessee, 7987

266. M. Eckersley , J.H. Bowie and R.N.Hayes , IntI. l. Møss Spectrom. Ion Processes , \989 ,93 ,799 References 327

267. LHoaç¡¿ (MeOCIIzCtri¡= 1557 kJ mol-1 ; ÁHo¿ç¡¿ (MeC!þCN)= 1570 kI mol-1 ; ÂHoacid (CII¡COzMe)= 1557 kJ mol-1 values taken from ref. 200

268. R.A.J. O'Hair, I.H. Bowie and R.N. Hayes, Rapid Commun. Mass Spectrom., 7988, 2, 275

269. R. Louw and E.C. Kooymans, RecI. Trar¿. Chim, Pays-Bas,196T, g6, 704i,

270. Unsuccessful attempts to investigate the ketene radical anion have been reported by l.g. collin and R. Locht, Intl. l. Mass spectrom. Ion phys., 1969, 37, 465 and V.I. Khvostenko, V.S. Falko and I.Kh, Zh. Org. Khim.,

1975,\1,665 Ë

271,. M.I. Alexander, I.H. Bowie and R.N. Hayes, Aust.l. Chem.,1990, 49, 2027

272. M.S. Silver, P.R. Shafer, I.E. Nordlander, C. Rüchardt and J.D. Roberts, I. Am. Chem. 5oc.,1960, 82,2646; A. Maercker and K.Weber, lustus Liebigs Ann. Chem., 1972, 43, 756

273. M.l.P. Harger, l. Chem. Soc., Chetn. Comm., 1979,990

274. K. Goda, R. Okazaki, K. Akiba and N. Inamoto, BulL Chem. Soc.

Çpn),1978, 5L,260

275. R.J. Waugh, P.C.H. Eichinger,J.H. Bowie and R.N. Hayes, Intl.l. Mass

Spectrom. Ion Proc., 7990, 96, 347

276. S.D. Lee, T.H. Chan and K.S. Kwon, Tetrahedron Letters,1984,3399 References 328

277. R.N. Nobes, D. Poppinger, w.K. Li, L. Radom, in "Comprehensirse Cørbanion ehemistry", (Edits. E. Buncel and T. Durst ), Elsevier, Amsterdam, 1987, Part C.

278. D.M. Hirst, J.D. Hopton and J.W. Lin¡rett, Tetrahedron Suppl.,!963, 2,

!5 ; J.J. Blomfield, ].R.s. Ireland and A.P. Marchand, Tetrahedron Letters, 'J,968, 5647, D. Bryce-Smith and A. Gilbert, l. Chem. Soc, Chem. Comm., !968, 1,319 ; D.L. Dean and H.Haft, l. Am. Chem. 5oc.,I972,94,687; H.-D. Scharf and R Klar, Chem. 9er.,1972,105,575

279. R.C. Haddon, Tetrahedron Letters, 1972, 3897 ; ]. Fleischhauer, M. Beckers and H.D. Scharf, Tetrahedron Letters, 1973, 4275 ; P. Lindner, Y. ç Ohrn, J.R. Sabin, Intl. l. Quantum Chem., 5ymp.,I973,7,26I ; N.H.F. Beebe, J.R. Sabin, Chem. Phys. Letts., \974, 24, 389 ; R.C. Haddon, D. Poppinger and L. Radom, I. Am.Chem. 9oc.,1975,97,1645

280. C.H. DePuy, J.J. Grabowski, V.M. Bierbaum, S. Ingemann, N.M.M Nibbering , I. Am. Chem. Soc., 1985, 107, 1093

281. I.H.J. Dawson and N.M.M. Nibbering, IntI. l. Mass Spectrom. Ion

Phys.,1980, 33,- 3

282. H.G. Viehe, R. Merényi, L. Stelle and Z. Janousek, Angew. Chem., L979,97,992

283. J.l. Grabowski, C.H. DePuy, V.M. Bierbaum,l.Am. Chem. Soc., 1983, L05,2565

284. R.B. Bates,T.J. Siahaan and K. Suvannachut,l.Org.Chem.,I990, 55,1328 References 329

285. J.s. Davies, v.H. Davies and C.H. Hassarr, f.chem. soc. (c), 1969,'J.BT3 ; C.H. Hassall and B.A. Morgan, l.Chun. soc. (D), 1920, r34s ; C.H. Hassall and B.A. Morgan, l. Chem. Soc., Perkin Trans. 1_,1973,2853 ; E.V. Vedejs, W.H. Miller and J.R. PribistU I. Org. Chem.,1983,48,36'J.1.; E.V. Vedejs and J.R. Pribish, l. Org, Chem., 1988,53,1593

286. R.W. Taft and R.D.Topsom, Prog. Phys. Org. Chem., I7BZ ,16,I ; T. Yokoyama, I. Hanazome, M. Mishima, M.J. Kamlet, R.W. Taft, I. Org. Chem. L987,52, 163; T. Yokoyama, T. Komaki, H. Sato, R.W. Taft and F. Anvia, I. Org. Chem., \989,54, 3983

Ê 287. R. Helwig and M. Hanack, lustus Liebigs Ann. Chem.,l977, 61.4

288. W. Lwowski and A. Milling, It.,l. Am. Chem. Soc.,L9Zg,95,609

289. B. Franzus, M.L. Scheinbaum, D.L. Waters and H.B. BowLin, l. Am.

Chem. Soc., 1976, 98, 124'1,

290. P.C.H. Eichinger, I.H. Bowie and R.N. Hayes, Aust. l. Chem.,Ig9g,42, 865

292.1. H. Bowie and T.Blumenthal , I. Am. Chem. Soc.,I9ZS ,97 ,295L ; I. E. Szulejko ,I. H. Bowie , I. Howe and J.H. Beynon ,IntI. l. Mass Spectrom. Ion Phys.,1980 ,73 ,76

293. V.G. Instruments, Wythenshaw, Manchester, M23 9LE, England

294. I.H.I. Dawson, T.A. Kaandorp and N.M.M. Nibbering, Org. Mass Spcectom.,1977,11, 330 ; M.B. Stringer, D.J. Underwood, J.H. Bowie,J.L. References 330

Holmes, A.A. Mommers and j.E. Szulejko, Can. l. Chem,,1986,64,T64 and, references cited therein.

295. W.C. Still, M. Khan and A. Mitra, l.Org.Chem,,I9Zg,43,2923

296.R.4. Ellison, R.Griffin and F.N. Kotsonis l.Organometallic Chem., 1972, 36(2), 209-2Ig

297. G. wittig and R. Clausnizer, lustus Liebigs Ann. chem., Lgs4^, sgg, 1.67.

298. D.s. Tarbell, "organic Reactions," wiley, New york, r9M, yor.2, p.26 ç

299. rcf .298, p 27

300. R. Delaby and L. Le Conte, BulI. Soc. Chim. Fr.,I93Z,4,T3g

301. W. Kimmel and A.C. Cope, I.Am. Chem. ioc.,1943, 65, T9g2

302. R Raddiffe, and P. Rodehurst, J. Org. Chem.,IïTO,3S, 4000

303. G. Klass, ].C. sheldon, and J.H. Bowie, l. chem. soc., perkin Trans. 2, 1983,1337

304. L.E. Friedrich and G.B. schuster, I. Am. chem. soc., !972,94, !!93

305. R.H. Baker, K.H. Cornell, and M.J. Cron,l.Am.Chem. Soc., I94g,70, 7490 Refercnces 331

306. G.G. Smith and K.J. Voohees, I. Org. Chem.,1970,35, ZlB2

307. M.R. Delaby, Compt. Rend.,I9JZ, Lg4,124g

308. w.H. Carothers C.F. Bickford and G.]. Hurwitz, I. Am. Chem. soc., 1927,49,2909

309. A.I. Meyers, R. Munavu and J. Durandetta, Tetrahedron Lett., Lgrz, 3929

310. J. Harvey, I.M. Heillbron, D.G. Wilkinsoru l. Chm. Soc., L930, 423

É

311. J. v. Braun, Ber. Dtsch. Chem. Ges.,1934,6Z,2IB

312. E.D. Thorsett and F.R. Stermitz, Synthetic Comm.,1972, Z, 3Ts

313. E. Bamberger and w. Lödter, lustus Liebigs Ann. chem.,rggs,2gg,7s

314. F. Strauss and L. Lemmel, Ber. Dtsch. Chem. Ges.,'J,92]-,54,32.

315. R. Stoermer and E. Laage, Ber. Dtsch. Chem. Ges.,I9\7, S0, 988

31,6. L. Brandsma and H.D.Verkruijsse, "synthesis of Allenes, Acetylenes and Cumulenes, A. Laborøtory Manual", Elsevier, Amsterdam, 1.98!, p.

159.

317. W.H. Carothers and G.]. Berchet,l. Am. Chem. 50c.,19j2,5q,4066

318. W. Hückel and W. Egerer, lustus Liebigs Ann. Chem.,196I, 6qS, 162 References 332

319. N. Meyer and D. seebach, Angew. Chem. Int. Ed. Engr., r97g, \7, szr

320. G.R. Lappin, I. Am.Chem. Soc.,\949,21^,3966; [Lappin has reported that phenylallene "on exposu¡e to air rapidly turned yellow and within a few hours polymerised to a soft brown solid.,']

321. P.F. Donaghue, P.Y. white, J.H. Bowie, B.D. Roney, and H.]. Rodda,

Org. Mass Spectrom.,1969,2,'1,06\; J.H. Bowie and p.y. White, Org. Mass.

Sp ectrom., !972, 6, 135.

322. G.I. Currie, J.H. Bowie, R.A. Massy - Westropp, and G.W. Adams, / Chem. Soc., Perkin Trans. 2, I9BB, 409 ç

323. A.P. Best and C.L. Wilson, I. Chem. 50c.,1.946,239 ; Vogel p 59g

324. A.r. vogel ; "Practicøl organic chemistry". Longmans and Green,

London, 3rd ed., p.758 ,1956

325. W.H. Watanabe and L.E. Conlor., I. Am. Chem. Soc., LISZ,79, Zg2g

326. r. Heilbron and H.M. Bunbury in "Dictionary of organic compounds",

Eyre and Spottiswoode, London, Vol. 1, 1953

327. A.E. Favorskii and M.F. schostakowski, Zhur. obshch. Khim., 1949, L3, 10 ; U.Gen.Chem.,I9M,38,3 (Engl.)l

328. C.D. Hurd and M. A. Pollack, I.Am. Chem. 50c.,1938, 60,Ig0g

329. S.G. Powell and R. Adams, I.Am.Chem. 5oc.,I920, q2,648 References 333

330. A. Reppe and v. Mitarbeiten, lustus Liebigs Ann. chem., r9s6,60t, g1

331. W.L. Ruigh and R.T. Major, I.Am.Chem. Soc.,I93I,Sg,2662

332. R.C. Huston and A.H. Agett, I. Org. Chern.,Ig41.,6, !29

333. M.J. Raftery and J.H. Bowie, Intl. l. Mass spectrom. Ion processes, rggT, 79,267

334.1. Collonge and P. Boisde, Bull. Soc. Chim.(Fr.),1956,89, IZ32

335. V. RautenstrauctL Helo. Chim. Acta,I97Z, SS, 594 Ê

336. A. Kh. Khusid, G.v. Kryshtal, v.F. Fucherov, L.A. Yanovskaya, Iza.

Akad, Nauk SSSR,Ser.Khim., 19T6,4LS-9 ; (EngL Trans.) BuU. Acad. Sci., ussR, 1976,396.

337. E.L. Allred, J. Sonnenberg and S. Winstein, l. Org. Chan., L960,25,26

338. K. Alder and F.H. Flock, Chem. Ber., \956,89, !722

339. P. Cressory Bull. Soc. Chim.(Fr,),1964,26L8

340. I.M. Heilbron, E.R.H. Jones, J.T. McCombie, and B.C.L. Weedon, / Chem. Soc.,1945,86

341. F.F. Blicke and H.M. Kaplary I.Am. Chem. Soc., t943,65,1969

342. G.M. Bennett and F.M. Reynolds, l. Chem. Soc., !935, L3L References 334

343. R.H. Manske, I. Am. Chem. Soc.,I93l, 53,II04

U4. H. Böhme and T.P. TeIE, lustus Liebigs Ann. Chem., 1959, 620,I.

345. H.S. Schultz, H.B. Freyermuth, and S.R. Buc, /. Org. Chem.,196g,2g, TL4O

346. W.E. Truce and L.B. Lindy, I. Org. Chem.,196!,26, !46J

347. T. Nambara, N. Matsushisa, Yaskugaku Zasshi, 196Z, gg, 642; G.M. Bennett and F. Heathcoat , l. Chem. Soc.,1929,2Tz

c

348. G.M. Bennett and J. Mosses, l. Chem. Soc.,I9J1-,2956

349. G.K. Abdullaev and F.A. Rustamov, lJch. Zap. Azerb. Gos. Llnir¿. ser.

Khim. Nøuk, I9&, 1.54,1-.60

350. E.N. Prilezhaeva, N.P. Petukhova and M.F. shostakovskii, Dokl.

Aknd. Nauk USSR, 7964, L54, T60

35L. M.I. \J¿z4rov, Sh.T. Akhmedov and M.A. Akhundova, lIch. Zap.

Azerb. Gos. Unio. Ser. Khim. Nauk, 1923, 4L

352. M.I. Nazarov, sh.T. Akhmedov and A.B. Kuliev, Izu. vyssh. Llcheb

Zaaed. Khim. Teknol. ,1973,16, 1530 ; Chem. Abst. L974, B0 : 47587p

353. M.I. Nazarov, Sh.T. Akhmedov and A.B. Kuliev, Zhur. Org. Khim., L973,9,28 (Eng p 25) References 335

354. K.w. Rosenmund and G. weiler, Ber. Dtsch. chem. Ges., 1923, s6, 1481

355. w.H. Bentley, E. Haworth and w.H. perkins þ.J. Chem. soc., rg96,69, 76L

356. S.G. Powell, I. Am.Chem. Soc., L923,45,2708

357. W.E. Hanford and R. Adams, I. Am. Chem. Soc.,I9iS, 57,9ZL

358. I.s. Isaev, v.N. Domrachev, M.I. Gorfinkel and V.A. Koptyug, DokI

Aknd. Nauk, 197'J.,20't, 10'1,4 (Engl. Trans.) q

359. Mitsubishi Chemical comp. Ltd., Br, patent gz}Bgs ; chem. Abstr. L9U,6'1,: I3240e

360. G.A. Ropp, l. Chem. 50c.,1950,22,2299

36l.. M. Fields, M.A. Leaffer, s. Rothchild and f. Rohan, I. Am. chem. soc., 7952,74,5499

362. F.K. Signaigo and P.L. Cramer, I.Am.Chem. Soc.,1933, SS, gg26

363. F. ullmann and J.B. uzbachian, Ber. Dtsch. chem. Ges., 1903,96, LzgT

364. R. Leuckart, l. Prakt. Chem., IB90, 41', IBT ; D.S. Tarbell and D.K. Fukushima, "Organic Synthesis", Coll. Vol. 3, 2nd. Edit., p. 809, Wiley, New York, 1941 References 336

365. c.D. Nenitzescu and N. scarlatescu, Ber. Dtsch. chem. Ges., r93s, 6g, 587

366. R.o. Clinton and s.C. Laskowski, l. Am. chem. soc., !948,70,3136

367. A. A. Goldberg and W. Kelly , l. Chem. Soc,,1947,I3T4

368. A. Adkins, R.M. Elofson, A.G. Rossow and C.c. Robinson , I. Am. Chem. 9oc.,7949 ,7'1, ,3624

369. c.A. Bischoff and P. walden , lustus Liebigs Ann. chem., tgg4 ,2Tg ,

100 t

370. E. schmidt and R. sachtleben , lustus Liebigs Ann. chem., rgzg ,1gg , 106

371. R. C. Fuson , H. Gray and J. J. Gouza , I. Am. Chem. Soc., !939 , 6l , L937

372- w. Reppe and v. Mitarbeitern, lustus Liebigs Ann. chem.,rgss, sg6 , 50

373. Y. Izumi, M. Imaidâ, H. Fukawa and s. Akabori, BuIt. chem. soc.

]pn.),1963,36 ,2'J,

374. A. Richard , Ann. Chim. Phys.,I9I0, 2L , g9B

375. S.F. Acree , Ber. Dtsch. Chem. Ges.,1904 , gZ, 2T6Z References JÔ/

376. S.H. Langer, S. Connell and I. Wender, I. Org. Chem. ,IgSg,2g, S0

377. D. J. costa , N. E. Boutin and J. G. Riess Tetrahedron ,7974 , g0 , gTgg

378. w. sakami , w. E. Evans and s. Gurin , I. Am. Chem. soc., 1947 ,69 , 1110

379. T.M. Bu¡ton and W.B. Fife, I. Am. Chem. Soc.,IgS2,74 ,3935

380. R. Anschütz and w. Bertram, Chem. Ber.,1903,g6,466. ( Et ester)

381. w. wislicenus and G. Andrejew, lustus Liebigs Ann. chem., rg6z, ç \33,2U

382. C.E. Rehberg and M.B. Dixory I.Am. Chem. Soc.,I9S0,72,1glrg

383. RBrettle and D.W. Latham, I. Chem. Soc. (C),1969,906.

384. P.D. Bartlett, M. Kuna and P.A. Levene, l. Biot. chem.,Lg3z,j.Lg,50g.

385. G. Castaldi, c. Giordano, Eur. pat. App. 299 4u (Chem. Abs. 19g9,111,

57472a); w.L. Richardson, w.T. stewart and M.R. Barush rJSp 3,s63,7rs

386. B. Neises and w. steglich, Angew. chem. IntL Ed. Engl., rgrï, tz, szz.

388. P. Karrer, J. Kebrle and R.M. Thakler, Hela. chem. Acta,1950, 33, r7rt

389. R.K. Hill, S. Sawada and S. Arfin, Biorg. Chem.,I9T9, B,IZs References 338

399. J.H. wagenkneckt, M.M. Baizer and J.L. chruma, synth. commun., 7972,2,2I5.

400. H.J.H. Fentory l. Chem. Soc., 1900,22, 1296

401. L. Henry and A. Dewael, Chem. Zent.,l9O4, Ll,IgTz

402. R. Buyle, HeIa. Chim. Acta.,1964, qZ,2444

403. B.Kotelko, R.Clinka, Acta PoI. pharm.,1973,30, L3S

404. W.H. Perkin, l. Chem. 50c.,7896, 69, 7IB7 ç

405. G.V. Asolkar and P.C. Guha, l. Indian Chem. Soc., L946,29,47

406. I.B. Niederl and S. Natelson, I. Am. Chem. Soc., L937, Sg,Ig2g

407.M.M. Béhal and E. Choay, BuII. Chim. Soc. (Fr.),Igg4,t7,2Og

408. P.F. Frankland and T. Turner, l. Chem. Soc., 1883, 4g, jíz

409. A. Märcker, lustus Liebigs Ann. Chem.,!869, L40,88

410. E. Baumann and E. Fromm, Dtsch. Chem. Ges. Ber., tBgS, Zg, gL}

411. P.A. Levene and L.A. Mikeska, l. Biol. Chem., ,T0, gZB alt Misubishi Petrochemical Co. Ltd., Jpn.Kokai Tokko Koho Jp 81, 1,3g,1,6s PUBLICATI ONS.

L. Gas Phase Ca¡banion Rearrangements. Deprotonated Benzyl and A[yl Ethe¡s.

Peter C.H. Eichinger and lohn H. Bowie, l. Chem. Soc., Perkin Trans. II, I9BB, 497 -506

2- Gas - Phase carbanion Rearrangements. Does the wittig

Rearrangement occur for Deprotonated Vinyl Ethers ?

Peter C.H. Eichinger and John H. Bowie, ! l. Chem. Soc., Perkin Trans.Il, 1990,IZ6g - B

3. Carbanion Rearrangements in the Gas phase. The Claisen Rearrangement of Deprotonated Allyl Vinyl Ether.

Peter C.H. Eichinger and John H. Bowie, Aust.l. Chem. , 1990,43(9),1.479 - Bs

4. The Gas - Phase smiles Rearrangement. A Heavy Atom Labeling Study.

Peter C.H. Eichinger , ]ohn H. Bowie and Roger N. Hayes, I. Am.Chem. Soc., L989,1L7(L2),4224 - 7

5. The Gas Phase Benzilic Acid Rearrangement.

Roger N. Hayes, Peter C.H. Eichinger and Jotrn H. Bowie,

Rapid communications in Mass spectrometry, 1990, 4(B),2g3 - 4 Publications 340

6. Anionic Rearrangement in the Gas phase. The Collision Induced Losses of carbon Monoxide from Deprotonated pyruvates and Hydroryacetates.

Peter C.H. Eichinger, Roger N. Hayes and John H. Bowie, l. Chem. Soc., Perkin Trans.II, 1990,1815 - 9

7.1,2 Anionic Rearrangements in the Gas phase. The Acylory Acetate - Acyl Hydroryacetate and related Rearrangements.

Peter C.H. Eichinger , Roger N. Hayes and Iotrn H. Bowie, l.Am.Chem. Soc. , 199't, ti.3 (6), 1949 -I9Sg

8. The Loss of 'Radicals' from Even-Electron Negative lons in the Gas F Phase.

Peter C.H. Eichinger and John H. Bowie, lntl. l. Mass spectrom. Ion Processes. (accepted. for publication) Eichinger, P. C. H. & Bowie, J. H. (1988). Gas-phase carbanion rearrangements. Deprotonated benzyl and allyl ethers. Journal of the Chemical Society, Perkin Transactions 2 (4) 497-506.

NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.

It is also available online to authorised users at: http://dx.doi.org/10.1039/p29880000497 Eichinger, P. C. H. & Bowie, J. H. (1990). Gas-phase carbanion rearrangements. Does the Wittig rearrangement occur for deprotonated vinyl ethers? Journal of the Chemical Society, Perkin Transactions 2 (11) 11763-1768763

NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.

It is also available online to authorised users at: http://dx.doi.org/10.1039/p29900001763 Eichinger, P. C. H. & Bowie, J. H. (1990). Carbanion rearrangements in the gas phase: The unusual Claisen rearrangement of deprotonated allyl vinyl ether. Australian Journal of Chemistry, 43(9), 1479-1485.

NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.

It is also available online to authorised users at: http://dx.doi.org/10.1071/CH9901479 Eichinger, P. C. H., Bowie, J. H. & Hayes, R. N. (1989). The gas-phase smiles rearrangement: A heavy atom labeling study. Journal of the American Chemical Society, 111(12), 4224-4227.

NOTE:

This publication is included in the print copy

of the thesis held in the University of Adelaide Library.

It is also available online to authorised users at:

http://dx.doi.org/10.1021/ja00194a011

Hayes, R. N., Eichinger, P. C. H. & Bowie, J. H. (1990). The gas phase benzilic acid rearrangement. Rapid Communications in Mass Spectrometry, 4(8), 283-284.

NOTE:

This publication is included in the print copy

of the thesis held in the University of Adelaide Library.

It is also available online to authorised users at:

http://dx.doi.org/10.1002/rcm.1290040805

Eichinger, P. C. H., Hayes, R. N. & Bowie, J. H. (1990). Anionic rearrangement in the gas phase. The collision-induced loss of carbon monoxide from deprotonated pyruvates and hydroxyacetates. Journal of the Chemical Society, Perkin Transactions 2 (11) 1815-1819.

NOTE: This publication is included in the print copy of the thesis held in the University of Adelaide Library.

It is also available online to authorised users at: http://dx.doi.org/10.1039/p29900001815 Eichinger, P. C. H., Bowie, J. H. & Hayes, R. N. (1991). 1,2 Anionic rearrangements in the gas phase. The (acyloxy)acetate-acylhydroxyacetate and related rearrangements. Journal of the American Chemical Society, 113(6), 1949-1953.

NOTE:

This publication is included in the print copy

of the thesis held in the University of Adelaide Library.

It is also available online to authorised users at:

http://dx.doi.org/10.1021/ja00006a013

" The readet of Mass Spectrometric literature should always beat in mind that most of the suggested ion structures are fictitious, ønd are merely based on plausible considerøtions regarding the most stable form of øn ion or the most suitable form for a decomposition process. None the less , the heuristic aalue tf such representations is ç undeniable,so long as speculøtion remains within acceptable limits.',

J. Müller, Angant. Chem.,7972,1-1-, 653

the ¡eader or this thesis rinds the specuration *,;" ;*.r;.J;ì,Ïrl*"